From vxwexplo-errs@csg.lbl.gov Sun Jul 1 04:03:36 2001 From: Vxworks Exploder Date: Sun Jul 1 04:03:38 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 1 04:03:03 PDT 2001 Subject: Re: why signal sample source code generates exception (VxWorks 5.4 for Subject: New pain with old Systems (vxWorks 5.2) Subject: Get the device name, given the fd Subject: Re: Rational ClearCase with Tornado & VxWorks? Subject: Re: How to get rid of const warning Subject: Re: Rational ClearCase with Tornado & VxWorks? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: why signal sample source code generates exception (VxWorks 5.4 for PPC) Date: Sat, 30 Jun 2001 04:51:35 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B3DBD47.88FA1741@covad.net> References: <9hk1qe$36e$1@overload.lbl.gov> Reply-To: drdiags@covad.net lhp, Have you applied the T2CP3 (Tornado 2 Cumulative Patch 3) release? This problem was fixed in that release. The following SPR may have been the one specifically for this (from T2CP3 release notes): 26584: Posix signals and timers cause general protection exceptions on simnt. lhp wrote: > > HI, > > I use Tornado II and VxWorks 5.4 for PPC. Why the following signal sample source code generates exception, no matter running in VxSim or running on an actual MPC860 board? > HELP! my email: LHP@gpte.com.cn > <...SNIP....> --------------------------- Newsgroups: comp.os.vxworks Subject: New pain with old Systems (vxWorks 5.2) Date: 30 Jun 2001 09:55:15 -0700 From: bob_k_2@my-deja.com (Bob K) Organization: http://groups.google.com/ Message-ID: <87dea4a0.0106300855.227af93d@posting.google.com> Hi, I have been forced to resurrect an old system to read some tapes. The old system is a Force CPU64. It is running vxWorks 5.2 with a vxWorks 5.1 BSP by force. The force unit supported scsi 1 hard drives without any problem (dosfs). We made it work with a SCSI II tape unit because the force BSP supported wide and synchronous transfers. I now have to get this system running again, but no longer have access to scsi 1 disks. I have attached SCSI 2 disks to the system and have got the system to recoginize that they exist with the following in usrScsi.c: if ((pSpdDDrv = scsiPhysDevCreate (pSysScsiCtrl, 0, 0, 0, 0, 0, 65536, 32768)) == (SCSI_PHYS_DEV *) NULL) { SCSI_DEBUG_MSG ("usrScsiConfig: d drive failed.\n", 0, 0, 0, 0, 0, 0); scsiTest |= 1; } else { if (ncr720ScsiSync(pSpdDDrv) != OK){ scsiTest |=2; } if (ncr720ScsiWide(pSpdDDrv) != OK){ scsiTest |= 4; } if ((pSbdDDrv = scsiBlkDevCreate (pSpdDDrv, 0, 63)) == NULL){ scsiTest |= 8; return (ERROR); } if ((pDvdDDrv = dosFsDevInit ("/ddrv/", pSbdDDrv, NULL)) == NULL){ scsiTest |=0x10; } } This executes except for the dosFsDevInit which routinely fails. I originally formatted these disks on a Win2K machine in DOSFS (FAT) format. I have had some success with running these commands from the command line. I have gotten the dosFsDevInit to work, dosFsMkfs works. dosFsConfigShow reports ok, but reading or writing to the disk fails. So anybody out there have any ideas of things I should try? Regards, Bob K --------------------------- Newsgroups: comp.os.vxworks Subject: Get the device name, given the fd Date: Fri, 29 Jun 2001 17:13:12 +0530 From: "Raghuram Y S" Organization: Hewlett Packard ISO, Bangalore. Message-ID: <9hhq0j$s04$1@pavan.india.hp.com> Hi, is there any way that one can get the device name just by providing the fd of the device The actual requirement is as below * Before 'open'ing a device, I want to know, if any other application or task has already 'open'ed the device. * and abort the call to open if the device has already opened. To do this, I think, I need to look into the FD_TABLE of the VxWorks and do some reverse engineering to get the name of the devcie per fd and compare them with the device name string that I intend to open. I think, there is no direct method for doing this In isoLib, there is an API isoFdValue(int fd) which validates the fd passed. Perhaps the source code of this will throw some light on how to access the FD_TABLE. But, after accessing the FD_TABLE, how do I get the device name out of fd is the question. Can somebody help me.. Thanks in advance - -Raghu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Rational ClearCase with Tornado & VxWorks? Date: 30 Jun 2001 19:56:11 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: molochai wrote: > Anybody using clearcase for maanging VxWorks projects? Amy experiences > you'd like yo share? I did some tests with it, and it works good. There are a couple of approaches to this. You can use CC to store the projects only, or to store your complete Tornado tree. Then you can choose to use dynamic or static views. But none of them gave any problems. Even from within Tornado, with the default command it works OK. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get rid of const warning Date: Sun, 01 Jul 2001 03:19:28 GMT From: pkockritz@home.com (Pete Kockritz) Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <9hhv0v$68c$1@overload.lbl.gov> In article <9hhv0v$68c$1@overload.lbl.gov>, Jeff David wrote: > Hello vxWorks users! > > When the code contains lines such as the following: > > if (ipAttach(0, "dc") != OK) > ... > > I get the following warning: > pniend.c:666: warning: passing arg 2 of `ipAttach' discards `const' from > pointer target type > > Does anybody know how to change the code to get rid of this warning? (Note: > Suppressing the warning is not an option.) How about this: char ifName[] = "dc"; ... if (ipAttach (0, ifName) != OK) or if (ipAttach (0, (char *)"dc") != OK) or Modify the prototype for ipAttach() in the Tornado header file to make the second argument 'const char *'. (probably not the best choice). - -- Pete Kockritz --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Rational ClearCase with Tornado & VxWorks? Date: 1 Jul 2001 07:54:59 GMT From: slava@qualcomm.com (Izoslav Tchigevsky) Organization: Qualcomm Israel Message-ID: References: molochai wrote in news:f3c67924a%molochai@vapour-trail.demon.co.uk: > Anybody using clearcase for maanging VxWorks projects? Amy experiences > you'd like yo share? > Just remember to define Tornado Project File (*.wpj) as binary file if you work with Windows. In other case ClearCase will transform this file to Windows text format (well-known "\n" problem), and Tornado will not read this file correctly. Also Remember that Tornado will not recompile you files in case of undo- checkout or when view config spec is changed to select older version of files. Everything else works fine. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 1 22:01:48 2001 From: "Mike Anderson" Date: Sun Jul 1 22:01:50 PDT 2001 Subject: Makefile problem Greetings! I'm trying to modify a VxWorks Makefile to handle the case when you build a bootrom vs. a normal VxWorks image. The MACH_EXTRA defines get included in both cases. This causes a problem in the bootroms because of another bug in VxWorks using the WDB_COMM_END communications type and T202. What I'm trying to do is to look at the type of target that is being built and adjust the MACH_EXTRA parameter accordingly. To that end, I have a little snippet of Makefile here: ################################################################### foundname = $(findstring boot, $@) ifneq ($(foundname), ) EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=NORMAL MACH_EXTRA = ../PentiumBSP/default/demo.o \ ../PentiumBSP/default/wan68.o \ ../PentiumBSP/default/osal.o \ ../PentiumBSP/default/sysWanicEnd.o \ ./usrAppInit.o \ ./fei82557End.o \ ./wdbEndPktDrv.o else EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=BOOTROM MACH_EXTRA = endif #################################################################### foundname is being set correctly. It contains "boot" if the target had the sequence "boot" in it (e.g., bootrom, bootrom_uncmp, etc.) or is blank otherwise. However, the comparisons just don't seem to be working. For some reason, I always get the else clause firing. If anyone has any ideas or pointers as to what could be going wrong, I'd sure appreciate the help. TIA, -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Mon Jul 2 04:03:14 2001 From: Vxworks Exploder Date: Mon Jul 2 04:03:16 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 2 04:03:04 PDT 2001 Subject: Re: How can I block VxWorks default shell (->)? Subject: Re: why signal sample source code generates exception (VxWorks 5.4 for PPC) Subject: serial management controller (SMC) / PPC823 Subject: Re: How can I get a Call Stack back trace? Subject: Re: How to get rid of const warning Subject: Re: How to get rid of const warning Subject: Makefile problem Subject: Re: strong ARM pointer align ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I block VxWorks default shell (->)? Date: 1 Jul 2001 12:56:45 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9hjt5a$ee2$1@news1.kornet.net> "±èÀÀ½Ä" wrote: > As far as I know windriver don't show the source of ``telnet'' and > ``shell''. > However I am looking forward to enter my own shell command interpreter > when I connect target throught telnet. There is an unsupported Telnet server available from WindRiver. Ask your FAE for it. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: why signal sample source code generates exception (VxWorks 5.4 for PPC) Date: 1 Jul 2001 13:08:12 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9hk1qe$36e$1@overload.lbl.gov> "lhp" wrote: > I use Tornado II and VxWorks 5.4 for PPC. Why the following signal > sample source code generates exception, no matter running in VxSim or > running on an actual MPC860 board? HELP! my email: LHP@gpte.com.cn As far as I know the signal function will be executed in the context of the task that created the signal, so if you leave your main task running it might work. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: serial management controller (SMC) / PPC823 Date: Sun, 01 Jul 2001 16:27:41 +0200 From: Heiko Schmidt Organization: T-Online Message-ID: <3B3F335D.3000404@t-online.de> Hello we are using a PPC823. We want to download a lot of bytes over the serial interface. For this we use the SMC in async. UART mode. I configered one rx buffer with the size of one byte. So i get a receive interrupt after every byte. For a lot of bytes it will work mostly. But sometimes i loose some bytes. Doese someone know something over the performance of the SMC ? We are using 19200 Baud. In my ISR routine i only copy the received byte into a buffer. Or is there a better solution for downloading a lot of bytes ? Thanks Heiko --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I get a Call Stack back trace? Date: Sun, 01 Jul 2001 15:10:32 GMT From: "Jin Qian" Organization: Prodigy Internet http://www.prodigy.com Message-ID: References: Hi Rob, If you target is still alive, you can use either your tShell or tornado shell (sometimes called windsh) to do the following: Suppose the name of the crashed task is called "xyz", by typping tt xyz you will get the backtrace of the task in question. Hope it helps. Jin "Robert H. Fritz" wrote in message news:tj4t4i2jcr3ma9@corp.supernews.com... > I'm generating a kernel exception which causes my vxWorks host to reboot. > From what I've been able to determine, I'm blowing up in bcopy which is used > by memcpy and probably a seemingly limitless number of things. I'm looking > for a tool that can provide me with a back-trace of my call stack so I can > determine what line of code is ultimately causing this problem. > > Any tool or technique suggestions appreciated. > > Thanks. > > Rob Fritz > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get rid of const warning Date: 01 Jul 2001 14:23:36 -1000 From: Jim Thomas Organization: Canada France Hawai`i Telescope Message-ID: References: <9hhv0v$68c$1@overload.lbl.gov> >>>>> "Pete" == Pete Kockritz writes: Pete> In article <9hhv0v$68c$1@overload.lbl.gov>, Jeff David Pete> wrote: >> When the code contains lines such as the following: >> >> if (ipAttach(0, "dc") != OK) >> ... >> >> I get the following warning: >> pniend.c:666: warning: passing arg 2 of `ipAttach' discards `const' from >> pointer target type Pete> How about this: Pete> char ifName[] = "dc"; Pete> ... Pete> if (ipAttach (0, ifName) != OK) Presumably you're compiling with -fwritable-strings ? If you can't flip that, then 'const char ifName[] = "dc";' should work. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get rid of const warning Date: 01 Jul 2001 14:56:09 -1000 From: Jim Thomas Organization: Canada France Hawai`i Telescope Message-ID: References: <9hhv0v$68c$1@overload.lbl.gov> >>>>> "Jim" == Jim Thomas writes: >>>>> "Pete" == Pete Kockritz writes: Pete> In article <9hhv0v$68c$1@overload.lbl.gov>, Jeff David Pete> wrote: >>> When the code contains lines such as the following: >>> >>> if (ipAttach(0, "dc") != OK) >>> ... >>> >>> I get the following warning: >>> pniend.c:666: warning: passing arg 2 of `ipAttach' discards `const' from >>> pointer target type Pete> How about this: Pete> char ifName[] = "dc"; Pete> ... Pete> if (ipAttach (0, ifName) != OK) Jim> Presumably you're compiling with -fwritable-strings ? If you can't flip Jim> that, then 'const char ifName[] = "dc";' should work. Arrgh, obviously I got that backwards. Do what Pete said. :-( --------------------------- Newsgroups: comp.os.vxworks Subject: Makefile problem Date: Mon, 2 Jul 2001 01:01:10 -0400 From: "Mike Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hp016$o73$1@overload.lbl.gov> Greetings! I'm trying to modify a VxWorks Makefile to handle the case when you build a bootrom vs. a normal VxWorks image. The MACH_EXTRA defines get included in both cases. This causes a problem in the bootroms because of another bug in VxWorks using the WDB_COMM_END communications type and T202. What I'm trying to do is to look at the type of target that is being built and adjust the MACH_EXTRA parameter accordingly. To that end, I have a little snippet of Makefile here: ################################################################### foundname = $(findstring boot, $@) ifneq ($(foundname), ) EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=NORMAL MACH_EXTRA = ../PentiumBSP/default/demo.o \ ../PentiumBSP/default/wan68.o \ ../PentiumBSP/default/osal.o \ ../PentiumBSP/default/sysWanicEnd.o \ ./usrAppInit.o \ ./fei82557End.o \ ./wdbEndPktDrv.o else EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=BOOTROM MACH_EXTRA = endif #################################################################### foundname is being set correctly. It contains "boot" if the target had the sequence "boot" in it (e.g., bootrom, bootrom_uncmp, etc.) or is blank otherwise. However, the comparisons just don't seem to be working. For some reason, I always get the else clause firing. If anyone has any ideas or pointers as to what could be going wrong, I'd sure appreciate the help. TIA, - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: Re: strong ARM pointer align Date: 2 Jul 2001 03:08:31 -0700 From: andy.birkett@transmitton.co.uk (Andy Birkett) Organization: http://groups.google.com/ Message-ID: <8ee5b646.0107020208.3b18d99d@posting.google.com> References: <9gqp7m$sj7$1@overload.lbl.gov> > I use intel strongARM SA1110 assabet board and Tornado 2.0. > I find if a pointer is not 4-byte align, data may be corrupted. For example > /* set p point with a non-4-byte align address */ > p = (int *) 0x08000002; > *p = 0xffffffff; > bytes with address 0x08000000 and 0x08000001 are corrupted!*/ There is a compiler option: - -funaligned-pointers -fpack-struct (I think you need both?) but the resultant code is massive (and much slower) so we never use it! You should be able to create a pointer to unaligned data by using code like: typedef unsigned short __attribute__ ((aligned(1))) *tpUnalignedShort; and then use: tpUnalignedShort px; short x; x=*px; etc. Unfortunately this generates the wrong code on ARM. I pointed this out to WindRiver in Dec '99 and they raised SPR#29782. We use the following workround: typedef union {unsigned short asUint16;} __attribute__ ((packed)) tsafeUint16; typedef tsafeUint16 *tpSafeUint16; tpSafeUint16 px; px= ; x=px->asUint16; Note the problem exists for all simple data types (i.e. 8,16 and 32 bit wide items) so we have a set of simular unions for each type. The problem does not occur with data in a structure provided you use __attribute__ ((packed)) in the declaration. The problem with this method is that you have to remember to do it and if you don't they effects may go un-noticed. If anyone knows how to trap alignment error and just throw an exception I'd like to know. Andy --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 2 09:19:46 2001 From: Anuradha Pani Date: Mon Jul 2 09:19:49 PDT 2001 Subject: Timers in vxworks hi, i am new to vxworks and am trying to implement basic timer routines. i'm not able to find any example code which tells me how to handle any signal upon timer expiry. i'm able to create timers but am not able to proceed further. could anyone give me any pointers? thanks in advance anuradha __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ From vxwexplo-errs@csg.lbl.gov Mon Jul 2 11:15:28 2001 From: fscheurer@manz-acs.de Date: Mon Jul 2 11:15:30 PDT 2001 Subject: Alternative cross-compiler for VxWorks Hi VxWorkers, I am trying to build an alternative cross tool-chain with configuration: host=i686-pc-cygwin target=i386-wrs-vxworks using the following versions and the cygwin tools: 29.06.01 15:47 binutils-2.11.2 29.06.01 15:53 gcc-2.95.3 29.06.01 16:00 newlib-1.9.0 29.06.01 17:50 gdb-4.18 Some problems could be corrected by the right Environment settings (PATH and LIB variables to cygwin-tools and setting the temporary files etc.), but now I ran into a problem that's: .. >sh .\build.sh Trying to build binutils ... Created "Makefile" in /cygdrive/p/PROJEKTE/GNU_NonWRS_Tools_sources/CROSSGCC/build_newlib1.9.0/build-binutils using "mh-frag" /cygdrive/c/usr/bin/ld: cannot open crt0.o: No such file or directory collect2: ld returned 1 exit status *** The command 'c:/usr/bin/gcc -o conftest -g -O2 conftest.c' failed. *** You must set the environment variable CC to a working compiler. --print-multi-lib: not found Configuring in ./libiberty loading cache ./config.cache checking host system type... i686-pc-cygwin checking build system type... i686-pc-cygwin checking for ar... i386-wrs-vxworks-ar checking for ranlib... i386-wrs-vxworks-ranlib checking for gcc... gcc checking whether we are using GNU C... yes checking whether gcc accepts -g... yes checking for POSIXized ISC... no checking for working const... yes checking for inline... inline checking for a BSD compatible install... /cygdrive/c/usr/bin/install -c cmp: not found checking how to run the C preprocessor... /lib/cpp checking for sys/file.h... grep: conftest.out: No such file or directory yes checking for sys/param.h... grep: conftest.out: No such file or directory yes checking for limits.h... grep: conftest.out: No such file or directory yes checking for stdlib.h... grep: conftest.out: No such file or directory yes checking for string.h... grep: conftest.out: No such file or directory yes checking for unistd.h... grep: conftest.out: No such file or directory yes checking for strings.h... grep: conftest.out: No such file or directory yes checking for sys/time.h... grep: conftest.out: No such file or directory yes checking for time.h... grep: conftest.out: No such file or directory yes checking for sys/resource.h... grep: conftest.out: No such file or directory yes checking for sys/stat.h... grep: conftest.out: No such file or directory yes checking for sys/mman.h... grep: conftest.out: No such file or directory yes checking for fcntl.h... grep: conftest.out: No such file or directory yes checking for sys/wait.h that is POSIX.1 compatible... no checking whether time.h and sys/time.h may both be included... no checking whether errno must be declared... yes checking whether the C compiler (gcc -g -O2 ) works... no configure: error: installation or configuration problem: C compiler cannot create executables. ... Why does my gcc not create right executables. Running a test program, shows the same problem with "crt0.o" Copying it to the local path let's us hang at "-luser32": .. >gcc -o conftest -g -O2 conftest.c /cygdrive/c/usr/bin/ld: cannot find -luser32 collect2: ld returned 1 exit status .. Someone out there having built already or having experiencs with that cross: host=i686-pc-cygwin target=i386-wrs-vxworks TIA MfG / Kind Regards Friedrich Scheurer, R&D -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de -------------------------------- From vxwexplo-errs@csg.lbl.gov Mon Jul 2 14:57:31 2001 From: "Michael Morrison" Date: Mon Jul 2 14:57:34 PDT 2001 Subject: Memory requirments.... Anyone know where I can find information on the memory requirments/footprint of VxWorks and it's various components? I just need a rough idea of how much memory our box will need. Thanks From vxwexplo-errs@csg.lbl.gov Mon Jul 2 17:58:33 2001 From: "Sunil Verma" Date: Mon Jul 2 17:58:35 PDT 2001 Subject: VxWorks Scheduling This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01C1035B.36316A80 Content-Type: text/plain; charset="iso-8859-1" Hi, Could anyone tell the behavior of VxWorks scheduler in following situation ? I have two tasks Task_A (priority 100) & Task_B (priority 90). 1. Task_A calls func_1() which does some processing and takes a semaphore (mutex). [it has still not called func_3()] 2. Task_B being higher priority is scheduled, func_2() starts executing and comes to semTake(), but since Task_A already has the semaphore, Task_B is blocked. [Task_B is placed the queue to take the semaphore next] 3. Task_A continues and func_3() is called which does a * TaskLock() * releases the semaphore * does some processing * tries to take the semaphore again Here lies my doubt, will the semaphore be given to Task_A (being in execution and having TaskLock) or will it be placed in the queue to take semaphore after Task_B() and TaskLock is released to avoid deadlock. I want to know how VxWorks handles this situation. The functions (func_1, func_2 & func_3) used in above example are detailed below. I would also appreciate if you guys could tell me where to find detailed implementation of VxWorks Scheduler ? I also wanna know if guys @ VxWorks fixed some bugs in memory manager recently ?? What is the latest released version of Vxworks ? Thanks, Sunil func_1() { ... semTake(mySemId, WAIT_FOREVER); ... ... func_3(); ... semGive(mySemId); } func_2() { ... semTake(mySemId, WAIT_FOREVER); ... } func_3() { taskLock(); semGive(mySemId); ... semTake(mySemId, WAIT_FOREVER); taskUnlock(); } ------_=_NextPart_001_01C1035B.36316A80 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable VxWorks Scheduling

Hi,

Could anyone tell the behavior = of VxWorks scheduler in following situation ?

I have two tasks Task_A = (priority 100) & Task_B (priority 90).
1. Task_A calls func_1() which = does some processing and takes a semaphore (mutex). [it has still not = called func_3()]
2. Task_B being higher = priority is scheduled, func_2() starts executing and comes to = semTake(), but since Task_A already has the semaphore, Task_B is = blocked. [Task_B is placed the queue to take the semaphore = next]

3. Task_A continues and = func_3() is called which does a
* TaskLock()
* releases the = semaphore
* does some processing
* tries to take the semaphore = again

Here lies my doubt, will the = semaphore be given to Task_A (being in execution and having TaskLock) = or will it be placed in the queue to take semaphore after Task_B() and = TaskLock is released to avoid deadlock. I want to know how VxWorks = handles this situation.

The functions (func_1, func_2 = & func_3) used in above example are detailed below.

I would also appreciate if you = guys could tell me where to find detailed implementation of VxWorks = Scheduler ?
I also wanna know if guys @ = VxWorks fixed some bugs in memory manager recently ??
What is the latest released = version of Vxworks ?

Thanks,
Sunil


func_1()
{
    ...
    = semTake(mySemId, WAIT_FOREVER);
    ...
    ...
    = func_3();
    ...
    = semGive(mySemId);
}

func_2()
{
    ...
    = semTake(mySemId, WAIT_FOREVER);
    ...
}

func_3()
{
    = taskLock();
    = semGive(mySemId);
    ...
    = semTake(mySemId, WAIT_FOREVER);
    = taskUnlock();
}

------_=_NextPart_001_01C1035B.36316A80-- From vxwexplo-errs@csg.lbl.gov Mon Jul 2 19:33:01 2001 From: "Mike Anderson" Date: Mon Jul 2 19:33:03 PDT 2001 Subject: mkboot.c that supports DosFsLib2? VxWorks Greetings! Has anyone out there in net land already modified the mkboot.c code to support installing the boot block on an IDE drive for use with DosFsLib2? The code included with the BSP works if your hard disk has a floppy disk format on it (ala DosFsLib1), but gives an error of Can't find primary DOS partition. Any help or pointers would be appreciated. TIA, Mike Anderson -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Mon Jul 2 21:47:34 2001 From: "Palaniar" Date: Mon Jul 2 21:47:36 PDT 2001 Subject: RE: Makefile Problem Hi, The ifneq will always obviously execute the else part, because the condition will always fail. (i.e), you have checked whether the foundname exists or not rather foundname has something or not. Hence, foundname will return true and that's why the execution of ifneq goes to the else part. Alternatively, use the following command: ifneq ($(foundname),"boot") else endif For the above set of instructions, if foundname == "boot", then else part will get executed, otherwise the first part will get executed. Regards ARP >From vxwexplo-errs@csg.lbl.gov Sun Jul 1 22:01:48 2001 From: "Mike Anderson" Date: Sun Jul 1 22:01:50 PDT 2001 Subject: Makefile problem Greetings! I'm trying to modify a VxWorks Makefile to handle the case when you build a bootrom vs. a normal VxWorks image. The MACH_EXTRA defines get included in both cases. This causes a problem in the bootroms because of another bug in VxWorks using the WDB_COMM_END communications type and T202. What I'm trying to do is to look at the type of target that is being built and adjust the MACH_EXTRA parameter accordingly. To that end, I have a little snippet of Makefile here: ################################################################### foundname = $(findstring boot, $@) ifneq ($(foundname), ) EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=NORMAL MACH_EXTRA = ../PentiumBSP/default/demo.o \ ../PentiumBSP/default/wan68.o \ ../PentiumBSP/default/osal.o \ ../PentiumBSP/default/sysWanicEnd.o \ ./usrAppInit.o \ ./fei82557End.o \ ./wdbEndPktDrv.o else EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=BOOTROM MACH_EXTRA = endif #################################################################### foundname is being set correctly. It contains "boot" if the target had the sequence "boot" in it (e.g., bootrom, bootrom_uncmp, etc.) or is blank otherwise. However, the comparisons just don't seem to be working. For some reason, I always get the else clause firing. If anyone has any ideas or pointers as to what could be going wrong, I'd sure appreciate the help. TIA, -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time. From vxwexplo-errs@csg.lbl.gov Tue Jul 3 04:03:21 2001 From: Vxworks Exploder Date: Tue Jul 3 04:03:23 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 3 04:03:06 PDT 2001 Subject: Re: Makefile problem Subject: Re: Get the device name, given the fd Subject: How to know which block (or memory address) is changed in ramDev (ramDisk) Subject: Re: vxWorks - developing a Shell on vxSim w/ no network package Subject: Re: writing a "Shell" to run on top of "tShell" of vxSim Subject: Re: problem with inet_aton Subject: Re: Help for a newbie with vxworks Subject: Re: WindView post-mortem: no luck yet Subject: Re: comp.os.vxworks newsdigest Subject: Re: SSH on VxWorks Subject: TFFS Alignment problem Subject: Can WindView lose events? Subject: Timers in vxworks Subject: Re: TFFS limitations? Subject: Re: Flash File System Alternatives Subject: Re: Flash File System Alternatives Subject: Malta BSP & usrNetInit string Subject: Alternative cross-compiler for VxWorks Subject: Re: TFFS Alignment problem Subject: how to reset the logical partition on a SCSI block device Subject: BSP question- device with 2 INT Subject: Defragmentation for TrueFFS Subject: Memory requirments.... Subject: Re: Can WindView lose events? Subject: Re: Makefile problem Subject: VxWorks Scheduling Subject: mkboot.c that supports DosFsLib2? Subject: Does Mibcomp also process "SEQUENCE" "Table " in SNMP? Subject: Vxsim very small fonts in output window Subject: à÷ãà Subject: RE: Makefile Problem Subject: Makefile problem Subject: Use of MemScope issue - TCB corruption Subject: Re: TFFS limitations? -- No limitition, here's what you do... Subject: Re: WindView post-mortem: no luck yet Subject: about "taskSpawn" Subject: I want to know "shell". Subject: Re: TFFS Alignment problem Subject: Re: TFFS Alignment problem Subject: Memory Management Subject: How can i print the stack call after an error ? Subject: porting of vxworks to threadx Subject: Re: vxworks configuration Subject: Re: about "taskSpawn" ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile problem Date: Mon, 02 Jul 2001 12:10:38 +0100 From: Graham Baxter Organization: Graham Baxter (Software) Limited Message-ID: <3B4056AE.76704284@NOSPAM.bcs.org.uk> References: <9hp016$o73$1@overload.lbl.gov> Mike, The variable $@ is usually used in recipe lines: x.o: x.c cc -c $< -o $@ $@ is the file name of the target x.o $< is the name of the first dependency x.c I suspect that $@ is null in your case. Regards, Graham Baxter Freelance Software Engineer gbaxter@NOSPAM.bcs.org.uk Mike Anderson wrote: > > Greetings! > > I'm trying to modify a VxWorks Makefile to handle the > case when you build a bootrom vs. a normal VxWorks image. > The MACH_EXTRA defines get included in both cases. This > causes a problem in the bootroms because of another bug in > VxWorks using the WDB_COMM_END communications type and T202. > What I'm trying to do is to look at the type of target that is > being built and adjust the MACH_EXTRA parameter accordingly. > To that end, I have a little snippet of Makefile here: > > ####################################################### > foundname is being set correctly. It contains "boot" if > the target had the sequence "boot" in it (e.g., bootrom, > bootrom_uncmp, etc.) or is blank otherwise. > > However, the comparisons just don't seem to be working. > For some reason, I always get the else clause firing. If > anyone has any ideas or pointers as to what could be going > wrong, I'd sure appreciate the help. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get the device name, given the fd Date: Mon, 2 Jul 2001 13:13:17 +0200 From: "Matthias Guelck" Organization: Heidelberg Digital Message-ID: <9hpl0e$f1r$1@news.hdpp.de> References: <9hhq0j$s04$1@pavan.india.hp.com> Hi Raghu, may this helps: ioctl(pFile->_file,FIOGETNAME,(int)pcCurrentFileName); Best regards Matthias "Raghuram Y S" wrote in message news:9hhq0j$s04$1@pavan.india.hp.com... > Hi, > is there any way that one can get the device name just by providing the fd > of the device > > The actual requirement is as below > > * Before 'open'ing a device, I want to know, if any other application or > task has already 'open'ed the device. > * and abort the call to open if the device has already opened. To do this, I > think, I need to look into the FD_TABLE of the VxWorks and do some reverse > engineering to get the name of the devcie per fd and compare them with the > device name string that I intend to open. > I think, there is no direct method for doing this > In isoLib, there is an API isoFdValue(int fd) which validates the fd > passed. Perhaps the source code of this will throw some light on how to > access the FD_TABLE. But, after accessing the FD_TABLE, how do I get the > device name out of fd is the question. > Can somebody help me.. > > Thanks in advance > -Raghu > > --------------------------- Newsgroups: comp.os.vxworks Subject: How to know which block (or memory address) is changed in ramDev (ramDisk) Date: 2 Jul 2001 06:54:50 -0700 From: roh_youngjin@hotmail.com (Youngjin Roh) Organization: http://groups.google.com/ Message-ID: <799f3d42.0107020554.7d35256@posting.google.com> Hello folks. I am using ppc750 Board which has 8MB FlashROM. While flash erase & write is too slow, I copy all FLASH data into memory when the board comes up, and mount it as RamDisk (ramDevCreate). But the problem is when I write the memory data to FLASH. I don't know which memory portion has been changed, I have to erase whole FLASH and write whole data of memory to FLASH. It takes about 1 minute. Is there any way to find out which block and address in DOS Block Dev has been changed ? If I can know the blocks, I can erase & write some which are changed instead of whole data. Help me folks~~~~ Thank you. Youngjin. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks - developing a Shell on vxSim w/ no network package Date: 2 Jul 2001 07:02:45 -0700 From: Yevgeni_Tunik@yahoo.com (Yevgeni Tunik) Organization: http://groups.google.com/ Message-ID: <9e9c2446.0107020602.60be9fd7@posting.google.com> References: <4.1.20010621095630.00b35330@mill.cisco.com> > What are the FDs to be used for in and out and err > > Any help, sample, suggestion is greatly appreciated. > > Thanks, > > > Bibhuti Kar > Cisco Systems Inc. > (978) 244 5740 > eFax (508)462-0368 The following fragment of "VxWorks Programmers Guide" answers: 3.3.2 Standard Input, Standard Output, and Standard Error Three file descriptors are reserved and have special meanings: 0 = standard input 1 = standard output 2 = standard error output You can download free Tornado2 Prototyper from www.windriver.com After install, you will have full set of original documentation. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: writing a "Shell" to run on top of "tShell" of vxSim Date: 2 Jul 2001 07:11:12 -0700 From: Yevgeni_Tunik@yahoo.com (Yevgeni Tunik) Organization: http://groups.google.com/ Message-ID: <9e9c2446.0107020611.5f103bb6@posting.google.com> References: > What FDs to be used for in/out/error? > Thanks a lot for any hint/suggestion/sample code. See "VxWorks Programmer'ss Guide": 3.3.2 Standard Input, Standard Output, and Standard Error Three file descriptors are reserved and have special meanings: 0 = standard input 1 = standard output 2 = standard error output --------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem with inet_aton Date: 2 Jul 2001 07:42:08 -0700 From: Yevgeni_Tunik@yahoo.com (Yevgeni Tunik) Organization: http://groups.google.com/ Message-ID: <9e9c2446.0107020642.70000103@posting.google.com> References: > I noticed VxWorks does not support "inet_addr" You need to include concerned component to the custom VxWorks build. This function works, but (at least on simulator) reverses the byte order. According to WindRiver documentation, The following example returns 0x5a000002 (in network order): inet_addr ("90.0.0.2"); But, actually, the function returns 0x0200005a. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help for a newbie with vxworks Date: 2 Jul 2001 07:54:18 -0700 From: Yevgeni_Tunik@yahoo.com (Yevgeni Tunik) Organization: http://groups.google.com/ Message-ID: <9e9c2446.0107020654.7c4aef1f@posting.google.com> References: <9g5rsv$4no$1@front5.grolier.fr> > I am intersted in compilers;... and perhaps simulator ?.. > > Could you help me ? :) > > thanks a lot! Download free Tornado2 Prototyper from www.windriver.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: WindView post-mortem: no luck yet Date: Mon, 2 Jul 2001 10:27:57 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9hq3te0usq@enews2.newsguy.com> References: <3B3C96DB.D012CF9E@oce.nl> > After a reboot Hard reboot or soft? The WindView logs are stored in RAM so if you do a hard reboot the info will be lost. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: comp.os.vxworks newsdigest Date: Mon, 2 Jul 2001 10:29:18 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9hq3us0v07@enews2.newsguy.com> References: <9hjjof$q1v$1@overload.lbl.gov> > Log Console I get a message "Error: clnttty_rcv: read timed out " .And I have seen this message when the baud rate is wrong. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SSH on VxWorks Date: 02 Jul 2001 12:24:25 -0400 From: "Paul D. Smith" Organization: Bay Networks, Inc. / Billerica, MA Message-ID: References: <7c4037a9.0106281550.732ff4a5@posting.google.com> <196667924a%molochai@vapour-trail.demon.co.uk> Reply-To: psmith@baynetworks.com %% molochai writes: m> In message <7c4037a9.0106281550.732ff4a5@posting.google.com> m> tapankarwa@yahoo.com (Tapan Karwa) wrote: m> OpenSSL ought to port quite easily. Not done so myself though. We did it, to VxWorks 5.3.1, and it was _NOT_ easy. Unfortunately I'm not familiar with enough details to be helpful to anyone attempting to do the same. Also, 5.3.1 is obviously quite old so maybe it would be easier on a newer version. Some things to consider: * Before you can port OpenSSH, you need OpenSSL. OpenSSL has a number of global variables. We actually stripped out all the algorithms except those actually needed by SSH. * OpenSSH (and OpenSSL, as above) use the standard UNIX "fork on connect" daemon model, which spawns a new process for each connection. * We ran into a nasty bug or two with VxWorks select/TTY handling that caused the terminal output to be buffered/displayed incorrectly (e.g., sometimes you'd have to hit an extra RETURN to see the output of the previous command). All in all, the port took a lot longer than we anticipated. FWIW. - -- - ------------------------------------------------------------------------------- Paul D. Smith HASMAT--HA Software Methods & Tools "Please remain calm...I may be mad, but I am a professional." --Mad Scientist - ------------------------------------------------------------------------------- These are my opinions---Nortel Networks takes no responsibility for them. --------------------------- Newsgroups: comp.os.vxworks Subject: TFFS Alignment problem Date: Mon, 2 Jul 2001 18:16:22 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: <3m6qh9.oqq.ln@gum.de> Hi there! We're using TFFS on Hitachi SH3 which has a 32Bit wide access to RAM and Flash memory. When writing data to the TFFS drive we get an exception as the function 'checkForWriteInplace' (ftllite.o) is trying to write to a 2byte aligned address instead of 4byte aligned. Does anyone know how to solve this problem? WRS support hasn't been of much help so far... :-( Thanks alot, Frank Wolf --------------------------- Newsgroups: comp.os.vxworks Subject: Can WindView lose events? Date: 2 Jul 2001 10:07:31 -0700 From: eyacobi@qualcomm.com (Eitan Yacobi) Organization: http://groups.google.com/ Message-ID: <28f198ce.0107020907.5dcff2c7@posting.google.com> Hi guys, Maybe you could help me with this one: I use Tornado version 2.0 and VxWorks version 5.4. I use WindView to monitor the performance of the system. Q1) When the CPU or the network are loaded, is possible that WindView will loose events (i.e. events that happened in the system will not be written to the WindView log)? Q2) I have selected the maximal event-mask on WindView, using the "WindView Selection Configuration" screen. Still, in the WindView log there are periods of time that no task or interrupt are in running state. What is the right way to specify an event-mask that will guaranty that during the log duration, everything that acquires the CPU will appear in the log? Thanks a lot, Eitan --------------------------- Newsgroups: comp.os.vxworks Subject: Timers in vxworks Date: Mon, 2 Jul 2001 09:19:44 -0700 (PDT) From: Anuradha Pani Organization: Lawrence Berkeley National Laboratory Message-ID: <9hqa6u$ge6$1@overload.lbl.gov> hi, i am new to vxworks and am trying to implement basic timer routines. i'm not able to find any example code which tells me how to handle any signal upon timer expiry. i'm able to create timers but am not able to proceed further. could anyone give me any pointers? thanks in advance anuradha __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS limitations? Date: Mon, 02 Jul 2001 13:29:02 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B40AF5E.C164FB4@digidata.com> References: <3B31BDA2.AEEA9B89@siemenscomms.co.uk> <3B3351AA.684FB899@digidata.com> <488e459a.0106270709.302cb0d1@posting.google.com> Couple of comments below. Thanks for the suggestions. Doug John wrote: > Doug Dotson wrote in message news:<3B3351AA.684FB899@digidata.com>... > > As far as I can tell, all you can do is specify an offset of where > > TFFS-managed flash begins. > > > > I can't even get the darn thing (TFFS) to try to identify my > > flash device. It throws an exception from flIdentifyFlash prior > > to attempting to call cfiscsIdentify. > > There's a couple of things to be careful of with this part of the > process: > > 1) Make sure that the flash is not marked read-only in sysPhysMemDesc > (or the region tables for AE) - this will cause the MMU mappings to be > set to read only, and the identify process needs to write to the flash > address to get the device into command mode. > No sysPhysMemDesc table, MMU not being used. > > 2) Make sure that you are specifying the actual start address of the > flash device (even if you are reserving a block for something else). > Obviously, make sure you get the offset right too if you have a boot > image in the same flash ;-) > Got that one. The flash sits in normal memory space does not need a window to access. > > HTH, > > John... > > > Hope this helps, > > Doug > > > > Owain Phillips wrote: > > > > > Hi, > > > > > > I have an 8MB device and the architectural limitation that the boot > > > image > > > has to lie at 4MB. > > > > > > With TFFS can I reserve arbitary regions of the prom that are not to be > > > used in filesystems? IE. Can I build a filesystem in that 4MB below the > > > bootimage? > > > > > > Or can I treat the 8MB device as though it is several devices and hence > > > use the lover 4MB as FSys? > > > > > > Or must I reserve 4MB+bootimageSize and have a filesystem > > > above the bootimage? > > > > > > I look forward to your responses... > > > > > > Regards, > > > Owain Phillips --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Flash File System Alternatives Date: Mon, 02 Jul 2001 13:31:39 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B40AFFB.73FD1CD0@digidata.com> References: <9hcuav$1gn$1@news1.xs4all.nl> There is TinyFS and it is free. Does not do wear leveling and cannlt be accessed through a DOS file system. We decided to use TFFS instead (still not working though). Doug molochai wrote: > In message <9hcuav$1gn$1@news1.xs4all.nl> > "Ton C. Jaspers" wrote: > > > Hi, > > > > Anyone know alternative flash file systems and their suppliers? > > Please don't tell me about M-systems TFFS, that is identical to the FFS > > supplied by Windriver. > > > > DIY depending on the subset of features of TFFS you want and how much time > you have, how many products it will be deployed on etc. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Flash File System Alternatives Date: 2 Jul 2001 10:56:02 -0700 From: aparker@vtlinx.com (parreg) Organization: http://groups.google.com/ Message-ID: <79ef8b60.0107020956.38ca8e3b@posting.google.com> References: <9hcuav$1gn$1@news1.xs4all.nl> molochai wrote in message news:... > In message <9hcuav$1gn$1@news1.xs4all.nl> > "Ton C. Jaspers" wrote: > > > Hi, > > > > Anyone know alternative flash file systems and their suppliers? > > Please don't tell me about M-systems TFFS, that is identical to the FFS > > supplied by Windriver. > > > > DIY depending on the subset of features of TFFS you want and how much time > you have, how many products it will be deployed on etc. Hello, There is the Virtual Small Block File Manager provided by Intel. You can visit their web site to get more info. - -parreg --------------------------- Newsgroups: comp.os.vxworks Subject: Malta BSP & usrNetInit string Date: Mon, 02 Jul 2001 13:59:19 -0400 From: Ben Kuhn Organization: Improv Systems, Inc. Message-ID: <3B40B677.CFA11BAC@improvsys.com> Reply-To: benk@improvsys.com This is a multi-part message in MIME format. - --------------F0C5853119A209409FA5A0EC Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hello all,

I'm relatively new to vxWorks, and am trying to get my system up with a user configured IP address.  I'm using the usrNetInit function, per the docs, but can't seem to get it to execute successfully.  I'm trying to configure a standalone system.  with the Malta4kc Beta BSP on a Malta rev 4 board.  Just coming out of boot into a shell (via the serial port) just about every string I try gets me a return code -1.  If anyone has successfully managed to init this board with usrNetInit, and could post a known good example string, I would be greatful.

I have managed to get an Arm Integrator working with usrNetInit, but just changing the device/ip in the string for Malta didn't seem to work.

Sample session with the Malta:
-> usrNetInit( "lnPci0(0,0) e=10.0.20.155 g=10.0.20.1" )
Host Name: bootHost
Target Name: vxTarget
User: target
value = -1 = 0xffffffff
->

version:
VxWorks (for Malta 4Kc) version VxWorks5.4.2.
Kernel: WIND version 2.5.

Thanks in advance,

Ben Kuhn
Improv Systems
 
  - --------------F0C5853119A209409FA5A0EC Content-Type: text/x-vcard; charset=us-ascii; name="benk.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Ben Kuhn Content-Disposition: attachment; filename="benk.vcf" begin:vcard n:Kuhn;Benjamin tel;cell:716 739 2171 tel;work:(716) 756 - 5411 x25 x-mozilla-html:FALSE adr:;;;;;; version:2.1 email;internet:benk@improvsys.com fn:Benjamin Kuhn end:vcard - --------------F0C5853119A209409FA5A0EC-- --------------------------- Newsgroups: comp.os.vxworks Subject: Alternative cross-compiler for VxWorks Date: 02 Jul 2001 18:15:26 UT From: fscheurer@manz-acs.de Organization: Lawrence Berkeley National Laboratory Message-ID: <9hqdnf$ibn$1@overload.lbl.gov> Hi VxWorkers, I am trying to build an alternative cross tool-chain with configuration: host=i686-pc-cygwin target=i386-wrs-vxworks using the following versions and the cygwin tools: 29.06.01 15:47

binutils-2.11.2 29.06.01 15:53 gcc-2.95.3 29.06.01 16:00 newlib-1.9.0 29.06.01 17:50 gdb-4.18 Some problems could be corrected by the right Environment settings (PATH and LIB variables to cygwin-tools and setting the temporary files etc.), but now I ran into a problem that's: .. >sh .\build.sh Trying to build binutils ... Created "Makefile" in /cygdrive/p/PROJEKTE/GNU_NonWRS_Tools_sources/CROSSGCC/build_newlib1.9.0/build-binutils using "mh-frag" /cygdrive/c/usr/bin/ld: cannot open crt0.o: No such file or directory collect2: ld returned 1 exit status *** The command 'c:/usr/bin/gcc -o conftest -g -O2 conftest.c' failed. *** You must set the environment variable CC to a working compiler. - --print-multi-lib: not found Configuring in ./libiberty loading cache ./config.cache checking host system type... i686-pc-cygwin checking build system type... i686-pc-cygwin checking for ar... i386-wrs-vxworks-ar checking for ranlib... i386-wrs-vxworks-ranlib checking for gcc... gcc checking whether we are using GNU C... yes checking whether gcc accepts -g... yes checking for POSIXized ISC... no checking for working const... yes checking for inline... inline checking for a BSD compatible install... /cygdrive/c/usr/bin/install -c cmp: not found checking how to run the C preprocessor... /lib/cpp checking for sys/file.h... grep: conftest.out: No such file or directory yes checking for sys/param.h... grep: conftest.out: No such file or directory yes checking for limits.h... grep: conftest.out: No such file or directory yes checking for stdlib.h... grep: conftest.out: No such file or directory yes checking for string.h... grep: conftest.out: No such file or directory yes checking for unistd.h... grep: conftest.out: No such file or directory yes checking for strings.h... grep: conftest.out: No such file or directory yes checking for sys/time.h... grep: conftest.out: No such file or directory yes checking for time.h... grep: conftest.out: No such file or directory yes checking for sys/resource.h... grep: conftest.out: No such file or directory yes checking for sys/stat.h... grep: conftest.out: No such file or directory yes checking for sys/mman.h... grep: conftest.out: No such file or directory yes checking for fcntl.h... grep: conftest.out: No such file or directory yes checking for sys/wait.h that is POSIX.1 compatible... no checking whether time.h and sys/time.h may both be included... no checking whether errno must be declared... yes checking whether the C compiler (gcc -g -O2 ) works... no configure: error: installation or configuration problem: C compiler cannot create executables. ... Why does my gcc not create right executables. Running a test program, shows the same problem with "crt0.o" Copying it to the local path let's us hang at "-luser32": .. >gcc -o conftest -g -O2 conftest.c /cygdrive/c/usr/bin/ld: cannot find -luser32 collect2: ld returned 1 exit status .. Someone out there having built already or having experiencs with that cross: host=i686-pc-cygwin target=i386-wrs-vxworks TIA MfG / Kind Regards Friedrich Scheurer, R&D - -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de - -------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS Alignment problem Date: 2 Jul 2001 18:44:07 GMT From: litwin@izzy.jpl.nasa.gov (Todd Litwin) Organization: jpl.nasa.gov Message-ID: <9hqfdn$cin$1@nntp1.jpl.nasa.gov> References: <3m6qh9.oqq.ln@gum.de> Frank Wolf wrote: >We're using TFFS on Hitachi SH3 which has a 32Bit wide access to >RAM and Flash memory. > >When writing data to the TFFS drive we get an exception as the You'll need to modify the MTD read and write functions to deal with this. TrueFFS will ask those functions for any number of bytes on any alignment. The read and write functions will need to interact with the hardware according to the 4-byte alignment requirements and then make the translation to what TrueFFS is requesting as needed. A small (4-byte) internal buffer is useful when implementing this. - -- Todd Litwin Jet Propulsion Laboratory Phone: (818) 354-5028; FAX: (818) 354-5028 Email: Todd.E.Litwin@jpl.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: how to reset the logical partition on a SCSI block device Date: 2 Jul 2001 11:59:40 -0700 From: emartin@zetec.com (Eric Martin) Organization: http://groups.google.com/ Message-ID: <158d62fa.0107021059.22c8a2a0@posting.google.com> We are using scsi MO drives with a PPC2603 cpu - vxworks 5.4 When a disk is changed we need to remove and re-create the SCSI physical device structure and do another scsiBlkDevCreate. Then we do a cbioIoct CBIO_RESET. This seems to work fine, the problem is that 80 bytes of ram is lost everytime we create a new scsiBlkDev. Does anyone know how to reset the logical partition on the existing block device using the new SCSI_PHYS_DEV structure? thanks in advance Eric Martin --------------------------- Newsgroups: comp.os.vxworks Subject: BSP question- device with 2 INT Date: Mon, 2 Jul 2001 12:47:54 -0700 From: "WJ" Organization: Posted via Supernews, http://www.supernews.com Message-ID: Hi, I got a device which has 2 different interrupt level sources. How should one deal with this kind of device(with ARM)? I can think of 1. the driver uses intConnect to connect a single ISR twice to two different level of interrupt, the ISR is reentrant 2. the driver uses intConnect to connect two different ISRs, the ISRs deal with resource sharing between each other if any. 3. the driver use intConnect to connect 1 ISR only. The BSP interrupt libary will return the correct level of interrupt for this source if current interrupt is not for this device, if current interrupt is for this device, then ignore this interrupt. The question is if there is any other approach to deal this kind of device? which approach is more proper for vxWorks or general operating systems? Is this kind of device(with 2 INT) design common? I seldom see that. Thanks, WJ --------------------------- Newsgroups: comp.os.vxworks Subject: Defragmentation for TrueFFS Date: Mon, 2 Jul 2001 14:04:08 -0700 From: "News Account" Organization: Epoch Internet Message-ID: <9hqng4$49e$1@nntp1-cm.news.eni.net> Does anybody know of any methodology or utility out there for defragmenting a VxWorks DOS system that is stored and maintained via TrueFFS? Also I have looked at the M-Systems TrueFFS code and noticed some sort of low level defragmentation option that I believe is supposed to be accessable either through the "tffsRawio" routine or some sort of "ioctl" call, however I was unable to get this to work at all. At this point we would even be willing to utilize a third party software to perform the defragmentation task. Please help!!! Regards, P. Eichler --------------------------- Newsgroups: comp.os.vxworks Subject: Memory requirments.... Date: Mon, 02 Jul 2001 14:59:44 -0700 From: "Michael Morrison" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hqrpd$rj0$1@overload.lbl.gov> Anyone know where I can find information on the memory requirments/footprint of VxWorks and it's various components? I just need a rough idea of how much memory our box will need. Thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can WindView lose events? Date: Mon, 02 Jul 2001 16:33:34 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B4104CE.92D3F18@covad.net> References: <28f198ce.0107020907.5dcff2c7@posting.google.com> Reply-To: drdiags@covad.net Eitan, Todd Sandor (sp) posted this same question 2 weeks ago, in regards to Q1. If you have a scenario that is repeatable, could you send it to me. I have been tracing a similar problem, but I cannot get a test case that will produce a consistent problem. This way, we can get an SPR submitted by WRS and have this fixed (hopefully). As far as Q2, sorry, I don't have a clue. Eitan Yacobi wrote: > > Hi guys, > > Maybe you could help me with this one: > > I use Tornado version 2.0 and VxWorks version 5.4. I use WindView to > monitor the performance of the system. > > Q1) When the CPU or the network are loaded, is possible that WindView > will loose events (i.e. events that happened in the system will not be > written to the WindView log)? > > Q2) I have selected the maximal event-mask on WindView, using the > "WindView Selection Configuration" screen. Still, in the WindView log > there are periods of time that no task or interrupt are in running > state. What is the right way to specify an event-mask that will > guaranty that during the log duration, everything that acquires the > CPU will appear in the log? > > Thanks a lot, > Eitan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile problem Date: 2 Jul 2001 17:59:04 -0700 From: khamon@lintek.com (krhohio) Organization: http://groups.google.com/ Message-ID: References: <9hp016$o73$1@overload.lbl.gov> Mike, I add a variable definition to my makefile invocation (we use a Microsoft Dev Studio "makefile project") to select the MACH_EXTRA setting; i.e. "make buildvxworksboot BuildType=bootrom" with the following code in makefile: ############################################################# MACH_EXTRA = if_esyf.o ifeq "$(BuildType)" "app" MACH_EXTRA += $(WIND_BASE)/target/config/vxipc/vxi/nivxi.o endif ############################################################# FWIW, Kent P.S. Thanks for all your other reply postings! "Mike Anderson" wrote in message news:<9hp016$o73$1@overload.lbl.gov>... > Greetings! > > I'm trying to modify a VxWorks Makefile to handle the > case when you build a bootrom vs. a normal VxWorks image. > The MACH_EXTRA defines get included in both cases. This > causes a problem in the bootroms because of another bug in > VxWorks using the WDB_COMM_END communications type and T202. > What I'm trying to do is to look at the type of target that is > being built and adjust the MACH_EXTRA parameter accordingly. > To that end, I have a little snippet of Makefile here: > > ################################################################### > foundname = $(findstring boot, $@) > > ifneq ($(foundname), ) > EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=NORMAL > MACH_EXTRA = ../PentiumBSP/default/demo.o \ > ../PentiumBSP/default/wan68.o \ > ../PentiumBSP/default/osal.o \ > ../PentiumBSP/default/sysWanicEnd.o \ > ./usrAppInit.o \ > ./fei82557End.o \ > ./wdbEndPktDrv.o > else > EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=BOOTROM > MACH_EXTRA = > endif > #################################################################### > > foundname is being set correctly. It contains "boot" if > the target had the sequence "boot" in it (e.g., bootrom, > bootrom_uncmp, etc.) or is blank otherwise. > > However, the comparisons just don't seem to be working. > For some reason, I always get the else clause firing. If > anyone has any ideas or pointers as to what could be going > wrong, I'd sure appreciate the help. > > TIA, > Mike Anderson --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Scheduling Date: Mon, 2 Jul 2001 17:58:02 -0700 From: "Sunil Verma" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hr6au$e30$1@overload.lbl.gov> This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_001_01C1035B.36316A80 Content-Type: text/plain; charset="iso-8859-1" Hi, Could anyone tell the behavior of VxWorks scheduler in following situation ? I have two tasks Task_A (priority 100) & Task_B (priority 90). 1. Task_A calls func_1() which does some processing and takes a semaphore (mutex). [it has still not called func_3()] 2. Task_B being higher priority is scheduled, func_2() starts executing and comes to semTake(), but since Task_A already has the semaphore, Task_B is blocked. [Task_B is placed the queue to take the semaphore next] 3. Task_A continues and func_3() is called which does a * TaskLock() * releases the semaphore * does some processing * tries to take the semaphore again Here lies my doubt, will the semaphore be given to Task_A (being in execution and having TaskLock) or will it be placed in the queue to take semaphore after Task_B() and TaskLock is released to avoid deadlock. I want to know how VxWorks handles this situation. The functions (func_1, func_2 & func_3) used in above example are detailed below. I would also appreciate if you guys could tell me where to find detailed implementation of VxWorks Scheduler ? I also wanna know if guys @ VxWorks fixed some bugs in memory manager recently ?? What is the latest released version of Vxworks ? Thanks, Sunil func_1() { ... semTake(mySemId, WAIT_FOREVER); ... ... func_3(); ... semGive(mySemId); } func_2() { ... semTake(mySemId, WAIT_FOREVER); ... } func_3() { taskLock(); semGive(mySemId); ... semTake(mySemId, WAIT_FOREVER); taskUnlock(); } - ------_=_NextPart_001_01C1035B.36316A80 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable VxWorks Scheduling

Hi,

Could anyone tell the behavior = of VxWorks scheduler in following situation ?

I have two tasks Task_A = (priority 100) & Task_B (priority 90).
1. Task_A calls func_1() which = does some processing and takes a semaphore (mutex). [it has still not = called func_3()]
2. Task_B being higher = priority is scheduled, func_2() starts executing and comes to = semTake(), but since Task_A already has the semaphore, Task_B is = blocked. [Task_B is placed the queue to take the semaphore = next]

3. Task_A continues and = func_3() is called which does a
* TaskLock()
* releases the = semaphore
* does some processing
* tries to take the semaphore = again

Here lies my doubt, will the = semaphore be given to Task_A (being in execution and having TaskLock) = or will it be placed in the queue to take semaphore after Task_B() and = TaskLock is released to avoid deadlock. I want to know how VxWorks = handles this situation.

The functions (func_1, func_2 = & func_3) used in above example are detailed below.

I would also appreciate if you = guys could tell me where to find detailed implementation of VxWorks = Scheduler ?
I also wanna know if guys @ = VxWorks fixed some bugs in memory manager recently ??
What is the latest released = version of Vxworks ?

Thanks,
Sunil


func_1()
{
    ...
    = semTake(mySemId, WAIT_FOREVER);
    ...
    ...
    = func_3();
    ...
    = semGive(mySemId);
}

func_2()
{
    ...
    = semTake(mySemId, WAIT_FOREVER);
    ...
}

func_3()
{
    = taskLock();
    = semGive(mySemId);
    ...
    = semTake(mySemId, WAIT_FOREVER);
    = taskUnlock();
}

- ------_=_NextPart_001_01C1035B.36316A80-- --------------------------- Newsgroups: comp.os.vxworks Subject: mkboot.c that supports DosFsLib2? Date: Mon, 2 Jul 2001 22:32:21 -0400 From: "Mike Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hrdc1$j9d$1@overload.lbl.gov> VxWorks Greetings! Has anyone out there in net land already modified the mkboot.c code to support installing the boot block on an IDE drive for use with DosFsLib2? The code included with the BSP works if your hard disk has a floppy disk format on it (ala DosFsLib1), but gives an error of Can't find primary DOS partition. Any help or pointers would be appreciated. TIA, Mike Anderson - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: Does Mibcomp also process "SEQUENCE" "Table " in SNMP? Date: Tue, 3 Jul 2001 12:34:15 +0800 From: "kumo Yeh" Organization: SEEDNet News Service Message-ID: <9hri4e$2go$1@news.seed.net.tw> Hi When I use mibcomp as down: .. ... C:\Tornado\target\src\snmpv1\mibs>mibcomp -l $WIND_BASE/target/src/snmpv1/mibs - stub -start proFile -o ppp.c rfc1155.smi rfc1213.mib ppp.mib Wind River Systems Emissary SNMP MIB Compiler, version 7.0. Copyright (c) 1989-1996, Wind River Systems Inc. C:\Tornado\target\src\snmpv1\mibs> ... .. The ppp.c gernerated,but doesn't has function with SEQUENCE or TABLE Why??? Thank you Best regards - -- Kumo,Yeh. Firmware Engineer GLORY TELECOM. Phone:(886)-3-3127832-386 Fax:(886)-3-3127831 http://www.glorytel.com.tw email:kumo@glorytel.com.tw --------------------------- Newsgroups: comp.os.vxworks Subject: Vxsim very small fonts in output window Date: Tue, 3 Jul 2001 07:42:16 +0300 From: "Eden Cohen" Organization: - Message-ID: <9hrijf$al1$1@news.netvision.net.il> Reply-To: "Eden Cohen" I am using Tornado 2 on Win98 and WinNT and get the same problem: when running Vxsim, the output window somehow gets a very small (and unreadable) font. Can I change this behaviour ? Thanks, Eden --------------------------- Newsgroups: comp.os.vxworks Subject: à÷ãà Date: Tue, 3 Jul 2001 07:41:00 +0300 From: "Eden Cohen" Organization: - Message-ID: <9hrih3$akv$1@news.netvision.net.il> Reply-To: "Eden Cohen" --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Makefile Problem Date: Tue, 3 Jul 2001 10:17:11 +0530 From: "Palaniar" Organization: Lawrence Berkeley National Laboratory Message-ID: <000101c1037b$3997e080$4106140a@future.futsoft.com> Hi, The ifneq will always obviously execute the else part, because the condition will always fail. (i.e), you have checked whether the foundname exists or not rather foundname has something or not. Hence, foundname will return true and that's why the execution of ifneq goes to the else part. Alternatively, use the following command: ifneq ($(foundname),"boot") else endif For the above set of instructions, if foundname == "boot", then else part will get executed, otherwise the first part will get executed. Regards ARP >From vxwexplo-errs@csg.lbl.gov Sun Jul 1 22:01:48 2001 From: "Mike Anderson" Date: Sun Jul 1 22:01:50 PDT 2001 Subject: Makefile problem Greetings! I'm trying to modify a VxWorks Makefile to handle the case when you build a bootrom vs. a normal VxWorks image. The MACH_EXTRA defines get included in both cases. This causes a problem in the bootroms because of another bug in VxWorks using the WDB_COMM_END communications type and T202. What I'm trying to do is to look at the type of target that is being built and adjust the MACH_EXTRA parameter accordingly. To that end, I have a little snippet of Makefile here: ################################################################### foundname = $(findstring boot, $@) ifneq ($(foundname), ) EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=NORMAL MACH_EXTRA = ../PentiumBSP/default/demo.o \ ../PentiumBSP/default/wan68.o \ ../PentiumBSP/default/osal.o \ ../PentiumBSP/default/sysWanicEnd.o \ ./usrAppInit.o \ ./fei82557End.o \ ./wdbEndPktDrv.o else EXTRA_DEFINE = -DCPU_VARIANT=PENTIUM -DBUILD_TYPE=BOOTROM MACH_EXTRA = endif #################################################################### foundname is being set correctly. It contains "boot" if the target had the sequence "boot" in it (e.g., bootrom, bootrom_uncmp, etc.) or is blank otherwise. However, the comparisons just don't seem to be working. For some reason, I always get the else clause firing. If anyone has any ideas or pointers as to what could be going wrong, I'd sure appreciate the help. TIA, - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time. --------------------------- Newsgroups: comp.os.vxworks Subject: Use of MemScope issue - TCB corruption Date: 2 Jul 2001 22:34:49 -0700 From: pignorel@yahoo.com (sonia pignorel) Organization: http://groups.google.com/ Message-ID: <36b73c7d.0107022134.5164157c@posting.google.com> Has someone already seen any issue using the MemScope utility? It seems that under certain circumstances having MemScope running would lead to corruption of TCB. Thanks by advance for any feedback. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS limitations? -- No limitition, here's what you do... Date: 2 Jul 2001 22:46:27 -0700 From: daveg@zaffire.com (Dave Gilchrist) Organization: http://groups.google.com/ Message-ID: References: <3B31BDA2.AEEA9B89@siemenscomms.co.uk> Owain Phillips wrote in message news:<3B31BDA2.AEEA9B89@siemenscomms.co.uk>... > Hi, > > I have an 8MB device and the architectural limitation that the boot > image > has to lie at 4MB. > > With TFFS can I reserve arbitary regions of the prom that are not to be > used in filesystems? IE. Can I build a filesystem in that 4MB below the > bootimage? > > Or can I treat the 8MB device as though it is several devices and hence > use the lover 4MB as FSys? > > Or must I reserve 4MB+bootimageSize and have a filesystem > above the bootimage? > > I look forward to your responses... > > Regards, > Owain Phillips > > > > -- We took an entirely different approach to this problem. We put our bootroms in the fallow area of flash, then set up the bootroms to mount tffs which then inflates our vxWorks image into ram and jumps into it. To clarify, we put our compressed vxWorks application image into tffs as a file (/RFA0/vxWorks.Z). Advantages: We put a compressed image into flash thus conserving flash resource. You should be able to get about a 70% reduction in image size doing this. Booting is lightning fast! And with symbols. This not only reduces bootup time but also streamlines the development cycle because the image does not have to get sucked in over the network. You just boot then attach the debugger and you are ready to go! Disadvantage: You gotta do a fair amount of stuff to your bootroms, but the effort is well worth it. BTW, this approach was originally used by Mars Pathfinder, whom we got help from thru WRS... See the vxWorks man page on 'inflate()' and the elftobin/deflate/etc... utilities for details. Also, in the bootroms, you will want to do a 'go 0x1000' or whatever. The '0x1000' is the location of the first executable statement (memory before that is reserved as a kernel scratch area). One other piece of advice. Don't chop flash up into chunks. Just go ahead and use on contiguous region. We made the mistake of treating each flash bank as a filesystem and now we regret it because we switched our flash hardware, resulting in side effects to application code which depended on the multiple device flash configuration. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: WindView post-mortem: no luck yet Date: Tue, 03 Jul 2001 08:52:39 +0200 From: Ton Janssen Organization: Oce Technologies B.V. Message-ID: <3B416BB7.C6465742@oce.nl> References: <3B3C96DB.D012CF9E@oce.nl> Ton Janssen wrote: > > Hi there, > > Anyone experience with post-mortem mode Windview logging: > > I've got the following problem: I'm trying to debug a situation > were the last thing that happens in a session is a hangup in > (probably) interrupt context. So I tried to use post-mortem > logging in WindView, together with the addition of a whole > bunch of user events (wvevent(nr,0,0)) in the suspected code > area. > > The problem is that when I hit the stop button, the target application > hangs at interrupt level. After a reboot (the target has disconnected > and connected again) I try to upload the windView buffer via NFS. > > However, the tool now complaints that the target buffer has not > been created; The buffer is created in an area which is not cleared > at boot-time, so it should work... > The exact error message was: " WV Buffer (0x1d9108) not yet created on target " Even after a warm reboot. I configured a post mortem upload buffer from 0x1f30000 upto 0x1fffff0. The settings are not the same as sysMemTop() and sysPhysMemTop() because our buffer is not contigious to the normal memory. The latter 2 funcs both return the value: 0xc1d000. So I did'nt use the USER_RESERVED_MEMORY macro also. I'm still using the old (Tornado 1.0.1) EVTBUFFER_ADDRESS and _SIZE. The specified buffer is battery backup-ed so it should work even after a cold reboot. Is it possible that the wvUploadTask uses sysMemTop and sysPhysMemTop instead of the values supplied at specification of the post mortem buffer? It probably does when trying first; see below... - --- answer --- I have the post-mortem log file now (;-) I still get the first error message after a reboot. But then, when I configure the upload mode again in WindView I get a pop-up window: Post mortem buffer exists. Obviously the tool then has the exact information concerning the ring buffer were the post mortem log was created. Then the upload succeeds. - --- end answer --- Thanks for all replies. - -- ############################################################ # Oce Technologies B.V. name: Ton Janssen (3N-38) # # P.O. Box 101 email: djan@oce.nl # # 5900 MA Venlo Private: dpc.janssen@hccnet.nl # # Tel: +31 (0)77-359 40 97 # # The Netherlands Fax: +31 (0)77-359 54 50 # ############################################################ This note does not necessarily represent the position of Oce Technoligies B.V. Therefore no liability or responsibility for whatever will be accepted. --------------------------- Newsgroups: comp.os.vxworks Subject: about "taskSpawn" Date: Tue, 3 Jul 2001 16:02:08 +0900 (KST) From: benjamin Organization: KORNET WEB NEWS Message-ID: <9hrqlg$mr0$1@news1.kornet.net> When creating a task with the function "taskSpawn", we must have information about "stack size". As vxworks's beginner, I like to know why the stack size is needed and what it is used for. I hope anyone who know the answers gives me those. - -------------- ÀÎÅÍ³Ý Ä«¸®½º¸¶ KORNET ------------- --------------------------- Newsgroups: comp.os.vxworks Subject: I want to know "shell". Date: Tue, 3 Jul 2001 16:14:34 +0900 From: "¿ÁÀ»¼®" Organization: Korea Telecom Message-ID: <9hrrep$qoc$1@news1.kornet.net> I use a standalone image without shell. but, because I must support "telnetd", I used 'ioTaskStdGet/ioTaskStdSet' instead of 'ioGlobalStdGet/ShellOrigStdSet". and restarted "a task like shell" but, not working. I must have restarted "Shell". I expect that "Shell" process fd(s). I want to know "Shell". any Information or shell source make me happy. Thanks in advanced. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS Alignment problem Date: Tue, 3 Jul 2001 09:03:21 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: <6lqrh9.k1u.ln@gum.de> References: <3m6qh9.oqq.ln@gum.de> <9hqfdn$cin$1@nntp1.jpl.nasa.gov> Hi again! We've modified the MTD to deal with the 32bit access - which is working (AFAIK). The problem is that TFFS access is causing an exception in kernel code not in our code! We've seen that TFFS is (internally) calling a function 'writeSector' which is calling 'checkForWriteInplace'. WriteSector is passing an odd aligned buffer (2byte aligned - not 4byte as needed) to the check function which is crashing then... Thanks again, Frank Wolf "Todd Litwin" schrieb im Newsbeitrag news:9hqfdn$cin$1@nntp1.jpl.nasa.gov... > Frank Wolf wrote: > > >We're using TFFS on Hitachi SH3 which has a 32Bit wide access to > >RAM and Flash memory. > > > >When writing data to the TFFS drive we get an exception as the > > You'll need to modify the MTD read and write functions to deal with this. > TrueFFS will ask those functions for any number of bytes on any alignment. The > read and write functions will need to interact with the hardware according to > the 4-byte alignment requirements and then make the translation to what TrueFFS > is requesting as needed. A small (4-byte) internal buffer is useful when > implementing this. > > -- > Todd Litwin > Jet Propulsion Laboratory > Phone: (818) 354-5028; FAX: (818) 354-5028 > Email: Todd.E.Litwin@jpl.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS Alignment problem Date: Tue, 3 Jul 2001 09:12:54 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: <27rrh9.i2u.ln@gum.de> References: <3m6qh9.oqq.ln@gum.de> <9hqfdn$cin$1@nntp1.jpl.nasa.gov> Just an addition to my last mail: We've also tried a configuration with 16Bit access to Flash memory using one of the predefined MTDs with slight modifications. This configuration is also crashing! Is there any possibility to set an alignment for the internal buffers of TFFS? (Maybe 'flSetWindowBusWidth' or another undocumented function?) Frank Wolf "Todd Litwin" schrieb im Newsbeitrag news:9hqfdn$cin$1@nntp1.jpl.nasa.gov... > Frank Wolf wrote: > > >We're using TFFS on Hitachi SH3 which has a 32Bit wide access to > >RAM and Flash memory. > > > >When writing data to the TFFS drive we get an exception as the > > You'll need to modify the MTD read and write functions to deal with this. > TrueFFS will ask those functions for any number of bytes on any alignment. The > read and write functions will need to interact with the hardware according to > the 4-byte alignment requirements and then make the translation to what TrueFFS > is requesting as needed. A small (4-byte) internal buffer is useful when > implementing this. > > -- > Todd Litwin > Jet Propulsion Laboratory > Phone: (818) 354-5028; FAX: (818) 354-5028 > Email: Todd.E.Litwin@jpl.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Memory Management Date: 3 Jul 2001 01:02:24 -0700 From: karl_ingstrom@email.com (=?ISO-8859-1?Q?Karl_Ingstr=F6m?=) Organization: http://groups.google.com/ Message-ID: Hi I am doing an investegation about how memory managemnt works in some rtos:s and wounder if any one knows about any memory management documentation for vxWorks. or any comments on the vxWorks memory management fuctionality is welcome. Also memory management in VRTX,eCos or OSE is of interest. Regards Karl Ingström --------------------------- Newsgroups: comp.os.vxworks Subject: How can i print the stack call after an error ? Date: Tue, 3 Jul 2001 08:17:03 +0000 (UTC) From: AsafH@Radlan.co.il (Asaf Harari) Organization: Mailgate.ORG Server - http://www.Mailgate.ORG Message-ID: <42AB6BE23C29D511831E0002B32024880FB0B7@MESSENGER> Hi , i need a code for printing the stack call after an error . For example when i write "os fatal error" . How can i use the pc/sp for finding the right lines of code ? Thanks. - -- Posted from sphere.barak.net.il [212.150.48.98] via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- Newsgroups: comp.os.vxworks Subject: porting of vxworks to threadx Date: 3 Jul 2001 03:00:16 -0700 From: m_nischal@mailcity.com (nish) Organization: http://groups.google.com/ Message-ID: anybody who has done porting of vxworks to threadx ,do respond . --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks configuration Date: Tue, 03 Jul 2001 11:48:12 +0100 From: Paul Caswell Organization: Land & Sea Systems Message-ID: <3B41A2EB.F095C704@baesystems.com> References: <9h76g0$k14$1@overload.lbl.gov> I think you need to do the following: In configure target server you should have selected wdbserial as the backend. From the drop-down list next to 'target server properties' select 'core file and symbols' then check 'file' and enter the path to the image that you have downloaded. Now try and lauch the target server again. It worked for me. Regards Paul Caswell --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Tue, 03 Jul 2001 11:20:54 +0100 From: Simon Farnsworth Message-ID: <3B419C86.2010104@snellwilcox.com> References: <9hrqlg$mr0$1@news1.kornet.net> benjamin wrote: > When creating a task with the function > "taskSpawn", we must have information > about "stack size". > > As vxworks's beginner, I like to know why the stack size is needed and what it is used for. The stack is used by the system for storing local variables and return addresses; to avoid the non-deterministic behaviour introduced by resizing a stack, VxWorks requires you to set a limit at startup. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 3 05:13:23 2001 From: fscheurer@manz-acs.de Date: Tue Jul 3 05:13:26 PDT 2001 Subject: Where is this telnetd from? Hi VxWorkers! some time ago I downloaded the attached source code. I also saw a HTML description and a Makefile but didn't store it right, so it's gone. Anyone knows where I found it? Probably the people of MBARI? Any hint appreciated! TIA MfG / Kind Regards Friedrich Scheurer, R&D -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de -------------------------------- /****************************************************************************/ /* Copyright 1992 to 1998 MBARI */ /****************************************************************************/ /* Summary : telnetd skeleton */ /* Filename : telnetd.c */ /* Author : Andrew Pearce */ /* Project : */ /* Version : Version 1.0 */ /* Created : 07/17/98 */ /* Modified : 07/17/98 */ /* Archived : */ /****************************************************************************/ /* Modification History: */ /* $Header: $ * $Log: $ */ /****************************************************************************/ #include /* vxWorks system declarations */ #include /* vxWorks task library functions */ #include /* vxWorks standard I/O functions */ #include /* vxWorks I/O library functions */ #include /* vxWorks string functions */ #include /* vxWorks Linked List library */ #include #include /* vxWorks internet protocol definitions*/ #include /* vxWorks network socket functions */ #include /* vxWorks select I/O functions */ #include /* mbari style guide type declarations */ #include /* Miscellaneous constants */ #define MAX_QUEUED_CONNECTIONS 4 /* Number of enqueued connect requests */ #define NET_TASK_PRIORITY 80 /* Telnet Task Priority */ #define NET_TASK_STACKSIZE 4000 /* Telnet Task Stacksize & TCB */ #define TELNET_PORT_NO 23 /* Default telnet port number */ /* Note this port will conflict with */ /* the vxWorks telnetd. Use a different */ /* port number or rebuild vxWorks with */ /* INCLUDE_TELNET undefined */ MLocal Void myTnTask( Int32 socketFd ) { Int32 on; on = TRUE; /* Turn on non-blocking for read */ ioctl(socketFd, FIONBIO, (int) &on); FOREVER { if (write(socketFd, "Message from telnet task tnTask\n", 32) == ERROR) { close(socketFd); return; } /* if */ taskDelay(sysClkRateGet()); } /* FOREVER */ } /* myTnTask() */ STATUS telnetDeamon( Void ) { struct sockaddr_in myAddr; struct sockaddr_in clientAddr; int clientAddrLen; Int32 socketFd; Int32 newFd; Int32 port = TELNET_PORT_NO; /* Default telnet port number */ /* create a socket for each port */ socketFd = socket(AF_INET, SOCK_STREAM, 0); bzero ((char *) &myAddr, sizeof (myAddr)); myAddr.sin_family = AF_INET; myAddr.sin_port = htons (port); if (bind (socketFd, (struct sockaddr *) &myAddr, sizeof (myAddr)) == ERROR) { printf ("telnet bind failed. 0x%x\n", errnoGet()); return(ERROR); } /* if */ /* Create client connect request queue */ if (listen (socketFd, MAX_QUEUED_CONNECTIONS) == ERROR) { printf ("telnet listen failed.\n"); return(ERROR); } /* if */ FOREVER { /* Wait for connection requests on all */ /* sockets using select */ errnoSet (OK); /* set error number to OK */ /* accept connection and spawn task to */ /* handle it */ clientAddrLen = sizeof (clientAddr); if ((newFd = accept (socketFd, &clientAddr, &clientAddrLen)) == ERROR) { printf ("telnet accept failed status = 0x%x\n", errnoGet ()); close (socketFd); } /* if */ if (taskSpawn("myTnTask", NET_TASK_PRIORITY, VX_FP_TASK, NET_TASK_STACKSIZE, (FUNCPTR) myTnTask, newFd, 0, 0, 0, 0, 0, 0, 0, 0, 0) == ERROR) { printf("taskSpawn failed\n"); close (newFd); } /* if */ } /* FOREVER */ } /* telnetDeamon() */ From vxwexplo-errs@csg.lbl.gov Tue Jul 3 06:25:39 2001 From: "Gary LeVan" Date: Tue Jul 3 06:25:42 PDT 2001 Subject: RE: Vx Exploder Digest --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Tue, 03 Jul 2001 11:20:54 +0100 From: Simon Farnsworth Message-ID: <3B419C86.2010104@snellwilcox.com> References: <9hrqlg$mr0$1@news1.kornet.net> benjamin wrote: > When creating a task with the function > "taskSpawn", we must have information > about "stack size". > > As vxworks's beginner, I like to know why the stack size is needed and what it is used for. The stack is used by the system for storing local variables and return addresses; to avoid the non-deterministic behaviour introduced by resizing a stack, VxWorks requires you to set a limit at startup. To start, Make the Stack size very large (how about 10000?) which of course depends on your application. Exercise your application fully and use checkStack( )to check amount used and adjust as necessary. Suggest leaving stack size at least double what you think you need. Exceeding maxStack size will walk on someone else and make you unhappy. From vxwexplo-errs@csg.lbl.gov Tue Jul 3 06:26:53 2001 From: "Manjunath Bhat" Date: Tue Jul 3 06:26:55 PDT 2001 Subject: Tornado simulator VxWorks Hi, Do anybody of you know how to increase the target memory size for the Tornado II simulator? Thanking you, Manjunath Bhat. _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. From vxwexplo-errs@csg.lbl.gov Tue Jul 3 07:32:45 2001 From: "lhp" Date: Tue Jul 3 07:32:48 PDT 2001 Subject: align malloc and free in VxWorks Hi, VxWorks 5.4 for PPC. I use following routine to malloc a block of cache-coherent memory which aligns parameter "Align". Align is a value which is power of 2. void* commonCacheDmaMallocAlign(UINT32 Align,UINT32 Size) { char* memory1, memory2; memory1 = (char*)cacheDmaMalloc(Size + Align); /* check if allocation successes */ if(memory1 == NULL) { printf("can"t allocate memory\n"); return NULL; } memory2 = ((char*)(((UINT32)(memory1 + Align)) & (~(Align-1)))); return (void*)memory2; } My Question: Is it safe and correct when I free the memory malloced by commCacheDmaMallocAlign() like this? void *p; p = commonCacheDmaMallocAlign(16, 1000); ... cacheDmaFree(p); Will the memory between "memory1" and "memory2" be lost? I don't know how the VxWorks kernel deal with it internally. Who can tell me? Certaily, we can save the value of original "memory1" first and use cacheDmaFree(memory1) to free the whole memory. I know this is absolutely safe and correct. MUST we do like this? I want an exact answer. From vxwexplo-errs@csg.lbl.gov Tue Jul 3 09:23:49 2001 From: Anuradha Pani Date: Tue Jul 3 09:23:51 PDT 2001 Subject: signal handling in vxworks. hi all, i am having some problem with catching signals. I have attached the following piece of code. I am calling Start to spawn the signal handler task, signalHandler and then sending a signal to that task, but nothing happens. I don't get any output from the sigHandler function. Please tell me what I'm doing wrong. Thanks Anuradha /* File: signalHandler.c */ /* Include files */ #include "stdio.h" #include "vxWorks.h" #include "sigLib.h" #include "taskLib.h" /* Function prototypes */ void signalProcess (int signalNum); void signalHandler(void) { struct sigaction act; int i; printf("Signal Handler started\n"); /* Install a signal handler */ act.sa_handler = &signalProcess; /* Function to be called */ sigfillset(&act.sa_mask); /* Block all other signals while signalProcess is executing */ act.sa_flags = 0; if ( sigaction(SIGUSR1, &act, NULL) == -1 ) { printf("Signal processor could not be established\n"); exit(1); } /* Remain in a loop */ for (i=0;i<1000;i++) { taskDelay(30); printf("Signal handler running\n"); } } /* End of signalHandler */ /* Signal Processor Function */ void signalProcess (int signalNum) { printf("SIGNAL CAUGHT\n"); } /* End of signal processor */ void Start(void) { int taskId; /* Spawn the signal handler process */ if ((taskId = taskSpawn("signal2", 100, 0, 2000, (FUNCPTR)signalHandler,0,0,0,0,0,0,0,0,0,0)) == -1) { printf("Start FAILED to launch signal handler\n"); } printf("Signal Handler Spawned\n"); taskDelay(30); kill(taskId, SIGUSR1); } /* End of start */ __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ From vxwexplo-errs@csg.lbl.gov Tue Jul 3 10:05:49 2001 From: francois.gorisse@modicon.com Date: Tue Jul 3 10:05:51 PDT 2001 Subject: Re : Assembler with VxWorks Hi Harman, May be the code below will help you. I use gcc for 486 under vxworks on a Windows NT PC platform with Tornado 2 T2P4. Note : Some compiler option are 02 -xc++. void outWrds(void) { unsigned short port = 0x378; unsigned char value = 0xff; __asm__ volatile ("outb %1,%0;" : : "d" (port) , "a" (value) ); } and you 'll get under the shell : -> l outWrds outWrds(void) 00030d60 55 PUSH EBP 00030d61 89 e5 MOV EBP, ESP 00030d63 ba 78 03 00 00 MOV EDX, 0x378 00030d68 b0 ff MOV AL, 0xff 00030d6a ee OUT DX, AL 00030d6b 89 ec MOV ESP, EBP 00030d6d 5d POP EBP 00030d6e c3 RET I hope it could serve you, Francois. From vxwexplo-errs@csg.lbl.gov Tue Jul 3 13:51:00 2001 From: Anuradha Pani Date: Tue Jul 3 13:51:02 PDT 2001 Subject: problem with timers --0-1957747793-994193458=:28046 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline hi all, i am new to vxworks. attached is a simple piece of code where i have tried to implement timers. but when i start the task, startTimer, nothing happens. i see no output at all. not even the "hi" i deliberately put in. can somebody help me? thanks anuradha __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ --0-1957747793-994193458=:28046 Content-Type: application/x-unknown; name="timerImp.c" Content-Transfer-Encoding: base64 Content-Description: timerImp.c Content-Disposition: attachment; filename="timerImp.c" LyogRmlsZTogdGltZXJJbXAuYyAtIENyZWF0ZXMgYW5kIHNldHMgYSB0aW1l ciAqLwoKLyogSW5jbHVkZSBmaWxlcyAqLwojaW5jbHVkZSAic3RkaW8uaCIK I2luY2x1ZGUgInZ4V29ya3MuaCIKI2luY2x1ZGUgInRhc2tMaWIuaCIKI2lu Y2x1ZGUgInNpZ0xpYi5oIgojaW5jbHVkZSAidGltZS5oIgoKCgovKiBGdW5j dGlvbjogc3RhcnRUaW1lcigpCiAgIFRoaXMgZnVuY3Rpb24gc3Bhd25zIGEg dGFzayB0byBjYXRjaCB1c2VyIHNpZ25hbHMuIAogICBJdCBzdGFydHMgYSB0 aW1lciB3aXRoIGEgdGltZS1wZXJpb2Qgb2YgNTAgbXMgYW5kIHVwb24gZXhw aXJ5CiAgIHNlbmRzIHVzZXIgc2lnbmFscyB0byB0aGUgdGFzayBzcGF3bmVk LgogICBJdCB0aGVuIHJlX2FybXMgdGhlIHRpbWVyLgoqLwp2b2lkIHN0YXJ0 VGltZXIodm9pZCkKewogIHN0cnVjdCBpdGltZXJzcGVjIHRpbWVQZXJpb2Q7 CiAgdGltZXJfdCB0aW1lcklkOwogIGludCBpOwoKCiAgLyogRnVuY3Rpb24g cHJvdG90eXBlcyAqLwogIHZvaWQgdG1yRXhwaXJ5SGFuZGxlcihpbnQgc2ln bmFsKTsKICAKICAgcHJpbnRmKCJIaVxuIik7CgogICAvKiBDcmVhdGUgYSB0 aW1lciAqLwogICBpZiggdGltZXJfY3JlYXRlKENMT0NLX1JFQUxUSU1FLCBO VUxMLCAmdGltZXJJZCkgPT0gLTEpCiAgICB7CiAgICAgICAgcHJpbnRmKCJF cnJvciBpbiBjcmVhdGluZyB0aW1lclxuIik7CglleGl0KDEpOwogICAgfQog ICBlbHNlCiAgICB7CiAgICAgIHByaW50ZigiVGltZXIgd2l0aCBpZCA9ICVk ICAgIGNyZWF0ZWQgc3VjY2Vzc2Z1bGx5XG4iLCB0aW1lcklkKTsKICAgICAg LyogQXR0YWNoIGEgdGltZXIgcm91dGluZSB0byB0aGUgdGltZXIgKi8KICAg ICAgaWYoIHRpbWVyX2Nvbm5lY3QodGltZXJJZCwgKHZvaWQqKSZ0bXJFeHBp cnlIYW5kbGVyLCAwKSA9PSAtMSkKICAgICAgIHsKCSBwcmludGYoIkVycm9y IGluIGF0dGFjaGluZyB0aW1lciByb3V0aW5lXG4iKTsKCSBleGl0KDEpOwog ICAgICAgfQogICAgICBlbHNlCiAgICAgICB7CgkgcHJpbnRmKCJUaW1lciBy b3V0aW5lIGF0dGFjaGVkIHN1Y2Nlc3NmdWxseVxuIik7CgkgLyogU2V0IHRp bWVyIHBlcmlvZCAqLwoJICB0aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X3Nl YyA9IDA7ICAgLyogTm8gcmVwZWF0YXRpb25zIGZvciBub3cgKi8KICAJICB0 aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X25zZWMgPSAwOyAgCiAgCSAgdGlt ZVBlcmlvZC5pdF92YWx1ZS50dl9zZWMgPSAwOwogICAgICAgICAgdGltZVBl cmlvZC5pdF92YWx1ZS50dl9uc2VjID0gNTAwMDAwMDA7IC8qIGZpcnN0IGV4 cGlyeSBhZnRlciA1MCBtcyAqLwoKCSAgLyogRGVsYXkgZm9yIHNvbWV0aW1l ICovCgkgIHRhc2tEZWxheSgzMCk7CgoJICBpZiggdGltZXJfc2V0dGltZSh0 aW1lcklkLCAhVElNRVJfQUJTVElNRSwgJnRpbWVQZXJpb2QsIE5VTEwpID09 IC0xKQogICAgICAgICAgIHsKICAgICAgCSAgICAgICBwcmludGYoIkVycm9y IGluIHNldHRpbmcgdGltZXJcbiIpOwogICAgICAJICAgICAgIGV4aXQoMSk7 CiAgICAgICAgICAgfQogICAgICAgICAgIHByaW50ZigiVGltZXIgc2V0IHN1 Y2Nlc3NmdWxseVxuIik7CiAgICAgICB9CiAgICB9CgoKICAgLyogVG8gbWFr ZSB0aGlzIHRhc2sgc3RheSBhbGl2ZSAqLwogICBmb3IoaT0wO2k8MTAwO2kr KykKICAgIHsKCXByaW50ZigiVGFzayBpcyBhbGl2ZVxuIik7CiAgICB9CgoK fSAvKiBFbmQgb2Ygc3RhcnRUaW1lciAqLwoKCi8qIEZ1bmN0aW9uOiB0bXJF eHBpcnlIYW5kbGVyKCkgdG8gaGFuZGxlIHRpbWVyRXhwaXJ5ICovCnZvaWQg dG1yRXhwaXJ5SGFuZGxlcihpbnQgc2lnbmFsKQp7CiAgcHJpbnRmKCJJbnNp ZGUgVGltZXIgRXhwaXJ5IEhhbmRsZXJcbiIpOwoKCn0gLyogRW5kIG9mIHRt ckV4cGlyeUhhbmRsZXIgKi8KCgoKCg== --0-1957747793-994193458=:28046-- From vxwexplo-errs@csg.lbl.gov Tue Jul 3 13:54:57 2001 From: Greg Milne Date: Tue Jul 3 13:55:00 PDT 2001 Subject: Tornado 2.0 BSP for MVME5100 boards. Hello gurus, I am attempting to port our system from the Motorola 2432 boards to the MVME5100. The system consists of master and slave VME boards ( both 5100s ) processing data fed to them from an NT server. The master and slave use shared memory and the proxy ARP network protocol. I am using the Early Access BSP supplied by Motorola. The system seems to run flawlessly for 7-12 hours at which point it will freeze. To unfreeze, I simply have to ping the slave or open a telnet session. Any ideas, suggestions ? Have I provided enough info ? Thanks in advance, Gregor Milne From vxwexplo-errs@csg.lbl.gov Tue Jul 3 21:32:17 2001 From: msudhir Date: Tue Jul 3 21:32:20 PDT 2001 Subject: subscribtion hello, can you please subscribe my mailid for VxWorks users Group? My mail Id is sudhir@cmcltd.com. Thanking you, M.Sudhir Engineer,R&D CMC Ltd. From vxwexplo-errs@csg.lbl.gov Wed Jul 4 00:10:27 2001 From: "Kim.Dokhac" Date: Wed Jul 4 00:10:29 PDT 2001 Subject: unsubscribtion hello, can you please unsubscribe my mailid for VxWorks users Group? My mail Id is kim.dokhac@meteo.fr Thank you, ================================================================================ Kim DO KHAC Tel : 01 30 13 61 84 METEO-FRANCE Fax : 01 30 13 60 62 DSO/CMR Email : kim.dokhac@meteo.fr 7 , Rue Teisserenc-de-Bort BP 202 78195 Trappes ================================================================================ From vxwexplo-errs@csg.lbl.gov Wed Jul 4 00:39:09 2001 From: fscheurer@manz-acs.de Date: Wed Jul 4 00:39:11 PDT 2001 Subject: Problems with Cygwin GCC WinNT cross i386-vxWorks Hi all, (if you answer please use my E-Mail address as well. Not only the mailing-list! TIA, F. Scheurer) I try to build the tool-chain mentioned above (explained somewhat by a HOWTO of Mumit Khan): #host=i686-cygwin-windowsnt host=i686-pc-cygwin ## FS 2001-07-03 build=$host #target=m68k-coff ## FS 2001-06-29 target=i386-wrs-vxworks ## FS 2001-07-03 #tmpdir=/tmp I passed configuration well using the following script (based on some contributions of CrossGCC and Mumit Khan) ... #prefix=/bar ## FS 2001-06-29 #prefix=../bar ## FS 2001-07-03 prefix=/usr/local/i386 i=$prefix/bin #mkdir build-binutils build-gcc build-newlib build-gdb # Configure, build and install binutils cd build-binutils #../binutils-2.9.1/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-02 echo "Trying to build binutils ..." #sh ../binutils-2.11.2/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 (see Khan-HOWTO) sh ../binutils-2.11.2/configure \ --with-included-gettext \ --target=$target --host=$host --build=$build \ --prefix=$prefix -v make all install # Configure, build and install gcc ONLY # ************************************************************************* # ATTENTION: # Huh, why the .exe extension? Well, it has to do with gcc hosted on # Cygwin, and you can always patch gcc sources to get rid of the automatic # .exe addition. Just comment out the EXECUTABLE_SUFFIX macro in # $/gcc-2.XX.Y-Z/gcc/config/i386/xm-cygwin.h and rebuild gcc. # ************************************************************************* cd ../build-gcc #../gcc-2.8.1/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-03 (see Khan-HOWTO Cygwin-Sources fixed for Cygwin platform) echo "Trying to build gcc ..." #sh ../gcc-2.95.3-5/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 (see Khan-HOWTO) sh ../gcc-2.95.3-5/configure \ --enable-languages=c,c++ \ --with-included-gettext --enable-shared --enable-threads \ --target=$target --host=$host --build=$build \ --with-newlib \ --prefix=$prefix -v #make all install make LANGUAGES=c all-gcc > make.log 2>&1 mv make.log make-c-only.log make LANGUAGES=c install-gcc > install.log 2>&1 mv install.log install-c-only.log # Configure, build and install newlib cd ../build-newlib #../newlib-1.8.1/configure --target=$target --prefix=$prefix -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. #make all install \ # CC_FOR_TARGET=$i/${target}-gcc \ # AS_FOR_TARGET=$i/${target}-as \ # LD_FOR_TARGET=$i/${target}-ld \ # AR_FOR_TARGET=$i/${target}-ar \ # RANLIB_FOR_TARGET=$i/${target}-ranlib ## FS 2001-07-02 (see Khan-HOWTO) echo "Trying to build newlib ..." #sh ../newlib-1.9.0/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. #make all \ # CC_FOR_TARGET=$i/${target}-gcc \ # AS_FOR_TARGET=$i/${target}-as \ # LD_FOR_TARGET=$i/${target}-ld \ # AR_FOR_TARGET=$i/${target}-ar \ # RANLIB_FOR_TARGET=$i/${target}-ranlib ## FS 2001-07-03 (see Khan-HOWTO) sh ../newlib-1.9.0/configure --target=$target --host=$host --build=$build \ --prefix=$prefix -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. make all install \ CC_FOR_TARGET=$i/${target}-gcc \ AS_FOR_TARGET=$i/${target}-as \ LD_FOR_TARGET=$i/${target}-ld \ AR_FOR_TARGET=$i/${target}-ar \ RANLIB_FOR_TARGET=$i/${target}-ranlib # Configure, build and install gcc "the rest" ## FS 2001-07-03 (see Khan-HOWTO) cd ../build-gcc make all install # Configure, build and install gdb cd ../build-gdb #../gdb-4.17/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-02 echo "Trying to build gdb ..." #sh ../gdb-4.18/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 sh ../gdb-4.18/configure --target=$target --host=$host --prefix=$prefix -v make all install All builds well - binutils - gcc-only - newlib until it comes to - gcc-all then problems occur with non-matching GNU/VxWorks or missing Header-Files 1. vxXXXXX.h are missing for name in _eh; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_new lib1.9.0/build-gcc/gcc/ -B/usr/local/i386/i386-wrs-vxworks/bin/ -I/usr/local/i386/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN _GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc -fexceptions -I. -I../.. /gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include -c \ -DL${name} ../../gcc-2.95.3-5/gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _eh In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:170, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxArch.h:88: arch/i86/archI86.h: No such file or directory In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:175, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxTypesOld.h:75: sys/types.h: No such file or directory make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 2. including them leads to conflicts with stddef.h for name in _eh; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_new lib1.9.0/build-gcc/gcc/ -B/usr/local/i386/i386-wrs-vxworks/bin/ -I/usr/local/i386/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN _GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc -fexceptions -I. -I../.. /gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include -c \ -DL${name} ../../gcc-2.95.3-5/gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _eh In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:174, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: redefinition of `ptrdiff_t' include/stddef.h:120: `ptrdiff_t' previously declared here ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: conflicting types for `size_t' include/stddef.h:172: previous declaration of `size_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: conflicting types for `wchar_t' include/stddef.h:257: previous declaration of `wchar_t' make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 3. Substituting stddef.h with the subset which comes with VxWorks leads us to: c:/usr/bin/gcc -c -DCROSS_COMPILE -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/conf ig -I../../gcc-2.95.3-5/gcc/../include ../../gcc-2.95.3-5/gcc/cp/g++spec.c In file included from ../../gcc-2.95.3-5/gcc/config/types/vxANSI.h:44, from ../../gcc-2.95.3-5/gcc/config/types/vxTypes.h:63, from ../../gcc-2.95.3-5/gcc/config/sys/types.h:39, from /usr/include/time.h:32, from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/cp/g++spec.c:22: ./../gcc-2.95.3-5/gcc/config/types/vxCpu.h:200: #error CPU is not defined correctly make[1]: *** [g++spec.o] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 4. Integrating CPU Macro/Define into highest level Makefile (again with original Cygwin-stddef.h in gcc sources) CPU = I80386 .. CC_FOR_BUILD = $(CC) -DCPU=$(CPU) .. leads us to conflicts in the build-tools: c:/usr/bin/gcc -DCPU=I80386 -c -DCROSS_COMPILE -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.3-5/gcc -I../../gcc-2.95 .3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include \ ../../gcc-2.95.3-5/gcc/gencheck.c In file included from ../../gcc-2.95.3-5/gcc/config/sys/types.h:39, from /usr/include/time.h:32, from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: warning: redefinition of `size_t' c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:172: warning: `size_t' previously declared here In file included from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/time.h:47: warning: `__cdecl__' attribute directive ignored /usr/include/time.h:47: parse error before `clock' In file included from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/stdio.h:50: warning: redefinition of `fpos_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:91: warning: `fpos_t' previously declared here In file included from /usr/include/stdlib.h:17, from ../../gcc-2.95.3-5/gcc/system.h:150, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:257: conflicting types for `wchar_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: previous declaration of `wchar_t' In file included from /usr/include/sys/unistd.h:12, from /usr/include/unistd.h:4, from ../../gcc-2.95.3-5/gcc/system.h:154, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:120: conflicting types for `ptrdiff_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: previous declaration of `ptrdiff_t' In file included from /usr/include/sys/time.h:27, from ../../gcc-2.95.3-5/gcc/system.h:194, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/sys/select.h:28: parse error before `fd_set' make[1]: *** [gencheck.o] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 What's the right way throug all this mess? Is the actual portability to vxworks not given? Where to find some more information about that X-ing i686-cygwin-X-x86-vxworks? Are all the adaptions to be made in the headers drawn from {WIND_BASE}/target/h? Any hint appreciated. Many thanks in advance! MfG / Kind Regards Friedrich Scheurer, R&D -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de -------------------------------- To: Steven_Snyder@3com.com aoliva@redhat.com Cc: gcc-help@gcc.gnu.org vxwexplo@lbl.gov From vxwexplo-errs@csg.lbl.gov Wed Jul 4 01:06:20 2001 From: "unnikrishnan purushothaman kartha" Date: Wed Jul 4 01:06:22 PDT 2001 Subject: Re: VxWorks Scheduling The behaviour of the scheduler would be like this: taskLock is no more valid when an operation involving scheduler is made (semGive in this case). Hence TaskB will be unblocked immediately and TaskA takes semaphore again. When func_3 returns, semGive is executed once more and TaskA returns. The current version of vxWorks is 5.4. I am not sure about which memory related problems you are referring to. Fragmentation problems remain. WindRiver has recently released a different OS called VxWorks AE which implements memory protected architecture. regards Unni. From vxwexplo-errs@csg.lbl.gov Wed Jul 4 04:03:33 2001 From: Vxworks Exploder Date: Wed Jul 4 04:03:35 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 4 04:03:07 PDT 2001 Subject: Where is this telnetd from? Subject: Re: Timers in vxworks Subject: sample code for PLX-PCI-9056 driver Subject: Re: about "taskSpawn" Subject: Inflate again !!! Guru code needed! Subject: RE: Vx Exploder Digest Subject: Re: about "taskSpawn" Subject: Tornado simulator VxWorks Subject: Re: Tornado simulator VxWorks Subject: align malloc and free in VxWorks Subject: Re: align malloc and free in VxWorks Subject: VxWorks FTP server Subject: Re: Tornado simulator VxWorks Subject: two IP stacks Subject: signal handling in vxworks. Subject: Re : Assembler with VxWorks Subject: Re: How can i print the stack call after an error ? Subject: Re: vxworks configuration Subject: Re: VxWorks FTP server Subject: Re: two IP stacks Subject: problem with timers Subject: Tornado 2.0 BSP for MVME5100 boards. Subject: Re: Defragmentation for TrueFFS Subject: standalone_vxworks with no network Subject: Re: about "taskSpawn" Subject: Re: How can i print the stack call after an error ? Subject: Re: standalone_vxworks with no network Subject: use cacheDmaFree to free aligned memory Subject: seperate interrupt stack Subject: abort button Subject: interface between asm and C language Subject: XModem protocol Subject: which network device does wdb agent use Subject: Re: Tornado 2.0 BSP for MVME5100 boards. Subject: How can I find the false with the image?? Subject: Re: standalone_vxworks with no network Subject: subscribtion Subject: Bringing up the ARM Integrator/CM7TDMI with vxWorks Subject: IP address of VxWorks Subject: two ethernet controllers in VxWorks Subject: nanosleep Subject: Re: two IP stacks Subject: unsubscribtion Subject: Re: two ethernet controllers in VxWorks Subject: How can I know the socket peer closed? Subject: FTP SERVER Subject: Re: signal handling in vxworks. Subject: Problems with Cygwin GCC WinNT cross i386-vxWorks Subject: re: VxWorks Scheduling Subject: Re: problem with inet_aton Subject: How to make my objects to library? Subject: Re: about "taskSpawn" Subject: Re: Memory requirments.... Subject: Re: problem with timers Subject: Re: interface between asm and C language Subject: Re: I want to know "shell". Subject: Re: How to make my objects to library? Subject: Re: FTP SERVER Subject: Re: which network device does wdb agent use Subject: Re: nanosleep Subject: Re: FTP SERVER Subject: why the "unit number" disappeared? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Where is this telnetd from? Date: 03 Jul 2001 12:13:23 UT From: fscheurer@manz-acs.de Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsd0d$4ge$1@overload.lbl.gov> Hi VxWorkers! some time ago I downloaded the attached source code. I also saw a HTML description and a Makefile but didn't store it right, so it's gone. Anyone knows where I found it? Probably the people of MBARI? Any hint appreciated! TIA MfG / Kind Regards Friedrich Scheurer, R&D - -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de - -------------------------------- /****************************************************************************/ /* Copyright 1992 to 1998 MBARI */ /****************************************************************************/ /* Summary : telnetd skeleton */ /* Filename : telnetd.c */ /* Author : Andrew Pearce */ /* Project : */ /* Version : Version 1.0 */ /* Created : 07/17/98 */ /* Modified : 07/17/98 */ /* Archived : */ /****************************************************************************/ /* Modification History: */ /* $Header: $ * $Log: $ */ /****************************************************************************/ #include /* vxWorks system declarations */ #include /* vxWorks task library functions */ #include /* vxWorks standard I/O functions */ #include /* vxWorks I/O library functions */ #include /* vxWorks string functions */ #include /* vxWorks Linked List library */ #include #include /* vxWorks internet protocol definitions*/ #include /* vxWorks network socket functions */ #include /* vxWorks select I/O functions */ #include /* mbari style guide type declarations */ #include /* Miscellaneous constants */ #define MAX_QUEUED_CONNECTIONS 4 /* Number of enqueued connect requests */ #define NET_TASK_PRIORITY 80 /* Telnet Task Priority */ #define NET_TASK_STACKSIZE 4000 /* Telnet Task Stacksize & TCB */ #define TELNET_PORT_NO 23 /* Default telnet port number */ /* Note this port will conflict with */ /* the vxWorks telnetd. Use a different */ /* port number or rebuild vxWorks with */ /* INCLUDE_TELNET undefined */ MLocal Void myTnTask( Int32 socketFd ) { Int32 on; on = TRUE; /* Turn on non-blocking for read */ ioctl(socketFd, FIONBIO, (int) &on); FOREVER { if (write(socketFd, "Message from telnet task tnTask\n", 32) == ERROR) { close(socketFd); return; } /* if */ taskDelay(sysClkRateGet()); } /* FOREVER */ } /* myTnTask() */ STATUS telnetDeamon( Void ) { struct sockaddr_in myAddr; struct sockaddr_in clientAddr; int clientAddrLen; Int32 socketFd; Int32 newFd; Int32 port = TELNET_PORT_NO; /* Default telnet port number */ /* create a socket for each port */ socketFd = socket(AF_INET, SOCK_STREAM, 0); bzero ((char *) &myAddr, sizeof (myAddr)); myAddr.sin_family = AF_INET; myAddr.sin_port = htons (port); if (bind (socketFd, (struct sockaddr *) &myAddr, sizeof (myAddr)) == ERROR) { printf ("telnet bind failed. 0x%x\n", errnoGet()); return(ERROR); } /* if */ /* Create client connect request queue */ if (listen (socketFd, MAX_QUEUED_CONNECTIONS) == ERROR) { printf ("telnet listen failed.\n"); return(ERROR); } /* if */ FOREVER { /* Wait for connection requests on all */ /* sockets using select */ errnoSet (OK); /* set error number to OK */ /* accept connection and spawn task to */ /* handle it */ clientAddrLen = sizeof (clientAddr); if ((newFd = accept (socketFd, &clientAddr, &clientAddrLen)) == ERROR) { printf ("telnet accept failed status = 0x%x\n", errnoGet ()); close (socketFd); } /* if */ if (taskSpawn("myTnTask", NET_TASK_PRIORITY, VX_FP_TASK, NET_TASK_STACKSIZE, (FUNCPTR) myTnTask, newFd, 0, 0, 0, 0, 0, 0, 0, 0, 0) == ERROR) { printf("taskSpawn failed\n"); close (newFd); } /* if */ } /* FOREVER */ } /* telnetDeamon() */ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Timers in vxworks Date: Tue, 3 Jul 2001 14:51:10 +0200 From: "Michel Hendriks" Organization: XS4ALL Internet BV Message-ID: <9hsf4k$afo$1@news1.xs4all.nl> References: <9hqa6u$ge6$1@overload.lbl.gov> Sender: tcj@gateway.pijnenburg.nl Try this, it should work without modifications... /*-------------------------------------------------------------------------- - - [TITLE] Demo [FILE] [VSN] [CREATED] [LASTCHNGD] [COPYRIGHT] Copyright 2001 (C) CPS Europe B.V. [AUTHOR] Michel Hendriks [PURPOSE] [NOTE] --------------------------------------------------------------------------- - - [VERSION HISTORY] [DATE] [WHO] [COMMENT] --------------------------------------------------------------------------- - -*/ /* Include files ------------------------------------------------------------*/ #include /* our OS */ #include /* task control */ #include /* signals */ #include /* printf */ #include /* sysClkRateGet */ #include /* timers */ #include /* logMsg */ #include /* watchdog routines */ /* Constants ----------------------------------------------------------------*/ /* Type Definitions ---------------------------------------------------------*/ /* External data ------------------------------------------------------------*/ /* Public data --------------------------------------------------------------*/ /* Private vars -------------------------------------------------------------*/ static BOOL wait; static WDOG_ID wdogId; /* Compile time checks ------------------------------------------------------*/ /* Prototyping for private routines -----------------------------------------*/ /* Implementation -----------------------------------------------------------*/ /*-------------------------------------------------------------------------- - --- Description : WatchDog ISR handler Parameters : Globals Changed: Returns : - ---------------------------------------------------------------------------- - -*/ static void watchDogISR(void) { logMsg( "timedemo: WatchDogISR called\n", 1, 2, 3, 4, 5, 6 ); } /*-------------------------------------------------------------------------- - --- Description : Signal handler for SIGALRM (Timer Handler) Parameters : Globals Changed: Returns : - ---------------------------------------------------------------------------- - -*/ static void timerHandler ( timer_t timerid, /* expired timer ID */ int arg /* user argument */ ) { logMsg( "timedemo: Timer 0x%x expired (arg = 0x%x)\n", (int)timerid, arg, 3, 4, 5, 6 ); wait = FALSE; } /*-------------------------------------------------------------------------- - --- Description : Main entry for this demo Parameters : Globals Changed: Returns : OK - ---------------------------------------------------------------------------- - -*/ STATUS timedemo( void ) { timer_t timerId ; /* id for the posix timer */ struct itimerspec timeToSet ; /* time to be set */ STATUS retval = OK; wait = TRUE; /* * Start POSIX timer demo */ /* Set the time to the desired value */ timeToSet.it_value.tv_sec = 5 ; /* initial (one shot) value */ timeToSet.it_value.tv_nsec = 0 ; timeToSet.it_interval.tv_sec = 0 ; /* reload (repetitive) value */ timeToSet.it_interval.tv_nsec = 0 ; if( timer_create( CLOCK_REALTIME, NULL, &timerId ) == ERROR ) { perror( "timedemo: Error in creating the timer\n" ); retval = ERROR; } if( timer_connect( timerId, &timerHandler, 2) == ERROR ) { perror( "timedemo: Error in connecting the timer\n" ); retval = ERROR; } else if( timer_settime( timerId, CLOCK_REALTIME, &timeToSet, NULL ) == ERROR ) { perror( "timedemo: Error in setting the timer\n" ); retval = ERROR; } else { if ( taskDelay(sysClkRateGet() * 20) == ERROR ) { perror( "timedemo: taskDelay expired early\n" ); retval = ERROR; } { struct timespec time_delay; struct timespec time_remaining; time_delay.tv_sec = 2; time_delay.tv_nsec = 500000000L; if ( nanosleep( &time_delay, &time_remaining ) == ERROR ) { perror( "timedemo: nanosleep expired early\n" ); retval = ERROR; } else { logMsg( "timedemo: nanosleep expired normally\n", 1, 2, 3, 4, 5, 6 ); } } /* while ( wait ) { } */ } /* Cleanup */ if (timer_delete (timerId) == ERROR) { perror( "timedemo: Error in deleting the timer\n" ); retval = ERROR; } /* * Start watchdog timer demo */ wdogId = wdCreate(); if ( wdogId == NULL ) { perror( "timedemo: couldn't create watchdog\n" ); retval = ERROR; } else if ( wdStart( wdogId, sysClkRateGet() * 5, (FUNCPTR)watchDogISR, 0 ) == ERROR ) { perror( "timedemo: couldn't start watchdog\n" ); retval = ERROR; } else { if ( taskDelay(sysClkRateGet() * 20) == ERROR ) { perror( "timedemo: taskDelay expired early\n" ); retval = ERROR; } } /* Cleanup */ if ( wdDelete( wdogId ) == ERROR ) { perror( "timedemo: couldn't delete watchdog\n" ); retval = ERROR; } return( retval ); } Anuradha Pani wrote in message news:9hqa6u$ge6$1@overload.lbl.gov... > hi, > > i am new to vxworks and am trying to implement basic > timer routines. i'm not able to find any example code > which tells me how to handle any signal upon timer > expiry. > > i'm able to create timers but am not able to proceed > further. > > could anyone give me any pointers? > > thanks in advance > anuradha > > __________________________________________________ > Do You Yahoo!? > Get personalized email addresses from Yahoo! Mail > http://personal.mail.yahoo.com/ --------------------------- Newsgroups: comp.os.vxworks Subject: sample code for PLX-PCI-9056 driver Date: Tue, 3 Jul 2001 14:30:13 +0200 From: "Joerg Engel" Organization: Customer of UUNET Deutschland GmbH Message-ID: <3b41c83d$0$227$4dbef881@businessnews.de.uu.net> Where can we get some source code samples for developing a driver using the PLX-PCI-9056 Controller. I don't want to develope a BSP but I need some sample code from where I can start . --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Tue, 3 Jul 2001 11:53:33 +0100 From: vxworksintent@tao-group.com (David Given) Organization: I do not speak for anyone but myself, and barely that. Message-ID: References: <9hrqlg$mr0$1@news1.kornet.net> In article <9hrqlg$mr0$1@news1.kornet.net>, benjamin writes: > When creating a task with the function > "taskSpawn", we must have information > about "stack size". > > As vxworks's beginner, I like to know why the stack size is needed and what it is used for. > > I hope anyone who know the answers gives me those. The stack is used by C to store things like the function call sequence and automatic (local) variables. Usually 8kB is a good size, but the more complicated the task is, the more it will need; deeply nested recursive functions will use lots of stack, etc. VxWorks does not (as far as I am aware) have the ability to dynamically grow stacks. - -- David Given dg@tao-group.com --------------------------- Newsgroups: comp.os.vxworks Subject: Inflate again !!! Guru code needed! Date: Tue, 3 Jul 2001 16:04:17 +0200 From: "T.J.Scheffel" Organization: Internet Access Eindhoven, the Netherlands Message-ID: <9hsjd3$429$1@news.IAEhv.nl> Hi Guru, Do YOU have any experience with re-inflating images on NT which were deflated?? I'm searching for some example code to do this. (Probably ZLIB based) If you have it, please help me out! I do have an inflate on the target, but i need it on NT in order to write some 'string changing' wrapper for different areas on this globe.. Thanx and keep up the good work! TJ --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Vx Exploder Digest Date: Tue, 3 Jul 2001 09:28:03 -0400 From: "Gary LeVan" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsk1d$7ku$1@overload.lbl.gov> - --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Tue, 03 Jul 2001 11:20:54 +0100 From: Simon Farnsworth Message-ID: <3B419C86.2010104@snellwilcox.com> References: <9hrqlg$mr0$1@news1.kornet.net> benjamin wrote: > When creating a task with the function > "taskSpawn", we must have information > about "stack size". > > As vxworks's beginner, I like to know why the stack size is needed and what it is used for. The stack is used by the system for storing local variables and return addresses; to avoid the non-deterministic behaviour introduced by resizing a stack, VxWorks requires you to set a limit at startup. To start, Make the Stack size very large (how about 10000?) which of course depends on your application. Exercise your application fully and use checkStack( )to check amount used and adjust as necessary. Suggest leaving stack size at least double what you think you need. Exceeding maxStack size will walk on someone else and make you unhappy. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado simulator VxWorks Date: Tue, 03 Jul 2001 13:26:46 -0000 From: "Manjunath Bhat" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsk1e$7kv$1@overload.lbl.gov> Hi, Do anybody of you know how to increase the target memory size for the Tornado II simulator? Thanking you, Manjunath Bhat. _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado simulator VxWorks Date: Tue, 03 Jul 2001 16:31:14 +0200 From: Stunic Domagoj Organization: Siemens AG Austria Message-ID: <3B41D732.2CFEB8B8@siemens.hr> References: <9hsk1e$7kv$1@overload.lbl.gov> Hi, add -m xxxxxxx, where xxxxxxx is amout of bytes that you need, in command line at the startup of the simulator. Manjunath Bhat wrote: > Hi, > Do anybody of you know how to increase the target memory size for the > Tornado II simulator? > > Thanking you, > Manjunath Bhat. > _________________________________________________________________________ > Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. --------------------------- Newsgroups: comp.os.vxworks Subject: align malloc and free in VxWorks Date: Tue, 3 Jul 2001 22:38:31 +0800 From: "lhp" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsnht$8rk$1@overload.lbl.gov> Hi, VxWorks 5.4 for PPC. I use following routine to malloc a block of cache-coherent memory which aligns parameter "Align". Align is a value which is power of 2. void* commonCacheDmaMallocAlign(UINT32 Align,UINT32 Size) { char* memory1, memory2; memory1 = (char*)cacheDmaMalloc(Size + Align); /* check if allocation successes */ if(memory1 == NULL) { printf("can"t allocate memory\n"); return NULL; } memory2 = ((char*)(((UINT32)(memory1 + Align)) & (~(Align-1)))); return (void*)memory2; } My Question: Is it safe and correct when I free the memory malloced by commCacheDmaMallocAlign() like this? void *p; p = commonCacheDmaMallocAlign(16, 1000); ... cacheDmaFree(p); Will the memory between "memory1" and "memory2" be lost? I don't know how the VxWorks kernel deal with it internally. Who can tell me? Certaily, we can save the value of original "memory1" first and use cacheDmaFree(memory1) to free the whole memory. I know this is absolutely safe and correct. MUST we do like this? I want an exact answer. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: align malloc and free in VxWorks Date: Tue, 03 Jul 2001 18:02:15 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B41EC87.A1B96983@ks.ericsson.se> References: <9hsnht$8rk$1@overload.lbl.gov> Hi Why do you try to reinvent the wheel? Have a look at memalign! Btw, you can only free a address you have received my a memory allocation, so your example might crash or maybe even reboot. You don't say what you need it for, but if its for DMA, use cacheDmaMalloc. If you really need to do what the example is intended to, you need to save the values returned from cacheDmaMalloc and if more than one task can call your functions you need to add some mutex and maybe a check that Align is a power of two. Urban lhp wrote: [Snip] > Is it safe and correct when I free the memory malloced by commCacheDmaMallocAlign() like this? > > void *p; > p = commonCacheDmaMallocAlign(16, 1000); > ... > cacheDmaFree(p); > > Will the memory between "memory1" and "memory2" be lost? I don't know how the VxWorks kernel > deal with it internally. Who can tell me? > > Certaily, we can save the value of original "memory1" first and use cacheDmaFree(memory1) to free the whole memory. I know this is absolutely safe and correct. MUST we do like this? I want an exact answer. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks FTP server Date: Tue, 03 Jul 2001 16:23:57 GMT From: "Christophe Belmont" Organization: Verio Message-ID: Hi, I'm running VxWorks FTP server (VxWorks 5.4). If I connect to the server using the command line, the list command works fine. If I try to use a graphic FTP client, it seems that the answer to the list command is not properly formatted by VxWorks and nothing appears in the Window. I was wondering if somebody else had this kind of problem? Regards, Christophe. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado simulator VxWorks Date: Tue, 3 Jul 2001 18:43:32 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9hssni$ld5$02$1@news.t-online.com> References: <9hsk1e$7kv$1@overload.lbl.gov> <3B41D732.2CFEB8B8@siemens.hr> Hello, the ramsize parameter is used as "-r " on windows. Ciao Martin "Stunic Domagoj" schrieb im Newsbeitrag news:3B41D732.2CFEB8B8@siemens.hr... > Hi, > > add -m xxxxxxx, where xxxxxxx is amout of bytes that you need, in command > line at the startup of the simulator. > > Manjunath Bhat wrote: > > > Hi, > > Do anybody of you know how to increase the target memory size for the > > Tornado II simulator? > > > > Thanking you, > > Manjunath Bhat. > > _________________________________________________________________________ > > Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. > --------------------------- Newsgroups: comp.os.vxworks Subject: two IP stacks Date: 3 Jul 2001 10:03:35 -0700 From: du_bin@yahoo.com (billy du) Organization: http://groups.google.com/ Message-ID: <340587ac.0107030903.5f999b75@posting.google.com> I do have quite and unusual request (but, yes I really do want two interfaces with the same IP address, which is possible in at least Linux and windows)? Emmanuel Herbreteau wrote in message news:<3B3C24E7.3783DE68@sepro-robotique.com>... > billy du wrote: > > Can VxWorks support two IP stacks inside one CPU? > > What ??? > > > Unfortuately, the outside world might use the private IP addresses. If > > we uses private IP addresses for internal connection, there might be > > conflict. Therefore, we plan to use two IP stacks in one card. > > You should use private IP address for your private network : Look at > RFC 1918 on the Internet (or a good book, like "TCP/IP" from MacMillan). > > ie : Private address (class C) "192.168.16.44" or private address (class > B) > "172.16.1.27" will never conflict with an Internet address. Sorry, my question is not clear. We want to support private IP addresses in the outside world as well because our equipment might connect to a private network using IP address like 10.x.x.x 192.168.x.x. Then we have a problem to assign IP address inside the equipment. That's why we need two IP stack or any other approach. --------------------------- Newsgroups: comp.os.vxworks Subject: signal handling in vxworks. Date: Tue, 3 Jul 2001 09:23:46 -0700 (PDT) From: Anuradha Pani Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsuiv$bdd$1@overload.lbl.gov> hi all, i am having some problem with catching signals. I have attached the following piece of code. I am calling Start to spawn the signal handler task, signalHandler and then sending a signal to that task, but nothing happens. I don't get any output from the sigHandler function. Please tell me what I'm doing wrong. Thanks Anuradha /* File: signalHandler.c */ /* Include files */ #include "stdio.h" #include "vxWorks.h" #include "sigLib.h" #include "taskLib.h" /* Function prototypes */ void signalProcess (int signalNum); void signalHandler(void) { struct sigaction act; int i; printf("Signal Handler started\n"); /* Install a signal handler */ act.sa_handler = &signalProcess; /* Function to be called */ sigfillset(&act.sa_mask); /* Block all other signals while signalProcess is executing */ act.sa_flags = 0; if ( sigaction(SIGUSR1, &act, NULL) == -1 ) { printf("Signal processor could not be established\n"); exit(1); } /* Remain in a loop */ for (i=0;i<1000;i++) { taskDelay(30); printf("Signal handler running\n"); } } /* End of signalHandler */ /* Signal Processor Function */ void signalProcess (int signalNum) { printf("SIGNAL CAUGHT\n"); } /* End of signal processor */ void Start(void) { int taskId; /* Spawn the signal handler process */ if ((taskId = taskSpawn("signal2", 100, 0, 2000, (FUNCPTR)signalHandler,0,0,0,0,0,0,0,0,0,0)) == -1) { printf("Start FAILED to launch signal handler\n"); } printf("Signal Handler Spawned\n"); taskDelay(30); kill(taskId, SIGUSR1); } /* End of start */ __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re : Assembler with VxWorks Date: Tue, 3 Jul 2001 18:59:55 +0200 From: francois.gorisse@modicon.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9hsuj3$bde$1@overload.lbl.gov> Hi Harman, May be the code below will help you. I use gcc for 486 under vxworks on a Windows NT PC platform with Tornado 2 T2P4. Note : Some compiler option are 02 -xc++. void outWrds(void) { unsigned short port = 0x378; unsigned char value = 0xff; __asm__ volatile ("outb %1,%0;" : : "d" (port) , "a" (value) ); } and you 'll get under the shell : - -> l outWrds outWrds(void) 00030d60 55 PUSH EBP 00030d61 89 e5 MOV EBP, ESP 00030d63 ba 78 03 00 00 MOV EDX, 0x378 00030d68 b0 ff MOV AL, 0xff 00030d6a ee OUT DX, AL 00030d6b 89 ec MOV ESP, EBP 00030d6d 5d POP EBP 00030d6e c3 RET I hope it could serve you, Francois. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can i print the stack call after an error ? Date: 3 Jul 2001 10:29:54 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107030929.2500a632@posting.google.com> References: <42AB6BE23C29D511831E0002B32024880FB0B7@MESSENGER> Hello, AsafH@Radlan.co.il (Asaf Harari) wrote in message news:<42AB6BE23C29D511831E0002B32024880FB0B7@MESSENGER>... > Hi , i need a code for printing the stack call after an error . > For example when i write "os fatal error" . > How can i use the pc/sp for finding the right lines of code ? > Thanks. As long as you still have the connection to the target for the Tornado environment, then the 'windsh' (shell) command tt should be able to help you. By default, it should use the task ID of the last task to hit an exception; if that doesn't work, try passing it the task ID of the task you are interested in. Also, the debugger should be able to connect to the task and let you examine the source code (assuming you compiled with -g). You might have to step up the stack a few times if the crash actually occurred in a VxWorks routine before you see C source code. If the connection to the host is failing, but the target is not rebooting, then you could try adding the target shell and target based debug tools. These make use of the serial port on the target to communicate, and are sometimes less affected by errant tasks than the WDB connection to the host. The target tools only support assembler level debugging, though they will be able to give you the function names (use tt to get the stack trace for the task that crashed, as for the host shell). HTH, John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks configuration Date: 3 Jul 2001 10:24:49 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107030924.6155f139@posting.google.com> References: <9h76g0$k14$1@overload.lbl.gov> Hello, > Ok, We've got an mv1604 with 16Mb RAM. > > I've looked in the User manual, and it gives no more information than > the config window; Unless I look in a specific section, and even then > the information is cursory at best. > > For example, let's take INCLUDE_CONSTANT_RDY_Q. Now that is not even > in the index; The description says: This option includes the constant > insert time reqdy queue management facility. I have NO idea what that > is, how am I supposed to know if I need it or not. The constant ready queue is a deterministic implementation of the priority queuing algorithm used for scheduling. It should be the default. The alternative is a simple list based queue which, while not being deterministic, has the advantage of using a little less RAM (the constant ready queue uses a fixed size bitmap to manage the queue). Also, the code for the list based queue will be present in the kernel anyway for pend queues in things like semaphores. So, if you are on a seriously memory constrained target, and absolute determinism is not important to you, removing the constant ready queue might help you out, otherwise you may as well leave it in there. > Considering the > number of INCLUDE's I'm loath to got for trial and error on each. > > Obviously some INCLUDES make sense to me but most don't. > > Say 'INCLUDE_FTP_SERVER'. And even this description is somewhat > terse. -> 'this option includes target-based ftp server > support'. Yes, but why would I want (or not want) that? If you have a local file system on your target (say a SCSI disk or a TFFS flash file system) and you'd like to access that file system from a remote machine (perhaps another VxWorks system, or perhaps just your desktop machine). If you have no local file system an ftp server makes no sense since it wil have nothing to serve ;-) > Something to the effect of 'If you want to ftp to the machine > running vxworks then include this option' may be better. At least > this one is in the index :-) > > Another example: INCLUDE_PING, again not in the index. Description: > This option includes ping() utility. Does that mean I can ping it or > it can ping others? It means you can ping other things; 'ping' is the name of the tool for probing other boards (same on Unix btw); the actual mechanism that is used I believe is an ICMP echo request, so it is ICMP support that you need to respond to a ping request (or any other ICMP request for that matter). HTH, John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks FTP server Date: Tue, 3 Jul 2001 21:11:01 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994187940.595627@newsmaster-04.atnet.at> References: If you look in the unsupported area of your Tornado installation, you should find \target\unsupported\src\netwrs\ftpdLib.c This is an improved version of the FTP Deamon for VxWorks. Using this version we are able to use a graphical FTP Client (namely Windows Commander, www.ghisler.com). hth werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: two IP stacks Date: Tue, 3 Jul 2001 21:18:09 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994188369.32158@newsmaster-04.atnet.at> References: <340587ac.0107030903.5f999b75@posting.google.com> Hi, wouldn't it be the easiest to take some private IP for your application as default and make it changeable if the need arises for one particular site? I think this would save you some routing troubles, if that is a concern. hth werner "billy du" wrote in message news:340587ac.0107030903.5f999b75@posting.google.com... > I do have quite and unusual request (but, yes I really do want > two interfaces with the same IP address, which is possible in at > least Linux and windows)? > > Emmanuel Herbreteau wrote in message news:<3B3C24E7.3783DE68@sepro-robotique.com>... > > billy du wrote: > > > Can VxWorks support two IP stacks inside one CPU? > > > > What ??? > > > > > Unfortuately, the outside world might use the private IP addresses. If > > > we uses private IP addresses for internal connection, there might be > > > conflict. Therefore, we plan to use two IP stacks in one card. > > > > You should use private IP address for your private network : Look at > > RFC 1918 on the Internet (or a good book, like "TCP/IP" from MacMillan). > > > > ie : Private address (class C) "192.168.16.44" or private address (class > > B) > > "172.16.1.27" will never conflict with an Internet address. > > Sorry, my question is not clear. We want to support private IP > addresses in the outside world as well because our equipment might > connect to a private network using IP address like 10.x.x.x > 192.168.x.x. Then we have a problem to assign IP address inside the > equipment. That's why we need two IP stack or any other approach. --------------------------- Newsgroups: comp.os.vxworks Subject: problem with timers Date: Tue, 3 Jul 2001 13:50:58 -0700 (PDT) From: Anuradha Pani Organization: Lawrence Berkeley National Laboratory Message-ID: <9htcku$jao$1@overload.lbl.gov> - --0-1957747793-994193458=:28046 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline hi all, i am new to vxworks. attached is a simple piece of code where i have tried to implement timers. but when i start the task, startTimer, nothing happens. i see no output at all. not even the "hi" i deliberately put in. can somebody help me? thanks anuradha __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ - --0-1957747793-994193458=:28046 Content-Type: application/x-unknown; name="timerImp.c" Content-Transfer-Encoding: base64 Content-Description: timerImp.c Content-Disposition: attachment; filename="timerImp.c" LyogRmlsZTogdGltZXJJbXAuYyAtIENyZWF0ZXMgYW5kIHNldHMgYSB0aW1l ciAqLwoKLyogSW5jbHVkZSBmaWxlcyAqLwojaW5jbHVkZSAic3RkaW8uaCIK I2luY2x1ZGUgInZ4V29ya3MuaCIKI2luY2x1ZGUgInRhc2tMaWIuaCIKI2lu Y2x1ZGUgInNpZ0xpYi5oIgojaW5jbHVkZSAidGltZS5oIgoKCgovKiBGdW5j dGlvbjogc3RhcnRUaW1lcigpCiAgIFRoaXMgZnVuY3Rpb24gc3Bhd25zIGEg dGFzayB0byBjYXRjaCB1c2VyIHNpZ25hbHMuIAogICBJdCBzdGFydHMgYSB0 aW1lciB3aXRoIGEgdGltZS1wZXJpb2Qgb2YgNTAgbXMgYW5kIHVwb24gZXhw aXJ5CiAgIHNlbmRzIHVzZXIgc2lnbmFscyB0byB0aGUgdGFzayBzcGF3bmVk LgogICBJdCB0aGVuIHJlX2FybXMgdGhlIHRpbWVyLgoqLwp2b2lkIHN0YXJ0 VGltZXIodm9pZCkKewogIHN0cnVjdCBpdGltZXJzcGVjIHRpbWVQZXJpb2Q7 CiAgdGltZXJfdCB0aW1lcklkOwogIGludCBpOwoKCiAgLyogRnVuY3Rpb24g cHJvdG90eXBlcyAqLwogIHZvaWQgdG1yRXhwaXJ5SGFuZGxlcihpbnQgc2ln bmFsKTsKICAKICAgcHJpbnRmKCJIaVxuIik7CgogICAvKiBDcmVhdGUgYSB0 aW1lciAqLwogICBpZiggdGltZXJfY3JlYXRlKENMT0NLX1JFQUxUSU1FLCBO VUxMLCAmdGltZXJJZCkgPT0gLTEpCiAgICB7CiAgICAgICAgcHJpbnRmKCJF cnJvciBpbiBjcmVhdGluZyB0aW1lclxuIik7CglleGl0KDEpOwogICAgfQog ICBlbHNlCiAgICB7CiAgICAgIHByaW50ZigiVGltZXIgd2l0aCBpZCA9ICVk ICAgIGNyZWF0ZWQgc3VjY2Vzc2Z1bGx5XG4iLCB0aW1lcklkKTsKICAgICAg LyogQXR0YWNoIGEgdGltZXIgcm91dGluZSB0byB0aGUgdGltZXIgKi8KICAg ICAgaWYoIHRpbWVyX2Nvbm5lY3QodGltZXJJZCwgKHZvaWQqKSZ0bXJFeHBp cnlIYW5kbGVyLCAwKSA9PSAtMSkKICAgICAgIHsKCSBwcmludGYoIkVycm9y IGluIGF0dGFjaGluZyB0aW1lciByb3V0aW5lXG4iKTsKCSBleGl0KDEpOwog ICAgICAgfQogICAgICBlbHNlCiAgICAgICB7CgkgcHJpbnRmKCJUaW1lciBy b3V0aW5lIGF0dGFjaGVkIHN1Y2Nlc3NmdWxseVxuIik7CgkgLyogU2V0IHRp bWVyIHBlcmlvZCAqLwoJICB0aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X3Nl YyA9IDA7ICAgLyogTm8gcmVwZWF0YXRpb25zIGZvciBub3cgKi8KICAJICB0 aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X25zZWMgPSAwOyAgCiAgCSAgdGlt ZVBlcmlvZC5pdF92YWx1ZS50dl9zZWMgPSAwOwogICAgICAgICAgdGltZVBl cmlvZC5pdF92YWx1ZS50dl9uc2VjID0gNTAwMDAwMDA7IC8qIGZpcnN0IGV4 cGlyeSBhZnRlciA1MCBtcyAqLwoKCSAgLyogRGVsYXkgZm9yIHNvbWV0aW1l ICovCgkgIHRhc2tEZWxheSgzMCk7CgoJICBpZiggdGltZXJfc2V0dGltZSh0 aW1lcklkLCAhVElNRVJfQUJTVElNRSwgJnRpbWVQZXJpb2QsIE5VTEwpID09 IC0xKQogICAgICAgICAgIHsKICAgICAgCSAgICAgICBwcmludGYoIkVycm9y IGluIHNldHRpbmcgdGltZXJcbiIpOwogICAgICAJICAgICAgIGV4aXQoMSk7 CiAgICAgICAgICAgfQogICAgICAgICAgIHByaW50ZigiVGltZXIgc2V0IHN1 Y2Nlc3NmdWxseVxuIik7CiAgICAgICB9CiAgICB9CgoKICAgLyogVG8gbWFr ZSB0aGlzIHRhc2sgc3RheSBhbGl2ZSAqLwogICBmb3IoaT0wO2k8MTAwO2kr KykKICAgIHsKCXByaW50ZigiVGFzayBpcyBhbGl2ZVxuIik7CiAgICB9CgoK fSAvKiBFbmQgb2Ygc3RhcnRUaW1lciAqLwoKCi8qIEZ1bmN0aW9uOiB0bXJF eHBpcnlIYW5kbGVyKCkgdG8gaGFuZGxlIHRpbWVyRXhwaXJ5ICovCnZvaWQg dG1yRXhwaXJ5SGFuZGxlcihpbnQgc2lnbmFsKQp7CiAgcHJpbnRmKCJJbnNp ZGUgVGltZXIgRXhwaXJ5IEhhbmRsZXJcbiIpOwoKCn0gLyogRW5kIG9mIHRt ckV4cGlyeUhhbmRsZXIgKi8KCgoKCg== - --0-1957747793-994193458=:28046-- --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 2.0 BSP for MVME5100 boards. Date: Tue, 3 Jul 2001 13:55:25 -0700 From: Greg Milne Organization: Lawrence Berkeley National Laboratory Message-ID: <9htcku$jap$1@overload.lbl.gov> Hello gurus, I am attempting to port our system from the Motorola 2432 boards to the MVME5100. The system consists of master and slave VME boards ( both 5100s ) processing data fed to them from an NT server. The master and slave use shared memory and the proxy ARP network protocol. I am using the Early Access BSP supplied by Motorola. The system seems to run flawlessly for 7-12 hours at which point it will freeze. To unfreeze, I simply have to ping the slave or open a telnet session. Any ideas, suggestions ? Have I provided enough info ? Thanks in advance, Gregor Milne --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Tue, 03 Jul 2001 23:15:26 GMT From: andreyk@home.com (Andray Kaganovsky) Organization: @Home Message-ID: References: <9hqng4$49e$1@nntp1-cm.news.eni.net> "News Account" wrote in news:9hqng4$49e$1@nntp1- cm.news.eni.net: > Does anybody know of any methodology or utility out there for defragmenting > a VxWorks DOS system that is stored and maintained via TrueFFS? Also I have > looked at the M-Systems TrueFFS code and noticed some sort of low level > defragmentation option that I believe is supposed to be accessable either > through the "tffsRawio" routine or some sort of "ioctl" call, however I was > unable to get this to work at all. At this point we would even be willing to > utilize a third party software to perform the defragmentation task. Please > help!!! > > Regards, > P. Eichler > [AndrayK]. To defragment TrueFFS flash array, you can use code fragment similar to the one below: int freeSectors = 0x70000000; /* impossible number of sectors */ (void) tffsRawio (0, TFFS_DEFRAGMENT_VOLUME, freeSectors, 0, 0); The call above could return error code, but, as far as I know, it is safe to ignore it. If media is badly fragmented, the call above can take very long time to complete, since it is doing complete defragmentation of the entire flash array. Andray --------------------------- Newsgroups: comp.os.vxworks Subject: standalone_vxworks with no network Date: Tue, 03 Jul 2001 23:21:27 GMT From: "Kelly Hornsby" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: hello I need some ideas of something that I can't figure out. I'm using T2 on a Sandpoint 750 system. I have a DEC 21143 NIC, and it works just fine with the bootloader/ ftp-Vxworks. However, I've created the standalone_vxworks.hex file, and flashed it into memory, but the network is not initialized now. the NIC is detected fine (pciHeaderShow indicates this) and the 21143 PCI initilization routine is executed (sysNetHwInit2), but there's no dev out there, and the device shows no devices for ifShow. I'm looking for suggestions or ideas for why this occurs. Any ideas? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Tue, 03 Jul 2001 06:36:43 -0700 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <3B41CA6B.3E584490@lmco.com> References: <9hrqlg$mr0$1@news1.kornet.net> The stack contains the parameters you send to the function, and any transient space it needs. You have to tell the system how much to allow because it has no ways to outguess you. There is a function somewhere in Tornado that will tell you how much stack a task is using. You could use that to tune the parameter later. I usually use 10K for starters, and if the task blows up, then I increase it. Speaking only for myself, Joe Durusau benjamin wrote: > > When creating a task with the function > "taskSpawn", we must have information > about "stack size". > > As vxworks's beginner, I like to know why the stack size is needed and what it is used for. > > I hope anyone who know the answers gives me those. > > -------------- ÀÎÅÍ³Ý Ä«¸®½º¸¶ KORNET ------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can i print the stack call after an error ? Date: Tue, 03 Jul 2001 06:42:47 -0700 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <3B41CBD7.6AA4AE4F@lmco.com> References: <42AB6BE23C29D511831E0002B32024880FB0B7@MESSENGER> Well, you could do something that creates a bus error or some such, and the sehll (assuming you are running run) will print registers. You might also consider looking at the stuff like taskRegsGet, taskRegsShow, or taskShow. Anothe approach is to use a code number for each of your code modules, and use the gcc macro __LINE__ in a logMsg or printf to make it tell you which line of source you are on. Note that the __LINE__ thing is not perfect, but it can help. Speaking only for myself, Joe Durusau Asaf Harari wrote: > > Hi , i need a code for printing the stack call after an error . > For example when i write "os fatal error" . > How can i use the pc/sp for finding the right lines of code ? > Thanks. > > -- > Posted from sphere.barak.net.il [212.150.48.98] > via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- Newsgroups: comp.os.vxworks Subject: Re: standalone_vxworks with no network Date: Wed, 4 Jul 2001 10:27:16 +0900 From: "¿ÁÀ»¼®" Organization: Korea Telecom Message-ID: <9htrfg$om7$1@news1.kornet.net> References: I want to know "error message". when booting, error message shows up. "Kelly Hornsby" wrote in message news:Xrs07.36578$w5.4378314@news1.rdc1.ga.home.com... > hello > > I need some ideas of something that I can't figure out. I'm using T2 on a > Sandpoint 750 system. I have a DEC 21143 NIC, and it works just fine with > the bootloader/ ftp-Vxworks. However, I've created the > standalone_vxworks.hex file, and flashed it into memory, but the network is > not initialized now. the NIC is detected fine (pciHeaderShow indicates > this) and the 21143 PCI initilization routine is executed (sysNetHwInit2), > but there's no dev out there, and the device shows no devices for ifShow. > > I'm looking for suggestions or ideas for why this occurs. Any ideas? > > --------------------------- Newsgroups: comp.os.vxworks Subject: use cacheDmaFree to free aligned memory Date: 3 Jul 2001 18:30:16 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031730.6d49192e@posting.google.com> I use following routine to malloc a block of cache-coherent memory which aligns parameter "Align". Align is a value which is power of 2. void* commonCacheDmaMallocAlign(UINT32 Align,UINT32 Size) { char* memory1, memory2; memory1 = (char*)cacheDmaMalloc(Size + Align); /* check if allocation successes */ if(memory1 == NULL) { printf("can"t allocate memory\n"); return NULL; } memory2 = ((char*)(((UINT32)(memory1 + Align)) & (~(Align-1)))); return (void*)memory2; } My Question: Is it safe and correct when I free the memory malloced by commCacheDmaMallocAlign() like this? void *p; p = commonCacheDmaMallocAlign(16, 1000); ... cacheDmaFree(p); Will the memory between "memory1" and "memory2" be lost? I don't know how the VxWorks kernel deal with it internally. Who can tell me? Certaily, we can save the value of original "memory1" first and use cacheDmaFree(memory1) to free the whole memory. I know this is absolutely safe and correct. MUST we do like this? I want an exact answer. lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: seperate interrupt stack Date: 3 Jul 2001 18:39:11 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031739.18a3d596@posting.google.com> We use VxWorks 5.4 for PPC and MPC860. (1)Does VxWorks use a seperate interrupt stack in this case? (2)If does, how to set the seperate stack size? (3)The benefit of seperate interrupt stack. I think the VxWorks kernel should also do task context switch(save and restore) in case of seperate interrupt stack. (4)The hardware(MPC860) support to seperate interrupt stack. Regards, lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: abort button Date: 3 Jul 2001 18:49:47 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031749.55de3fe5@posting.google.com> We use VxWorks 5.4 for PPC and MPC860. Some system may use "abort button" to do some special analytic work to find out why the system crashes (e.g. because of infinite loop). In my opinion, the abort button will trigger a NMI(non maskable interrupt), the analytic work is done in the ISR. Does anyone have experience on "abort button"? Can anyone post me any sample code of the NMI ISR triggered by "abort button"? I don't know how to analyze the cause of system crashes in the ISR. lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: interface between asm and C language Date: 3 Jul 2001 19:00:05 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031800.18b3244e@posting.google.com> Does any one know the interface between asm and C language when I use VxWorks and gcc? For example, when I invoke(jump to) a C routin in an asm file, how to pass parameters to the C routin? How the stack changes? lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: XModem protocol Date: 3 Jul 2001 19:03:36 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031803.2d2c73e6@posting.google.com> XModem protocol is used to transfer file. Does anyone can tell me where to find/download the XModem protocol specification? Does anyone can provide source code of the implement to XModem protocol? lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: which network device does wdb agent use Date: 3 Jul 2001 19:11:37 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107031811.65960870@posting.google.com> We use VxWorks 5.4 for PPC and MPC860. Which network device does wdb agent use when configured to use END driver? The one specified in DEFAULT_BOOT_LINE or the one which is the first entry of endDevTbl[]? Regards, lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 2.0 BSP for MVME5100 boards. Date: Wed, 04 Jul 2001 02:20:12 GMT From: "Mark Fanara" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <9htcku$jap$1@overload.lbl.gov> Greg - We had similar problems using shared memory queues between 5101 boards and 2304 boards. We found that the 5100 BSP did not contain a fix to a known problem. Our FSE tested our equipment for weeks and found it!. I would assume that the FCS (first customer shipment) version of the BSP has this fix. If not, contact me and I will send it to you. Mark Greg Milne wrote in message news:9htcku$jap$1@overload.lbl.gov... > Hello gurus, > > I am attempting to port our system from the Motorola 2432 boards to the > MVME5100. The system consists of master and slave VME boards ( both 5100s ) > processing data fed to them from an NT server. The master and slave use > shared memory and the proxy ARP network protocol. I am using the Early > Access BSP supplied by Motorola. The system seems to run flawlessly for > 7-12 hours at which point it will freeze. To unfreeze, I simply have to > ping the slave or open a telnet session. > > Any ideas, suggestions ? Have I provided enough info ? > > Thanks in advance, > > Gregor Milne --------------------------- Newsgroups: comp.os.vxworks Subject: How can I find the false with the image?? Date: Wed, 4 Jul 2001 12:11:58 +0800 From: "kumo Yeh" Organization: SEEDNet News Service Message-ID: <9hu565$88i$1@news.seed.net.tw> Hi: When Idownload an image& I got this How can I know where the problem is?? Can someone tell me plz! /***************************************/ VxWorks Copyright 1984-1998 Wind River Systems, Inc. CPU: EST est8240 Evaluation Board VxWorks: 5.4 BSP version: 1.2/4 Creation date: Jul 4 2001 WDB: Ready. data access Exception current instruction address: 0x0012fda0 Machine Status Register: 0x0000a030 Data Access Register: 0x06e20023 Condition Register: 0x44000080 Data storage interrupt Register: 0x0000a030 Task: 0x1fffdf8 "tRootTask" /*****************************************/ - -- Kumo,Yeh. email:kumo@ms42.url.com.tw --------------------------- Newsgroups: comp.os.vxworks Subject: Re: standalone_vxworks with no network Date: Tue, 03 Jul 2001 21:49:14 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B42A04A.CFDC0357@covad.net> References: Reply-To: drdiags@covad.net Kelly, So using the "other=" field did not help with getting your network initialized? I thought the fix for this was to do something similiar to setting "other=motfcc0" or the like. Did you check the vxWorks FAQ? http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html Kelly Hornsby wrote: > > hello > > I need some ideas of something that I can't figure out. I'm using T2 on a > Sandpoint 750 system. I have a DEC 21143 NIC, and it works just fine with > the bootloader/ ftp-Vxworks. However, I've created the > standalone_vxworks.hex file, and flashed it into memory, but the network is > not initialized now. the NIC is detected fine (pciHeaderShow indicates > this) and the 21143 PCI initilization routine is executed (sysNetHwInit2), > but there's no dev out there, and the device shows no devices for ifShow. > > I'm looking for suggestions or ideas for why this occurs. Any ideas? --------------------------- Newsgroups: comp.os.vxworks Subject: subscribtion Date: Wed, 04 Jul 2001 10:02:07 +0530 From: msudhir Organization: Lawrence Berkeley National Laboratory Message-ID: <9hu8ot$o8v$1@overload.lbl.gov> hello, can you please subscribe my mailid for VxWorks users Group? My mail Id is sudhir@cmcltd.com. Thanking you, M.Sudhir Engineer,R&D CMC Ltd. --------------------------- Newsgroups: comp.os.vxworks,comp.sys.arm Subject: Bringing up the ARM Integrator/CM7TDMI with vxWorks Date: 3 Jul 2001 22:25:03 -0700 From: suman@mistralsolutions.com (Suman Anil) Organization: http://groups.google.com/ Message-ID: Hi , I have an ARM Integrator/CM7TDMI AP development motherboard , the core microprocessor being ARM7TDMI . I need to bring this board up with vxWorks . I installed T2 with the BSP pid7t . Built bootrom.hex ( making appropriate changes to get M32 S-Record format )& flashed it . I failed to see anything on the hyper terminal . Am I using the right BSP ? If pid7t is the correct BSP , are there any changes I need to make to bring up this board ? Any help is greatly appreciated . Thanx in advance, Suman --------------------------- Newsgroups: comp.os.vxworks Subject: IP address of VxWorks Date: 3 Jul 2001 22:31:43 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107032131.640f74f7@posting.google.com> Our VxWorks image is indepencent of bootROM. We burn it into FLASH directly use flash programmer, not load it via bootROM. We can set an IP address for the system in DEFAULT_BOOT_LINE. I find only when "network components --> basic network initialization components --> bootline processing components network initializaton - --> initialize network at boot time" is included, the VxWorks will initialize the network device specified in DEFAULT_BOOT_LINE and ipAttach() and usrNetIfConfig() to it. This means that, if I don't include "initialize network at boot time", none ipAttach() and usrNetIfConfig() will happen, the target system will not be endrowed with an IP address acturally. We cann't "ping" it successfully. At this time, if we want to make the IP stack up, we have to invoke ipAttach() and usrNetIfConfig() manually. Is my understanding right? Best Regards, lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: two ethernet controllers in VxWorks Date: 3 Jul 2001 23:38:52 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107032238.943af64@posting.google.com> VxWorks 5.4 for PPC. Our VxWorks image is indepencent of bootROM. We burn it into FLASH directly use flash programmer, not load it via bootROM. I know we can set an IP address for the system in DEFAULT_BOOT_LINE. Now I want to make SCC1 and SCC2 of MPC860 as etnernet controller. Thus there are two ethernet controllers (network devices) in VxWorks. Should I prepare two IP address for our system in this case? How to set DEFAULT_BOOT_LINE since there are two network device? Which network device should I ipAttach() and usrNetIfConfig()? If two network dvice both need to do so, then when I create a socket and bind(), which IP address is bound to? lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: nanosleep Date: 4 Jul 2001 00:00:37 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107032300.1015546c@posting.google.com> How POSIX routine nanosleep() implement nanosecond level sleep internally? Does it also base on ticks? Our system's system frequency is 50MHz, and we make it 100 ticks/second. A tick is 10ms. Then how can nanosleep() effect? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: two IP stacks Date: Wed, 4 Jul 2001 09:10:20 +0200 From: "Michel Hendriks" Organization: XS4ALL Internet BV Message-ID: <9hufh8$k51$1@news1.xs4all.nl> References: <340587ac.0107030903.5f999b75@posting.google.com> Sender: tcj@gateway.pijnenburg.nl 1- 2 Stacks, 1 IP address: how do you plan on separating incoming connections? 2 - You want to provide private IP addresses to the outside world? They're called private for a reason 3 - You cannot connect directly to another network's private machines. That network would need 1 public machine that acts as a firewall/proxy with NAT. Then, you tell the public machine to which local machine to forward incoming connections (e.g. port 80 to 192.168.0.1, port 21 to 192.168.0.2, etc) But maybe I don't understand what you're asking, and am I just blabbing here Greets billy du wrote in message news:340587ac.0107030903.5f999b75@posting.google.com... > I do have quite and unusual request (but, yes I really do want > two interfaces with the same IP address, which is possible in at > least Linux and windows)? > > Emmanuel Herbreteau wrote in message news:<3B3C24E7.3783DE68@sepro-robotique.com>... > > billy du wrote: > > > Can VxWorks support two IP stacks inside one CPU? > > > > What ??? > > > > > Unfortuately, the outside world might use the private IP addresses. If > > > we uses private IP addresses for internal connection, there might be > > > conflict. Therefore, we plan to use two IP stacks in one card. > > > > You should use private IP address for your private network : Look at > > RFC 1918 on the Internet (or a good book, like "TCP/IP" from MacMillan). > > > > ie : Private address (class C) "192.168.16.44" or private address (class > > B) > > "172.16.1.27" will never conflict with an Internet address. > > Sorry, my question is not clear. We want to support private IP > addresses in the outside world as well because our equipment might > connect to a private network using IP address like 10.x.x.x > 192.168.x.x. Then we have a problem to assign IP address inside the > equipment. That's why we need two IP stack or any other approach. --------------------------- Newsgroups: comp.os.vxworks Subject: unsubscribtion Date: Wed, 4 Jul 2001 08:09:41 +0100 (WET DST) From: "Kim.Dokhac" Organization: Lawrence Berkeley National Laboratory Message-ID: <200107040709.IAA28772@machina.dso.meteo.fr> hello, can you please unsubscribe my mailid for VxWorks users Group? My mail Id is kim.dokhac@meteo.fr Thank you, ================================================================================ Kim DO KHAC Tel : 01 30 13 61 84 METEO-FRANCE Fax : 01 30 13 60 62 DSO/CMR Email : kim.dokhac@meteo.fr 7 , Rue Teisserenc-de-Bort BP 202 78195 Trappes ================================================================================ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: two ethernet controllers in VxWorks Date: Wed, 4 Jul 2001 16:14:51 +0900 From: "???" Organization: Korea Telecom Message-ID: <9hufr5$t8$1@news1.kornet.net> References: <1d411918.0107032238.943af64@posting.google.com> I use BOOT_LINE and another NVRAM memory for system INFORMATION. you can store two network device INFORAMTION in other NVRAM space. and, whenever booting, you read it first. and set the system. "DragonSpring" wrote in message news:1d411918.0107032238.943af64@posting.google.com... > VxWorks 5.4 for PPC. Our VxWorks image is indepencent of bootROM. We > burn it into FLASH directly use flash programmer, not load it via > bootROM. > > I know we can set an IP address for the system in DEFAULT_BOOT_LINE. > Now I want to make SCC1 and SCC2 of MPC860 as etnernet controller. > Thus there are two ethernet controllers (network devices) in VxWorks. > > Should I prepare two IP address for our system in this case? How to > set DEFAULT_BOOT_LINE since there are two network device? > > Which network device should I ipAttach() and usrNetIfConfig()? If two > network dvice both need to do so, then when I create a socket and > bind(), which IP address is bound to? > > lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: How can I know the socket peer closed? Date: 4 Jul 2001 00:49:18 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107032349.71a4864f@posting.google.com> VxWorks 5.4 for PPC. I create a socket and use it to establish a TCP connect. Then I create a new task to read from the socket. This is infinite loop like this: while ((n = read (sock, buf, sizeof (buf))) > 0) { /* deal with buf */ } /* n<=0 means that the remote close the TCP connect */ This "while" loop terminate when the remote closes TCP connect or the remote crashes. To ensure this, I use /* turn on KEEPALIVE so if the remote crashes, we'll know about it */ optval = 1; setsockopt (sock, SOL_SOCKET, SO_KEEPALIVE, (char *) &optval, sizeof (optval)); But I find it is no use. When I terminal remote normally(i.e. the remote close() the socket explicitly), then all is OK, the "while" loop terminate. But when I shutdown the remote host suddenly or pull out the network cable, the "while" loop will not terminate. How can I know that the TCP connect has lost in this case? why setsockopt(..., SO_KEEPALIVE,...) is no use? Regards, lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: FTP SERVER Date: Wed, 04 Jul 2001 08:53:19 +0100 From: Graham Baxter Organization: Graham Baxter (Software) Limited Message-ID: <3B42CB6F.60697C5D@NOSPAM.bcs.org.uk> VxWorkers, Is there a hook inside the FTP server so that a function can be called after a file is received? I need to do some validation and processing of received files. I know that I can look at the FTP directories say every minute, but there may be some part written files of FTPs in progress. Thanks for any ideas. - -- Regards, Graham Baxter Freelance Software Engineer gbaxter@NOSPAM.bcs.org.uk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: signal handling in vxworks. Date: Wed, 04 Jul 2001 09:14:01 +0100 From: Graham Baxter Organization: Graham Baxter (Software) Limited Message-ID: <3B42D049.6E990ADB@NOSPAM.bcs.org.uk> References: <9hsuiv$bdd$1@overload.lbl.gov> You are allowed to call IO (printf) from a signal handler. (VxWorks Programmer's Guide 5.4 Edition 1 section 2.4.7 p82). Try incrementing a semaphore which is waited for by a task which does the printf. - -- Regards, Graham Baxter Freelance Software Engineer gbaxter@NOSPAM.bcs.org.uk --------------------------- Newsgroups: comp.os.vxworks Subject: Problems with Cygwin GCC WinNT cross i386-vxWorks Date: 04 Jul 2001 07:39:06 UT From: fscheurer@manz-acs.de Organization: Lawrence Berkeley National Laboratory Message-ID: <9hujad$76u$1@overload.lbl.gov> Hi all, (if you answer please use my E-Mail address as well. Not only the mailing-list! TIA, F. Scheurer) I try to build the tool-chain mentioned above (explained somewhat by a HOWTO of Mumit Khan): #host=i686-cygwin-windowsnt host=i686-pc-cygwin ## FS 2001-07-03 build=$host #target=m68k-coff ## FS 2001-06-29 target=i386-wrs-vxworks ## FS 2001-07-03 #tmpdir=/tmp I passed configuration well using the following script (based on some contributions of CrossGCC and Mumit Khan) ... #prefix=/bar ## FS 2001-06-29 #prefix=../bar ## FS 2001-07-03 prefix=/usr/local/i386 i=$prefix/bin #mkdir build-binutils build-gcc build-newlib build-gdb # Configure, build and install binutils cd build-binutils #../binutils-2.9.1/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-02 echo "Trying to build binutils ..." #sh ../binutils-2.11.2/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 (see Khan-HOWTO) sh ../binutils-2.11.2/configure \ --with-included-gettext \ --target=$target --host=$host --build=$build \ --prefix=$prefix -v make all install # Configure, build and install gcc ONLY # ************************************************************************* # ATTENTION: # Huh, why the .exe extension? Well, it has to do with gcc hosted on # Cygwin, and you can always patch gcc sources to get rid of the automatic # .exe addition. Just comment out the EXECUTABLE_SUFFIX macro in # $/gcc-2.XX.Y-Z/gcc/config/i386/xm-cygwin.h and rebuild gcc. # ************************************************************************* cd ../build-gcc #../gcc-2.8.1/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-03 (see Khan-HOWTO Cygwin-Sources fixed for Cygwin platform) echo "Trying to build gcc ..." #sh ../gcc-2.95.3-5/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 (see Khan-HOWTO) sh ../gcc-2.95.3-5/configure \ --enable-languages=c,c++ \ --with-included-gettext --enable-shared --enable-threads \ --target=$target --host=$host --build=$build \ --with-newlib \ --prefix=$prefix -v #make all install make LANGUAGES=c all-gcc > make.log 2>&1 mv make.log make-c-only.log make LANGUAGES=c install-gcc > install.log 2>&1 mv install.log install-c-only.log # Configure, build and install newlib cd ../build-newlib #../newlib-1.8.1/configure --target=$target --prefix=$prefix -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. #make all install \ # CC_FOR_TARGET=$i/${target}-gcc \ # AS_FOR_TARGET=$i/${target}-as \ # LD_FOR_TARGET=$i/${target}-ld \ # AR_FOR_TARGET=$i/${target}-ar \ # RANLIB_FOR_TARGET=$i/${target}-ranlib ## FS 2001-07-02 (see Khan-HOWTO) echo "Trying to build newlib ..." #sh ../newlib-1.9.0/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. #make all \ # CC_FOR_TARGET=$i/${target}-gcc \ # AS_FOR_TARGET=$i/${target}-as \ # LD_FOR_TARGET=$i/${target}-ld \ # AR_FOR_TARGET=$i/${target}-ar \ # RANLIB_FOR_TARGET=$i/${target}-ranlib ## FS 2001-07-03 (see Khan-HOWTO) sh ../newlib-1.9.0/configure --target=$target --host=$host --build=$build \ --prefix=$prefix -v # The settings for FOO_FOR_TARGET aren't necessary if you put $prefix/bin # in your path before running this. make all install \ CC_FOR_TARGET=$i/${target}-gcc \ AS_FOR_TARGET=$i/${target}-as \ LD_FOR_TARGET=$i/${target}-ld \ AR_FOR_TARGET=$i/${target}-ar \ RANLIB_FOR_TARGET=$i/${target}-ranlib # Configure, build and install gcc "the rest" ## FS 2001-07-03 (see Khan-HOWTO) cd ../build-gcc make all install # Configure, build and install gdb cd ../build-gdb #../gdb-4.17/configure --target=$target --prefix=$prefix -v #make all install ## FS 2001-07-02 echo "Trying to build gdb ..." #sh ../gdb-4.18/configure --target=$target --host=$host --prefix=$prefix --tmpdir=$tmpdir -v #make all ## FS 2001-07-03 sh ../gdb-4.18/configure --target=$target --host=$host --prefix=$prefix -v make all install All builds well - - binutils - - gcc-only - - newlib until it comes to - - gcc-all then problems occur with non-matching GNU/VxWorks or missing Header-Files 1. vxXXXXX.h are missing for name in _eh; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_new lib1.9.0/build-gcc/gcc/ -B/usr/local/i386/i386-wrs-vxworks/bin/ -I/usr/local/i386/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN _GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc -fexceptions -I. -I../.. /gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include -c \ -DL${name} ../../gcc-2.95.3-5/gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _eh In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:170, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxArch.h:88: arch/i86/archI86.h: No such file or directory In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:175, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxTypesOld.h:75: sys/types.h: No such file or directory make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 2. including them leads to conflicts with stddef.h for name in _eh; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_new lib1.9.0/build-gcc/gcc/ -B/usr/local/i386/i386-wrs-vxworks/bin/ -I/usr/local/i386/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN _GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -Dinhibit_libc -fexceptions -I. -I../.. /gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include -c \ -DL${name} ../../gcc-2.95.3-5/gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _eh In file included from ../../gcc-2.95.3-5/gcc/config/vxWorks.h:174, from ../../gcc-2.95.3-5/gcc/gthr-vxworks.h:38, from gthr-default.h:1, from ../../gcc-2.95.3-5/gcc/gthr.h:98, from ../../gcc-2.95.3-5/gcc/libgcc2.c:3058: ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: redefinition of `ptrdiff_t' include/stddef.h:120: `ptrdiff_t' previously declared here ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: conflicting types for `size_t' include/stddef.h:172: previous declaration of `size_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: conflicting types for `wchar_t' include/stddef.h:257: previous declaration of `wchar_t' make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 3. Substituting stddef.h with the subset which comes with VxWorks leads us to: c:/usr/bin/gcc -c -DCROSS_COMPILE -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.3-5/gcc -I../../gcc-2.95.3-5/gcc/conf ig -I../../gcc-2.95.3-5/gcc/../include ../../gcc-2.95.3-5/gcc/cp/g++spec.c In file included from ../../gcc-2.95.3-5/gcc/config/types/vxANSI.h:44, from ../../gcc-2.95.3-5/gcc/config/types/vxTypes.h:63, from ../../gcc-2.95.3-5/gcc/config/sys/types.h:39, from /usr/include/time.h:32, from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/cp/g++spec.c:22: ./../gcc-2.95.3-5/gcc/config/types/vxCpu.h:200: #error CPU is not defined correctly make[1]: *** [g++spec.o] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 4. Integrating CPU Macro/Define into highest level Makefile (again with original Cygwin-stddef.h in gcc sources) CPU = I80386 .. CC_FOR_BUILD = $(CC) -DCPU=$(CPU) .. leads us to conflicts in the build-tools: c:/usr/bin/gcc -DCPU=I80386 -c -DCROSS_COMPILE -DIN_GCC -g -O2 -DHAVE_CONFIG_H -I. -I../../gcc-2.95.3-5/gcc -I../../gcc-2.95 .3-5/gcc/config -I../../gcc-2.95.3-5/gcc/../include \ ../../gcc-2.95.3-5/gcc/gencheck.c In file included from ../../gcc-2.95.3-5/gcc/config/sys/types.h:39, from /usr/include/time.h:32, from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: warning: redefinition of `size_t' c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:172: warning: `size_t' previously declared here In file included from /usr/include/sys/reent.h:14, from /usr/include/stdio.h:48, from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/time.h:47: warning: `__cdecl__' attribute directive ignored /usr/include/time.h:47: parse error before `clock' In file included from ../../gcc-2.95.3-5/gcc/system.h:37, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/stdio.h:50: warning: redefinition of `fpos_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:91: warning: `fpos_t' previously declared here In file included from /usr/include/stdlib.h:17, from ../../gcc-2.95.3-5/gcc/system.h:150, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:257: conflicting types for `wchar_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: previous declaration of `wchar_t' In file included from /usr/include/sys/unistd.h:12, from /usr/include/unistd.h:4, from ../../gcc-2.95.3-5/gcc/system.h:154, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: c:/usr/bin/../lib/gcc-lib/i686-pc-cygwin/2.95.3-5/include/stddef.h:120: conflicting types for `ptrdiff_t' ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: previous declaration of `ptrdiff_t' In file included from /usr/include/sys/time.h:27, from ../../gcc-2.95.3-5/gcc/system.h:194, from ../../gcc-2.95.3-5/gcc/gencheck.c:22: /usr/include/sys/select.h:28: parse error before `fd_set' make[1]: *** [gencheck.o] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/CrossGCC/build_newlib1.9.0/build-gcc/gcc' make: *** [all-gcc] Error 2 What's the right way throug all this mess? Is the actual portability to vxworks not given? Where to find some more information about that X-ing i686-cygwin-X-x86-vxworks? Are all the adaptions to be made in the headers drawn from {WIND_BASE}/target/h? Any hint appreciated. Many thanks in advance! MfG / Kind Regards Friedrich Scheurer, R&D - -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de - -------------------------------- To: Steven_Snyder@3com.com aoliva@redhat.com Cc: gcc-help@gcc.gnu.org vxwexplo@lbl.gov --------------------------- Newsgroups: comp.os.vxworks Subject: re: VxWorks Scheduling Date: Wed, 4 Jul 2001 00:38:26 -0700 From: "unnikrishnan purushothaman kartha" Organization: Lawrence Berkeley National Laboratory Message-ID: <9hujaf$76v$1@overload.lbl.gov> The behaviour of the scheduler would be like this: taskLock is no more valid when an operation involving scheduler is made (semGive in this case). Hence TaskB will be unblocked immediately and TaskA takes semaphore again. When func_3 returns, semGive is executed once more and TaskA returns. The current version of vxWorks is 5.4. I am not sure about which memory related problems you are referring to. Fragmentation problems remain. WindRiver has recently released a different OS called VxWorks AE which implements memory protected architecture. regards Unni. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem with inet_aton Date: 4 Jul 2001 01:20:15 -0700 From: yevgeni_tunik@yahoo.com (Yevgeni Tunik) Organization: http://groups.google.com/ Message-ID: <203c9486.0107040020.15d45f70@posting.google.com> References: <9e9c2446.0107020642.70000103@posting.google.com> Yevgeni_Tunik@yahoo.com (Yevgeni Tunik) wrote in message news:<9e9c2446.0107020642.70000103@posting.google.com>... > > I noticed VxWorks does not support "inet_addr" > > This function works, but (at least on simulator) reverses the byte > order. This possibly because of difference between Pentium and Network byte ordering. The communication works, at least between client and server sockets within the same VxSim on NT4. --------------------------- Newsgroups: comp.os.vxworks Subject: How to make my objects to library? Date: 4 Jul 2001 01:30:33 -0700 From: eaichul@ntl30.com (Eaichul Choung) Organization: http://groups.google.com/ Message-ID: <3da3205f.0107040030.2fe1bb7b@posting.google.com> Hi there, I am using vxWorks now. and.. I hope to make Makefile for building library. I heard I should use arppc. How could I make Makefile for Library building? Is there anyone who have an example Makefile for it? regards, Eaichul Choung eaichul@ntl30.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about "taskSpawn" Date: Wed, 04 Jul 2001 11:36:50 +0200 From: Amit Raam Organization: Verio Message-ID: References: <9hrqlg$mr0$1@news1.kornet.net> vxworksintent@tao-group.com (David Given) wrote: >In article <9hrqlg$mr0$1@news1.kornet.net>, > benjamin writes: >> When creating a task with the function >> "taskSpawn", we must have information >> about "stack size". >> >> As vxworks's beginner, I like to know why the stack size is needed and what it is used for. >> >> I hope anyone who know the answers gives me those. > >The stack is used by C to store things like the function call sequence and >automatic (local) variables. Usually 8kB is a good size, but the more >complicated the task is, the more it will need; deeply nested recursive >functions will use lots of stack, etc. VxWorks does not (as far as I am >aware) have the ability to dynamically grow stacks. You can type checkStack at your vxWorks terminal to see the status of your task stack usage. Note the Margin column, it tells you how much of the allocated stack was never used. Amit Raam Email address: raama at indigo.co.il - -------------------------------------------------- ".sigless and proud of it" (R) (C) (tm) (look-and-feel) Std. disclaimer:I don't represent the opinions of anyone but myself. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Memory requirments.... Date: Tue, 3 Jul 2001 21:29:52 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994236581.137210@newsmaster-04.atnet.at> References: <9hqrpd$rj0$1@overload.lbl.gov> For the VxWorks image itself you can see some rough numbers from the project facility. Compose your image as you think you'll need it and view the properties of the root (your project) node in the VxWorks tab. It shows you approximately how large your image will be. This does not tell you anything about dynamic memory usage (malloced storage, task stacks and so on). I don't know of any good documentation of the latter. You could try run the image (and maybe your app, if it already exists) in the simulator and use memShow() and friends to find out how much it requires. Keep in mind that it is not the same platform, though hth werner "Michael Morrison" wrote in message news:9hqrpd$rj0$1@overload.lbl.gov... > Anyone know where I can find information on the memory > requirments/footprint of VxWorks and it's > various components? I just need a rough idea of how much memory our > box will need. > > > Thanks > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem with timers Date: Wed, 4 Jul 2001 10:59:06 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994237625.252395@newsmaster-04.atnet.at> References: <9htcku$jao$1@overload.lbl.gov> at least my newsreader only presents character salad instead of the file(s) you promised. maybe you try post it again with the code in the body regards werner "Anuradha Pani" wrote in message news:9htcku$jao$1@overload.lbl.gov... > --0-1957747793-994193458=:28046 > Content-Type: text/plain; charset=us-ascii > Content-Disposition: inline > > hi all, > > i am new to vxworks. > attached is a simple piece of code where i have tried > to implement timers. but when i start the task, > startTimer, nothing happens. i see no output at all. > not even the "hi" i deliberately put in. > > can somebody help me? > > thanks > anuradha > > __________________________________________________ > Do You Yahoo!? > Get personalized email addresses from Yahoo! Mail > http://personal.mail.yahoo.com/ > --0-1957747793-994193458=:28046 > Content-Type: application/x-unknown; name="timerImp.c" > Content-Transfer-Encoding: base64 > Content-Description: timerImp.c > Content-Disposition: attachment; filename="timerImp.c" > > LyogRmlsZTogdGltZXJJbXAuYyAtIENyZWF0ZXMgYW5kIHNldHMgYSB0aW1l > ciAqLwoKLyogSW5jbHVkZSBmaWxlcyAqLwojaW5jbHVkZSAic3RkaW8uaCIK > I2luY2x1ZGUgInZ4V29ya3MuaCIKI2luY2x1ZGUgInRhc2tMaWIuaCIKI2lu > Y2x1ZGUgInNpZ0xpYi5oIgojaW5jbHVkZSAidGltZS5oIgoKCgovKiBGdW5j > dGlvbjogc3RhcnRUaW1lcigpCiAgIFRoaXMgZnVuY3Rpb24gc3Bhd25zIGEg > dGFzayB0byBjYXRjaCB1c2VyIHNpZ25hbHMuIAogICBJdCBzdGFydHMgYSB0 > aW1lciB3aXRoIGEgdGltZS1wZXJpb2Qgb2YgNTAgbXMgYW5kIHVwb24gZXhw > aXJ5CiAgIHNlbmRzIHVzZXIgc2lnbmFscyB0byB0aGUgdGFzayBzcGF3bmVk > LgogICBJdCB0aGVuIHJlX2FybXMgdGhlIHRpbWVyLgoqLwp2b2lkIHN0YXJ0 > VGltZXIodm9pZCkKewogIHN0cnVjdCBpdGltZXJzcGVjIHRpbWVQZXJpb2Q7 > CiAgdGltZXJfdCB0aW1lcklkOwogIGludCBpOwoKCiAgLyogRnVuY3Rpb24g > cHJvdG90eXBlcyAqLwogIHZvaWQgdG1yRXhwaXJ5SGFuZGxlcihpbnQgc2ln > bmFsKTsKICAKICAgcHJpbnRmKCJIaVxuIik7CgogICAvKiBDcmVhdGUgYSB0 > aW1lciAqLwogICBpZiggdGltZXJfY3JlYXRlKENMT0NLX1JFQUxUSU1FLCBO > VUxMLCAmdGltZXJJZCkgPT0gLTEpCiAgICB7CiAgICAgICAgcHJpbnRmKCJF > cnJvciBpbiBjcmVhdGluZyB0aW1lclxuIik7CglleGl0KDEpOwogICAgfQog > ICBlbHNlCiAgICB7CiAgICAgIHByaW50ZigiVGltZXIgd2l0aCBpZCA9ICVk > ICAgIGNyZWF0ZWQgc3VjY2Vzc2Z1bGx5XG4iLCB0aW1lcklkKTsKICAgICAg > LyogQXR0YWNoIGEgdGltZXIgcm91dGluZSB0byB0aGUgdGltZXIgKi8KICAg > ICAgaWYoIHRpbWVyX2Nvbm5lY3QodGltZXJJZCwgKHZvaWQqKSZ0bXJFeHBp > cnlIYW5kbGVyLCAwKSA9PSAtMSkKICAgICAgIHsKCSBwcmludGYoIkVycm9y > IGluIGF0dGFjaGluZyB0aW1lciByb3V0aW5lXG4iKTsKCSBleGl0KDEpOwog > ICAgICAgfQogICAgICBlbHNlCiAgICAgICB7CgkgcHJpbnRmKCJUaW1lciBy > b3V0aW5lIGF0dGFjaGVkIHN1Y2Nlc3NmdWxseVxuIik7CgkgLyogU2V0IHRp > bWVyIHBlcmlvZCAqLwoJICB0aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X3Nl > YyA9IDA7ICAgLyogTm8gcmVwZWF0YXRpb25zIGZvciBub3cgKi8KICAJICB0 > aW1lUGVyaW9kLml0X2ludGVydmFsLnR2X25zZWMgPSAwOyAgCiAgCSAgdGlt > ZVBlcmlvZC5pdF92YWx1ZS50dl9zZWMgPSAwOwogICAgICAgICAgdGltZVBl > cmlvZC5pdF92YWx1ZS50dl9uc2VjID0gNTAwMDAwMDA7IC8qIGZpcnN0IGV4 > cGlyeSBhZnRlciA1MCBtcyAqLwoKCSAgLyogRGVsYXkgZm9yIHNvbWV0aW1l > ICovCgkgIHRhc2tEZWxheSgzMCk7CgoJICBpZiggdGltZXJfc2V0dGltZSh0 > aW1lcklkLCAhVElNRVJfQUJTVElNRSwgJnRpbWVQZXJpb2QsIE5VTEwpID09 > IC0xKQogICAgICAgICAgIHsKICAgICAgCSAgICAgICBwcmludGYoIkVycm9y > IGluIHNldHRpbmcgdGltZXJcbiIpOwogICAgICAJICAgICAgIGV4aXQoMSk7 > CiAgICAgICAgICAgfQogICAgICAgICAgIHByaW50ZigiVGltZXIgc2V0IHN1 > Y2Nlc3NmdWxseVxuIik7CiAgICAgICB9CiAgICB9CgoKICAgLyogVG8gbWFr > ZSB0aGlzIHRhc2sgc3RheSBhbGl2ZSAqLwogICBmb3IoaT0wO2k8MTAwO2kr > KykKICAgIHsKCXByaW50ZigiVGFzayBpcyBhbGl2ZVxuIik7CiAgICB9CgoK > fSAvKiBFbmQgb2Ygc3RhcnRUaW1lciAqLwoKCi8qIEZ1bmN0aW9uOiB0bXJF > eHBpcnlIYW5kbGVyKCkgdG8gaGFuZGxlIHRpbWVyRXhwaXJ5ICovCnZvaWQg > dG1yRXhwaXJ5SGFuZGxlcihpbnQgc2lnbmFsKQp7CiAgcHJpbnRmKCJJbnNp > ZGUgVGltZXIgRXhwaXJ5IEhhbmRsZXJcbiIpOwoKCn0gLyogRW5kIG9mIHRt > ckV4cGlyeUhhbmRsZXIgKi8KCgoKCg== > > --0-1957747793-994193458=:28046-- > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: interface between asm and C language Date: Wed, 4 Jul 2001 11:07:08 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994238106.754754@newsmaster-04.atnet.at> References: <1d411918.0107031800.18b3244e@posting.google.com> Maybe this depends on the architecture in question For x86 C calling convention is used, parameters are put on the stack from right to left (the last argument is put on the stack first). The stack is cleaned up by the caller, your asm routine may just RETurn. I'd suggest using data types for passing between ASM and C where the alignment is obvious (like 4 byte int and pointers in x86 architecture). For details I'd recommend the GCC documentation. regards werner "DragonSpring" wrote in message news:1d411918.0107031800.18b3244e@posting.google.com... > Does any one know the interface between asm and C language when I use > VxWorks and gcc? For example, when I invoke(jump to) a C routin in an > asm file, how to pass parameters to the C routin? How the stack > changes? > > lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I want to know "shell". Date: Wed, 4 Jul 2001 11:09:12 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994238230.681601@newsmaster-04.atnet.at> References: <9hrrep$qoc$1@news1.kornet.net> There is unsupported source code for a telnet server your WRS FAE might be willing to give you. Ask him/her. regards werner "¿ÁÀ»¼®" wrote in message news:9hrrep$qoc$1@news1.kornet.net... > I use a standalone image without shell. > but, because I must support "telnetd", I used 'ioTaskStdGet/ioTaskStdSet' > instead of 'ioGlobalStdGet/ShellOrigStdSet". > and restarted "a task like shell" > but, not working. > > I must have restarted "Shell". > I expect that "Shell" process fd(s). > > I want to know "Shell". > any Information or shell source make me happy. > > Thanks in advanced. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to make my objects to library? Date: Wed, 4 Jul 2001 11:22:54 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994239053.856725@newsmaster-04.atnet.at> References: <3da3205f.0107040030.2fe1bb7b@posting.google.com> You are using Tornado 2, are you? The easiest way to get a library is to use the project facility. Open the Build Properties, Rules Tab Select the rule 'archive' Build the stuff and you have a nice archive with all your objects in there. hth werner "Eaichul Choung" wrote in message news:3da3205f.0107040030.2fe1bb7b@posting.google.com... > Hi there, > > I am using vxWorks now. and.. I hope to make Makefile for building library. > I heard I should use arppc. How could I make Makefile for Library building? > Is there anyone who have an example Makefile for it? > > regards, > > Eaichul Choung > eaichul@ntl30.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FTP SERVER Date: Wed, 4 Jul 2001 11:24:54 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994239172.231953@newsmaster-04.atnet.at> References: <3B42CB6F.60697C5D@NOSPAM.bcs.org.uk> In target/unsupported/src/netwrs you should find source code for an FTP server. Use this instead of the default and add your special processing. I don't know of any hook in the default FTP server. hth werner "Graham Baxter" wrote in message news:3B42CB6F.60697C5D@NOSPAM.bcs.org.uk... > VxWorkers, > > Is there a hook inside the FTP server so that a function can be called > after a file is received? > > I need to do some validation and processing of received files. I know > that I can look at the FTP directories say every minute, but there may > be some part written files of FTPs in progress. > > > Thanks for any ideas. > > > -- > > Regards, > > > Graham Baxter > Freelance Software Engineer > gbaxter@NOSPAM.bcs.org.uk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: which network device does wdb agent use Date: Wed, 4 Jul 2001 11:26:07 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994239245.404002@newsmaster-04.atnet.at> References: <1d411918.0107031811.65960870@posting.google.com> I guess the first one in the table (no matter if it actually works). regards werner "DragonSpring" wrote in message news:1d411918.0107031811.65960870@posting.google.com... > We use VxWorks 5.4 for PPC and MPC860. > > Which network device does wdb agent use when configured to use END > driver? The one specified in DEFAULT_BOOT_LINE or the one which is the > first entry of endDevTbl[]? > > Regards, > > lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: nanosleep Date: 4 Jul 2001 09:32:33 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <1d411918.0107032300.1015546c@posting.google.com> lihuapeng@21cn.com (DragonSpring) wrote: > How POSIX routine nanosleep() implement nanosecond level sleep > internally? Does it also base on ticks? Our system's system frequency > is 50MHz, and we make it 100 ticks/second. A tick is 10ms. Then how > can nanosleep() effect? This routine is based on the system tick, so you don't get nanosecond sleeps. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FTP SERVER Date: 4 Jul 2001 09:34:18 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <3B42CB6F.60697C5D@NOSPAM.bcs.org.uk> <994239172.231953@newsmaster-04.atnet.at> "Werner Schiendl" wrote: > In target/unsupported/src/netwrs > you should find source code for an FTP server. > > Use this instead of the default and add your special processing. > > I don't know of any hook in the default FTP server. This server is not in the standard distribution, it is part of DosFS2 Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: why the "unit number" disappeared? Date: 4 Jul 2001 03:09:49 -0700 From: gcai@netease.com (cai) Organization: http://groups.google.com/ Message-ID: <31f1a7e2.0107040209.41c46dfa@posting.google.com> hi all, I was confused by the follow question. I found the same thing in configurations of El3c90x. in dec21x4xEnd.c ,it says: The format of the parameter string is "::::::\ ::" but I found these example codes in configNet.h: #define DEC_LOAD_FUNC dec21x40EndLoad #define DEC_BUFF_LOAN 1 #define DEC_LOAD_STRING \ "0x81020000:0x80000000:0x12:0x12:-1:-1:-1:0:0x80000000" /* ::::: ::: */ extern END_OBJ* DEC_LOAD_FUNC (char*, void*); END_TBL_ENTRY endDevTbl [] = { { 0, DEC_LOAD_FUNC, DEC_LOAD_STRING, DEC_BUFF_LOAN, NULL, FALSE}, { 0, END_TBL_END, NULL, 0, NULL, FALSE}, }; Thanks in advance. cai --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 4 23:56:27 2001 From: fscheurer@manz-acs.de Date: Wed Jul 4 23:56:30 PDT 2001 Subject: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Hi VxWorks-Gurus, I ran into the problem, that the integrated target-headers seem to be not confor with GNU expectations. This means e.g.: 1.) CPU is not defined as expected by the target-headers .. ./../gcc-2.95.3-5/gcc/config/types/vxCpu.h:200: #error CPU is not defined correctly .. 2.) vxTypes.h redefines and has other parse errors .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: warning: redefinition of `size_t' .. /usr/include/time.h:47: warning: `__cdecl__' attribute directive ignored /usr/include/time.h:47: parse error before `clock' .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:91: warning: `fpos_t' previously declared here .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: previous declaration of `wchar_t' .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: previous declaration of `ptrdiff_t' .. /usr/include/sys/select.h:28: parse error before `fd_set' .. MY QUESTION: Where is the most central place in - the sources, - the configure-script or - the Makefile to "repair" it AND to get the right types to be compatible to objects build with older gcc-version? Any hint is gratefully accepted! MfG / Kind Regards Friedrich Scheurer, R&D -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de -------------------------------- From vxwexplo-errs@csg.lbl.gov Thu Jul 5 04:03:13 2001 From: Vxworks Exploder Date: Thu Jul 5 04:03:15 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 5 04:03:05 PDT 2001 Subject: Re: FTP SERVER Subject: Re: problem with timers Subject: Help creating VxWorks Image. Subject: Re: Help creating VxWorks Image. Subject: target resident versus windSh versions of loadModule Subject: Serial Port initialisation Subject: Re: IP address of VxWorks Subject: Re: target resident versus windSh versions of loadModule Subject: 860 clock date time forever....... Subject: Re: target resident versus windSh versions of loadModule Subject: Do you know howto make library? Subject: Good Project Chance Subject: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Subject: ioctl(Q) Subject: Re: interface between asm and C language Subject: Re: signal handling in vxworks. ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: FTP SERVER Date: Wed, 4 Jul 2001 13:28:28 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994246591.276689@newsmaster-04.atnet.at> References: <3B42CB6F.60697C5D@NOSPAM.bcs.org.uk> <994239172.231953@newsmaster-04.atnet.at> Hi, I think you can get it from your FAE, if you do not have DosFS 2.0 regards werner "Johan Borkhuis" wrote in message news:Xns90D475C36F646borkhuisagerecom@135.7.153.101... > "Werner Schiendl" wrote: > > This server is not in the standard distribution, it is part of DosFS2 > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem with timers Date: Wed, 04 Jul 2001 10:20:37 +0100 From: Simon Farnsworth Message-ID: <3B42DFE5.6030704@startrekmail.com> References: <9htcku$jao$1@overload.lbl.gov> Anuradha Pani wrote: > --0-1957747793-994193458=:28046 > Content-Type: text/plain; charset=us-ascii > Content-Disposition: inline > attached is a simple piece of code where i have tried Could you please resend the code as plain text; the attachment is unviewable in many newsreaders. > --0-1957747793-994193458=:28046 > Content-Transfer-Encoding: base64 > Content-Disposition: attachment; filename="timerImp.c" > > LyogRmlsZTogdGltZXJJbXAuYyAtIENyZWF0ZXMgYW5kIHNldHMgYSB0aW1l > --0-1957747793-994193458=:28046-- > --------------------------- Newsgroups: comp.os.vxworks Subject: Help creating VxWorks Image. Date: 4 Jul 2001 07:17:24 -0700 From: sanchez.garcia@sener.es (Carlos Sanchez) Organization: http://groups.google.com/ Message-ID: <47c1e53c.0107040617.4a609cce@posting.google.com> Hi guys. My enviroment Tornado 2, VxW 5.4 HW (Target) PPC405 Walnut board AMD79C97x Ethernet Controller (uses ln79x driver) As newbie in VxWorks I find this problem hard to fix, but I sure that for you it will be very easy. My board has an AMD79C97X Ethernet Controlled (embedded into it) This Ethernet controller uses a different driver (ln79x). When creating a new VxWorks Image how can I tell VxWorks to use this driver when this driver is not selectable using the Project Facilty Tool? I guess that editig config.h? Any help, advise? NOTE: To download applications to the Target I´m using serial connection. Thanks. - -CSG- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help creating VxWorks Image. Date: Wed, 4 Jul 2001 17:02:11 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994259415.233571@newsmaster-04.atnet.at> References: <47c1e53c.0107040617.4a609cce@posting.google.com> Hi, If your BSP vendor just 'forgot' to add the driver to the CDF files (which are responsible for what is selectable in the project facility) your easiest workaround is to edit your BSP's config.h file. Lookup your documentation what macro(s) to define. Usually there is a macro like INCLUDE_LN79X that controls inclusion of the driver. Maybe there are additional parameters that can be/must be set. hth werner "Carlos Sanchez" wrote in message news:47c1e53c.0107040617.4a609cce@posting.google.com... > Hi guys. > > My enviroment > > Tornado 2, VxW 5.4 > HW (Target) PPC405 Walnut board > AMD79C97x Ethernet Controller (uses ln79x driver) > > As newbie in VxWorks I find this problem hard to fix, but I sure that > for you it will be very easy. > > My board has an AMD79C97X Ethernet Controlled (embedded into it) > This Ethernet controller uses a different driver (ln79x). > When creating a new VxWorks Image how can I tell VxWorks to use this > driver when this driver is not selectable using the Project Facilty > Tool? > I guess that editig config.h? > Any help, advise? > > NOTE: To download applications to the Target I´m using serial > connection. > > Thanks. > -CSG- --------------------------- Newsgroups: comp.os.vxworks Subject: target resident versus windSh versions of loadModule Date: Wed, 4 Jul 2001 17:38:21 +0100 From: "R Grommet" Organization: Defence Evaluation & Research Agency Message-ID: <9hvchh$jrp$1@trog.dera.gov.uk> Hi An easy question, I hope. I'm have an MVME5100 PPC board with 64MB of memory. I start my application using the following script (called foobar.scr) : ld 1,0,"/foo/bar.o" sp hello If I run the script from the target resident shell I get a whole bunch of linker relocation error messages. I presume its because the executable is being put too high up in memory for PPC short call addressing to work. However, if I run the script from the windSh interface as follows: Hi, I read a lot of message about init, but not about what I need: parity: none mode: full duplex data: 8 bits start bit 1 stop bit 1 baud rate 9600 I write: rs232_fd = open(DEVICE, O_RDWR | O_NOCTTY,0777); status = ioctl(rs232_fd,FIOBAUDRATE,9600); status = ioctl(rs232_fd,FIOFLUSH,0); status = ioctl(rs232_fd,FIOSETOPTIONS,OPT_RAW); miss somthing? or something wrong? in the open, what's the meaning of O_NOCTTY what's the meaning of the 777? what's the meaning of OPT_RAW? thanks a lot Patrick Boulay --------------------------- Newsgroups: comp.os.vxworks Subject: Re: IP address of VxWorks Date: Wed, 04 Jul 2001 09:09:54 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B433FD2.5F407BD2@covad.net> References: <1d411918.0107032131.640f74f7@posting.google.com> Reply-To: drdiags@covad.net DragonSpring, What about using the "other=" field? I believe that the vxWorks_rom images can have the network setup if the "other=" field contains the name of its network interface. This is mentioned somewhere in the documentation. HTH DragonSpring wrote: > > Our VxWorks image is indepencent of bootROM. We burn it into FLASH > directly use flash programmer, not load it via bootROM. > > We can set an IP address for the system in DEFAULT_BOOT_LINE. > > I find only when "network components --> basic network initialization > components --> bootline processing components network initializaton > --> initialize network at boot time" is included, the VxWorks will > initialize the network device specified in DEFAULT_BOOT_LINE and > ipAttach() and usrNetIfConfig() to it. This means that, if I don't > include "initialize network at boot time", none ipAttach() and > usrNetIfConfig() will happen, the target system will not be endrowed > with an IP address acturally. We cann't "ping" it successfully. At > this time, if we want to make the IP stack up, we have to invoke > ipAttach() and usrNetIfConfig() manually. > > Is my understanding right? > > Best Regards, > > lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: target resident versus windSh versions of loadModule Date: Wed, 04 Jul 2001 09:26:52 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B4343CC.F7AE0B9D@covad.net> References: <9hvchh$jrp$1@trog.dera.gov.uk> Reply-To: drdiags@covad.net If this is the T2/vxWorks 5.4 version, the PPC Toolchain allows you to compile with the -longcall (-mlongcall) (?)flag, which will work for you. Look on the vxWorks FAQ site to read more on this (if you were not already aware of this option). HTH http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ page) R Grommet wrote: > > Hi > > An easy question, I hope. I'm have an MVME5100 PPC board with 64MB of > memory. I start my application using the following script (called > foobar.scr) : > > ld 1,0,"/foo/bar.o" > sp hello > > If I run the script from the target resident shell I get a whole bunch of > linker relocation error messages. I presume its because the executable is > being put too high up in memory for PPC short call addressing to work. > However, if I run the script from the windSh interface as follows: > > > then it works without any errors. My question is, how do I get the target > resident shell to load my program in the same manner that windSh does it? I > know that I can hide the top 32MB of memory from VxWorks and get it to work, > but I need to store data in the full 64MB. > > TIA --------------------------- Newsgroups: comp.os.vxworks Subject: 860 clock date time forever....... Date: 4 Jul 2001 09:30:28 -0700 From: leonardmatherson@hotmail.com (leonard matherson) Organization: http://groups.google.com/ Message-ID: Hello everyone, I would like to keep the RTC clock register running even when the chip is powered off. I have enabled the RTE flag of the RTCSC register, and the RTC increments nicely. When the chip is powered off the RTC remains at the Powered off time. I have an MDP8xx pro evaluation board, and the battery power is being fed into the JP7 Battery Backup Jumper. How do I keep the time / date value updated when the main power is off. We use vxworks 5.4 and Tornado 2.0 Thanks, Leonard. LeonardMatherson@hotmail.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: target resident versus windSh versions of loadModule Date: Wed, 4 Jul 2001 21:15:09 +0100 From: "R Grommet" Organization: Defence Evaluation & Research Agency Message-ID: <9hvp83$nfj$1@trog.dera.gov.uk> References: <9hvchh$jrp$1@trog.dera.gov.uk> <3B4343CC.F7AE0B9D@covad.net> I did know about -mlongcall, but I thought there might be performance issues associated with it (after all, why would 26-bit addressing exist if it wasn't faster?) Your link was also very useful. I didn't think of hiding the top 32M and calling memAddToPool(sysMemTop(),0x2000000) to allocate the rest. Thanks for your prompt reply. It is much appreciated! "DrDiags" wrote in message news:3B4343CC.F7AE0B9D@covad.net... > If this is the T2/vxWorks 5.4 version, the PPC Toolchain allows > you to compile with the -longcall (-mlongcall) (?)flag, which will work > for you. Look on the vxWorks FAQ site to read more on this (if you > were not already aware of this option). HTH > > http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ page) > --------------------------- Newsgroups: comp.os.vxworks Subject: Do you know howto make library? Date: 4 Jul 2001 23:13:09 -0700 From: eaichul@ntl30.com (Eaichul Choung) Organization: http://groups.google.com/ Message-ID: <3da3205f.0107042213.7d9e54ab@posting.google.com> Hi guru! I'd like to make a library using arppc, but I don't have any reference Makefile. If someone have it,let me know please. I am trying to make *.a using Makefile. Thanks in advance. Regards, --------------------------- Newsgroups: comp.os.vxworks Subject: Good Project Chance Date: 4 Jul 2001 23:35:06 -0700 From: eaichul@ntl30.com (Eaichul Choung) Organization: http://groups.google.com/ Message-ID: <3da3205f.0107042235.192e00a6@posting.google.com> I developed Switch Board and ported vxWorks on it. I am finding someone or company can port Wind Web Server. The project time is 1 month. I am in Seoul,Korea. If you have interests, Call or email me. eaichul eaichul@ntl30.com +82-19-272-7942 --------------------------- Newsgroups: comp.os.vxworks Subject: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Date: 05 Jul 2001 06:56:26 UT From: fscheurer@manz-acs.de Organization: Lawrence Berkeley National Laboratory Message-ID: <9i145t$65d$1@overload.lbl.gov> Hi VxWorks-Gurus, I ran into the problem, that the integrated target-headers seem to be not confor with GNU expectations. This means e.g.: 1.) CPU is not defined as expected by the target-headers .. ./../gcc-2.95.3-5/gcc/config/types/vxCpu.h:200: #error CPU is not defined correctly .. 2.) vxTypes.h redefines and has other parse errors .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:71: warning: redefinition of `size_t' .. /usr/include/time.h:47: warning: `__cdecl__' attribute directive ignored /usr/include/time.h:47: parse error before `clock' .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:91: warning: `fpos_t' previously declared here .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:76: previous declaration of `wchar_t' .. ./../gcc-2.95.3-5/gcc/config/types/vxTypes.h:66: previous declaration of `ptrdiff_t' .. /usr/include/sys/select.h:28: parse error before `fd_set' .. MY QUESTION: Where is the most central place in - - the sources, - - the configure-script or - - the Makefile to "repair" it AND to get the right types to be compatible to objects build with older gcc-version? Any hint is gratefully accepted! MfG / Kind Regards Friedrich Scheurer, R&D - -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de - -------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: ioctl(Q) Date: Thu, 5 Jul 2001 08:37:16 +0000 (UTC) From: essenlee@hanmail.net (ms) Organization: Mailgate.ORG Server - http://www.Mailgate.ORG Message-ID: <20010705173409.HM.40000000002GDTn@www5.hanmail.net> In I/O Function, I want to know why ioctl() is needed and what it is used for. I hope anyone who know the answers gives me those.. - - thank you for read.. ================================================== ¿ì¸® ÀÎÅͳÝ, Daum Æò»ý ¾²´Â ¹«·á E-mail ÁÖ¼Ò ÇѸÞÀÏ³Ý Áö±¸ÃÌ ÇÑ±Û °Ë»ö¼­ºñ½º Daum°Ë»ö http://www.daum.net - -- Posted from [210.108.207.90] by way of [211.233.29.57] via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- Newsgroups: comp.os.vxworks Subject: Re: interface between asm and C language Date: 5 Jul 2001 02:24:24 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107050124.213b2900@posting.google.com> References: <1d411918.0107031800.18b3244e@posting.google.com> <994238106.754754@newsmaster-04.atnet.at> I use VxWorks 5.4 for PPC, the CPU is MPC860. I just want to know the asm file how to pass parameters to the C routin in MPC860? What is the use of r3? "Werner Schiendl" wrote in message news:<994238106.754754@newsmaster-04.atnet.at>... > Maybe this depends on the architecture in question > > For x86 C calling convention is used, parameters are put on the stack from > right to left (the last argument is put on the stack first). The stack is > cleaned up by the caller, your asm routine may just RETurn. > > I'd suggest using data types for passing between ASM and C where the > alignment is obvious (like 4 byte int and pointers in x86 architecture). > > For details I'd recommend the GCC documentation. > > regards > werner > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: signal handling in vxworks. Date: Thu, 5 Jul 2001 17:46:12 +0800 From: "Ren Hantao" Organization: Lucent Technologies, Columbus, Ohio Message-ID: <9i1d0b$i1c@nntpa.cb.lucent.com> References: <9hsuiv$bdd$1@overload.lbl.gov> The codes runs OK on my computer(Simulator) Ren Hantao "Anuradha Pani" wrote in message news:9hsuiv$bdd$1@overload.lbl.gov... > hi all, > > i am having some problem with catching signals. I have > attached the following piece of code. > > I am calling Start to spawn the signal handler task, > signalHandler and then sending a signal to that task, > but nothing happens. I don't get any output from the > sigHandler function. > > Please tell me what I'm doing wrong. > > Thanks > Anuradha --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jul 5 05:13:42 2001 From: "Jayanth" Date: Thu Jul 5 05:13:44 PDT 2001 Subject: RELATING TO OPEN CALL IN VXWORKS Hi, 1 )I have Installed Tronado in C drive and want's to open the file from another drive.... which i'm unable to do .... So i request U to get me out of this Issue..... 2 ) In Unix we got .cshrc file So what is that equalent in VxWorks 3 ) Where can i found a programming examples in VxWorks Other then http://www.rt.db.erau.edu/experiments/vx/toc/TableOfContents.html Yeah More In next mail........... Thx and Reply as early as possible Regs Jayanth From vxwexplo-errs@csg.lbl.gov Thu Jul 5 07:15:47 2001 From: "David Laight" Date: Thu Jul 5 07:15:49 PDT 2001 Subject: Re: How can I know the socket peer closed? > But when I shutdown the remote host suddenly or pull > out the network cable, the "while" loop will not terminate. > How can I know that the TCP connect has lost in this case? > why setsockopt(..., SO_KEEPALIVE,...) is no use? You are probably not waiting long enough! The TCP protocol stack was designed (lobbed together?) many years ago and has a few problems which are not exhibited by more recent protocols (eg ISO class 4 transport). 'Keepalive' messages are not part of the TCP protocol, I believe that they were not felt necessary by the guys that wrote the original protocol spec. There ommision removes the network traffic caused by idle connections. In some senarions the idle connection traffic can become a significant fraction of the network bandwidth (anyone seen a large corporate LAN with a lot of systems running LAN Mangler or Novell Netware and the shere number of 'I am here' broadcasts generated?). In any case SO_KEEPALIVE generates a keepalive packet after 2 hours idle! It is reconned that if there is no data to send, and no unacked data, it doesn't matter whether the connection is detected as being dead. Unfortunately no one allowed for 'denial of service' events... In some systems it is possible to hack the keepalive timer to a shorter value. (a few minutes) however this is not recommended and I don't know where you would have to hack the vxworks stack. David From vxwexplo-errs@csg.lbl.gov Thu Jul 5 07:15:49 2001 From: "David Laight" Date: Thu Jul 5 07:15:57 PDT 2001 Subject: Re : Assembler with VxWorks > void outWrds(void) > { > unsigned short port = 0x378; > unsigned char value = 0xff; > __asm__ volatile ("outb %1,%0;" : : "d" (port) , "a" (value) ); > } It is MUCH easier (and safer) to write the code in a .s file! For the above: .code .globl _outWrds _outWrds: mov dx,0x378 mov al,0xff out dx,al ret will (probably) suffice. (check in (say) romInit.s for the syntax!) If you include xxx.o in the MACH_EXTRA for the bsp build, the vxWorks makefile will assemble xxx.s for you. No more difficult than adding a C file. I don't know how good gcc is at asm functions. But in my experience compilers and optimsers don't behave very well with asm functions. Even one that had a compilcated 'feature' to inline asm functions! Years ago we did the following: 1) compile unoptimised to a .s file 2) process the .s file to inline some functions - eg: jsr outb was replaced by: pop dx pop ax out dx,al 3) push the resultant file through the peephole optimiser this collapsed the push xxx, push yyy, pop dx, pop ax sequence 4) assemble This isn't possible with most modern compilers as they won't optimise an assembler file. David From vxwexplo-errs@csg.lbl.gov Thu Jul 5 07:15:54 2001 From: "David Laight" Date: Thu Jul 5 07:15:59 PDT 2001 Subject: Re: VxWorks Scheduling > Can a low priority task can re-acquire a semaphore released after taskLock? What do we EXPECT to happen - run through the senarios: First, what is the behaviour of semGive, two (obvious) choices: 1) The state of the sema4 is marked 'free' and the first task waiting for sema4 is marked 'ready'. When it runs it will re-attempt to acquire the sema4. 2) The sema4 is marked as owned by the first task that is waiting, and that task is marked 'ready'. When it runs it already owns the sema4. Now if the second is true, the semTake will block (in spite of the taskLock) So assume it does the former. What about semTake? What does it check first, the state of the sema4 or the list of 'waiters'. If it checks the list of 'waiters' the current task will be queued behind the other task - so will be blocked (SEM_Q_PRIORITY and SEM_INVERSION_SAVE will give the tasks equal priority...) Only if semTake checks the sema4 state first will the low priority task be able to re-acquire the sema4. I don't (off hand) know what the vxworks sema4 stuff does. It is fairly tricky code to get right, and changed between 5.3.1. and 5.4. I have a gut feeling that semGive does schedule the waiting task - giving it the sema4, but might be wrong. Some careful testing will show what happens on a given system, but I suspect the high priority task is likely to get the sema4. It might depend on the type of sema4, and might change between vxWorks releases. On the memory front, the vxWorks memPartAlignedAlloc is documented as using a 'first fit' algorithm. This leads to serious memory fragmentation. Indeed if you were trying to fragment memory this would be a good approach! It is possible to implement a 'best fit' algorithm using the same free list structure. This is infinitely better, the shorter free list may even mean that the allocations are faster (less dross to scan past). It is certainly no less 'deterministic' and could be implemented by the same code. Maybe memPartOptionsSet() could be enhanced to select the mode? David From vxwexplo-errs@csg.lbl.gov Thu Jul 5 07:15:54 2001 From: "David Laight" Date: Thu Jul 5 07:16:00 PDT 2001 Subject: Re: align malloc and free in VxWorks > Is it safe and correct when I free the memory malloced by > commCacheDmaMallocAlign() like this? > > void *p; > p = commonCacheDmaMallocAlign(16, 1000); > ... > cacheDmaFree(p); > > Will the memory between "memory1" and "memory2" be lost? Not a chance - this is guaranteed to at least generate an error message from the memPartAlloc code, it might generate a fault in the task freeing the memory. If really unlucky your system might freeze solid (accessing a random memory location). At alternative is to allocate the memory area you need uncached from outside the area used by the vxWorks kernel. Two options: 1) The end of physical memory - by setting USER_RESERVED_MEM. But beware at least one WRS BSP (the StrongARM one) uses an area allocated there for an intermediate page table - then frees an equivalent space from the other end of the area (ie not the same address if you reserve more space). 2) Low memory below MEM_LOW_ADDR. This has the advantage that you can grope physical memory and still have it contiguous. In both cases you can get the standard system page tables set to give this memory the required properties. The fixed addresses make diagnostics easier! David From vxwexplo-errs@csg.lbl.gov Thu Jul 5 09:02:26 2001 From: Sankar.Jagannathan@infineon.com Date: Thu Jul 5 09:02:29 PDT 2001 Subject: Binary Image and BOOTROM Hi, I am working on tornado II and snds100 borad. I want to know the following clarifications. 1) How to create a plain binary file (executable which runs on target board when it is placed at specific memory base address ) 2) How to vxworks bootroms for target board Please clarify and do the needful. Regards, Jagannathan(Infineon, US) From vxwexplo-errs@csg.lbl.gov Thu Jul 5 12:24:39 2001 From: "CICON,JAMES (HP-Vancouver,ex1)" Date: Thu Jul 5 12:24:41 PDT 2001 Subject: WindView not showing Context Switches I need some help with WindView. It seems to be working part way, but the Task Context switches are missing. It shows tWdbTask as the only executing task. * The timestamp information is correct. * An accurate list of all tasks are listed, but none of these tasks are ever shown as executing. * Semgives and Semtakes (and other object status) are shown as all occurring in the tWdbTask context. I am using Tornado 1.0.1 and VxWorks 5.3.1 Jim Cicon Hewlett Packard Company From vxwexplo-errs@csg.lbl.gov Thu Jul 5 20:45:18 2001 From: "Jun Wu" Date: Thu Jul 5 20:45:20 PDT 2001 Subject: VxWorks net task data abort Hi vxWorks experts I encounter a strange problem: My application have a download task which download data from memory to flash. If the download task's priority ( 100 ) is lower than that of vxWorks system task tNetTask ( 50 ), in the downloading procedure, tNetTask will throw a data abort exception, Data abort Exception address: 0x00064c60 Current process status register: 0xc0000013 Task: 0x1f0a270 "tNetTask" If the download task's priority ( 40 ) is higher than that of vxWorks system task tNetTask, exception will never occur! PS: 1. Exception address: 0x00064c60 seems to be owned by a functin called ip_slowtimo 2. the download task never call any vxWorks net function. How can troubleshoot this problem? BR Jun From vxwexplo-errs@csg.lbl.gov Fri Jul 6 04:03:22 2001 From: Vxworks Exploder Date: Fri Jul 6 04:03:24 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 6 04:03:05 PDT 2001 Subject: RELATING TO OPEN CALL IN VXWORKS Subject: Unable to receive multicast packets on a raw socket in VxWorks. Subject: How to do inline assembly in C BSP ??? Subject: Re: How can I know the socket peer closed? Subject: Database for VxWorks Subject: Re: align malloc and free in VxWorks Subject: Re: align malloc and free in VxWorks Subject: re: How can I know the socket peer closed? Subject: Re : Assembler with VxWorks Subject: re: VxWorks Scheduling Subject: re: align malloc and free in VxWorks Subject: Re: Assembler with VxWorks Subject: Re: nest mpc 8260 interrupt? Subject: Binary Image and BOOTROM Subject: ATA Harddidsk Error Subject: Re: Bringing up the ARM Integrator/CM7TDMI with vxWorks Subject: Re: RELATING TO OPEN CALL IN VXWORKS Subject: Re: standalone_vxworks with no network Subject: PPP fail to re-connect from NT to vxWorks Subject: Accessing Host files from VxSim? Subject: WindView not showing Context Switches Subject: How to get patches Subject: un-numbered interfaces Subject: Re: How to do inline assembly in C BSP ??? Subject: Re: How can I know the socket peer closed? Subject: Re: Defragmentation for TrueFFS Subject: Re: Binary Image and BOOTROM Subject: Re: How can I know the socket peer closed? Subject: Re: seperate interrupt stack Subject: Re: PPP fail to re-connect from NT to vxWorks Subject: VxWorks net task data abort Subject: Re: Bringing up the ARM Integrator/CM7TDMI with vxWorks Subject: Re: How to get patches Subject: Re: How to get patches Subject: Re: ATA Harddidsk Error Subject: Re: align malloc and free in VxWorks Subject: How old is VxWorks ? Subject: Re: How to do inline assembly in C BSP ??? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: RELATING TO OPEN CALL IN VXWORKS Date: Thu, 5 Jul 2001 17:25:09 +0530 From: "Jayanth" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i1loj$shr$1@overload.lbl.gov> Hi, 1 )I have Installed Tronado in C drive and want's to open the file from another drive.... which i'm unable to do .... So i request U to get me out of this Issue..... 2 ) In Unix we got .cshrc file So what is that equalent in VxWorks 3 ) Where can i found a programming examples in VxWorks Other then http://www.rt.db.erau.edu/experiments/vx/toc/TableOfContents.html Yeah More In next mail........... Thx and Reply as early as possible Regs Jayanth --------------------------- Newsgroups: comp.os.vxworks Subject: Unable to receive multicast packets on a raw socket in VxWorks. Date: 5 Jul 2001 05:52:36 -0700 From: kavleenk@delhi.tcs.co.in (Kavleen Bhatia) Organization: http://groups.google.com/ Message-ID: <3c2606e4.0107050452.230478a5@posting.google.com> I need to send & receive raw ethernet packets off the network. I am able to send to the multicast group but am facing problem in joining the multicast group for receiving the packet. The "setsockopt" function does not work if the interface does not have a valid IP address. I need to know how to attach my Interface(by specifying only the MAC address) to the multicast group, to receive multicast packets. I would like to know if somebody has successfully done the same --------------------------- Newsgroups: comp.os.vxworks Subject: How to do inline assembly in C BSP ??? Date: Thu, 05 Jul 2001 08:54:34 -0400 From: "Granum, Allen [CAR:CF81:EXCH]" Organization: Nortel Networks Message-ID: <3B44638A.1E60DC64@americasm01.nt.com> I'm trying to add some inline assembly into a C function at the BSP level. I have not had much success so far. Here is what I've tried (forget about the functionallity of the assembly. I'm just trying to illustrate what works and what doesn't): This code works..... void myFunct( BOOL flag ) { /* blah, blah, blah... */ /* Useful C code goes here */ /*-------------------------------------------------------------*/ __asm__ ("sync"); /*-------------------------------------------------------------*/ /* blah, blah, blah... */ /* More useful C code goes here */ } however, if the asm statement reads as it does in the example below, I get the error: /var/tmp/ccQyaa5E.s: Assembler messages: /var/tmp/ccQyaa5E.s:4463: Error: unresolved expression that must be resolved /var/tmp/ccQyaa5E.s:4463: Error: unresolved expression that must be resolved *** Error code 1 void myFunct( BOOL flag ) { /* blah, blah, blah... */ /* Useful C code goes here */ /*-------------------------------------------------------------*/ __asm__ ("ori r5,r5,0x01"); /*-------------------------------------------------------------*/ /* blah, blah, blah... */ /* More useful C code goes here */ } It doesn't seem to like the 'r5' value. PS: I've tried placing a '%' in front of the 'r5' so the statement reads: __asm__ ("ori %r5,%r5,0x01"). This gives the error: /var/tmp/ccqZaygG.s: Assembler messages: /var/tmp/ccqZaygG.s:4463: Error: Bad expression /var/tmp/ccqZaygG.s:4463: Error: syntax error; found `r' but expected `,' /var/tmp/ccqZaygG.s:4463: Error: junk at end of line: `r5,%r5,0x01' *** Error code 1 Can anyone share some advice? Thanks - -- Allen Granum Nortel Networks Wireless Internet BSP Development Team Email: agranum@nortelnetworks.com Phone: (613)763-4324 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I know the socket peer closed? Date: 5 Jul 2001 06:10:01 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107050510.6daeace0@posting.google.com> References: <1d411918.0107032349.71a4864f@posting.google.com> I now know that the setsockopt(..., SO_KEEPALIVE, ...) acturally takes affect. But the interval before the are detected is too long. But I don't want to modify the value of TCPTV_KEEPALIVE, because it is global,base on VxWorks kernel. (1) can I modify TCPTV_KEEPALIVE base on socket? How to? (2)Is there any other method to detect peer crashes or network failure (i.e. the TCP connect is not valid anymore)? --------------------------- Newsgroups: comp.os.vxworks Subject: Database for VxWorks Date: Thu, 5 Jul 2001 15:21:35 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: <8ip1i9.vlc.ln@gum.de> Hi there! Does someone know about a simple to use, fast and reliable database for VxWorks (using the TFFS) ? (Might be a commercial product!) Thanks, Frank Wolf --------------------------- Newsgroups: comp.os.vxworks Subject: Re: align malloc and free in VxWorks Date: 5 Jul 2001 06:40:12 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107050540.7e34e87a@posting.google.com> References: <9hsnht$8rk$1@overload.lbl.gov> <3B41EC87.A1B96983@ks.ericsson.se> This piece of source code is from Galileo Technology's SDK for Galnet II ethernet switch chip. The BD of the DMA controller in the chip need a block of buffer which must be cache-coherent and must align at long word (16 bytes) bound. So it first use cacheDmaMalloc() to alloc a block of memory which is cache-coherent, then align it. It cannot use memalign here. Is there any other good method to achieve this aim? Please give piece of source code. Regards, Urban Lindberg wrote in message news:<3B41EC87.A1B96983@ks.ericsson.se>... > Hi > > Why do you try to reinvent the wheel? Have a look at memalign! Btw, you > can only free a address you have received my a memory allocation, so > your example might crash or maybe even reboot. You don't say what you > need it for, but if its for DMA, use cacheDmaMalloc. If you really need > to do what the example is intended to, you need to save the values > returned from cacheDmaMalloc and if more than one task can call your > functions you need to add some mutex and maybe a check that Align is a > power of two. > > Urban > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: align malloc and free in VxWorks Date: 5 Jul 2001 06:45:30 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107050545.1b047039@posting.google.com> References: <9hsnht$8rk$1@overload.lbl.gov> <3B41EC87.A1B96983@ks.ericsson.se> Yes, we really use it for DMA and the DMA requires the buffer to be aligned. "your example might crash or maybe even reboot." Are you sure? Why should we add some mutex? I think it is unnecessary. Urban Lindberg wrote in message news:<3B41EC87.A1B96983@ks.ericsson.se>... > Hi > > Why do you try to reinvent the wheel? Have a look at memalign! Btw, you > can only free a address you have received my a memory allocation, so > your example might crash or maybe even reboot. You don't say what you > need it for, but if its for DMA, use cacheDmaMalloc. If you really need > to do what the example is intended to, you need to save the values > returned from cacheDmaMalloc and if more than one task can call your > functions you need to add some mutex and maybe a check that Align is a > power of two. > > Urban --------------------------- Newsgroups: comp.os.vxworks Subject: re: How can I know the socket peer closed? Date: Thu, 5 Jul 2001 14:48:37 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i1spn$4gn$1@overload.lbl.gov> > But when I shutdown the remote host suddenly or pull > out the network cable, the "while" loop will not terminate. > How can I know that the TCP connect has lost in this case? > why setsockopt(..., SO_KEEPALIVE,...) is no use? You are probably not waiting long enough! The TCP protocol stack was designed (lobbed together?) many years ago and has a few problems which are not exhibited by more recent protocols (eg ISO class 4 transport). 'Keepalive' messages are not part of the TCP protocol, I believe that they were not felt necessary by the guys that wrote the original protocol spec. There ommision removes the network traffic caused by idle connections. In some senarions the idle connection traffic can become a significant fraction of the network bandwidth (anyone seen a large corporate LAN with a lot of systems running LAN Mangler or Novell Netware and the shere number of 'I am here' broadcasts generated?). In any case SO_KEEPALIVE generates a keepalive packet after 2 hours idle! It is reconned that if there is no data to send, and no unacked data, it doesn't matter whether the connection is detected as being dead. Unfortunately no one allowed for 'denial of service' events... In some systems it is possible to hack the keepalive timer to a shorter value. (a few minutes) however this is not recommended and I don't know where you would have to hack the vxworks stack. David --------------------------- Newsgroups: comp.os.vxworks Subject: Re : Assembler with VxWorks Date: Thu, 5 Jul 2001 15:05:57 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i1sqg$4go$1@overload.lbl.gov> > void outWrds(void) > { > unsigned short port = 0x378; > unsigned char value = 0xff; > __asm__ volatile ("outb %1,%0;" : : "d" (port) , "a" (value) ); > } It is MUCH easier (and safer) to write the code in a .s file! For the above: .code .globl _outWrds _outWrds: mov dx,0x378 mov al,0xff out dx,al ret will (probably) suffice. (check in (say) romInit.s for the syntax!) If you include xxx.o in the MACH_EXTRA for the bsp build, the vxWorks makefile will assemble xxx.s for you. No more difficult than adding a C file. I don't know how good gcc is at asm functions. But in my experience compilers and optimsers don't behave very well with asm functions. Even one that had a compilcated 'feature' to inline asm functions! Years ago we did the following: 1) compile unoptimised to a .s file 2) process the .s file to inline some functions - eg: jsr outb was replaced by: pop dx pop ax out dx,al 3) push the resultant file through the peephole optimiser this collapsed the push xxx, push yyy, pop dx, pop ax sequence 4) assemble This isn't possible with most modern compilers as they won't optimise an assembler file. David --------------------------- Newsgroups: comp.os.vxworks Subject: re: VxWorks Scheduling Date: Thu, 5 Jul 2001 14:30:31 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i1sqq$4gp$1@overload.lbl.gov> > Can a low priority task can re-acquire a semaphore released after taskLock? What do we EXPECT to happen - run through the senarios: First, what is the behaviour of semGive, two (obvious) choices: 1) The state of the sema4 is marked 'free' and the first task waiting for sema4 is marked 'ready'. When it runs it will re-attempt to acquire the sema4. 2) The sema4 is marked as owned by the first task that is waiting, and that task is marked 'ready'. When it runs it already owns the sema4. Now if the second is true, the semTake will block (in spite of the taskLock) So assume it does the former. What about semTake? What does it check first, the state of the sema4 or the list of 'waiters'. If it checks the list of 'waiters' the current task will be queued behind the other task - so will be blocked (SEM_Q_PRIORITY and SEM_INVERSION_SAVE will give the tasks equal priority...) Only if semTake checks the sema4 state first will the low priority task be able to re-acquire the sema4. I don't (off hand) know what the vxworks sema4 stuff does. It is fairly tricky code to get right, and changed between 5.3.1. and 5.4. I have a gut feeling that semGive does schedule the waiting task - giving it the sema4, but might be wrong. Some careful testing will show what happens on a given system, but I suspect the high priority task is likely to get the sema4. It might depend on the type of sema4, and might change between vxWorks releases. On the memory front, the vxWorks memPartAlignedAlloc is documented as using a 'first fit' algorithm. This leads to serious memory fragmentation. Indeed if you were trying to fragment memory this would be a good approach! It is possible to implement a 'best fit' algorithm using the same free list structure. This is infinitely better, the shorter free list may even mean that the allocations are faster (less dross to scan past). It is certainly no less 'deterministic' and could be implemented by the same code. Maybe memPartOptionsSet() could be enhanced to select the mode? David --------------------------- Newsgroups: comp.os.vxworks Subject: re: align malloc and free in VxWorks Date: Thu, 5 Jul 2001 15:19:42 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i1srf$4gq$1@overload.lbl.gov> > Is it safe and correct when I free the memory malloced by > commCacheDmaMallocAlign() like this? > > void *p; > p = commonCacheDmaMallocAlign(16, 1000); > ... > cacheDmaFree(p); > > Will the memory between "memory1" and "memory2" be lost? Not a chance - this is guaranteed to at least generate an error message from the memPartAlloc code, it might generate a fault in the task freeing the memory. If really unlucky your system might freeze solid (accessing a random memory location). At alternative is to allocate the memory area you need uncached from outside the area used by the vxWorks kernel. Two options: 1) The end of physical memory - by setting USER_RESERVED_MEM. But beware at least one WRS BSP (the StrongARM one) uses an area allocated there for an intermediate page table - then frees an equivalent space from the other end of the area (ie not the same address if you reserve more space). 2) Low memory below MEM_LOW_ADDR. This has the advantage that you can grope physical memory and still have it contiguous. In both cases you can get the standard system page tables set to give this memory the required properties. The fixed addresses make diagnostics easier! David --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Assembler with VxWorks Date: Thu, 05 Jul 2001 17:59:34 +0200 From: Frank Marquardt Organization: Draeger Medizintechnik GmbH Message-ID: <3B448EE5.EABB0C14@draeger.com> References: <9hb1mi$8g2$1@panther.uwo.ca> Hi , we use Tornado 2.0 with GNU GCC. Try these lines: __asm__ (" mtspr 562,4 /* assembler lines */ " : /* output */ : /* input */ : "4" /* modified register */ ); Looking for __asm__ in GNU TollKit User's Guide DOC-12912-ZD-00 printed by Wind River. Bye Harman Bassan schrieb: > Hi everyone, > Can some one give me some idea on How can i do inline assembly in vxWorks > with GCC. I looked at the GCC toolkit but things didn't look that clear to > me. For example , the following lines which are valid with Visual C++, How > can they be conerted to run with GCC on VxWorks. > > _asm > { > mov dx,378h > mov al,0ffh > out dx,al > } > > Thanks in advance, > Harman. - -- Mit freundlichen Grüßen ##################################################### # Frank Marquardt # # Draeger Medizintechnik GmbH # # mt-i-euk, KSt. 5610 # # Moislinger Allee 53/55 # # 23542 Luebeck # # Tel.: 0451/882-1810 Fax: -3733 # # E-Mail: Frank.Marquardt@draeger.com # ##################################################### --------------------------- Newsgroups: comp.os.vxworks Subject: Re: nest mpc 8260 interrupt? Date: Thu, 05 Jul 2001 18:04:30 +0200 From: Frank Marquardt Organization: Draeger Medizintechnik GmbH Message-ID: <3B44900E.F95B099D@draeger.com> References: Hi , I do not know the mpc 8260, but we use the mpc 850. There we can unlock the interrupts by setting the EE bit in the MSR register. But you have to save the critical registers before. Bye u687591552@spawnkill.ip-mobilphone.net schrieb: > I was told that for mpc 8260, an interrupt can not preempt another > interrupt, if one ISR starts, it goes until finished. I am not sure > if that's true at least I can not find support in the MPC8260 > manual. If anyone has such experience can tell me something about > it. > > > -- > Sent by guangbutun from hotmail included in com > This is a spam protected message. Please answer with reference header. > Posted via http://www.usenet-replayer.com/cgi/content/new - -- Mit freundlichen Grüßen ##################################################### # Frank Marquardt # # Draeger Medizintechnik GmbH # # mt-i-euk, KSt. 5610 # # Moislinger Allee 53/55 # # 23542 Luebeck # # Tel.: 0451/882-1810 Fax: -3733 # # E-Mail: Frank.Marquardt@draeger.com # ##################################################### --------------------------- Newsgroups: comp.os.vxworks Subject: Binary Image and BOOTROM Date: Thu, 5 Jul 2001 08:59:42 -0700 From: Sankar.Jagannathan@infineon.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9i23qg$7m3$1@overload.lbl.gov> Hi, I am working on tornado II and snds100 borad. I want to know the following clarifications. 1) How to create a plain binary file (executable which runs on target board when it is placed at specific memory base address ) 2) How to vxworks bootroms for target board Please clarify and do the needful. Regards, Jagannathan(Infineon, US) --------------------------- Newsgroups: comp.os.vxworks Subject: ATA Harddidsk Error Date: Thu, 5 Jul 2001 09:35:04 -0700 From: "Shesha Bhushan" Organization: Arizona State University Message-ID: <9i250j$n9t$1@news.asu.edu> Hi Gurus, When i include ATA Hard Drive under hard disks on VxWorks tab in Tornado Environment, and i try to create a bootable image then i get the following error. Can u help me in debugging this error. Your help is highly valuable for the success of our project Thanks Shesha B Sreenivasamurthy dataSegPad.o partialImage.o ctdt.o symTbl.o -o vxWorks partialImage.o: In function `usrAtaInit': partialImage.o(.text+0x12f46): undefined reference to `ataResources' partialImage.o(.text+0x12f4a): undefined reference to `ataResources' partialImage.o: In function `ataDrv': partialImage.o(.text+0x2e73a): undefined reference to `ataResources' partialImage.o(.text+0x2e73e): undefined reference to `ataResources' partialImage.o(.text+0x2e994): undefined reference to `sysIntEnablePIC' partialImage.o(.text+0x2ea3a): undefined reference to `ataTypes' partialImage.o(.text+0x2ea3e): undefined reference to `ataTypes' partialImage.o: In function `ataDevCreate': partialImage.o(.text+0x2f336): undefined reference to `ataTypes' partialImage.o(.text+0x2f33a): undefined reference to `ataTypes' partialImage.o: In function `ataRawio': partialImage.o(.text+0x2f6de): undefined reference to `ataTypes' partialImage.o(.text+0x2f6e2): more undefined references to `ataTypes' follow partialImage.o: In function `ataIntr': partialImage.o(.text+0x30190): undefined reference to `sysInByte' partialImage.o: In function `ataWait': partialImage.o(.text+0x302e0): undefined reference to `sysInByte' partialImage.o(.text+0x30320): undefined reference to `sysInByte' partialImage.o(.text+0x30394): undefined reference to `sysInByte' partialImage.o(.text+0x303c4): undefined reference to `sysInByte' partialImage.o(.text+0x303f4): more undefined references to `sysInByte' follow partialImage.o: In function `ataInit': partialImage.o(.text+0x30490): undefined reference to `sysOutByte' partialImage.o(.text+0x304ac): undefined reference to `sysDelay' partialImage.o(.text+0x304d0): undefined reference to `sysOutByte' partialImage.o(.text+0x304ec): undefined reference to `sysDelay' partialImage.o(.text+0x30548): undefined reference to `sysInByte' partialImage.o(.text+0x30588): undefined reference to `sysOutByte' partialImage.o(.text+0x305a4): undefined reference to `sysDelay' partialImage.o(.text+0x305c8): undefined reference to `sysInByte' partialImage.o: In function `ataCmd': partialImage.o(.text+0x30712): undefined reference to `ataTypes' partialImage.o(.text+0x30716): undefined reference to `ataTypes' partialImage.o(.text+0x307c0): undefined reference to `sysOutByte' partialImage.o(.text+0x307d8): undefined reference to `sysOutByte' partialImage.o(.text+0x307f4): undefined reference to `sysOutByte' partialImage.o(.text+0x30808): undefined reference to `sysOutByte' partialImage.o(.text+0x3084c): undefined reference to `sysOutByte' partialImage.o(.text+0x30874): more undefined references to `sysOutByte' follow partialImage.o: In function `ataPread': partialImage.o(.text+0x30bb0): undefined reference to `sysInWordStringRev' partialImage.o: In function `ataRW': partialImage.o(.text+0x30c96): undefined reference to `ataTypes' partialImage.o(.text+0x30c9a): undefined reference to `ataTypes' partialImage.o(.text+0x30ce4): undefined reference to `sysOutByte' partialImage.o(.text+0x30cf4): undefined reference to `sysOutByte' partialImage.o(.text+0x30d04): undefined reference to `sysOutByte' partialImage.o(.text+0x30d14): undefined reference to `sysOutByte' partialImage.o(.text+0x30d2c): undefined reference to `sysOutByte' partialImage.o(.text+0x30d78): more undefined references to `sysOutByte' follow partialImage.o: In function `ataRW': partialImage.o(.text+0x30ecc): undefined reference to `sysOutWordString' partialImage.o(.text+0x30eec): undefined reference to `sysOutLongString' partialImage.o(.text+0x30f9c): undefined reference to `sysOutByte' partialImage.o(.text+0x30fb0): undefined reference to `sysOutByte' partialImage.o(.text+0x310a4): undefined reference to `sysInWordString' partialImage.o(.text+0x310c4): undefined reference to `sysInLongString' make: *** [vxWorks] Error 0x1 --------------------------- Newsgroups: comp.os.vxworks,comp.sys.arm Subject: Re: Bringing up the ARM Integrator/CM7TDMI with vxWorks Date: 5 Jul 2001 09:37:59 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107050837.2f4af9a4@posting.google.com> References: Hello, > I have an ARM Integrator/CM7TDMI AP development motherboard , the core > microprocessor being ARM7TDMI . I need to bring this board up with > vxWorks . > > I installed T2 with the BSP pid7t . Built bootrom.hex ( making > appropriate changes to get M32 S-Record format )& flashed it . I > failed to see anything on the hyper terminal . > > Am I using the right BSP ? If pid7t is the correct BSP , are there any > changes I need to make to bring up this board ? No, you are starting from the wrong BSP, and perhaps the wrong version of Tornado 2. You should try to get the latest release of Tornado 2 for ARM (I believe that 2.0.1 is FCS and 2.1 is in beta - speak to your local Wind River sales guys and ask about those, especially 2.1 since that should have the BSP for the Integrator + 7TDMI in it). The pid7t BSP is definitely not the place to start for an Integrator board... HTH, John... > Any help is greatly appreciated . > > Thanx in advance, > Suman --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RELATING TO OPEN CALL IN VXWORKS Date: Thu, 05 Jul 2001 13:32:57 -0500 From: Luc Message-ID: <3B44B2D9.E666108F@qwest.net> References: <9i1loj$shr$1@overload.lbl.gov> You made one small mistake : you believed the salesperson! Tornado2 for NT has so many bugs they'll never fix it. Then they will tell you, you need vxWorksAE Java bugs included! Jayanth wrote: > Hi, > > 1 )I have Installed Tronado in C drive and want's to open the file from > another drive.... which i'm unable to do .... > > So i request U to get me out of this Issue..... > > 2 ) In Unix we got .cshrc file So what is that equalent in VxWorks > > 3 ) Where can i found a programming examples in VxWorks Other then > http://www.rt.db.erau.edu/experiments/vx/toc/TableOfContents.html > > Yeah More In next mail........... > > Thx and Reply as early as possible > > Regs > Jayanth --------------------------- Newsgroups: comp.os.vxworks Subject: Re: standalone_vxworks with no network Date: Thu, 05 Jul 2001 13:42:43 -0500 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <3B44B523.7A1C18B7@aps.anl.gov> References: Kelly Hornsby wrote: > > I need some ideas of something that I can't figure out. I'm using T2 on a > Sandpoint 750 system. I have a DEC 21143 NIC, and it works just fine with > the bootloader/ ftp-Vxworks. However, I've created the > standalone_vxworks.hex file, and flashed it into memory, but the network is > not initialized now. the NIC is detected fine (pciHeaderShow indicates > this) and the 21143 PCI initilization routine is executed (sysNetHwInit2), > but there's no dev out there, and the device shows no devices for ifShow. > > I'm looking for suggestions or ideas for why this occurs. Any ideas? I can't explain *why* WRS decided not to initialize the network in a standalone target, but I can tell you how to get it to change that: You need to add #define STANDALONE_NET to your configAll.h file and rebuild. I don't know how you do that if you're using the Tornado project facility to build vxWorks, but I expect there's some way to do it. - - Andrew - -- The world is such a cheerful place when viewed from upside-down It makes a rise of every fall, a smile of every frown --------------------------- Newsgroups: comp.os.vxworks Subject: PPP fail to re-connect from NT to vxWorks Date: Thu, 5 Jul 2001 15:29:13 -0400 From: "Jason Ke" Organization: Nortel Message-ID: <9i2f2r$70$1@bcarh8ab.ca.nortel.com> Hi, I am trying to establish the PPP connection to the vxWorks from my NT4.0 SP6 workstation. I have read the windriver WTN-50 w.r.t. the PPP with Win NT. I have performed all the steps. But I have following problem. 0) start the usrPPPinit on vxWorks (code included below) 1) initiate the PPP from the NT to vxWorks .. the connection establish properly ... verified by ping and other apps .... 2) disconnect it with "hangup" from NT side with the Dial-up network monitor ... now according to WTN50 page 16 first paragraph. I should be able to re-dial to the vxWorks and get connect again. But I was not able to. I always get error 651 (the same as on WTN50 section 7.5) Note: 1) the ppp tasks still alive on the vxWorks (since I have not called ppDelete) - -> i NAME ENTRY TID PRI STATUS PC SP ERRNO DELAY - ---------- ------------ -------- --- ---------- -------- -------- ------- -- - --- tExcTask excTask ee9018 0 PEND 19f4cc ee8f38 0 0 tLogTask logTask ee6690 0 PEND 19f4cc ee65b8 0 0 tWdbTask wdbCmdLoop e90b48 3 READY 16d934 e909e0 0 0 tNetTask netTask eb3a38 50 READY 16ddd4 eb3968 0 0 tPPP1 ppp_task eff2c0 55 PEND 155f64 eff0e0 0 0 tPPP1Wrt ppp_wrt_task e89028 60 PEND 16beb4 e88f60 0 0 2) if I call the pppDelete on vxWorks, and restart it over again with usrPPPinit and I will be able to connect from NT smoothly again. Question: 1) Am I supposed to call pppDelete, then usrPPPinit, for NT to re-connect? 2) From vxWorks side, how do I know NT side has disconnected 3) Is there any way for NT reconnect without having to do anything on vxWorks side. ============================================== Here is my code to initiate the ppp from vxWorks STATUS startPPP () PPP_OPTIONS pppOpt; /* PPP configuration options */ PPP_INFO pppInfo; STATUS ret; memset (&pppOpt, 0, sizeof(PPP_OPTIONS)); pppOpt.flags = OPT_NO_PAP | OPT_NO_CHAP | OPT_NO_VJ | OPT_PASSIVE_MODE | OPT_DRIVER_DEBUG | OPT_DEBUG | OPT_SILENT_MODE; ret = pppInit (1, "/tyCo/0","90.0.0.50", "90.0.0.60", 19200, &pppOpt, NULL); if (ret == ERROR) return ERROR; if ((pppInfoGet (1, &pppInfo) == OK) && (pppInfo.ipcp_fsm.state == OPENED)) return (OK); /* link established */ else return (ERROR); /* link down */ } --------------------------- Newsgroups: comp.os.vxworks Subject: Accessing Host files from VxSim? Date: 5 Jul 2001 12:39:48 -0700 From: prenom_nomus@yahoo.com (Mark) Organization: http://groups.google.com/ Message-ID: Is there a way to open and read a file from the c: drive from a VxSim program? 1) I tried using the full pathname "c:\thedir\myfile.dat" 2) I also tried in windshell cd c:\thedir ls - myfile.dat yes file is there fp = open("myfile.dat", 0, 0); neither worked thanks Mark --------------------------- Newsgroups: comp.os.vxworks Subject: WindView not showing Context Switches Date: Thu, 5 Jul 2001 15:24:28 -0400 From: "CICON,JAMES (HP-Vancouver,ex1)" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i2hsc$50m$1@overload.lbl.gov> I need some help with WindView. It seems to be working part way, but the Task Context switches are missing. It shows tWdbTask as the only executing task. * The timestamp information is correct. * An accurate list of all tasks are listed, but none of these tasks are ever shown as executing. * Semgives and Semtakes (and other object status) are shown as all occurring in the tWdbTask context. I am using Tornado 1.0.1 and VxWorks 5.3.1 Jim Cicon Hewlett Packard Company --------------------------- Newsgroups: comp.os.vxworks Subject: How to get patches Date: 5 Jul 2001 13:37:19 -0700 From: joewoodbury@mail.com (Joe Woodbury) Organization: http://groups.google.com/ Message-ID: I see suggestions on this board to check Wind River Systems Windsurf site for SPR#xxxxx, which is fixed by Cumulative Patch x. I am a legitimate customer, have a customer ID and access to WindSurf but I am unable to find either of these things. What am I doing wrong? Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: un-numbered interfaces Date: 5 Jul 2001 14:31:43 -0700 From: du_bin@yahoo.com (billy du) Organization: http://groups.google.com/ Message-ID: <340587ac.0107051331.6dc4232c@posting.google.com> Does anyone know whether VxWorks IP stack(SENS) support un-numbered interfaces? It means we can attach an interface to MUX without the interface IP address. Also we can add a route without the next hop IP address? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to do inline assembly in C BSP ??? Date: 5 Jul 2001 15:09:37 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107051409.3a7caabe@posting.google.com> References: <3B44638A.1E60DC64@americasm01.nt.com> Hello, I don't understand why the sync example did not work, although it could be the white space in the line (some of these parsers can be a little particular about spaces). Here's an example from a PPC BSP (looks like PPC code that you are writing ;-): __asm__(" sync"); Now, the register case is much simpler to fix. There are no defined register names for the assembler, so simply use the register number: __asm__(" ori 5,5,0x01"); That should get you what you wanted. In all the examples that I have, the open parenthesis is not space separated from the __asm__ and there is at least one space after the opening quote. I don't know if either of those are actually needed to make it work. Also, in cases where there is more than one instruction, they are separated like this example: __asm__(" eieio; sync"); HTH, John... "Granum, Allen [CAR:CF81:EXCH]" wrote in message news:<3B44638A.1E60DC64@americasm01.nt.com>... > I'm trying to add some inline > assembly into a C function at > the BSP level. > I have not had much success so > far. > > Here is what I've tried > (forget about the > functionallity of the > assembly. I'm just trying to > illustrate what works and what > doesn't): > > > This code works..... > > void myFunct( BOOL flag ) > { > /* blah, blah, blah... */ > /* Useful C code goes here > */ > > /*-------------------------------------------------------------*/ > > __asm__ ("sync"); > /*-------------------------------------------------------------*/ > > /* blah, blah, blah... */ > /* More useful C code goes > here */ > } > > however, if the asm statement > reads as it does in the > example below, I get the > error: > > /var/tmp/ccQyaa5E.s: Assembler > messages: > /var/tmp/ccQyaa5E.s:4463: > Error: unresolved expression > that must be resolved > /var/tmp/ccQyaa5E.s:4463: > Error: unresolved expression > that must be resolved > *** Error code 1 > > > void myFunct( BOOL flag ) > { > /* blah, blah, blah... */ > /* Useful C code goes here > */ > > /*-------------------------------------------------------------*/ > > __asm__ ("ori r5,r5,0x01"); > /*-------------------------------------------------------------*/ > > /* blah, blah, blah... */ > /* More useful C code goes > here */ > } > > > It doesn't seem to like the > 'r5' value. > > PS: I've tried placing a '%' > in front of the 'r5' so the > statement reads: __asm__ ("ori > %r5,%r5,0x01"). > This gives the error: > > /var/tmp/ccqZaygG.s: Assembler > messages: > /var/tmp/ccqZaygG.s:4463: > Error: Bad expression > /var/tmp/ccqZaygG.s:4463: > Error: syntax error; found `r' > but expected `,' > /var/tmp/ccqZaygG.s:4463: > Error: junk at end of line: > `r5,%r5,0x01' > *** Error code 1 > > > Can anyone share some advice? > > Thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I know the socket peer closed? Date: 5 Jul 2001 15:12:42 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107051412.3f768b71@posting.google.com> References: <1d411918.0107032349.71a4864f@posting.google.com> <1d411918.0107050510.6daeace0@posting.google.com> Hello, > (2)Is there any other method to detect peer crashes or network failure > (i.e. the TCP connect is not valid anymore)? You could always generate your own "keep-alive" message at whatever frequency you like (as long as you control both ends of the link, or there is something you can send that will be treated as a NOP by the remote end). Just a thought, John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Thu, 5 Jul 2001 16:40:17 -0700 From: "P.Eichler" Organization: Evolve Products, Inc. Message-ID: <9i2tol$r84$1@nntp2-cm.news.eni.net> References: <9hqng4$49e$1@nntp1-cm.news.eni.net> Reply-To: "P.Eichler" Thank you for your input. I've looked into the provided TFFS option (which is not standard in the VxWorks version of TrueFFS). Unfortunately it only seems to deal with "recollecting" and "packing" sectors at the low flash level. It does NOT seem to deal with de-fragmenting at the DOS file level, which is really what I need to improve the overall speed of writing and deleting files on a very fragmented system. "Andray Kaganovsky" wrote in message news:Xns90D3C43FA4E11andreykhomecom@24.2.9.58... > "News Account" wrote in news:9hqng4$49e$1@nntp1- > cm.news.eni.net: > > > Does anybody know of any methodology or utility out there for defragmenting > > a VxWorks DOS system that is stored and maintained via TrueFFS? Also I have > > looked at the M-Systems TrueFFS code and noticed some sort of low level > > defragmentation option that I believe is supposed to be accessable either > > through the "tffsRawio" routine or some sort of "ioctl" call, however I was > > unable to get this to work at all. At this point we would even be willing to > > utilize a third party software to perform the defragmentation task. Please > > help!!! > > > > Regards, > > P. Eichler > > > [AndrayK]. To defragment TrueFFS flash array, you can use code fragment > similar to the one below: > > int freeSectors = 0x70000000; /* impossible number of sectors */ > > (void) tffsRawio (0, TFFS_DEFRAGMENT_VOLUME, freeSectors, 0, 0); > > The call above could return error code, but, as far as I know, it is safe > to ignore it. If media is badly fragmented, the call above can take very > long time to complete, since it is doing complete defragmentation of the > entire flash array. > > Andray > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Binary Image and BOOTROM Date: 5 Jul 2001 17:09:02 -0700 From: eaichul@ntl30.com (Eaichul Choung) Organization: http://groups.google.com/ Message-ID: <3da3205f.0107051609.cf2b2c5@posting.google.com> References: <9i23qg$7m3$1@overload.lbl.gov> Sankar.Jagannathan@infineon.com wrote in message news:<9i23qg$7m3$1@overload.lbl.gov>... > Hi, > I am working on tornado II and snds100 borad. I want to know the following clarifications. > > 1) How to create a plain binary file (executable which runs on target board when it is placed at specific memory base address ) When you use Makefile provided by WindRiver, It produce binary file("vxWorks" file is binary).Binary format is different from the target to the target. > > 2) How to vxworks bootroms for target board > Try , make bootrom.hex and burn it.As you know, 'bootrom' is also binary.:) > Please clarify and do the needful. > Regards, > Jagannathan(Infineon, US) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I know the socket peer closed? Date: 5 Jul 2001 19:09:26 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107051809.6bae786a@posting.google.com> References: <1d411918.0107032349.71a4864f@posting.google.com> <1d411918.0107050510.6daeace0@posting.google.com> <488e459a.0107051412.3f768b71@posting.google.com> I don't want to bother and affect the peer application and the peer application is not written by me. I cann't control the behavior of the peer. I just want to solve it at TCP level, not application level. How to do? Any piece of source code? john_94501@yahoo.com (John) wrote in message news:<488e459a.0107051412.3f768b71@posting.google.com>... > Hello, > > > (2)Is there any other method to detect peer crashes or network failure > > (i.e. the TCP connect is not valid anymore)? > > You could always generate your own "keep-alive" message at whatever > frequency you like (as long as you control both ends of the link, or > there is something you can send that will be treated as a NOP by the > remote end). > > Just a thought, > > John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: seperate interrupt stack Date: 5 Jul 2001 19:19:42 -0700 From: lihuapeng@21cn.com (DragonSpring) Organization: http://groups.google.com/ Message-ID: <1d411918.0107051819.29247b61@posting.google.com> References: <1d411918.0107031739.18a3d596@posting.google.com> Sorry, should be *separate* not *seperate*. lihuapeng@21cn.com (DragonSpring) wrote in message news:<1d411918.0107031739.18a3d596@posting.google.com>... > We use VxWorks 5.4 for PPC and MPC860. > > (1)Does VxWorks use a seperate interrupt stack in this case? > (2)If does, how to set the seperate stack size? > (3)The benefit of seperate interrupt stack. I think the VxWorks kernel > should also do task context switch(save and restore) in case of > seperate interrupt stack. > (4)The hardware(MPC860) support to seperate interrupt stack. > > Regards, > > lihuapeng@21cn.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPP fail to re-connect from NT to vxWorks Date: Thu, 05 Jul 2001 20:38:04 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B45329C.B25B91C0@covad.net> References: <9i2f2r$70$1@bcarh8ab.ca.nortel.com> Reply-To: drdiags@covad.net Jason, Jason Ke wrote: > <....SNIP.....> > 2) From vxWorks side, how do I know NT side has disconnected I found this in the latest release of the VxWorks Network 5.4 Programmers Guide Edition 2 pg 37: ======================================================================= The return value of pppDelete( ) does not indicate the status of the PPP link. To check whether the link is actually terminated, call pppInfoGet( ) and make sure the return value is ERROR. The following code fragment demonstrates the usage of this mechanism for PPP unit 4: PPP_INFO pppInfo; if (pppInfoGet (4, &pppInfo) == ERROR) return (OK); /* link terminated */ else return (ERROR); /* link still up */ <....SNIP....> - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks net task data abort Date: Fri, 6 Jul 2001 11:43:54 +0800 From: "Jun Wu" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i3e0c$hob$1@overload.lbl.gov> Hi vxWorks experts I encounter a strange problem: My application have a download task which download data from memory to flash. If the download task's priority ( 100 ) is lower than that of vxWorks system task tNetTask ( 50 ), in the downloading procedure, tNetTask will throw a data abort exception, Data abort Exception address: 0x00064c60 Current process status register: 0xc0000013 Task: 0x1f0a270 "tNetTask" If the download task's priority ( 40 ) is higher than that of vxWorks system task tNetTask, exception will never occur! PS: 1. Exception address: 0x00064c60 seems to be owned by a functin called ip_slowtimo 2. the download task never call any vxWorks net function. How can troubleshoot this problem? BR Jun --------------------------- Newsgroups: comp.os.vxworks,comp.sys.arm Subject: Re: Bringing up the ARM Integrator/CM7TDMI with vxWorks Date: 5 Jul 2001 21:42:19 -0700 From: suman@mistralsolutions.com (Suman Anil) Organization: http://groups.google.com/ Message-ID: References: <488e459a.0107050837.2f4af9a4@posting.google.com> john_94501@yahoo.com (John) wrote in message news:<488e459a.0107050837.2f4af9a4@posting.google.com>... > Hello, > > > I have an ARM Integrator/CM7TDMI AP development motherboard , the core > > microprocessor being ARM7TDMI . I need to bring this board up with > > vxWorks . > > > > I installed T2 with the BSP pid7t . Built bootrom.hex ( making > > appropriate changes to get M32 S-Record format )& flashed it . I > > failed to see anything on the hyper terminal . > > > > Am I using the right BSP ? If pid7t is the correct BSP , are there any > > changes I need to make to bring up this board ? > > No, you are starting from the wrong BSP, and perhaps the wrong version > of Tornado 2. You should try to get the latest release of Tornado 2 > for ARM (I believe that 2.0.1 is FCS and 2.1 is in beta - speak to > your local Wind River sales guys and ask about those, especially 2.1 > since that should have the BSP for the Integrator + 7TDMI in it). The > pid7t BSP is definitely not the place to start for an Integrator > board... > > HTH, > > John... > > > Any help is greatly appreciated . > > > > Thanx in advance, > > Suman Hi John , Thank for your response . Tornado2.1 for ARM , FCS will be out by mid July . I need to wait till then . Thanx once again, Suman --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Fri, 6 Jul 2001 08:24:16 +0200 From: "Matthias Guelck" Organization: Heidelberg Digital Message-ID: <9i3lif$4pg$1@news.hdpp.de> References: Hello Joe, on http://www.wrs.com/corporate/support/private/validate.html you might see the reason: either you purchased your Wind River product more than 1 year ago or your support contract is no longer valid (like mine). Kind regards Matthias "Joe Woodbury" wrote in message news:c8c13723.0107051237.4700018@posting.google.com... > I see suggestions on this board to check Wind River Systems Windsurf > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > I am a legitimate customer, have a customer ID and access to WindSurf > but I am unable to find either of these things. What am I doing wrong? > > Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Fri, 6 Jul 2001 08:34:04 +0200 From: "Matthias Guelck" Organization: Heidelberg Digital Message-ID: <9i3m4s$4r8$1@news.hdpp.de> References: <9i3lif$4pg$1@news.hdpp.de> Sorry, the only question is: do you have a valid support contract? Kind regards Matthias "Matthias Guelck" wrote in message news:9i3lif$4pg$1@news.hdpp.de... > Hello Joe, > > on http://www.wrs.com/corporate/support/private/validate.html you might see the reason: > either you purchased your Wind River product more than 1 year ago or your support contract is no longer valid (like mine). > > Kind regards > > Matthias > > "Joe Woodbury" wrote in message news:c8c13723.0107051237.4700018@posting.google.com... > > I see suggestions on this board to check Wind River Systems Windsurf > > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > > > I am a legitimate customer, have a customer ID and access to WindSurf > > but I am unable to find either of these things. What am I doing wrong? > > > > Joe Woodbury > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ATA Harddidsk Error Date: Fri, 06 Jul 2001 09:01:19 +0200 From: Emmanuel Herbreteau Organization: SEPRO Robotique (BEP) Message-ID: <3B45623F.1DFFBF69@sepro-robotique.com> References: <9i250j$n9t$1@news.asu.edu> Shesha Bhushan wrote: Hi, > partialImage.o: In function `usrAtaInit': > partialImage.o(.text+0x12f46): undefined reference to `ataResources' > partialImage.o: In function `ataDevCreate': > partialImage.o(.text+0x2f336): undefined reference to `ataTypes' ataResources and ataTypes have to be defined in "sysLib.c". This is a sample from MBX821's BSP from WRS : #ifdef INCLUDE_ATA /* global data structures used by the EIDE/ATA driver */ /* * The first two entries are populated in sysAtaInit() for the * MBX860 PCI ATA Controller. The third entry is setup for a * PCMCIA Ata Device. * */ ATA_RESOURCE ataResources[ATA_MAX_CTRLS] = { { { 5, 0, {0, 0}, {0, 0}, 0, 0, 0, 0, 0, 0 }, IDE_LOCAL, 0, 0, 0, 0, 0, 0, 0, 0 }, /* PCI Ata Controller 0 */ { { 5, 0, {0, 0}, {0, 0}, 0, 0, 0, 0, 0, 0 }, IDE_LOCAL, 0, 0, 0, 0, 0, 0, 0, 0 }, /* PCI Ata Controller 1 */ #ifdef INCLUDE_PCMCIA { { 5, 0, {ATA_PCMCIA_IO_START0, ATA_PCMCIA_IO_START1}, {ATA_PCMCIA_IO_STOP0, ATA_PCMCIA_IO_STOP1}, 0, 0, 0, 0, 0, 0 }, ATA_PCMCIA, 1, (int) ATA_PCMCIA_INT_VEC,(int) ATA_PCMCIA_INT_LVL, ATA_PCMCIA_CONFIG, ATA_PCMCIA_SEM_TIMEOUT, ATA_PCMCIA_WDG_TIMEOUT, 0, 0 } /* PCMCIA - ctrl 2 */ #endif /* INCLUDE_PCMCIA */ }; ATA_TYPE ataTypes [ATA_MAX_CTRLS][ATA_MAX_DRIVES] = { { { ATA_DEV0_CYL, 0, 0, 512, 0xff}, /* controller 0, drive 0 */ { ATA_DEV1_CYL, 0, 0, 512, 0xff}, /* controller 0, drive 1 */ }, { { ATA_DEV2_CYL, 0, 0, 512, 0xff}, /* controller 1, drive 0 */ { ATA_DEV3_CYL, 0, 0, 512, 0xff}, /* controller 1, drive 1 */ }, { { ATA_DEV4_CYL, 0, 0, 512, 0xff}, /* controller 0, drive 0 */ { ATA_DEV5_CYL, 0, 0, 512, 0xff}, /* controller 0, drive 1 */ } }; #endif /* INCLUDE_ATA */ > partialImage.o: In function `ataInit': > partialImage.o(.text+0x30490): undefined reference to `sysOutByte' > partialImage.o(.text+0x304ac): undefined reference to `sysDelay' > partialImage.o(.text+0x30548): undefined reference to `sysInByte' The "sysXXXXX" function must be in your BSP. The SysDelay is a wait loop, located in 'sysLib.c', which wait for 1 millisecond. The sysOutByte, sysInByte, etc... are defined in 'sysALib.s'. It's small assembly routines from your BSP. In our BSP (PPC assembly), it looks like : sysInByte: /* Read byte from address */ lbzx r3,r0,r3 /* Sync I/O operation */ eieio /* Return to caller */ bclr 20,0 > Can u help me in debugging this error. Your help is highly valuable > for the success of our project Don't forget it's a public newsgroup where people are trying to help other people for free... If you want me to send you those files (sysLib.c and sysALib.s), send me an email (it's too big for this forum). - -- Regards Emmanuel (SEPRO Robotique, France) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: align malloc and free in VxWorks Date: Fri, 06 Jul 2001 09:53:36 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B456E80.708ADA6C@ks.ericsson.se> References: <9hsnht$8rk$1@overload.lbl.gov> <3B41EC87.A1B96983@ks.ericsson.se> <1d411918.0107050545.1b047039@posting.google.com> Hi, We also use DMA and we use memalign for the buffers and it works as it should (sorry, I was a bit in error regarding cacheDmaMalloc). I tried calling cacheDmaMalloc (on a PPC604 VI-board) from the shell and it seems to return memory align on 2^12, so it should work without any extra alignment. Yes, I am very sure it will crash or reboot. The allocator functions actually allocate some extra bytes for accounting of the memory. The allocator function return the address following the accounting information. The deallocator takes the address given to it and subtacts the size of the accounting information and if you don't return exactly the address you recieved from tha allocator the bookkeeping will go wrong and will crash or reboot your system. The accounting information is often one pointer and a long int for the size of the block and the pointer is followed by the deallocator and depending of what kind of allocation you made it could access a NULL pointer (crash), uninitialized pointer (possibly modifying the stack or even code) which might reboot the system put in an infinite loop or even correct a bug in your program (I have actually seen all of them happen). HTH Urban DragonSpring wrote: > > Yes, we really use it for DMA and the DMA requires the buffer to be aligned. > > "your example might crash or maybe even reboot." Are you sure? > > Why should we add some mutex? I think it is unnecessary. > [snip] --------------------------- Newsgroups: comp.os.vxworks Subject: How old is VxWorks ? Date: Fri, 6 Jul 2001 11:03:33 +0200 From: "Jonathan ILIAS" Organization: University of Angers, France. Message-ID: <9i3urd$2lg4$1@news.univ-angers.fr> Hi, Does anyone know what is VxWorks birth date ? I'm looking for this information on WindRiver web site but I could'nt find it. I'm still looking for it and I will post this information here if I find it, but if you already know that... Thanks, Jonathan ILIAS PS : sorry for the mistakes, I'm french and not very good in english writing... :,-( --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to do inline assembly in C BSP ??? Date: Thu, 5 Jul 2001 18:48:46 +0100 From: vxworksintent@tao-group.com (David Given) Organization: I do not speak for anyone but myself, and barely that. Message-ID: References: <3B44638A.1E60DC64@americasm01.nt.com> Sender: dg@pearl.tao.co.uk In article <3B44638A.1E60DC64@americasm01.nt.com>, "Granum, Allen [CAR:CF81:EXCH]" writes: > I'm trying to add some inline > assembly into a C function at > the BSP level. > I have not had much success so > far. > > Here is what I've tried > (forget about the > functionallity of the > assembly. I'm just trying to > illustrate what works and what > doesn't): [...] > __asm__ ("ori r5,r5,0x01"); gcc targets gas, which was designed specifically to assemble the output given by compilers, rather than humans. As a result the syntax can be... disturbing. You're using the PowerPC, right? This: int foo(int a, int b) { return a+b+42; } ...compiles into... foo: add 3, 3, 4 addi 3, 4, 42 blr (with optimisation). You may draw your own conclusions. Frankly, I'd recommend using as little inline assembly as possible. If you give input and output constraints to the __asm__ directive, you can give enough information to let the optimiser cope with your assembly; which means you can create a bunch of pseudo-functions to wrap each funny opcode, and then embed them in ordinary C, and with lots of optimisation turned on you'll get pretty good code without having to write any assembler. Look at the `Assembler with VxWorks' thread for a discussion of input and output constraints. If you really have to write big chunks of assembly, I'd strongly recommend putting it in a seperate .S file and defining constants to represent things like registers. - -- David Given dg@tao-group.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Fri Jul 6 05:08:37 2001 From: Vikrant Mittal Date: Fri Jul 6 05:08:40 PDT 2001 Subject: Problem in porting Solaris project to vxWorks Hi , We have a problem in porting an application from Solaris OS to VxWorks OS. The application is running fine on Solaris, however we are unable to compile the application on VxWorks. This application uses "pthread"s and we think that we are facing problems in compilation due to these "pthread"s only. Can any one of you please give us some insight as to how to get these incorporated on a VxWorks machine as well. Looking for more, Regards -- Vikrant Mittal Tata Consultancy Services 231, Udyog Vihar Phase I, Gurgaon - 122 016 ,Haryana (India) Tel (91-124) 6347270, 6347271, 6346935 (extn 233) _____________________________________________________ From vxwexplo-errs@csg.lbl.gov Fri Jul 6 08:23:50 2001 From: "Mike Anderson" Date: Fri Jul 6 08:23:52 PDT 2001 Subject: Integrating OSPF MIB into VxWorks SNMP V1/V2c VxWorks Greetings! I have the WRS SNMP & OSPF implementations and I'm trying to ge the OSPF MIB into the SNMP MIB on the target. I execute the: -> snmpdTreeAdd("1.3.6.1.2.1.14", &ospf_mib) value = 0 = 0x0 and get a success status. However, the OSPF MIB info doesn't show up when I walk the MIB tree. Anyone got any clues to this behavior that they'd care to share? TIA, Mike -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Fri Jul 6 08:46:27 2001 From: "Ben Kelly" Date: Fri Jul 6 08:46:30 PDT 2001 Subject: RE: Problem in porting Solaris project to vxWorks VxWorks uses tasks to accomplish threading. It supports parts of the POSIX definition but it does not support pthreads. You can buy a third part pthreads library for VxWorks. Look here: http://www.windriver.com/cgi-bin/corporate/tradeWinds/viewProd.cgi?id=203 Note that even if you get a pthreads library there are a number of features in the specification that are optional (ie nonstandard). Depending on how the code was originally written you may have some work to get these features to play nice on VxWorks. Enjoy. - Ben >> Hi , > We have a problem in porting an application from Solaris OS to > VxWorks OS. The application is running fine on Solaris, however we are > unable to compile the application on VxWorks. > > This application uses "pthread"s and we think that we are facing > problems in compilation due to these "pthread"s only. Can any one of you > > please give us some insight as to how to get these incorporated on a > VxWorks machine as well. From vxwexplo-errs@csg.lbl.gov Fri Jul 6 15:19:49 2001 From: s.eckardt@ixthos.com Date: Fri Jul 6 15:19:51 PDT 2001 Subject: Symbol Table Sync Problem I've been successfully running a VxWorks image/tgtsvr with symbol table synchronization enabled, on a networked Sun Solaris host. Upon moving to a non-networked Windows NT host, I consistently get the following error: Fatal WTX error (0x1012F), synchronization stopped. Any idea why? If it matters, my board is configured as a Slave on a VME Shared Memory network. As such, its Ethernet traffic goes through another board which maintains the physical Ethernet connection to the host. Suzie Eckardt Chief Software Engineer Ixthos, Inc. s.eckardt@ixthos.com From vxwexplo-errs@csg.lbl.gov Fri Jun 29 20:03:02 2001 From: "David Anderson" Date: Fri Jul 6 16:27:20 PDT 2001 Subject: FW: Network pool running out... -----Original Message----- From: David Anderson [SMTP:david.anderson@dtrack.com] Sent: 27 June 2001 16:05 To: 'Vxworks_Users (E-mail) Subject: Network pool running out... vxWorks 5.3.1 Tornado 1.0.1 Hello all, I've run into a problem which produces an apparent 'freeze up' of a PPP link, but still allows 'normal' Ethernet style TCP/IP connections to proceed. We are using the Tornado 2.0 ppp daemon. It appears that running an FTP session and a TELNET session concurrently over the PPP link causes the network pool to run out of buffers ( error 0x73000a ). However, the PPP link does not recover, the daemon tasks are running but you can no longer ping from either end to the other, or establish a new socket connection across it. The buffer error appears only to be a soft error - an Ethernet TCP/IP connection can still be set-up OK. Any ideas about why the PPP link should be 'frozen out' although all status indications suggest it is open and up. David Anderson. NEW E-MAIL ADDRESS: mailto:david.anderson@dtrack.com ------------------------------------- FF: 2B + ~2B , that is the question ? ------------------------------------- Disclaimer ---------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please destroy and notify Data Track Technology Plc +44 1425 271900. ------------------------------------------------------- From vxwexplo-errs@csg.lbl.gov Fri Jul 6 16:27:18 2001 Received: from postal1.lbl.gov (postal1.lbl.gov [128.3.7.82]) by csg.lbl.gov (8.9.3/8.9.3) with ESMTP id QAA18009 for ; Fri, 6 Jul 2001 16:27:18 -0700 (PDT) Received: from SpamWall.lbl.gov (localhost [127.0.0.1]) by postal1.lbl.gov (8.11.2/8.11.2) with ESMTP id f66NRHA20605 for ; Fri, 6 Jul 2001 16:27:17 -0700 (PDT) Received: from kilby.brackett.com (mail.brackett.com [209.183.211.3]) by SpamWall.lbl.gov (8.11.2/8.11.2) with ESMTP id f66NRG920601 for ; Fri, 6 Jul 2001 16:27:16 -0700 (PDT) Received: from defiant (sdsl-216-36-86-22.dsl.iad.megapath.net [216.36.86.22]) by kilby.brackett.com (8.9.3/8.9.3) with SMTP id TAA17129 for ; Fri, 6 Jul 2001 19:27:20 -0400 From: "Mike Anderson" To: "the vxWorks Users Group Exploder" Subject: RE: Symbol Table Sync Problem Date: Fri, 6 Jul 2001 19:26:24 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2911.0) X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Importance: Normal In-Reply-To: <200107062232.PAA17492@csg.lbl.gov> Hi Suzie, > Submitted-by: s.eckardt@ixthos.com > > I've been successfully running a VxWorks image/tgtsvr with symbol table > synchronization enabled, on a networked Sun Solaris host. Upon > moving to a > non-networked Windows NT host, I consistently get the following error: > Fatal WTX error (0x1012F), synchronization stopped. Any idea why? > > If it matters, my board is configured as a Slave on a VME Shared Memory > network. As such, its Ethernet traffic goes through another board which > maintains the physical Ethernet connection to the host. > The problem is that your Windows host isn't listening for the target agent connection on a socket. The way to fix it is to go to the options tab in Tornado and and set the Wind Registry up to use a network address (the address of the machine running the Registry server) instead of selecting the local registry entry. This allows the target agent to connect. HTH, Mike -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Fri Jul 6 18:26:09 2001 From: "MicroSat" Date: Fri Jul 6 18:26:11 PDT 2001 Subject: how to obtain vxworks software for demo? Hi, I am a student. I want to use vxworks but don't know where there is vxworks software for demo. Regards From vxwexplo-errs@csg.lbl.gov Sat Jul 7 04:03:29 2001 From: Vxworks Exploder Date: Sat Jul 7 04:03:31 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 7 04:03:05 PDT 2001 Subject: Problem in porting Solaris project to vxWorks Subject: Re: Problem in porting Solaris project to vxWorks Subject: Integrating OSPF MIB into VxWorks SNMP V1/V2c Subject: RE: Problem in porting Solaris project to vxWorks Subject: routing table doesn't show the entries with mask 255.255.255.255 Subject: Re: How old is VxWorks ? Subject: How is "..." used in the body of a VxWorks function? Subject: Re: How to get patches Subject: test Subject: Re: Accessing Host files from VxSim? Subject: Tornado II / VxWorks 5.4 / Linux ??? Subject: Trying to replace LXT970 by LXT971A HELP ! Subject: Can I put all files in a directory to remote FTP server one time? Subject: Re: How to get patches Subject: Re: PPP fail to re-connect from NT to vxWorks Subject: Re: Can I put all files in a directory to remote FTP server one time? Subject: Re: Tornado II / VxWorks 5.4 / Linux ??? Subject: Symbol Table Sync Problem Subject: Re: Symbol Table Sync Problem Subject: Re: How can I block VxWorks default shell (->)? Subject: FW: Network pool running out... Subject: Network pool running out... Subject: RE: Symbol Table Sync Problem Subject: Re: Symbol Table Sync Problem Subject: Check This Out!!! 1712 Subject: Re: Can I put all files in a directory to remote FTP server one time? Subject: Check This Out!!! 8134 Subject: ISR callable routines Subject: Re: Unable to receive multicast packets on a raw socket in VxWorks. Subject: how to obtain vxworks software for demo? Subject: Re: how to obtain vxworks software for demo? Subject: How to setup a PPP link through telephone connection. Subject: missing timer_connect() and friends ... ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Problem in porting Solaris project to vxWorks Date: Fri, 06 Jul 2001 17:39:28 +0530 From: Vikrant Mittal Organization: Lawrence Berkeley National Laboratory Message-ID: <9i4a4c$a4t$1@overload.lbl.gov> Hi , We have a problem in porting an application from Solaris OS to VxWorks OS. The application is running fine on Solaris, however we are unable to compile the application on VxWorks. This application uses "pthread"s and we think that we are facing problems in compilation due to these "pthread"s only. Can any one of you please give us some insight as to how to get these incorporated on a VxWorks machine as well. Looking for more, Regards - -- Vikrant Mittal Tata Consultancy Services 231, Udyog Vihar Phase I, Gurgaon - 122 016 ,Haryana (India) Tel (91-124) 6347270, 6347271, 6346935 (extn 233) _____________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem in porting Solaris project to vxWorks Date: 6 Jul 2001 13:10:22 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9i4a4c$a4t$1@overload.lbl.gov> Vikrant Mittal wrote: > This application uses "pthread"s and we think that we are facing > problems in compilation due to these "pthread"s only. Can any one of you > > please give us some insight as to how to get these incorporated on a > VxWorks machine as well. This might help: http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt5.html#5.5 Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Integrating OSPF MIB into VxWorks SNMP V1/V2c Date: Fri, 6 Jul 2001 11:22:53 -0400 From: "Mike Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i4o6c$fol$1@overload.lbl.gov> VxWorks Greetings! I have the WRS SNMP & OSPF implementations and I'm trying to ge the OSPF MIB into the SNMP MIB on the target. I execute the: - -> snmpdTreeAdd("1.3.6.1.2.1.14", &ospf_mib) value = 0 = 0x0 and get a success status. However, the OSPF MIB info doesn't show up when I walk the MIB tree. Anyone got any clues to this behavior that they'd care to share? TIA, Mike - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Problem in porting Solaris project to vxWorks Date: Fri, 6 Jul 2001 11:46:24 -0400 From: "Ben Kelly" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i4o6d$fom$1@overload.lbl.gov> VxWorks uses tasks to accomplish threading. It supports parts of the POSIX definition but it does not support pthreads. You can buy a third part pthreads library for VxWorks. Look here: http://www.windriver.com/cgi-bin/corporate/tradeWinds/viewProd.cgi?id=203 Note that even if you get a pthreads library there are a number of features in the specification that are optional (ie nonstandard). Depending on how the code was originally written you may have some work to get these features to play nice on VxWorks. Enjoy. - - Ben >> Hi , > We have a problem in porting an application from Solaris OS to > VxWorks OS. The application is running fine on Solaris, however we are > unable to compile the application on VxWorks. > > This application uses "pthread"s and we think that we are facing > problems in compilation due to these "pthread"s only. Can any one of you > > please give us some insight as to how to get these incorporated on a > VxWorks machine as well. --------------------------- Newsgroups: comp.os.vxworks Subject: routing table doesn't show the entries with mask 255.255.255.255 Date: 6 Jul 2001 10:39:12 -0700 From: bala@innocor.com (balendran bala) Organization: http://groups.google.com/ Message-ID: Hi vxWorks gurus, I am having a routing problem with my target (MC68360 QUICC) and I need your help. I need my target to learn the network with the mask of 255.255.255.255 via RIP V2. So when a RIP update comes from the remote target with the mask of 255.255.255.255, I expect that the entry should be included into my target routing table. However, the entry never included into my target routing table. Does anyone know why this doesn't work? Note that if I add the entry manually, it works!! Sample network setup: Ethernet---MY TARGET----[Frame Relay]-----remote target(third party)---{??} thank you Balendran --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How old is VxWorks ? Date: Fri, 06 Jul 2001 13:00:11 -0500 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <3B45FCAB.3FE95040@aps.anl.gov> References: <9i3urd$2lg4$1@news.univ-angers.fr> Jonathan ILIAS wrote: > > Does anyone know what is VxWorks birth date ? > I'm looking for this information on WindRiver web site but I could'nt find > it. I'm still looking for it and I will post this information here if I find > it, but if you already know that... This isn't definitive but the first modification history entry in the header file vxWorks.h from vxWorks 4.0.2 is dated 15Feb83 (yes, we still have that version online, although we're not using it!). Here is the whole mod. history from 4.0.2's vxWorks.h: modification history - -------------------- 01t,08apr89,dnw added ifdef to prevent inclusion of vxWorks.h more than once. 01s,22jun88,dnw moved READ, WRITE, and UPDATE back here from ioLib.h. 01r,22apr88,gae oops! forgot some #endif's in 01q. 01q,12apr88,gae removed QUICK & WAIT; added STD_{IN,OUT,ERR}. fixed #define's of FALSE, TRUE, etc. moved READ, WRITE, and UPDATE to ioLib.h. 01p,04dec87,dnw added undefine of MC68000 to get around Green Hills bug that pre-defines MC68000. 01o,12nov87,ecs added type ULONG. 01n,08feb86,dnw added types INSTR, UINT, USHORT. 01m,14oct85,rdc added BUS types. 01l,16jul85,jlf added conditional for NULL and EOF. 01k,24jun85,rdc installed condtional compile so we can include in assembly language files. See instructions below. Added System type macro and CPU type macro. 01j,13jun85,dnw cleaned-up, removed more obsolete stuff to wrs.h 01i,11sep84,jlf changed name from wrs.h to vxWorks.h. removed GLOBAL. 01h,03jun84,dnw removed IGNORE declaration. 01g,09apr84,jlf added MEMBER_SIZE macro. 01f,14dec83,dnw added MSB, LSB macros 01e,17nov83,jlf added STATUS type, for routines which return a status. 01d,13jul83,dnw added NELEMENTS macro 01c,14May83,dnw added OFFSET macro 01b,17Feb83,dnw added stuff from Whitesmiths std.h 01a,15Feb83,dnw written The other files having initial modification history entries in 1983 are: i208.h:01a,24mar83,jlf written i534.h:01a,10mar83,jlf written memLib.h:01a,24may83,dnw written rt11Lib.h:01a,14Feb83,dnw written semLib.h:01a,15nov83,ecs written symLib.h:01a,03aug83,dnw written tyLib.h:01a,24jun83,ecs written vxWorks.h:01a,15Feb83,dnw written - - Andrew - -- The world is such a cheerful place when viewed from upside-down It makes a rise of every fall, a smile of every frown --------------------------- Newsgroups: comp.os.vxworks Subject: How is "..." used in the body of a VxWorks function? Date: Fri, 6 Jul 2001 14:02:34 -0400 From: "Adam J. Caruso" Organization: Marconi Message-ID: <9i4um1$2l7$1@newsfeed.pit.comms.marconi.com> Hello, Several VxWorks function use a "..." in the body. I'm not sure what "..." means in the function's body or how it is able to compile. Could someone please enlighten me? Example code is found below. STATUS intConnect ( VOIDFUNCPTR * vector, /* interrupt vector to attach to */ VOIDFUNCPTR routine, /* routine to be called */ int parameter /* parameter to be passed to routine */ ) { ... } void excVecSet ( FUNCPTR *vector, /* vector offset */ FUNCPTR function /* address to place in vector */ ) { ... } Thanks, Adam --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: 6 Jul 2001 11:17:01 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107061017.509ac515@posting.google.com> References: Hello Joe, Try checking the following link (you'll need your WindSurf user name & password). At the top of this page you'll find the link to cumulative patch 3 for T2.0. http://www.wrs.com/csdocs/kplocator/patchProduct.shtml To find these from the WindSurf front page, click on the download tab at the top, and then the Patches link. This will take you to the SPR list; then select the "patches by product" link to get to the link above. HTH, John... joewoodbury@mail.com (Joe Woodbury) wrote in message news:... > I see suggestions on this board to check Wind River Systems Windsurf > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > I am a legitimate customer, have a customer ID and access to WindSurf > but I am unable to find either of these things. What am I doing wrong? > > Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: test Date: Fri, 6 Jul 2001 21:03:59 +0200 From: "Uri Yosef" Organization: NetVision Israel Message-ID: <9i4um6$m56$1@news.netvision.net.il> --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Accessing Host files from VxSim? Date: 6 Jul 2001 13:19:58 -0700 From: ddiinnxx@YAHOO.COM (Dinker Charak) Organization: http://groups.google.com/ Message-ID: <58e54b3.0107061219.463d1bd6@posting.google.com> References: > 1) I tried using the full pathname "c:\thedir\myfile.dat" > 2) I also tried in windshell > cd c:\thedir > ls - myfile.dat yes file is there > fp = open("myfile.dat", 0, 0); I have not worked with VxSim ... how does it talk to the remote file system. Like do you need to have a FTP server running or the VxSim does not need it? If you have a FTP server running, what does the message on the server log show? if after ls you do pwd, what does it show? try appending that to the file name. Or copy myfile.dat to the default local dir. So that eleminates the need for path prefixing to the file name. Random thoughts ... hope they help dinx --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado II / VxWorks 5.4 / Linux ??? Date: 6 Jul 2001 13:23:59 -0700 From: ddiinnxx@YAHOO.COM (Dinker Charak) Organization: http://groups.google.com/ Message-ID: <58e54b3.0107061223.374a1baf@posting.google.com> Hello, Is Tornado II for Linux available? Dinx --------------------------- Newsgroups: comp.os.vxworks Subject: Trying to replace LXT970 by LXT971A HELP ! Date: Fri, 06 Jul 2001 20:52:51 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: Hi ! I am trying to integrate the motFecEnd Driver to support the LXT971A. All the configuration registers are correct. When data comes in, the processor reboot ... Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e2100 (tNetTask): motFecHandleRecvInt: 0x8e2100 (tNetTask): motFecHandleRecvInt... Done, 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e2100 (tNetTask): motFecHandleRecvInt: 0x8e2100 (tNetTask): motFecHandleRecvInt... Done, 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e2100 (tNetTask): motFecHandleRecvInt: 0x8e2100 (tNetTask): motFecReceive rbdIndex=0 0x8e2100 (tNetTask): data in : 64 0x8e2100 (tNetTask): motFecRbdClean... Done, 0x8e2100 (tNetTask): motFecReceive... Done, 0x8e2100 (tNetTask): motFecHandleRecvInt... Done, 0x8e2100 (tNetTask): motFecRxTxHandle: /* The CPU reboot here */ Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e /* The CPU reboot here */ Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e2100 (tNetTask): motFecHandleRecvInt: 0x8e2100 (tNet /* The CPU reboot here */ Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e /* The CPU reboot here */ Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e2100 /* The CPU reboot here */ Configuring network interface motfec0... done. 0x8e2100 (tNetTask): motFecRxTxHandle: 0x8e --------------------------- Newsgroups: comp.os.vxworks Subject: Can I put all files in a directory to remote FTP server one time? Date: 6 Jul 2001 14:19:56 -0700 From: yxiu@salira.com (Yan Xiu) Organization: http://groups.google.com/ Message-ID: <22159022.0107061319.50e7fd8c@posting.google.com> When I put files in TFFS to remote FTP server, I want to put all files in a directory to server one time, can I do it? Yan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Fri, 06 Jul 2001 14:26:20 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B462CFC.1956AD3B@covad.net> References: Reply-To: drdiags@covad.net Joe, There was a window where some of the SPRs included in previous T2CPx releases were removed from public access. If there is a particular SPR that you want to know about and cannot find it on the WindSurf webpage, you may have to contact your WindRiver support personnel to get access to this. I thought I had heard that this was being corrected, so if you are not able to find your topic of interest, contact Wind River and let them give you the reasoning behind this. HTH. Joe Woodbury wrote: > > I see suggestions on this board to check Wind River Systems Windsurf > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > I am a legitimate customer, have a customer ID and access to WindSurf > but I am unable to find either of these things. What am I doing wrong? > > Joe Woodbury - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPP fail to re-connect from NT to vxWorks Date: Fri, 6 Jul 2001 17:19:50 -0400 From: "Jason Ke" Organization: Nortel Message-ID: <9i59ua$lk9$1@bcarh8ab.ca.nortel.com> References: <9i2f2r$70$1@bcarh8ab.ca.nortel.com> <3B45329C.B25B91C0@covad.net> So does this mean that on the vxWorks side I have to keep monitoring if the link is up or not? Does NT send some "hungup" signall to PPP stack on vxWorks? "DrDiags" wrote in message news:3B45329C.B25B91C0@covad.net... > Jason, > > Jason Ke wrote: > > > <....SNIP.....> > > 2) From vxWorks side, how do I know NT side has disconnected > > I found this in the latest release of the VxWorks Network 5.4 > Programmers Guide Edition 2 pg 37: > ======================================================================= > > The return value of pppDelete( ) does not indicate the status of the PPP > link. To check whether the link is actually terminated, call pppInfoGet( > ) and make sure the return value is ERROR. The following code fragment > demonstrates the usage of this mechanism for PPP unit 4: > > > PPP_INFO pppInfo; > > if (pppInfoGet (4, &pppInfo) == ERROR) > return (OK); /* link terminated */ > else > return (ERROR); /* link still up */ > > <....SNIP....> > -- > http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home > page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can I put all files in a directory to remote FTP server one time? Date: Fri, 06 Jul 2001 14:35:13 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B462F11.5195DEFA@covad.net> References: <22159022.0107061319.50e7fd8c@posting.google.com> Reply-To: drdiags@covad.net Yan, Sorry, I don't understand the question. Do you mean is it possible to do a mput *.* from a remote FTP client, so this is a vxWorks FTP Server question? Or do you mean is it possible for TFFS to handle multiple file copies and this is a TFFS question. Or are you asking if the vxWorks FTP client supports "mput *.*". Is it possible to narrow down the choices? Sorry if I missed the obvious here, plus I probably don't have an answer, just wanted to let you know my confusion...... Yan Xiu wrote: > > When I put files in TFFS to remote FTP server, I want to put all files > in a directory to server one time, can I do it? > > Yan - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado II / VxWorks 5.4 / Linux ??? Date: Fri, 06 Jul 2001 14:30:06 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B462DDE.F35B060F@covad.net> References: <58e54b3.0107061223.374a1baf@posting.google.com> Reply-To: drdiags@covad.net Dinker, Don't believe that Linux support was added until vxWorks AE. You probably will have to do a port or ask your FAE if a Professional Services release was done for this. Just my opinion...... Dinker Charak wrote: > > Hello, > > Is Tornado II for Linux available? > > Dinx - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Symbol Table Sync Problem Date: Fri, 6 Jul 2001 18:19:24 -0400 From: s.eckardt@ixthos.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9i5gq1$2k8$1@overload.lbl.gov> I've been successfully running a VxWorks image/tgtsvr with symbol table synchronization enabled, on a networked Sun Solaris host. Upon moving to a non-networked Windows NT host, I consistently get the following error: Fatal WTX error (0x1012F), synchronization stopped. Any idea why? If it matters, my board is configured as a Slave on a VME Shared Memory network. As such, its Ethernet traffic goes through another board which maintains the physical Ethernet connection to the host. Suzie Eckardt Chief Software Engineer Ixthos, Inc. s.eckardt@ixthos.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Symbol Table Sync Problem Date: Fri, 06 Jul 2001 16:51:48 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B464F14.A6F896A7@covad.net> References: <9i5gq1$2k8$1@overload.lbl.gov> Reply-To: drdiags@covad.net Suzie, Here is what I remember about this. I constantly had the same problem on a WinNT box. I was told that I had to set the Tornado->Tornado Registry from Local Registry to Remote Registry and put my Host IP address or name if it was in my Host file/DNS server. There was some gibberish that the Target Server would try to use the localhost entry under some unknown condition that would make it attempt to connect to the Registry using its (Target) loopback interface. Well, enough of this rambling. Put your non-networked NT box IP address in your Host file and then put that name in your Options Remote Registry entry. HTH s.eckardt@ixthos.com wrote: > > I've been successfully running a VxWorks image/tgtsvr with symbol table > synchronization enabled, on a networked Sun Solaris host. Upon moving to a > non-networked Windows NT host, I consistently get the following error: > Fatal WTX error (0x1012F), synchronization stopped. Any idea why? > > If it matters, my board is configured as a Slave on a VME Shared Memory > network. As such, its Ethernet traffic goes through another board which > maintains the physical Ethernet connection to the host. > > Suzie Eckardt > Chief Software Engineer > Ixthos, Inc. > s.eckardt@ixthos.com - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I block VxWorks default shell (->)? Date: Fri, 6 Jul 2001 18:57:58 -0500 From: "John Poplett" Organization: MindSpring Enterprises Message-ID: <9i5jb4$lfs$1@slb0.atl.mindspring.net> References: <9hjt5a$ee2$1@news1.kornet.net> I tried the Telnet server Johan refers to. It works but in our multi-threaded application, only the CLI thread had its standard output mapped to theTelnet session. I tried to forward output to the CLI thread but this was awkward and the whole solution was not reliable in practice. The unsupported Telnet server also redefines several functions in telnetLib.h. I feared the risk of clobbering the new telnetLib.h with each Tornado update, a problem that compounds if you have multiple Tornado installations. You also host Telnet sessions on different ports. The standard Telnet port maps to the VxWorks shell and you access your CLI through some alternate port. I am having good luck with an alternative strategy: export a a single entrypoint that invokes your CLI's command processing loop and call it from the standard VxWorks shell. Implement an exit command that terminates the processing loop, e.g.: - -> myCLI() My CLI V1.0 Jan 1, 1970 [My CLI] exit - -> This approach will work with the standard Telnet implementation provided with Tornado, is simple to implement and maintain and is completely accessible from the standard Telnet port. John I had difficulty mapping standard output from all processes to Telnet session since their handles "Johan Borkhuis" wrote in message news:Xns90D198216D0FDborkhuisagerecom@135.7.153.102... > "±èÀÀ½Ä" wrote: > > As far as I know windriver don't show the source of ``telnet'' and > > ``shell''. > > However I am looking forward to enter my own shell command interpreter > > when I connect target throught telnet. > > There is an unsupported Telnet server available from WindRiver. Ask your > FAE for it. > > Groeten, > Johan > > -- > o o o o o o o . . . _____________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | borkhuis@agere.com | > >(________|__|_[_________]_|__________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === > --------------------------- Newsgroups: comp.os.vxworks Subject: FW: Network pool running out... Date: Fri, 29 Jun 2001 16:53:20 +0100 From: "David Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i5kac$3rv$1@overload.lbl.gov> - -----Original Message----- From: David Anderson [SMTP:david.anderson@dtrack.com] Sent: 27 June 2001 16:05 To: 'Vxworks_Users (E-mail) Subject: Network pool running out... vxWorks 5.3.1 Tornado 1.0.1 Hello all, I've run into a problem which produces an apparent 'freeze up' of a PPP link, but still allows 'normal' Ethernet style TCP/IP connections to proceed. We are using the Tornado 2.0 ppp daemon. It appears that running an FTP session and a TELNET session concurrently over the PPP link causes the network pool to run out of buffers ( error 0x73000a ). However, the PPP link does not recover, the daemon tasks are running but you can no longer ping from either end to the other, or establish a new socket connection across it. The buffer error appears only to be a soft error - an Ethernet TCP/IP connection can still be set-up OK. Any ideas about why the PPP link should be 'frozen out' although all status indications suggest it is open and up. David Anderson. NEW E-MAIL ADDRESS: mailto:david.anderson@dtrack.com - ------------------------------------- FF: 2B + ~2B , that is the question ? - ------------------------------------- Disclaimer - ---------- This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please destroy and notify Data Track Technology Plc +44 1425 271900. - ------------------------------------------------------- From vxwexplo-errs@csg.lbl.gov Fri Jul 6 16:27:18 2001 Received: from postal1.lbl.gov (postal1.lbl.gov [128.3.7.82]) by csg.lbl.gov (8.9.3/8.9.3) with ESMTP id QAA18009 for ; Fri, 6 Jul 2001 16:27:18 -0700 (PDT) Received: from SpamWall.lbl.gov (localhost [127.0.0.1]) by postal1.lbl.gov (8.11.2/8.11.2) with ESMTP id f66NRHA20605 for ; Fri, 6 Jul 2001 16:27:17 -0700 (PDT) Received: from kilby.brackett.com (mail.brackett.com [209.183.211.3]) by SpamWall.lbl.gov (8.11.2/8.11.2) with ESMTP id f66NRG920601 for ; Fri, 6 Jul 2001 16:27:16 -0700 (PDT) Received: from defiant (sdsl-216-36-86-22.dsl.iad.megapath.net [216.36.86.22]) by kilby.brackett.com (8.9.3/8.9.3) with SMTP id TAA17129 for ; Fri, 6 Jul 2001 19:27:20 -0400 From: "Mike Anderson" To: "the vxWorks Users Group Exploder" Subject: RE: Symbol Table Sync Problem Date: Fri, 6 Jul 2001 19:26:24 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2911.0) X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Importance: Normal In-Reply-To: <200107062232.PAA17492@csg.lbl.gov> Hi Suzie, > Submitted-by: s.eckardt@ixthos.com > > I've been successfully running a VxWorks image/tgtsvr with symbol table > synchronization enabled, on a networked Sun Solaris host. Upon > moving to a > non-networked Windows NT host, I consistently get the following error: > Fatal WTX error (0x1012F), synchronization stopped. Any idea why? > > If it matters, my board is configured as a Slave on a VME Shared Memory > network. As such, its Ethernet traffic goes through another board which > maintains the physical Ethernet connection to the host. > The problem is that your Windows host isn't listening for the target agent connection on a socket. The way to fix it is to go to the options tab in Tornado and and set the Wind Registry up to use a network address (the address of the machine running the Registry server) instead of selecting the local registry entry. This allows the target agent to connect. HTH, Mike - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Symbol Table Sync Problem Date: Fri, 06 Jul 2001 17:25:32 -0700 From: Lori Fraleigh Organization: Real-Time Innovations Message-ID: <3B4656FC.A40790D@rti.com> References: <9i5gq1$2k8$1@overload.lbl.gov> Suzie - There is a bug in the Windows version of Tornado that prevents symbol table synchronization from working on of the box. Bring up Tornado and go to the Tools menu. Select Options Select Tornado Registry at the far right Change the registry from local machine to remote machine and enter either the name or ip address of your Windows machine. When you specify local machine, 127.0.0.1 gets passed down to the target and the target tries to talk to itself, not the host. Oops! Regards, Lori s.eckardt@ixthos.com wrote: > > I've been successfully running a VxWorks image/tgtsvr with symbol table > synchronization enabled, on a networked Sun Solaris host. Upon moving to a > non-networked Windows NT host, I consistently get the following error: > Fatal WTX error (0x1012F), synchronization stopped. Any idea why? > > If it matters, my board is configured as a Slave on a VME Shared Memory > network. As such, its Ethernet traffic goes through another board which > maintains the physical Ethernet connection to the host. > > Suzie Eckardt > Chief Software Engineer > Ixthos, Inc. > s.eckardt@ixthos.com - -- Lori Fraleigh Director of Engineering Real-Time Innovations, Inc. http://www.rti.com Phone: 408-734-4200 x121 Fax: 408-734-5009 --------------------------- Newsgroups: comp.os.vxworks Subject: Check This Out!!! 1712 Date: Sat, 07 Jul 2001 00:38:24 GMT From: kqtzys@my.com Organization: RoadRunner - Midsouth Message-ID: <4Ss17.78754$By.1839411@typhoon.midsouth.rr.com> Check out our new store at www.our-superstore.com bkdqncwfqsdhyghxisnhw --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can I put all files in a directory to remote FTP server one time? Date: 6 Jul 2001 17:56:52 -0700 From: yxiu@salira.com (Yan Xiu) Organization: http://groups.google.com/ Message-ID: <22159022.0107061656.72365639@posting.google.com> References: <22159022.0107061319.50e7fd8c@posting.google.com> <3B462F11.5195DEFA@covad.net> Hi, DrDiags, My question is if vxWorks FTP client supports "mput *.*", and including all the files and sub-directories in a whole directory. Thanks, Yan --------------------------- Newsgroups: comp.os.vxworks Subject: Check This Out!!! 8134 Date: Sat, 07 Jul 2001 01:07:02 GMT From: ivveld@my.com Organization: RoadRunner - Midsouth Message-ID: Check out our new store at www.our-superstore.com wcdbdoxwimbetqncxjygiyzbonwwowqejzdp --------------------------- Newsgroups: comp.os.vxworks Subject: ISR callable routines Date: 6 Jul 2001 18:48:44 -0700 From: pepsi_000@hotmail.com (vxJunkie) Organization: http://groups.google.com/ Message-ID: <676eb679.0107061748.2cbd27c0@posting.google.com> hi folks, Can i call functions like a] netClusterGet b] netClBlkGet c] netMblkGet from the ISR context? I am not sure if these are blocking calls. Thanks in advance --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Unable to receive multicast packets on a raw socket in VxWorks. Date: 6 Jul 2001 18:59:25 -0700 From: pepsi_000@hotmail.com (vxJunkie) Organization: http://groups.google.com/ Message-ID: <676eb679.0107061759.2f9e0133@posting.google.com> References: <3c2606e4.0107050452.230478a5@posting.google.com> kavleenk@delhi.tcs.co.in (Kavleen Bhatia) wrote in message news:<3c2606e4.0107050452.230478a5@posting.google.com>... > I need to send & receive raw ethernet packets off the network. I am > able to send to the multicast group but am facing problem in joining > the multicast group for receiving the packet. The "setsockopt" > function does not work if the interface does not have a valid IP > address. > > I need to know how to attach my Interface(by specifying only the MAC > address) to the multicast group, to receive multicast packets. > > I would like to know if somebody has successfully done the same hi kavleen, Have you ever tried using MUX layer for sending and receiveing ethernet packets....This is a very nice architecture from tornado...check it out with your FAE...You do not need sockets to send and receive raw ethernet packets. Bye, vxJunkie --------------------------- Newsgroups: comp.os.vxworks Subject: how to obtain vxworks software for demo? Date: Sat, 7 Jul 2001 09:31:04 +0800 From: "MicroSat" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i5rbc$69u$1@overload.lbl.gov> Hi, I am a student. I want to use vxworks but don't know where there is vxworks software for demo. Regards --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to obtain vxworks software for demo? Date: Sat, 07 Jul 2001 00:09:47 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B46B5BB.C2F69E1F@covad.net> References: <9i5rbc$69u$1@overload.lbl.gov> Reply-To: drdiags@covad.net MicroSat, Visit the Wind River WindSurf website. You should be able to download the Tornado 2 Prototyper, which is a time based version of Tornado 2 and tools with vxSim (simulator) capability. I believe that this may not have the full networking capability, but other than that enough to get you started. MicroSat wrote: > > Hi, > I am a student. I want to use vxworks but don't know where there is vxworks software for demo. > > Regards - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: How to setup a PPP link through telephone connection. Date: 7 Jul 2001 00:28:50 -0700 From: gujc@huawei.com (Gu, Jiangchun) Organization: http://groups.google.com/ Message-ID: Hi, everyone: I want to setup a PPP link on target board running vxWorks v5.4, but I don't know when, where and how to control the modem. The target works as a server and waits the other peer to dial in. Any help will be appreciated, thanks in advance! Gu, Jiangchun --------------------------- Newsgroups: comp.os.vxworks Subject: missing timer_connect() and friends ... Date: Sat, 07 Jul 2001 08:34:18 GMT From: Julian Sun Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: <3B46C989.C47931F5@yahoo.com> Nothing really special here, but when I tried to download my downloadable app to the NTsim, it complains about missing timer_connect(), timer_create() and timer_settime(), etc. Everything otherwise works fine (i.e., if I remove those functions the app works fine.) Did I miss anything here? Thanks. Julian --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 8 04:03:18 2001 From: Vxworks Exploder Date: Sun Jul 8 04:03:20 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 8 04:03:04 PDT 2001 Subject: Re: How to get patches Subject: Re: How old is VxWorks ? Subject: Is Wind River really this pathetic Subject: project multiprocessor dsp Subject: Re: How is "..." used in the body of a VxWorks function? Subject: 5.4 release notes? Subject: Re: How to get patches Subject: Re: How to get patches ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Sat, 7 Jul 2001 11:02:13 -0600 From: "Joe Woodbury" Message-ID: References: <9i3lif$4pg$1@news.hdpp.de> <9i3m4s$4r8$1@news.hdpp.de> I do have a valid support contract. As it turns out Wind River is a rather pathetic company--they messed up our original order, giving us the wrong keys and incorrect access to Wind Surf. After a heated discussion on the phone with the local field engineer and his boss back in California, they promised to fix everything, but have yet to do so. Exactly how hard is it to administer a simple database? Apparently quite hard. Needless to say, we've put a hold on our payment to Wind River. Are we alone in being treated like sh*t by this company? Joe Woodbury "Matthias Guelck" wrote in message news:9i3m4s$4r8$1@news.hdpp.de... > Sorry, > > the only question is: do you have a valid support contract? > > Kind regards > > Matthias > > "Matthias Guelck" wrote in message news:9i3lif$4pg$1@news.hdpp.de... > > Hello Joe, > > > > on http://www.wrs.com/corporate/support/private/validate.html you might see the reason: > > either you purchased your Wind River product more than 1 year ago or your support contract is no longer valid (like mine). > > > > Kind regards > > > > Matthias > > > > "Joe Woodbury" wrote in message news:c8c13723.0107051237.4700018@posting.google.com... > > > I see suggestions on this board to check Wind River Systems Windsurf > > > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > > > > > I am a legitimate customer, have a customer ID and access to WindSurf > > > but I am unable to find either of these things. What am I doing wrong? > > > > > > Joe Woodbury > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How old is VxWorks ? Date: 7 Jul 2001 10:06:21 -0700 From: espin@idiom.com (Geoffrey Espin) Organization: http://groups.google.com/ Message-ID: <51f8f6c6.0107070906.41e0efaa@posting.google.com> References: <9i3urd$2lg4$1@news.univ-angers.fr> <3B45FCAB.3FE95040@aps.anl.gov> Andrew Johnson wrote in message news:<3B45FCAB.3FE95040@aps.anl.gov>... > Jonathan ILIAS wrote: > > > > Does anyone know what is VxWorks birth date ? > This isn't definitive but the first modification history entry in the > header file vxWorks.h from vxWorks 4.0.2 is dated 15Feb83 (yes, we still > modification history Good answer. I (gae) better write this down while I can still fuzzily remember. WRS was incorporated in 1981, the same year as Sun Micro, by Jerry Fiddler (jlf). He hired and made a partner David Wilner (dnw) who took over a contract for a "real-time" video editing application. Jerry and his wife-to- be, went on around the world for ~6? months. Dave & Jerry had met up at LBL Real-Time Systems Group in Berkeley. When Jerry came back, Dave had written VxWorks. :-) It was the shell (brilliant use of nm, sed and yacc), some good Makefiles and an in-memory relational database all on top of VRTX (hence "vx"). Eric Stromberg (ecs) was hired to do utility application development. I think the 3 of them toiled on mainly video editing applications for years. Control Video Corporation, owned by Larry Seehorn, was considered the Godfather of WRS as he got Jerry & Co their first several jobs. Finally, CVC failed to pay their bills and the application and nascent VxWorks was sold to a Canadian company, Editec, where I was working in '85. I got to come to 67th St. in Emeryville to learn about the editing application and VxWorks. Bob Cohen (rdc) was hired around '85. He worked on a new video story-board contract for Francis Ford Coppola (Zoetrope) later he brought up BSD 4.2 TCP/IP with dnw's help. In 1986 Leslie Kirby (llk) was added (RPC&NFS). My Canadian company folded, so I got sqeezed in, being the first & only VxWorks programmer outside of WRS (& first WRS H1B visa). I finished off another video editing application for Lake Systems (LaKart), while almost all engineering started focusing on VxWorks "The Product". The first h/w I recall was Multibus-I with 68000, then 68010, but now 68020 on VME was the rage -- we must have done 20 BSPs -- with no tools but an EPROM UV eraser. The ISI (Integrated Solutions Inc -- Sun contemporary - -- VME/68k vendor) was the host system with BSD4.2 Unix -- it was great! John Fogelin (jcf), bud of Bob's, had just dropped out of UCB and was hired to do odd-jobs. He first made a Multibus-I and VME board holders out of plywood. This must have been good training for BSP work. A month later he was bringing up VxWorks on a Sun-2 (68020)! Within a year, he was writing a generic kernel layer and added the pSOS option, then he wrote the WIND kernel, with guidance from dnw. Early '87 we got our first Sun-3 server with Sunview & monitor. I recall thinking for many months that my WISE terminal was much better and efficient. The first VP of Sales Tim Wilkins came from Heurikon, a VME board manufacturer. WRS real success was built by enthusistic VME board manufacturers' salesmen. Tim took WRS from $1M/year to $4M/year (~89?)! :-) Kent Long (kdl) joined from Ready Systems and did a "clean-room" DosFS. Jim Talbot(jwt) came from Force and took over SPARC. Mark Fox joined as the GNU guru. Many more bodies were added (see Hwa-Jin Bae's (hjb) old FAQ for some names). Here's a partial list pre-90 from memory (apologies in advance for those forgotten). dzb - Don Brooks -- bsp ajm - Al Maillet -- ported MIPS single-handedly, now in WRS Sales nfs - Nigel Standing -- when support was great! (deceased '2000) yp - Yogu Ponnambalam -- ditto jmm - James Moore -- ditto elh - Emily Hipp -- networking (with hjb) jdi - Jim Irwin -- tech writer sh - Scott Herzinger -- C++ hdn - Hdei Nunoe -- TRON, :-) x86 ms - Marc Shepard -- caf - Chris Ford -- bsp ccc - Carl Chesbrough -- bsp yao - Yaoyao Chen - bsp jcc - jason cotton -- scsi rrr - Robert Rose -- kernel & fixed signals on all arch's del - David Larsen -- 960 wmd - William Deer -- bsp mcl - Mark Linimon -- sparc @ mizar(?) (VME vendor) rbr - roger rohrbach -- tools scl - Simon Leong -- bsp elr - Eric Rabinowitz -- bsp Anyways... yeah, the answer was '81. Geoff - -- Geoffrey Espin espin@idiom.com --------------------------- Newsgroups: comp.os.vxworks Subject: Is Wind River really this pathetic Date: Sat, 7 Jul 2001 11:12:45 -0600 From: "Joe Woodbury" Message-ID: <6qH17.79$5f6.133962@news.uswest.net> Since we ordered VxWorks/Tornado II, we've had nothing but problems with Wind River. They filled the order incorrectly and I had to resort to a lot of yelling and threats to not pay our purchase order to get their attention. Yet, two days later, they still have yet to do anything but email the correct BSP. On top of all this, the highest level manager at Wind River kept blaming us for some of the problems. So far, I've concluded that Wind River employees are given an IQ test. If they fail, they are given a personal relationship test. If they fail that also, they are hired. And heaven forbid they actually do something so obvious like creating a compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. The shame is that VxWorks proper is such an excellent product. The questions are: Are our experiences unique? Did our company just happen to run into the one group of bastards working at Wind River? Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: project multiprocessor dsp Date: Sat, 7 Jul 2001 18:57:01 +0200 From: "federico torretta" Organization: Tiscali Spa Message-ID: <9i7g27$a5k$1@pegasus.tiscalinet.it> federico wrote in message ... > my name is Federico and I am a major in > >>Electronic Engineering > > >>My thesis in microelectronics is about the > >>implementation in hardware of an algorhythm that > >>simulates the reaction of a neuron network exposed to > >>outer electomagnetic fields. > >>The above mentioned algorhythm first of all models the > >>behaviour of single ionic channels of the cellular > >>membrane, then it joins them as to simulate the > >>bahaviour of the cellular membrane (determination of > >>its action potential when exposed to an > >>electromagnetic field, referring above all to the > >>Hadgin-Huxley approach)and finally the single cellular > >>membranes are integrated in order to describe the > >>behavoiur of a neuron network. > >>I kindly ask you if you know any hardware device which > >>already implements algorhythm that are similar to the > >>above mentioned. Or more simply how I can connect DSP > >>processors (which simulate the effect of the membrane > >>of a single neuron) without using share memories, (foe > >>example with a bus ...) > >> > >>Thanking you in advance I remain > >> > >> Sincerely Yours > >> Federico > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How is "..." used in the body of a VxWorks function? Date: 7 Jul 2001 10:48:49 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107070948.6c1210ff@posting.google.com> References: <9i4um1$2l7$1@newsfeed.pit.comms.marconi.com> Hello, These are not functions that are compiled; they a 'fake' C versions of functions that are implemented in assembler and are used to generate the reference manuals. The directory you are looking in is no doubt called 'doc' for documentation ;-) HTH, John... "Adam J. Caruso" wrote in message news:<9i4um1$2l7$1@newsfeed.pit.comms.marconi.com>... > Hello, > > Several VxWorks function use a "..." in the body. I'm not sure what "..." > means in the function's body or how it is able to compile. Could someone > please enlighten me? Example code is found below. > > STATUS intConnect > ( > VOIDFUNCPTR * vector, /* interrupt vector to attach to > */ > VOIDFUNCPTR routine, /* routine to be called > */ > int parameter /* parameter to be passed > to routine */ > ) > { > ... > } > > void excVecSet > ( > FUNCPTR *vector, /* vector offset */ > FUNCPTR function /* address to place in vector */ > ) > { > ... > } > > Thanks, > Adam --------------------------- Newsgroups: comp.os.vxworks Subject: 5.4 release notes? Date: Sat, 7 Jul 2001 20:52:28 +0100 From: "Archerfish" Message-ID: <9i7pan$n3d$1@news5.svr.pol.co.uk> I've searched Wind River's website for something approximating the VxWorks 5.4 release notes, without success. Not even a press release. Can anybody recommend upgrading from 5.3.1 to 5.4 ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Sat, 07 Jul 2001 21:14:32 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B47DE28.A61D12DB@covad.net> References: <9i3lif$4pg$1@news.hdpp.de> <9i3m4s$4r8$1@news.hdpp.de> Reply-To: drdiags@covad.net Joe, The Sales guy who sold your company the product has the obligation to insure that you are getting what he/she promised you. Your field person has about as much power to change anything as you and I. Remember, if you cannot ship product because of Wind River's products, you have a very large stick to fight with. I would push to have your upper management raise this up the Wind River management chain to the CEO, Tom St. Dennis. Do you think arrogance and greed is only marketed by select companies? This is a trend that seems to exist up and down the line with many companies. I don't think WRS has sole ownership on this one. Ineptitude, just not an Wind River commodity. We all are aware of having to follow the "process" and instead of just fixing the darn problem, have meetings to talk the process to death. Me, I never produce bad code, it always works the first time. No test engineer has ever been able to find fault in my product and I always work in close harmony with .... But seriously, I am surprised to hear that your Sales person isn't getting involved with this. Seems he/she would like to keep a good working relationship with your company, so he/she can try to sell you more stuff. (when pigs fly.....?) <....SNIP....> - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Sat, 7 Jul 2001 22:51:33 -0600 From: "Joe Woodbury" Message-ID: References: <9i3lif$4pg$1@news.hdpp.de> <9i3m4s$4r8$1@news.hdpp.de> <3B47DE28.A61D12DB@covad.net> Unfortunately, my sales person is on vacation. Maybe you don't produce bad code, but I produce it all the time. Except I take responsibility for my mistakes and don't treat my customers like sh*t. (Yes, I know you were being sarcastic.) "DrDiags" wrote in message news:3B47DE28.A61D12DB@covad.net... > Joe, > > The Sales guy who sold your company the product has the obligation > to insure that you are getting what he/she promised you. Your field > person has about as much power to change anything as you and I. > Remember, if you cannot ship product because of Wind River's products, > you have a very large stick to fight with. I would push to have your > upper management raise this up the Wind River management chain to > the CEO, Tom St. Dennis. > Do you think arrogance and greed is only marketed by select > companies? This is a trend that seems to exist up and down the line > with many companies. I don't think WRS has sole ownership on this one. > Ineptitude, just not an Wind River commodity. We all are aware of > having to follow the "process" and instead of just fixing the darn > problem, have meetings to talk the process to death. > Me, I never produce bad code, it always works the first time. > No test engineer has ever been able to find fault in my product and > I always work in close harmony with .... > But seriously, I am surprised to hear that your Sales person isn't > getting involved with this. Seems he/she would like to keep a good > working relationship with your company, so he/she can try to sell > you more stuff. (when pigs fly.....?) > > > <....SNIP....> > > -- > http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home > page) --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 8 06:54:56 2001 From: vxworks_user Date: Sun Jul 8 06:54:59 PDT 2001 Subject: I want to join the mail list of vxWorks ______________________________________ =================================================================== ÐÂÀËÃâ·Ñµç×ÓÓÊÏä (http://mail.sina.com.cn) ¶©ÔÄÊÖ»ú¶ÌÐÅÍ·ÌõÐÂÎÅ£¬ÌìÌì½±¶à¿îʱÉÐÊÖ»ú£¡ (http://dailynews.sina.com.cn/c/272235.html) ¶©ÔÄÊÖ»ú¶ÌÐŶ¥¼¶ÐÂÎÅÿÌìµÃпîÊÖ»ú´ó½±£¡ (http://dailynews.sina.com.cn/c/266499.html) From vxwexplo-errs@csg.lbl.gov Sun Jul 8 07:51:04 2001 From: "David Laight" Date: Sun Jul 8 07:51:06 PDT 2001 Subject: Re: standalone_vxworks with no network > I can't explain *why* WRS decided not to initialize the network in a > standalone target, but I can tell you how to get it to change that: You > need to add #define STANDALONE_NET to your configAll.h file and rebuild. Or just rewrite chunks of usrNetwork.c to initialise the things you need to do in the right order with parameters taken from somewhere handy. The vxWorks 5.4 file is a much clearer than the 5.3.1 copy - and will compile for either version. For instance you may decide that you want to start TCP even if you can't find a network card! The 'lo' interface will work fine, the daemon listeners can be started etc. If a network card appears later (eg a pcmcia card is inserted, or a modem PPP connection dials) then everything is ready. Also some parts of the system will use TCP for local connections, and they wil also work. Rather than trying to start the network described in the boot line, I used the boot line as the parameters for the named interface. Any other interface would use DHCP. The network card driver code caused the interface to be initialised (as opposed to the boot parameters indicating which interface to bring up). There is an awful lot of 'dross' in usrNetwork.c ... David From vxwexplo-errs@csg.lbl.gov Sun Jul 8 08:02:21 2001 From: "David Laight" Date: Sun Jul 8 08:02:23 PDT 2001 Subject: Re: PPP fail to re-connect from NT to vxWorks > 1) Am I supposed to call pppDelete, then usrPPPinit, for NT to re-connect? > 2) From vxWorks side, how do I know NT side has disconnected This is typical wrs code - 'tick in the box' but not robust! Turn on Debug in the ppp_config file - might help. To detect PPP disconnects I added a hook to my serial driver that would call code (to do pppDelete) when the serial port was closed. I found the the vxworks ppp disconnect hook was never called! (It is also impossible to use the connect hook for modem dialling - due to the brain-dead way the ppp stack is given receive data) Some words of warning! - the PPP code does not contain adequate locks against pppDelete being actioned in inappropriate states. You need to delay until the state has 'settled'. (it is, however, impossible to obtain this state information with the standard library code) - pppInit returns after creating the ppp task, it does not wait for the ppp data areas (eg that checked by pppInfoGet() to be set up). If the setup code blocks your initialisation code may be rescheduled before this has happened! I resolved this be checking with the serial device driver that the ppp code had reopened the serial fd - special serial driver... David From vxwexplo-errs@csg.lbl.gov Sun Jul 8 08:10:20 2001 From: "David Laight" Date: Sun Jul 8 08:10:23 PDT 2001 Subject: Re: un-numbered interfaces > Does anyone know whether VxWorks IP stack(SENS) support un-numbered > interfaces? It means we can attach an interface to MUX without the > interface IP address. Not supported by vxWorks. > Also we can add a route without the next hop IP address? There is a 'cludge' that works on many Unix systems. If you know the MAC address of a gateway, but not it's IP address (or, more usually, the IP address isn't in the same subnet as your own system so cannot be ARPed) then do the following: - allocate an IP address in your local subnet that is otherwise unused - add a static ARP table entry with that IP address and the gateway's MAC address - add the route to the allocated address The IP address you allocate NEVER appears in any packet, and is only used internally within the sending system (the gateway doesn't have to know you have done this). Maybe the vxworks mRouteEntryAdd can do it in a single entry? but I doubt it! (See if an appropriate entry is generated for a normal gateway - and see if you can replicate it.... The local rtTblShow() function might display extra info). Davod From vxwexplo-errs@csg.lbl.gov Sun Jul 8 08:20:35 2001 From: "David Laight" Date: Sun Jul 8 08:20:37 PDT 2001 Subject: Re: ISR callable routines > Can i call functions like > a] netClusterGet > b] netClBlkGet > c] netMblkGet > from the ISR context? I am not sure if these are blocking calls No - not a chance! (The routines uses semaphores to control access to the free list - which don't lock out interrupts.) vxWorks expects you to use netJobAdd to get tNetTask to go this work. Remember to add a flag so that the ISR doesn't schedule the 'work' routine more than once - or you WILL exceed the size of the kernel work Q and/or tNetTasks list. David From vxwexplo-errs@csg.lbl.gov Sun Jul 8 08:23:00 2001 From: "David Laight" Date: Sun Jul 8 08:23:03 PDT 2001 Subject: Re: How to get patches > Are we alone in being treated like sh*t by this company? Probably not. If windriver and vxWorks were what the salesmen claimed them to be then this newsgroup wouldn't have to exist..... David From vxwexplo-errs@csg.lbl.gov Mon Jul 9 04:03:27 2001 From: Vxworks Exploder Date: Mon Jul 9 04:03:30 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 9 04:03:05 PDT 2001 Subject: I want to join the mail list of vxWorks Subject: re: standalone_vxworks with no network Subject: re: PPP fail to re-connect from NT to vxWorks Subject: re: un-numbered interfaces Subject: re: ISR callable routines Subject: re: How to get patches Subject: Re: ISR callable routines Subject: Re: Accessing Host files from VxSim? Subject: Unsupported telnet server compile error... Subject: Re: How old is VxWorks ? Subject: driver for sh4 companion chip HD64465 Subject: Re: 5.4 release notes? Subject: Re: Is Wind River really this pathetic Subject: Re: How to setup a PPP link through telephone connection. Subject: Re: missing timer_connect() and friends ... Subject: QLogic SCSI/FC Driver Subject: IPv6 stack Subject: Absolute Linking Subject: Absolute Linking ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: I want to join the mail list of vxWorks Date: Sun, 08 Jul 2001 21:53:35 +0800 From: vxworks_user Organization: Lawrence Berkeley National Laboratory Message-ID: <9i9ptb$6qs$1@overload.lbl.gov> ______________________________________ =================================================================== ÐÂÀËÃâ·Ñµç×ÓÓÊÏä (http://mail.sina.com.cn) ¶©ÔÄÊÖ»ú¶ÌÐÅÍ·ÌõÐÂÎÅ£¬ÌìÌì½±¶à¿îʱÉÐÊÖ»ú£¡ (http://dailynews.sina.com.cn/c/272235.html) ¶©ÔÄÊÖ»ú¶ÌÐŶ¥¼¶ÐÂÎÅÿÌìµÃпîÊÖ»ú´ó½±£¡ (http://dailynews.sina.com.cn/c/266499.html) --------------------------- Newsgroups: comp.os.vxworks Subject: re: standalone_vxworks with no network Date: Sun, 8 Jul 2001 15:55:53 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i9te2$8o7$1@overload.lbl.gov> > I can't explain *why* WRS decided not to initialize the network in a > standalone target, but I can tell you how to get it to change that: You > need to add #define STANDALONE_NET to your configAll.h file and rebuild. Or just rewrite chunks of usrNetwork.c to initialise the things you need to do in the right order with parameters taken from somewhere handy. The vxWorks 5.4 file is a much clearer than the 5.3.1 copy - and will compile for either version. For instance you may decide that you want to start TCP even if you can't find a network card! The 'lo' interface will work fine, the daemon listeners can be started etc. If a network card appears later (eg a pcmcia card is inserted, or a modem PPP connection dials) then everything is ready. Also some parts of the system will use TCP for local connections, and they wil also work. Rather than trying to start the network described in the boot line, I used the boot line as the parameters for the named interface. Any other interface would use DHCP. The network card driver code caused the interface to be initialised (as opposed to the boot parameters indicating which interface to bring up). There is an awful lot of 'dross' in usrNetwork.c ... David --------------------------- Newsgroups: comp.os.vxworks Subject: re: PPP fail to re-connect from NT to vxWorks Date: Sun, 8 Jul 2001 16:07:17 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i9te9$8o8$1@overload.lbl.gov> > 1) Am I supposed to call pppDelete, then usrPPPinit, for NT to re-connect? > 2) From vxWorks side, how do I know NT side has disconnected This is typical wrs code - 'tick in the box' but not robust! Turn on Debug in the ppp_config file - might help. To detect PPP disconnects I added a hook to my serial driver that would call code (to do pppDelete) when the serial port was closed. I found the the vxworks ppp disconnect hook was never called! (It is also impossible to use the connect hook for modem dialling - due to the brain-dead way the ppp stack is given receive data) Some words of warning! - - the PPP code does not contain adequate locks against pppDelete being actioned in inappropriate states. You need to delay until the state has 'settled'. (it is, however, impossible to obtain this state information with the standard library code) - - pppInit returns after creating the ppp task, it does not wait for the ppp data areas (eg that checked by pppInfoGet() to be set up). If the setup code blocks your initialisation code may be rescheduled before this has happened! I resolved this be checking with the serial device driver that the ppp code had reopened the serial fd - special serial driver... David --------------------------- Newsgroups: comp.os.vxworks Subject: re: un-numbered interfaces Date: Sun, 8 Jul 2001 16:15:14 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9i9tei$8o9$1@overload.lbl.gov> > Does anyone know whether VxWorks IP stack(SENS) support un-numbered > interfaces? It means we can attach an interface to MUX without the > interface IP address. Not supported by vxWorks. > Also we can add a route without the next hop IP address? There is a 'cludge' that works on many Unix systems. If you know the MAC address of a gateway, but not it's IP address (or, more usually, the IP address isn't in the same subnet as your own system so cannot be ARPed) then do the following: - - allocate an IP address in your local subnet that is otherwise unused - - add a static ARP table entry with that IP address and the gateway's MAC address - - add the route to the allocated address The IP address you allocate NEVER appears in any packet, and is only used internally within the sending system (the gateway doesn't have to know you have done this). Maybe the vxworks mRouteEntryAdd can do it in a single entry? but I doubt it! (See if an appropriate entry is generated for a normal gateway - and see if you can replicate it.... The local rtTblShow() function might display extra info). Davod --------------------------- Newsgroups: comp.os.vxworks Subject: re: ISR callable routines Date: Sun, 8 Jul 2001 16:25:31 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9ia0uc$am6$1@overload.lbl.gov> > Can i call functions like > a] netClusterGet > b] netClBlkGet > c] netMblkGet > from the ISR context? I am not sure if these are blocking calls No - not a chance! (The routines uses semaphores to control access to the free list - which don't lock out interrupts.) vxWorks expects you to use netJobAdd to get tNetTask to go this work. Remember to add a flag so that the ISR doesn't schedule the 'work' routine more than once - or you WILL exceed the size of the kernel work Q and/or tNetTasks list. David --------------------------- Newsgroups: comp.os.vxworks Subject: re: How to get patches Date: Sun, 8 Jul 2001 16:27:56 +0100 From: "David Laight" Organization: Lawrence Berkeley National Laboratory Message-ID: <9ia0ud$am7$1@overload.lbl.gov> > Are we alone in being treated like sh*t by this company? Probably not. If windriver and vxWorks were what the salesmen claimed them to be then this newsgroup wouldn't have to exist..... David --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ISR callable routines Date: 8 Jul 2001 14:14:16 -0700 From: pepsi_000@hotmail.com (vxJunkie) Organization: http://groups.google.com/ Message-ID: <676eb679.0107081314.26e89809@posting.google.com> References: <9ia0uc$am6$1@overload.lbl.gov> hi david, Thanks for the info. What do you mean my "add a flag so that the ISR doesn't schedule the 'work' routine more than once " Does netJobAdd schedule the same job more than once? Thanks "David Laight" wrote in message news:<9ia0uc$am6$1@overload.lbl.gov>... > > Can i call functions like > > > a] netClusterGet > > b] netClBlkGet > > c] netMblkGet > > from the ISR context? I am not sure if these are blocking calls > > No - not a chance! > (The routines uses semaphores to control access to the free list - which > don't lock out interrupts.) > > vxWorks expects you to use netJobAdd to get tNetTask to go this work. > > Remember to add a flag so that the ISR doesn't schedule the 'work' routine > more than once - or you WILL exceed the size of the kernel work Q and/or > tNetTasks list. > > David --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Accessing Host files from VxSim? Date: Mon, 9 Jul 2001 10:38:07 +0800 From: "tanghaoyu" Organization: Bentium Ltd. Message-ID: <9ib5aq$9na$1@mail.cn99.com> References: <58e54b3.0107061219.463d1bd6@posting.google.com> try these steps: 1)"host:c:\thedir\myfile.dat" ,host is your host name in bootline. You can see it by "devs" in windshell. 2) @cd "host:" @cd "c:\thedir" @ls - myfile.dat yes file is there fp = open("myfile.dat", 0, 0); "Dinker Charak" wrote in message news:58e54b3.0107061219.463d1bd6@posting.google.com... > > 1) I tried using the full pathname "c:\thedir\myfile.dat" > > 2) I also tried in windshell > > cd c:\thedir > > ls - myfile.dat yes file is there > > fp = open("myfile.dat", 0, 0); > > I have not worked with VxSim ... how does it talk to the remote file > system. Like do you need to have a FTP server running or the VxSim > does not need it? If you have a FTP server running, what does the > message on the server log show? > > if after ls you do pwd, what does it show? try appending that to the > file name. Or copy myfile.dat to the default local dir. So that > eleminates the need for path prefixing to the file name. > > Random thoughts ... hope they help > > dinx > --------------------------- Newsgroups: comp.os.vxworks Subject: Unsupported telnet server compile error... Date: Mon, 9 Jul 2001 14:04:40 +0900 From: "±èÀÀ½Ä" Organization: Korea Telecom Message-ID: <9ibe3k$7cp$1@news1.kornet.net> I requested my FAE with regard to 'unsupported telnet server from windriver', and I've got the source. however I do not have a clue why following errors come out when compiling... /*************************************************************************** **************** * Compile error messages **************************************************************************** ****************/ E:/Tornado2/target/src/config/../../src/config/telnetLib.c: In function `tnInput': In file included from E:/Tornado2/target/src/config/usrExtra.c:217, from E:/Tornado2/target/config/all/usrConfig.c:123: E:/Tornado2/target/src/config/../../src/config/telnetLib.c:670: syntax error before `.' E:/Tornado2/target/src/config/../../src/config/telnetLib.c:675: `field' undeclared (first use this function) E:/Tornado2/target/src/config/../../src/config/telnetLib.c:675: (Each undeclared identifier is reported only once E:/Tornado2/target/src/config/../../src/config/telnetLib.c:675: for each function it appears in.) make[1]: *** [usrConfig.o] Error 1 I've looked for source code but there's no ``.'' or ``field'' variable. Plz share your how-to with me... ====================================== Eung-Shik Kim (Ed) ed@ebridgecom.com PowerVoIP Gateway Hardware Devl. Dept. 467-12 Dogok-Dong Kangnam-Gu MMAA Bldg 29th Fl. Seoul (zip 135-270) KOREA, Republic of Tel+82 2 575 6476 ext 517, 3713 Fax+82 2 575 6478 C.P. +82 11 9704 0840 ICQ#61658172 / Send me plain text please... ======================================= --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How old is VxWorks ? Date: Mon, 9 Jul 2001 09:59:01 +0200 From: "Jonathan ILIAS" Organization: University of Angers, France. Message-ID: <9ibo6e$10e7$1@news.univ-angers.fr> References: <9i3urd$2lg4$1@news.univ-angers.fr> <3B45FCAB.3FE95040@aps.anl.gov> <51f8f6c6.0107070906.41e0efaa@posting.google.com> > > > Does anyone know what is VxWorks birth date ? > > This isn't definitive but the first modification history entry in the > > header file vxWorks.h from vxWorks 4.0.2 is dated 15Feb83 (yes, we still > > modification history > > Anyways... yeah, the answer was '81. OK, thanks but there is something I don't understand. The software started in 1981 and came up to release 4.0.2 early in 1983 !!!??? That's a very fast development, isn't it ? --------------------------- Newsgroups: comp.os.vxworks Subject: driver for sh4 companion chip HD64465 Date: 9 Jul 2001 01:21:59 -0700 From: mnylund@mail.com (Martin Nylund) Organization: http://groups.google.com/ Message-ID: <4113af3.0107090021.7d4f1e44@posting.google.com> Hi, I'm doing research, which operating system should we introduce to our pc104 SH4 module. We are using the companion chip HD64465, and I'm interseted of the availability of VxWorks drivers for this chip. I would really appreciate if somebody could inform me where to look for these drivers if they exist. Thank You Martin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 5.4 release notes? Date: 9 Jul 2001 08:42:07 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9i7pan$n3d$1@news5.svr.pol.co.uk> "Archerfish" wrote: > I've searched Wind River's website for something approximating the > VxWorks 5.4 release notes, without success. Not even a press release. > Can anybody recommend upgrading from 5.3.1 to 5.4 ? As far as I know the main difference between 5.3.1 and 5.4 is that it changed from Tornado 1.0.1 to Tornado 2. There were probably some other changes also, but the compiler was the same. That will be changed from T2 to T2.1. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: 9 Jul 2001 08:39:23 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> "Joe Woodbury" wrote: > The questions are: Are our experiences unique? Did our company just > happen to run into the one group of bastards working at Wind River? My experience with WindRiver has been very good, no problems with delivery or other major problems. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to setup a PPP link through telephone connection. Date: 9 Jul 2001 08:51:43 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: gujc@huawei.com (Gu, Jiangchun) wrote: > Hi, everyone: > I want to setup a PPP link on target board running vxWorks v5.4, > but I don't know when, where and how to control the modem. The target > works as a server and waits the other peer to dial in. Have a task waiting for input on your serial line. When a "RING" comes in have the modem answer the call (or have it on auto-answer). When the "CONNECT" is seen startup the PPP-task. When the PPP-link drops reset or restart thet modem, and wait for the next call. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: missing timer_connect() and friends ... Date: 9 Jul 2001 08:49:56 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <3B46C989.C47931F5@yahoo.com> Julian Sun wrote: > > Nothing really special here, but when I tried to download my > downloadable app to the NTsim, it complains about missing > timer_connect(), timer_create() and timer_settime(), etc. You probably forgot to include the timer functions in your simulator. You can include it from the VxWorks tab in you workspace window. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: QLogic SCSI/FC Driver Date: Mon, 9 Jul 2001 11:09:01 +0200 From: "Carlos Sánchez" Organization: Sener Message-ID: <9ibs9n$a20$1@taliesin.netcom.net.uk> Sender: sanchez.garcia@sener.es Hello. I need to develop the driver for the Fiber Channel Host Bus Adapter QLogic2200. I´m intended to use the SCSI library as initiator, but I'm rather lost about what the Target involves. It will be very helpfull if someone will be able to provide me the source code of previous SCSI drivers like (ncr810Lib, ncr710Lib). Any kind of help will be appreciated. Regards Carlos. --------------------------- Newsgroups: comp.os.vxworks Subject: IPv6 stack Date: 9 Jul 2001 17:25:55 +0800 From: Message-ID: <3b4978a2@newsgate.hknet.com> Hi all, Anyone know anyone is working on the IPv6 on VxWorks? If I'm want to work on it, is it possible to get any toolkit to add the IPv6 protocol support on top of the current IPv4 stack? SK Cheung --------------------------- Newsgroups: comp.os.vxworks Subject: Absolute Linking Date: Mon, 9 Jul 2001 12:13:40 +0200 From: "Michael Lawnick" Organization: Buergernetz Dillingen Message-ID: <9ic0dt$9il$1@snoopy.bndlg.de> Reply-To: "Michael Lawnick" Sender: mlawnick@pd9e65644.dip.t-dialin.net Hi VxWorks Gurus, has anybody ever tried to link some code/data to absolute location in bootrom ? The background: We have a PPC405, whose H/W reset vector is 0xfffffffc. VxWorks bootrom starts at ROM_TEXT_ADRS in Makefile. Normally we edit the hex code, but now search for a 'make' solution. We tried to create a section, put the jump instruction to ROM_TEXT_ADRS in it and locate the section at 0xfffffffc. This shouldn't be too difficult, but we couldn't get it run by now. Has anybody done this already and can give me some advice ? - -- Mit freundlichen Grüßen, Michael Lawnick ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Absolute Linking Date: Mon, 9 Jul 2001 12:13:40 +0200 From: "Michael Lawnick" Organization: Buergernetz Dillingen Message-ID: <9ic0j4$9ji$1@snoopy.bndlg.de> Reply-To: "Michael Lawnick" Sender: mlawnick@pd9e65644.dip.t-dialin.net Hi VxWorks Gurus, has anybody ever tried to link some code/data to absolute location in bootrom ? The background: We have a PPC405, whose H/W reset vector is 0xfffffffc. VxWorks bootrom starts at ROM_TEXT_ADRS in Makefile. Normally we edit the hex code, but now search for a 'make' solution. We tried to create a section, put the jump instruction to ROM_TEXT_ADRS in it and locate the section at 0xfffffffc. This shouldn't be too difficult, but we couldn't get it run by now. Has anybody done this already and can give me some advice ? - -- Mit freundlichen Grüßen, Michael Lawnick ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 9 05:03:50 2001 From: deejoshi@hss.hns.com Date: Mon Jul 9 05:03:52 PDT 2001 Subject: a query Hello to VxWorks community This is my first question . I want to know that how can I trace a stack compiled by gcc for MIPS R4000, so that through a stack dump(of a task) I can have functions back trace with their arguments for this first I want to know the stack layout I'll be thankful if somebody can answer my question regards Deepak Joshi From vxwexplo-errs@csg.lbl.gov Mon Jul 9 06:25:34 2001 From: Geoffrey Espin Date: Mon Jul 9 06:25:37 PDT 2001 Subject: Re: a query Deepak, > I want to know that how can I trace a stack compiled by gcc for MIPS R4000, > so that through a stack dump(of a task) I can have functions back trace with > their arguments for this first I want to know the stack layout Generic trace function attached. Geoff -- -------------------------------------------------------------------------------- /* xxxTrace.c - user trace function */ /* DESCRIPTION There is no description of the API needed by trcStack() for hooking up your own dbgTrace functions... here it is! (Not compiled or tested recently.) */ #include "vxWorks.h" #include "logLib.h" /******************************************************************************* * * xxxTracePrint - stack trace print function * * RETURNS: OK or ERROR */ void xxxTracePrint ( INSTR *caller, int func, int nargs, int *args ) { char buf [250]; int ix; int len = 0; len += sprintf (&buf [len], "%#10x: %#10x (", (int)caller, func); for (ix = 0; ix < nargs; ix++) { if (ix != 0) len += sprintf (&buf [len], ", "); len += sprintf (&buf [len], "%#x", args [ix]); } len += sprintf (&buf [len], ")\n"); logMsg (buf, 0, 0, 0, 0, 0, 0); } /******************************************************************************* * * xxxTrace - stack trace * * RETURNS: OK or ERROR */ int xxxTrace ( WIND_TCB *pTcb ) { REG_SET regs; if (pTcb == NULL) return (ERROR); taskRegsGet ((int)pTcb, ®s); trcStack (®s, (FUNCPTR) xxxTracePrint, (int)pTcb); return (OK); } -------------------------------------------------------------------------------- From vxwexplo-errs@csg.lbl.gov Mon Jul 9 10:22:24 2001 From: "Gold, Phillip W" Date: Mon Jul 9 10:22:26 PDT 2001 Subject: Can target obtain time-of-day from PC? I am using Tornado on a PC with the target (a BAJA 750 PPC running VxWorks) connected via the target server. Does anyone know how to have the target obtain the time from the PC time-of-day clock? Thanks in advance. Phil Gold phillip.w.gold@lmco.com From vxwexplo-errs@csg.lbl.gov Mon Jul 9 12:21:44 2001 From: Rainer Stelzer Date: Mon Jul 9 12:21:46 PDT 2001 Subject: Re: Can target obtain time-of-day from PC? > I am using Tornado on a PC with the target (a BAJA 750 PPC running VxWorks) > connected via the target server. > Does anyone know how to have the target obtain the time from the PC > time-of-day clock? have a look at the SNTP client component delivered with vxworks and look for a free SNTP server for your PC's OS. regards Rainer ===================================================== Gebrüder Stoye GmbH department: traffic control systems Name: Rainer Stelzer Adr: Longericher Strasse 177 city: Cologne, Germany Zip: 50739 phone: ++49/221/2616-551 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== From vxwexplo-errs@csg.lbl.gov Mon Jul 9 23:15:35 2001 From: venkatk Date: Mon Jul 9 23:15:37 PDT 2001 Subject: Board Support: Greetings, Can anyone help where to buy the eval boards based on i960 with a bsp for vxworks. Regards From vxwexplo-errs@csg.lbl.gov Tue Jul 10 02:42:44 2001 From: prakashar@in.huawei.com Date: Tue Jul 10 02:42:47 PDT 2001 Subject: Task information This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01C10929.46F53A30 Content-Type: text/plain; charset="iso-8859-1" Hello, Is it possible to access the ready/suspended/blocked queue information maintained by the VxWorks scheduler? Is there any ready API which I can use in my program to get the aforementioned information? Also, in VxWorks, when a task is executing, how does the scheduler wakes up? Is it through a clock interrupt? Thanks in advance, Cheers, Prakash ------_=_NextPart_001_01C10929.46F53A30 Content-Type: text/html; charset="iso-8859-1" Task information

Hello,

Is it possible to access the ready/suspended/blocked
queue information maintained by the VxWorks scheduler?
Is there any ready API which I can use in my program to
get the aforementioned information?

Also, in VxWorks, when a task is executing, how does
the scheduler wakes up? Is it through a clock interrupt?

Thanks in advance,

Cheers,
Prakash

------_=_NextPart_001_01C10929.46F53A30-- From vxwexplo-errs@csg.lbl.gov Tue Jul 10 03:56:18 2001 From: "Mike Anderson" Date: Tue Jul 10 03:56:21 PDT 2001 Subject: RE: Board Support: VxWorks Greetings! > Submitted-by: venkatk > > Greetings, > > Can anyone help where to buy the eval boards based on i960 with a > bsp for vxworks. > You might try Cyclone Microsystems (http://www.cyclone.com). They have a number of i960 boards for PCI, cPCI, etc. that have VxWorks BSPs available for them. I've had good luck with their equipment in the past. HTH, Mike -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Tue Jul 10 04:03:35 2001 From: Vxworks Exploder Date: Tue Jul 10 04:03:37 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 10 04:03:21 PDT 2001 Subject: Inconsistency while mibcomp -man -trap.c Subject: a query Subject: Increase maximum line length of Target Shell Subject: Tornado Prototyper Question Subject: re: a query Subject: two 3com 3c905-TX Ethernet cards on the target machine Subject: Re: How old is VxWorks ? Subject: Re: Accessing Host files from VxSim? Subject: Re: Symbol Table Sync Problem Subject: Re: missing timer_connect() and friends ... Subject: Re: Increase maximum line length of Target Shell Subject: Can target obtain time-of-day from PC? Subject: SMC / PPC823 Subject: Help - Wind Web Server 2.0 Subject: Help - Is there any VxWorks discussion mail lists/discussion groups other than this one? Subject: Re: Defragmentation for TrueFFS Subject: Help - Does VxWorks's Wind Web Server support Java applets? Subject: Help - Which one is better, VxWorks or Windows CE? Subject: Re: Can target obtain time-of-day from PC? Subject: re: Can target obtain time-of-day from PC? Subject: Re: Tornado Prototyper Question Subject: Re: Tornado Prototyper Question Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Subject: Re: Defragmentation for TrueFFS Subject: Re: Help - Which one is better, VxWorks or Windows CE? Subject: Re: Can target obtain time-of-day from PC? Subject: Re: Increase maximum line length of Target Shell Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Subject: Re: Is Wind River really this pathetic Subject: Re: 5.4 release notes? Subject: Re: How to get patches Subject: MPC107 PCI bridge driver for MPC755 Subject: Board Support: Subject: Re: Board Support: Subject: Re: Absolute Linking Subject: Re: How to get patches Subject: DirLib and DosFS Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Subject: Re: Increase maximum line length of Target Shell Subject: Re: How to setup a PPP link through telephone connection. Subject: Task information ------------------------------------------------------- Newsgroups: comp.protocols.snmp,comp.os.psos,comp.os.vxworks Subject: Inconsistency while mibcomp -man -trap.c Date: Mon, 09 Jul 2001 14:52:04 +0200 From: Alex Vinokur Message-ID: <3B49A8F3.5CE751AC@bigfoot.com> ============================ pRISM+ for pSOSystem/PowerPC pRISM+ 1.2.3 . pSOSystem 2.2.2 Diab Data D-CC Optimizing C compiler SNMP Release 7.4 (SNMPv1) Epilogue Technology Emissary SNMP MIB Compiler, version 6.3 Host : Windows98 Target : PowerPC (MPC860) ============================ Files leaves.c and trap.c have been generated with mibcomp. mibcomp -man -o leaves.c mibcomp -trap.c -o trap.c //------- Fragment of leaves.c ------- OIDC_T ID_totalStarted[] = { 1, 3, 6, 1, 4, 1, 4466, 1, 5, 1, 1, 5, 0 }; unsigned int len_totalStarted = 12; //------------------------------------ //------- Fragment of trap.c --------- static OIDC_T ID_totalStarted[] = { 1, 3, 6, 1, 4, 1, 4466, 1, 5, 1, 1, 5, 0 }; static int len_totalStarted = 13; //------------------------------------ We can see that ID_totalStarted[] has the same value in both files. However len_totalStarted has different values in the same files. Is it a bug of mibcomp? ================================== Alex Vinokur mailto:alexvn@bigfoot.com http://up.to/alexvn http://go.to/alexv_math ================================== --------------------------- Newsgroups: comp.os.vxworks Subject: a query Date: Mon, 9 Jul 2001 17:37:00 +0530 From: deejoshi@hss.hns.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9ic78b$3qe$1@overload.lbl.gov> Hello to VxWorks community This is my first question . I want to know that how can I trace a stack compiled by gcc for MIPS R4000, so that through a stack dump(of a task) I can have functions back trace with their arguments for this first I want to know the stack layout I'll be thankful if somebody can answer my question regards Deepak Joshi --------------------------- Newsgroups: comp.os.vxworks Subject: Increase maximum line length of Target Shell Date: Mon, 9 Jul 2001 14:18:46 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994681619.314539@newsmaster-04.atnet.at> Hi VxWorkers does there exist a way to increase the maximum line length the target shell will accept? Currently it seems to have a hard-coded 128 character line limit. I want to run some class path initialization with a startup script, but this cannot be done with that limit. Any ideas? Thanks Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado Prototyper Question Date: Mon, 9 Jul 2001 14:22:04 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994681815.874726@newsmaster-04.atnet.at> Hi, Does anyone know, if Tornado Prototyper is able to run a target server and WindShell for a pre-made VxWorks image? The image runs on I80486 hardware and should be connected to the host with Ethernet connection. I just need to use the WindShell (and maybe the Browser for Spy Chart,...). Thanks Werner --------------------------- Newsgroups: comp.os.vxworks Subject: re: a query Date: Mon, 9 Jul 2001 06:25:31 -0700 From: Geoffrey Espin Organization: Lawrence Berkeley National Laboratory Message-ID: <9ice9b$6uq$1@overload.lbl.gov> Deepak, > I want to know that how can I trace a stack compiled by gcc for MIPS R4000, > so that through a stack dump(of a task) I can have functions back trace with > their arguments for this first I want to know the stack layout Generic trace function attached. Geoff - -- - -------------------------------------------------------------------------------- /* xxxTrace.c - user trace function */ /* DESCRIPTION There is no description of the API needed by trcStack() for hooking up your own dbgTrace functions... here it is! (Not compiled or tested recently.) */ #include "vxWorks.h" #include "logLib.h" /******************************************************************************* * * xxxTracePrint - stack trace print function * * RETURNS: OK or ERROR */ void xxxTracePrint ( INSTR *caller, int func, int nargs, int *args ) { char buf [250]; int ix; int len = 0; len += sprintf (&buf [len], "%#10x: %#10x (", (int)caller, func); for (ix = 0; ix < nargs; ix++) { if (ix != 0) len += sprintf (&buf [len], ", "); len += sprintf (&buf [len], "%#x", args [ix]); } len += sprintf (&buf [len], ")\n"); logMsg (buf, 0, 0, 0, 0, 0, 0); } /******************************************************************************* * * xxxTrace - stack trace * * RETURNS: OK or ERROR */ int xxxTrace ( WIND_TCB *pTcb ) { REG_SET regs; if (pTcb == NULL) return (ERROR); taskRegsGet ((int)pTcb, ®s); trcStack (®s, (FUNCPTR) xxxTracePrint, (int)pTcb); return (OK); } - -------------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: two 3com 3c905-TX Ethernet cards on the target machine Date: Mon, 9 Jul 2001 11:14:28 -0400 From: "Fangzhe Chang" Organization: Bell Labs Message-ID: <9ichai$jhm1@news.research.bell-labs.com> Reply-To: "Fangzhe Chang" Hi, I'm trying to install two 3COM 3c905-TX FastEthernet cards on my target machine (VxWorks 5.4). However, I could not get it recognize both of them. Could someone do me a favor and give some suggestion? The first card (i.e., elPci0) comes up by default. To install the second card, I add a line to the "endDevTbl" network device table, as {1, sysEl3c90xEndload, "", 1, NULL, FALSE} I modify the sysEl3c90xEndLoad() a little bit to detect the PCI devices and build the correct parameter strings. It would try to load the two devices with the following parameters (both when mulDevLoad() calls it the second time), el3c90xEndLoad("0:0xffffffff:0xee80:0x0:41:9:-1:-1:-1:0x0:0xffffffff"); el3c90xEndLoad("1:0xffffffff:0xef00:0x0:42:10:-1:-1:-1:0x0:0xffffffff"); Both of them returns the same END object 0x7ffec24. However, the system still does not know about second interface after this. My attempt to configure it using ifMaskSet("elPci1", exffffff00) and ifAddrAdd("elPci1", "172.16.1.10") fails; ifShow only tells about the first interface. Any help is greatly appreciated. Fangzhe Chang --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How old is VxWorks ? Date: 9 Jul 2001 09:39:11 -0700 From: espin@idiom.com (Geoffrey Espin) Organization: http://groups.google.com/ Message-ID: <51f8f6c6.0107090839.febb056@posting.google.com> References: <9i3urd$2lg4$1@news.univ-angers.fr> <3B45FCAB.3FE95040@aps.anl.gov> <51f8f6c6.0107070906.41e0efaa@posting.google.com> <9ibo6e$10e7$1@news.univ-angers.fr> "Jonathan ILIAS" wrote in message news:<9ibo6e$10e7$1@news.univ-angers.fr>... > > > > Does anyone know what is VxWorks birth date ? > > > This isn't definitive but the first modification history entry in the > > > header file vxWorks.h from vxWorks 4.0.2 is dated 15Feb83 (yes, we still > > > modification history > > Anyways... yeah, the answer was '81. > OK, thanks but there is something I don't understand. The software started > in 1981 and came up to release 4.0.2 early in 1983 !!!??? That's a very fast > development, isn't it ? In 1985 it was 3.21. 4.0.2 was more like 1989 or so. 3.x was just vrtx (pre vrtx). 4.x was wind1.0/psos/vrtx-32. 5.x was wind2.0 only. Geoff P.S. Sorry Don Brooks (dab) and Dana Bird (dzb). --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Accessing Host files from VxSim? Date: Mon, 9 Jul 2001 11:38:15 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9icmk10c87@enews2.newsguy.com> References: The best way is to use Target Server File System. When you use "host:c:\..." you are using passFS and there are problems with passFS. Something about the permission flags not matching up between VxWorks and NT..... "Mark" wrote in message news:c5b88987.0107051139.6f694df4@posting.google.com... > Is there a way to open and read a file from the c: drive from a VxSim program? > > 1) I tried using the full pathname "c:\thedir\myfile.dat" > 2) I also tried in windshell > cd c:\thedir > ls - myfile.dat yes file is there > fp = open("myfile.dat", 0, 0); > > neither worked > > thanks > Mark --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Symbol Table Sync Problem Date: Mon, 9 Jul 2001 11:52:07 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9icne10d9j@enews2.newsguy.com> References: <9i5gq1$2k8$1@overload.lbl.gov> > non-networked Windows NT host SymSync doesn't work without networking. Read the docs for symSyncLib. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: missing timer_connect() and friends ... Date: Mon, 9 Jul 2001 11:53:15 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9icng40dch@enews2.newsguy.com> References: <3B46C989.C47931F5@yahoo.com> The standard simulator doesn't have POSIX timers built in. You have to build a custom image. "Julian Sun" wrote in message news:3B46C989.C47931F5@yahoo.com... > > Nothing really special here, but when I tried to download my > downloadable app to the NTsim, it complains about missing > timer_connect(), timer_create() and timer_settime(), etc. > > Everything otherwise works fine (i.e., if I remove those functions the > app works fine.) > > Did I miss anything here? > > Thanks. > > Julian --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Increase maximum line length of Target Shell Date: Mon, 9 Jul 2001 11:55:17 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9icnjv0di6@enews2.newsguy.com> References: <994681619.314539@newsmaster-04.atnet.at> That maximum is hard coded. You would have to get WRS to recompile it for you. "Werner Schiendl" wrote in message news:994681619.314539@newsmaster-04.atnet.at... > Hi VxWorkers > > does there exist a way to increase the maximum line length the target shell > will accept? > Currently it seems to have a hard-coded 128 character line limit. > > I want to run some class path initialization with a startup script, but this > cannot be done with that limit. > > Any ideas? > > Thanks > Werner > > --------------------------- Newsgroups: comp.os.vxworks Subject: Can target obtain time-of-day from PC? Date: Mon, 09 Jul 2001 11:23:09 -0600 From: "Gold, Phillip W" Organization: Lawrence Berkeley National Laboratory Message-ID: <9icsbg$d8k$1@overload.lbl.gov> I am using Tornado on a PC with the target (a BAJA 750 PPC running VxWorks) connected via the target server. Does anyone know how to have the target obtain the time from the PC time-of-day clock? Thanks in advance. Phil Gold phillip.w.gold@lmco.com --------------------------- Newsgroups: comp.os.vxworks Subject: SMC / PPC823 Date: Mon, 09 Jul 2001 20:13:24 +0200 From: Heiko Schmidt Organization: T-Online Message-ID: <3B49F444.9324CC1D@t-online.de> Hello a question to the serial management controller (SMC) in the PPC823: If i receive a lot of datas (for ex. software downloading), after a short time the frameing bit (FR) in the status register of the RX buffer descriptor is set. After this, i think the received datas in the buffer are invalid. In the PPC documentation this error means, that a stop bit is missed. Does anyone know how this can happen and what can i do ? The baudrate is correct, i have checked it with an Oszi. This error only occurs if i get continously a lot of datas. The buffer is big enough and the software to copy the datas from the buffer is also fast enough. Our Baudrate is 38400, 8 data, no parity 1 stopbit. Thanks Heiko --------------------------- Newsgroups: comp.os.vxworks Subject: Help - Wind Web Server 2.0 Date: Mon, 9 Jul 2001 13:30:28 -0500 From: "Lee" Message-ID: <9ict86$htc1a$1@ID-92003.news.dfncis.de> Hi, I would be appreciate it if someone could help me on this, or give me a hint. We've been using Tornado 2.0 with Wind Web Server 1.0 (WWS1.0) for a while. Everything works OK. We are trying to upgrade WWS1.0 to WWS2.0, and find that WWS2.0 is not fully backward compatible with our code. After rebuild our code with WWS2.0, web server doesn't seem work any more. Does anyone have experience with WWS2.0? How can we make the migration from WWS1.0 to WWS2.0 smooth? Thank you in advance. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Help - Is there any VxWorks discussion mail lists/discussion groups other than this one? Date: Mon, 9 Jul 2001 13:32:11 -0500 From: "Lee" Message-ID: <9ictbd$e9mhr$1@ID-92003.news.dfncis.de> Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Mon, 09 Jul 2001 14:49:49 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B49FCCD.D96F7754@digidata.com> References: <9hqng4$49e$1@nntp1-cm.news.eni.net> I'm curious why you would want to defragment a flash file system. Defragentation is a performance enhancement in mechanical devices to help overcome rotational latency and seek delay. It would not apply in a solid state device. I'm interested because if there is some advantage then I also would be interested in a solution. Doug News Account wrote: > Does anybody know of any methodology or utility out there for defragmenting > a VxWorks DOS system that is stored and maintained via TrueFFS? Also I have > looked at the M-Systems TrueFFS code and noticed some sort of low level > defragmentation option that I believe is supposed to be accessable either > through the "tffsRawio" routine or some sort of "ioctl" call, however I was > unable to get this to work at all. At this point we would even be willing to > utilize a third party software to perform the defragmentation task. Please > help!!! > > Regards, > P. Eichler --------------------------- Newsgroups: comp.os.vxworks Subject: Help - Does VxWorks's Wind Web Server support Java applets? Date: Mon, 9 Jul 2001 13:54:03 -0500 From: "Lee" Message-ID: <9icukd$hppmb$1@ID-92003.news.dfncis.de> Hi, Could someone tell me if VxWorks' Wind Web Server 1.0 supports Java applets or not? I don't have full version of Wind Web Server 1.0 document. Thank you. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Help - Which one is better, VxWorks or Windows CE? Date: Mon, 9 Jul 2001 14:02:34 -0500 From: "Lee" Message-ID: <9icv4b$i1r35$1@ID-92003.news.dfncis.de> Hi, Could someone tell me the pros & cons of VxWorks and Windows CE. They both are real time operating systems. Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can target obtain time-of-day from PC? Date: 9 Jul 2001 19:25:42 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9icsbg$d8k$1@overload.lbl.gov> "Gold, Phillip W" wrote: > I am using Tornado on a PC with the target (a BAJA 750 PPC running > VxWorks) connected via the target server. > Does anyone know how to have the target obtain the time from the PC > time-of-day clock? Yes, have a look at the FAQ, there are some examples of that. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: re: Can target obtain time-of-day from PC? Date: Mon, 09 Jul 2001 21:26:06 +0200 From: Rainer Stelzer Organization: Lawrence Berkeley National Laboratory Message-ID: <9id3cb$fo8$1@overload.lbl.gov> > I am using Tornado on a PC with the target (a BAJA 750 PPC running VxWorks) > connected via the target server. > Does anyone know how to have the target obtain the time from the PC > time-of-day clock? have a look at the SNTP client component delivered with vxworks and look for a free SNTP server for your PC's OS. regards Rainer ===================================================== Gebrüder Stoye GmbH department: traffic control systems Name: Rainer Stelzer Adr: Longericher Strasse 177 city: Cologne, Germany Zip: 50739 phone: ++49/221/2616-551 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Prototyper Question Date: Mon, 09 Jul 2001 20:46:26 GMT From: "Merlin" Organization: [Infostrada] Message-ID: References: <994681815.874726@newsmaster-04.atnet.at> No it does not. You won' t be able to connect to your target . ( if this were not the case,why should you buy Tornado ? ) Bye "Werner Schiendl" ha scritto nel messaggio news:994681815.874726@newsmaster-04.atnet.at... > Hi, > > Does anyone know, if Tornado Prototyper is able to run a target server and > WindShell for a pre-made VxWorks image? > The image runs on I80486 hardware and should be connected to the host with > Ethernet connection. > I just need to use the WindShell (and maybe the Browser for Spy Chart,...). > > Thanks > Werner > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Prototyper Question Date: Mon, 9 Jul 2001 22:57:32 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994712734.371394@newsmaster-04.atnet.at> References: <994681815.874726@newsmaster-04.atnet.at> Thanks for your reply, "Merlin" wrote in message news:CKo27.2600$1u2.100657@news.infostrada.it... > No it does not. > You won' t be able to connect to your target . ( if this were not the > case,why should you buy Tornado ? ) > Why? Hm... not a bad argument ;-) Anyway you can do a lot of things with the target resident shell and it would not harm to be able to do the same things with Tornado Prototyper. They could add a passus to the licence that you may not use the prototyper for production use. Thanks again Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Date: Mon, 9 Jul 2001 23:01:34 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994712976.913719@newsmaster-04.atnet.at> References: <9icukd$hppmb$1@ID-92003.news.dfncis.de> Hi, Applets run in the browser, so if you are able to transfer files of arbitrary mime type, applets should work. If you are thinking about servlets, I think the answer is no. However, WindRiver has a Java product too, and maybe the products can be combined to run servlets (don't know any details about this). If you want to be able to run (a subset of) JavaScript (not Java) on the server, you could have a look at the GoAhead Webserver (www.goahead.com) hth werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Mon, 9 Jul 2001 17:07:25 -0400 From: "Kevin Livingston" Message-ID: References: <9hqng4$49e$1@nntp1-cm.news.eni.net> <3B49FCCD.D96F7754@digidata.com> remember it also saves space too... if you have a lot of half used juked up sectors etc. if your drive is only 32 meg or something and you do a lot of IO it _could_ help you, then again it could do nothing too. you should do tests in your specific app to find out for sure how useful it would be. Kevin "Doug Dotson" wrote in message news:3B49FCCD.D96F7754@digidata.com... > I'm curious why you would want to defragment a flash file system. > Defragentation is a performance enhancement in mechanical > devices to help overcome rotational latency and seek delay. It > would not apply in a solid state device. I'm interested because > if there is some advantage then I also would be interested in a > solution. > > Doug > > > News Account wrote: > > > Does anybody know of any methodology or utility out there for defragmenting > > a VxWorks DOS system that is stored and maintained via TrueFFS? Also I have > > looked at the M-Systems TrueFFS code and noticed some sort of low level > > defragmentation option that I believe is supposed to be accessable either > > through the "tffsRawio" routine or some sort of "ioctl" call, however I was > > unable to get this to work at all. At this point we would even be willing to > > utilize a third party software to perform the defragmentation task. Please > > help!!! > > > > Regards, > > P. Eichler > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help - Which one is better, VxWorks or Windows CE? Date: Mon, 9 Jul 2001 23:03:23 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994713087.356574@newsmaster-04.atnet.at> References: <9icv4b$i1r35$1@ID-92003.news.dfncis.de> Hi, IIRC they both are embedded systems, but Windows CE is not a real time OS. It really depends on your app, which one will fit better. Maybe you can provide more details, what is of particular interest for you. regards werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can target obtain time-of-day from PC? Date: 9 Jul 2001 14:45:45 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107091345.77e40454@posting.google.com> References: <9icsbg$d8k$1@overload.lbl.gov> Hello, You could use SNTP to get the time from the PC (or any other machine), as long as the host machine is running an SNTP server. There are some commercially available SNTP servers for Windows (most Unix machines come with a network time protocol daemon, ntpd, so if you have a unix box anywhere that could act as your time server for free). Alternatively, as long as the target has access to the internet, you could use one of the following: ntp2.usno.navy.mil at 192.5.41.209 tock.usno.navy.mil at 192.5.41.41 Both of these are operated by the U.S. Naval Observatory, so they're probably pretty accurate ;-) For VxWorks, you will need to include the SNTP client code, and then use sntpcTimeGet() to obtain the time, and clock_settime(), or one of the other clock setting functions, to set the target's idea of real time. HTH, John... "Gold, Phillip W" wrote in message news:<9icsbg$d8k$1@overload.lbl.gov>... > I am using Tornado on a PC with the target (a BAJA 750 PPC running VxWorks) > connected via the target server. > Does anyone know how to have the target obtain the time from the PC > time-of-day clock? > > Thanks in advance. > > Phil Gold > phillip.w.gold@lmco.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Increase maximum line length of Target Shell Date: Mon, 09 Jul 2001 16:04:33 -0600 From: Don Small Organization: Sandia National Laboratories Message-ID: <3B4A2A71.B2A58EB3@sandia.gov> References: <994681619.314539@newsmaster-04.atnet.at> I'm not sure what you need to do exactly but I've had a similar problem. I put some of the longer paths into a string and then passed the string as the variable on the command line. This allowed us to get by the 128 character limit. Also consider setting environment variables. The first is easier. path = "/usr/local/wind" function1 (path) Hope this helps! Don Werner Schiendl wrote: > > Hi VxWorkers > > does there exist a way to increase the maximum line length the target shell > will accept? > Currently it seems to have a hard-coded 128 character line limit. > > I want to run some class path initialization with a startup script, but this > cannot be done with that limit. > > Any ideas? > > Thanks > Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Date: Mon, 9 Jul 2001 17:25:10 -0500 From: "Lee" Message-ID: <9idb08$ibsgk$1@ID-92003.news.dfncis.de> References: <9icukd$hppmb$1@ID-92003.news.dfncis.de> <994712976.913719@newsmaster-04.atnet.at> Hi, Thank you for the info. Here is more info on the issues we see. A firmware (*.dlx file) contains Tornado 2.0, Wind Web Server 1.0, application files, *.html files and jpg files. It's uploaded into a flash memory on the target system. We can use web browsers on client systems to access the web server on the target system and show the web pages and jpg files. This means both *.html files and jpg files are downloaded OK from the target system. Now, we add a Java applet onto a web page. The applet is very simple and has only a button. Its size is about 800bytes. When we access the *.html file which contains the applet, we only see a gray box showing on the browser. Browser's Java console shows an error message "Truncated class file". This means (I think) web browser does find the Java applet (a class file) on the target system, but the web server does not send it as a whole piece to web browser. We just couldn't figure out why. I would appreciate it if someone could help. Lee "Werner Schiendl" wrote in message news:994712976.913719@newsmaster-04.atnet.at... > Hi, > > Applets run in the browser, so if you are able to transfer files of > arbitrary mime type, applets should work. > > If you are thinking about servlets, I think the answer is no. > However, WindRiver has a Java product too, and maybe the products can be > combined to run servlets (don't know any details about this). > > If you want to be able to run (a subset of) JavaScript (not Java) on the > server, you could have a look at the GoAhead Webserver (www.goahead.com) > > hth > werner > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Mon, 09 Jul 2001 23:11:51 GMT From: "Joe Kerkes" Organization: RoadRunner - TampaBay Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> Joe, They are all bastards if you ask me! We also have had a lot of problems. Order mistakes, license keys, bsp and eval boards we have had problems with them all! Joe Kerkes "Joe Woodbury" wrote in message news:6qH17.79$5f6.133962@news.uswest.net... > Since we ordered VxWorks/Tornado II, we've had nothing but problems with > Wind River. They filled the order incorrectly and I had to resort to a lot > of yelling and threats to not pay our purchase order to get their attention. > Yet, two days later, they still have yet to do anything but email the > correct BSP. On top of all this, the highest level manager at Wind River > kept blaming us for some of the problems. > > So far, I've concluded that Wind River employees are given an IQ test. If > they fail, they are given a personal relationship test. If they fail that > also, they are hired. > > And heaven forbid they actually do something so obvious like creating a > compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. > > The shame is that VxWorks proper is such an excellent product. > > The questions are: Are our experiences unique? Did our company just happen > to run into the one group of bastards working at Wind River? > > Joe Woodbury > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 5.4 release notes? Date: 10 Jul 2001 01:23:38 GMT From: stnchris@xmission.com (Steve Christensen) Organization: XMission Internet, LLC Message-ID: References: <9i7pan$n3d$1@news5.svr.pol.co.uk> Reply-To: stnchris@xmission.com On 9 Jul 2001 08:42:07 GMT, Johan Borkhuis wrote: >"Archerfish" wrote: > >> I've searched Wind River's website for something approximating the >> VxWorks 5.4 release notes, without success. Not even a press release. >> Can anybody recommend upgrading from 5.3.1 to 5.4 ? > >As far as I know the main difference between 5.3.1 and 5.4 is that it >changed from Tornado 1.0.1 to Tornado 2. There were probably some other >changes also, but the compiler was the same. That will be changed from T2 >to T2.1. The compiler is not quite the same. Specifically, the GCC version with Tornado 2 / VxWorks 5.4 for PPC has a '-mlongcall' option. The '-mlongcall' option is available for Tornado1/VxW 5.3 as a patch. I guess that point is pretty nit-picky, as you could consider the compiler that comes bundled with Tornado 2 / VxWorks 5.4 to just have the 5.3 patches applied to it. With regard to other differences between 5.3 and 5.4, in addition to the IDE changes Johan noted (Tornado 1 -> Tornado 2), there were significant changes to the network stack between 5.3 and 5.4. - -Steve - -- Steve Christensen | "It's like getting a scalp massage from Lucifer" stnchris@xmission.com | -- Space Ghost, SGCC --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Fri, 6 Jul 2001 09:11:09 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <9ie3si$ngv$1@at-vie-newsmaster01.nextra.at> References: on the windsurf homepage there are (in one of the first lines) some buttons home find manage download .... click on download and then on patches. if you cant find this, maybe you havent enough rights on windsurf and then you have to click on contact. - -- regards, Mario Semo. http://www.kirchnersoft.com Joe Woodbury wrote in message news:c8c13723.0107051237.4700018@posting.google.com... > I see suggestions on this board to check Wind River Systems Windsurf > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > I am a legitimate customer, have a customer ID and access to WindSurf > but I am unable to find either of these things. What am I doing wrong? > > Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: MPC107 PCI bridge driver for MPC755 Date: 9 Jul 2001 22:56:34 GMT From: afernando@adventnetworks.com Organization: NewsOne.Net - Free Usenet News via the Web - http://newsone.net/ Message-ID: <9idcr2$3m0$1@news.netmar.com> Followup-To: comp.os.vxworks Hi All, In our board, the MPC107 pci bridge has many devices (82559ER Intel eth, HLDC FREEDM32, uart, fpga etc..) My question for vxworks, the pciAutoConfig() routine is it enough to config the MPC107 bridge? we have ethernet mux driver, trying get a hdlc driver from pmc-seirra, etc. Or where can i get MPC107 config s/w? In general, we would need the PCI configuration s/w for MPC107, and for each device on the PCI bus, we would need a driver. I am bit new to bsp type work, so any advice is helpful. many thanks. ----- Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web ----- http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups NewsOne.Net prohibits users from posting spam. If this or other posts made through NewsOne.Net violate posting guidelines, email abuse@newsone.net --------------------------- Newsgroups: comp.os.vxworks Subject: Board Support: Date: Tue, 10 Jul 2001 12:02:31 +0530 From: venkatk Organization: Lawrence Berkeley National Laboratory Message-ID: <9ie6hb$53v$1@overload.lbl.gov> Greetings, Can anyone help where to buy the eval boards based on i960 with a bsp for vxworks. Regards --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Board Support: Date: 10 Jul 2001 07:11:12 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9ie6hb$53v$1@overload.lbl.gov> venkatk wrote: > Greetings, > > Can anyone help where to buy the eval boards based on i960 with a bsp > for vxworks. Check out this page: http://www.wrs.com/cgi- bin/corporate/tradeWinds/navProd.cgi?category=hardware§ion=4 or this page for more info: http://www.wrs.com/corporate/html/twintro.html. Or ask your sales rep or FAE for information on supported BSP's. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Absolute Linking Date: Tue, 10 Jul 2001 07:31:00 GMT From: Diego Serafin Organization: http://www.newsranger.com Message-ID: References: <9ic0j4$9ji$1@snoopy.bndlg.de> In article <9ic0j4$9ji$1@snoopy.bndlg.de>, Michael Lawnick says... > >Hi VxWorks Gurus, > >has anybody ever tried to link some code/data to absolute location in >bootrom ? > >The background: >We have a PPC405, whose H/W reset vector is 0xfffffffc. >VxWorks bootrom starts at ROM_TEXT_ADRS in Makefile. >Normally we edit the hex code, but now search for a 'make' solution. > >We tried to create a section, put the jump instruction to ROM_TEXT_ADRS in >it and locate the section at 0xfffffffc. This shouldn't be too difficult, >but we couldn't get it run by now. > >Has anybody done this already and can give me some advice ? I don't know how PPC405 is fetching the first instruction from reset vector, but here it works on my MPC860: The reset vector is at 0x100 and, at reset, the CPU holds ROM CS active, i.e. when, at reset, CPU is fetching from 0x100 it really fetches at FFF00100 where ROM is located. Code at FFF00100 is the VxWorks entry point (_romInit) linked by ldppc at (for my image)... address RAM_LOW_ADRS (for me 0x100100). The solution to make all this mess running is to write the first routines (romInit and romStart) as PIC (Position Independent Code), all jumps must be Program Counter Relative. I try to be more clear: - - First istruction is fetched from 0x100 but ROM CS is selected and higher address lines are not connected so CPU is really fetching from ROM, at offset 0x100 from its beginning (set for my target at FFF00000). - - First istruction is a relative jump to _cold label for cold boot and then an absolute jump is computed as relative to ROM_TEXT_ADRS to go straight into ROM space (jump to _start label is computed subtracting _romInit from _start and then adding ROM_TEXT_ADRS to find the offset in ROM space where _start label resides) - - Then setup register and all the reboot stuff in assemply and jump to _romStart with proper calculation of its address (same as above) - - romStart will copy ROM into RAM and, finally, will jump to RAM at code loaded at proper address (RAM_HIGH_ADRS for bootRom if I remember well) freeing programmer from being PIC (!) . Take a look at "Tornado BSP Developer's Kit for VxWorks User's guide" for further details ... HTH, Ciao, Diego. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Sun, 08 Jul 2001 11:29:20 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B48A680.91FFA33C@covad.net> References: <9i3lif$4pg$1@news.hdpp.de> <9i3m4s$4r8$1@news.hdpp.de> <3B47DE28.A61D12DB@covad.net> Reply-To: drdiags@covad.net Joe, Let me try this again. First of all, you have the right after investing thousands/millions of dollars in a product to have it perform out of the box the way you expected it to. This is a given. It is not too much to expect that product A built to support architecture B with added components C-F, made by the same company, to install on support Host G and build for architecture B. That it doesn't raises concerns about the quality of what you have based your design around. Asking the supplier to rectify this shouldn't be met with excuses but action. I would hope this is not a systematic problem with Wind River but just a case of getting caught in unnecessary red tape. My poorly worded pun was an attempt to remove the justifiable emotions attached to your present situation by trying to have you step back and see how you can get what you need without getting more frustrated. The reality is that what you are experiencing is not unique and should concern WRS. I have seen many posts on this newsgroup by people who have had similiar experiences and feelings about how they are treated by WRS. When enough folks have had enough, then the company will attempt to find out why they are loosing out to other vendors. After many surveys, they may discover that the problem was not the product line, but the quality and services attached to them. By then, it will be too late to reverse the tide of negative experiences and contempt built up over time so some of the same folks who don't seem to care will be working for a competitor and possibly soiling their reputation for customer care. Bottom line, until I could find a replacement for what WRS is providing, I would attempt to hold their feet to the fire by taking the fight up to the highest level. Joe Woodbury wrote: > > Unfortunately, my sales person is on vacation. > > Maybe you don't produce bad code, but I produce it all the time. Except I > take responsibility for my mistakes and don't treat my customers like sh*t. > (Yes, I know you were being sarcastic.) - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: DirLib and DosFS Date: Tue, 10 Jul 2001 11:22:44 +0200 From: Emmanuel Herbreteau Organization: SEPRO Robotique (BEP) Message-ID: <3B4AC964.42D96941@sepro-robotique.com> Hi vxWorkers, I have some troubles using the "dirLib" with DOS file system (1.0) and ATA driver on PPC 860 (dirLib is used to scan a directory tree recursively). My code is : - ----------------- DIR *myDir; struct dirent *myDirEnt; struct stat s; char name[256]; myDir=opendir("/pcmcia/"); if (myDir==NULL) return(ERROR); FOREVER { myDirEnt=readdir(myDir); if (myDirEnt==NULL) break; strcpy(name,"/pcmcia/"); strcat(name,(myDir->dd_dirent).d_name); stat(name,&s); if (S_ISDIR(s.st_mode)) ....... - ----------------- It seems to work... But sometimes, I get strange file names like : name = "/pcmcia/BOOT/Av." or : name = "/pcmcia/UGL/Ap." (It's always "A*.") The stat function return s.st_mode=0x41FF But, there is no directory or file name with this name ! Any ideas ??? - -- Regards Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Date: Tue, 10 Jul 2001 11:29:27 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994757941.260795@newsmaster-04.atnet.at> References: <9icukd$hppmb$1@ID-92003.news.dfncis.de> <994712976.913719@newsmaster-04.atnet.at> <9idb08$ibsgk$1@ID-92003.news.dfncis.de> I would first check if the mime type of the file is transmitted correctly. Probably the WindWebServer has some possibility to set the mime types for files to be transmitted. A wrong mime type could cause the browser (or maybe even the server) to treat it in a wrong way. If that does not seem to be the problem, try to download the applet (not the HTML file that contains it, since this seems to download ok) with a download manager like GetRight (or with a browser that just lets you save the file to disk). Compare the files (the original you put in the ROM and what you downloaded). Then you'll see what went wrong. Maybe you got some LF<->CRLF translation, maybe the file is truncated on a special character. Anyway this should lighten the problem. If you find out what's wrong, please post your findings hth werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Increase maximum line length of Target Shell Date: Tue, 10 Jul 2001 11:33:07 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994758069.549597@newsmaster-04.atnet.at> References: <994681619.314539@newsmaster-04.atnet.at> <3B4A2A71.B2A58EB3@sandia.gov> Hi Don, thanks for your reply. Allocating a buffer and strcpy()ing / strcat()ing the path is in fact what I ended with by now. However, this is not the most elegant way to handle the problem. Anyway it is for evaluation only and thus its fine for now. Thanks Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to setup a PPP link through telephone connection. Date: 10 Jul 2001 03:01:56 -0700 From: gujc@huawei.com (Gu, Jiangchun) Organization: http://groups.google.com/ Message-ID: References: "Johan Borkhuis" wrote in message news:... > gujc@huawei.com (Gu, Jiangchun) wrote: > > > Hi, everyone: > > I want to setup a PPP link on target board running vxWorks v5.4, > > but I don't know when, where and how to control the modem. The target > > works as a server and waits the other peer to dial in. > > Have a task waiting for input on your serial line. When a "RING" comes in > have the modem answer the call (or have it on auto-answer). When the > "CONNECT" is seen startup the PPP-task. When the PPP-link drops reset or > restart thet modem, and wait for the next call. > > Groeten, > Johan Thanks a lot for your help! I'm a newbie, and I have some more questions to ask. I use SCC port on MPC8260 to support PPP. Should I use standard IO routines such as fopen, fgetc/read and fputc/write to communicate with the modem? SCC doesn't control DTR,DSR and RI signals on serial port. Need I write codes to control these signals? How can I detect dropping of PPP link? Gu, Jiangchun --------------------------- Newsgroups: comp.os.vxworks Subject: Task information Date: Tue, 10 Jul 2001 15:15:43 +0500 From: prakashar@in.huawei.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9iekjb$cbl$1@overload.lbl.gov> This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_001_01C10929.46F53A30 Content-Type: text/plain; charset="iso-8859-1" Hello, Is it possible to access the ready/suspended/blocked queue information maintained by the VxWorks scheduler? Is there any ready API which I can use in my program to get the aforementioned information? Also, in VxWorks, when a task is executing, how does the scheduler wakes up? Is it through a clock interrupt? Thanks in advance, Cheers, Prakash - ------_=_NextPart_001_01C10929.46F53A30 Content-Type: text/html; charset="iso-8859-1" Task information

Hello,

Is it possible to access the ready/suspended/blocked
queue information maintained by the VxWorks scheduler?
Is there any ready API which I can use in my program to
get the aforementioned information?

Also, in VxWorks, when a task is executing, how does
the scheduler wakes up? Is it through a clock interrupt?

Thanks in advance,

Cheers,
Prakash

- ------_=_NextPart_001_01C10929.46F53A30-- --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 10 04:17:40 2001 From: "Mike Anderson" Date: Tue Jul 10 04:17:42 PDT 2001 Subject: RE: Task information VxWorks Greetings! > Submitted-by: prakashar@in.huawei.com > > > Is it possible to access the ready/suspended/blocked > queue information maintained by the VxWorks scheduler? > Is there any ready API which I can use in my program to > get the aforementioned information? > > Also, in VxWorks, when a task is executing, how does > the scheduler wakes up? Is it through a clock interrupt? > There are many ways to access this info, but you don't mention what sort of info you need. Do you need info from the TCBs of tasks, or do you want to manipulate the queues themselves? taskIdListGet() will get a list of all of the tasks that are currently in the system. Then taskInfoGet(...) will return the task descriptor information that will describe the current state (i.e., what queue is it in?) of the task. windTcb(...) will return a pointer to the TCB itself. You can change priorities, move to/from the suspend queue, etc. All of this is in taskLib. There you will also find the taskSwitchHook... calls that allow your code to be called on context switches. The scheduler wakes up in one of several cases. 1) The currently executing task calls something that requires kernel intervention such as a semTake or taskDelay. 2) The scheduler always gets called after the return from ISR context. 3) After an exception has occured 4) There is another that I'm missing, but I can't recall what. HTH, Mike -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." From vxwexplo-errs@csg.lbl.gov Tue Jul 10 07:18:36 2001 From: prakashar@in.huawei.com Date: Tue Jul 10 07:18:38 PDT 2001 Subject: RE: Task information This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01C1094F.D0E2DEC0 Content-Type: text/plain; charset="iso-8859-1" Hello Mike, Thanks for the information. I am mainly looking for the order in which each task is put in the ready/ suspended/blocked queues. Is it possible for me to get this information ? Thanks in advance, Cheers, Prakash -----Original Message----- From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] Sent: Tuesday, July 10, 2001 5:02 PM To: vxworks_users@csg.lbl.gov Subject: RE: Task information > Submitted-by: prakashar@in.huawei.com > > > Is it possible to access the ready/suspended/blocked > queue information maintained by the VxWorks scheduler? > Is there any ready API which I can use in my program to > get the aforementioned information? > > Also, in VxWorks, when a task is executing, how does > the scheduler wakes up? Is it through a clock interrupt? > There are many ways to access this info, but you don't mention what sort of info you need. Do you need info from ------_=_NextPart_001_01C1094F.D0E2DEC0 Content-Type: text/html; charset="iso-8859-1" RE: Task information

Hello Mike,

Thanks for the information. I am mainly looking for
the order in which each task is put in the ready/
suspended/blocked queues. Is it possible for me to
get this information ?

Thanks in advance,

Cheers,
Prakash

-----Original Message-----
From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov]
Sent: Tuesday, July 10, 2001 5:02 PM
To: vxworks_users@csg.lbl.gov
Subject: RE: Task information

<snip>

> Submitted-by: prakashar@in.huawei.com
>
>
> Is it possible to access the ready/suspended/blocked
> queue information maintained by the VxWorks scheduler?
> Is there any ready API which I can use in my program to
> get the aforementioned information?
>
> Also, in VxWorks, when a task is executing, how does
> the scheduler wakes up? Is it through a clock interrupt?
>

  There are many ways to access this info, but you don't
mention what sort of info you need.  Do you need info from

<snip>

------_=_NextPart_001_01C1094F.D0E2DEC0-- From vxwexplo-errs@csg.lbl.gov Tue Jul 10 08:13:38 2001 From: Diego Novellon Date: Tue Jul 10 08:13:40 PDT 2001 Subject: RE: two 3com 3c905-TX Ethernet cards on the target machine > -----Mensaje original----- > De: Vxworks Exploder [SMTP:vx_explo@csg.lbl.gov] > Enviado el: martes 10 de julio de 2001 13:18 > Para: vxworks_users@csg.lbl.gov > Asunto: Vx Exploder Digest > > Newsgroups: comp.os.vxworks > Subject: two 3com 3c905-TX Ethernet cards on the target machine > Date: Mon, 9 Jul 2001 11:14:28 -0400 > From: "Fangzhe Chang" > Organization: Bell Labs > Message-ID: <9ichai$jhm1@news.research.bell-labs.com> > Reply-To: "Fangzhe Chang" > > Hi, > > I'm trying to install two 3COM 3c905-TX FastEthernet cards on my > target machine (VxWorks 5.4). However, I could not get it recognize both of > them. Could someone do me a favor and give some suggestion? > > The first card (i.e., elPci0) comes up by default. To install the > second card, I add a line to the "endDevTbl" network device table, as > {1, sysEl3c90xEndload, "", 1, NULL, FALSE} > > I modify the sysEl3c90xEndLoad() a little bit to detect the PCI > devices and build the correct parameter strings. It would try to load the two > devices with the following parameters (both when mulDevLoad() calls it the > second time), el3c90xEndLoad("0:0xffffffff:0xee80:0x0:41:9:-1:-1:- > 1:0x0:0xffffffff"); el3c90xEndLoad("1:0xffffffff:0xef00:0x0:42:10:-1:-1:- > 1:0x0:0xffffffff"); > > Both of them returns the same END object 0x7ffec24. However, the > system still does not know about second interface after this. My attempt to > configure it using ifMaskSet("elPci1", exffffff00) and ifAddrAdd > ("elPci1", "172.16.1.10") fails; ifShow only tells about the first interface. > > Any help is greatly appreciated. > > Fangzhe Chang > Hi, I remember having modified sysEl3c90xEnd.c as well to fix a couple of bugs in sysEl3c90xPciInit() and sysEl3c90xEndLoad(). I don't have the code in front of me right now, but according to my notes it was something like changing: * in sysEl3c90xPciInit(): 'netEndUnits' to '0' as a parameter to pciFindDevice() pEl3c90xActualRsc[unit] to pEl3c90xActualRsc[netEndUnits] * in sysEl3c90xEndLoad(): Use the unit number instead of '0' as an index to the pEl3c90xActualRsrc[] array. I guess it from pParamStr. If you're interested I can probably post tomorrow the actual code. Regards, ///////////////////////////////////////////////// // Diego Novellón Latre, Computer Engineer // // TETRA Infr., R&D Dept, Teltronic S.A.U. // // Polígono Malpica, C/ F-Oeste, parcela 12 // // E50057 Zaragoza - Spain // // Tel +34.976.465656 #174, fax +34.976.465722 // // http://www.teltronic.es/ // ///////////////////////////////////////////////// From vxwexplo-errs@csg.lbl.gov Tue Jul 10 09:30:57 2001 From: Chandwani Date: Tue Jul 10 09:30:59 PDT 2001 Subject: memScope Startup Errors..... Hi All !!, I am running Tornado 2.0 on Sun Solaris 2.5.1 host and the target is 8260 based Single Board Computer. When I click on the memScope button on the Tornado launch window ( once I am connected to the target ), I get following errors. I have tried all configurations but couldn't make ot work. Can someone help me : ld.so.1:/home/vxtools/vxworks/rti/scopetools.2.0b/Java/INSTALLED/jre/sparcSol2.5/bin/../bin/sparc/native_threads/java: fatal: relocation error: file /usr/openwin/lib/locale/en_US/XLIBI18N/xlibi18n.so.1: symbol _XlcDefaultInitLocale: referenced symbol not found Rakesh (rexrc@lucent.com) From vxwexplo-errs@csg.lbl.gov Tue Jul 10 12:02:18 2001 From: "Chang, Mark" Date: Tue Jul 10 12:02:20 PDT 2001 Subject: ftpd lockup problems vxworks 5.3.1 Our FTPd is locking up after repeated (about 10 in row) quick DIRs. Is this a known bug? Maybe increasing the mbufs? I know the data connection is probably not closing fast enough, but should I shorten the timeout for that? Advice appreciated. Thanks, Mark ---------------------------------------------------------- Mark Chang Senior Engineer Network Management Broadband Access and Transport Group ADC Wireline Systems Division Mark_Chang@ADC.com (714)730-2827 Learn about ADC - The Broadband Company at www.adc.com From vxwexplo-errs@csg.lbl.gov Wed Jul 11 04:03:32 2001 From: Vxworks Exploder Date: Wed Jul 11 04:03:34 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 11 04:03:08 PDT 2001 Subject: Re: How to setup a PPP link through telephone connection. Subject: RE: Board Support: Subject: Tools questions. Subject: Re: Is Wind River really this pathetic Subject: Re: TFFS limitations? -- No limitition, here's what you do... Subject: RE: Task information Subject: Problem of reading of the serial port Subject: Re: DirLib and DosFS Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Subject: Re: TFFS limitations? -- No limitition, here's what you do... Subject: Re: Defragmentation for TrueFFS Subject: vxWorks dosFs2.0 debug message Subject: Re: Defragmentation for TrueFFS Subject: how does stack work in vxWorks on PPC Subject: RE: Task information Subject: RE: Task information Subject: RE: two 3com 3c905-TX Ethernet cards on the target machine Subject: how to use vxsim? Subject: build boot disk for Pentium with SCSI hard disk driver Subject: memScope Startup Errors..... Subject: Re: Is Wind River really this pathetic Subject: Re: build boot disk for Pentium with SCSI hard disk driver Subject: ftpd lockup problems Subject: WDB Exception: 256 Subject: Re: Tools questions. Subject: Re: standalone_vxworks with no network Subject: HDLC and LABD questions? on MPC8260 Subject: Re: Is Wind River really this pathetic Subject: Re: HDLC and LABD questions? on MPC8260 Subject: Does BOOTROM always load the first file in TFFS, when I want to startup from TFFS? Subject: Re: 2 Rational Apex questions...#2 Subject: Re: Defragmentation for TrueFFS Subject: Re: Does BOOTROM always load the first file in TFFS, when I want to startup from TFFS? Subject: Re: how to use vxsim? Subject: Re: ftpd lockup problems Subject: I wanna change mbuf configuration. Subject: Re: how to use vxsim? Subject: Re: Accessing Host files from VxSim? Subject: Re: Absolute Linking Subject: tNetTask Subject: Re: Problem of reading of the serial port Subject: Re: I wanna change mbuf configuration. Subject: Re: tNetTask Subject: Help:Request Intel TE28F320C3 driver. ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to setup a PPP link through telephone connection. Date: 10 Jul 2001 11:05:17 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: gujc@huawei.com (Gu, Jiangchun) wrote: > I use SCC port on MPC8260 to support PPP. Should I use standard IO > routines such as fopen, fgetc/read and fputc/write to communicate with > the modem? Yes, just open the port and then you can read and write to it. After that you close it and give the same port name to the pppInit call. > SCC doesn't control DTR,DSR and RI signals on serial port. Need I > write codes to control these signals? > How can I detect dropping of PPP link? The standard serial drivers do not support these signals. If you need these you should modify the device driver for the serial port. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Board Support: Date: Tue, 10 Jul 2001 06:55:11 -0400 From: "Mike Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9ieo3r$e9h$1@overload.lbl.gov> VxWorks Greetings! > Submitted-by: venkatk > > Greetings, > > Can anyone help where to buy the eval boards based on i960 with a > bsp for vxworks. > You might try Cyclone Microsystems (http://www.cyclone.com). They have a number of i960 boards for PCI, cPCI, etc. that have VxWorks BSPs available for them. I've had good luck with their equipment in the past. HTH, Mike - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: Tools questions. Date: 10 Jul 2001 04:16:47 -0700 From: dlong@wvintenltd.com (Darren Long) Organization: http://groups.google.com/ Message-ID: <5c34b056.0107100316.10f30da0@posting.google.com> My experience with the Windriver sales people was that they were overwhelmed with the sheer number of their own and third party tools available and they could not seem to recommend a well specified but useful set of tools for use to use. For example, it looked like they were suggesting that we bought 3 debuggers. I would like comments on the following issues to see if I can work out a useful, minimally complete set of tools to order. Our project is based on the MPC555 PowerPC processor. We code exclusively in C++ and want to use full ANSI C++ language, including namespaces, exceptions, C++ standard library (STL etc.) and RTTI. 1)Does the Windriver version of the GNU compiler support these features? 2) Does the Diab compiler support these features? 3) Is Visual Slick Edit any good (comparable to MSDevStudio 6)? 4) Is Sniff+ worth having if we've got Artisan RTS (I don't think so)? 5)Debuggers...we get GNU tools by default...if we want Single Step we need to buy Diab compiler.We also need Diab for VisionClick. Windriver guys recommended VisionClick for "hardware bringup" of our new CPU card. Seems reasonable, but that's 3 debuggers that we get then! Any suggestions? 5) Windview looks invaluable. Any comments? 6) Is look worth having? 7) Is Diab RTA worth having (requires Diab compiler)? 8) Any experience with VissionProbe2 for debugging PowerPC via BDM port? 9) The datasheet says that the Wind Foundation Class libraries (class wrappers for system calls, etc) are only compatible with GNU Tools. Is this true? What about wrappers for use with DIab? 10) Why did I seriously look at adopting other RTOS's with a more straightforward tool chain? 11) Why am I now back here looking at Windriver stuff again? Any constructive comments appreciated. Darren Long Principal Software Engineer W Vinten Ltd. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Tue, 10 Jul 2001 14:19:08 +0200 From: no@spam.please Organization: Verio Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> I've tried a couple of time to get a demo license, with a CLEAR indication that I'm going to BUY, WindView. I'm still waiting.. for some reason, they don't want our money. Their technical support is not what I would expect from someone with access to the actual source code. My project has more that enough bypasses for bugs in their TCP/IP layer. no wonder people call them the microsoft of real-time :-( AR "Joe Woodbury" wrote: >Since we ordered VxWorks/Tornado II, we've had nothing but problems with >Wind River. They filled the order incorrectly and I had to resort to a lot >of yelling and threats to not pay our purchase order to get their attention. >Yet, two days later, they still have yet to do anything but email the >correct BSP. On top of all this, the highest level manager at Wind River >kept blaming us for some of the problems. > >So far, I've concluded that Wind River employees are given an IQ test. If >they fail, they are given a personal relationship test. If they fail that >also, they are hired. > >And heaven forbid they actually do something so obvious like creating a >compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. > >The shame is that VxWorks proper is such an excellent product. > >The questions are: Are our experiences unique? Did our company just happen >to run into the one group of bastards working at Wind River? > >Joe Woodbury > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS limitations? -- No limitition, here's what you do... Date: Tue, 10 Jul 2001 13:15:21 +0100 From: Owain Phillips Organization: Siemens Inc. Message-ID: <3B4AF1D9.EF59BD9B@siemenscomms.co.uk> References: <3B31BDA2.AEEA9B89@siemenscomms.co.uk> This is a multi-part message in MIME format. - --------------7FC63ADC93A18DB75C7414D3 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Dave, Ahhhahhh....See my comments below... > > We took an entirely different approach to this problem. We put > our bootroms in the fallow area of flash, then set up the bootroms > to mount tffs which then inflates our vxWorks image into ram and > jumps into it. To clarify, we put our compressed vxWorks application > image into tffs as a file (/RFA0/vxWorks.Z). > We use such an aproach on some products but we used an uncompressed image stored in the file-system. My problem is actually that I have an 8Meg device and that my bootrom image has to lie at the 4Meg watermark; this stems from an architectural restiction that means the physical boot address is mid-way through the prom. As far as I know in tffs you can reserve an area of prom for your bootrom image from the prom start to the end of the bootprom image. This unfortunately means I have to reserve the space below my bootprom image. IE I end up with about 5Meg reserved for bootrom and the remainding 3 Meg for a Flash FS ;-( Are there ways around this? - --------------7FC63ADC93A18DB75C7414D3 Content-Type: text/x-vcard; charset=us-ascii; name="owain.phillips.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Owain Phillips Content-Disposition: attachment; filename="owain.phillips.vcf" begin:vcard n:Phillips;Owain tel;fax:+44 115 943 4969 tel;work:+44 115 943 2167 x-mozilla-html:TRUE url:http://www.siemenscomms.co.uk org:Siemens Comunications Ltd version:2.1 email;internet:owain.phillips@siemenscomms.co.uk title:Software Engineer adr;quoted-printable:;;Technology Drive=0D=0ABeeston;Nottingham;;NG9 1LA;United Kingdom fn:Owain Phillips end:vcard - --------------7FC63ADC93A18DB75C7414D3-- --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Task information Date: Tue, 10 Jul 2001 07:16:33 -0400 From: "Mike Anderson" Organization: Lawrence Berkeley National Laboratory Message-ID: <9ierkb$g6f$1@overload.lbl.gov> VxWorks Greetings! > Submitted-by: prakashar@in.huawei.com > > > Is it possible to access the ready/suspended/blocked > queue information maintained by the VxWorks scheduler? > Is there any ready API which I can use in my program to > get the aforementioned information? > > Also, in VxWorks, when a task is executing, how does > the scheduler wakes up? Is it through a clock interrupt? > There are many ways to access this info, but you don't mention what sort of info you need. Do you need info from the TCBs of tasks, or do you want to manipulate the queues themselves? taskIdListGet() will get a list of all of the tasks that are currently in the system. Then taskInfoGet(...) will return the task descriptor information that will describe the current state (i.e., what queue is it in?) of the task. windTcb(...) will return a pointer to the TCB itself. You can change priorities, move to/from the suspend queue, etc. All of this is in taskLib. There you will also find the taskSwitchHook... calls that allow your code to be called on context switches. The scheduler wakes up in one of several cases. 1) The currently executing task calls something that requires kernel intervention such as a semTake or taskDelay. 2) The scheduler always gets called after the return from ISR context. 3) After an exception has occured 4) There is another that I'm missing, but I can't recall what. HTH, Mike - -- ======================================================= T h e P T R G r o u p, I n c. ======================================================= Embedded, Real-Time Solutions and Training Mike Anderson mailto:mike@ThePTRGroup.com Chief Scientist http://www.ThePTRGroup.com V: (703) 585-9384 F: (703) 430-3748 ======================================================= "Software development is like making a baby... You can't make a baby in one month by impregnating nine women. Some things just take time." --------------------------- Newsgroups: comp.os.vxworks Subject: Problem of reading of the serial port Date: 10 Jul 2001 05:26:30 -0700 From: ilandrei@mail.ru (Andrey Iljitchev) Organization: http://groups.google.com/ Message-ID: <28d45ddc.0107100426.4a5d8f40@posting.google.com> Hi I am trying to connect two MBX860-cards with null-modem through the serial port. I am succesfully writing to the serial port, but I can not read anything on the other end. How it is possible to solve this problem? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: DirLib and DosFS Date: Tue, 10 Jul 2001 15:50:10 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B4B0812.3B541773@ks.ericsson.se> References: <3B4AC964.42D96941@sepro-robotique.com> Emmanuel Herbreteau wrote: [snip] > strcat(name,(myDir->dd_dirent).d_name); [snip] change (myDir->dd_dirent) to myDirEnt and I think it will work better HTH Urban --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help - Does VxWorks's Wind Web Server support Java applets? Date: Tue, 10 Jul 2001 09:58:22 -0400 From: "Kevin Livingston" Message-ID: References: <9icukd$hppmb$1@ID-92003.news.dfncis.de> <994712976.913719@newsmaster-04.atnet.at> <9idb08$ibsgk$1@ID-92003.news.dfncis.de> are you serving a *.class file or a *.jar file. some JVM's are sensitive about this for some "odd" reason. try packaging your file in a .jar and see if that helps... disclaimer: I have not worked with the webserver from WindRiver. but one of my colleges has done it with GoAhead Kevin "Lee" wrote in message news:9idb08$ibsgk$1@ID-92003.news.dfncis.de... > Hi, > > Thank you for the info. > > Here is more info on the issues we see. > > A firmware (*.dlx file) contains Tornado 2.0, Wind Web Server 1.0, > application files, *.html files and jpg files. It's uploaded into a flash > memory on the target system. We can use web browsers on client systems to > access the web server on the target system and show the web pages and jpg > files. This means both *.html files and jpg files are downloaded OK from > the target system. > > Now, we add a Java applet onto a web page. The applet is very simple and > has only a button. Its size is about 800bytes. When we access the *.html > file which contains the applet, we only see a gray box showing on the > browser. Browser's Java console shows an error message "Truncated class > file". This means (I think) web browser does find the Java applet (a class > file) on the target system, but the web server does not send it as a whole > piece to web browser. We just couldn't figure out why. I would appreciate > it if someone could help. > > > Lee > > > > "Werner Schiendl" wrote in message > news:994712976.913719@newsmaster-04.atnet.at... > > Hi, > > > > Applets run in the browser, so if you are able to transfer files of > > arbitrary mime type, applets should work. > > > > If you are thinking about servlets, I think the answer is no. > > However, WindRiver has a Java product too, and maybe the products can be > > combined to run servlets (don't know any details about this). > > > > If you want to be able to run (a subset of) JavaScript (not Java) on the > > server, you could have a look at the GoAhead Webserver (www.goahead.com) > > > > hth > > werner > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS limitations? -- No limitition, here's what you do... Date: Tue, 10 Jul 2001 10:04:19 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B4B0B63.DB42640A@digidata.com> References: <3B31BDA2.AEEA9B89@siemenscomms.co.uk> <3B4AF1D9.EF59BD9B@siemenscomms.co.uk> You might try having the MTD describe the device as half its actual size. That way only the first helf will be known to TFFS. Doug Owain Phillips wrote: > Hi Dave, > > Ahhhahhh....See my comments below... > > > > > We took an entirely different approach to this problem. We put > > our bootroms in the fallow area of flash, then set up the bootroms > > to mount tffs which then inflates our vxWorks image into ram and > > jumps into it. To clarify, we put our compressed vxWorks application > > image into tffs as a file (/RFA0/vxWorks.Z). > > > > We use such an aproach on some products but we used an uncompressed > image stored in the file-system. > > My problem is actually that I have an 8Meg device and that my bootrom > image has to lie at the 4Meg watermark; this stems from an architectural > restiction that means the physical boot address is mid-way through the prom. > > As far as I know in tffs you can reserve an area of prom for your bootrom > image from the prom start to the end of the bootprom image. This unfortunately > means I have to reserve the space below my bootprom image. > > IE I end up with about 5Meg reserved for bootrom and the remainding > 3 Meg for a Flash FS ;-( > > Are there ways around this? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Tue, 10 Jul 2001 10:08:16 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B4B0C50.50C5529C@digidata.com> References: <9hqng4$49e$1@nntp1-cm.news.eni.net> <3B49FCCD.D96F7754@digidata.com> Not sure how that can happen. The only sectors that would have partial information would be the last sector of each file would't it? Defragmentation would not fix that unless TFFS is doing some sort of magic underneath. Doug Kevin Livingston wrote: > remember it also saves space too... > if you have a lot of half used juked up sectors etc. > if your drive is only 32 meg or something and you do a lot of IO it _could_ > help you, then again it could do nothing too. you should do tests in your > specific app to find out for sure how useful it would be. > > Kevin > > "Doug Dotson" wrote in message > news:3B49FCCD.D96F7754@digidata.com... > > I'm curious why you would want to defragment a flash file system. > > Defragentation is a performance enhancement in mechanical > > devices to help overcome rotational latency and seek delay. It > > would not apply in a solid state device. I'm interested because > > if there is some advantage then I also would be interested in a > > solution. > > > > Doug > > > > > > News Account wrote: > > > > > Does anybody know of any methodology or utility out there for > defragmenting > > > a VxWorks DOS system that is stored and maintained via TrueFFS? Also I > have > > > looked at the M-Systems TrueFFS code and noticed some sort of low level > > > defragmentation option that I believe is supposed to be accessable > either > > > through the "tffsRawio" routine or some sort of "ioctl" call, however I > was > > > unable to get this to work at all. At this point we would even be > willing to > > > utilize a third party software to perform the defragmentation task. > Please > > > help!!! > > > > > > Regards, > > > P. Eichler > > --------------------------- Newsgroups: comp.os.vxworks Subject: vxWorks dosFs2.0 debug message Date: Tue, 10 Jul 2001 14:08:11 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: I use an empty boot sector on flash and a printout message in printed infinitely on console : 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. 0x12f0b38 (tFlashCheck): dosFsLib.c : Malformed boot sector. Offset 0, value 255. is this possible to remove this printout by setting an extern variable to 0 ?? Thanks Steeve Gagne --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Tue, 10 Jul 2001 16:14:10 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B4B0DB2.B2B9A067@ks.ericsson.se> References: <9hqng4$49e$1@nntp1-cm.news.eni.net> <3B49FCCD.D96F7754@digidata.com> I'm also curious why you would do something like that. It will decrease the write performance, shorten the life of the flash and increase the risk of loosing data due to powerfailures etc. The space you could save is limited to directory repacking (or whatever you like to call it). The speed you might gain is that it could read more sectors in one call to the TrueFFS. Creating a disk cache (or using a ramdisk) would probably do more to performance than defragmenting. The best way to defrag the device is to read all files into memory, reformat the filesystem and write all files back to the disk. The there is the question: when should i defragment the disk? You can't allow disk access while defragmenting the disk (yes, I know M$ does, but I don't like the 5 restarts it makes sometimes). Defragmenting a flash file system is like tuning up a Ferrari, it won't go much faster, but it might break down a lot sooner than expected. I'm quite sure M-System would have written a defragmenter if they thought there was a need for it. There is an option to create contigous files if you really need the extra performance. Urban Kevin Livingston wrote: > > remember it also saves space too... > if you have a lot of half used juked up sectors etc. > if your drive is only 32 meg or something and you do a lot of IO it _could_ > help you, then again it could do nothing too. you should do tests in your > specific app to find out for sure how useful it would be. > > Kevin > > "Doug Dotson" wrote in message > news:3B49FCCD.D96F7754@digidata.com... > > I'm curious why you would want to defragment a flash file system. [snip] --------------------------- Newsgroups: comp.os.vxworks Subject: how does stack work in vxWorks on PPC Date: 10 Jul 2001 08:02:30 -0700 From: kswap@usa.net (kswap) Organization: http://groups.google.com/ Message-ID: i tried a task which does invoke checkStack() only and the print of every check Stack() shows that the current SP has grown further. Can anyone tell me how the stack operations are done in vxWorks (actually checkStack shouldnot eatup somuch of stack that we get unreliable results), any alternative to get the SP only without eating up much stack..? Swap --------------------------- Newsgroups: comp.os.vxworks Subject: RE: Task information Date: Tue, 10 Jul 2001 19:51:35 +0500 From: prakashar@in.huawei.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9if65r$jrl$1@overload.lbl.gov> This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_001_01C1094F.D0E2DEC0 Content-Type: text/plain; charset="iso-8859-1" Hello Mike, Thanks for the information. I am mainly looking for the order in which each task is put in the ready/ suspended/blocked queues. Is it possible for me to get this information ? Thanks in advance, Cheers, Prakash - -----Original Message----- From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] Sent: Tuesday, July 10, 2001 5:02 PM To: vxworks_users@csg.lbl.gov Subject: RE: Task information > Submitted-by: prakashar@in.huawei.com > > > Is it possible to access the ready/suspended/blocked > queue information maintained by the VxWorks scheduler? > Is there any ready API which I can use in my program to > get the aforementioned information? > > Also, in VxWorks, when a task is executing, how does > the scheduler wakes up? Is it through a clock interrupt? > There are many ways to access this info, but you don't mention what sort of info you need. Do you need info from - ------_=_NextPart_001_01C1094F.D0E2DEC0 Content-Type: text/html; charset="iso-8859-1" RE: Task information

Hello Mike,

Thanks for the information. I am mainly looking for
the order in which each task is put in the ready/
suspended/blocked queues. Is it possible for me to
get this information ?

Thanks in advance,

Cheers,
Prakash

-----Original Message-----
From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov]
Sent: Tuesday, July 10, 2001 5:02 PM
To: vxworks_users@csg.lbl.gov
Subject: RE: Task information

<snip>

> Submitted-by: prakashar@in.huawei.com
>
>
> Is it possible to access the ready/suspended/blocked
> queue information maintained by the VxWorks scheduler?
> Is there any ready API which I can use in my program to
> get the aforementioned information?
>
> Also, in VxWorks, when a task is executing, how does
> the scheduler wakes up? Is it through a clock interrupt?
>

  There are many ways to access this info, but you don't
mention what sort of info you need.  Do you need info from

<snip>

- ------_=_NextPart_001_01C1094F.D0E2DEC0-- --------------------------- Newsgroups: comp.os.vxworks Subject: RE: two 3com 3c905-TX Ethernet cards on the target machine Date: Tue, 10 Jul 2001 17:14:04 +0200 From: Diego Novellon Organization: Lawrence Berkeley National Laboratory Message-ID: <9if65r$jrm$1@overload.lbl.gov> > -----Mensaje original----- > De: Vxworks Exploder [SMTP:vx_explo@csg.lbl.gov] > Enviado el: martes 10 de julio de 2001 13:18 > Para: vxworks_users@csg.lbl.gov > Asunto: Vx Exploder Digest > > Newsgroups: comp.os.vxworks > Subject: two 3com 3c905-TX Ethernet cards on the target machine > Date: Mon, 9 Jul 2001 11:14:28 -0400 > From: "Fangzhe Chang" > Organization: Bell Labs > Message-ID: <9ichai$jhm1@news.research.bell-labs.com> > Reply-To: "Fangzhe Chang" > > Hi, > > I'm trying to install two 3COM 3c905-TX FastEthernet cards on my > target machine (VxWorks 5.4). However, I could not get it recognize both of > them. Could someone do me a favor and give some suggestion? > > The first card (i.e., elPci0) comes up by default. To install the > second card, I add a line to the "endDevTbl" network device table, as > {1, sysEl3c90xEndload, "", 1, NULL, FALSE} > > I modify the sysEl3c90xEndLoad() a little bit to detect the PCI > devices and build the correct parameter strings. It would try to load the two > devices with the following parameters (both when mulDevLoad() calls it the > second time), el3c90xEndLoad("0:0xffffffff:0xee80:0x0:41:9:-1:-1:- > 1:0x0:0xffffffff"); el3c90xEndLoad("1:0xffffffff:0xef00:0x0:42:10:-1:-1:- > 1:0x0:0xffffffff"); > > Both of them returns the same END object 0x7ffec24. However, the > system still does not know about second interface after this. My attempt to > configure it using ifMaskSet("elPci1", exffffff00) and ifAddrAdd > ("elPci1", "172.16.1.10") fails; ifShow only tells about the first interface. > > Any help is greatly appreciated. > > Fangzhe Chang > Hi, I remember having modified sysEl3c90xEnd.c as well to fix a couple of bugs in sysEl3c90xPciInit() and sysEl3c90xEndLoad(). I don't have the code in front of me right now, but according to my notes it was something like changing: * in sysEl3c90xPciInit(): 'netEndUnits' to '0' as a parameter to pciFindDevice() pEl3c90xActualRsc[unit] to pEl3c90xActualRsc[netEndUnits] * in sysEl3c90xEndLoad(): Use the unit number instead of '0' as an index to the pEl3c90xActualRsrc[] array. I guess it from pParamStr. If you're interested I can probably post tomorrow the actual code. Regards, ///////////////////////////////////////////////// // Diego Novellón Latre, Computer Engineer // // TETRA Infr., R&D Dept, Teltronic S.A.U. // // Polígono Malpica, C/ F-Oeste, parcela 12 // // E50057 Zaragoza - Spain // // Tel +34.976.465656 #174, fax +34.976.465722 // // http://www.teltronic.es/ // ///////////////////////////////////////////////// --------------------------- Newsgroups: comp.os.vxworks Subject: how to use vxsim? Date: 10 Jul 2001 09:39:22 -0700 From: xing@cs.uga.edu (G Xing) Organization: http://groups.google.com/ Message-ID: <30c8c76f.0107100839.4b83667e@posting.google.com> I am a beginner of vxsim, and I am tring to simulate mips code on sparc machine, is this possible to use vxsim, which image should I launch for vxsim: vxworks for sparc or vxworks for mips. Thanks, G Xing --------------------------- Newsgroups: comp.os.vxworks Subject: build boot disk for Pentium with SCSI hard disk driver Date: Tue, 10 Jul 2001 17:01:01 +0000 (UTC) From: fwang@redswitch.com (Feng Wang) Organization: Mailgate.ORG Server - http://www.Mailgate.ORG Message-ID: <3B4B34BF.95DD0CBA@redswitch.com> This is a multi-part message in MIME format. - --------------78BFDD5B41591B6BB1CFA15A Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I'm trying to build a boot disk for a pentium machine with SCSI hard disk. According to the manual, I modified the config.h to define INCLUDE_SCSI. However, the boot disk just hanging around while load vxworks from the floppy disk. Anyone know how to boot a machine with SCSI disk? Thanks, Feng - --------------78BFDD5B41591B6BB1CFA15A Content-Type: text/x-vcard; charset=us-ascii; name="fwang.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Feng Wang Content-Disposition: attachment; filename="fwang.vcf" begin:vcard n:Wang;Feng tel;fax:408-719-4800 tel;work:408-719-4752 x-mozilla-html:FALSE url:http://www.redswitch.com org:RedSwitch Inc.;Software adr:;;1815 McCandless Drive;Milpitas;CA;95035;USA version:2.1 email;internet:fwang@redswitch.com title:Management Software Engineer end:vcard - --------------78BFDD5B41591B6BB1CFA15A-- - -- Posted from 206-14-68-143.ncal.verio.com [206.14.68.143] (may be forged) via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- Newsgroups: comp.os.vxworks Subject: memScope Startup Errors..... Date: Tue, 10 Jul 2001 12:28:30 -0400 From: Chandwani Organization: Lawrence Berkeley National Laboratory Message-ID: <9ifd6r$mbb$1@overload.lbl.gov> Hi All !!, I am running Tornado 2.0 on Sun Solaris 2.5.1 host and the target is 8260 based Single Board Computer. When I click on the memScope button on the Tornado launch window ( once I am connected to the target ), I get following errors. I have tried all configurations but couldn't make ot work. Can someone help me : ld.so.1:/home/vxtools/vxworks/rti/scopetools.2.0b/Java/INSTALLED/jre/sparcSol2.5/bin/../bin/sparc/native_threads/java: fatal: relocation error: file /usr/openwin/lib/locale/en_US/XLIBI18N/xlibi18n.so.1: symbol _XlcDefaultInitLocale: referenced symbol not found Rakesh (rexrc@lucent.com) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Tue, 10 Jul 2001 13:09:22 -0400 From: Tim White Organization: Johns Hopkins University Applied Physics Lab Message-ID: <3B4B36C2.D327CC1D@aplexus.jhuapl.edu> References: <6qH17.79$5f6.133962@news.uswest.net> This is a multi-part message in MIME format. - --------------12945A6D8BE86C51207A2923 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Joe, VxWorks has a very good product, but their customer support really stinks. I have contacted them with problems with proxy net on the shared memory network for 5.3.1 and 5.4 for a couple of months now and have not resolved a single one. The company is aware of their shortfall in this area (I have a friend who went to work for them about a year ago) and hopefully they will get their support at the same level as their product. Tim - --------------12945A6D8BE86C51207A2923 Content-Type: text/x-vcard; charset=us-ascii; name="twhite.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Tim White Content-Disposition: attachment; filename="twhite.vcf" begin:vcard n:White;Timothy x-mozilla-html:FALSE org:Johns Hopkins University Applied Physics Lab;Battle Force Integration Group (A3C) adr:;;;;;; version:2.1 email;internet:Tim.White@jhuapl.edu title:Real Time Software Engineer x-mozilla-cpt:;-18792 fn:Timothy White end:vcard - --------------12945A6D8BE86C51207A2923-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: build boot disk for Pentium with SCSI hard disk driver Date: Tue, 10 Jul 2001 19:39:52 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994787273.209133@newsmaster-04.atnet.at> References: <3B4B34BF.95DD0CBA@redswitch.com> Do you mean your Bootrom still loads the image from the floppy? You need to modify the bootline as well, to reflect the correct boot device. You can do so at startup (if you have a terminal or screen/keyboard connected) Otherwise you can modify DEFAULT_BOOT_LINE hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: ftpd lockup problems Date: Tue, 10 Jul 2001 14:01:45 -0500 From: "Chang, Mark" Organization: Lawrence Berkeley National Laboratory Message-ID: <9ifk7q$ph7$1@overload.lbl.gov> vxworks 5.3.1 Our FTPd is locking up after repeated (about 10 in row) quick DIRs. Is this a known bug? Maybe increasing the mbufs? I know the data connection is probably not closing fast enough, but should I shorten the timeout for that? Advice appreciated. Thanks, Mark - ---------------------------------------------------------- Mark Chang Senior Engineer Network Management Broadband Access and Transport Group ADC Wireline Systems Division Mark_Chang@ADC.com (714)730-2827 Learn about ADC - The Broadband Company at www.adc.com --------------------------- Newsgroups: comp.os.vxworks Subject: WDB Exception: 256 Date: Tue, 10 Jul 2001 12:50:10 -0700 From: Satheesha Rangegowda Organization: Agilent Technologies Message-ID: <3B4B5C72.9BFA9E69@agilent.com> Hi, I am trying to load a module using windSh, after loading this module windSh prints out following exception message. - -> Exception number 256: Task: 0x18e8a40 (tWdbTask) system reset program counter: 0 machine status: 0x10 I am trying to load a module which is compiled with -mlongcall option & i have about 64Meg RAM on my target. Any info. on why it is giving this exception just after loading the module? Thanks in advance. - -- - ------------------------------------------------------------- Satheesha Rangegowda satheesha_rangegowda@agilent.com R&D Engineer (509) 921 3960 Agilent Technologies Spokane, WA - 99019. - ------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tools questions. Date: Tue, 10 Jul 2001 21:08:01 +0100 From: molochai Message-ID: <3f29b984a%molochai@vapour-trail.demon.co.uk> References: <5c34b056.0107100316.10f30da0@posting.google.com> In message <5c34b056.0107100316.10f30da0@posting.google.com> dlong@wvintenltd.com (Darren Long) wrote: > My experience with the Windriver sales people was that they were > overwhelmed with the sheer number of their own and third party tools > available and they could not seem to recommend a well specified but useful > set of tools for use to use. For example, it looked like they were > suggesting that we bought 3 debuggers. > > I would like comments on the following issues to see if I can work out a > useful, minimally complete set of tools to order. > > Our project is based on the MPC555 PowerPC processor. We code exclusively > in C++ and want to use full ANSI C++ language, including namespaces, > exceptions, C++ standard library (STL etc.) and RTTI. > > 1)Does the Windriver version of the GNU compiler support these features? Depends which version of GCC: 2.7.2 supports everything but namespaces (although not necessarily correctly in all cases, and is not strictly ANSI C++ compatible either). GCC 2.95.2 apparently does the lot, but is not shipped with T2 at the moment. It is supposed to be imminent as a upgrade.. It is also AFAIK the standard compiler with T3/VxworksAE so its only a matter of time before the promised cummulative patch apears. You can get the source code and build the compiler yourself, but its the simple things like new name demanglers in shells & debuggers, and the standard libraries and C++ runtime support that all need changing that are the problem. FWIW my team use GCC 2.7.2 at the moment, and are waiting for an evaulation of Diab which actually works at runtime. Inicidentally, GCC is *much* faster at compiling code than Diab, even if the code is not as good. > 2) Does the Diab compiler support these features? Yes. > 3) Is Visual Slick Edit any good (comparable to MSDevStudio 6)? Yes. very good, especially for C++. Haven't used MS Dev Studio 6 much, but it seemed very similar to Slick Edit, although I think Slick Edit has the edge in code completion personally. VSE also reads Tornado II workspace files, and allows you to build within the editor - without any extra configuration. It loads project files from practically everything else too. It also contains an FTP client which can cope with the standard vxWorks ftp server. A text editor is probably be the bit of development software you'll use the most. VSE certainly massively reduces typing. > 4) Is Sniff+ worth having if we've got Artisan RTS (I don't think so)? I never thought so when I evaluated it. VSE will provide some of the browsing facilities anyway. > 5)Debuggers...we get GNU tools by default...if we want Single Step we need > to buy Diab compiler.We also need Diab for VisionClick. I think Visionclick works with both compilers, and SingleStep is being made to do so.. In any case, any BDM which can act as a target agent should be usable under CrossWinds (?). I'd hazard a guess that Cross Winds will either gain the name demangler for diab or be replaced with one of the other two in the near future. > Windriver guys recommended VisionClick for "hardware bringup" of our > new CPU card. Seems reasonable, but that's 3 debuggers that we get > then! Any suggestions? > 5) Windview looks invaluable. Any comments? Absolutely. WindView has payed itself off many times over in my team. > 6) Is look worth having? In a word, NO. I Think it's thrown in with the C++ Tornado seat anyway. I wouldn't pay extra for it. > 7) Is Diab RTA worth having (requires Diab compiler)? > 8) Any experience with VissionProbe2 for debugging PowerPC via BDM > port? > 9) The datasheet says that the Wind Foundation Class libraries (class > wrappers for system calls, etc) are only compatible with GNU Tools. Probably Unless you get source and recompile for Diab. Seriously, you could always write your own - we did - there's not much there, and you might like to enforce house style through the implementation of your wrappers. > Is this true? What about wrappers for use with DIab? > 10) Why did I seriously look at adopting other RTOS's with a more > straightforward tool chain? Probably because there's less incentive for tool vendors to get the tools to actually work when there's no competition. At least when there's competition, you have choice. TBPH, buying tools seperately does at least let you avoid the duff ones (and yes, there are plenty of crap tools out there) > 11) Why am I now back here looking at Windriver stuff again? > > Any constructive comments appreciated. > Worth checking what is bundled with what, this seems to change frequently as WRS acquire more of their competitors. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: standalone_vxworks with no network Date: 10 Jul 2001 13:14:50 -0700 From: chrisanc1967@yahoo.com (Christopher Sanchez) Organization: http://groups.google.com/ Message-ID: References: What we did is add #define STANDALONE_NET to our config.h file This initialized the network. "Kelly Hornsby" wrote in message news:... > hello > > I need some ideas of something that I can't figure out. I'm using T2 on a > Sandpoint 750 system. I have a DEC 21143 NIC, and it works just fine with > the bootloader/ ftp-Vxworks. However, I've created the > standalone_vxworks.hex file, and flashed it into memory, but the network is > not initialized now. the NIC is detected fine (pciHeaderShow indicates > this) and the 21143 PCI initilization routine is executed (sysNetHwInit2), > but there's no dev out there, and the device shows no devices for ifShow. > > I'm looking for suggestions or ideas for why this occurs. Any ideas? --------------------------- Newsgroups: at.telekomm.technik.isdn,cloud9.slip-ppp,comp.dcom.isdn,comp.os.vxworks,comp.protocols.ppp,news.admin.os.vxworks,ppc.general Subject: HDLC and LABD questions? on MPC8260 Date: Tue, 10 Jul 2001 20:10:29 GMT From: u687591552@spawnkill.ip-mobilphone.net Message-ID: Sender: u687591552@spawnkill.ip-mobilphone.net 1) Why do we say LAPD is a HDLC type protocol? How different are they? 2) For specific for MPC8260, somewhere in user manaul it mentions that, configure SCC to handle HDLC and LAPD, page 14-31, I don't understand how let it handle LAPD. Help? BT - -- Sent by guangbutun from hotmail piece from com This is a spam protected message. Please answer with reference header. Posted via http://www.usenet-replayer.com/cgi/content/new --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Tue, 10 Jul 2001 15:51:07 -0500 From: John McClenny Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B4B36C2.D327CC1D@aplexus.jhuapl.edu> In article <3B4B36C2.D327CC1D@aplexus.jhuapl.edu>, twhite@aplexus.jhuapl.edu says... > Joe, > > VxWorks has a very good product, but their customer support really stinks. I > have contacted them with problems with proxy net on the shared memory network > for 5.3.1 and 5.4 for a couple of months now and have not resolved a single > one. > VxWorks has a very good kernel, but the farther away you get from the kernel the less mature (buggier) the code is. The network stack is crap and a lot of the standard BSP packages you get from them have very worthless drivers for any real application. > The company is aware of their shortfall in this area (I have a friend who went > to work for them about a year ago) and hopefully they will get their support > at the same level as their product. > Support can be good if you have the right local support person. --------------------------- Newsgroups: at.telekomm.technik.isdn,comp.dcom.isdn,comp.os.vxworks,news.admin.os.vxworks,ppc.general Subject: Re: HDLC and LABD questions? on MPC8260 Date: 10 Jul 2001 17:06:23 -0400 From: James Carlson Organization: Sun Microsystems Inc. - BDC Message-ID: References: u687591552@spawnkill.ip-mobilphone.net writes: > 1) Why do we say LAPD is a HDLC type protocol? How different are they? HDLC is (generally) just the framing and addressing. LAP-D includes error control (ARQ; windowing and retransmission). > 2) For specific for MPC8260, somewhere in user manaul it mentions that, > configure SCC to handle HDLC and LAPD, page 14-31, I don't understand > how let it handle LAPD. Help? Configure the SCC for HDLC as needed for your physical interface. Implement LAP-D in software. Note the trimmed newsgroup list. Neither of thse questsions really belongs on comp.protocols.ppp ... - -- James Carlson, Internet Engineering SUN Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084 MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677 --------------------------- Newsgroups: comp.os.vxworks Subject: Does BOOTROM always load the first file in TFFS, when I want to startup from TFFS? Date: 10 Jul 2001 16:10:24 -0700 From: yxiu@salira.com (Yan Xiu) Organization: http://groups.google.com/ Message-ID: <22159022.0107101510.5ffa8e48@posting.google.com> hi, guys, I have made up a DOSFS over flash by TFFS, and the device name is /TFFS0/. I copied two VxWorks image file into /TFFS0/, first is vxworks0, the other is vxworks1. When I boot vxwroks from /TFFS0/vxworks0, it works. But when I boot vxworks from /TFFS0/vxworks1, it run with /TFFS0/vxworks0 again! It had not been normal until I deleted the file vxworks0. What's happened? Is this a feature of DOS/TFFS or anthing else? Thanks, Yan --------------------------- Newsgroups: comp.os.vxworks,comp.lang.ada Subject: Re: 2 Rational Apex questions...#2 Date: Tue, 10 Jul 2001 19:36:23 -0400 From: "Corey Ashford" Message-ID: References: <3B466FD7.38D17AAF@mailandnews.com> Tornado 2.02 is designed to work with VxWorks 5.4.1 Tornado 2.02 won't play nice with VxWorks 5.1. Are you maybe thinking of 5.3.1? That was what was shipped with Tornado 1.0.1. You're best off upgrading to VxWorks 5.4.1. That said, Tornado 2.02 doesn't currently work with Apex 3.2.0b, but Tornado 2.0 does. A patch for that is pending. Contact support@rational.com for patch availability. Corey Ashford Rational Software Corp. "Bigg K" wrote in message news:3B466FD7.38D17AAF@mailandnews.com... > I'm not sure if these questions are compiler or OS related, > so I'm posting it to two groups. Thanks again for any help > you can offer... > > #2 > > Here's the environment: > > VxWorks 5.1 as target running on Motorola 68040 > Host: sun4_solaris2 > Tornado version 2.02 > Rational Apex Duo 3.2.0b [build date: 10/11/99] > > Is it possible to run a program compiled and linked under > this version of Apex on VxWorks 5.1? Is an upgrade to 5.4 > required? > > > ______________________________________________________________________ > Posted Via Uncensored-News.Com - Still Only $9.95 - http://www.uncensored-news.com > With Seven Servers In California And Texas - The Worlds Uncensored News Source > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Defragmentation for TrueFFS Date: Tue, 10 Jul 2001 23:54:15 GMT From: andreyk@home.com (Andray Kaganovsky) Organization: @Home Message-ID: References: <9hqng4$49e$1@nntp1-cm.news.eni.net> <9i2tol$r84$1@nntp2-cm.news.eni.net> "P.Eichler" wrote in news:9i2tol$r84$1@nntp2-cm.news.eni.net: > Thank you for your input. I've looked into the provided TFFS option (which > is not standard in the VxWorks version of TrueFFS). Unfortunately it only > seems to deal with "recollecting" and "packing" sectors at the low flash > level. It does NOT seem to deal with de-fragmenting at the DOS file level, > which is really what I need to improve the overall speed of writing and > deleting files on a very fragmented system. [AndrayK 2]. Are you sure that there is any significant impact that fragmentation at DOS level will have on I/O performance ? The reason why fragmenation at DOS level impacts I/O performance of conventional hard disks, is that the 'seek' time (time required to move disk's heads between cylinders) becomes significant. This is not the case with TFFS volumes, where 'seek' time is virtually zero. Assuming that dosFs caches the entire FAT in RAM, it seems logical to assume that there should be no major impact of defragmentation at DOS level on TFFS performance. > "Andray Kaganovsky" wrote in message > news:Xns90D3C43FA4E11andreykhomecom@24.2.9.58... >> "News Account" wrote in news:9hqng4$49e$1@nntp1- >> cm.news.eni.net: >> >> > Does anybody know of any methodology or utility out there for >> > defragmenting a VxWorks DOS system that is stored and maintained via >> > TrueFFS? Also I have looked at the M-Systems TrueFFS code and noticed >> > some sort of low level defragmentation option that I believe is >> > supposed to be accessable either through the "tffsRawio" routine or >> > some sort of "ioctl" call, however I was unable to get this to work at >> > all. At this point we would even be willing to utilize a third party >> > software to perform the defragmentation task. Please help!!! >> > >> > Regards, >> > P. Eichler >> > >> [AndrayK]. To defragment TrueFFS flash array, you can use code fragment >> similar to the one below: >> >> int freeSectors = 0x70000000; /* impossible number of sectors */ >> >> (void) tffsRawio (0, TFFS_DEFRAGMENT_VOLUME, freeSectors, 0, 0); >> >> The call above could return error code, but, as far as I know, it is safe >> to ignore it. If media is badly fragmented, the call above can take very >> long time to complete, since it is doing complete defragmentation of the >> entire flash array. >> >> Andray >> > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Does BOOTROM always load the first file in TFFS, when I want to startup from TFFS? Date: Wed, 11 Jul 2001 00:16:34 GMT From: andreyk@home.com (Andray Kaganovsky) Organization: @Home Message-ID: References: <22159022.0107101510.5ffa8e48@posting.google.com> yxiu@salira.com (Yan Xiu) wrote in news:22159022.0107101510.5ffa8e48 @posting.google.com: > hi, guys, > > I have made up a DOSFS over flash by TFFS, and the device name is > /TFFS0/. > I copied two VxWorks image file into /TFFS0/, first is vxworks0, the > other is vxworks1. > > When I boot vxwroks from /TFFS0/vxworks0, it works. But when I boot > vxworks from /TFFS0/vxworks1, it run with /TFFS0/vxworks0 again! It > had not been normal until I deleted the file vxworks0. > > What's happened? Is this a feature of DOS/TFFS or anthing else? > > Thanks, > Yan > [AndrayK]. As far as I know, TFFS simply serves filesystem's requests to read sector, so if there is any problem, then it's likely to be in dosFs rather then in TFFS. Did you happen to have console attached to your board ? If you do, could you hit any key during the boot's downcount, then hit 'c', and edit boot parameters to make sure that you actually booting /tffs0/vxWorks1 image. My understanding is that this way you will override boot parameters that are embedded in the bootrom, as well as stored in board's NOVRAM. Andray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use vxsim? Date: Wed, 11 Jul 2001 00:23:00 GMT From: andreyk@home.com (Andray Kaganovsky) Organization: @Home Message-ID: References: <30c8c76f.0107100839.4b83667e@posting.google.com> xing@cs.uga.edu (G Xing) wrote in news:30c8c76f.0107100839.4b83667e@posting.google.com: > I am a beginner of vxsim, and I am tring to simulate mips > code on sparc machine, is this possible to use vxsim, > which image should I launch for vxsim: vxworks for sparc > or vxworks for mips. > Thanks, > G Xing > [AndrayK]. As far as I know, vxSim runs as the process on Tornado development host (as opposed to running on the target board), i.e. you can't run MIPS binaries on SUN hosts. Andray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ftpd lockup problems Date: 10 Jul 2001 18:29:46 -0700 From: tlksk@hotmail.com (Kumar) Organization: http://groups.google.com/ Message-ID: References: <9ifk7q$ph7$1@overload.lbl.gov> Hi Chang Check whether the files opened are closed properly use "iosFdShow" for this. Also do an "inetstatShow" to see what state the previous ftp sessions are in. You can also check what is the maximum number of ftp clients that can connect to your system. You can check using the "ftpsMaxClients" variable. Kumar "Chang, Mark" wrote in message news:<9ifk7q$ph7$1@overload.lbl.gov>... > vxworks 5.3.1 > > Our FTPd is locking up after repeated (about 10 in row) > quick DIRs. > > Is this a known bug? Maybe increasing the mbufs? > I know the data connection is probably not closing > fast enough, but should I shorten the timeout for that? > > Advice appreciated. > Thanks, > Mark > > > > ---------------------------------------------------------- > Mark Chang > Senior Engineer Network Management > Broadband Access and Transport Group > ADC Wireline Systems Division > Mark_Chang@ADC.com > (714)730-2827 > > Learn about ADC - The Broadband Company at www.adc.com --------------------------- Newsgroups: comp.os.vxworks Subject: I wanna change mbuf configuration. Date: Wed, 11 Jul 2001 11:27:49 +0900 From: "Ogre" Organization: Korea Telecom Message-ID: <9igdki$cq9$1@news1.kornet.net> I wanto to change mbuf conf. my system is following ~~ - ------------------------------------------------------------------------- - -> memShow(1) FREE LIST: num addr size --- ---------- ---------- 1 0x3fb1e8 9704 2 0x13c970 2002736 3 0x3fb000 472 4 0x3f9000 4088 5 0x3f7000 4088 6 0x3f5000 4088 7 0x3f3000 4088 8 0x3f1000 4088 9 0x3ef000 4088 10 0x3ed000 4088 11 0x3eb000 4088 12 0x3e9000 4088 13 0x3e7000 4088 14 0x3e5000 4088 15 0x3e3000 4088 16 0x3e1000 4088 17 0x3df000 4088 18 0x3dd000 4088 19 0x3db000 4088 20 0x3d9000 4088 21 0x3d7000 4088 22 0x3d5000 4088 23 0x3d3000 4088 24 0x3d1000 4088 25 0x3cf000 4088 26 0x3cd000 4088 27 0x3cb000 4088 28 0x3c9000 4088 29 0x3c7000 4088 30 0x3c5000 4088 31 0x3c3000 4072 SUMMARY: status bytes blocks avg block max block ------ --------- -------- ---------- ---------- current free 2127360 31 68624 2002736 alloc 770168 156 4936 - cumulative alloc 901328 159 5668 - - ------------------------------------------------------------------------- but, I want to end mbuf list at addr 0x3e5000. because I'm using that addr. following 'memShow(1)' shows ... - ------------------------------------------------------------------------- - ->memShow(1) FREE LIST: num addr size --- ---------- ---------- 1 0x3fb1e8 192 2 0x13c970 1833512 3 0x3fb000 472 4 0x3f9000 4088 5 0x3f7000 4088 6 0x3f5000 4088 7 0x3f3000 4088 8 0x3f1000 4088 9 0x3ef000 4088 10 0x3ed000 4088 11 0x3eb000 4088 12 0x3e9000 4088 13 0x3e7000 4088 14 0x3e5000 4088 15 0x3e3000 4088 16 0x3e1000 4088 17 0x3df000 4088 18 0x3dd000 4088 19 0x3db000 4088 20 0x3d9000 4088 21 0x3d7000 4088 22 0x3d5000 4088 23 0x3d3000 4088 24 0x3d1000 4088 invalid block at 0x3cf000 deleted - ------------------------------------------------------------------------ error occured. so, system broke down. Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use vxsim? Date: 10 Jul 2001 20:42:00 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107101941.50fcf7aa@posting.google.com> References: <30c8c76f.0107100839.4b83667e@posting.google.com> Hello, The VxWorks simulator is an API simulator for the operating system, not a CPU/instruction level simulator. To simulate MIPS code you would need to find an instruction level simulator. If such a thing was available, it might then be possible to write a board support package for VxWorks that allowed you to run a MIPS version of the VxWorks kernel under the instruction level simulator. I don't know if anybody has done this kind of thing, or even if an instruction level simulator for MIPS is readily available. You might be better off just getting a MIPS eval board and an ICE/JTAG debugger ;-) HTH, John... xing@cs.uga.edu (G Xing) wrote in message news:<30c8c76f.0107100839.4b83667e@posting.google.com>... > I am a beginner of vxsim, and I am tring to simulate mips > code on sparc machine, is this possible to use vxsim, > which image should I launch for vxsim: vxworks for sparc > or vxworks for mips. > Thanks, > G Xing --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Accessing Host files from VxSim? Date: Tue, 10 Jul 2001 09:14:31 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <9igova$ige$1@at-vie-newsmaster01.nextra.at> References: <9icmk10c87@enews2.newsguy.com> the open problem can be fixed very easy by installing all the fixpacks. i posted the list of SPRs here about 1.5 ago. just search WindSurf. Maybe (i expect so) all the fixes are part of C-SP3. - -- regards, Mario Semo. http://www.kirchnersoft.com tazboy78 wrote in message news:9icmk10c87@enews2.newsguy.com... > The best way is to use Target Server File System. When you use "host:c:\..." > you are using passFS and there are problems with passFS. Something about the > permission flags not matching up between VxWorks and NT..... > > "Mark" wrote in message > news:c5b88987.0107051139.6f694df4@posting.google.com... > > Is there a way to open and read a file from the c: drive from a VxSim > program? > > > > 1) I tried using the full pathname "c:\thedir\myfile.dat" > > 2) I also tried in windshell > > cd c:\thedir > > ls - myfile.dat yes file is there > > fp = open("myfile.dat", 0, 0); > > > > neither worked > > > > thanks > > Mark > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Absolute Linking Date: Wed, 11 Jul 2001 08:58:17 +0200 From: "Michael Lawnick" Organization: Buergernetz Dillingen Message-ID: <9igtlq$tae$1@snoopy.bndlg.de> References: <9ic0j4$9ji$1@snoopy.bndlg.de> Reply-To: "Michael Lawnick" Sender: mlawnick@pd9e65608.dip.t-dialin.net Hi Diego, nice discourse, but this wasn't the question. I know how its done on Motorola boards, but my problem is my IBM PPC405 which starts code execution at 0xFFFFFFFC. The way it's done now, is to edit bootrom_uncmp.hex and insert a line before last (S9) record. What I want to do, is to generate a new section and assign code to it. Again: who has already used additional section in make/cc/ln ? BTW: Pls no answers like "haven't done yet, but ..." - -- Mit freundlichen Grüßen, Michael Lawnick ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== "Diego Serafin" schrieb im Newsbeitrag news:Uay27.13627$Kf3.147424@www.newsranger.com... > In article <9ic0j4$9ji$1@snoopy.bndlg.de>, Michael Lawnick says... > > > >Hi VxWorks Gurus, > > > >has anybody ever tried to link some code/data to absolute location in > >bootrom ? > > > >The background: > >We have a PPC405, whose H/W reset vector is 0xfffffffc. > >VxWorks bootrom starts at ROM_TEXT_ADRS in Makefile. > >Normally we edit the hex code, but now search for a 'make' solution. > > > >We tried to create a section, put the jump instruction to ROM_TEXT_ADRS in > >it and locate the section at 0xfffffffc. This shouldn't be too difficult, > >but we couldn't get it run by now. > > > >Has anybody done this already and can give me some advice ? > > I don't know how PPC405 is fetching the first instruction from reset vector, but > here it works on my MPC860: > > The reset vector is at 0x100 and, at reset, the CPU holds ROM CS active, i.e. > when, at reset, CPU is fetching from 0x100 it really fetches > at FFF00100 where ROM is located. Code at FFF00100 is the VxWorks entry point > (_romInit) linked by ldppc at (for my image)... address RAM_LOW_ADRS (for me > 0x100100). > The solution to make all this mess running is to write the first routines > (romInit and romStart) as PIC (Position > Independent Code), all jumps must be Program Counter Relative. > > I try to be more clear: > > - First istruction is fetched from 0x100 but ROM CS is selected and higher > address lines are not connected so CPU is really fetching from ROM, at offset > 0x100 from its beginning (set for my target at FFF00000). > - First istruction is a relative jump to _cold label for cold boot and then an > absolute jump is computed as relative to ROM_TEXT_ADRS to go straight into ROM > space (jump to _start label is computed subtracting _romInit from _start and > then adding ROM_TEXT_ADRS to find the offset in ROM space where _start label > resides) > - Then setup register and all the reboot stuff in assemply and jump to _romStart > with proper calculation of its address (same as above) > - romStart will copy ROM into RAM and, finally, will jump to RAM at code loaded > at proper address (RAM_HIGH_ADRS for bootRom if I remember well) freeing > programmer from being PIC (!) . > > Take a look at "Tornado BSP Developer's Kit for VxWorks User's guide" for > further details ... > > HTH, > Ciao, > Diego. > > --------------------------- Newsgroups: comp.os.vxworks Subject: tNetTask Date: Wed, 11 Jul 2001 16:41:08 +0900 From: "Ogre" Organization: Korea Telecom Message-ID: <9ih000$4dm$1@news1.kornet.net> I can't decide tNetTask's priority. My system has following tasks. tShell = priority 1 tTelnetd = priority 2 tNetTask = priority 50 tWeb = priority 51 A task = priority 60 and my system connects with PC through LAN. Now, I executed 'ping to my system'. It's OK. But, if I browse Web Pages in 'my System', my tNetTask or tWeb task go SUSPEND status - ------------------------------------------- Data access Exception current instruction address: 0x00064e6c Machine Status Register: 0x00001030 Data Access Register: 0xaaaa0300 Condition Register: 0x48222042 Data storage interrupt Register: 0x0000001f Task: 0x345a98 "tWeb" - -------------------------------------------- what's problem? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem of reading of the serial port Date: 11 Jul 2001 07:48:22 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <28d45ddc.0107100426.4a5d8f40@posting.google.com> ilandrei@mail.ru (Andrey Iljitchev) wrote: > Hi > > I am trying to connect two MBX860-cards with null-modem through the > serial port. I am succesfully writing to the serial port, but I can not > read anything on the other end. > > How it is possible to solve this problem? A suggestion: connect a PC to the serial port of the board, check if you can read and write from this port. Do the same with the other board, and see if you can find a problem. If this does not show any problem, check your cable. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I wanna change mbuf configuration. Date: Wed, 11 Jul 2001 11:18:22 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994843586.465722@newsmaster-04.atnet.at> References: <9igdki$cq9$1@news1.kornet.net> Hi, I think you should not include the memory range you need (for whatever reason) to the system memory pool. Probably you need to modify your sysLib.c to not include the memory. However, be sure to still map it with the MMU, if your board has one. hth werner "Ogre" wrote in message news:9igdki$cq9$1@news1.kornet.net... > I wanto to change mbuf conf. > > my system is following ~~ > ------------------------------------------------------------------------- > -> memShow(1) > FREE LIST: > num addr size > --- ---------- ---------- > 1 0x3fb1e8 9704 > 2 0x13c970 2002736 > 3 0x3fb000 472 > 4 0x3f9000 4088 > 5 0x3f7000 4088 > 6 0x3f5000 4088 > 7 0x3f3000 4088 > 8 0x3f1000 4088 > 9 0x3ef000 4088 > 10 0x3ed000 4088 > 11 0x3eb000 4088 > 12 0x3e9000 4088 > 13 0x3e7000 4088 > 14 0x3e5000 4088 > 15 0x3e3000 4088 > 16 0x3e1000 4088 > 17 0x3df000 4088 > 18 0x3dd000 4088 > 19 0x3db000 4088 > 20 0x3d9000 4088 > 21 0x3d7000 4088 > 22 0x3d5000 4088 > 23 0x3d3000 4088 > 24 0x3d1000 4088 > 25 0x3cf000 4088 > 26 0x3cd000 4088 > 27 0x3cb000 4088 > 28 0x3c9000 4088 > 29 0x3c7000 4088 > 30 0x3c5000 4088 > 31 0x3c3000 4072 > > SUMMARY: > status bytes blocks avg block max block > ------ --------- -------- ---------- ---------- > current > free 2127360 31 68624 2002736 > alloc 770168 156 4936 - > cumulative > alloc 901328 159 5668 - > ------------------------------------------------------------------------- > > but, I want to end mbuf list at addr 0x3e5000. > because I'm using that addr. > > following 'memShow(1)' shows ... > ------------------------------------------------------------------------- > ->memShow(1) > FREE LIST: > num addr size > --- ---------- ---------- > 1 0x3fb1e8 192 > 2 0x13c970 1833512 > 3 0x3fb000 472 > 4 0x3f9000 4088 > 5 0x3f7000 4088 > 6 0x3f5000 4088 > 7 0x3f3000 4088 > 8 0x3f1000 4088 > 9 0x3ef000 4088 > 10 0x3ed000 4088 > 11 0x3eb000 4088 > 12 0x3e9000 4088 > 13 0x3e7000 4088 > 14 0x3e5000 4088 > 15 0x3e3000 4088 > 16 0x3e1000 4088 > 17 0x3df000 4088 > 18 0x3dd000 4088 > 19 0x3db000 4088 > 20 0x3d9000 4088 > 21 0x3d7000 4088 > 22 0x3d5000 4088 > 23 0x3d3000 4088 > 24 0x3d1000 4088 > invalid block at 0x3cf000 deleted > ------------------------------------------------------------------------ > > error occured. so, system broke down. > > Thanks in advance. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tNetTask Date: Wed, 11 Jul 2001 11:22:04 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994843809.663265@newsmaster-04.atnet.at> References: <9ih000$4dm$1@news1.kornet.net> > But, if I browse Web Pages in 'my System', > my tNetTask or tWeb task go SUSPEND status > > ------------------------------------------- > Data access > Exception current instruction address: 0x00064e6c > Machine Status Register: 0x00001030 > Data Access Register: 0xaaaa0300 > Condition Register: 0x48222042 > Data storage interrupt Register: 0x0000001f > Task: 0x345a98 "tWeb" > -------------------------------------------- > > what's problem? To me this looks like a exception has been raised due to an error in your tWeb task. You can try to get a stack backtrace to see what happended just when the error occured. The Shell Command to get a backtrace is tt It has (most probably) nothing to do with priorities. hth werner > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Help:Request Intel TE28F320C3 driver. Date: 11 Jul 2001 02:54:29 -0700 From: old-lian@163.net (lian) Organization: http://groups.google.com/ Message-ID: Hi, Does anyone have the driver for Intel TE28F320? Please help! Thank you! --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 11 10:25:05 2001 From: Michael.Ben-Ari@ecitele.com Date: Wed Jul 11 10:25:07 PDT 2001 Subject: VxWork & Proxy Arp I have 2 separate CPUs between which there exists IP support via Dual Port Ram (NOT USING THE Shared Memory Driver- our hardware can't support use of SM driver over the bp). IP Works fine! Network setup is as follows: Class C - Network CPU #1 (vx1)is multi-homed: Ethernet Address 192.168.17.26 DPRam Address 192.168.17.129 CPU #2 (vx2) DPRam Address 192.168.17.130 Host #1 (pc) Ethernet Address 192.168.17.61 vx1's routing tables are as follows: -> routeShow ROUTE NET TABLE destination gateway flags Refcnt Use Interface ---------------------------------------------------------------------------- 0.0.0.0 192.168.17.254 3 0 1 lnPci0 192.168.17.0 192.168.17.26 101 0 0 lnPci0 ---------------------------------------------------------------------------- ROUTE HOST TABLE destination gateway flags Refcnt Use Interface ---------------------------------------------------------------------------- 127.0.0.1 127.0.0.1 5 0 0 lo0 192.168.17.129 192.168.17.130 7 0 0 ipodpr4 ---------------------------------------------------------------------------- vx2's routing tables are as follows: -> routeShow ROUTE NET TABLE destination gateway flags Refcnt Use Interface ---------------------------------------------------------------------------- 0.0.0.0 192.168.17.129 3 0 0 ipodpr4 192.168.17.0 192.168.17.27 101 0 0 motfcc0 ---------------------------------------------------------------------------- ROUTE HOST TABLE destination gateway flags Refcnt Use Interface ---------------------------------------------------------------------------- 127.0.0.1 127.0.0.1 5 0 0 lo0 192.168.17.129 192.168.17.130 5 0 177 ipodpr4 ---------------------------------------------------------------------------- I can setup the PC's routing tables to access vx2 by setting up the following route 192.168.17.128 192.168.17.26. However, I would like to allow for the gateway in the middle to recognize vx2 and implement "Proxy Arp" on vx1. Subnet Masks are not possible with the Class 'C' address (as far as I know). "Proxy Arp" support in VxWorks is only supported when using the SharedMemory Support. Question: Anyone have any suggestions as to where I can 1. Find public domain "Proxy Arp" support, or 2. Get VxWorks to support "Proxy Arp" without using the SharedMemory Driver. 3. Any other pointers to suggestions. All help would be appreciated. -Mike ----------------------------------------------- Mike Ben-Ari ECI Telecom - NGTS Ltd. Cellular Division - Celtro Mail: mbenari@ecitele.com Tel: +972-3-926-6794 Fax: +972-3-926-6460 ----------------------------------------------- From vxwexplo-errs@csg.lbl.gov Thu Jul 12 02:23:27 2001 From: Ilia Bosis Date: Thu Jul 12 02:23:29 PDT 2001 Subject: RE: VxWork & Proxy Arp AFAIK, the vxWorks proxy ARP library does not check for the actual Shared Memory driver, just for device_name == "sm" So try to rename Your Dual port RAM driver to "sm" - that may do the trick. Regards Ilia. -------------------------------------------- Ilia Bosis Software Engineer PacketLight Networks Tel: +972-9-7644228 Fax: +972-9-7645444 E-mail: Ilia_Bosis@packetlight.com Web: www.packetlight.com > -----Original Message----- > From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] > Sent: Wednesday, July 11, 2001 7:32 PM > To: vxworks_users@csg.lbl.gov > Subject: VxWork & Proxy Arp > > > Submitted-by vxwexplo-errs@csg.lbl.gov Wed Jul 11 10:25:05 2001 > Submitted-by: Michael.Ben-Ari@ecitele.com > > > I have 2 separate CPUs between which there exists IP support > via Dual Port Ram (NOT USING THE Shared Memory Driver- > our hardware can't support use of SM driver over the bp). > IP Works fine! > > > Network setup is as follows: Class C - Network > > CPU #1 (vx1)is multi-homed: Ethernet Address 192.168.17.26 > DPRam Address 192.168.17.129 > > CPU #2 (vx2) DPRam Address 192.168.17.130 > > Host #1 (pc) Ethernet Address 192.168.17.61 > > > vx1's routing tables are as follows: > > > -> routeShow > ROUTE NET TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 0.0.0.0 192.168.17.254 3 0 1 > lnPci0 > 192.168.17.0 192.168.17.26 101 0 0 > lnPci0 > -------------------------------------------------------------- > -------------- > > ROUTE HOST TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 127.0.0.1 127.0.0.1 5 0 0 lo0 > 192.168.17.129 192.168.17.130 7 0 0 > ipodpr4 > -------------------------------------------------------------- > -------------- > > > vx2's routing tables are as follows: > > > -> routeShow > > ROUTE NET TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 0.0.0.0 192.168.17.129 3 0 0 > ipodpr4 > 192.168.17.0 192.168.17.27 101 0 0 > motfcc0 > -------------------------------------------------------------- > -------------- > > ROUTE HOST TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 127.0.0.1 127.0.0.1 5 0 0 lo0 > 192.168.17.129 192.168.17.130 5 0 177 > ipodpr4 > -------------------------------------------------------------- > -------------- > > > I can setup the PC's routing tables to access vx2 by setting up the > following > route 192.168.17.128 192.168.17.26. However, I would like to > allow for the > gateway > in the middle to recognize vx2 and implement "Proxy Arp" on > vx1. Subnet > Masks > are not possible with the Class 'C' address (as far as I know). > > "Proxy Arp" support in VxWorks is only supported when using the > SharedMemory Support. > > Question: > > Anyone have any suggestions as to where I can > > 1. Find public domain "Proxy Arp" support, or > 2. Get VxWorks to support "Proxy Arp" without using > the SharedMemory Driver. > 3. Any other pointers to suggestions. > > All help would be appreciated. > > -Mike > > ----------------------------------------------- > Mike Ben-Ari > ECI Telecom - NGTS Ltd. > Cellular Division - Celtro > Mail: mbenari@ecitele.com > Tel: +972-3-926-6794 > Fax: +972-3-926-6460 > ----------------------------------------------- From vxwexplo-errs@csg.lbl.gov Thu Jul 12 04:03:25 2001 From: Vxworks Exploder Date: Thu Jul 12 04:03:27 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 12 04:03:20 PDT 2001 Subject: Need Help - PowerPC MPC8xx SDRAM Setup Subject: Help: VxWorks 5.4 with if_fei driver Subject: about windML2.0 Subject: Re: Problem of reading of the serial port Subject: Command line for building Subject: Re: Command line for building Subject: Re: Problem of reading of the serial port Subject: Re: about windML2.0 Subject: booting PC with VxWorks from Hard Disk - how to? Subject: Re: Problem of reading of the serial port Subject: tffsRawio Subject: Re: TFFS Alignment problem Subject: Re: booting PC with VxWorks from Hard Disk - how to? Subject: Re: How to do inline assembly in C BSP ??? Subject: Re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Subject: VxWork & Proxy Arp Subject: Re: tNetTask Subject: Re: Help: VxWorks 5.4 with if_fei driver Subject: Re: tffsRawio Subject: Re: Is Wind River really this pathetic Subject: Additional Sections in Bootrom (was: Absolute Linking) Subject: Re: Need Help - PowerPC MPC8xx SDRAM Setup Subject: RE: VxWork & Proxy Arp Subject: HDLC Freedm-32 chip, MUX Driver? ------------------------------------------------------- Newsgroups: comp.arch.embedded,comp.os.linux.hardware,comp.os.linux.powerpc,comp.os.vxworks,comp.sys.powerpc.tech Subject: Need Help - PowerPC MPC8xx SDRAM Setup Date: Wed, 11 Jul 2001 11:36:04 GMT From: "Jeff Gentry" Organization: Verio Message-ID: <3b4c3dae@news.integrityonline.com> Anyone using Micron Semiconductor's 64 Mbit SDRAM, MT48LC4M16A2TG, 4M x 16, with a PowerPC design? If so, I am working with a hardware design that is having problems executing out of SDRAM. It may be an initialization or UPM setup problem. I have init code and UPM tables and want to verify them. Thanks, Jeff Gentry, Sr. Software Design Engineer PEI Electronics, An Integrated Defense Technologies Company Email: JGentry@pei-idt.com Voice: (256) 895-2106 Fax: (256) 895-2222 --------------------------- Newsgroups: comp.os.vxworks Subject: Help: VxWorks 5.4 with if_fei driver Date: Wed, 11 Jul 2001 13:40:52 +0200 From: "Stefan Trampe" Organization: Telelogic Germany Message-ID: <9ihe67$tdk$1@rot.telelogic.de> Hi, please can somebody help me ? I have a PMC 3101 Dual-Ethernet card with a Intel 82559ER Ethernet chip. I want to use this card with VxWorks 5.4. .VxWorks includes the if_fei driver for the Intel chip. But I don't know how I must use this driver. Especially the meaning of the parameter of the user defineable function "STATUS sys557Init(int unit, BOARD_INFO *pBoard)". Many thanks for help Stefan --------------------------- Newsgroups: comp.os.vxworks Subject: about windML2.0 Date: 11 Jul 2001 04:57:44 -0700 From: gcai@netease.com (cai) Organization: http://groups.google.com/ Message-ID: <31f1a7e2.0107110357.7457394b@posting.google.com> hi, Is there anyone has used windML 2.0 (UGL 2.0)? I have two questions: 1. How to configure windML to use touch screen? In windML programmer's guide,It says Touch screen device is /touchscreen/0 .but I can't find it in input device's pointer configuration. 2. My distributor give me a unicode font which is ugly.Can you tell me where I can find another one? thanks in advance. gcai --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem of reading of the serial port Date: 11 Jul 2001 05:53:07 -0700 From: ilandrei@mail.ru (Andrey Iljitchev) Organization: http://groups.google.com/ Message-ID: <28d45ddc.0107110453.6c6f391d@posting.google.com> References: <28d45ddc.0107100426.4a5d8f40@posting.google.com> Yes, I already tried to incorporate with a PC, was convinced, that recording in the serial port passes - the PC on the other end reads that was written. But I can nothing read from the serial port. I used also gets() function for stdin and read() function for the device/tyCo/0 - no results. Prompt me, that I do incorrectly "Johan Borkhuis" wrote in message news:... > ilandrei@mail.ru (Andrey Iljitchev) wrote: > > > Hi > > > > I am trying to connect two MBX860-cards with null-modem through the > > serial port. I am succesfully writing to the serial port, but I can not > > read anything on the other end. > > > > How it is possible to solve this problem? > > A suggestion: connect a PC to the serial port of the board, check if you > can read and write from this port. Do the same with the other board, and > see if you can find a problem. If this does not show any problem, check > your cable. > > Groeten, > Johan --------------------------- Newsgroups: comp.os.vxworks Subject: Command line for building Date: Wed, 11 Jul 2001 14:02:32 GMT From: "Stefano" Message-ID: Hi, I am a begginer with Tornado, I would like to compile different projects using a DOS shell at the same time. These projects are build with Tornado 2.0 and I don't know how to produce the make file from the .wpj files without using the GUI. Could anyone help me? Thank you in advance. Bye Stefano --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Command line for building Date: Wed, 11 Jul 2001 07:36:24 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B4C6468.8A9D4923@covad.net> References: Reply-To: drdiags@covad.net Stefano, The technique is documented on the vxWorks FAQ list maintained at the link: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html Look in the section on the Project Facility (I think). HTH. Stefano wrote: > > Hi, > > I am a begginer with Tornado, I would like to compile different projects > using a DOS shell at the same time. These projects are build with Tornado > 2.0 and I don't know how to produce the make file from the .wpj files > without using the GUI. Could anyone help me? > Thank you in advance. > > Bye > Stefano - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem of reading of the serial port Date: 11 Jul 2001 14:56:31 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <28d45ddc.0107100426.4a5d8f40@posting.google.com> <28d45ddc.0107110453.6c6f391d@posting.google.com> ilandrei@mail.ru (Andrey Iljitchev) wrote: > Yes, I already tried to incorporate with a PC, was convinced, that > recording in the serial port passes - the PC on the other end reads > that was written. But I can nothing read from the serial port. I used > also gets() function for stdin and read() function for the > device/tyCo/0 - no results. Prompt me, that I do incorrectly Looks to me like a problem with the serial driver. It works the same on both 860-boards? If that is the case then there are 2 possibilities: 1. Problem with your routine. Could it be that your read-function does not work properly? 2. Problem with the driver. You might have to dive into this. One easy way to check if the driver is working is to add a logMsg call to the receive interrupt routine. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about windML2.0 Date: Wed, 11 Jul 2001 17:10:05 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994864682.712499@newsmaster-04.atnet.at> References: <31f1a7e2.0107110357.7457394b@posting.google.com> > 1. How to configure windML to use touch screen? > In windML programmer's guide,It says Touch screen device is > /touchscreen/0 .but I can't find it in input device's pointer > configuration. /touchscreen/0 ist just a name for the input device that will be created for the touch screen driver provided with VxWorks. This name does not really matter. If your touchscreen is not of the type of driver that comes with VxWorks (Assabet Touch Screen) you need to write your own driver. Forget the configuration utility, it cannot cope with custom drivers. You need to build your UGL image on the command line (or, at your option, hack their config tool ;-) > > 2. My distributor give me a unicode font which is ugly.Can you tell me > where I can find another one? > Agfa provides Font Engines that can render vector fonts and sells the appropriate fonts. Their fonts are quite nice in look and there is lots of choice, but pricing is "adequate" *g*. Is that Unicode font you got from WRS? Or a free font? Just curious, because we are looking for one for test purposes. regards werner --------------------------- Newsgroups: comp.os.vxworks Subject: booting PC with VxWorks from Hard Disk - how to? Date: Wed, 11 Jul 2001 11:31:56 -0700 From: "Sean Bose" Organization: Lucent Technologies, Columbus, Ohio Message-ID: <9ihr60$n03@nntpb.cb.lucent.com> Hi Friends I can make a stand alone VxWorks image, vxworks.st, for an x86 PC , which I can boot from a floppy drive. I would like to know, what I need to do to boot a PC with VxWorks image from a hard disk. My image with drivers is too big for a floppy. thanks for your help, Sean --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem of reading of the serial port Date: Wed, 11 Jul 2001 11:40:30 -0700 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <3B4C9D9E.9538A9F7@lmco.com> References: <28d45ddc.0107100426.4a5d8f40@posting.google.com> <28d45ddc.0107110453.6c6f391d@posting.google.com> Just a SWAG, but you might check the state of the control signals on the serial port. Some drivers, by default will not receive data unless they get the right handshakes from the other end. Most of the time a null-modem solves these problems, but maybe not in all cases. You might want to see, at the receiving device, the state of pins 1, 4, 6, and 8 (assuming the use of 9-pin D connectors). Depending on whether your card is configured DTE or DCE, some of these will be driven by the card, others will be controlled by the other end. Speaking only for myself, Joe Durusau Andrey Iljitchev wrote: > > Yes, I already tried to incorporate with a PC, was convinced, that recording > in the serial port passes - the PC on the other end reads that was written. > But I can nothing read from the serial port. I used also gets() function > for stdin and read() function for the device/tyCo/0 - no results. > Prompt me, that I do incorrectly > > "Johan Borkhuis" wrote in message news:... > > ilandrei@mail.ru (Andrey Iljitchev) wrote: > > > > > Hi > > > > > > I am trying to connect two MBX860-cards with null-modem through the > > > serial port. I am succesfully writing to the serial port, but I can not > > > read anything on the other end. > > > > > > How it is possible to solve this problem? > > > > A suggestion: connect a PC to the serial port of the board, check if you > > can read and write from this port. Do the same with the other board, and > > see if you can find a problem. If this does not show any problem, check > > your cable. > > > > Groeten, > > Johan --------------------------- Newsgroups: comp.os.vxworks Subject: tffsRawio Date: 11 Jul 2001 09:41:21 -0700 From: pathomas40@hotmail.com (Philip Thomas) Organization: http://groups.google.com/ Message-ID: I want to use tffsRawio to write to a fallow area of FLASH before the TFFS, but TFFS_PHYSICAL_WRITE and TFFS_PHYSICAL_ERASE always return - -1. I discovered that the operations succeed if called after tffsDrv, but before tffsDevCreate, but not after tffsDevCreate. I may have to live with this limitation, but I would like to be able to write to the fallow area after the TFFS has been mounted. Does anyone know how I can have my cake and eat it too?? Thanks Phil --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS Alignment problem Date: Wed, 11 Jul 2001 11:06:12 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B4C6B64.21B8A665@digidata.com> References: <3m6qh9.oqq.ln@gum.de> You have to modify the MTD for 32-bit access. TFFS does not support it directly. Our FAE offered to give me some unsupported MTD source code that has been modified for 32-bit access. I think it is the CFI/SCS MTD. You can reach him at mark.hamilton@windriver.com Doug Frank Wolf wrote: > Hi there! > > We're using TFFS on Hitachi SH3 which has a 32Bit wide access to > RAM and Flash memory. > > When writing data to the TFFS drive we get an exception as the > function 'checkForWriteInplace' (ftllite.o) is trying to write to a > 2byte aligned address instead of 4byte aligned. > > Does anyone know how to solve this problem? WRS support hasn't > been of much help so far... :-( > > Thanks alot, > > Frank Wolf --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting PC with VxWorks from Hard Disk - how to? Date: Wed, 11 Jul 2001 19:03:17 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994871475.298837@newsmaster-04.atnet.at> References: <9ihr60$n03@nntpb.cb.lucent.com> Hi, In your bootrom: Set the boot device to ata=0,0 and the file name to /ata0/ (you can do so interactively at startup, but I don't know if the default Bootrom contains the ATA driver) be sure, INCLUDE_ATA is defined when compiling the bootrom hth werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to do inline assembly in C BSP ??? Date: Wed, 11 Jul 2001 16:43:17 +0100 From: "Dave Korn" Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser refresh) Message-ID: <8P%27.103$os1.458204@newsr2.u-net.net> References: <3B44638A.1E60DC64@americasm01.nt.com> David Given wrote in message ... >You may draw your own conclusions. >If you really have to write big chunks of assembly, I'd strongly recommend >putting it in a seperate .S file and defining constants to represent >things like registers. My own conclusion is that you should use the compiler flag -xassembler-with-cpp and then #include which gives you loads of lovely #defines for your registers. DaveK - -- They laughed at Galileo. They laughed at Copernicus. They laughed at Columbus. But remember, they also laughed at Bozo the Clown. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Date: Wed, 11 Jul 2001 16:51:00 +0100 From: "Dave Korn" Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser refresh) Message-ID: <8P%27.104$os1.458451@newsr2.u-net.net> References: <9i145t$65d$1@overload.lbl.gov> fscheurer@manz-acs.de wrote in message <9i145t$65d$1@overload.lbl.gov>... >Hi VxWorks-Gurus, > >I ran into the problem, that the integrated target-headers >seem to be not confor with GNU expectations. Did you use the --with-headers= flag when configuring to supply the target headers, or did you copy them into $prefix/$target/sys-include, or into $prefix/$target/include ? The first two of those methods will cause the Gcc build process to invoke the 'fixincs' script, which runs through all the supplied headers, checking for and fixing things that need to be made GNU-compatible. Although it is recommended in many pieces of documentation that you install target headers into $prefix/$target/include when building a cross compiler, doing so prevents the fixincs from being run: fixincs will only ever process files from the sys-include directory. The error messages you posted look a lot like fixincs hasn't run. Can you confirm or deny this? >MY QUESTION: >Where is the most central place in >- the sources, >- the configure-script >or >- the Makefile > >to "repair" it >AND to get the right types to be compatible to objects build with >older gcc-version? I suspect that the fixes already exist in the fixincs script, and merely haven't been applied to your target headers in the course of building the cross-compiler. However, the VxWorks targets for Gcc haven't been well maintained of late, and it is possible that the WRS headers have changed in such a way that the fixincs test-pattern no longer matches the WRS header file, causing a fix to be missed; or that new code in the headers mean that new fixes need to be added. Give it another try, being sure to use one of the first two methods I mentioned above to supply the headers, and tell us if it works any better. DaveK - -- They laughed at Galileo. They laughed at Copernicus. They laughed at Columbus. But remember, they also laughed at Bozo the Clown. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWork & Proxy Arp Date: Wed, 11 Jul 2001 20:24:49 +0300 From: Michael.Ben-Ari@ecitele.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9ii53o$5is$1@overload.lbl.gov> I have 2 separate CPUs between which there exists IP support via Dual Port Ram (NOT USING THE Shared Memory Driver- our hardware can't support use of SM driver over the bp). IP Works fine! Network setup is as follows: Class C - Network CPU #1 (vx1)is multi-homed: Ethernet Address 192.168.17.26 DPRam Address 192.168.17.129 CPU #2 (vx2) DPRam Address 192.168.17.130 Host #1 (pc) Ethernet Address 192.168.17.61 vx1's routing tables are as follows: - -> routeShow ROUTE NET TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 0.0.0.0 192.168.17.254 3 0 1 lnPci0 192.168.17.0 192.168.17.26 101 0 0 lnPci0 - ---------------------------------------------------------------------------- ROUTE HOST TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 127.0.0.1 127.0.0.1 5 0 0 lo0 192.168.17.129 192.168.17.130 7 0 0 ipodpr4 - ---------------------------------------------------------------------------- vx2's routing tables are as follows: - -> routeShow ROUTE NET TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 0.0.0.0 192.168.17.129 3 0 0 ipodpr4 192.168.17.0 192.168.17.27 101 0 0 motfcc0 - ---------------------------------------------------------------------------- ROUTE HOST TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 127.0.0.1 127.0.0.1 5 0 0 lo0 192.168.17.129 192.168.17.130 5 0 177 ipodpr4 - ---------------------------------------------------------------------------- I can setup the PC's routing tables to access vx2 by setting up the following route 192.168.17.128 192.168.17.26. However, I would like to allow for the gateway in the middle to recognize vx2 and implement "Proxy Arp" on vx1. Subnet Masks are not possible with the Class 'C' address (as far as I know). "Proxy Arp" support in VxWorks is only supported when using the SharedMemory Support. Question: Anyone have any suggestions as to where I can 1. Find public domain "Proxy Arp" support, or 2. Get VxWorks to support "Proxy Arp" without using the SharedMemory Driver. 3. Any other pointers to suggestions. All help would be appreciated. - -Mike - ----------------------------------------------- Mike Ben-Ari ECI Telecom - NGTS Ltd. Cellular Division - Celtro Mail: mbenari@ecitele.com Tel: +972-3-926-6794 Fax: +972-3-926-6460 - ----------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tNetTask Date: 11 Jul 2001 11:50:20 -0700 From: tlksk@hotmail.com (Kumar) Organization: http://groups.google.com/ Message-ID: References: <9ih000$4dm$1@news1.kornet.net> <994843809.663265@newsmaster-04.atnet.at> Hi It is better to put all the necesary info and ask what is the problem. Rather than just sending the exception message. Do either "tt tNetTask" or "tt tWeb" and send the info to the group. Kumar "Werner Schiendl" wrote in message news:<994843809.663265@newsmaster-04.atnet.at>... > > But, if I browse Web Pages in 'my System', > > my tNetTask or tWeb task go SUSPEND status > > > > ------------------------------------------- > > Data access > > Exception current instruction address: 0x00064e6c > > Machine Status Register: 0x00001030 > > Data Access Register: 0xaaaa0300 > > Condition Register: 0x48222042 > > Data storage interrupt Register: 0x0000001f > > Task: 0x345a98 "tWeb" > > -------------------------------------------- > > > > what's problem? > > To me this looks like a exception has been raised due to an error in your > tWeb task. > You can try to get a stack backtrace to see what happended just when the > error occured. > The Shell Command to get a backtrace is tt > > It has (most probably) nothing to do with priorities. > > hth > werner > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Help: VxWorks 5.4 with if_fei driver Date: 11 Jul 2001 12:06:35 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107111106.46fea21d@posting.google.com> References: <9ihe67$tdk$1@rot.telelogic.de> Hello, Search the archives of comp.os.vxworks at google.com or on WindSurf for this. Support for the ER variant of the 82559 is a topic that has come up before. It has a different PCI ID from the standard 82559 which means that it is not seen by the default fei support code, but I believe that it can be changed, and details of how to do so have been posted to this group in the past. HTH, John... "Stefan Trampe" wrote in message news:<9ihe67$tdk$1@rot.telelogic.de>... > Hi, > please can somebody help me ? > I have a PMC 3101 Dual-Ethernet card with a Intel 82559ER Ethernet chip. I > want to use this card with VxWorks 5.4. .VxWorks includes the if_fei driver > for the Intel chip. But I don't know how I must use this driver. Especially > the meaning of the parameter of the user defineable function "STATUS > sys557Init(int unit, BOARD_INFO *pBoard)". > > Many thanks for help > Stefan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tffsRawio Date: Wed, 11 Jul 2001 20:08:01 GMT From: andreyk@home.com (Andray Kaganovsky) Organization: @Home Message-ID: References: pathomas40@hotmail.com (Philip Thomas) wrote in news:a98a1bc8.0107110841.4f6e231b@posting.google.com: > I want to use tffsRawio to write to a fallow area of FLASH before the > TFFS, but TFFS_PHYSICAL_WRITE and TFFS_PHYSICAL_ERASE always return > -1. I discovered that the operations succeed if called after tffsDrv, > but before tffsDevCreate, but not after tffsDevCreate. I may have to > live with this limitation, but I would like to be able to write to the > fallow area after the TFFS has been mounted. Does anyone know how I > can have my cake and eat it too?? > > Thanks > > Phil > [AndrayK]. It's true that TFFS blocks these two _PHYSICAL_ calls after tffsDevCreate() has been called. Here is an idea that I hope will be of interest to you. Try installing second socket component that is absolutely identical to the first one (i.e. it points to exactly the same flash array). You will be using first socket components as usual (i.e. tffsDevCreate() etc.), and the second one exclusively for _PHYSICAL_ calls. Please make sure to flush filesystem cache, and hold all application's tasks off the first socket while using second socket for _PHYSICAL_ calls. Andray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Thu, 12 Jul 2001 03:21:27 GMT From: "ag" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> "Joe Woodbury" wrote in message news:6qH17.79$5f6.133962@news.uswest.net... > Since we ordered VxWorks/Tornado II, we've had nothing but problems with > Wind River. They filled the order incorrectly and I had to resort to a lot > of yelling and threats to not pay our purchase order to get their attention. > Yet, two days later, they still have yet to do anything but email the > correct BSP. On top of all this, the highest level manager at Wind River > kept blaming us for some of the problems. > > So far, I've concluded that Wind River employees are given an IQ test. If > they fail, they are given a personal relationship test. If they fail that > also, they are hired. > > And heaven forbid they actually do something so obvious like creating a > compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. > > The shame is that VxWorks proper is such an excellent product. > > The questions are: Are our experiences unique? Did our company just happen > to run into the one group of bastards working at Wind River? > > Joe Woodbury > > No, this type of experience is not unique to you. --------------------------- Newsgroups: comp.os.vxworks Subject: Additional Sections in Bootrom (was: Absolute Linking) Date: Thu, 12 Jul 2001 08:57:02 +0200 From: "Michael Lawnick" Organization: Buergernetz Dillingen Message-ID: <9ijhvf$1fp$1@snoopy.bndlg.de> Reply-To: "Michael Lawnick" Sender: mlawnick@pd9e65689.dip.t-dialin.net Hi VxWorks Gurus, since the old thread (Absolute Linking) starved, again the question: Is there anybody, who added additional sections to bootrom for locating data or code to a fixed place ? The problem: Our PPC405 starts at 0xFFFFFFFC and bootcode starts at 0xFFF80100. So we have to extend bootrom.hex with an additional S2-record that contains the jump code. By now this is done by hand, inserting the line with sed is not so easy (DOS/UNIX file format -> CR/LF problems). Our idea is to create a new linker section, assigning the jump instruction to it and tell linker to locate the section to 0xFFFFFFFC. We fail to do this, as it seems, when breaking the default setup of sections, we have to redefine all ! Again the question: has anybody already done this or is there any other idea to get the instruction to the end of bootrom in a host OS independant way ? - -- Mit freundlichen Grüßen, Michael Lawnick ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== --------------------------- Newsgroups: comp.arch.embedded,comp.os.linux.hardware,comp.os.linux.powerpc,comp.os.vxworks,comp.sys.powerpc.tech Subject: Re: Need Help - PowerPC MPC8xx SDRAM Setup Date: Thu, 12 Jul 2001 09:58:11 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: References: <3b4c3dae@news.integrityonline.com> Hi Jeff! We're using PPC823 with VxWorks. Maybe it helps when I send you our 'romInit.s' module which does the initialization of the PPC? Frank "Jeff Gentry" schrieb im Newsbeitrag news:3b4c3dae@news.integrityonline.com... > Anyone using Micron Semiconductor's 64 Mbit SDRAM, > MT48LC4M16A2TG, 4M x 16, with a PowerPC design? > > If so, I am working with a hardware design that is > having problems executing out of SDRAM. It may be > an initialization or UPM setup problem. > > I have init code and UPM tables and want to verify > them. > > Thanks, > > Jeff Gentry, > Sr. Software Design Engineer > PEI Electronics, > An Integrated Defense Technologies Company > Email: JGentry@pei-idt.com > Voice: (256) 895-2106 > Fax: (256) 895-2222 > > > --------------------------- Newsgroups: comp.os.vxworks Subject: RE: VxWork & Proxy Arp Date: Thu, 12 Jul 2001 12:21:31 +0200 From: Ilia Bosis Organization: Lawrence Berkeley National Laboratory Message-ID: <9ijtba$18s$1@overload.lbl.gov> AFAIK, the vxWorks proxy ARP library does not check for the actual Shared Memory driver, just for device_name == "sm" So try to rename Your Dual port RAM driver to "sm" - that may do the trick. Regards Ilia. - -------------------------------------------- Ilia Bosis Software Engineer PacketLight Networks Tel: +972-9-7644228 Fax: +972-9-7645444 E-mail: Ilia_Bosis@packetlight.com Web: www.packetlight.com > -----Original Message----- > From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] > Sent: Wednesday, July 11, 2001 7:32 PM > To: vxworks_users@csg.lbl.gov > Subject: VxWork & Proxy Arp > > > Submitted-by vxwexplo-errs@csg.lbl.gov Wed Jul 11 10:25:05 2001 > Submitted-by: Michael.Ben-Ari@ecitele.com > > > I have 2 separate CPUs between which there exists IP support > via Dual Port Ram (NOT USING THE Shared Memory Driver- > our hardware can't support use of SM driver over the bp). > IP Works fine! > > > Network setup is as follows: Class C - Network > > CPU #1 (vx1)is multi-homed: Ethernet Address 192.168.17.26 > DPRam Address 192.168.17.129 > > CPU #2 (vx2) DPRam Address 192.168.17.130 > > Host #1 (pc) Ethernet Address 192.168.17.61 > > > vx1's routing tables are as follows: > > > -> routeShow > ROUTE NET TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 0.0.0.0 192.168.17.254 3 0 1 > lnPci0 > 192.168.17.0 192.168.17.26 101 0 0 > lnPci0 > -------------------------------------------------------------- > -------------- > > ROUTE HOST TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 127.0.0.1 127.0.0.1 5 0 0 lo0 > 192.168.17.129 192.168.17.130 7 0 0 > ipodpr4 > -------------------------------------------------------------- > -------------- > > > vx2's routing tables are as follows: > > > -> routeShow > > ROUTE NET TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 0.0.0.0 192.168.17.129 3 0 0 > ipodpr4 > 192.168.17.0 192.168.17.27 101 0 0 > motfcc0 > -------------------------------------------------------------- > -------------- > > ROUTE HOST TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 127.0.0.1 127.0.0.1 5 0 0 lo0 > 192.168.17.129 192.168.17.130 5 0 177 > ipodpr4 > -------------------------------------------------------------- > -------------- > > > I can setup the PC's routing tables to access vx2 by setting up the > following > route 192.168.17.128 192.168.17.26. However, I would like to > allow for the > gateway > in the middle to recognize vx2 and implement "Proxy Arp" on > vx1. Subnet > Masks > are not possible with the Class 'C' address (as far as I know). > > "Proxy Arp" support in VxWorks is only supported when using the > SharedMemory Support. > > Question: > > Anyone have any suggestions as to where I can > > 1. Find public domain "Proxy Arp" support, or > 2. Get VxWorks to support "Proxy Arp" without using > the SharedMemory Driver. > 3. Any other pointers to suggestions. > > All help would be appreciated. > > -Mike > > ----------------------------------------------- > Mike Ben-Ari > ECI Telecom - NGTS Ltd. > Cellular Division - Celtro > Mail: mbenari@ecitele.com > Tel: +972-3-926-6794 > Fax: +972-3-926-6460 > ----------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: HDLC Freedm-32 chip, MUX Driver? Date: 11 Jul 2001 16:01:22 GMT From: afernando@adventnetworks.com Organization: NewsOne.Net - Free Usenet News via the Web - http://newsone.net/ Message-ID: <9iht8i$pl7$1@news.netmar.com> Followup-To: comp.os.vxworks hi all, Does PMC-Sierra provide its HDLC framing engine (FREEDM32 chip) with a MUX driver for Vxworks (ppc750)? Has anyone seen a MUX driver a HDLC chips in general? please advice.. Anton Austin, TX. ----- Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web ----- http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups NewsOne.Net prohibits users from posting spam. If this or other posts made through NewsOne.Net violate posting guidelines, email abuse@newsone.net --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jul 12 07:48:18 2001 From: Kumaraswamy Mudide Date: Thu Jul 12 07:48:20 PDT 2001 Subject: CD-ROM file system Hi All, I have seen some where in the documentation, vxWorks cd-rom file system does not support write feature. Is this true? If this is true, and if I want to write to CD-ROM form vxWorks OS what should I do. Please help us. Thanks, Swamy __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ From vxwexplo-errs@csg.lbl.gov Thu Jul 12 21:32:48 2001 From: "rajesh biswal" Date: Thu Jul 12 21:32:51 PDT 2001 Subject: VxWorks - Booting Problem This is a multi-part message in MIME format. --------------InterScan_NT_MIME_Boundary Content-Type: multipart/alternative; boundary="----=_NextPart_000_0007_01C10B82.F049B1A0" ------=_NextPart_000_0007_01C10B82.F049B1A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello, We are trying to boot the Pentium Target through ethernet. The specification of Driver is: AcerLAN ALN-325 10/100 Base-TX Fast Ethernet Device Type: Network Adapters. Manufacturer: Acer Netxus Inc. H/w Version: 016 Interrupt Request: 10 Memory Range : 09100000-091000FF Input/Output Range: 7000-70FF What are the settings required in VxWorks Configuration. We are using T2. Thanx in advance, Regards, Rajesh Biswal. **************************************************************************** ************ Embedded & Internet Access Division Wipro Technologies Address : 30, Mission Road 1st Main, S.R Nagar Bangalore, 560027 Tel : 91-80-2275590/2241730 Extn 1145 Fax : 91-80-2241769 E-mail : rajesh.biswal@wipro.com Web Page : www.wipro.com The World's First SEI CMM Level 5 Software Services company **************************************************************************** ************ Silence and thinking never hurt anyone. ------=_NextPart_000_0007_01C10B82.F049B1A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello,
We are = trying to=20 boot the Pentium Target through ethernet.
The = specification of=20 Driver is:
 
AcerLAN ALN-325=20 10/100 Base-TX Fast Ethernet
Device = Type: Network=20 Adapters.
Manufacturer: Acer=20 Netxus Inc.
H/w = Version:=20 016
 
Interrupt=20 Request:     10
Memory = Range :       09100000-091000FF
Input/Output=20 Range:  7000-70FF
 
What = are the=20 settings required in VxWorks Configuration.
We are = using=20 T2.
 
Thanx = in=20 advance,
Regards,
Rajesh Biswal.
****************************************************************= ************************=20
Embedded & Internet Access Division
Wipro=20 Technologies
Address    : 30, Mission = Road
          &nbs= p;      =20 1st Main, S.R Nagar
          &nbs= p;      =20 Bangalore, 560027
Tel          &= nbsp; :=20 91-80-2275590/2241730 Extn 1145
Fax           = :=20 91-80-2241769
E-mail       : rajesh.biswal@wipro.com
Web = Page : www.wipro.com
 
The World's First SEI CMM Level 5 = Software Services=20 company
****************************************************************= ************************=20
 Silence and thinking never hurt = anyone.
 
------=_NextPart_000_0007_01C10B82.F049B1A0-- --------------InterScan_NT_MIME_Boundary Content-Type: text/plain; name="Wipro_Disclaimer.txt" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="Wipro_Disclaimer.txt" The Information contained and transmitted by this E-MAIL is proprietary to Wipro Limited and is intended for use only by the individual or entity to which it is addressed, and may contain information that is privileged, confidential or exempt from disclosure under applicable law. If this is a forwarded message, the content of this E-MAIL may not have been sent with the authority of the Company. If you are not the intended recipient, an agent of the intended recipient or a person responsible for delivering the information to the named recipient, you are notified that any use, distribution, transmission, printing, copying or dissemination of this information in any way or in any manner is strictly prohibited. If you have received this communication in error, please delete this mail & notify us immediately at mailadmin@wipro.com --------------InterScan_NT_MIME_Boundary-- From vxwexplo-errs@csg.lbl.gov Fri Jul 13 01:32:07 2001 From: fscheurer@manz-acs.de Date: Fri Jul 13 01:32:09 PDT 2001 Subject: Re-Re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Hi Dave, Thanx for your answer. Based on your Homepage and with the help of J. Pisano (he did it for Linux host or Solaris) I advanced another step. I apologize the long posting, but now I ran into the detail problems. So here my two answers and the state where I am now. I am grateful for every hint, about the options I forgot or what Headers to copy where. I suppose no one tried that host-target X-ing before, otherwise the person should have been run into the same problems. Everyone is happy if it works for his own host. Especially the FIND/SORT error messages I cannot track anymore. Where's their source? Details below: -------- Original Message -------- .. > > Newsgroups: comp.os.vxworks > Subject: Re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 > Date: Wed, 11 Jul 2001 16:51:00 +0100 > From: "Dave Korn" > Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser > refresh) > Message-ID: <8P%27.104$os1.458451@newsr2.u-net.net> > References: <9i145t$65d$1@overload.lbl.gov> > > fscheurer@manz-acs.de wrote in message <9i145t$65d$1@overload.lbl.gov>... > >Hi VxWorks-Gurus, > > > >I ran into the problem, that the integrated target-headers > >seem to be not confor with GNU expectations. > > Did you use the --with-headers= flag when configuring to supply the > target > headers, or did you copy them into $prefix/$target/sys-include, or into > $prefix/$target/include ? Both no in that first stage! Is it to directly include $WIND_BASE/target/h there with all it's sub-directories? > > The first two of those methods will cause the Gcc build process to invoke > the 'fixincs' script, which runs through all the supplied headers, checking > for and fixing things that need to be made GNU-compatible. > > Although it is recommended in many pieces of documentation that you > install target headers into $prefix/$target/include when building a cross > compiler, doing so prevents the fixincs from being run: fixincs will only > ever process files from the sys-include directory. > > The error messages you posted look a lot like fixincs hasn't run. Can > you > confirm or deny this? > In that previously mentionned first stage probably not. In Jim's script this is done: .. #cd gcc-2.95.3/ cd ./gcc-2.95.3/ # first, fix includes with patches #cd gcc/fixinc cd ./gcc/fixinc ./genfixes cd ../.. mkdir build #cd build cd ./build ./configure --prefix=$WIND_BASE/host/$WIND_HOST_TYPE --target=$VX_TARGET --with-gnu-as --with-gnu-ld --with-headers=$WIND_BASE/target/h --enable-languages=c,c++ -v 2>&1 | tee config.log ## FS 2001-07-10 make clean # make LANGUAGES="c c++" all install 2>&1 | tee build1.log .. > >MY QUESTION: > >Where is the most central place in > >- the sources, > >- the configure-script > >or > >- the Makefile > > > >to "repair" it > >AND to get the right types to be compatible to objects build with > >older gcc-version? > > I suspect that the fixes already exist in the fixincs script, and merely > haven't been applied to your target headers in the course of building the > cross-compiler. However, the VxWorks targets for Gcc haven't been well > maintained of late, and it is possible that the WRS headers have changed in > such a way that the fixincs test-pattern no longer matches the WRS header > file, causing a fix to be missed; or that new code in the headers mean that > new fixes need to be added. Give it another try, being sure to use one of > the first two methods I mentioned above to supply the headers, and tell us > if it works any better. > Jim's script uses autogen which I had to build in advance. I let it all run through but encounter some other problems using gcc. First I didn't realise that autogen wasn't build right. I saw during build of cross-compiler in build.log: .. for name in _muldi3 _divdi3 _moddi3 _udivdi3 _umoddi3 _negdi2 _lshrdi3 _ashldi3 _ashrdi3 _ffsdi2 _udiv_w_sdiv _udivmoddi4 _cmpdi2 _ucmpdi2 _floatdidf _floatdisf _fixunsdfsi _fixunssfsi _fixunsdfdi _fixdfdi _fixunssfdi _fixsfdi _fixxfdi _fixunsxfdi _floatdixf _fixunsxfsi _fixtfdi _fixunstfdi _floatditf __gcc_bcmp _varargs __dummy _eprintf _bb _shtab _clear_cache _trampoline __main _exit _ctors _pure; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/ -B/Tornado2/host/winnt/i386-wrs-vxworks/bin/ -I/Tornado2/host/winnt/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN_GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -I. -I../../gcc -I../../gcc/config -I../../gcc/../include -c -DL${name} \ ../../gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _muldi3 ./../gcc/libgcc2.c:41: stdlib.h: No such file or directory ./../gcc/libgcc2.c:42: unistd.h: No such file or directory make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc' make: *** [all-gcc] Error 2 .. The headers are really not there! Some steps before I found: ... make[2]: Entering directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc' autogen: not found You need to install autogen Fixing headers into /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc for i386-wrs-vxworks target Finding directories and links to directories Searching /usr/include/. All directories (including links to directories): FIND: Parameterformat nicht korrekt mkdir: cannot create directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc/FIND:': No such file or directory Making symbolic directory links Fixing directory /usr/include into /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc Checking header files Cleaning up DONE files. FIND: Parameterformat nicht korrekt Removing unneeded directories: SORT: Unzul"ssige Option autogen: not found You need to install autogen .. To shorten it: The FIND/SORT problem of make/gcc or anything is also the problem that autogen hasn't been build. I found out that there's a problem with depending libraries: .. *** Warning: inter-library dependencies are not known to be supported. *** All declared inter-library dependencies are being dropped. *** The inter-library dependencies that have been dropped here will be *** automatically added whenever a program is linked with this library *** or is declared to -dlopen it. generating symbol list for `libopts.la' dlltool --export-all --exclude-symbols DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12,DllMainCRTStartup@12,DllEntryPoint@12 --output-def .libs/libopts-3-2-2.dll-def autoopts.lo pgusage.lo putshell.lo save.lo stack.lo usage.lo version.lo streqvcmp.lo makeshell.lo ../snprintfv/snprintfv/.libs/libsnprintfvc.a sed -e "1,/EXPORTS/d" -e "s/ @ [0-9]*//" -e "s/ *;.*$//" < .libs/libopts-3-2-2.dll-def > .libs/libopts.exp rm -fr .libs/libopts.lax mkdir .libs/libopts.lax rm -fr .libs/libopts.lax/libsnprintfvc.a mkdir .libs/libopts.lax/libsnprintfvc.a (cd .libs/libopts.lax/libsnprintfvc.a && ar x /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14/autoopts/../snprintfv/snprintfv/.libs/libsnprintfvc.a) echo EXPORTS > .libs/libopts-3-2-2.dll-def _lt_hint=1; cat .libs/libopts.exp | while read symbol; do set dummy $symbol; case $# in 2) echo " $2 @ $_lt_hint ; " >> .libs/libopts-3-2-2.dll-def;; *) echo " $2 @ $_lt_hint $3 ; " >> .libs/libopts-3-2-2.dll-def;; esac; _lt_hint=`expr 1 + $_lt_hint`; done gcc -Wl,--base-file,.libs/libopts-3-2-2.dll-base -Wl,--dll -nostartfiles -Wl,-e,__cygwin_dll_entry@12 -o .libs/libopts-3-2-2.dll autoopts.lo pgusage.lo putshell.lo save.lo stack.lo usage.lo version.lo streqvcmp.lo makeshell.lo FIND: Parameterformat nicht korrekt gcc: FIND:: No such file or directory gcc: Parameterformat: No such file or directory gcc: nicht: No such file or directory gcc: korrekt: No such file or directory make[2]: *** [libopts.la] Error 1 make[2]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14/autoopts' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14' make: *** [all-recursive-am] Error 2 .. This is caused by the following passage out of autogen-4.5.14\snprintfv\config\ltconfig where for the case host == cygwin* the $deplibs is obviously set to "FIND: Parameterformat nicht korrekt": .. cygwin* | mingw*) # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. hardcode_libdir_flag_spec='-L$libdir' allow_undefined_flag=unsupported always_export_symbols=yes extract_expsyms_cmds='test -f $output_objdir/impgen.c || \ sed -e "/^# \/\* impgen\.c starts here \*\//,/^# \/\* impgen.c ends here \*\// { s/^# //; p; }" -e d < $0 > $output_objdir/impgen.c~ test -f $output_objdir/impgen.exe || (cd $output_objdir && \ if test "x$HOST_CC" != "x" ; then $HOST_CC -o impgen impgen.c ; \ else $CC -o impgen impgen.c ; fi)~ $output_objdir/impgen $dir/$soname > $output_objdir/$soname-def' old_archive_from_expsyms_cmds='$DLLTOOL --as=$AS --dllname $soname --def $output_objdir/$soname-def --output-lib $output_objdir/$newlib' # cygwin and mingw dlls have different entry points and sets of symbols # to exclude. # FIXME: what about values for MSVC? dll_entry=__cygwin_dll_entry@12 dll_exclude_symbols=DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12~ case "$host_os" in mingw*) # mingw values dll_entry=_DllMainCRTStartup@12 dll_exclude_symbols=DllMain@12,DllMainCRTStartup@12,DllEntryPoint@12~ ;; esac # mingw and cygwin differ, and it's simplest to just exclude the union # of the two symbol sets. dll_exclude_symbols=DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12,DllMainCRTStartup@12,DllEntryPoint@12 # recent cygwin and mingw systems supply a stub DllMain which the user # can override, but on older systems we have to supply one (in ltdll.c) if test "x$lt_cv_need_dllmain" = "xyes"; then ltdll_obj='$output_objdir/$soname-ltdll.'"$objext " ltdll_cmds='test -f $output_objdir/$soname-ltdll.c || sed -e "/^# \/\* ltdll\.c starts here \*\//,/^# \/\* ltdll.c ends here \*\// { s/^# //; p; }" -e d < $0 > $output_objdir/$soname-ltdll.c~ test -f $output_objdir/$soname-ltdll.$objext || (cd $output_objdir && $CC -c $soname-ltdll.c)~' else ltdll_obj= ltdll_cmds= fi # Extract the symbol export list from an `--export-all' def file, # then regenerate the def file from the symbol export list, so that # the compiled dll only exports the symbol export list. # Be careful not to strip the DATA tag left be newer dlltools. export_symbols_cmds="$ltdll_cmds"' $DLLTOOL --export-all --exclude-symbols '$dll_exclude_symbols' --output-def $output_objdir/$soname-def '$ltdll_obj'$libobjs $convenience~ sed -e "1,/EXPORTS/d" -e "s/ @ [0-9]*//" -e "s/ *;.*$//" < $output_objdir/$soname-def > $export_symbols' # If DATA tags from a recent dlltool are present, honour them! archive_expsym_cmds='echo EXPORTS > $output_objdir/$soname-def~ _lt_hint=1; cat $export_symbols | while read symbol; do set dummy \$symbol; case \$# in 2) echo " \$2 @ \$_lt_hint ; " >> $output_objdir/$soname-def;; *) echo " \$2 @ \$_lt_hint \$3 ; " >> $output_objdir/$soname-def;; esac; _lt_hint=`expr 1 + \$_lt_hint`; done~ '"$ltdll_cmds"' $CC -Wl,--base-file,$output_objdir/$soname-base '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags~ $DLLTOOL --as=$AS --dllname $soname --exclude-symbols '$dll_exclude_symbols' --def $output_objdir/$soname-def --base-file $output_objdir/$soname-base --output-exp $output_objdir/$soname-exp~ $CC -Wl,--base-file,$output_objdir/$soname-base $output_objdir/$soname-exp '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags~ $DLLTOOL --as=$AS --dllname $soname --exclude-symbols '$dll_exclude_symbols' --def $output_objdir/$soname-def --base-file $output_objdir/$soname-base --output-exp $output_objdir/$soname-exp~ $CC $output_objdir/$soname-exp '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags' ;; .. And there's no obvious place where $deplibs is set within that script. MfG / Kind Regards Friedrich Scheurer, R&D -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de -------------------------------- From vxwexplo-errs@csg.lbl.gov Fri Jul 13 04:03:26 2001 From: Vxworks Exploder Date: Fri Jul 13 04:03:29 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 13 04:03:19 PDT 2001 Subject: Re: how to use vxsim? Subject: sockets - increasing number available Subject: resolvGetHostByName Timeout Subject: Re: Is Wind River really this pathetic Subject: Re: how to use vxsim? Subject: CD-ROM file system Subject: Re: How common is SNiFF usage? Subject: gdb (Crosswind) w/ 2.95 & C++ problems Subject: Kermit for VxWorks Subject: Browser Subject: Re: how to use vxsim? Subject: sendto problem:ENOBUFS Subject: Re: Kermit for VxWorks Subject: Re: Kermit for VxWorks Subject: Re: sockets - increasing number available Subject: Serial Communication Subject: Re: sockets - increasing number available Subject: Multicast Receive Problem Subject: Re: How to get patches Subject: Re: CD-ROM file system Subject: What's the latest version of VxWorks? Subject: Does VxWorks use User Mode & Kernel Mode? Subject: Re: What's the latest version of VxWorks? Subject: MBX860 and DIMM Subject: Re: What's the latest version of VxWorks? Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: PPC405 PCMCIA driver with ATA support Subject: VxWorks - Booting Problem Subject: Re: sockets - increasing number available Subject: --split-by-reloc option results in undefined references Subject: ROM based vxworks build with T2 Project: any success ? Subject: Re: --split-by-reloc option results in undefined references Subject: Re-re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Subject: ataDrv arguments ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use vxsim? Date: 12 Jul 2001 05:22:39 -0700 From: jwalicee@altavista.com (Jonathan) Organization: http://groups.google.com/ Message-ID: References: <30c8c76f.0107100839.4b83667e@posting.google.com> <488e459a.0107101941.50fcf7aa@posting.google.com> john_94501@yahoo.com (John) wrote in message news:<488e459a.0107101941.50fcf7aa@posting.google.com>... [snip] > I don't know if anybody has done this kind of thing, or even if an > instruction level simulator for MIPS is readily available. You might > be better off just getting a MIPS eval board and an ICE/JTAG debugger [snip] Isn't SPIM[0] an instruction level simulator for MIPS? > xing@cs.uga.edu (G Xing) wrote in message news:<30c8c76f.0107100839.4b83667e@posting.google.com>... > > I am a beginner of vxsim, and I am tring to simulate mips > > code on sparc machine, is this possible to use vxsim, > > which image should I launch for vxsim: vxworks for sparc > > or vxworks for mips. > > Thanks, > > G Xing I think he, Xing, is trying to run object code, no matter the type as long as VxSim/Tornado can handle it, on a solaris and determine what its effectiveness would be on a MIPS board. Isn't this in the realm of Tornado/VxSim? I think Shade[1] is more along the lines of what he needs but it currently doesn't support MIPS IV as far as I have found. Does anyone know of a version of Shade that does? [0] http://www.cs.wisc.edu/~larus/spim.html [1] http://www.cs.washington.edu/research/compiler/papers.d/shade.html --------------------------- Newsgroups: comp.os.vxworks Subject: sockets - increasing number available Date: Thu, 12 Jul 2001 09:37:11 -0400 From: "Kevin Livingston" Message-ID: Is there a way to increase the available number of available sockets in VxWorks. Right now we are being capped at aprox 30 sockets (I think that is the correct number) and new connections to the "sever" are blocked until one of those sockets is "freed"/"recycled". Causing the number of socket conections to our VxWorks board to be severly limited. Kevin --------------------------- Newsgroups: comp.os.vxworks Subject: resolvGetHostByName Timeout Date: 12 Jul 2001 06:41:51 -0700 From: tyagel@bigfoot.com (tyagel) Organization: http://groups.google.com/ Message-ID: <6ac4e513.0107120541.6758f1bb@posting.google.com> Is there any way to set how long it takes for resolvGetHostByName to timeout when it does not receive any response from a DNS server. Currently it seems to take ~5 minutes. I am using Tornado 2 and running vxWorks on an 8260 processor. Thanks, Thomas --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Thu, 12 Jul 2001 14:15:19 +0100 From: "Andy Ward" Organization: MINX Message-ID: <9ikchr$ou1$1@localhost.localdomain> References: <6qH17.79$5f6.133962@news.uswest.net> Mediocre support and service. "Joe Woodbury" wrote in message news:6qH17.79$5f6.133962@news.uswest.net... > Since we ordered VxWorks/Tornado II, we've had nothing but problems with > Wind River. They filled the order incorrectly and I had to resort to a lot > of yelling and threats to not pay our purchase order to get their attention. > Yet, two days later, they still have yet to do anything but email the > correct BSP. On top of all this, the highest level manager at Wind River > kept blaming us for some of the problems. > > So far, I've concluded that Wind River employees are given an IQ test. If > they fail, they are given a personal relationship test. If they fail that > also, they are hired. > > And heaven forbid they actually do something so obvious like creating a > compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. > > The shame is that VxWorks proper is such an excellent product. > > The questions are: Are our experiences unique? Did our company just happen > to run into the one group of bastards working at Wind River? > > Joe Woodbury > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use vxsim? Date: Thu, 12 Jul 2001 10:43:02 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B4DB776.7F443FF5@digidata.com> References: <30c8c76f.0107100839.4b83667e@posting.google.com> <488e459a.0107101941.50fcf7aa@posting.google.com> We are using the MIPS 7000 and there is no JTAG provision and no ICE available. Kind of a pain getting things started (burn and test...). Not aware of a simulator but if you find one, please drop me a line. Doug Jonathan wrote: > john_94501@yahoo.com (John) wrote in message news:<488e459a.0107101941.50fcf7aa@posting.google.com>... > [snip] > > I don't know if anybody has done this kind of thing, or even if an > > instruction level simulator for MIPS is readily available. You might > > be better off just getting a MIPS eval board and an ICE/JTAG debugger > [snip] > > Isn't SPIM[0] an instruction level simulator for MIPS? > > > xing@cs.uga.edu (G Xing) wrote in message news:<30c8c76f.0107100839.4b83667e@posting.google.com>... > > > I am a beginner of vxsim, and I am tring to simulate mips > > > code on sparc machine, is this possible to use vxsim, > > > which image should I launch for vxsim: vxworks for sparc > > > or vxworks for mips. > > > Thanks, > > > G Xing > > I think he, Xing, is trying to run object code, no matter the type as > long as VxSim/Tornado can handle it, on a solaris and determine what > its effectiveness would be on a MIPS board. Isn't this in the realm > of Tornado/VxSim? I think Shade[1] is more along the lines of what he > needs but it currently doesn't support MIPS IV as far as I have found. > Does anyone know of a version of Shade that does? > > [0] http://www.cs.wisc.edu/~larus/spim.html > [1] http://www.cs.washington.edu/research/compiler/papers.d/shade.html --------------------------- Newsgroups: comp.os.vxworks Subject: CD-ROM file system Date: Thu, 12 Jul 2001 07:48:16 -0700 (PDT) From: Kumaraswamy Mudide Organization: Lawrence Berkeley National Laboratory Message-ID: <9iketq$ar7$1@overload.lbl.gov> Hi All, I have seen some where in the documentation, vxWorks cd-rom file system does not support write feature. Is this true? If this is true, and if I want to write to CD-ROM form vxWorks OS what should I do. Please help us. Thanks, Swamy __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail http://personal.mail.yahoo.com/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How common is SNiFF usage? Date: Thu, 12 Jul 2001 14:55:05 +0100 From: "Dave Korn" Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser refresh) Message-ID: References: <9gijt8$abo$03$1@news.t-online.com> <9goe18$frm$05$1@news.t-online.com> <9gq81t$c2v$1@news.hdpp.de> Matthias Guelck wrote in message <9gq81t$c2v$1@news.hdpp.de>... >Complaining about some debugger crashes (we have got hundreds of crash log files) and problems (default window size for watch, >variables,…), WRS told us, that the whole world is very satisfied, and we are the only people in the world complaining about their >famous products! If I was to send all my debugger crash logs to WRS in one go, I'd be thrown off the internet for mailbombing them! The debugger is sometime just barely useable and sometimes completely disfunctional on a project of the size I'm working on (~300 source files, 640k lines of code, compiled exe size 18Mb without debug info, 98Mb with). DaveK - -- They laughed at Galileo. They laughed at Copernicus. They laughed at Columbus. But remember, they also laughed at Bozo the Clown. --------------------------- Newsgroups: comp.os.vxworks Subject: gdb (Crosswind) w/ 2.95 & C++ problems Date: Thu, 12 Jul 2001 12:27:47 -0400 From: "Jim Pisano" Organization: NRAO Message-ID: <9ikj65$8qq$1@newsfeed.cv.nrao.edu> Greetings, I have successfully built the gcc 2.95.3 cross compiler for power-pc-vxworks hosted on Solaris, but have had a conflict with C++ and the WRS-supplied version of gdb (Crosswind). It appears that straight C code is okay, but for C++ the symbol tables are screwed up. That is, when I set a breakpoint in the source code view, it's never reached & the break point is actually set somewhere else in the binary code of the target application. I know that C++ name mangling was different between 2.72 & 2.95, but I'm not sure that this is the problem as I can issue the command "list Foo::someMemberFunctionOfFoo" and Crosswind correctly shows the function. It's when I set a breakpoint in that source code view & continue the application that the breakpoint is never reached. My application seems to run okay, but I've got a bug somewhere that has cropped up that I'd like to debug but I can't debug it (Catch-22). If anyone is interested in the what hoops I had to go through to get gcc built on Solaris, drop me an email. Related to my problem is this posting (which didn't fix anything for me as the "-gdwarf" option excludes C++ info :-( ): http://groups.google.com/groups?hl=en&safe=off&ic=1&th=a00ed6212efa0e88,2&se ekm=8qfetf%24no7%241%40snipp.uninett.no#p --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Kermit for VxWorks Date: 12 Jul 2001 12:39:22 -0400 From: fdc@watsun.cc.columbia.edu (Frank da Cruz) Organization: Columbia University Message-ID: <9ikjrq$asm@watsun.cc.columbia.edu> From time to time, we get requests for Kermit file transfer software for VxWorks. I'd like to be able satisfy them. Does VxWorks have a C compiler, the basic minimum well-known header files, a C library, and a make program? Does it support the UNIX notion of standard i/o? If so, then maybe it's just a simple matter of getting the G-Kermit sources files onto VxWorks and typing "make": http://www.columbia.edu/kermit/gkermit.html Would anybody care to give it a try and report back? Thanks! Frank da Cruz The Kermit Project Columbia University http://www.columbia.edu/kermit/ --------------------------- Newsgroups: comp.os.vxworks Subject: Browser Date: Thu, 12 Jul 2001 17:49:11 +0100 From: "Andy Ward" Organization: MINX Message-ID: <9ikkiu$vhi$1@localhost.localdomain> Hello, We want to use a browser on VxWorks - ideally HTML 4.0 with applet support - any user experience? regards, Andy --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use vxsim? Date: 12 Jul 2001 10:06:58 -0700 From: jwalicee@altavista.com (Jonathan) Organization: http://groups.google.com/ Message-ID: References: <30c8c76f.0107100839.4b83667e@posting.google.com> <488e459a.0107101941.50fcf7aa@posting.google.com> <3B4DB776.7F443FF5@digidata.com> Doug Dotson wrote in message news:<3B4DB776.7F443FF5@digidata.com>... > We are using the MIPS 7000 and there is no JTAG > provision and no ICE available. Kind of a pain getting > things started (burn and test...). Not aware of a simulator > but if you find one, please drop me a line. [snip] Have you considered SimOS? I just learned about that today therefore I do not know if it is what you would be looking for. http://simos.stanford.edu/ Jonathan --------------------------- Newsgroups: comp.os.vxworks Subject: sendto problem:ENOBUFS Date: Thu, 12 Jul 2001 17:14:32 GMT From: Tony Rodayer Organization: Guest of ProXad - France Message-ID: <3B4DDB7D.3050503@free.fr> Hy, Is the sendto function works as under Unix(sendto block)? Someone says me that the sendto in vxWorks doesn't block as under Unix but he is not sure. I have a problem when i try to send data at High Speed. I send data(100000 packets of 1k) to 127.0.0.1 and i have some ENOBUFS(2000) result of the sendto. When a sendto problem(ENOBUF) appears, my test program does a taskDelay(1) and then re-send the data but i don't think that ENOBUFS warnings is good. I try to increase the mbuf but there is no changes. Thanks. --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Thu, 12 Jul 2001 13:30:03 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B4DDE9B.C12DEF3@digidata.com> References: <9ikjrq$asm@watsun.cc.columbia.edu> It shouldn't be too much of a problem to port it. vxWorks is a bit different in that you don't build executables in the UNIX sense. You build .o files and they are linked when loaded. There is no main(). But if you changed main() to kermit() you should be OK. Argument typed on the command line are mapped into the function arguments. You will have to change some header files and the makefile won't work as it is. DOug Frank da Cruz wrote: > From time to time, we get requests for Kermit file transfer software > for VxWorks. I'd like to be able satisfy them. > > Does VxWorks have a C compiler, the basic minimum well-known header > files, a C library, and a make program? Does it support the UNIX > notion of standard i/o? If so, then maybe it's just a simple matter > of getting the G-Kermit sources files onto VxWorks and typing "make": > > http://www.columbia.edu/kermit/gkermit.html > > Would anybody care to give it a try and report back? Thanks! > > Frank da Cruz > The Kermit Project > Columbia University > http://www.columbia.edu/kermit/ --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: 12 Jul 2001 18:00:27 GMT From: fdc@watsun.cc.columbia.edu (Frank da Cruz) Organization: Columbia University Message-ID: <9ikojr$r9c$1@newsmaster.cc.columbia.edu> References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> In article <3B4DDE9B.C12DEF3@digidata.com>, Doug Dotson wrote: : It shouldn't be too much of a problem to port it. vxWorks is : a bit different in that you don't build executables in the UNIX : sense. You build .o files and they are linked when loaded. There : is no main(). But if you changed main() to kermit() you should be : OK. Argument typed on the command line are mapped into the : function arguments. You will have to change some header files : and the makefile won't work as it is. : I'd be glad to give it a shot if I had VxWorks developer tools to build the kermit function and VxWorks itself to test it, but I have neither. Would someone who does have them like to try adapting G-Kermit to VxWorks? There's clearly a demand, at least among those who don't mind devoting somewhere between 30 and 80K, depending on the architecture, to a file-transfer function, and it doesn't sound like too much work. - - Frank --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sockets - increasing number available Date: Thu, 12 Jul 2001 13:09:32 -0500 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <3B4DE7DC.EE05EA86@aps.anl.gov> References: Kevin Livingston wrote: > > Is there a way to increase the available number of available sockets in > VxWorks. Right now we are being capped at aprox 30 sockets (I think that is > the correct number) and new connections to the "sever" are blocked until one > of those sockets is "freed"/"recycled". Causing the number of socket > conections to our VxWorks board to be severly limited. Yes, increase the #define NUM_FILES in your configAll.h file (or there will be an equivalent if you build using the Tornado GUI, I don't know the details of that though). It defaults to 50 IIRC, and can go up to about 250 if you're using select, more if not. - - Andrew - -- The world is such a cheerful place when viewed from upside-down It makes a rise of every fall, a smile of every frown --------------------------- Newsgroups: comp.os.vxworks Subject: Serial Communication Date: 12 Jul 2001 11:31:04 -0700 From: fiddooo@yahoo.com (fiddooo) Organization: http://groups.google.com/ Message-ID: <11061abf.0107121031.493333b6@posting.google.com> I am using the serial port /tyCo/1 of VxWorks box. and i use write() call to write to the port. However I am puzzled to see that when i run my code from a shell, sometimes the printf() statements are written to the serial port instead of shell console. At times the printfs are missing totally not visible even on the serial port. I guess this is a problem with redirecting the io. but i am not sure about it and am looking for a solution. Please help. regards, fiddooo . --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sockets - increasing number available Date: 12 Jul 2001 11:52:05 -0700 From: tlksk@hotmail.com (Kumar) Organization: http://groups.google.com/ Message-ID: References: Hi Kevin You can do the above by configuring the NUM_FILES parameter in the "core TCP/IP components" component folder in the launch window. Check to see if that is 30 and increase that to the maximum you need. Kumar "Kevin Livingston" wrote in message news:... > Is there a way to increase the available number of available sockets in > VxWorks. Right now we are being capped at aprox 30 sockets (I think that is > the correct number) and new connections to the "sever" are blocked until one > of those sockets is "freed"/"recycled". Causing the number of socket > conections to our VxWorks board to be severly limited. > > Kevin --------------------------- Newsgroups: comp.os.vxworks Subject: Multicast Receive Problem Date: Thu, 12 Jul 2001 16:12:02 -0400 From: "VxWorks" Message-ID: Hi, We have a MVME-5101 (PPC750) with and i82559 chipset (er0) that will not receive multicast IP. We are confident that we have multicast receive working properly because it works on a MVME-2604 board and pcPentium with i82559. The problem is that when our server broadcasts to multicast address 225.0.0.30 (which the 5101 is a member of) it does not receive the data. It keeps sitting at recv(). We have setup membership and are using the VxWorks sample code from the VxWorks Network Programmer's Guide. I am pretty sure that er0 is an END interface as well? Why does it hang on recv() on the 5101 and receive the data successfully on other PPC and Intel CPU's? Regards, Todd --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get patches Date: Thu, 12 Jul 2001 20:52:52 GMT From: "Peter @ Home" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: Hi Joe, Here is the exact procedure to access the WindSurf patches: www.windriver.com - --> Click on "support" at the top of the page - --> Clock on "Windsurf Login" at the left of the next page - --> enter your username / password (if you do not have a login, you may create one at this time you will require your Wind River license #) - --> Click on the "download" tab - --> Click on the "patches" link You will have ended up here: http://www.windriver.com/csdocs/kplocator/patchList.shtml Here you will find all of the currently publicly available SPRs. (Software Problem Reports - or patches). If you are using Tornado 2.0, you can upgrade to Tornado 2.0.2 by downloading Tornado 2 Cumulative Patch 4 (T2CP4). This also contains all of the previous patches. Individual patches are removed from WindSurf if: - they are rolled into the cumulative patch or -- they are found to be defective in any way. If you ever require an individual patch (eg. if your project is in its late stages and it would be risky to upgrade an entire cumulative patch level), you can contact support@windriver.com and they will work hard to get you the individual patch that you require. I hope that this helps. Regards, Peter. "Joe Woodbury" wrote in message news:c8c13723.0107051237.4700018@posting.google.com... > I see suggestions on this board to check Wind River Systems Windsurf > site for SPR#xxxxx, which is fixed by Cumulative Patch x. > > I am a legitimate customer, have a customer ID and access to WindSurf > but I am unable to find either of these things. What am I doing wrong? > > Joe Woodbury --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CD-ROM file system Date: Thu, 12 Jul 2001 23:17:37 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <994973132.997983@newsmaster-04.atnet.at> References: <9iketq$ar7$1@overload.lbl.gov> Hi, To write to a CD you need a CD recording software. This is significantly more than just a driver that pops out some bits through some IO address. I don't know of any such for VxWorks. Maybe you find something in the Linux/FreeBSD corner. However, if you not only want to write one image, but data files as they come up, you'll need a packet writing software. Maybe you should consider some more 'traditional' storage device... regards werner --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: What's the latest version of VxWorks? Date: Thu, 12 Jul 2001 17:32:16 -0500 From: "Lee" Message-ID: <9il8hi$j1ktd$1@ID-92003.news.dfncis.de> Hi, Got two general questions here. Q1: what is the latest version of VxWorks. Q2: what is the latest version of Tornado? Q3: which versions of VxWorks are included in Tornado 2.0 and Tornado 2.1? Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Does VxWorks use User Mode & Kernel Mode? Date: Thu, 12 Jul 2001 17:57:21 -0500 From: "Lee" Message-ID: <9ila0j$k005n$1@ID-92003.news.dfncis.de> Hi, Just wondering if VxWorks uses User Mode and Kernel Mode to separate user code from hardware-related code, or does it have only one mode? If it has only one mode, then high level application code can access hardware IO space directly. Are there any potential problems? Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: What's the latest version of VxWorks? Date: Fri, 13 Jul 2001 01:14:02 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9ilb0l$tc5$03$1@news.t-online.com> References: <9il8hi$j1ktd$1@ID-92003.news.dfncis.de> Hello Lee, the latest stuff is vxWorks AE 1.0 coming with Tornado AE 1.0 If you talk about Tornado 2.x: vxWorks 5.4 coming with Tornado 2.1 (not released yet for all processor architectures) Ciao Martin "Lee" schrieb im Newsbeitrag news:9il8hi$j1ktd$1@ID-92003.news.dfncis.de... > Hi, > > Got two general questions here. > > Q1: what is the latest version of VxWorks. > Q2: what is the latest version of Tornado? > Q3: which versions of VxWorks are included in Tornado 2.0 and Tornado 2.1? > > Thanks. > > > Lee > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: MBX860 and DIMM Date: 12 Jul 2001 16:38:55 -0700 From: jbundy@megsinet.net (John Bundy) Organization: http://groups.google.com/ Message-ID: Hi, Has anybody had success adding DRAM DIMM memory to the MBX860? I would be interested in what DIMM part number works and general bsp changes. I am using pSOS, but it looks like a lot more vxworks people have used the MBX860. John Bundy --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: What's the latest version of VxWorks? Date: Fri, 13 Jul 2001 01:26:37 GMT From: Andray Kaganovsky Organization: @Home Message-ID: References: <9il8hi$j1ktd$1@ID-92003.news.dfncis.de> <9ilb0l$tc5$03$1@news.t-online.com> Isn't the latest version of VxWorks AE is 1.1 (Tornado Tools 3.1) ? Andray "Martin Raabe" wrote in news:9ilb0l$tc5$03$1 @news.t-online.com: > Hello Lee, > the latest stuff is vxWorks AE 1.0 coming with Tornado AE 1.0 > > If you talk about Tornado 2.x: > vxWorks 5.4 coming with Tornado 2.1 (not released yet for all processor > architectures) > > Ciao > > Martin > > "Lee" schrieb im Newsbeitrag > news:9il8hi$j1ktd$1@ID-92003.news.dfncis.de... >> Hi, >> >> Got two general questions here. >> >> Q1: what is the latest version of VxWorks. >> Q2: what is the latest version of Tornado? >> Q3: which versions of VxWorks are included in Tornado 2.0 and Tornado 2.1? >> >> Thanks. >> >> >> Lee >> >> >> >> > > > --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Fri, 13 Jul 2001 01:29:51 GMT From: Tim Sohacki Organization: Road Runner - NC Message-ID: <3B4E4FF5.66545477@nc.rr.com> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> Lee wrote: > > Hi, > > Just wondering if VxWorks uses User Mode and Kernel Mode to separate user > code from hardware-related code, or does it have only one mode? If it has > only one mode, then high level application code can access hardware IO space > directly. Are there any potential problems? > > Thanks. > > Lee All tasks running under VxWorks 5.4 run in supervisor (Kernel) mode. I am sure this is true for all processor architectures, as this was a tenet of VxWorks for a long time. VxWorks AE (basically, the version after VxWorks 5.4) may utilize User Mode to accomplish the memory protection that it offers. And yes, out of the box VxWorks 5.4 does allow application code to access hardware IO space directly, or any memory address that is configured R/W. There are lots of potential problems when any task can access any part of memory ... trashing the heap, crashing the system, confusing a peripheral device, etc. The WRS rationale I've heard is that always running in Supervisor or Kernel mode allows for more deterministic and much faster context switches than using User Mode. Cheers, Tim. --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Fri, 13 Jul 2001 02:00:59 GMT From: joegwinn@mediaone.net (Joe Gwinn) Organization: Gwinn Instruments Message-ID: References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> In article <9ila0j$k005n$1@ID-92003.news.dfncis.de>, "Lee" wrote: > Just wondering if VxWorks uses User Mode and Kernel Mode to separate user > code from hardware-related code, or does it have only one mode? There is one mode. In fact, user tasks are directly linked to the kernel, and call kernel services using ordinary subroutine calls (not trap interrupts). Most other RTOSes at least separate user code from kernel code to the extent of requiring a trap to invoke kernel services, but still everything is in one big address space, and everything typically runs in kernel mode. Some RTOSes allow one to run threads in user mode, but don't require it. > ... If it has > only one mode, then high level application code can access hardware IO space > directly. True. > Are there any potential problems? Basically, any task (~thread) can look at anything, stomp anything, in other tasks or the kernel or I/O space or stackspace or whatever. That said, this kind of a one-shared-address-space architecture has been traditional in RTOSes for many decades, and it does work, with billions of dollars of fielded systems still in everyday use. There are some RTOSes that isolate the tasks (processes) from one another and from the kernel and perhaps the I/O drivers as well. The advantage is that it's easier to figure out why the system went face-down, and to recover from bugs. The disadvantage is lower performance and increased overall complexity. If you want to hear all about the problems of one-space RTOSes, look into QNX, which is a stringent member of the isolate-everything school. www.qnx.com If you want to hear why isolate-everything is a terrible idea, look into VxWorks. Which model is best is something of a religious issue; I like both, each in their own way, having no vested interest in either. Joe Gwinn --------------------------- Newsgroups: comp.os.vxworks Subject: PPC405 PCMCIA driver with ATA support Date: Thu, 12 Jul 2001 15:30:33 +0200 From: Adam Abrams Organization: Bezeq International Ltd. Message-ID: <01554D7E32E7D1118A8800805F31D8A3012C9B0D@tiger.seabridge.co.il> Sender: Adam Abrams It seems to me that VxWorks under PPC has no built-in support for PCMCIA, as the driver doesn't compile and link - and seems very Intel/PCI-ISA dependant. 1. Is that at all true? 2. Does anyone have experience with porting PCMCIA support to PPC, with ATA support (on dosFS) 3. Even better, anyone with DMA experience? Thanks, Adam. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks - Booting Problem Date: Fri, 13 Jul 2001 10:02:34 +0530 From: "rajesh biswal" Organization: Lawrence Berkeley National Laboratory Message-ID: <9im04q$4v5$1@overload.lbl.gov> This is a multi-part message in MIME format. - --------------InterScan_NT_MIME_Boundary Content-Type: multipart/alternative; boundary="----=_NextPart_000_0007_01C10B82.F049B1A0" - ------=_NextPart_000_0007_01C10B82.F049B1A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello, We are trying to boot the Pentium Target through ethernet. The specification of Driver is: AcerLAN ALN-325 10/100 Base-TX Fast Ethernet Device Type: Network Adapters. Manufacturer: Acer Netxus Inc. H/w Version: 016 Interrupt Request: 10 Memory Range : 09100000-091000FF Input/Output Range: 7000-70FF What are the settings required in VxWorks Configuration. We are using T2. Thanx in advance, Regards, Rajesh Biswal. **************************************************************************** ************ Embedded & Internet Access Division Wipro Technologies Address : 30, Mission Road 1st Main, S.R Nagar Bangalore, 560027 Tel : 91-80-2275590/2241730 Extn 1145 Fax : 91-80-2241769 E-mail : rajesh.biswal@wipro.com Web Page : www.wipro.com The World's First SEI CMM Level 5 Software Services company **************************************************************************** ************ Silence and thinking never hurt anyone. - ------=_NextPart_000_0007_01C10B82.F049B1A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello,
We are = trying to=20 boot the Pentium Target through ethernet.
The = specification of=20 Driver is:
 
AcerLAN ALN-325=20 10/100 Base-TX Fast Ethernet
Device = Type: Network=20 Adapters.
Manufacturer: Acer=20 Netxus Inc.
H/w = Version:=20 016
 
Interrupt=20 Request:     10
Memory = Range :       09100000-091000FF
Input/Output=20 Range:  7000-70FF
 
What = are the=20 settings required in VxWorks Configuration.
We are = using=20 T2.
 
Thanx = in=20 advance,
Regards,
Rajesh Biswal.
****************************************************************= ************************=20
Embedded & Internet Access Division
Wipro=20 Technologies
Address    : 30, Mission = Road
          &nbs= p;      =20 1st Main, S.R Nagar
          &nbs= p;      =20 Bangalore, 560027
Tel          &= nbsp; :=20 91-80-2275590/2241730 Extn 1145
Fax           = :=20 91-80-2241769
E-mail       : rajesh.biswal@wipro.com
Web = Page : www.wipro.com
 
The World's First SEI CMM Level 5 = Software Services=20 company
****************************************************************= ************************=20
 Silence and thinking never hurt = anyone.
 
- ------=_NextPart_000_0007_01C10B82.F049B1A0-- - --------------InterScan_NT_MIME_Boundary Content-Type: text/plain; name="Wipro_Disclaimer.txt" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="Wipro_Disclaimer.txt" The Information contained and transmitted by this E-MAIL is proprietary to Wipro Limited and is intended for use only by the individual or entity to which it is addressed, and may contain information that is privileged, confidential or exempt from disclosure under applicable law. If this is a forwarded message, the content of this E-MAIL may not have been sent with the authority of the Company. If you are not the intended recipient, an agent of the intended recipient or a person responsible for delivering the information to the named recipient, you are notified that any use, distribution, transmission, printing, copying or dissemination of this information in any way or in any manner is strictly prohibited. If you have received this communication in error, please delete this mail & notify us immediately at mailadmin@wipro.com - --------------InterScan_NT_MIME_Boundary-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sockets - increasing number available Date: Thu, 12 Jul 2001 22:20:18 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B4E8512.79C23A8F@covad.net> References: Reply-To: drdiags@covad.net Kevin, You can increase the number of sockets as others have indicated but you should also be aware that you may have to adjust your mbuf configuration. I think Johan has added the tech tip provided by WR Networks support folks to the FAQ, if not, you should be able to search this newsgroup's archive to find the post showing how to make the adjustments if you are not aware of how to do it. HTH. Kevin Livingston wrote: > > Is there a way to increase the available number of available sockets in > VxWorks. Right now we are being capped at aprox 30 sockets (I think that is > the correct number) and new connections to the "sever" are blocked until one > of those sockets is "freed"/"recycled". Causing the number of socket > conections to our VxWorks board to be severly limited. > > Kevin - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: --split-by-reloc option results in undefined references Date: Fri, 13 Jul 2001 12:02:55 +0530 From: jp Organization: Lucent Technologies Message-ID: <3B4E9617.E5CB7136@lucent.com> Hi Vxworkers, I am running vxsim on Windows NT. My application is huge one and I have to use - --split-by-reloc option flag in order to recover from reloc overflow error. But including this option results in some strange undefined references for some routines used by crt1.o (main,__cexit, winMainCrtStartup etc.). Please tell me what all I am doing wrong. Thanks in advance. Regards gaurav. --------------------------- Newsgroups: comp.os.vxworks Subject: ROM based vxworks build with T2 Project: any success ? Date: Fri, 13 Jul 2001 07:47:15 GMT From: Diego Serafin Organization: http://www.newsranger.com Message-ID: <7Ix37.17948$Kf3.233634@www.newsranger.com> Hi folks ! I'm trying to build a ROM based version of VxWorks (vxworkx_rom and vxworks_romCompress images) using the Tornado 2 Project tool, but found that image I produced did not boot properly. What happens: 1) romInit works fine. 2) romStart works fine but, at the end, the jump to absEntry that should fire usrEntry() crashes system. I saw that vxworks module, stored at (rom address of) binArrayStart does not start with code, but with .ELF header so that when romStart jumps to RAM_DST_ADRS it fetches rubbish (...".ELF"). The real usrEntry code sits at 0x100 bytes from RAM_DST_ADRS. I found that elfToBin tool, during image make-ing, outputs a module that still starts with ".ELF" ... is this correct ? If I modify romStart's copyLong() call to start copying flash 0x100 bytes ahead the systems comes up correctly. Where am I wrong ? Thank You, Ciao. Diego. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: --split-by-reloc option results in undefined references Date: Fri, 13 Jul 2001 09:59:26 +0200 From: "Frank Wolf" Organization: T-Online Message-ID: References: <3B4E9617.E5CB7136@lucent.com> Hi there! Please take a look at your linker script... most probably you've got one .text segment declared. But due to 'split-by-reloc' you need more than just one .text segment to link your application. Frank "jp" schrieb im Newsbeitrag news:3B4E9617.E5CB7136@lucent.com... > Hi Vxworkers, > > I am running vxsim on Windows NT. My application is huge one and I have to use > --split-by-reloc option flag in order to recover from reloc overflow error. > > But including this option results in some strange undefined references for some > routines used by crt1.o (main,__cexit, winMainCrtStartup etc.). > > Please tell me what all I am doing wrong. > > Thanks in advance. > > Regards > gaurav. --------------------------- Newsgroups: comp.os.vxworks Subject: Re-re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Date: 13 Jul 2001 08:32:02 UT From: fscheurer@manz-acs.de Organization: Lawrence Berkeley National Laboratory Message-ID: <9ime6p$a80$1@overload.lbl.gov> Hi Dave, Thanx for your answer. Based on your Homepage and with the help of J. Pisano (he did it for Linux host or Solaris) I advanced another step. I apologize the long posting, but now I ran into the detail problems. So here my two answers and the state where I am now. I am grateful for every hint, about the options I forgot or what Headers to copy where. I suppose no one tried that host-target X-ing before, otherwise the person should have been run into the same problems. Everyone is happy if it works for his own host. Especially the FIND/SORT error messages I cannot track anymore. Where's their source? Details below: - -------- Original Message -------- .. > > Newsgroups: comp.os.vxworks > Subject: Re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 > Date: Wed, 11 Jul 2001 16:51:00 +0100 > From: "Dave Korn" > Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser > refresh) > Message-ID: <8P%27.104$os1.458451@newsr2.u-net.net> > References: <9i145t$65d$1@overload.lbl.gov> > > fscheurer@manz-acs.de wrote in message <9i145t$65d$1@overload.lbl.gov>... > >Hi VxWorks-Gurus, > > > >I ran into the problem, that the integrated target-headers > >seem to be not confor with GNU expectations. > > Did you use the --with-headers= flag when configuring to supply the > target > headers, or did you copy them into $prefix/$target/sys-include, or into > $prefix/$target/include ? Both no in that first stage! Is it to directly include $WIND_BASE/target/h there with all it's sub-directories? > > The first two of those methods will cause the Gcc build process to invoke > the 'fixincs' script, which runs through all the supplied headers, checking > for and fixing things that need to be made GNU-compatible. > > Although it is recommended in many pieces of documentation that you > install target headers into $prefix/$target/include when building a cross > compiler, doing so prevents the fixincs from being run: fixincs will only > ever process files from the sys-include directory. > > The error messages you posted look a lot like fixincs hasn't run. Can > you > confirm or deny this? > In that previously mentionned first stage probably not. In Jim's script this is done: .. #cd gcc-2.95.3/ cd ./gcc-2.95.3/ # first, fix includes with patches #cd gcc/fixinc cd ./gcc/fixinc ./genfixes cd ../.. mkdir build #cd build cd ./build ./configure --prefix=$WIND_BASE/host/$WIND_HOST_TYPE --target=$VX_TARGET --with-gnu-as --with-gnu-ld --with-headers=$WIND_BASE/target/h --enable-languages=c,c++ -v 2>&1 | tee config.log ## FS 2001-07-10 make clean # make LANGUAGES="c c++" all install 2>&1 | tee build1.log .. > >MY QUESTION: > >Where is the most central place in > >- the sources, > >- the configure-script > >or > >- the Makefile > > > >to "repair" it > >AND to get the right types to be compatible to objects build with > >older gcc-version? > > I suspect that the fixes already exist in the fixincs script, and merely > haven't been applied to your target headers in the course of building the > cross-compiler. However, the VxWorks targets for Gcc haven't been well > maintained of late, and it is possible that the WRS headers have changed in > such a way that the fixincs test-pattern no longer matches the WRS header > file, causing a fix to be missed; or that new code in the headers mean that > new fixes need to be added. Give it another try, being sure to use one of > the first two methods I mentioned above to supply the headers, and tell us > if it works any better. > Jim's script uses autogen which I had to build in advance. I let it all run through but encounter some other problems using gcc. First I didn't realise that autogen wasn't build right. I saw during build of cross-compiler in build.log: .. for name in _muldi3 _divdi3 _moddi3 _udivdi3 _umoddi3 _negdi2 _lshrdi3 _ashldi3 _ashrdi3 _ffsdi2 _udiv_w_sdiv _udivmoddi4 _cmpdi2 _ucmpdi2 _floatdidf _floatdisf _fixunsdfsi _fixunssfsi _fixunsdfdi _fixdfdi _fixunssfdi _fixsfdi _fixxfdi _fixunsxfdi _floatdixf _fixunsxfsi _fixtfdi _fixunstfdi _floatditf __gcc_bcmp _varargs __dummy _eprintf _bb _shtab _clear_cache _trampoline __main _exit _ctors _pure; \ do \ echo ${name}; \ /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/xgcc -B/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/ -B/Tornado2/host/winnt/i386-wrs-vxworks/bin/ -I/Tornado2/host/winnt/i386-wrs-vxworks/include -O2 -DCROSS_COMPILE -DIN_GCC -g -O2 -I./include -g1 -DHAVE_GTHR_DEFAULT -DIN_LIBGCC2 -D__GCC_FLOAT_NOT_NEEDED -I. -I../../gcc -I../../gcc/config -I../../gcc/../include -c -DL${name} \ ../../gcc/libgcc2.c -o ${name}.o; \ if [ $? -eq 0 ] ; then true; else exit 1; fi; \ i386-wrs-vxworks-ar rc tmplibgcc2.a ${name}.o; \ rm -f ${name}.o; \ done _muldi3 ./../gcc/libgcc2.c:41: stdlib.h: No such file or directory ./../gcc/libgcc2.c:42: unistd.h: No such file or directory make[1]: *** [libgcc2.a] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc' make: *** [all-gcc] Error 2 .. The headers are really not there! Some steps before I found: ... make[2]: Entering directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc' autogen: not found You need to install autogen Fixing headers into /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc for i386-wrs-vxworks target Finding directories and links to directories Searching /usr/include/. All directories (including links to directories): FIND: Parameterformat nicht korrekt mkdir: cannot create directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc/FIND:': No such file or directory Making symbolic directory links Fixing directory /usr/include into /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/gcc-2.95.3/build/gcc/fixinc Checking header files Cleaning up DONE files. FIND: Parameterformat nicht korrekt Removing unneeded directories: SORT: Unzul"ssige Option autogen: not found You need to install autogen .. To shorten it: The FIND/SORT problem of make/gcc or anything is also the problem that autogen hasn't been build. I found out that there's a problem with depending libraries: .. *** Warning: inter-library dependencies are not known to be supported. *** All declared inter-library dependencies are being dropped. *** The inter-library dependencies that have been dropped here will be *** automatically added whenever a program is linked with this library *** or is declared to -dlopen it. generating symbol list for `libopts.la' dlltool --export-all --exclude-symbols DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12,DllMainCRTStartup@12,DllEntryPoint@12 --output-def .libs/libopts-3-2-2.dll-def autoopts.lo pgusage.lo putshell.lo save.lo stack.lo usage.lo version.lo streqvcmp.lo makeshell.lo ../snprintfv/snprintfv/.libs/libsnprintfvc.a sed -e "1,/EXPORTS/d" -e "s/ @ [0-9]*//" -e "s/ *;.*$//" < .libs/libopts-3-2-2.dll-def > .libs/libopts.exp rm -fr .libs/libopts.lax mkdir .libs/libopts.lax rm -fr .libs/libopts.lax/libsnprintfvc.a mkdir .libs/libopts.lax/libsnprintfvc.a (cd .libs/libopts.lax/libsnprintfvc.a && ar x /cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14/autoopts/../snprintfv/snprintfv/.libs/libsnprintfvc.a) echo EXPORTS > .libs/libopts-3-2-2.dll-def _lt_hint=1; cat .libs/libopts.exp | while read symbol; do set dummy $symbol; case $# in 2) echo " $2 @ $_lt_hint ; " >> .libs/libopts-3-2-2.dll-def;; *) echo " $2 @ $_lt_hint $3 ; " >> .libs/libopts-3-2-2.dll-def;; esac; _lt_hint=`expr 1 + $_lt_hint`; done gcc -Wl,--base-file,.libs/libopts-3-2-2.dll-base -Wl,--dll -nostartfiles -Wl,-e,__cygwin_dll_entry@12 -o .libs/libopts-3-2-2.dll autoopts.lo pgusage.lo putshell.lo save.lo stack.lo usage.lo version.lo streqvcmp.lo makeshell.lo FIND: Parameterformat nicht korrekt gcc: FIND:: No such file or directory gcc: Parameterformat: No such file or directory gcc: nicht: No such file or directory gcc: korrekt: No such file or directory make[2]: *** [libopts.la] Error 1 make[2]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14/autoopts' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/cygdrive/c/GNU_CrossGCC_Cygwin/DaveK_JimPisano/build/autogen-4.5.14' make: *** [all-recursive-am] Error 2 .. This is caused by the following passage out of autogen-4.5.14\snprintfv\config\ltconfig where for the case host == cygwin* the $deplibs is obviously set to "FIND: Parameterformat nicht korrekt": .. cygwin* | mingw*) # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. hardcode_libdir_flag_spec='-L$libdir' allow_undefined_flag=unsupported always_export_symbols=yes extract_expsyms_cmds='test -f $output_objdir/impgen.c || \ sed -e "/^# \/\* impgen\.c starts here \*\//,/^# \/\* impgen.c ends here \*\// { s/^# //; p; }" -e d < $0 > $output_objdir/impgen.c~ test -f $output_objdir/impgen.exe || (cd $output_objdir && \ if test "x$HOST_CC" != "x" ; then $HOST_CC -o impgen impgen.c ; \ else $CC -o impgen impgen.c ; fi)~ $output_objdir/impgen $dir/$soname > $output_objdir/$soname-def' old_archive_from_expsyms_cmds='$DLLTOOL --as=$AS --dllname $soname --def $output_objdir/$soname-def --output-lib $output_objdir/$newlib' # cygwin and mingw dlls have different entry points and sets of symbols # to exclude. # FIXME: what about values for MSVC? dll_entry=__cygwin_dll_entry@12 dll_exclude_symbols=DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12~ case "$host_os" in mingw*) # mingw values dll_entry=_DllMainCRTStartup@12 dll_exclude_symbols=DllMain@12,DllMainCRTStartup@12,DllEntryPoint@12~ ;; esac # mingw and cygwin differ, and it's simplest to just exclude the union # of the two symbol sets. dll_exclude_symbols=DllMain@12,_cygwin_dll_entry@12,_cygwin_noncygwin_dll_entry@12,DllMainCRTStartup@12,DllEntryPoint@12 # recent cygwin and mingw systems supply a stub DllMain which the user # can override, but on older systems we have to supply one (in ltdll.c) if test "x$lt_cv_need_dllmain" = "xyes"; then ltdll_obj='$output_objdir/$soname-ltdll.'"$objext " ltdll_cmds='test -f $output_objdir/$soname-ltdll.c || sed -e "/^# \/\* ltdll\.c starts here \*\//,/^# \/\* ltdll.c ends here \*\// { s/^# //; p; }" -e d < $0 > $output_objdir/$soname-ltdll.c~ test -f $output_objdir/$soname-ltdll.$objext || (cd $output_objdir && $CC -c $soname-ltdll.c)~' else ltdll_obj= ltdll_cmds= fi # Extract the symbol export list from an `--export-all' def file, # then regenerate the def file from the symbol export list, so that # the compiled dll only exports the symbol export list. # Be careful not to strip the DATA tag left be newer dlltools. export_symbols_cmds="$ltdll_cmds"' $DLLTOOL --export-all --exclude-symbols '$dll_exclude_symbols' --output-def $output_objdir/$soname-def '$ltdll_obj'$libobjs $convenience~ sed -e "1,/EXPORTS/d" -e "s/ @ [0-9]*//" -e "s/ *;.*$//" < $output_objdir/$soname-def > $export_symbols' # If DATA tags from a recent dlltool are present, honour them! archive_expsym_cmds='echo EXPORTS > $output_objdir/$soname-def~ _lt_hint=1; cat $export_symbols | while read symbol; do set dummy \$symbol; case \$# in 2) echo " \$2 @ \$_lt_hint ; " >> $output_objdir/$soname-def;; *) echo " \$2 @ \$_lt_hint \$3 ; " >> $output_objdir/$soname-def;; esac; _lt_hint=`expr 1 + \$_lt_hint`; done~ '"$ltdll_cmds"' $CC -Wl,--base-file,$output_objdir/$soname-base '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags~ $DLLTOOL --as=$AS --dllname $soname --exclude-symbols '$dll_exclude_symbols' --def $output_objdir/$soname-def --base-file $output_objdir/$soname-base --output-exp $output_objdir/$soname-exp~ $CC -Wl,--base-file,$output_objdir/$soname-base $output_objdir/$soname-exp '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags~ $DLLTOOL --as=$AS --dllname $soname --exclude-symbols '$dll_exclude_symbols' --def $output_objdir/$soname-def --base-file $output_objdir/$soname-base --output-exp $output_objdir/$soname-exp~ $CC $output_objdir/$soname-exp '$lt_cv_cc_dll_switch' -Wl,-e,'$dll_entry' -o $lib '$ltdll_obj'$libobjs $deplibs $compiler_flags' ;; .. And there's no obvious place where $deplibs is set within that script. MfG / Kind Regards Friedrich Scheurer, R&D - -------------------------------- acs GmbH Steigaeckerstrasse 13 72768 Reutlingen Telefon (07121) 9019-23 Telefax (07121) 9019-98 Internet www.manz-acs.de e-mail: fscheurer@manz-acs.de - -------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: ataDrv arguments Date: 13 Jul 2001 03:36:11 -0700 From: junkmailnotread@yahoo.com (lost distance) Organization: http://groups.google.com/ Message-ID: <22ba361d.0107130236.76378b0b@posting.google.com> The manual page for ataDrv (ATA/IDE disk device driver) states that it takes a configuration flag as a parameter. Good - I want to use this to configure PIO mode 4. However the formal parameter - "configType" - is defined as a BOOL. Bad - how can I possibly pass in the various configuration flag values to a BOOL? Any enlightenment will be gratefully accepted. Thanks --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jul 14 04:03:25 2001 From: Vxworks Exploder Date: Sat Jul 14 04:03:28 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 14 04:03:08 PDT 2001 Subject: tExcTask Subject: Re: Is Wind River really this pathetic Subject: Re: ataDrv arguments Subject: Re: PPC405 PCMCIA driver with ATA support Subject: Re: Need Help - PowerPC MPC8xx SDRAM Setup Subject: Re: ISR callable routines Subject: Re: MBX860 and DIMM Subject: state machine implement Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: MBX860 and DIMM Subject: What is the main differences between i386 and i960? Subject: "File System" on VxWorks Subject: Re: state machine implement Subject: unresolved symbol pciHeaderShow when downloading Subject: Rebuilding C++ String (STL) library in Tornado? Subject: Re: Is Wind River really this pathetic Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: What's the latest version of VxWorks? Subject: Re: ROM based vxworks build with T2 Project: any success ? Subject: Re: ROM based vxworks build with T2 Project: any success ? Subject: Re: Re-re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Subject: Re: What is the main differences between i386 and i960? Subject: Re: PPC405 PCMCIA driver with ATA support Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: Re : Assembler with VxWorks Subject: Re: Kermit for VxWorks Subject: Re: unresolved symbol pciHeaderShow when downloading Subject: Re: Inconsistency while mibcomp -man -trap.c ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: tExcTask Date: Fri, 13 Jul 2001 12:19:59 +0100 From: Graham Baxter Organization: Graham Baxter (Software) Limited Message-ID: <3B4ED95F.B7B11A91@NOSPAM.bcs.org.uk> VxWorkers, Has anyone ever seen an exception in VxWorks' tExcTask? The address 0x312e3138 is not in our address space. All our tasks were running. There were no stack errors. I guess some of its data has been corrupted. instruction access Exception next instruction address: 0x312e3138 Machine Status Register: 0x40009030 Condition Register: 0x44000040 Task: 0x1ef4cc8 "tExcTask" Our target in PPC 8260. Thanks for any ideas. - -- Regards, Graham Baxter gbaxter@NOSPAM.bcs.org.uk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Fri, 13 Jul 2001 08:08:20 -0400 From: "Chris Slominski" Organization: Jefferson Lab Message-ID: <9imocq$inu$1@inn.jlab.org> References: <6qH17.79$5f6.133962@news.uswest.net> VxWorks Users, I have had much success with using their OS, I don't bother with many of their tools, and their customer support is marginal. Chris "Joe Woodbury" wrote in message news:6qH17.79$5f6.133962@news.uswest.net... > Since we ordered VxWorks/Tornado II, we've had nothing but problems with > Wind River. They filled the order incorrectly and I had to resort to a lot > of yelling and threats to not pay our purchase order to get their attention. > Yet, two days later, they still have yet to do anything but email the > correct BSP. On top of all this, the highest level manager at Wind River > kept blaming us for some of the problems. > > So far, I've concluded that Wind River employees are given an IQ test. If > they fail, they are given a personal relationship test. If they fail that > also, they are hired. > > And heaven forbid they actually do something so obvious like creating a > compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. > > The shame is that VxWorks proper is such an excellent product. > > The questions are: Are our experiences unique? Did our company just happen > to run into the one group of bastards working at Wind River? > > Joe Woodbury > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ataDrv arguments Date: Fri, 13 Jul 2001 14:15:18 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B4EE656.EAA44251@ks.ericsson.se> References: <22ba361d.0107130236.76378b0b@posting.google.com> Go ahead and use it as if configType was an int, it's a typo in headerfiles, printed doc, online doc (and maybe some more places). Check the source code for ataDrv in: $WIND_BASE/src/drv/hdisk/ataDrv.c and se for yourself. HTH Urban lost distance wrote: > > The manual page for ataDrv (ATA/IDE disk device driver) states that it > takes a configuration flag as a parameter. Good - I want to use this > to configure PIO mode 4. > > However the formal parameter - "configType" - is defined as a BOOL. > Bad - how can I possibly pass in the various configuration flag values > to a BOOL? > > Any enlightenment will be gratefully accepted. > > Thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPC405 PCMCIA driver with ATA support Date: Fri, 13 Jul 2001 12:27:02 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: References: <01554D7E32E7D1118A8800805F31D8A3012C9B0D@tiger.seabridge.co.il> > 1. Is that at all true? False > 2. Does anyone have experience with porting PCMCIA support to PPC, with > ATA support (on dosFS) I port it on PPC860T and PPC855T > 3. Even better, anyone with DMA experience? Sorry ... not yet. "Adam Abrams" wrote in message news:01554D7E32E7D1118A8800805F31D8A3012C9B0D@tiger.seabridge.co.il... > It seems to me that VxWorks under PPC has no built-in support for > PCMCIA, as the driver doesn't compile and link - and seems very > Intel/PCI-ISA dependant. > > 1. Is that at all true? > > 2. Does anyone have experience with porting PCMCIA support to PPC, with > ATA support (on dosFS) > > 3. Even better, anyone with DMA experience? > > Thanks, > > Adam. > --------------------------- Newsgroups: comp.arch.embedded,comp.os.linux.hardware,comp.os.linux.powerpc,comp.os.vxworks,comp.sys.powerpc.tech Subject: Re: Need Help - PowerPC MPC8xx SDRAM Setup Date: Fri, 13 Jul 2001 12:29:27 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: References: <3b4c3dae@news.integrityonline.com> I use 860T/855T and I use the same Micron SDRAM. I use the same init table from the CPV3060 BSP (romInit.s) /* UPM-A table data ( SDRAMs @ 50MHZ) */ cpvUpma50Mhz: ... "Frank Wolf" wrote in message news:l7lji9.ckm.ln@gum.de... > Hi Jeff! > > We're using PPC823 with VxWorks. > Maybe it helps when I send you our 'romInit.s' module which > does the initialization of the PPC? > > Frank > > > "Jeff Gentry" schrieb im Newsbeitrag > news:3b4c3dae@news.integrityonline.com... > > Anyone using Micron Semiconductor's 64 Mbit SDRAM, > > MT48LC4M16A2TG, 4M x 16, with a PowerPC design? > > > > If so, I am working with a hardware design that is > > having problems executing out of SDRAM. It may be > > an initialization or UPM setup problem. > > > > I have init code and UPM tables and want to verify > > them. > > > > Thanks, > > > > Jeff Gentry, > > Sr. Software Design Engineer > > PEI Electronics, > > An Integrated Defense Technologies Company > > Email: JGentry@pei-idt.com > > Voice: (256) 895-2106 > > Fax: (256) 895-2222 > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ISR callable routines Date: 13 Jul 2001 05:49:50 -0700 From: kumarpavan@lycos.com (pavan) Organization: http://groups.google.com/ Message-ID: <307af50b.0107130449.78a8db29@posting.google.com> References: <9ia0uc$am6$1@overload.lbl.gov> <676eb679.0107081314.26e89809@posting.google.com> hi, david has once posted a small snippet of code long back how to do that with a flag. pepsi_000@hotmail.com (vxJunkie) wrote in message news:<676eb679.0107081314.26e89809@posting.google.com>... > hi david, > > Thanks for the info. > What do you mean my > "add a flag so that the ISR doesn't schedule the 'work' routine > more than once " > Does netJobAdd schedule the same job more than once? > > Thanks > > "David Laight" wrote in message news:<9ia0uc$am6$1@overload.lbl.gov>... > > > Can i call functions like > > > > a] netClusterGet > > > b] netClBlkGet > > > c] netMblkGet > > > from the ISR context? I am not sure if these are blocking calls > > > > No - not a chance! > > (The routines uses semaphores to control access to the free list - which > > don't lock out interrupts.) > > > > vxWorks expects you to use netJobAdd to get tNetTask to go this work. > > > > Remember to add a flag so that the ISR doesn't schedule the 'work' routine > > more than once - or you WILL exceed the size of the kernel work Q and/or > > tNetTasks list. > > > > David --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MBX860 and DIMM Date: Fri, 13 Jul 2001 13:03:29 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: References: I don't know any brand but I can tell you information written on schematic page of MBX860 DI MM Connection 4K Refresh ONLY supported on : 2M x 64/72 DIMM Module 4M x 64/72 DIMM Module 8M x 64/72 DIMM Module 1K, 2K & 4K Refresh supported on 1M x 64/72 DIMM Module "John Bundy" wrote in message news:a14421a1.0107121538.5522f64b@posting.google.com... > Hi, > > Has anybody had success adding DRAM DIMM memory to the MBX860? > > I would be interested in what DIMM part number works and general bsp > changes. I am using pSOS, but it looks like a lot more vxworks > people have used the MBX860. > > John Bundy --------------------------- Newsgroups: comp.os.vxworks Subject: state machine implement Date: Fri, 13 Jul 2001 09:42:05 -0400 From: "Yong Zhang" Organization: Marconi Message-ID: <9imtr7$dn3$1@newsfeed.pit.comms.marconi.com> Hi, I will need to implement 7 state machines. These state machines are related. The state change in one state machine make some variables change, which will cause other state machines to change. Is there a easy way to implement this? Thanks! Yong --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Fri, 13 Jul 2001 09:03:29 -0500 From: "Lee" Message-ID: <9imv3j$js1sj$1@ID-92003.news.dfncis.de> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> Hi, Thank you for the info. It helps a lot. One more question here. If VxWorks has only one mode - kernel mode, then every piece of code is able to access all memory space and all hardware IO space, therefore every piece of code can do drivers's job. Why does VxWorks have use drivers? What are VxWorks' drivers exactly, and what type of jobs do they do? Thank you. Lee "Joe Gwinn" wrote in message news:joegwinn-1207012200580001@192.168.1.100... > In article <9ila0j$k005n$1@ID-92003.news.dfncis.de>, "Lee" > wrote: > > > Just wondering if VxWorks uses User Mode and Kernel Mode to separate user > > code from hardware-related code, or does it have only one mode? > > There is one mode. In fact, user tasks are directly linked to the kernel, > and call kernel services using ordinary subroutine calls (not trap > interrupts). > > Most other RTOSes at least separate user code from kernel code to the > extent of requiring a trap to invoke kernel services, but still everything > is in one big address space, and everything typically runs in kernel > mode. Some RTOSes allow one to run threads in user mode, but don't > require it. > > > > ... If it has > > only one mode, then high level application code can access hardware IO space > > directly. > > True. > > > Are there any potential problems? > > Basically, any task (~thread) can look at anything, stomp anything, in > other tasks or the kernel or I/O space or stackspace or whatever. That > said, this kind of a one-shared-address-space architecture has been > traditional in RTOSes for many decades, and it does work, with billions of > dollars of fielded systems still in everyday use. > > There are some RTOSes that isolate the tasks (processes) from one another > and from the kernel and perhaps the I/O drivers as well. The advantage is > that it's easier to figure out why the system went face-down, and to > recover from bugs. The disadvantage is lower performance and increased > overall complexity. > > If you want to hear all about the problems of one-space RTOSes, look into > QNX, which is a stringent member of the isolate-everything school. > www.qnx.com > > If you want to hear why isolate-everything is a terrible idea, look into > VxWorks. > > Which model is best is something of a religious issue; I like both, each > in their own way, having no vested interest in either. > > > Joe Gwinn --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MBX860 and DIMM Date: Fri, 13 Jul 2001 12:04:17 +0100 From: "Jeremy" Organization: Customer of Energis Squared Message-ID: References: Yes, we installed 64MB EDO DIMM single bank, 3.3V, 4K refresh. Can't remember where we got them. I seem to remember the jumper settings were different depending upon the revision of the board. There are eight HY51V65804A on each dimm. Using MBX bootmonitor you can read contents of DIMMs 24c02 nvram. Sorry can't be more help. John Bundy wrote in message news:a14421a1.0107121538.5522f64b@posting.google.com... > Hi, > > Has anybody had success adding DRAM DIMM memory to the MBX860? > > I would be interested in what DIMM part number works and general bsp > changes. I am using pSOS, but it looks like a lot more vxworks > people have used the MBX860. > > John Bundy --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: What is the main differences between i386 and i960? Date: Fri, 13 Jul 2001 10:05:14 -0500 From: "Lee" Message-ID: <9in2nd$jsr3p$1@ID-92003.news.dfncis.de> Just wondering what are the main differences between i386 and i960. Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: "File System" on VxWorks Date: Fri, 13 Jul 2001 10:56:10 -0500 From: "Lee" Message-ID: <9in5ms$jtkbu$1@ID-92003.news.dfncis.de> Hi, Could someone give me some info on "file systems" used on VxWorks? Is there any articles online addressing this? Thanks. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Re: state machine implement Date: Fri, 13 Jul 2001 13:20:35 -0400 From: "Kevin Livingston" Message-ID: References: <9imtr7$dn3$1@newsfeed.pit.comms.marconi.com> Without knowing more about your problem, message queues might be a good solution, if there is a lot of data flying back and forth. Otherwise a much simpilier synchronization solution might work like mutexs or semaphores. In all cases once reaching a state they could block on the call that takes them to the next state. Then when the approiate other machine changes it could unblock them - message, semaphore give, etc. (That is all assuming each machine is in it's own thread, if not it becomes much simplier, and you can just poll the variables etc.) When you have multiple threads accessing one data source you need to make sure you have some kind of protection on the common data to prevent corruption. Is that kinda what you were looking for? Kevin "Yong Zhang" wrote in message news:9imtr7$dn3$1@newsfeed.pit.comms.marconi.com... > Hi, > > I will need to implement 7 state machines. These state machines are related. > The state change in one state machine make some variables change, which will > cause other state machines to change. Is there a easy way to implement this? > > Thanks! > > Yong > > --------------------------- Newsgroups: comp.os.vxworks Subject: unresolved symbol pciHeaderShow when downloading Date: Fri, 13 Jul 2001 13:37:17 -0400 From: Michael Reed Message-ID: <3B4F31CD.31EBDE0C@ll.mit.edu> Hello, My app compiles without problems, but when I go to download it to the target (MVME2400), it fails with an Unresolved Symbol error for pciHeaderShow. Do I have to rebuild my vxWorks image to include this pciHeaderShow code so that it's on the target at download time? Thanks, Mike --------------------------- Newsgroups: comp.os.vxworks Subject: Rebuilding C++ String (STL) library in Tornado? Date: 13 Jul 2001 10:50:05 -0700 From: vannevar@op.net (Andy Kaiser) Organization: http://groups.google.com/ Message-ID: <905086e3.0107130950.71db6efd@posting.google.com> I've been working with Wind River customer service on this for over a month now, and they're clueless on the problem, so if anyone has any help, I'd appreciate it! We've got a bug in our application somewhere that causes a length assertion error to be fired in the program bastring, part of the C++ String class (I think). Basically, it's when one String is copied into another, and the length of the first is insufficient to hold the second, it seems. Probably easy enough to fix, but the catch is that we have no idea who's causing it to happen! The assertion error tells you were the error is in bastring.cc, but that function is embedded deep down in STL-land, so it could be called by any number of things. (Strings, of course, but also possibly Vectors and other stuff.) What I'd like to do is compile my own version of the STL library into the kernel, and include a stack dump with the assertion printout to find out the source of the offending function call. I've got the modified bastring.o file ready to be included, but I see no way of removing the existing bastring to replace it. (It's not in the lib...gnuvx.a archive.) Does anyone have any clue how to do this? I've tried about a dozen suggestions from Wind River support, and none of those have worked. Thanks in advance for any advice anyone out there may have! Andy Kaiser Software Engineer L-3 Communications In case you need to know: Host: Win32 (Windows NT) Target: PPC604 Version: Tornado 2.0.2 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Fri, 13 Jul 2001 12:14:53 -0500 From: "tazboy78" Organization: http://extra.newsguy.com Message-ID: <9ina7v01uqd@enews1.newsguy.com> References: <6qH17.79$5f6.133962@news.uswest.net> <9ikchr$ou1$1@localhost.localdomain> > Mediocre support and service. Compared to who? --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: 13 Jul 2001 11:41:01 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107131041.503bce9@posting.google.com> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <3B4E4FF5.66545477@nc.rr.com> Hello all, Tim Sohacki wrote in message news:<3B4E4FF5.66545477@nc.rr.com>... > > All tasks running under VxWorks 5.4 run in supervisor > (Kernel) mode. I am sure this is true for all processor > architectures, as this was a tenet of VxWorks for a long > time. VxWorks AE (basically, the version after VxWorks 5.4) > may utilize User Mode to accomplish the memory protection > that it offers. Indeed it does make use of user mode, but in keeping with VxWorks traditions the choice to do so is yours as the system designer. By default, non-kernel domains execute in user mode to ensure maximum protection, but VxWorks AE also supports a concept of "privileged" domains outside the kernel where all tasks run in supervisor mode too. HTH, John... --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Fri, 13 Jul 2001 20:48:11 +0200 From: "Mikkel Elmholdt" Organization: CyberCity Internet Message-ID: <9infpn$p73$1@news.cybercity.dk> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> "Lee" wrote in message news:9imv3j$js1sj$1@ID-92003.news.dfncis.de... > Hi, > > Thank you for the info. It helps a lot. > > One more question here. If VxWorks has only one mode - kernel mode, then > every piece of code is able to access all memory space and all hardware IO > space, therefore every piece of code can do drivers's job. Why does VxWorks > have use drivers? > What are VxWorks' drivers exactly, and what type of jobs do they do? The same as in every other RTOS, which does not have virtual memory support (i.e. 95 out of 100). It mainly provides a nice separation of functionality into modules, which increases maintainability. Your point is correct however. There is nothing stopping you from doing driver-like stuff in "application programs" if you so desire. Mikkel --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: 13 Jul 2001 11:48:25 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107131048.53654628@posting.google.com> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> Hello Lee, > One more question here. If VxWorks has only one mode - kernel mode, then > every piece of code is able to access all memory space and all hardware IO > space, therefore every piece of code can do drivers's job. Why does VxWorks > have use drivers? The advantage to having formalised drivers is that your application code can make use of standard, portable APIs: open(), read(), write(), ioctl() and close(). You can also insert a layer of abstraction between devices and the application. For example, suppose you needed to support two different camera devices on different systems with the same application, by creating a device driver for each camera that installs itself using the name, say, "/camera" then your application only has to open "/camera" to get to whichever of the cameras is connected to the target. > What are VxWorks' drivers exactly, and what type of jobs do they do? The drivers are basically providing the standard API functions mentioned above, or a subset of them if they don't all make sense. Converting your non I/O system driver code into an I/O system driver should be relatively straightforward for most types of devices. The more complex ones are the block devices, such as disks. I would recommend reading the VxWorks Programmer's Guide which, if you don't already have it, can be downloaded either from Wind Surf or by completing the form at: http://www.wrs.com/products/html/manual-form.html There is a chapter in there on the I/O system that includes an explanation of how to write simple drivers for the I/O system. HTH, John... --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: What's the latest version of VxWorks? Date: Fri, 13 Jul 2001 20:50:17 +0200 From: "Mikkel Elmholdt" Organization: CyberCity Internet Message-ID: <9inftl$pb1$1@news.cybercity.dk> References: <9il8hi$j1ktd$1@ID-92003.news.dfncis.de> "Lee" wrote in message news:9il8hi$j1ktd$1@ID-92003.news.dfncis.de... > Hi, > > Got two general questions here. > > Q1: what is the latest version of VxWorks. > Q2: what is the latest version of Tornado? > Q3: which versions of VxWorks are included in Tornado 2.0 and Tornado 2.1? In your place I would contact the nearest Wind River office or partner, or perhaps e-mail WRS directly for such factual questions. Mikkel --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ROM based vxworks build with T2 Project: any success ? Date: 13 Jul 2001 12:03:13 -0700 From: pfliv@yahoo.com (Paul Livingston) Organization: http://groups.google.com/ Message-ID: <7b1623b6.0107131103.410c599@posting.google.com> References: <7Ix37.17948$Kf3.233634@www.newsranger.com> We ran into similar problems making vxworks_rom images (PPC targets). Finally, WRS admitted there is a bug in elftobin.exe. We were able to get elftobinrom.exe from WRS, modify rules.vxworks to call elftobinrom when making rom images and everything has worked since. Good Luck! Diego Serafin wrote in message news:<7Ix37.17948$Kf3.233634@www.newsranger.com>... > Hi folks ! > I'm trying to build a ROM based version of VxWorks (vxworkx_rom and > vxworks_romCompress images) using the Tornado 2 Project tool, but found that > image I produced did not boot properly. What happens: > > 1) romInit works fine. > 2) romStart works fine but, at the end, the jump to absEntry that should fire > usrEntry() crashes system. > > I saw that vxworks module, stored at (rom address of) binArrayStart does not > start with code, but with .ELF header so that when romStart jumps to > RAM_DST_ADRS it fetches rubbish (...".ELF"). The real usrEntry code sits at > 0x100 bytes from RAM_DST_ADRS. > > I found that elfToBin tool, during image make-ing, outputs a module that still > starts with ".ELF" ... is this correct ? > > If I modify romStart's copyLong() call to start copying flash 0x100 bytes ahead > the systems comes up correctly. > > Where am I wrong ? > > Thank You, > Ciao. > Diego. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ROM based vxworks build with T2 Project: any success ? Date: Fri, 13 Jul 2001 14:18:24 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B4F4980.946C2F3F@fnal.gov> References: <7Ix37.17948$Kf3.233634@www.newsranger.com> <7b1623b6.0107131103.410c599@posting.google.com> > > I saw that vxworks module, stored at (rom address of) binArrayStart does not > > start with code, but with .ELF header so that when romStart jumps to > > RAM_DST_ADRS it fetches rubbish (...".ELF"). The real usrEntry code sits at > > 0x100 bytes from RAM_DST_ADRS. i faced the same problem ... i used to snip off the header off the image ... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Re-re: Anyone any experiences building gcc 2.95.X Cygwin cross x86 Date: 13 Jul 2001 12:34:55 -0700 From: nebraskin@yahoo.com (Tim Nebraskin) Organization: http://groups.google.com/ Message-ID: <5a958b4a.0107131134.48408092@posting.google.com> References: <9ime6p$a80$1@overload.lbl.gov> fscheurer@manz-acs.de wrote: > Especially the FIND/SORT error messages I cannot track > anymore. Where's their source? Details below: > > FIND: Parameterformat nicht korrekt Removing unneeded directories: > SORT: Unzul"ssige Option autogen: not found > You need to install autogen > .. > > To shorten it: The FIND/SORT problem of make/gcc or anything is > also the problem that autogen hasn't been build. It seems to me that make runs windows/dos's FIND and SORT, but not the cygwin's. You probably need to check the value of $PATH variable from your bash prompt. It must have /usr/local/bin:/usr/bin:/bin before any /cygdrive/c/WINNT/SYSTEM32 in it... Hopefully, this will help you a bit... --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: What is the main differences between i386 and i960? Date: Fri, 13 Jul 2001 15:10:53 -0500 From: "Stanley R. Allen" Organization: Raytheon Company Message-ID: <3B4F55CD.EB2AC359@gsde.hou.us.ray.com> References: <9in2nd$jsr3p$1@ID-92003.news.dfncis.de> Reply-To: Stanley_R_Allen-NR@Raytheon.com Lee wrote: > > Just wondering what are the main differences between i386 and i960. > In spite of the fact that these were made by the same company, they are totally different. - -- Stanley Allen mailto:Stanley_R_Allen-NR@Raytheon.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPC405 PCMCIA driver with ATA support Date: Fri, 13 Jul 2001 22:10:27 +0100 From: molochai Message-ID: <63629c994a%molochai@vapour-trail.demon.co.uk> References: <01554D7E32E7D1118A8800805F31D8A3012C9B0D@tiger.seabridge.co.il> In message <01554D7E32E7D1118A8800805F31D8A3012C9B0D@tiger.seabridge.co.il> Adam Abrams wrote: > It seems to me that VxWorks under PPC has no built-in support for > PCMCIA, as the driver doesn't compile and link - and seems very > Intel/PCI-ISA dependant. > > 1. Is that at all true? > > 2. Does anyone have experience with porting PCMCIA support to PPC, with > ATA support (on dosFS) I've got the WindRiver PCMCIA driver working on an 8240 after some fairly substantial alterations. I'm using PCMIA ATA flash cards with dosFs2. The code that is shipped with Tornado II is very little endian specific and is unsupported on other architectures (not one of WRS's better efforts in places to TBH). Not surprisingly, WRS don't support it on non x86 targets. I got it to work using a TI PCI->Cardbus bridge [1] (which supports 16 bit PCCards - I didn't get round to getting cardbus to work, but it didn't look hard). In the source code there are drivers for two types of bridges (and their clones) I guess yours will be one or the other. In both cases, they appear to originally have been ISA-style parts (16 PCCard is practically ISA anyway, and Cardbus is PCI with a few extras). This is of course not a problem for any type of PCI bridge, although there will be stuff outside the scope of the driver that needs setting up. Most of the source code you'll need to fix is in cisLib.c. Practically every access of a 16 or 32 bit CIS record needs fixing up work on a big-endian architecture, although I recall there being a few other places... The other thing that may need attention was interrupts. The driver rather assumes that each of the cards and the bridge have seperate interrupt lines - something which is not the case with the TI part over PCI. If you have to share lines, a pseudo-interrupt controller is in order as you'll want to detach ISRs when cards are removed. If you're using dosFs2, you may find that the PPC upgrade doesn't include the modified source code files as documented in the release notes - these merely fix a gross memory leak associated with the original dosFs. > > 3. Even better, anyone with DMA experience? I didn't, but then the WRS generic ATA driver doesn't use it either (in any case, you'll need to get WRS to release the source code for this to you). It also looked like a bitch to get PCCard style DMA working. For Cardbus, it's easy. However, some PowerPC parts have DMA controllers that ought to be capable of initiating transfers from the processor end using the ATA controller in polled mode. Good luck.. [1] Can't remember the part number of the top of my head - and haven't got the source code to hand right now. In any case, it's a clone of an Intel part whose part number also escapes me. However, it works with one of the two bridge drivers present. --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Fri, 13 Jul 2001 16:11:00 -0500 From: "Lee" Message-ID: <9ino55$jccd8$1@ID-92003.news.dfncis.de> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> <488e459a.0107131048.53654628@posting.google.com> Hi John and Everyone, Thank you for the info. It helps a lot. Regards, Lee "John" wrote in message news:488e459a.0107131048.53654628@posting.google.com... > Hello Lee, > > > One more question here. If VxWorks has only one mode - kernel mode, then > > every piece of code is able to access all memory space and all hardware IO > > space, therefore every piece of code can do drivers's job. Why does VxWorks > > have use drivers? > > The advantage to having formalised drivers is that your application > code can make use of standard, portable APIs: open(), read(), write(), > ioctl() and close(). > > You can also insert a layer of abstraction between devices and the > application. For example, suppose you needed to support two different > camera devices on different systems with the same application, by > creating a device driver for each camera that installs itself using > the name, say, "/camera" then your application only has to open > "/camera" to get to whichever of the cameras is connected to the > target. > > > What are VxWorks' drivers exactly, and what type of jobs do they do? > > The drivers are basically providing the standard API functions > mentioned above, or a subset of them if they don't all make sense. > Converting your non I/O system driver code into an I/O system driver > should be relatively straightforward for most types of devices. The > more complex ones are the block devices, such as disks. > > I would recommend reading the VxWorks Programmer's Guide which, if you > don't already have it, can be downloaded either from Wind Surf or by > completing the form at: > http://www.wrs.com/products/html/manual-form.html > > There is a chapter in there on the I/O system that includes an > explanation of how to write simple drivers for the I/O system. > > HTH, > > John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Re : Assembler with VxWorks Date: 13 Jul 2001 16:01:07 -0700 From: Chris Torek Organization: None of The Above Message-ID: <9inujj$go4$1@elf.eng.bsdi.com> References: <9i1sqg$4go$1@overload.lbl.gov> Sender: torek@elf.eng.bsdi.com In article <9i1sqg$4go$1@overload.lbl.gov> David Laight wrote: >I don't know how good gcc is at asm functions. GCC is *very* good at inline "asm". In fact, the original example probably should be redone as at least one #define: >> void outWrds(void) >> { >> unsigned short port = 0x378; >> unsigned char value = 0xff; >> __asm__ volatile ("outb %1,%0;" : : "d" (port) , "a" (value) ); >> } /* "inline.h", i386-specific */ #define outb(port, val) \ __asm__ volatile("outb %%al,%%dx" :: "d"(port), "a"(val)) /* the file that uses outWrds() can then just do this, if there is no reason to require it to be an actual function */ #define outWrds() outb(/*port*/0x378, /*value*/0xff) The reason gcc is so good at inline "asm" constructs is that the syntax above -- using the "d" and "a" constraints to force the value and port to wind up in the %dx and %al registers (actually %edx and %eax) is the *same* as the internal stuff gcc uses to understand how to produce machine-code in the first place. In other words, by expressing the constraints on the instruction, gcc winds up "knowing" what the instruction "means". In the absence of the "volatile" keyword, gcc will even move or combine the instruction as appropriate, e.g., hoist it out of a loop. (This is one reason it is important to use the volatile keyword for instructions with side effects.) >It is MUCH easier (and safer) to write the code in a .s file! This is true for other compilers, or for large functions. With gcc and small code snippets, it tends to be just as easy and safe to devise an inline __asm__ construct -- but first you have to wade through the gcc info file, and sometimes the .md file for the target machine, to learn all about its constraints. Note: always inspect the resulting assembly (use "gcc -S ...") the first few times, to make sure you got it right. :-) - -- In-Real-Life: Chris Torek, Wind River Systems El Cerrito, CA, USA Domain: torek@bsdi.com +1 510 234 3167 http://claw.eng.bsdi.com/torek/ (not always up) I report spam to abuse@. --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Fri, 13 Jul 2001 19:29:08 GMT From: Adam Dunkels Organization: Telia Internet Message-ID: <8_H37.5136$z21.527498@newsc.telia.net> References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> Followup-To: comp.os.vxworks Frank da Cruz wrote: > In article <3B4DDE9B.C12DEF3@digidata.com>, > Doug Dotson wrote: > : It shouldn't be too much of a problem to port it. vxWorks is > : a bit different in that you don't build executables in the UNIX > : sense. You build .o files and they are linked when loaded. There > : is no main(). But if you changed main() to kermit() you should be > : OK. Argument typed on the command line are mapped into the > : function arguments. You will have to change some header files > : and the makefile won't work as it is. > : > I'd be glad to give it a shot if I had VxWorks developer tools to > build the kermit function and VxWorks itself to test it, but I > have neither. Would someone who does have them like to try adapting > G-Kermit to VxWorks? There's clearly a demand, at least among those > who don't mind devoting somewhere between 30 and 80K, depending on > the architecture, to a file-transfer function, and it doesn't sound > like too much work. But gkermit is released under the GPL, which means that you cannot distribute a product that includes gkermit unless you are willing to release and distibute all source code for your project, including the source code to VxWorks. I don't know much about VxWorks, but I assume that it is not free software licensed under the GPL. /adam - -- Adam Dunkels (spambait) http://dunkels.com/adam/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: unresolved symbol pciHeaderShow when downloading Date: 13 Jul 2001 19:32:31 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107131832.7352c097@posting.google.com> References: <3B4F31CD.31EBDE0C@ll.mit.edu> Hello, > My app compiles without problems, but when I go to download it to the > target (MVME2400), it fails with an Unresolved Symbol error for > pciHeaderShow. Do I have to rebuild my vxWorks image to include this > pciHeaderShow code so that it's on the target at download time? Yes, if your application needs that feature. HTH, John... --------------------------- Newsgroups: comp.protocols.snmp,comp.os.psos,comp.os.vxworks Subject: Re: Inconsistency while mibcomp -man -trap.c Date: Sat, 14 Jul 2001 00:09:21 -0500 From: Jim Jones Organization: netINS, Inc. Message-ID: <3B4FD3FB.4120635E@icon-labs.com> References: <3B49A8F3.5CE751AC@bigfoot.com> Reply-To: jim_jones@icon-labs.com Alex, No.. it is not a bug. At least not in the trap.c output. If you look at how the static value len_totalStarted is USED within trap.c you will see that it should be 13, not 12. It will fill in the zero-initialized instance part with the proper value... and the 13 is the size of the entire OID (instance part included). The -man (and -man.h) file is just one of several outputs that can be generated that you COULD use if you find it useful for some programming effort. The len_Foo in this case excludes the instance part (but I would look at some columnar examples with multi-component indexes to be sure). If so, then the instance length equals the array length minus the len_Foo value for it. JJ Alex Vinokur wrote: > ============================ > pRISM+ for pSOSystem/PowerPC > pRISM+ 1.2.3 . pSOSystem 2.2.2 > Diab Data D-CC Optimizing C compiler > SNMP Release 7.4 (SNMPv1) > Epilogue Technology Emissary SNMP MIB Compiler, version 6.3 > Host : Windows98 > Target : PowerPC (MPC860) > ============================ > > Files leaves.c and trap.c have been generated with mibcomp. > > mibcomp -man -o leaves.c > mibcomp -trap.c -o trap.c > > //------- Fragment of leaves.c ------- > > OIDC_T ID_totalStarted[] = { 1, 3, 6, 1, 4, 1, 4466, 1, 5, 1, 1, 5, 0 }; > > unsigned int len_totalStarted = 12; > > //------------------------------------ > > //------- Fragment of trap.c --------- > > static OIDC_T ID_totalStarted[] = { 1, 3, 6, 1, 4, 1, 4466, 1, 5, 1, 1, > 5, 0 }; > static int len_totalStarted = 13; > > //------------------------------------ > > We can see that ID_totalStarted[] has the same value in both files. > However len_totalStarted has different values in the same files. > > Is it a bug of mibcomp? > > ================================== > Alex Vinokur > mailto:alexvn@bigfoot.com > http://up.to/alexvn > http://go.to/alexv_math > ================================== --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 15 04:03:27 2001 From: Vxworks Exploder Date: Sun Jul 15 04:03:29 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 15 04:03:12 PDT 2001 Subject: Re: unresolved symbol pciHeaderShow when downloading Subject: Re: What is the main differences between i386 and i960? Subject: Re: Kermit for VxWorks Subject: Re: Kermit for VxWorks Subject: Re: Kermit for VxWorks Subject: Re: Kermit for VxWorks Subject: Re: Is Wind River really this pathetic Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Exception Handler Question... [excVecSet() vs excHookAdd() on a 8260] Subject: Re: Does VxWorks use User Mode & Kernel Mode? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: unresolved symbol pciHeaderShow when downloading Date: Sat, 14 Jul 2001 12:08:34 GMT From: "Kelly Hornsby" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: <6DW37.51528$w5.5253609@news1.rdc1.ga.home.com> References: <3B4F31CD.31EBDE0C@ll.mit.edu> For a while, this symbol (and others) was missing, and there was no way to include the pci show functions because of a bug in T2. Make sure your Tornado has all patches. Otherwise, you need to include the pci show suite, and then do a lkup on pciHeaderShow when you've rebuilt. If it still doesn't show up, then you need to check your patches AGAIN to see if the pciHeaderShow patch is actually applied. I went through this, but took a while to finally determine that the pciHeaderShow was not being included (and being declared as non-STATIC). "Michael Reed" wrote in message news:3B4F31CD.31EBDE0C@ll.mit.edu... --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: What is the main differences between i386 and i960? Date: Sat, 14 Jul 2001 09:41:47 -0600 From: Dan Henry Organization: Independent Message-ID: References: <9in2nd$jsr3p$1@ID-92003.news.dfncis.de> <3B4F55CD.EB2AC359@gsde.hou.us.ray.com> On Fri, 13 Jul 2001 15:10:53 -0500, "Stanley R. Allen" wrote: >Lee wrote: >> >> Just wondering what are the main differences between i386 and i960. >> > >In spite of the fact that these were made by the same company, >they are totally different. The 386 is a 32-bit CISC CPU that evolved from the 8086, an architecture that has continued to evolve into today's Pentium class CPU's. The 960 is a 32-bit RISC CPU; a different architecture entirely. - --Dan Henry (dhenry@sprynet.com) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Kermit for VxWorks Date: 14 Jul 2001 15:42:07 GMT From: fdc@watsun.cc.columbia.edu (Frank da Cruz) Organization: Columbia University Message-ID: <9ipp8f$2ob$1@newsmaster.cc.columbia.edu> References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> <8_H37.5136$z21.527498@newsc.telia.net> In article <8_H37.5136$z21.527498@newsc.telia.net>, Adam Dunkels wrote: : ... : But gkermit is released under the GPL, which means that you cannot : distribute a product that includes gkermit unless you are willing to : release and distibute all source code for your project, including the : source code to VxWorks. I don't know much about VxWorks, but I assume that : it is not free software licensed under the GPL. : I'm not a lawyer but I think that's an extreme interpretation. After all, many commercial UNIX vendors distributed gcc with their proprietary OS's: DG/UX and NeXTSTEP spring to mind (in which gcc was the *only* C compiler), and they certainly did not also distribute source code for their OS's. Anyway, it's 2001 and license discussions have been done to death. Let me only say this about G-Kermit's license: there was going to be a GPL'd Kermit program one way or another. Better that it come from the Kermit Project, so it interoperates with other Kermit programs. The important thing is to get something working on VxWorks. Licensing can be sorted out later. In fact, as the author of G-Kermit, I believe I have the right to issue multiple copies of it, each under a different license. - - Frank --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Sat, 14 Jul 2001 09:00:45 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B506CAD.D85AC443@covad.net> References: <9ikjrq$asm@watsun.cc.columbia.edu> Reply-To: drdiags@covad.net Frank, Some points I have noticed during the porting of gkermit. There are several functions and structures which I probably would declare static, such as lkup() in the gwart.c file. vxWorks has a similar function. So I am going through all the source and leaving global those functions that make sense for me. I may not complete this in a timely manner, so hopefully someone else has already moved further along than I. Frank da Cruz wrote: > > From time to time, we get requests for Kermit file transfer software > for VxWorks. I'd like to be able satisfy them. > > Does VxWorks have a C compiler, the basic minimum well-known header > files, a C library, and a make program? Does it support the UNIX > notion of standard i/o? If so, then maybe it's just a simple matter > of getting the G-Kermit sources files onto VxWorks and typing "make": > > http://www.columbia.edu/kermit/gkermit.html > > Would anybody care to give it a try and report back? Thanks! > > Frank da Cruz > The Kermit Project > Columbia University > http://www.columbia.edu/kermit/ - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Sat, 14 Jul 2001 13:33:01 GMT From: fred smith Organization: None! Message-ID: References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> <8_H37.5136$z21.527498@newsc.telia.net> Followup-To: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Sender: fredex@fcshome.stoneham.ma.us Adam Dunkels wrote: : Frank da Cruz wrote: :> In article <3B4DDE9B.C12DEF3@digidata.com>, :> Doug Dotson wrote: :> I'd be glad to give it a shot if I had VxWorks developer tools to :> build the kermit function and VxWorks itself to test it, but I :> have neither. Would someone who does have them like to try adapting :> G-Kermit to VxWorks? There's clearly a demand, at least among those :> who don't mind devoting somewhere between 30 and 80K, depending on :> the architecture, to a file-transfer function, and it doesn't sound :> like too much work. : But gkermit is released under the GPL, which means that you cannot : distribute a product that includes gkermit unless you are willing to : release and distibute all source code for your project, including the : source code to VxWorks. I don't know much about VxWorks, but I assume that : it is not free software licensed under the GPL. Er, no, I don't think that's what the GPL says. If you distribute a GPL'd work you must make its source available. If you distribute a work DERIVED FROM a GPL'd work you must make its source available. It doesn't say anything about distributing source for everything else in the package accompanying the GPL'd work. I don't think the discussion was about adding gkermit to the VxWorks package (though I don't see anything wrong with that either), but merely about trying to use a kermit protocol engine to do file transfers to/from VxWorks. Don't listen to MS's FUD about how the GLL infects everything it touches, because in an example like this it doesn't. - -- - ---- Fred Smith -- fredex@fcshome.stoneham.ma.us ---------------------------- "For the word of God is living and active. Sharper than any double-edged sword, it penetrates even to dividing soul and spirit, joints and marrow; it judges the thoughts and attitudes of the heart." - ---------------------------- Hebrews 4:12 (niv) ------------------------------ --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: 14 Jul 2001 16:26:45 GMT From: fdc@watsun.cc.columbia.edu (Frank da Cruz) Organization: Columbia University Message-ID: <9iprs5$49g$1@newsmaster.cc.columbia.edu> References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B506CAD.D85AC443@covad.net> In article <3B506CAD.D85AC443@covad.net>, wrote: : Some points I have noticed during the porting of gkermit. There : are several functions and structures which I probably would declare : static, such as lkup() in the gwart.c file. vxWorks has a similar : function. So I am going through all the source and leaving global : those functions that make sense for me. I may not complete this in : a timely manner, so hopefully someone else has already moved further : along than I. : Yes, at least one other person is working this. If anybody else is working on it too, please get in touch with me so I can coordinate. Obviously I underestimated the differences between VxWorks and other UNIXlike RTOSs like QNX and OS-9. It appears that traditionally structured C programs like G-Kermit (and C-Kermit) are not a good fit to the VxWorks model. G-Kermit was not designed for embedding. Since my original posting I have learned that: . Software for VxWorks is cross-built on other platforms, such as Sun. . One doesn't have programs in VxWorks, one has functions. Thus there can be no main(). . When a function is invoked from the VxWorks shell, command-line arguments are mapped to function arguments. Thus the handling of argc/argv must be reworked. . All memory is global, there is no separate address space for different tasks (thus the need to be careful with global variables and buffers). . VxWorks public functions are object files, not executables; they are linked when loaded. Another guy who tried building G-Kermit on VxWorks got clean compiles for all modules except gunixio.c, where the following functions had errors. Here is what those functions should do: . sysinit() gets and saves the console terminal modes (if such a concept even exists on VxWorks) and then sets the console terminal to ignore keyboard interrupts. If VxWorks does not support console terminal keyboard interrupts then sysinit() has nothing to do. . ttpkt() sets the console terminal modes (if necessary) to allow Kermit to read a stream of bytes literally, without the console driver getting in the way. It also needs to enable some form of flow control, if VxWorks supports such a thing, such as RTS/CTS or Xon/Xoff. It also instructs the OS not to echo incoming characters. . ttflui() reads and discards any characters that have arrived and are sitting in the console buffer, but have not yet been read by Kermit. It must not block. . ttinl() reads a Kermit packet with a given timeout -- that is, if the packet does not arrive within "timo" seconds, it's supposed to give up and return -1. In UNIX this is done with alarm() and signal(), which apparently don't exist in VxWorks? Also, note that gwart.c and gproto.w can probably be ignored. gproto.w is the Kermit protocol state table interpreter, written in a Lex-like language called "wart". gwart.c is compiled and saved as "wart", and then gproto.w is run through wart to produce gproto.c. But gproto.c is already included. You don't need to mess with wart and gproto.w unless you are changing the protocol module. In any case, the wart program is run only on the development system, not on VxWorks. - - Frank --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Sat, 14 Jul 2001 10:36:48 -0600 From: "Joe Woodbury" Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> Follow up from original poster: As an FYI, I found that after yelling alot at my local reps and their bosses and threatening to halt payment of our PO, they started acting very nice. Mind you, I'm still at the beginning of the project, so I'll have plenty of time to curse them again later. In the vain hope someone who actually matters a damn at Wind River monitors this forum, I have some simple questions: Are you completely stupid? Most companies act very nice and reserve the screwing over until the customer is fully committed. You alienate your customers from the moment you get the purchase order, if not before. In case you haven't noticed, the computer industry is growing up; if you continue acting like this, competitors are going to roll right over you. If you make the mistake of thinking that could never happen, you've already lost. And; Why the hell did you buy the Diab compiler and yet continue dinking around with GCC? Does anyone in your company actually have an MBA or even a business degree? Does anyone there even comprehend the concept of resource allocation and the problem of maintaining self-competing product lines in a low volume industry? Beyond that, why would you spent money to develop something you don't even own? --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Sat, 14 Jul 2001 22:51:45 GMT From: joegwinn@mediaone.net (Joe Gwinn) Organization: Gwinn Instruments Message-ID: References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> In article <9imv3j$js1sj$1@ID-92003.news.dfncis.de>, "Lee" wrote: > Hi, > > Thank you for the info. It helps a lot. Welcome. > One more question here. If VxWorks has only one mode - kernel mode, then > every piece of code is able to access all memory space and all hardware IO > space, therefore every piece of code can do drivers's job. True, and many people do just that. > Why does VxWorks [even] use drivers? As has been ably explained by other posters, to provide modularization. The basic issue not mentioned is that most hardware devices require a single stream of commands, and cannot deal with multiple overlapping streams. So, one must have a bit of software to sort out the perhaps conflicting streams of commands and reorder them into something the device can deal with. This bit of software is the driver, although it may be called something else. > What are VxWorks' drivers exactly, and what type of jobs do they do? This has been ably described by other posters, and I have nothing to add. One thing that may not have been clear in the other postings is the true distinction between kernel mode and user mode. It generally has nothing to do with memory and address spaces, the difference is in the machine instructions that are permitted. In kernel mode, all instructions are permitted, no matter how dangerous. In user mode, such instructions are forbidden, causing an illegal-instruction trap if attempted. Back when I built and altered RTOSes, I would add a special instruction for a user-mode task to turn kernel mode on and off, but with some bookkeeping. The strategy was to run in the safety of user mode until it was time to do something draconian, request kernel mode, do the deed, then relinquish kernel mode. If something bad happened, the memory dump would reveal who had the sharp knife checked out, focusing the search for the perp. It was quite rare for there to be a misadventure with the knife though, probably because the time window of vulnerability was so short. Joe Gwinn > "Joe Gwinn" wrote in message > news:joegwinn-1207012200580001@192.168.1.100... > > In article <9ila0j$k005n$1@ID-92003.news.dfncis.de>, "Lee" > > wrote: > > > > > Just wondering if VxWorks uses User Mode and Kernel Mode to separate user > > > code from hardware-related code, or does it have only one mode? > > > > There is one mode. In fact, user tasks are directly linked to the kernel, > > and call kernel services using ordinary subroutine calls (not trap > > interrupts). > > > > Most other RTOSes at least separate user code from kernel code to the > > extent of requiring a trap to invoke kernel services, but still everything > > is in one big address space, and everything typically runs in kernel > > mode. Some RTOSes allow one to run threads in user mode, but don't > > require it. > > > > > > > ... If it has > > > only one mode, then high level application code can access hardware I/O space > > > directly. > > > > True. > > > > > Are there any potential problems? > > > > Basically, any task (~thread) can look at anything, stomp anything, in > > other tasks or the kernel or I/O space or stackspace or whatever. That > > said, this kind of a one-shared-address-space architecture has been > > traditional in RTOSes for many decades, and it does work, with billions of > > dollars of fielded systems still in everyday use. > > > > There are some RTOSes that isolate the tasks (processes) from one another > > and from the kernel and perhaps the I/O drivers as well. The advantage is > > that it's easier to figure out why the system went face-down, and to > > recover from bugs. The disadvantage is lower performance and increased > > overall complexity. > > > > If you want to hear all about the problems of one-space RTOSes, look into > > QNX, which is a stringent member of the isolate-everything school. > > www.qnx.com > > > > If you want to hear why isolate-everything is a terrible idea, look into > > VxWorks. > > > > Which model is best is something of a religious issue; I like both, each > > in their own way, having no vested interest in either. > > > > > > Joe Gwinn --------------------------- Newsgroups: comp.os.vxworks Subject: Exception Handler Question... [excVecSet() vs excHookAdd() on a 8260] Date: Sun, 15 Jul 2001 06:07:33 GMT From: Todd Sandor Organization: @home Message-ID: <3B51368C.7832D79F@home.com> Hi: I'm looking for help - I'm having trouble getting excHookAdd() to work on a 8260 (ppc) based system. I'm trying to perform some additional operations when a [task] exception [machine check] occurs. I'ld prefer not to replace the existing default exception handler (excIntHandler), unless I can determine how to get information normally displayed by the default exception hander via the exception stack frame pointer [Exception next instruction address, Task name, etc]. I've been able to successfully replace the machine check exception handler via the call: excVecSet((FUNCPTR *)_EXC_OFF_MACH, FUNCREF(myhdlerEh)); When a machine check exception occurs, "myhdlerEh" is called and is passed a pointer to the exception stack frame. This is OK, but I don't know how I can get the taskId/name and next instruction address with the stack frame pointer - does anyone have any hints? The method I'ld prefer [or I think I would prefer] to use is the excHookAdd() method that is suppose to call the supplied routine after normal exception handling is performed. I want to do this for machine check exceptions and possibly data access exception - although I don't really know the difference between a machine check and data access exception on a 8260, anyone? The supplied routine is called with the taskId, the exception vector number and the stack frame pointer as parameters. When I try to use excHookAdd() method, via the call: excHookAdd((FUNCPTR)handler); the "hander" functions isn't called when an exception occurs - I'm not sure why? When I check the address pointed to by the "_func_excIntHook" variable, it is 0 after the excHookAdd() call - I believe this is suppose to be set after the excHookAdd() call is made. If I set the value of "_func_excIntHook" to the address of my "hander" function and then generate an exception, my hander function still doesn't get called. The problem could either I''m not doing something correctly/not-using-the-correct-approach or the default exception hander "excIntHander()" is not working as documented. [if I knew where the code was, I look at it, but I couldn't find it anywhere]. If anyone can give me any hints, it would be appreciated...Cheers... --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Sun, 15 Jul 2001 10:14:08 +0300 From: Paul Keinanen Organization: SAUNALAHDEN asiakas Message-ID: References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> On Sat, 14 Jul 2001 22:51:45 GMT, joegwinn@mediaone.net (Joe Gwinn) wrote: >One thing that may not have been clear in the other postings is the true >distinction between kernel mode and user mode. It generally has nothing >to do with memory and address spaces, the difference is in the machine >instructions that are permitted. In kernel mode, all instructions are >permitted, no matter how dangerous. In user mode, such instructions are >forbidden, causing an illegal-instruction trap if attempted. Some architectures use separate stack pointers for user mode and privilege mode, so you do not need to allocate space for interrupts in each stack for each user mode task. With on-chip memory management units, the processor mode usually affects what mapping registers are used. Even with a simple one to one logical to physical address mapping, if the application code is executed from RAM (i.e. it is loaded from a disk or flash) it can be a good idea to set the application code (and some system regions) to readonly in user mode. Thus, the kernel mode can load applications into RAM, but the application can not overwrite the code, making it easier to trap such errors during development and also increase the likelihood of a successful warm restart. Of course, such protection is not needed, if the application code is executed directly from (E)PROMs:-). Paul --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 16 04:03:29 2001 From: Vxworks Exploder Date: Mon Jul 16 04:03:31 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 16 04:03:07 PDT 2001 Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: Does VxWorks use User Mode & Kernel Mode? Subject: Re: help required on linker configuration on tornado Subject: Problem with Connect and select Subject: how to redirect pseudo terminal output? Subject: Re: Kermit for VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: 15 Jul 2001 06:14:50 -0700 From: Chris Torek Organization: None of The Above Message-ID: <9is50a$jpi$1@elf.eng.bsdi.com> References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> Sender: torek@elf.eng.bsdi.com In addition to everything that has been said so far, I will add a few general remarks. Ultimately, "modes" are really all about privileges. This may mean instruction privileges -- such as whether the "halt" instruction really halts, or traps instead to a monitor -- or addressing privileges or access privileges or any other set of privileges you can think of. There is a sort of "wheel of reincarnation" in the computing world, in which big, heavy :-) , expensive systems like supercomputers or giant multi-user mainframe IBM designers invent some technology such as "multiprocessors" or "superscalar execution" or "Thomasulo scheduling" or "RISC" (though they did not call it that) or whatever, and then over the years it all eventually trickles down into the $1.59 microchips, in some form or another. Modes -- user and kernel, or user + executive + supervisor + kernel, or "capabilities", or whatnot -- have been around since at least the 1960s. The exact implementations vary considerably, so one always has to consult the documentation for the CPU(s) in question; but ultimately they always boil down to privilege: there is something some mode can do that another mode cannot. These modes do not *have* to be arranged hierarchically (for instance, one could have a CPU in which I/O instructions are allowed only when the CPU is in "I/O mode" and halts are allowed only when it is in "halt mode" and so on), but hierarchies are traditional, and are easy to work with and to think about. The simplest case is the traditional "user vs kernel", completely unprivileged vs completely privileged. As soon as you commit to such a division, you can (if the hardware is done right) acquire a new ability, the "virtual machine". If privileged instructions in user mode always "trap safely", and if any instruction that reads the current mode is considered privileged, one can write a small kernel -- often called a "nucleus" in the old days -- that catches such traps and emulates the instruction safely, making this user code *think* it is kernel code. An emulated "read mode" instruction says "kernel", and an emulated "do I/O or memory-mapped I/O" instruction talks to emulated hardware. This is a wonderful debugging trick. Unfortunately, it tends to be slow. Indeed, this is an Achilles Heel of modes of any kind: they tend to be slow. Traps themselves are usually slow: on modern CPUs they either require pipeline flushes, or hairy cleanup code to figure out where the trap really occurred and what instructions were partway through the pipeline and what state they should get on restart/resume or abort. Moreover, if any particular operation or instruction may be off-limits to any particular mode, some extra check must occur every time that instruction is executed. Finally, to use these modes, the system has to do some kind of mode-switching operation. Depending on the CPU and desired effect (such as memory protection), this can require things like TLB flushes, which may cost several hundred or even thousand clock cycles when you add all the subsequent refill penalties. Processor that are designed up front with modes in mind can avoid (or at least mitigate) many of these problems. For instance, each TLB entry can be marked with an "address space tag", so that changing address spaces (e.g., user <-> kernel, or userA <-> userB in more finely divided systems) does not require a TLB flush after all. Alternatively, the system can give up the protections such divisions provide. In the end, you have to make some tradeoffs. Exactly which features you want is up to you; their price depends heavily on the CPU. - -- In-Real-Life: Chris Torek, Wind River Systems El Cerrito, CA, USA Domain: torek@bsdi.com +1 510 234 3167 http://claw.eng.bsdi.com/torek/ (not always up) I report spam to abuse@. --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Does VxWorks use User Mode & Kernel Mode? Date: Sun, 15 Jul 2001 15:58:21 GMT From: joegwinn@mediaone.net (Joe Gwinn) Organization: Gwinn Instruments Message-ID: References: <9ila0j$k005n$1@ID-92003.news.dfncis.de> <9imv3j$js1sj$1@ID-92003.news.dfncis.de> In article , Paul Keinanen wrote: > On Sat, 14 Jul 2001 22:51:45 GMT, joegwinn@mediaone.net (Joe Gwinn) > wrote: > > > >One thing that may not have been clear in the other postings is the true > >distinction between kernel mode and user mode. It generally has nothing > >to do with memory and address spaces, the difference is in the machine > >instructions that are permitted. In kernel mode, all instructions are > >permitted, no matter how dangerous. In user mode, such instructions are > >forbidden, causing an illegal-instruction trap if attempted. > > Some architectures use separate stack pointers for user mode and > privilege mode, so you do not need to allocate space for interrupts in > each stack for each user mode task. True. > With on-chip memory management units, the processor mode usually > affects what mapping registers are used. Even with a simple one to one > logical to physical address mapping, if the application code is > executed from RAM (i.e. it is loaded from a disk or flash) it can be a > good idea to set the application code (and some system regions) to > readonly in user mode. Thus, the kernel mode can load applications > into RAM, but the application can not overwrite the code, making it > easier to trap such errors during development and also increase the > likelihood of a successful warm restart. Of course, such protection is > not needed, if the application code is executed directly from > (E)PROMs:-). While it's true that the hardware usually works or can work as you describe, it's traditional for RTOSes to 1:1 map everything, and thus not use the memory mapping unit to create memory spaces. An RTOS need not make use of all that a hardware mode permits. The standard justifications are simplicity and speed. As Chris Torek points out, modes are basically defined by their privileges, but CPUs differ greatly in their details. Joe Gwinn --------------------------- Newsgroups: comp.os.vxworks Subject: Re: help required on linker configuration on tornado Date: Sun, 15 Jul 2001 22:11:13 +0200 From: "Juergen Sievers" Organization: PPP Internetdienstleistungen GmbH Message-ID: References: <9hd7rm$4hs$1@overload.lbl.gov> <3B3AE355.96503767@wipro.com> May be it is helpfully :) Usage: ar [-X32_64] [-]{dmpqrstx}[abcfilNoPsSuvV] [member-name] [count] archive- file file... ar -M [ "Vijay Kumar Peshkar" schrieb im Newsbeitrag news:3B3AE355.96503767@wipro.com... > Sujeet, > > Did you look at MACH_EXTRA option?. Guess it should help. > > regards, > Vijay > > Sujeet Naik wrote: > > > Hi everyone, > > I have perhaps a strange requirement. I have a project with 4 files viz., > > f1.c, f2.c, f3.c and f4.c (all are C source files). When I make the > > dependencies and build, I compile all these four files and create an archive > > MYPROJ.a. This is later linked with other archives to create a final binary. > > > > My requirement is, I want to force the functions from f2.c to start from a > > fixed location, while the other files should be linked and located > > contiguously with other archive files. For this, I have my linker directive > > (.ld file) forcing f2.o to start from a fixed address. The problem is, when > > I am doing this, I am not supposed to have f2.o as part of MYPROJ.a. > > Otherwise, I will have multiple instances of all the functions in f2.c since > > the .ld file already has forced f2.o from a fixed location. > > > > Can someone help me on deciding how to eliminate f2.o from MYPROJ.a. I cant > > manually go and edit makefile of this project since it is dynamically > > created. > > > > Also, if this cant be done, is there any other way I can achieve my goal. > > > > Thanks in advance, > > > > Sujeet > --------------------------- Newsgroups: comp.os.vxworks Subject: Problem with Connect and select Date: 15 Jul 2001 20:21:47 -0700 From: tomar27@msn.com (Mahesh Tomar) Organization: http://groups.google.com/ Message-ID: <4647d063.0107151921.569fee68@posting.google.com> Hi, I'm facing a strange problem with socket system calls i.e. connect and select. Before starting the server, i try to start my client, which try to connect to this server(which is not running at this moment) in non blocking mode. connect returns immediatly with EINPROGRESS err_code, then i make a select call, where i specify a 2 minutes time to connect to the server, strangly the select calls returns immediatly with no error, but when i try to write on that socket, it gives me EPIPE error, i don't understand why select doesn't wait for 2 minutes and try to connect to the server. Could somebody explain me why doesn't select wait for specified time. But if i start my server before starting my client, everything goes fine. Can somebody throw some light on it. Thanks MT --------------------------- Newsgroups: comp.os.vxworks Subject: how to redirect pseudo terminal output? Date: Mon, 16 Jul 2001 16:03:18 +0900 From: "ozlab" Organization: A poorly-installed InterNetNews site Message-ID: <9iu39q$mce$1@news.kornet.net> My source as followings: int fd; /* fd to local pseudo-terminal */ fd = ioGlobalStdGet (STD_OUT); if ( (status = ioctl (fd, FIOGETNAME, &ptyFd)) == TRUE ) { ioOptions = OPT_7_BIT | OPT_ABORT | OPT_TANDEM | OPT_LINE; /* disable echo output to pty */ (void) ioctl (ptyFd, FIOOPTIONS, ioOptions); } So far I cannot understood 'pty' well enough, I think that above works but not. I want to know how to redirect psuedo terminal output. ====================================== Eung-Shik Kim (Ed) ed@ebridgecom.com PowerVoIP Gateway Hardware Devl. Dept. 467-12 Dogok-Dong Kangnam-Gu MMAA Bldg 29th Fl. Seoul (zip 135-270) KOREA, Republic of Tel+82 2 575 6476 ext 517, 3713 Fax+82 2 575 6478 C.P. +82 11 9704 0840 ICQ#61658172 / Send me plain text please... ======================================= --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Mon, 16 Jul 2001 07:45:40 -0000 From: caf@omen.com (Chuck Forsberg WA7KGX) Organization: Omen Technology INC Message-ID: References: <9ikjrq$asm@watsun.cc.columbia.edu> In article <9ikjrq$asm@watsun.cc.columbia.edu>, Frank da Cruz wrote: > >From time to time, we get requests for Kermit file transfer software >for VxWorks. I'd like to be able satisfy them. What many of these embedded applications require is a memory efficient implementation devoid of Unix-isms that gives excellent results with software the customer has already paid for (i.e., the Hyperterm that comes with Windows). Industrial ZMODEM has been used with chips as primitive as the 8-bit 8051 family. It does not need an operating system, just a small number of user supplied routines. For undemanding applications, an "Industrial Kermit" could be developed, subject to the limitations of non-Columbia Kermit implementations as described in Kermit News. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 16 04:48:51 2001 From: rsrinivasarao@123india.com Date: Mon Jul 16 04:48:53 PDT 2001 Subject: Boot Image Hi, 1 How do I create bootrom_uncrom floppy image for PC486 platfrom in vxworks from an unix shell. Can any one tell steps I need to follow for to build boot image. 2 Can I use command "fdimage bootrom_umcmp a:" to crate bootable floppy from windows dos prompt. Thanks & Regards Srini ______________________________________________________ 123India.com - India's Premier Portal Get your Free Email Account at http://www.123india.com From vxwexplo-errs@csg.lbl.gov Mon Jul 16 09:24:34 2001 From: Chris James Date: Mon Jul 16 09:24:37 PDT 2001 Subject: Booting Tornado II on an MVME2431 through use of the PPC4-Bug "romboot" mechanism Has anyone successfully configured an MVME2431 to boot vxWorks 5.4 (Tornado II) from flash using the PPC4-BUG "romboot" facility? I seem to be having problems with the PCI autoconfiguration process conflicting with PPC4-BUG's initial setup of the PCI bus. David. C. James@Boeing.com From vxwexplo-errs@csg.lbl.gov Tue Jul 17 04:03:38 2001 From: Vxworks Exploder Date: Tue Jul 17 04:03:40 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 17 04:03:17 PDT 2001 Subject: Re: Kermit for VxWorks Subject: Boot Image Subject: Re: Is Wind River really this pathetic Subject: Re: ROM based vxworks build with T2 Project: any success ? Subject: Re: Kermit for VxWorks Subject: Re: Kermit for VxWorks Subject: Booting Tornado II on an MVME2431 through use of the PPC4-Bug "romboot" mechanism Subject: Logging to file from Vxsim Subject: Re: Exception Handler Question... [excVecSet() vs excHookAdd() on a 8260] Subject: location of the BSS text and the Data segment Subject: tornado tools - dependency tracking Subject: Re: Browser Subject: Re: Browser Subject: 82378 timer on mvme1600 (mv1604) Subject: Problems with setjmp.h in 5.4 Subject: Re: Kermit for VxWorks Subject: Re: tornado tools - dependency tracking Subject: missing functions Subject: Re: Kermit for VxWorks Subject: PPC 8240 and Power Management in C Subject: Newbie Subject: Link errors in the build window Subject: Recommend a quad serial I/O PMC Subject: Re: Is Wind River really this pathetic Subject: Re: Newbie ------------------------------------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Mon, 16 Jul 2001 14:05:07 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995285625.348908@newsmaster-04.atnet.at> References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> <8_H37.5136$z21.527498@newsc.telia.net> > It doesn't say anything about distributing source for everything else in > the package accompanying the GPL'd work. I don't think the discussion was > about adding gkermit to the VxWorks package (though I don't see anything > wrong with that either), but merely about trying to use a kermit protocol > engine to do file transfers to/from VxWorks. Depends on the point of view. Usually, you compile everything into one big binary and burn that beast to Flash or whatever starage the device has. So you would in fact add the gkermit sources to your VxWorks image. VxWorks has no concept of processes and executables stored in files out of the box. You can load object modules at runtime, but they are linked against the target too. I don't think that the GPL handles that particular case very good. regards werner --------------------------- Newsgroups: comp.os.vxworks Subject: Boot Image Date: 16 Jul 2001 04:48:22 -0700 From: rsrinivasarao@123india.com Organization: Lawrence Berkeley National Laboratory Message-ID: <9iuls9$3fk$1@overload.lbl.gov> Hi, 1 How do I create bootrom_uncrom floppy image for PC486 platfrom in vxworks from an unix shell. Can any one tell steps I need to follow for to build boot image. 2 Can I use command "fdimage bootrom_umcmp a:" to crate bootable floppy from windows dos prompt. Thanks & Regards Srini ______________________________________________________ 123India.com - India's Premier Portal Get your Free Email Account at http://www.123india.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Mon, 16 Jul 2001 13:21:11 GMT From: Lee DeRaud Organization: None to speak of Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> Sender: nntp@news.boeing.com (Boeing NNTP News Access) On Sat, 14 Jul 2001 10:36:48 -0600, "Joe Woodbury" wrote: >In the vain hope someone who actually matters a damn at Wind River monitors >this forum, I have some simple questions: > >Beyond that, why would you spent money to develop >something you don't even own? All things considered, that's better than the alternative: buying a product line and then discontinuing support for it. And yes, they've done that too. Lee --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ROM based vxworks build with T2 Project: any success ? Date: Mon, 16 Jul 2001 13:52:06 GMT From: Diego Serafin Organization: http://www.newsranger.com Message-ID: References: <7Ix37.17948$Kf3.233634@www.newsranger.com> <7b1623b6.0107131103.410c599@posting.google.com> <3B4F4980.946C2F3F@fnal.gov> Thank You Guys, waiting for WRS support to send me the corrected elfToBin I wrote a small C filter program to snip off the 0x100 bytes header and modified rules.vxWorks. I did not think it could be possible there was such a bug in elfToBin and nobody at WRS cared to send in a patch for it ... Bye, Ciao. Diego. In article <3B4F4980.946C2F3F@fnal.gov>, Dinker Charak says... > >> > I saw that vxworks module, stored at (rom address of) binArrayStart does not >> > start with code, but with .ELF header so that when romStart jumps to >> > RAM_DST_ADRS it fetches rubbish (...".ELF"). The real usrEntry code sits at >> > 0x100 bytes from RAM_DST_ADRS. > >i faced the same problem ... i used to snip off the header off the image >... --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Mon, 16 Jul 2001 14:07:17 GMT From: Adam Dunkels Organization: Telia Internet Message-ID: References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> <8_H37.5136$z21.527498@newsc.telia.net> Followup-To: comp.os.vxworks fred smith wrote: > : But gkermit is released under the GPL, which means that you cannot > : distribute a product that includes gkermit unless you are willing to > : release and distibute all source code for your project, including the > : source code to VxWorks. I don't know much about VxWorks, but I assume > : that it is not free software licensed under the GPL. > > Er, no, I don't think that's what the GPL says. > If you distribute a GPL'd work you must make its source available. > If you distribute a work DERIVED FROM a GPL'd work you must make its > source available. > It doesn't say anything about distributing source for everything else in > the package accompanying the GPL'd work. I don't think the discussion was > about adding gkermit to the VxWorks package (though I don't see anything > wrong with that either), but merely about trying to use a kermit protocol > engine to do file transfers to/from VxWorks. From the GPL FAQ : You have a GPL'ed program that I'd like to link with my code to build a proprietary program. Does the fact that I link with your program mean I have to GPL my program? Yes. /adam - -- Adam Dunkels (spambait) http://dunkels.com/adam/ --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Mon, 16 Jul 2001 14:34:04 GMT From: CBFalconer Organization: Ched Research Message-ID: <3B52EB7D.2238FE2B@my-deja.com> References: <9ikjrq$asm@watsun.cc.columbia.edu> Reply-To: cbfalconer@worldnet.att.net Chuck Forsberg WA7KGX wrote: > > In article <9ikjrq$asm@watsun.cc.columbia.edu>, Frank da Cruz wrote: > > > >From time to time, we get requests for Kermit file transfer software > >for VxWorks. I'd like to be able satisfy them. > > What many of these embedded applications require is a memory > efficient implementation devoid of Unix-isms that gives > excellent results with software the customer has already > paid for (i.e., the Hyperterm that comes with Windows). > > Industrial ZMODEM has been used with chips as primitive as > the 8-bit 8051 family. It does not need an operating > system, just a small number of user supplied routines. Some years ago ZMODEM was the standard file transfer protocol on Fido and BBS systems, and quite a few other places. I don't know what underlies FTP protocol, but I have the impression that it is less efficient than ZMODEM. I believe you have published the ZMODEM specifications, but have you ever published complete sources for both receive and transmit modules, and if so how portable are they? Can the system be mapped into the streaming capabilities of the C file system, or do they require char by char hardware interfaces? - -- Chuck F (cbfalconer@my-deja.com) (cbfalconer@XXXXworldnet.att.net) (Remove "XXXX" from reply address. my-deja works unmodified) mailto:uce@ftc.gov (for spambots to harvest) --------------------------- Newsgroups: comp.os.vxworks Subject: Booting Tornado II on an MVME2431 through use of the PPC4-Bug "romboot" mechanism Date: Mon, 16 Jul 2001 09:22:44 -0700 From: Chris James Organization: Lawrence Berkeley National Laboratory Message-ID: <3B5314D4.EF62D70E@sim.ssd.bna.boeing.com> Has anyone successfully configured an MVME2431 to boot vxWorks 5.4 (Tornado II) from flash using the PPC4-BUG "romboot" facility? I seem to be having problems with the PCI autoconfiguration process conflicting with PPC4-BUG's initial setup of the PCI bus. David. C. James@Boeing.com --------------------------- Newsgroups: comp.os.vxworks Subject: Logging to file from Vxsim Date: Mon, 16 Jul 2001 14:11:24 -0500 From: Johnny Schmittou Organization: Alcatel USA Message-ID: <3B533C5C.E51F8D38@usa.alcatel.com> Hi: We're using Tornado 2.0 and VxWorks 5.4 mostly in a simulation environment right now - thus we're currently making heavy use of Vxsim. We develop on Sun/Solaris/CDE/Sparc WSs and thus run Tornado/Vxsim in that environment for now. My question is, for every xterm I use, I can use a key combination of control-Mouse1 and select the "Log to File" option. xterm puts a check mark beside the option and from then on, everything output to that window is captured into a file in my root directory of the format: XtermLog. In Vxsim, I can do control-Mouse1 and try to select the "Log to File" option, but it never lets me select the option and never logs anything to a file. How does one get the "Log to File" to work under Vxsim? I would imagine there's a file some place w/ Xterm resource settings, perhaps similar to one's own ~/.Xdefaults. Thanks, Johnny If you see the word "cheese" in my e-mail address, please replace that word with "usa" in order to send me an e-mail reply. - -- ______________"I always tell the truth except for when I lie."_____________ Internet....: Johnny.Schmittou@cheese.alcatel.com Real life...: Johnny Schmittou Virtual life: Q-man ______I'd rather be drumming.______ Dallas, Texas USA --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Exception Handler Question... [excVecSet() vs excHookAdd() on a 8260] Date: Mon, 16 Jul 2001 19:28:32 GMT From: Dan Gold Organization: http://www.newsranger.com Message-ID: References: <3B51368C.7832D79F@home.com> In article <3B51368C.7832D79F@home.com>, Todd Sandor says... > >Hi: >I'm looking for help - I'm having trouble getting excHookAdd() to work >on a 8260 (ppc) based system. > >I'm trying to perform some additional operations when a [task] exception >[machine check] occurs. I'ld prefer not to replace the existing default >exception handler (excIntHandler), unless I can determine how to get >information normally displayed by the default exception hander via the >exception stack frame pointer [Exception next instruction address, Task >name, etc]. It depends on the particular exception. External interrupt (0x500) and the decrementer (0x900) call excIntHandle(). The others (i.e. machine check, dsi, etc..) call excExcHandle(). >The method I'ld prefer [or I think I would prefer] to use is the >excHookAdd() method that is suppose to call the supplied routine after >normal exception handling is performed. I want to do this for machine >check exceptions and possibly data access exception - although I don't >really know the difference between a machine check and data access >exception on a 8260, anyone? The supplied routine is called with the >taskId, the exception vector number and the stack frame pointer as >parameters. Machine check can be caused by multiple things, i.e. TEA, etc... data access is usually called in conjunction with the MMU. Either there's no translation for the desired address, the write-protect bits were set on the page, user/supervisor access, etc... >When I try to use excHookAdd() method, via the call: >excHookAdd((FUNCPTR)handler); >the "hander" functions isn't called when an exception occurs - I'm not >sure why? >When I check the address pointed to by the "_func_excIntHook" variable, >it is 0 after the excHookAdd() call - I believe this is suppose to be >set after the excHookAdd() call is made. Because _func_excIntHook is checked in excIntHandle(). Try modifying _func_excBaseHook instead. >If anyone can give me any hints, it would be appreciated...Cheers... hth, Dan Gold gold@ensemble.com --------------------------- Newsgroups: comp.os.vxworks Subject: location of the BSS text and the Data segment Date: Mon, 16 Jul 2001 22:08:21 +0300 From: "Alon Horesh" Organization: NetVision Israel Message-ID: <9ive08$isf$1@news.netvision.net.il> hello VxWorkers can someone please tell me how do i set the BSS Text and the Data segment location on the PowerPC ?? Where can i configure the address of the segments ?? Thank you very much , Alon . --------------------------- Newsgroups: comp.os.vxworks Subject: tornado tools - dependency tracking Date: Mon, 16 Jul 2001 16:08:09 -0400 From: "Kevin Livingston" Message-ID: <2OH47.21$s52.1282@client> Hello, I was wondering if anyone else was having trouble with the dependency tracking feature of the Tornado tools. (I am using Tornado 2.0.2 with patchs DosFS, WindML(and patch), Personal JWorks.) When I add files to a project, sometimes some of the dependencies are missing. Then if I remove a file from a project and re-add it etc. I frequently loose some of the dependencies that were originally present. The dependencies that I loose are included from a directory exteral to the project directory (eg. shared source with another project etc.) That directory is included on my compile line with the -I option. I also have it listed with the VPATH macro. WRS recomended reinstalling Tornado to correct the problem, I have not attempted this yet because it is not a short process with all the patches listed above. I have also been able to replicate the problem on a coworker's machine, so I am not convinced that a reinstall will have any affect. Does anyone else experience anything like this? or have a potential workaround for such behavior? thanks Kevin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Browser Date: 16 Jul 2001 20:34:03 GMT From: Richard Gibbons Message-ID: References: <9ikkiu$vhi$1@localhost.localdomain> Hi Andy, > We want to use a browser on VxWorks - ideally HTML 4.0 with applet > support - any user experience? We have been going through some pain with this one. We initially looked for a native solution but did not find any good native browsers for VxWorks - they typicaly lack HTML 4.0 and applet support. OIur application required access to internally generated content plus full web access. We did look at the ICE Java browser, but we found it unstable on our platform. We did not seem to get good support and new updates on the product. It seems that Wind have gone very quiet on this now - does anyone know what is happening with this product? We are currently evaluating the Espial Escape browser - Wind River badged this product before they purchased ICE and it seems pretty good so far. I would be interested to hear if others have had the same experience. Richard --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Browser Date: Mon, 16 Jul 2001 17:25:09 -0500 From: John McClenny Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <9ikkiu$vhi$1@localhost.localdomain> In article , gibb49@yahoo.com says... > Hi Andy, > > > We want to use a browser on VxWorks - ideally HTML 4.0 with applet > > support - any user experience? > > We have been going through some pain with this one. We initially looked for > a native solution but did not find any good native browsers for VxWorks - > they typicaly lack HTML 4.0 and applet support. OIur application required > access to internally generated content plus full web access. > > We did look at the ICE Java browser, but we found it unstable on our > platform. We did not seem to get good support and new updates on the > product. It seems that Wind have gone very quiet on this now - does anyone > know what is happening with this product? > > We are currently evaluating the Espial Escape browser - Wind River badged > this product before they purchased ICE and it seems pretty good so far. We went through both the Espial and ICE browsers before giving up on the PJava path. Most of the problems seemed to be PJava related and not necessarily the browser. Both ran fine on a PC JVM. In either case, the performance was going to suck compared to a native browser and was not going to be acceptable in our application. If you have lower requirements or tons of CPU speed, YMMV. We are now porting the ANT browser ( www.antlimited.com ). We just started this week and have the basic graphics up and are working on the socket code. So far, it looks good. The documentation is ok and the source code looks clean and interfaces are well defined. I will post a followup in a couple of weeks when we have the basic browser working over the Internet. john mc --------------------------- Newsgroups: comp.os.vxworks Subject: 82378 timer on mvme1600 (mv1604) Date: Tue, 17 Jul 2001 00:05:07 GMT From: pat.dirienzo@verizon.net Message-ID: <3B538277.AF0A6F7B@verizon.net> Reply-To: pat.dirienzo@verizon.net Has anyone written a driver for the 82378 timer on the mvme1600 board (they might share) WindRiver's mv1604 bsp does not provide one for this timer. They do provide one generic driver for the 8254 timer chip (p/o 82378), but it must be modified for this bsp. We want to use this timer because it has the highest interrupt priority of all timers on the board. Another answer to our problem might be to change the z8536 timer priority (at the 82378) to a higher priority and use it, but I am uncertain how or if this can be done. any help here is appreciated thanks Pat D. --------------------------- Newsgroups: comp.os.vxworks Subject: Problems with setjmp.h in 5.4 Date: Tue, 17 Jul 2001 02:05:20 GMT From: "David Beck" Organization: RoadRunner - Central Florida Message-ID: It seems that the setjmp.h header file requires the inclusion of vxWorks.h. This is certainly non-ANSI, has anyone else been able to use the setjmp facility without resorting to including other header files? David Beck --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Tue, 17 Jul 2001 03:43:17 -0000 From: caf@omen.com (Chuck Forsberg WA7KGX) Organization: Omen Technology INC Message-ID: References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B52EB7D.2238FE2B@my-deja.com> In article <3B52EB7D.2238FE2B@my-deja.com>, CBFalconer wrote: >I believe you have published the ZMODEM specifications, but have >you ever published complete sources for both receive and transmit >modules, and if so how portable are they? Can the system be >mapped into the streaming capabilities of the C file system, or do >they require char by char hardware interfaces? I wrote Indusrtrial ZMODEM(Tm) to support embedded applications where the data source or sink may be a region of memory or even a process. The only limit is the engineer's imagination. It is possible to cobble something together from legacy rzsz 1.xx versions and a few man months of hacking, testing, and debugging. Some engineers have decided the modest one time instant licensing for Industrial ZMODEM is better suited to their needs. Check http://www.omen.com/ind.html for the details. - -- Chuck Forsberg WA7KGX N2469R caf@omen.COM www.omen.com Widget need a Brain Drain? Check out Industrial ZMODEM(Tm) Omen Technology Inc "The High Reliability Software" POB 4681 Portland OR 97208 503-614-0430 FAX:503-629-0665 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tornado tools - dependency tracking Date: 16 Jul 2001 22:43:14 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107162143.6f4bd373@posting.google.com> References: <2OH47.21$s52.1282@client> We met the same problem before, you can change the include definition from "#inlcude "xxx\xxx\xxx.h" to "#include "xxx/xxx/xxx.h" ,then test it again. Wish it helpful. ellin "Kevin Livingston" wrote in message news:<2OH47.21$s52.1282@client>... > Hello, > > I was wondering if anyone else was having trouble with the dependency > tracking feature of the Tornado tools. (I am using Tornado 2.0.2 with > patchs DosFS, WindML(and patch), Personal JWorks.) > > When I add files to a project, sometimes some of the dependencies are > missing. Then if I remove a file from a project and re-add it etc. I > frequently loose some of the dependencies that were originally present. The > dependencies that I loose are included from a directory exteral to the > project directory (eg. shared source with another project etc.) That > directory is included on my compile line with the -I option. I also have it > listed with the VPATH macro. > > WRS recomended reinstalling Tornado to correct the problem, I have not > attempted this yet because it is not a short process with all the patches > listed above. I have also been able to replicate the problem on a > coworker's machine, so I am not convinced that a reinstall will have any > affect. > > Does anyone else experience anything like this? or have a potential > workaround for such behavior? > thanks > Kevin --------------------------- Newsgroups: comp.os.vxworks Subject: missing functions Date: Tue, 17 Jul 2001 07:09:25 GMT From: "Tri VU KHAC" Organization: KPNQwest customer news service Message-ID: Hi folks, I received a project from someone in my company. This project has been implemented in VxWorks for MIPS. The problem is when I download the compiled code on my target, it says that some functions are missing. These functions are sysIntLock(), sysIntUnlock(), and sysIntEnablePIC(). I can see the declarations of such functions in some header file of VxWorks, but not their definitions. Did anyone get the same problem ? Do you know what they are ? Can I replace them by the other functions which are documented by WindRiver ? Thank you in advance. Tri VU KHAC R&D software engineer. --------------------------- Newsgroups: comp.os.vxworks,comp.arch.embedded,comp.protocols.kermit.misc Subject: Re: Kermit for VxWorks Date: Mon, 16 Jul 2001 22:05:28 GMT From: fred smith Organization: None! Message-ID: References: <9ikjrq$asm@watsun.cc.columbia.edu> <3B4DDE9B.C12DEF3@digidata.com> <9ikojr$r9c$1@newsmaster.cc.columbia.edu> <8_H37.5136$z21.527498@newsc.telia.net> <995285625.348908@newsmaster-04.atnet.at> Sender: fredex@fcshome.stoneham.ma.us Werner Schiendl wrote: :> It doesn't say anything about distributing source for everything else in :> the package accompanying the GPL'd work. I don't think the discussion was :> about adding gkermit to the VxWorks package (though I don't see anything :> wrong with that either), but merely about trying to use a kermit protocol :> engine to do file transfers to/from VxWorks. : Depends on the point of view. Usually, you compile everything into one big : binary and burn that beast to Flash or whatever starage the device has. So : you would in fact add the gkermit sources to your VxWorks image. VxWorks has : no concept of processes and executables stored in files out of the box. You : can load object modules at runtime, but they are linked against the target : too. I don't think that the GPL handles that particular case very good. OK, obviously I was speaking out of ignorance. Not knowing anything about Vxworks I didn't realize it was an embedded system and that you couldn't use gkemrit on it WITHOUT having them all linked/distributed together. Open mouth, insert food, chew vigorously. :^( - -- - ---- Fred Smith -- fredex@fcshome.stoneham.ma.us ---------------------------- But God demonstrates his own love for us in this: While we were still sinners, Christ died for us. - ------------------------------- Romans 5:8 (niv) ------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: PPC 8240 and Power Management in C Date: Tue, 17 Jul 2001 12:22:16 +0200 From: internet Organization: keine Angabe Message-ID: <3B5411D7.8664BB61@internet.de> Hello, we use the PPC 8240 with vxWorks. It is possible to get an C example to use the Power Management of the PPC 8240. I will use the doze Mode , for example. regards Oliver Faessler --------------------------- Newsgroups: comp.os.vxworks Subject: Newbie Date: 17 Jul 2001 01:48:20 -0700 From: wys205@yahoo.com (Tom Jameson) Organization: http://groups.google.com/ Message-ID: <688f9159.0107170048.b15bcab@posting.google.com> Hi forks, I am a newbie in this field, and want to learn Vxworks. Tom --------------------------- Newsgroups: comp.os.vxworks Subject: Link errors in the build window Date: Tue, 17 Jul 2001 11:05:09 +0200 From: "Carsten Schwartz" Organization: TDC Internet Message-ID: <3b54008d$0$876$edfadb0f@dspool01.news.tele.dk> Reply-To: "Carsten Schwartz" I would like to see the 'Unresolved symbol errors' in the build window before I download to target from Tornado. How can you make the linking happen at build time on the Work Station.? Carsten Schwartz, Denmark --------------------------- Newsgroups: comp.os.vxworks Subject: Recommend a quad serial I/O PMC Date: Tue, 17 Jul 2001 11:34:35 +0100 From: "R Grommet" Organization: Defence Evaluation & Research Agency Message-ID: <9j102l$h72$1@trog.dera.gov.uk> Hi Can anyone recommend a good quad RS232/RS422 serial I/O PMC with driver support for T2 and the Motorola MVME5100? I'm having amazing problems with my current card and I'm wondering if there is a better one out there. TIA Ralph G --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Tue, 17 Jul 2001 12:10:13 +0200 From: Emmanuel Herbreteau Organization: SEPRO Robotique (BEP) Message-ID: <3B540F05.BE9B26EB@sepro-robotique.com> References: <6qH17.79$5f6.133962@news.uswest.net> Joe Woodbury wrote: > Why the hell did you buy the Diab compiler and yet continue dinking around > with GCC? Does anyone in your company actually have an MBA or even a > business degree? > Beyond that, why would you spent money to develop something you don't even own? Hi, GCC is a GNU software and is not related to WindRiver System. WRS only provide support for Tornado 2.0 and VxWorks. - -- Regards Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Newbie Date: 17 Jul 2001 10:18:06 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <688f9159.0107170048.b15bcab@posting.google.com> wys205@yahoo.com (Tom Jameson) wrote: > I am a newbie in this field, and want to learn Vxworks. Go to the site of WindRiver, click on Tornado Prototyper and download the SW. This will install a simulation environment and there you can learn a lot about VxWorks. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 17 10:46:32 2001 From: "Chang, Mark" Date: Tue Jul 17 10:46:34 PDT 2001 Subject: RTI tools Is there any possible way to have the RTI tools running and sampling, ready to go when a system powers up? It seems impossible since the TCP/IP connection is not ready and many other components are not up yet. But I shouldn't rule out the smallest possibility that someone has found a way to do so. I would like to see how my task/functions etc... are executing during power up. If there is a way without the RTI, I would like to know too. Even if it requires modifying the task/memory functions. Which is what RTI tools do anyway in real time. Thanks, Mark Vxworks 5.3.1 T1.0 ---------------------------------------------------------- Mark Chang Senior Engineer Network Management Broadband Access and Transport Group ADC Wireline Systems Division Mark_Chang@ADC.com (714)730-2827 Learn about ADC - The Broadband Company at www.adc.com From vxwexplo-errs@csg.lbl.gov Wed Jul 18 00:28:32 2001 From: Michael.Ben-Ari@ecitele.com Date: Wed Jul 18 00:28:35 PDT 2001 Subject: RE: VxWork & Proxy Arp >AFAIK, the vxWorks proxy ARP library does not check for the actual >Shared Memory driver, just for device_name == "sm" > >So try to rename Your Dual port RAM driver to "sm" - >that may do the trick. Thanks! Changing the Driver name to "sm" does the trick - sorta. I have the driver now failing upon proxyArpAttach(), but there is another issue involved here (when I figure out the problem/where it fails) that has nothing to do with the "Proxy ARP" driver from wind. According to someone else (who responded directly to me and not to the list), there is a patch that can be gotten from WRS that no longer requires "sm" to run Proxy ARP Driver. -Mike ----------------------------------------------- Mike Ben-Ari ECI Telecom - NGTS Ltd. Cellular Division - Celtro Mail: mbenari@ecitele.com Tel: +972-3-926-6794 Fax: +972-3-926-6460 ----------------------------------------------- From vxwexplo-errs@csg.lbl.gov Wed Jul 18 04:03:29 2001 From: Vxworks Exploder Date: Wed Jul 18 04:03:32 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 18 04:03:07 PDT 2001 Subject: Re: C++ base pointer access causes a crash Subject: Where to find most updated VxWorks FAQ? Subject: Re: Exception Handler Question... [excVecSet() vs excHookAdd() on a Subject: Re: Is Wind River really this pathetic Subject: how to use ld() to load c++ module with new or delete Subject: binding a network service Subject: How to Solove Memory leak problems? Subject: LongJmp Subject: Re: ROM based vxworks build with T2 Project: any success ? Subject: Re: ROM based vxworks build with T2 Project: any success ? Subject: Re: Where to find most updated VxWorks FAQ? Subject: Re: "File System" on VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks,comp.lang.c++ Subject: Re: C++ base pointer access causes a crash Date: 17 Jul 2001 17:45:57 -0700 From: davlet_panech@my-deja.com (davlet) Organization: http://groups.google.com/ Message-ID: <86b42096.0107171645.4547fbc0@posting.google.com> References: <655ac6b.0107171059.34d6d01a@posting.google.com> prakash_padmanabhan@yahoo.com (Prakash Padmanabhan) wrote in message news:<655ac6b.0107171059.34d6d01a@posting.google.com>... > Hello, > > I have a class Y which is derived from class X and implements one of class > X's virtual functions f. I've declared a static object of type Y and I'm > trying to call the virtual function using a pointer to X. Although the > address of the object (i.e, pointer value) is correct, the call crashes. > I find that the virtual function table has a null entry. > > The declarations look like this > > class X > { > public: > virtual void f(void); > }; > > class Y Surely you meant: class Y: public X { Other than that looks OK -- if it doesn't work blame the compiler. D.P. > { > public: > void f(void); > private: > int v1; > int v2; > }; > > The access is like > Y y; > X *pX = &y; > pX->f(); > > Does anyone have some idea as to what might be causing the crash? > > TIA, > Prakash > > P.S: I've munched the .o output file to a .out and have verified that > tha base and derived class constructors are getting called immediately > after the file is loaded. --------------------------- Newsgroups: comp.os.vxworks Subject: Where to find most updated VxWorks FAQ? Date: Wed, 18 Jul 2001 01:59:46 GMT From: "Kli88" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: thanks for the pointer, K --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Exception Handler Question... [excVecSet() vs excHookAdd() on a 8260] Date: Wed, 18 Jul 2001 02:08:19 GMT From: Todd Sandor Organization: @home Message-ID: <3B54F2F7.F0EADF7C@home.com> References: <3B51368C.7832D79F@home.com> It turned out that the excHookAdd() method was working all along. When I use this method and then forced a "task" have an exception the specified routine was called. My initial tests involved invoking commands/routines that performed an exception in the context of the vxWorks "shell" which has special characteristics [the shell is respawn'ed if an exception occurs]. Thanks for the responses... Cheers... Dan Gold wrote: > In article <3B51368C.7832D79F@home.com>, Todd Sandor says... > > > >Hi: > >I'm looking for help - I'm having trouble getting excHookAdd() to work > >on a 8260 (ppc) based system. > > > >I'm trying to perform some additional operations when a [task] exception > >[machine check] occurs. I'ld prefer not to replace the existing default > >exception handler (excIntHandler), unless I can determine how to get > >information normally displayed by the default exception hander via the > >exception stack frame pointer [Exception next instruction address, Task > >name, etc]. > > It depends on the particular exception. External interrupt (0x500) and the > decrementer (0x900) call excIntHandle(). The others (i.e. machine check, dsi, > etc..) call excExcHandle(). > > >The method I'ld prefer [or I think I would prefer] to use is the > >excHookAdd() method that is suppose to call the supplied routine after > >normal exception handling is performed. I want to do this for machine > >check exceptions and possibly data access exception - although I don't > >really know the difference between a machine check and data access > >exception on a 8260, anyone? The supplied routine is called with the > >taskId, the exception vector number and the stack frame pointer as > >parameters. > > Machine check can be caused by multiple things, i.e. TEA, etc... data access is > usually called in conjunction with the MMU. Either there's no translation for > the desired address, the write-protect bits were set on the page, > user/supervisor access, etc... > > >When I try to use excHookAdd() method, via the call: > >excHookAdd((FUNCPTR)handler); > >the "hander" functions isn't called when an exception occurs - I'm not > >sure why? > >When I check the address pointed to by the "_func_excIntHook" variable, > >it is 0 after the excHookAdd() call - I believe this is suppose to be > >set after the excHookAdd() call is made. > > Because _func_excIntHook is checked in excIntHandle(). Try modifying > _func_excBaseHook instead. > > >If anyone can give me any hints, it would be appreciated...Cheers... > > hth, > > Dan Gold gold@ensemble.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Tue, 17 Jul 2001 22:26:45 -0600 From: "Joe Woodbury" Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> I think you misunderstand. WindRiver ships GCC with Tornado. To do so, they make, and are continuing to make, modifications to it. All those changes fall under the GPL which means WindRiver is giving away it's labor and any trade secrets and/or proprietary algorithms it may develop. This makes no business sense, especially when they have an excellent product in the Diab compiler. Why not concentrate on making Diab even better? "Emmanuel Herbreteau" wrote in message news:3B540F05.BE9B26EB@sepro-robotique.com... > Joe Woodbury wrote: > > Why the hell did you buy the Diab compiler and yet continue dinking around > > with GCC? Does anyone in your company actually have an MBA or even a > > business degree? > > Beyond that, why would you spent money to develop something you don't even own? > > Hi, > > GCC is a GNU software and is not related to WindRiver System. > WRS only provide support for Tornado 2.0 and VxWorks. > > -- > Regards > Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: how to use ld() to load c++ module with new or delete Date: Wed, 18 Jul 2001 12:43:32 +0800 From: "Nixvue Systems" Organization: Singapore Telecommunications Ltd Message-ID: <9j345j$214$1@clematis.singnet.com.sg> Below is some cpp code that we want to use ld() function to load into the memory. ld(0,0,"projectcpp.out"); However it always give below error message: Undefined symbol: __builtin_vec_new (binding 1 type 0) Undefined symbol: __builtin_vec_delete (binding 1 type 0) Undefined symbol: __builtin_delete (binding 1 type 0) Undefined symbol: __builtin_new (binding 1 type 0) How do we go around this problem? If we only using c function, the ld() perform no problem. However if we code in cpp, the ld() always have above mention problem. - ---------------------------------------------------- /*projectcpp.cpp*/ /*after compile it produce projectcpp.out*/ #include "string.h" #include "iostream.h" #include "stdlib.h" #include "stdio.h" void realtest(void); void realtest(void) { int *i; int *j; cout << "testing start\n\n" << endl; i=new int; printErr("i address=0x%x\n",i); delete i; j=new int[100]; printErr("j address=0x%x\n",i); delete[] j; } extern "C" void testcpp(void) { printErr("test cpp load module\n\n\n"); realtest(); } --------------------------- Newsgroups: comp.os.vxworks Subject: binding a network service Date: Wed, 18 Jul 2001 07:48:34 +0200 From: "Pär Sjöholm" Message-ID: <3b55232c@grendel.df.lth.se> Hi, I am having trouble binding a network service to the motfcc END on a ADS8260 eval board. The 'VxWorks Network Protocol Toolkit' only describes what to do superficially (at least to me). Are there any help to be had from anywhere on the Internet (except WindRivers WindSurf)? I am able to use the muxTkBind(...) function. I bind it as MUX_PROTO_SNARF. I return FALSE from my receive routine, but it seems all pakets are eaten anyway, since I am not able to ping after I bind. I've found some references to a function generically refered to as fooAttach(). Is this something I have to worry about? How is this function invoked? /Paer Sjoeholm --------------------------- Newsgroups: comp.os.vxworks Subject: How to Solove Memory leak problems? Date: 17 Jul 2001 22:59:26 -0700 From: chwin@china.com (Adom) Organization: http://groups.google.com/ Message-ID: <92dbccc1.0107172159.c86d6b2@posting.google.com> Hi,All: i have developed a program in vxworks. Now it can run correctly .But i found there are memory leak in it . I have try all i can think,but it still is there. Can anybody tell me how to solove this problem? Any suggest is appreciate! Thanks! Adom chwin@china.com --------------------------- Newsgroups: comp.os.vxworks Subject: LongJmp Date: Wed, 18 Jul 2001 11:35:26 +0530 From: jp Organization: Lucent Technologies Message-ID: <3B552726.A8804859@lucent.com> Hi While downloading my application to vxSim, longjmp function is giving undefined reference But strangely, setjmp is not giving any problem. I have included both setjmp.h and vxworks.h. Is it bug in vxworks. Please help me as my application is using so many longjmp and setjmp calls. Regds gk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ROM based vxworks build with T2 Project: any success ? Date: 17 Jul 2001 23:53:55 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107172253.38868917@posting.google.com> References: <7Ix37.17948$Kf3.233634@www.newsranger.com> <7b1623b6.0107131103.410c599@posting.google.com> <3B4F4980.946C2F3F@fnal.gov> I created a vxworks_rom with a bootable project, but when I used elftobin to make it, I got fail message as below: seg1: Expected load address 0xfff00100 but file address is 0x003e20cc what's the matter? Thx. ellin Diego Serafin wrote in message news:... > Thank You Guys, > waiting for WRS support to send me the corrected elfToBin I wrote a small C > filter program to snip off the 0x100 bytes header and modified rules.vxWorks. > > I did not think it could be possible there was such a bug in elfToBin and nobody > at WRS cared to send in a patch for it ... > > Bye, > Ciao. > > Diego. > > > In article <3B4F4980.946C2F3F@fnal.gov>, Dinker Charak says... > > > >> > I saw that vxworks module, stored at (rom address of) binArrayStart does not > >> > start with code, but with .ELF header so that when romStart jumps to > >> > RAM_DST_ADRS it fetches rubbish (...".ELF"). The real usrEntry code sits at > >> > 0x100 bytes from RAM_DST_ADRS. > > > >i faced the same problem ... i used to snip off the header off the image > >... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ROM based vxworks build with T2 Project: any success ? Date: Wed, 18 Jul 2001 07:52:44 GMT From: Diego Serafin Organization: http://www.newsranger.com Message-ID: References: <7Ix37.17948$Kf3.233634@www.newsranger.com> <7b1623b6.0107131103.410c599@posting.google.com> <3B4F4980.946C2F3F@fnal.gov> <44a01eba.0107172253.38868917@posting.google.com> In article <44a01eba.0107172253.38868917@posting.google.com>, ellin says... > >I created a vxworks_rom with a bootable project, but when I used >elftobin to make it, I got fail message as below: >seg1: Expected load address 0xfff00100 but file address is 0x003e20cc >what's the matter? elfToBin for PPC is broken ! Your particular problem has already been reported in newsgroups some time ago. Yesterday I phoned WRS to have a functioning version... it seems that they did not know about the problem (?). Ciao, Diego. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Where to find most updated VxWorks FAQ? Date: 18 Jul 2001 08:07:09 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: "Kli88" wrote: > thanks for the pointer, Try this link: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: "File System" on VxWorks Date: Wed, 18 Jul 2001 10:24:49 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995445191.662263@newsmaster-04.atnet.at> References: <9in5ms$jtkbu$1@ID-92003.news.dfncis.de> VxWorks has basically the following file systems available - - RawFs a very basic file system, that treats the whole disk space as one big file, very little overhead but also very little features - - DosFs A 'normal' DOS compatible file system. The version that shipped with Tornado II did only support FAT-12 and FAT-16 partitions (thus, has a 2 GB partition size limit) and has hardly support for more than one partition. An upgrade (DosFs 2.0) is available, that supports FAT-32 (larger partitions, less wasted disk space due to smaller clusters) and better (still not ideal) support for partitions. Be warned, that DosFs is not that completely compatible with DOS and other systems as the name might suggest. E. g. you cannot access a partition created with VxWorks well with an other system. However, the other way round works. - - CDROM Fs Never tested that, reads ISO 9660 CDROMs (for obvious reasons, this will be a read only FS) - - TFFS An optional components to handle areas of flash space as disks. Depending on your hardware, this might be easier and/or cheaper than a real harddisk. If you already have an ATA controller, you can also use a ready-made flash card instead (one with ATA interface). - - Third party VxWorks supports third party file systems to some extent. Cannot provide more details on this issue, though. hth werner --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 18 05:00:50 2001 From: "Jeffrey A Angielski" Date: Wed Jul 18 05:00:52 PDT 2001 Subject: Porting InterNiche network stack to vxWorks Has anybody out there ported the InterNiche TCP/IP Protocol Stack to vxWorks? ------------------------------------------------------ Jeff Angielski The PTR Group, Inc. Embedded, Real-time Solutions and Services mailto:jeff@theptrgroup.com http://www.theptrgroup.com V:703.405.6896 ------------------------------------------------------ From vxwexplo-errs@csg.lbl.gov Thu Jul 19 03:05:11 2001 From: "ANIL KUMAR K.K." Date: Thu Jul 19 03:05:13 PDT 2001 Subject: doubt in vxsim networking Hi, Here in my system I set up networking in my vxworks simulator as said in http://www.xs4all.nl/~borkhuis/vxworks/multsimnt.html. I ran 2 simulators with IP addresses 90.0.0.1 and 90.0.0.2 and tried ping in between them inside my system and it fails. I am unable to ping within my system itself and how are u able to communicate across systems I have attached a copy of the mail which u sent to the mailing list thanks Anil From vxwexplo-errs@csg.lbl.gov Thu Jul 19 05:36:29 2001 From: "rajesh biswal" Date: Thu Jul 19 05:36:31 PDT 2001 Subject: VxWorks - Context Save This is a multi-part message in MIME format. --------------InterScan_NT_MIME_Boundary Content-Type: multipart/alternative; boundary="----=_NextPart_000_002A_01C1107D.7A12E6A0" ------=_NextPart_000_002A_01C1107D.7A12E6A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hi, 1) I want to know the name of the function which is called to saves/retrieves the context during a function call/return. 2) Also I want to know whether is it possible to get the name of the function a task is currently executing. Thanx, Rajesh Biswal. **************************************************************************** ************ Embedded & Internet Access Division Wipro Technologies Address : 30, Mission Road 1st Main, S.R Nagar Bangalore, 560027 Tel : 91-80-2275590/2241730 Extn 1145 Fax : 91-80-2241769 E-mail : rajesh.biswal@wipro.com Web Page : www.wipro.com The World's First SEI CMM Level 5 Software Services company **************************************************************************** ************ Silence and thinking never hurt anyone. ------=_NextPart_000_002A_01C1107D.7A12E6A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi,
1) I = want to know=20 the name of the function which is called to saves/retrieves the context = during a=20 function call/return.
2) = Also I want to=20 know whether is it possible to get the name of the function a task is=20 currently  executing.
 
Thanx,
Rajesh Biswal.
****************************************************************= ************************=20
Embedded & Internet Access Division
Wipro=20 Technologies
Address    : 30, Mission = Road
          &nbs= p;      =20 1st Main, S.R Nagar
          &nbs= p;      =20 Bangalore, 560027
Tel          &= nbsp; :=20 91-80-2275590/2241730 Extn 1145
Fax           = :=20 91-80-2241769
E-mail       : rajesh.biswal@wipro.com
Web = Page : www.wipro.com
 
The World's First SEI CMM Level 5 = Software Services=20 company
****************************************************************= ************************=20
 Silence and thinking never hurt = anyone.
 
------=_NextPart_000_002A_01C1107D.7A12E6A0-- --------------InterScan_NT_MIME_Boundary Content-Type: text/plain; name="Wipro_Disclaimer.txt" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="Wipro_Disclaimer.txt" The Information contained and transmitted by this E-MAIL is proprietary to Wipro Limited and is intended for use only by the individual or entity to which it is addressed, and may contain information that is privileged, confidential or exempt from disclosure under applicable law. If this is a forwarded message, the content of this E-MAIL may not have been sent with the authority of the Company. If you are not the intended recipient, an agent of the intended recipient or a person responsible for delivering the information to the named recipient, you are notified that any use, distribution, transmission, printing, copying or dissemination of this information in any way or in any manner is strictly prohibited. If you have received this communication in error, please delete this mail & notify us immediately at mailadmin@wipro.com --------------InterScan_NT_MIME_Boundary-- From vxwexplo-errs@csg.lbl.gov Fri Jul 20 01:25:17 2001 From: "Burkin, Valery" Date: Fri Jul 20 01:25:19 PDT 2001 Subject: EtherLink III ISA/PCI setup Hi, I've got problem configuring EtherLink III ISA or PCI cards under vxWorks. The basic target is Pentium PC... Actually I'm trying to get them up during boot stage in order to download kernel image from host via TFTP. config.h for ISA card: #define INCLUDE_NETWORK #define INCLUDE_ELT #undef IO_ADRS_ELT #define IO_ADRS_ELT 0x300 #undef INT_LVL_ELT #define INT_LVL_ELT 0x07 #undef CONFIG_ELT #define CONFIG_ELT 1 /* 0=EEPROM 1=AUI 2=BNC 3=RJ45 */ I/O base, IRQ level, transmiter type configured with DOS 3c5x9cfg. and tested under Win95/NT with those parameters. at the vxworks boot prompt: I try to get up elt0 interface with: "n elt" here's the output: Attaching network interface elt0... done. Error setting inet address of elt0 to 0, errno = 0 Cannot initialize interface named "elt0" in the boot line inet address: e=172.16.98.26, i.e. not null....... And...Oh, of course I cannot trace down this problem because of known reasons the only way is printf... Any help is welcome, valery. From vxwexplo-errs@csg.lbl.gov Fri Jul 20 01:27:49 2001 From: "AMASS" Date: Fri Jul 20 01:27:51 PDT 2001 Subject: Tornado project creation wizard This is a multi-part message in MIME format. ------=_NextPart_000_000E_01C110FD.90682840 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi All When using the Tornado project creation wizard, what rules/limitations = are to be applied when speicfying path names for the location of the = project? I seem to be encountering difficulties when using longer path = names (but well under the limit imposed by the host OS) - any help = appreciated. We are on a NT host. Cheers Andy vxworks ------=_NextPart_000_000E_01C110FD.90682840 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi All
 
When using the Tornado project creation = wizard,=20 what rules/limitations are to be applied when speicfying path names for = the=20 location of the project? I seem to be encountering difficulties when = using=20 longer path names (but well under the limit imposed by the host = OS) - any=20 help appreciated.
 
We are on a NT host.
 
Cheers
 
Andy
 
vxworks
------=_NextPart_000_000E_01C110FD.90682840-- From vxwexplo-errs@csg.lbl.gov Fri Jul 20 04:03:30 2001 From: Vxworks Exploder Date: Fri Jul 20 04:03:33 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 20 04:03:13 PDT 2001 Subject: Re: Vxsim very small fonts in output window Subject: Re: Vxsim very small fonts in output window Subject: Decent C++ compiler for Tornado2/VxWorks? Subject: Re: Vxsim very small fonts in output window Subject: Re: Vxsim very small fonts in output window Subject: Re: Vxsim very small fonts in output window Subject: Re: Vxsim very small fonts in output window Subject: Re: Vxsim very small fonts in output window Subject: Re: port to VxWorks Subject: Elf Executables on VxWork5.4 for I386 Subject: How to use Tornado 1.0 makefile in Tornado 2.0 IDE? Subject: Re: shut down VxSim window from script? Subject: Re: port to VxWorks Subject: Re: Works in shell, not in task Subject: Re: "File System" on VxWorks Subject: Re: LoadLib and unldLib Subject: LoadLib and unldLib Subject: Re: How to Solove Memory leak problems? Subject: Robotics Engineering Excellence && RTC (TCP/shared memory IPC package) Subject: PCI_AUTOCONFIG, does it work? Subject: file i/o w/multiple tasks Subject: Re: LoadLib and unldLib Subject: Re: file i/o w/multiple tasks Subject: Re: LongJmp Subject: Re: how to use ld() to load c++ module with new or delete Subject: using Java,C++,C mixture Subject: Zinc and setFocus() Subject: Re: "File System" on VxWorks Subject: Re: Works in shell, not in task Subject: simulator and objects Subject: Re: How to Solove Memory leak problems? Subject: Re: simulator and objects Subject: Re: using Java,C++,C mixture Subject: Re: Works in shell, not in task Subject: simulator again ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 15:52:50 +0300 From: "Eden Cohen" Organization: - Message-ID: <9j6lc2$t9e$1@news.netvision.net.il> References: <9hrijf$al1$1@news.netvision.net.il> Reply-To: "Eden Cohen" Hi You probably use hebrew version of windows, well what you need to do is to change your default setting to english. In win2000 control panel ->reagional options ->your local ->english Good Luck, Uri "Eden Cohen" wrote in message news:9hrijf$al1$1@news.netvision.net.il... > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 15:54:19 +0300 From: "Eden Cohen" Organization: - Message-ID: <9j6ler$l$1@news.netvision.net.il> References: <9hrijf$al1$1@news.netvision.net.il> Reply-To: "Eden Cohen" You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way if this doesn't help. tempering with the code of VxWorx.exe _________________________________ SIEMENS PSE SWS MSA Dipl. Ing. Tomislav Razov tel: 00385 1 6105 101 tomislav.razov@siemens.hr _________________________________ "Eden Cohen" wrote in message news:9hrijf$al1$1@news.netvision.net.il... > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Decent C++ compiler for Tornado2/VxWorks? Date: Thu, 19 Jul 2001 14:52:46 +1200 From: Ian Collins Organization: Masuma Message-ID: <3B564B7E.BDB4C67A@masuma.com> I have been attempting to port an application to VxWorks on PowerPC and I have been derailed by the lack of a decent C++ compiler for this platform. The version of gcc supplied (2.7) is very old, Diab 4.4b has some nasty template bugs (and a non-standard STL). I have been told there is a new release of Tornado with gcc 2.95, but I have been unable to get a release date. So, does any one know of a gcc port that works with Tornado?? Ian - -- Ian Collins Masuma Ltd, Christchurch New Zealand. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 13:48:47 +0200 From: Tomislav Razov Organization: Siemens Message-ID: <3B56C91F.4D970AEC@siemens.hr> References: <9hrijf$al1$1@news.netvision.net.il> You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way where you don't have to change regional settings and it involves tempering with the code of VxWorx.exe. Eden Cohen wrote: > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 13:49:59 +0200 From: Tomislav Razov Organization: Siemens Message-ID: <3B56C967.D3C175E@siemens.hr> References: <9hrijf$al1$1@news.netvision.net.il> You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way where you don't have to change regional settings and it involves tempering with the code of VxWorx.exe. Eden Cohen wrote: > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 13:55:37 +0200 From: Tomislav Razov Organization: Siemens Message-ID: <3B56CAB9.D5DFEBC5@siemens.hr> References: <9hrijf$al1$1@news.netvision.net.il> You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way where you don't have to change regional settings and it involves tempering with the code of VxWorx.exe. Eden Cohen wrote: > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 13:56:08 +0200 From: Tomislav Razov Organization: Siemens Message-ID: <3B56CAD8.59EC3856@siemens.hr> References: <9hrijf$al1$1@news.netvision.net.il> You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way where you don't have to change regional settings and it involves tempering with the code of VxWorx.exe. Eden Cohen wrote: > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxsim very small fonts in output window Date: Thu, 19 Jul 2001 13:56:28 +0200 From: Tomislav Razov Organization: Siemens Message-ID: <3B56CAEC.E21DC8E9@siemens.hr> References: <9hrijf$al1$1@news.netvision.net.il> You are probably using Regional settings other than English. If you change it to English, it should work. There is also another way where you don't have to change regional settings and it involves tempering with the code of VxWorx.exe. Eden Cohen wrote: > I am using Tornado 2 on Win98 and WinNT and get the same problem: when > running Vxsim, the output window > somehow gets a very small (and unreadable) font. > Can I change this behaviour ? > > Thanks, > Eden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: port to VxWorks Date: 19 Jul 2001 12:31:16 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <3B56CB83.4070803@sybernet.com> Frederic Lebouc wrote: > Is there a newsgroup that talks about porting stuff to VxWorks ? This newsgoup talks about all aspects of VxWorks. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Elf Executables on VxWork5.4 for I386 Date: Thu, 19 Jul 2001 10:40:03 -0400 From: "Daniel G Waddington" Organization: Bell Laboratories, Lucent Technologies Message-ID: <9j6rgn$pu82@news.research.bell-labs.com> Hi, Can anyone tell me if it is possible to build and use ELF linkables with Tornado 2 and VxWorks 5.4 on a I386 platform? A.out's are really cack. Ta' DAN --------------------------- Newsgroups: comp.os.vxworks Subject: How to use Tornado 1.0 makefile in Tornado 2.0 IDE? Date: Thu, 19 Jul 2001 15:33:03 +0000 (UTC) From: "Chih-Ming.Liao" Organization: Dept. of Computer & Information Science, NCTU, Taiwan Message-ID: <9j6ujf$i7r$1@news.cis.nctu.edu.tw> Sender: Chih-Ming.Liao Does anybody know how to convert the Tornado 1.0 makefile to Tornado 2.0 IDE setting? thanks in advance! ChihMing --------------------------- Newsgroups: comp.os.vxworks Subject: Re: shut down VxSim window from script? Date: Thu, 19 Jul 2001 17:35:27 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995557435.797243@newsmaster-04.atnet.at> References: <995492817.49988@cswreg.cos.agilent.com> Hi, Using the Windows API (I think it's FindWindow or something near to that) you could search for the Window with that particular title and send it a WM_CLOSE event. This should do the job. However it requires you to hack a little tool for that. hth werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: port to VxWorks Date: Thu, 19 Jul 2001 17:37:13 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995557539.876209@newsmaster-04.atnet.at> References: <3B56CB83.4070803@sybernet.com> Hi, Depends on what you want to know. If it is related to the application you want to port, it's probably best to ask in the (development) newsgroup for that app. For VxWorks, as to my knowledge, there is no other newsgroup that this one. So feel free to ask your VxWorks specific porting questions here. regards werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Works in shell, not in task Date: Thu, 19 Jul 2001 17:40:17 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995557723.939510@newsmaster-04.atnet.at> References: <3B56CB2B.8010206@sybernet.com> For the stack, you can use the Stack Check facility of your Tornado Browser. Did you try this? Are you sure the stack is set correctly for your task? (The Stack Check diagram will point out such problem as well). Besides from that, there is no much difference between invoking the thing from your code or from the shell. However, the shell passes always 10 parameters to the function and sets parameters that are not satisfied from the input to 0. Just in case your prototype is not correct. hth werner --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: "File System" on VxWorks Date: Thu, 19 Jul 2001 17:46:59 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995558126.363157@newsmaster-04.atnet.at> References: <9in5ms$jtkbu$1@ID-92003.news.dfncis.de> <995445191.662263@newsmaster-04.atnet.at> <9j4pp9$ki3iv$1@ID-92003.news.dfncis.de> Hi, "Lee" wrote in message news:9j4pp9$ki3iv$1@ID-92003.news.dfncis.de... > Thanks. It helps. > > Suppose I use DosFs. Can I see file structure like the real DOS after I > upload the firmware into a flash memory? In other words, can I use function > calls (in a user application) like fopen() to open a user file (it's also > part of the firmware) with full path such as "fs\path1\path2\userFile.txt"? You can use most of the standard C file related functions, including the fopen, fread, etc. However, you should use (or must, not sure) forward slashes '/' instead of backslashes '\' for the path name. > One thing I'm not very clear is that everything (VxWorks, user applications, > user files) is in a single firmware file, how can we use DOS file structure > inside a ** single firmware file **, and how does VxWorks control this? No, you cannot. The DOS file system works on the raw disk, not on the VxWorks image. You can load your image from the ATA (or SCSI) disk with appropriate setup in the bootrom. But the VxWorks image is just a normal file, like anything else on the disk. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: LoadLib and unldLib Date: Thu, 19 Jul 2001 09:45:01 -0600 From: Don Small Organization: Sandia National Laboratories Message-ID: <3B57007D.BCDF3636@sandia.gov> References: <3B5690DD.FE84ED2E@sepro-robotique.com> The ability to ld and then unld a program is wonderful during the software development stage. I have had no problems at all as long as you don't attempt to unld a program that has hooked into an interrupt vector, i.e. an ISR. I suppose that if you hook into any kernel service it would cause problems unless you unhook your service routine before unloading. With the above noted caution in mind, you can divide most problems into the ISR function and the work related functions and connect them with queue, semaphores, etc. Place these separate functions into unique files for convenient ld'ing and unld'ing. If this is done properly you can still ld and unld the work related portion without rebooting the processor. From my experience, I would reboot the processor board when modifying the ISR routines. BTW, I never explained the details of removing an interrupt service routine (ISR). If you ld and run a function that uses intConnect to connect an interrupt service routine to an external interrupt, and then unld your code, the kernel still assumes that valid code exists at the memory location where the old ISR existed. While it may stay there for a short time, there is no guarantee. The first time this interrupt occurs and your code has been overwritten, the system usually just hangs. The same problem can occur for any other vxWorks 'hook' (taskCreateHookAdd, taskSwitchHookAdd, ipFilterHookAdd, ...) but at least you have the option with most of these hooks to perform a xxxHookDelete function to clear things up before unld'ing them. Emmanuel Herbreteau wrote: > > Hi, > > Is there someone here who had already used the > loadLib ("loadModule()") and unldLib ("unld()") > to load and unload program (from ATA disk) > when other real time task are running ? > > What are the pitfalls ? > > Is it reliable ? > > -- > Regards > Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: LoadLib and unldLib Date: Thu, 19 Jul 2001 09:48:45 +0200 From: Emmanuel Herbreteau Organization: SEPRO Robotique (BEP) Message-ID: <3B5690DD.FE84ED2E@sepro-robotique.com> Hi, Is there someone here who had already used the loadLib ("loadModule()") and unldLib ("unld()") to load and unload program (from ATA disk) when other real time task are running ? What are the pitfalls ? Is it reliable ? - -- Regards Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to Solove Memory leak problems? Date: Thu, 19 Jul 2001 16:13:28 +0800 From: "tanghaoyu" Organization: Bentium Ltd. Message-ID: <9j64r5$duu$1@mail.cn99.com> References: <92dbccc1.0107172159.c86d6b2@posting.google.com> CodeTest maybe can help you. "Adom" wrote in message news:92dbccc1.0107172159.c86d6b2@posting.google.com... > Hi,All: > i have developed a program in vxworks. > Now it can run correctly .But i found there > are memory leak in it . > I have try all i can think,but it still is there. > Can anybody tell me how to solove this problem? > Any suggest is appreciate! > > Thanks! > Adom > chwin@china.com > --------------------------- Newsgroups: comp.os.vxworks Subject: Robotics Engineering Excellence && RTC (TCP/shared memory IPC package) Date: Thu, 19 Jul 2001 12:55:41 -0400 From: Jorgen Pedersen Organization: National Robotics Engineering Consortium Message-ID: <3B57110D.142474AD@rec.ri.cmu.edu> ANNOUNCEMENT: RE Squared (Robotics Engineering Excellence) is a new start-up company based in Pittsburgh, PA. Our engineers are seasoned robotics engineers with many years of both theoretical and practical robotics experience. Our expertise includes, but is not limited to vehicle retrofits, hardware control systems, electrical design and fabrication, sensor interfaces and integration, ethernet and serial communications, control algorithms, path planning, computer vision, and artificial intelligence. We also offer core software technologies, such as the RTC. RTC is a middleware IPC communications package built on top of TCP (and shared memory under VxWorks) that allows you to send C data structures between multiple computers. RTC is available for Linux, Windows NT, VxWorks, and QNX. Older versions of RTC also support HP-UX, SUN, Solaris, Extended DOS, and Irix. RTC has been extensively tested for over 4 years and is used by many organizations/companies (e.g. Carnegie Mellon's Robotics Institute, NASA, Caterpillar, Network Photonics, etc.). Other core technologies will also be offered soon (e.g. SICK laser interface code, KVH gyro interface code, Crossbow tilt sensor interface code, Novatel GPS interface code, etc.). If you require experienced, efficient robotics engineers, RE Squared has the resources you need. If you have the resources, but want to increase your efficiency by using core robotics technologies, we probably have the technologies you need. Feel free to contact us via email at: info@resquared.com or visit our website at: http://www.resquared.com/index.html For more information on RTC, please visit the RTC web page: http://www.resquared.com/products/rtc/general.html --------------------------- Newsgroups: comp.os.vxworks Subject: PCI_AUTOCONFIG, does it work? Date: Thu, 19 Jul 2001 14:22:50 -0400 From: Michael Reed Message-ID: <3B57257A.FDEEC54B@ll.mit.edu> Hello, TORNADO = Tor2.0.2 KERNEL = vxWorks5.4 HOST = Solaris2.6 BSP = mv2400 TARGET = mvme2400 CPU ARCH = ppc750 I'm porting a PCI/PMC driver which was written for MVME2300 board to MVME2400 board. It was written with the standard form of: pciFindDevice(...) pciDevConfig(...) set up int levels But I keep reading snippets about INCLUDE_PCI_AUTOCONFIG. Can I use this instead of pciFindDevice & pciDevConfig? I tried removing pciFindDevice & pciDevConfig and running pciHeaderShow, and it appeared that the Config Header has some valid BAR addresses and int lines in it. Does this mean that pciAutoConfig ran? Thanks, Mike R. --------------------------- Newsgroups: comp.os.vxworks Subject: file i/o w/multiple tasks Date: Thu, 19 Jul 2001 13:28:00 -0500 From: csallstr Organization: Lucent Technologies Message-ID: <3B5726B0.F6AADBBF@lucent.com> Group: I am rather new to VxWorks, and as I research my own answer, was hoping others can more quickly point out where/what/how etc. I have a problem I have to look at in which 1 task is continualy writing to a file. Occasionally another task must read this file and print the file (ascii only) Using fopen/fprintf, fgetc, etc. I am seeing a problem where the reading task sees non-ascii characters because of what I am assuming is a buffering problem. The current code checks for eof, but sometimes sees non-ascii, but if you re-read later you get past that previous point and may continue. Its like the reader sees the actual contents of the file plus garbage in some system buffer that hasn't been flushed. Any ideas... or pitfalls I need to check for and thus correct. BTW, still reading manuals..... Thankx: Chris S. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: LoadLib and unldLib Date: Thu, 19 Jul 2001 16:49:13 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B5755D9.6408BE84@fnal.gov> References: <3B5690DD.FE84ED2E@sepro-robotique.com> <3B57007D.BCDF3636@sandia.gov> > The ability to ld and then unld a program is wonderful during the > software development stage. I have had no problems at all as long as Something related to ld. Can we restrict the size of max size of a file that ld can download? So of configuration, etc so no one is allowed to download files more than say 500k? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: file i/o w/multiple tasks Date: Fri, 20 Jul 2001 00:59:57 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995584101.809506@newsmaster-04.atnet.at> References: <3B5726B0.F6AADBBF@lucent.com> Hi, I don't use the f... set of functions much, so maybe this does not cause your problems. However, I'm pretty sure that the underlying functions open(), close(), read(), write(), etc. do not care for mutual exclusion and serializing access of multiple tasks. Moreover I think you cannot (or at least should not) open the same file twice (if you open it for _writing_, in particular). If you use the same file handle (pointer) for both activities, I would try to synchronize the two tasks. Do not write to the file while reading and be sure to have the file pointer (seek) where it was before. If you open the file separately, I recommend you make sure only one task has opened the file at any time. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: LongJmp Date: Thu, 19 Jul 2001 08:25:57 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <9j8g8g$eke$1@at-vie-newsmaster01.nextra.at> References: <3B552726.A8804859@lucent.com> Hello, first i have to say that the integrated vxSim in T2 (which is the only one i have) is unbelievable buggy. its imo the most buggy application i have seen in the last years. i use setjmp/longjmp in my real vxWorks , target=x86 & ppc application without problem. i tried - just a very fast testcase - a setjmp/longjmp in my vxsim and it i am even not able to load the code without problems. i think i have integrated complete CRT. #include #include #include static jmp_buf mark; extern "C" int vxmain() { int rc; if ((rc=setjmp(mark))!=0) { printf("RC=%d\n",rc); } else { longjmp(mark,123); } return 1; } and vxSim: value = 0 = 0x0 - -> ld < setjmp2.bin Loading S:/test/vxw/setjmp/setjmp2.bin | Undefined symbols: _longjmp _setjmp Warning: object module may not be usable because of undefined symbols. value = 17384272 = 0x1094350 what version of vxSim do you use to get over this? i have T2 with SP2 installed. - -- regards, Mario Semo. http://www.kirchnersoft.com jp wrote in message news:3B552726.A8804859@lucent.com... > Hi > > While downloading my application to vxSim, longjmp function is giving undefined > reference But strangely, setjmp is not giving any problem. > I have included both setjmp.h and vxworks.h. > Is it bug in vxworks. Please help me as my application is using so many longjmp > and setjmp calls. > > > > Regds > gk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to use ld() to load c++ module with new or delete Date: Thu, 19 Jul 2001 08:28:40 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <9j8g8i$eke$2@at-vie-newsmaster01.nextra.at> References: <9j345j$214$1@clematis.singnet.com.sg> search for your problem in a) this newsgroup with something like Deja... b) in the patch area of WindSurf this is a very very old very well known fixed problem. about once all 2 weeks we post this here. about 3 weeks ago i posted a complete fix (implementation of all missing functions). sorry, currently no time to put it together again. - -- regards, Mario Semo. http://www.kirchnersoft.com Nixvue Systems wrote in message news:9j345j$214$1@clematis.singnet.com.sg... > Below is some cpp code that we want to use ld() function to load into the > memory. > > ld(0,0,"projectcpp.out"); > > However it always give below error message: > > Undefined symbol: __builtin_vec_new (binding 1 type 0) > Undefined symbol: __builtin_vec_delete (binding 1 type 0) > Undefined symbol: __builtin_delete (binding 1 type 0) > Undefined symbol: __builtin_new (binding 1 type 0) > > > How do we go around this problem? > > If we only using c function, the ld() perform no problem. However if we code > in cpp, the ld() always have above mention problem. > > ---------------------------------------------------- > > /*projectcpp.cpp*/ > /*after compile it produce projectcpp.out*/ > > #include "string.h" > #include "iostream.h" > #include "stdlib.h" > #include "stdio.h" > > void realtest(void); > void realtest(void) > { > int *i; > int *j; > > cout << "testing start\n\n" << endl; > i=new int; > printErr("i address=0x%x\n",i); > delete i; > > j=new int[100]; > printErr("j address=0x%x\n",i); > delete[] j; > > } > > extern "C" void testcpp(void) > { > printErr("test cpp load module\n\n\n"); > realtest(); > } > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: using Java,C++,C mixture Date: Fri, 20 Jul 2001 08:16:54 +0200 From: Joern Tietjen Organization: =?iso-8859-1?Q?Dr=E4ger?= Medical AG & Co. KGaA Message-ID: <3B57CCD6.D821710C@draeger.com> Hi, I´d like to know whether anybody uses VxWorks/Tornado in a mixed programming language development environment. Does Java scale for realtime systems ? I´m quite convinced that you allways need a bit of code other than Java in an embedded realtime application (drivers, time critical apps). So, how does the tool chain look like in mixed language environments ? Is Tornado an appropriate IDE in this case ? What JVM would you recommend ? Any experiences,performance figures, pointers or other information are very much appreciated. TIA, Joern Tietjen Draeger Medical AG & Co. KGaA --------------------------- Newsgroups: comp.os.vxworks Subject: Zinc and setFocus() Date: Fri, 20 Jul 2001 08:37:43 +0200 From: Emmanuel Herbreteau Organization: SEPRO Robotique (BEP) Message-ID: <3B57D1B7.DBB82A54@sepro-robotique.com> Hi, We are using Zinc 6.0 and VxWorks 5.4. the setFocus() function give us troubles : when our task "taskXYZ" call setFocus(), sometimes, randomly, it crash our target (PPC). It seems that setFocus() must not be called from any task but the event manager. ???? Is there an other way to do a "setFocus" without this #@$*#*@ function ? ;-) I suspect our GCC command line ? What should I check ? (support for re-entrant function ?) Thanks a lot - -- regards Emmanuel --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: "File System" on VxWorks Date: Wed, 18 Jul 2001 17:49:19 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995471861.649847@newsmaster-04.atnet.at> References: <9in5ms$jtkbu$1@ID-92003.news.dfncis.de> <995445191.662263@newsmaster-04.atnet.at> Hi, Hi, > Then I prayed nothing happened during the data transfer! Yes, this is basically your best bet. Unfortunately, there is no fail-safe file system. If you experience a power fail while writing data to disk you may loose the entire directory, or even worse. At least they added come check functionality to DosFs 2.0. (like chkdsk) > > Things may have changed with DosFs 2.0; it's been a few years since I looked > at this. You are right, however, that WRS has never offered good support > for DosFs. Again, it could be different with 2.0, but the earlier stuff > seemed to be a quick add-on, so that they could advertise that they > supported DosFs. > You should not over-estimate the improvements. Besides the more partitions and FAT-32 features, changes are quire rare. regards Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Works in shell, not in task Date: Fri, 20 Jul 2001 08:26:09 +0000 From: Frederic Lebouc Message-ID: <3B57EB21.9000200@sybernet.com> References: <3B56CB2B.8010206@sybernet.com> <3B576641.40A1729B@support.nl> Kees van der Bent wrote: > > Please post the smallest (comment away everything upto where the problems > disappear) version of your code that still has (at least one of) these problems. > Also show the exact shell command you used. > > Kees I'm porting the lame library. I wrote a wrapper around the frontend/main.c/main (a standard C main)function of lame and renamed it lameLibMain #define MAX_ARG 20 int lameLibMainWrapper(char* arg) { char* argv[MAX_ARG]; int argc = 0; char* main_name = "lameLibMain"; char* delimiter = " "; char* pLast = NULL; int retVal = 0; /* clean-up */ bzero((void*)argv, MAX_ARG * sizeof(char*)); /* first arg is the function itself */ argv[argc++] = main_name; /* then we use the horrible strtok */ argv[argc] = strtok_r(arg, delimiter, &pLast); while(argv[argc]) { argv[++argc] = strtok_r(NULL, delimiter, &pLast); } /* now call lameLibMain */ retVal = lameLibMain(argc, argv); return retVal; } int lameTask() { lameLibMainWrapper("--preset phone -s8 -b16 -x /AG/TEST.PCM /AG/test.mp3"); return 0; } int tryMe() { taskSpawn("tryMe", /* task name */ 151, /* priority */ VX_FP_TASK, /* options */ 0x2000, /* stack size */ (FUNCPTR)lameTask, /* function */ 0, /* parameters */ 0, 0,0,0,0,0,0,0,0); return 0; } *** END *** From the shell I type: - ->lameLibMainWrapper "--preset phone -s8 -b16 -x /AG/TEST.PCM /AG/test.mp3" Frederic. --------------------------- Newsgroups: comp.os.vxworks Subject: simulator and objects Date: Fri, 20 Jul 2001 09:32:55 +0200 From: "Marcin Nicpon" Organization: ACI - Szczecin http://aci.com.pl Message-ID: <9j8jgv$oqo$1@zeus.man.szczecin.pl> I`ve a problem. I working on Tornado 2 for PPC and I using a simulator. I wrote a source code in C++ (include object). While downloading some errors is occured (concerned with object). What I have to do to be able to use a simulator? Does simulator support a C++ object? - -- Greetings Marcin Nicpoñ RTS, Szczecin, Poland nicpon@rts.com.pl mobile (+48)602-533132 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to Solove Memory leak problems? Date: Fri, 20 Jul 2001 09:24:01 +0100 From: Kees van der Bent Organization: news-service.com Message-ID: <3B57EAA1.F766A0AD@support.nl> References: <92dbccc1.0107172159.c86d6b2@posting.google.com> Adom wrote: > Hi,All: > i have developed a program in vxworks. > Now it can run correctly .But i found there > are memory leak in it . > I have try all i can think,but it still is there. > Can anybody tell me how to solove this problem? > Any suggest is appreciate! Before using 'fancy' tools I always (first) use my brain: Find all places where you allocate memory malloc(), calloc(), new ... but also the VxWorks xxxCreate() routines like semBCreate(). Then simply go through the list and see if there is a matching free()/xxxDelete() routine in ALL possible circumstances (use your imagination and ask yourself 'what if ...' questions). Even when you would use automated tools and find the leak must quicker, you still need to understand why/how you (or a colleague) made this programming error. Although code inspection takes more time (when you start doing it), it has the potential of giving you a better understanding of memory allocation issues (because you spent time thinking) and may give you a stronger focus/will to prevent these problems the next time. All the best, Kees --------------------------- Newsgroups: comp.os.vxworks Subject: Re: simulator and objects Date: Fri, 20 Jul 2001 09:39:54 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9j8n8t$t0a$01$1@news.t-online.com> References: <9j8jgv$oqo$1@zeus.man.szczecin.pl> Hello Marcin, you need to recompile the Simulator and configure it to support C++ environment. Further details see the Tornado Manual. Ciao Martin Raabe PS: Speaking for my own, not for my employer "Marcin Nicpon" schrieb im Newsbeitrag news:9j8jgv$oqo$1@zeus.man.szczecin.pl... > I`ve a problem. I working on Tornado 2 for PPC and I using a simulator. I > wrote a source code in C++ (include object). > While downloading some errors is occured (concerned with object). What I > have to do to be able to use a simulator? > Does simulator support a C++ object? > > -- > > > > Greetings > > > Marcin Nicpoñ > RTS, Szczecin, Poland > nicpon@rts.com.pl > mobile (+48)602-533132 > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using Java,C++,C mixture Date: Fri, 20 Jul 2001 09:38:30 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9j8n6a$puk$06$1@news.t-online.com> References: <3B57CCD6.D821710C@draeger.com> Hello Joern, there is an Add-On Tool for Tornado II, that can handle Java and C Sourcecode and supports VxWorks and pJWorks, which is the java support for Tornado. This Add-On Tool is called SNiFF+ and helps you handle Java and C code (Browsing, Projectmanagement and build). For further details ask your local sales representative. HIH Martin Raabe PS: Speaking for my own, not for my employer! "Joern Tietjen" schrieb im Newsbeitrag news:3B57CCD6.D821710C@draeger.com... > Hi, > > I´d like to know whether anybody uses VxWorks/Tornado in a mixed > programming language development environment. > Does Java scale for realtime systems ? > I´m quite convinced that you allways need a bit of code other than Java > in an embedded realtime application (drivers, time critical apps). So, > how does the tool chain look like in mixed language environments ? > Is Tornado an appropriate IDE in this case ? > What JVM would you recommend ? > Any experiences,performance figures, pointers or other information are > very much appreciated. > > TIA, > > Joern Tietjen > Draeger Medical AG & Co. KGaA --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Works in shell, not in task Date: Fri, 20 Jul 2001 10:17:12 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B57E908.F34A6CCB@ks.ericsson.se> References: <3B56CB2B.8010206@sybernet.com> <3B576641.40A1729B@support.nl> <3B57EB21.9000200@sybernet.com> Hi You could try and use memPartOptionsSet with the option MEM_ALLOC_ERROR_SUSPEND_FLAG on the partition you get allocation errors in (0x29998c from your first post), and connect the debugger and do a backtrace to see where it was called from. Btw, is 0x29998c a valid memory partition (use memPartShow(0x29998c, 1))? I'm guessing you are using the simulator (GPF and page faults is not something I have seen in VxWorks) and a couple of post in this NG indicates it's a bit buggy. Will you get the same error if you call lameTask or tryMe from the shell? If you are using the sim. does it work in a real target? HTH Urban --------------------------- Newsgroups: comp.os.vxworks Subject: simulator again Date: Fri, 20 Jul 2001 11:03:14 +0200 From: "Marcin Nicpon" Organization: ACI - Szczecin http://aci.com.pl Message-ID: <9j8oq9$rdu$1@zeus.man.szczecin.pl> I found #define INCLUDE_CPLUS #define INCLUDE_CPLUS_IOSTREAMS #define INCLUDE_CPLUS_STL in one of the configuration files in my simulator project, does it means that the C++ is implemented ? or not ? If yes then I don`t know what I need to do to my simulator downloading C++ objects propertly - -- Greetings Marcin Nicpoñ RTS, Szczecin, Poland nicpon@rts.com.pl mobile (+48)602-533132 --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 22 04:03:32 2001 From: Vxworks Exploder Date: Sun Jul 22 04:03:34 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 22 04:03:16 PDT 2001 Subject: Re: PCI_AUTOCONFIG, does it work? Subject: Re: using Java,C++,C mixture Subject: menuItemInsert menuItemAppend Subject: Re: Is Wind River really this pathetic Subject: Re: about windML2.0 Subject: Interrup question: Please if you can answer Subject: Re: Zinc and setFocus() Subject: VxWorks Interrupt question for you experts out there....... Subject: Re: Interrup question: Please if you can answer Subject: Timezone calc utility Subject: Re: Link errors in the build window Subject: unresolved symbols and bad timings. Subject: Re: simulator again Subject: The PC and Software Museum Subject: Re: How to Solove Memory leak problems? Subject: How to communicate using Serial Port in Vxworks Subject: about bootable application project Subject: What is the proper size of the .out file? Subject: Can I set a few timer In vxworks system?How to do? Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Subject: sntpcLib: sntpcTimeGet error??? Subject: waking only 1 task out o many blocked on a semaphore Subject: Re: about bootable application project Subject: semaphore in vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Subject: Boot sequence problem Subject: Re: PCI_AUTOCONFIG, does it work? Subject: Is VxWorks a RTOS? Subject: Multiple Watchdog timeouts Subject: ccppc problem Subject: Re: How can I block VxWorks default shell (->)? Subject: Re: ccppc problem Subject: Re: Is Wind River really this pathetic Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Subject: Re: PCI_AUTOCONFIG, does it work? Subject: Re: port to VxWorks Subject: Re: How to communicate using Serial Port in Vxworks Subject: Re: Recommend a quad serial I/O PMC Subject: Re: Is Wind River really this pathetic Subject: Re: waking only 1 task out o many blocked on a semaphore Subject: Re: semaphore in vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Subject: NFS issue between VxWorks 5.4 and Linux 2.4 Subject: DHCPc ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Date: Fri, 20 Jul 2001 13:05:20 +0200 From: "Michael Lawnick" Organization: Buergernetz Dillingen Message-ID: <9j93h7$60f$1@snoopy.bndlg.de> References: <3B57257A.FDEEC54B@ll.mit.edu> Reply-To: "Michael Lawnick" Sender: mlawnick@pd9e65848.dip.t-dialin.net Hi Michael, this is not sufficient. Power down your target or, much better, change some PCI-headers and press reset button. Grep for INCLUDE_PCI_AUTOCONFIG in your BSP and project. May be its selectable in workspace. It works in sysLib, where you should find a call to sysPciAutoConfig() in sysHwInit(). HTH - -- Mit freundlichen Grüßen, Michael Lawnick ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== "Michael Reed" schrieb im Newsbeitrag news:3B57257A.FDEEC54B@ll.mit.edu... > Hello, > > TORNADO = Tor2.0.2 > KERNEL = vxWorks5.4 > HOST = Solaris2.6 > BSP = mv2400 > TARGET = mvme2400 > CPU ARCH = ppc750 > > I'm porting a PCI/PMC driver which was written for MVME2300 board to > MVME2400 board. It was written with the standard form of: > > pciFindDevice(...) > > pciDevConfig(...) > > set up int levels > > But I keep reading snippets about INCLUDE_PCI_AUTOCONFIG. Can I use > this instead of pciFindDevice & pciDevConfig? I tried removing > pciFindDevice & pciDevConfig and running pciHeaderShow, and it appeared > that the Config Header has some valid BAR addresses and int lines in > it. Does this mean that pciAutoConfig ran? > > Thanks, > Mike R. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using Java,C++,C mixture Date: Fri, 20 Jul 2001 08:26:09 -0400 From: "Mike Kordik" Message-ID: References: <3B57CCD6.D821710C@draeger.com> We are using Java for our GUI and database. We are using C/C++ for all our data processing. Our system contains a fieldbus back end network to communicate with various sensors. We are doing data logging, statistical calculations, feedback control and other proprietary data manipulations. All this will be done in C/C++. We will be communicating between the C/C++ and Java code using JNI (Java Native Interface). So far we have been unable to get it to work reliably. WRS claims it works and we sent them samples but as of yet have not heard anything from their engineers. BTW calling C++ from Java works fine. I think that is what JNI was intended for. We want C++ to call Java and that is not working well at all. The alternative is to use sockets or just have the Java side poll for data. - -Mike "Joern Tietjen" wrote in message news:3B57CCD6.D821710C@draeger.com... > Hi, > > I´d like to know whether anybody uses VxWorks/Tornado in a mixed > programming language development environment. > Does Java scale for realtime systems ? > I´m quite convinced that you allways need a bit of code other than Java > in an embedded realtime application (drivers, time critical apps). So, > how does the tool chain look like in mixed language environments ? > Is Tornado an appropriate IDE in this case ? > What JVM would you recommend ? > Any experiences,performance figures, pointers or other information are > very much appreciated. > > TIA, > > Joern Tietjen > Draeger Medical AG & Co. KGaA --------------------------- Newsgroups: comp.os.vxworks Subject: menuItemInsert menuItemAppend Date: Fri, 20 Jul 2001 18:02:52 +0530 From: Nagendra Singh Organization: Philips Semiconductors Message-ID: <3B5824F4.EDF29F2A@philips.com> Hi all, i am working on tornado customizations. There are two gui-tcl api's called menuItemInsert and menuItemAppend. These two were well documented in the release 1.0.1 of tornado. But with the release of tornado 2.0 these have undergone some changes but they have not been documented properly. Even tough the release notes of tornado 2.0 says (in section 11.7 gui tcl library...) that the documentation for these can be found in the online documents that come with tornado, i have not been able to find the documentation for these two api's. however other api's documented online refer to these api's freely. so if any of you have worked on this , can you help me out in understanding these api's. thanks in advance, nagendra. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: 20 Jul 2001 11:25:04 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> Sender: bpringlemeir@DeadDuck >>>>> "Joe" == Joe Woodbury writes: Joe> I think you misunderstand. WindRiver ships GCC with Tornado. To Joe> do so, they make, and are continuing to make, modifications to Joe> it. All those changes fall under the GPL which means WindRiver Joe> is giving away it's labor and any trade secrets and/or Joe> proprietary algorithms it may develop. This makes no business Joe> sense, especially when they have an excellent product in the Joe> Diab compiler. Why not concentrate on making Diab even better? What processors does the Diab compiler support? What processors does GCC support? What processors does vxWorks support? Please shut-up. regards, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about windML2.0 Date: 20 Jul 2001 11:50:41 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <31f1a7e2.0107110357.7457394b@posting.google.com> <994864682.712499@newsmaster-04.atnet.at> Sender: bpringlemeir@DeadDuck >> 1. How to configure windML to use touch screen? In windML >> programmer's guide,It says Touch screen device is /touchscreen/0 >> .but I can't find it in input device's pointer configuration. Werner> /touchscreen/0 ist just a name for the input device that will Werner> be created for the touch screen driver provided with Werner> VxWorks. This name does not really matter. Werner> If your touchscreen is not of the type of driver that comes Werner> with VxWorks (Assabet Touch Screen) you need to write your Werner> own driver. Forget the configuration utility, it cannot cope Werner> with custom drivers. You need to build your UGL image on the Werner> command line (or, at your option, hack their config tool ;-) I have written a driver for the `Burr Brown' ADS7846 on a CL-PS7111. These are basically the two parts of the equation for a resistive touch driver. You have an A/D chip and some means of communicating serially with the main processor. This gets the raw touch values. I didn't really like the WindML method of `cooking' the touch driver output. I used a `projective transformation'. I recently looked at a Linux touch driver as well. You can browse around "www.handhelds.org" and other sights to find several Linux touch drivers. WindML used a piece wise interpolation to map the touch value to a screen co-ordinate. A projective transform looks like this, Xs = a*Xt + b*Yt + c Ys = d*Xt + e*Yt + f a,b,c,d,e,f are determined from calibration. Xt,Yt are touch values and Xs,Ys are screen values. You probably determine the constants by solving a matrix. At first I thought the determinate would be good to scale things by, but I think that the A/D bits is a better choice (ie, 8 or 12 bits). >> 2. My distributor give me a unicode font which is ugly.Can you >> tell me where I can find another one? >> Werner> Agfa provides Font Engines that can render vector fonts and Werner> sells the appropriate fonts. Their fonts are quite nice in Werner> look and there is lots of choice, but pricing is "adequate" Werner> *g*. Werner> Is that Unicode font you got from WRS? Or a free font? Just Werner> curious, because we are looking for one for test purposes. WRS has a perl utility to convert BFD fonts to WindML format (see the windSurf sight). I think that there are utilities to convert TrueType fonts to BSD on the web as well. However, many TrueType fonts are copyrighted. I think that the BSD fonts are a good source as well as any XFree86 fonts. (although they might be one and the same). I have actually converted some Agfa fonts to WindML. I think Agfa provides utilities to convert the vector fonts to bitmap. Bitmap fonts are much more efficient than the vector engine... obviously. I think you would need a pretty hefty processor (300mips+) to make the vector engine work nicely. You still might want to avoid this if you are battery powered. regards, Bill Pringlemeir. - -- Do you want to fiddle with what they call insanity? You're allowed! Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Interrup question: Please if you can answer Date: 20 Jul 2001 15:52:32 GMT From: PC@User.com (PC User) Organization: ExecPC Internet - Milwaukee, WI Message-ID: <3b5853c0$0$18890$272ea4a1@news.execpc.com> Could any of you out there help me here. I am trying to generate an interrupt and see if my handler would catch it. This is being done on a Power PC 603 processor. But some how my example code does not seem to be working. Would you please look at this and provide comments or an alternative like an example of your own. Thanks. /* includes */ #include #include #include "vxWorks.h" #include "intLib.h" #include "taskLib.h" #include "iv.h" #include "logLib.h" #include "Logger.h" /* function prototypes */ void interruptHandler(int); void interruptCatcher(void); /* globals */ #define INTERRUPT_NUM 2 #define INTERRUPT_LEVEL 64 #define ITER1 40 #define LONG_TIME 1000000 #define PRIORITY 100 #define ONE_SECOND 100 /* task to generate the SIGINT signal */ void interruptGenerator (void) { int i, j, taskId, priority; STATUS taskAlive; if ( ( taskId = taskSpawn ( "interruptCatcher", PRIORITY, 0x100, 20000, (FUNCPTR) interruptCatcher,0,0,0,0,0,0,0,0,0,0 ) ) == ERROR ) { logMsg ( "taskSpawn interruptCatcher failed\n", 0,0,0,0,0,0); } for ( i = 0; i < ITER1; i++ ) { /* suspend interruptGenerator for one second */ taskDelay ( ONE_SECOND ); sysIntEnable ( INTERRUPT_NUM ); /* check to see if interruptCatcher task is alive! */ if ( ( taskAlive = taskIdVerify ( taskId ) ) == OK ) { logMsg ( "++++++++++++++++++++++++++Interrupt generated\n", 0,0,0,0,0,0 ); /* generate hardware interrupt 2 */ if ( ( sysBusIntGen ( INTERRUPT_NUM, INTERRUPT_LEVEL ) ) == ERROR ) logMsg ( "Interrupt not generated\n", 0,0,0,0,0,0 ); } else /* interruptCatcher is dead */ break; } logMsg ( "\n***************interruptGenerator Exited***************\n", 0,0,0,0,0,0 ); } /* task to handle the interrupt */ void interruptCatcher(void) { int i = 0, j; STATUS connected; /* connect the interrupt vector, INTERRUPT_LEVEL, to a specific interrupt */ /* handler routine , interruptHandler, and pass an argument, i */ if (( connected = intConnect ( INUM_TO_IVEC ( INTERRUPT_LEVEL ), (VOIDFUNCPTR) interruptHandler, i ) ) == ERROR ) logMsg ( "intConnect failed\n", 0,0,0,0,0,0 ); for ( i = 0; i < ITER1; i++ ) { for (j = 0; j < LONG_TIME; j++) logMsg ( "Normal processing in interruptCatcher\n", 0,0,0,0,0,0 ); } logMsg ( "\n+++++++++++++++interruptCatcher Exited+++++++++++++++\n", 0,0,0,0,0,0 ); } /* signal handler code */ void interruptHandler ( int arg) { int i; logMsg ( "-------------------------------interrupt caught\n", 0,0,0,0,0,0 ); for ( i = 0; i < 5; i++ ) logMsg ( "interrupt processing\n", 0,0,0,0,0,0 ); } --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Zinc and setFocus() Date: 20 Jul 2001 12:01:03 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <3B57D1B7.DBB82A54@sepro-robotique.com> Sender: bpringlemeir@DeadDuck >>>>> "Emmanuel" == Emmanuel Herbreteau writes: Emmanuel> Hi, We are using Zinc 6.0 and VxWorks 5.4. the setFocus() Emmanuel> function give us troubles : when our task "taskXYZ" call Emmanuel> setFocus(), sometimes, randomly, it crash our target (PPC). Emmanuel> It seems that setFocus() must not be called from any task Emmanuel> but the event manager. ???? Emmanuel> Is there an other way to do a "setFocus" without this Emmanuel> #@$*#*@ function ? ;-) Emmanuel> I suspect our GCC command line ? What should I check ? Emmanuel> (support for re-entrant function ?) Normally Zinc is not re-entrant. You can not set focus from another task. Even when you enable `Zinc re-entrancy' there are still restrictions. This is explained in the Zinc manuals. You can edit the configuration to add a define for re-entrancy. This means adding `-DZAF_REENTRANT' and rebuilding the Zinc library. See page 166 of the "Zinc programmer's guide v6.0". However, I think a better option is to post an event to the UGL queue. This queue is processed by Zinc. You may even be able to post a setFocus() type event. Certainly it should be obvious that setting focus needs some sort of synchronization as the mouse and keyboard (and other things) could also be setting focus concurrently. hth, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Interrupt question for you experts out there....... Date: 20 Jul 2001 16:08:51 GMT From: PC@User.com (PC User) Organization: ExecPC Internet - Milwaukee, WI Message-ID: <3b585793$0$18890$272ea4a1@news.execpc.com> Hello, I am trying to test the interupt capabilities on a Power PC 603 board using vxWorks. I am using the example below but my handler does not seem to be able to catch the interrupt generated. I am not sure about the vector and interrupt level. Could you all look at the following example and see if you can offer a correction or a solution of your own. I appreciate your time. /* includes */ #include #include #include "vxWorks.h" #include "intLib.h" #include "taskLib.h" #include "iv.h" #include "logLib.h" #include "Logger.h" /* function prototypes */ void interruptHandler(int); void interruptCatcher(void); /* globals */ #define INTERRUPT_NUM 2 #define INTERRUPT_LEVEL 1 #define ITER1 40 #define LONG_TIME 1000000 #define PRIORITY 100 #define ONE_SECOND 100 /* task to generate the SIGINT signal */ void interruptGenerator (void) { int i, j, taskId, priority; STATUS taskAlive; if ( ( taskId = taskSpawn ( "interruptCatcher", PRIORITY, 0x100, 20000, (FUNCPTR) interruptCatcher,0,0,0,0,0,0,0,0,0,0 ) ) == ERROR ) { logMsg ( "taskSpawn interruptCatcher failed\n", 0,0,0,0,0,0); } for ( i = 0; i < ITER1; i++ ) { /* suspend interruptGenerator for one second */ taskDelay ( ONE_SECOND ); sysIntEnable ( INTERRUPT_NUM ); /* check to see if interruptCatcher task is alive! */ if ( ( taskAlive = taskIdVerify ( taskId ) ) == OK ) { logMsg ( "++++++++++++++++++++++++++Interrupt generated\n", 0,0,0,0,0,0 ); /* generate hardware interrupt 2 */ if ( ( sysBusIntGen ( INTERRUPT_NUM, INTERRUPT_LEVEL ) ) == ERROR ) logMsg ( "Interrupt not generated\n", 0,0,0,0,0,0 ); } else /* interruptCatcher is dead */ break; } logMsg ( "\n***************interruptGenerator Exited***************\n", 0,0,0,0,0,0 ); } /* task to handle the interrupt */ void interruptCatcher(void) { int i = 0, j; STATUS connected; /* connect the interrupt vector, INTERRUPT_LEVEL, to a specific interrupt */ /* handler routine , interruptHandler, and pass an argument, i */ if (( connected = intConnect ( INUM_TO_IVEC ( INTERRUPT_LEVEL ), (VOIDFUNCPTR) interruptHandler, i ) ) == ERROR ) logMsg ( "intConnect failed\n", 0,0,0,0,0,0 ); for ( i = 0; i < ITER1; i++ ) { for (j = 0; j < LONG_TIME; j++) logMsg ( "Normal processing in interruptCatcher\n", 0,0,0,0,0,0 ); } logMsg ( "\n+++++++++++++++interruptCatcher Exited+++++++++++++++\n", 0,0,0,0,0,0 ); } /* signal handler code */ void interruptHandler ( int arg) { int i; logMsg ( "-------------------------------interrupt caught\n", 0,0,0,0,0,0 ); for ( i = 0; i < 5; i++ ) logMsg ( "interrupt processing\n", 0,0,0,0,0,0 ); } --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Interrup question: Please if you can answer Date: Fri, 20 Jul 2001 17:59:28 GMT From: Dan Gold Organization: http://www.newsranger.com Message-ID: <4k_57.1098$ar1.4744@www.newsranger.com> References: <3b5853c0$0$18890$272ea4a1@news.execpc.com> In article <3b5853c0$0$18890$272ea4a1@news.execpc.com>, PC User says... > >Could any of you out there help me here. I am trying to generate an interrupt >and see if my handler would catch it. This is being done on a Power PC 603 >processor. But some how my example code does not seem to be working. Would >you please look at this and provide comments or an alternative like an example >of your own. A little more information would be useful. Ok, a lot more information. I.e. what hardware platform are you running on ? Is it your own custom PPC design, or an off the shelf board ? Seeing sysBusIntGen() implies it may be a VME design ? Again, off the shelf (with a well-known PPC-VME chip) or your own custom one (and related HW description) ? Have you actually looked at sysBusIntGen() for your particular BSP to see what it does ? Do you understand it ? What version (and BSP) of vxWorks are you running ? Are there bugs in the BSP related to PPC/VME access or setup ? Have you checked with your FAE or windsurf too see if you may have a problem ? Dan Gold gold@ensemble.com --------------------------- Newsgroups: comp.os.vxworks Subject: Timezone calc utility Date: Fri, 20 Jul 2001 14:12:49 -0400 From: "Yi Qin" Organization: Storm Internet Services Message-ID: <9j9t0q$bro$1@news.storm.ca> Sender: ss8@mail.ss8.ca Hi. Is there any timezone calculation utility I can find somewhere? I am trying to make my software work with all the standard timezones. The current difficulty is the calculation of the start and end of the daylight saving time for each timezone. Any help will be rgeatly appreciated. Yi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Link errors in the build window Date: Wed, 18 Jul 2001 17:56:00 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <995472260.57070@newsmaster-04.atnet.at> References: <3b54008d$0$876$edfadb0f@dspool01.news.tele.dk> Hi, you can just link your code to the VxWorks image to achieve that. Otherwise the build process has no information about the VxWorks image and symbol table. (A downloadable application can be downloaded to any image of that architecture) regards Werner --------------------------- Newsgroups: comp.os.vxworks Subject: unresolved symbols and bad timings. Date: Fri, 20 Jul 2001 14:48:02 -0400 From: "Patrick Brochu" Organization: Nortel Message-ID: <9j9ud1$m8i$1@bcarh8ab.ca.nortel.com> Hi all: We are running into a quirk with vxWorks and the MPC8260. Our code runs on very tight timings and is set to the highest priority (lowest priority number) right after the exception task. The code runs perfectly fine for days if nobody plays with either the serial console or a tShell through Tornado. Once in a while, if I go and mistype a command from the shell, for example: - -> lkupp undefined symbol: lkupp After that, my code seems to lose sync with the hardware it is controlling, as if something was keeping my code to run at the proper intervals. Anybody has any idea why a simple symbol lookup can prevent my code from running as it's been doing for the past 2 days? Thanks in advance! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: simulator again Date: Sat, 21 Jul 2001 00:18:05 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9jaanm$l8g$02$1@news.t-online.com> References: <9j8oq9$rdu$1@zeus.man.szczecin.pl> Hello Marcin, what file do you download? abc.c is the source Do you download abc.o or abc.out? Ciao Martin PS: Speaking for my self, not for my employer. "Marcin Nicpon" schrieb im Newsbeitrag news:9j8oq9$rdu$1@zeus.man.szczecin.pl... > I found > > #define INCLUDE_CPLUS > #define INCLUDE_CPLUS_IOSTREAMS > #define INCLUDE_CPLUS_STL > > in one of the configuration files > > in my simulator project, does it means that the C++ is implemented ? or not > ? > > If yes then I don`t know what I need to do to my simulator downloading C++ > objects propertly > -- > > > > Greetings > > > Marcin Nicpoñ > RTS, Szczecin, Poland > nicpon@rts.com.pl > mobile (+48)602-533132 > > --------------------------- Newsgroups: comp.os.vxworks Subject: The PC and Software Museum Date: Fri, 20 Jul 2001 22:39:10 +0000 (UTC) From: rznody@nowasia.net Organization: An OTEnet S.A. customer Message-ID: <9jabue$5cc$259@usenet.otenet.gr> - - Do you like OLD computers and software ? - - Would you like to download some OLD Windows Versions ? ** THE NUMBER 1 MUSEUM FOR PERSONAL COMPUTERS AND SOFTWARE ** /////////////////////////////////////// / www.pcmuseum.8m.net / /////////////////////////////////////// Featuring: * Personal Computers * Terminals * Dumb Terminal Support Center * Software Section * Trade & Buy Market * and MORE ! http://www.pcmuseum.8m.net mutegqkereizu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to Solove Memory leak problems? Date: 18 Jul 2001 18:48:49 -0700 From: chwin@china.com (Adom) Organization: http://groups.google.com/ Message-ID: <92dbccc1.0107181748.5bec9853@posting.google.com> References: <92dbccc1.0107172159.c86d6b2@posting.google.com> thanks ,all: the reply is very useful my project is to implemetion IPSEC and IKE on vxworks. the problem is rasied when process ike message . i have found some possible reasons using WindView and Triggering. Adom 19th,July --------------------------- Newsgroups: comp.os.vxworks Subject: How to communicate using Serial Port in Vxworks Date: 19 Jul 2001 01:24:27 -0700 From: chwin@china.com (Adom) Organization: http://groups.google.com/ Message-ID: <92dbccc1.0107190024.42f7c61f@posting.google.com> Hi, In my project ,i must use Terminal to config my system parameters (PPC860,Vxwors),but i don't know how to communicate using Serial Port. Can you give a suggest? Thanks! Adom 2001-7-19 --------------------------- Newsgroups: comp.os.vxworks Subject: about bootable application project Date: 19 Jul 2001 00:11:57 -0700 From: gcai@netease.com (cai) Organization: http://groups.google.com/ Message-ID: <31f1a7e2.0107182311.603ec815@posting.google.com> hi, I have just finished my downloadable application modules(*.o).Can you tell me how to link them to the bootable module? I can't compile my C++ files with those bootable module files,it seems the compiler does not support C++ in that project. (I can compiler these files in a downloadable project.) how can I set it up to use c++? thanks in advance. cai --------------------------- Newsgroups: comp.os.vxworks Subject: What is the proper size of the .out file? Date: 19 Jul 2001 02:54:54 -0700 From: virsala@yahoo.com (robert chen) Organization: http://groups.google.com/ Message-ID: Hi all, I am a newbie to vxworks. We compile a project and get a out file which size is about 3MB (is not debug version!). But the board memory is 1.4 MB. The size seems too big:). What should I do? Any help is appreciated! Robert Chen --------------------------- Newsgroups: comp.os.vxworks Subject: Can I set a few timer In vxworks system?How to do? Date: 19 Jul 2001 04:27:19 -0700 From: mingxinmu@hotmail.com (mumx) Organization: http://groups.google.com/ Message-ID: <75f406a7.0107190327.78a065cd@posting.google.com> In unix,I can use delta list to set timer and their handlers. vxworks system is real time system,Does it have any advantages in timers? I want to set timers as delta list . Any other method? Thank you in advance! mumx --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Date: 19 Jul 2001 08:55:01 -0700 From: sandra.a.nielsen@boeing.com (Sandra Nielsen) Organization: http://groups.google.com/ Message-ID: <7820a34b.0107190755.43b03102@posting.google.com> References: <3B564B7E.BDB4C67A@masuma.com> Try Green Hills MULTI. We've been using it for over a year to to PowerPC (inc. Altivec) C++ code. It's more than just a compiler, it's an entire Integrated Development Environment (IDE) for real-time embedded development. Works great. Goto http://www.ghs.com/. Ian Collins wrote in message news:<3B564B7E.BDB4C67A@masuma.com>... > I have been attempting to port an application to VxWorks on PowerPC and > I have been derailed by the lack of a decent C++ compiler for this > platform. > > The version of gcc supplied (2.7) is very old, Diab 4.4b has some nasty > template bugs (and a non-standard STL). > > I have been told there is a new release of Tornado with gcc 2.95, but I > have been unable to get a release date. > > So, does any one know of a gcc port that works with Tornado?? > > Ian --------------------------- Newsgroups: comp.os.vxworks Subject: sntpcLib: sntpcTimeGet error??? Date: 19 Jul 2001 11:40:47 -0700 From: jhannah@nortelnetworks.com (Jay Hannah) Organization: http://groups.google.com/ Message-ID: <4394e178.0107191040.344c175e@posting.google.com> Hello all, I'm just starting to get familiar with the SNTP protocol and I'm trying to get the 2 libraries, sntpcLib and sntpsLib working in the vxSim on an NT machine. However, I seem to have hit a roadblock. I can get the time synced between two sims when I use the PASSIVE mode but I've been asked to get it working in ACTIVE mode. The problem is, when I run sntpcTimeGet, it returns and error and I'm pulling my hair out trying to figure out why. Here is the code that I'm using to try and get it to work: struct timespec myTime; STATUS sntpc_status; sntpc_status = sntpcTimeGet(NULL, WAIT_FOREVER, &myTime); The code seems simple enough but the call to sntpcTimeGet returns and error (-1) every time I call it. It also sets the errno value to 48. I've tried just about everything I can think of to get this working, without success. The only thing I can think of now is that it has something to do with the simulated network connecting my two simulators, of which I know very little, however, if that is the case, why would PASSIVE mode work? Like I said, I am totally stumped, so any tips, tricks, advice, comments or suggestions would be greatly appreciated. Thanks in advance. Jay Hannah --------------------------- Newsgroups: comp.os.vxworks Subject: waking only 1 task out o many blocked on a semaphore Date: 19 Jul 2001 17:49:48 -0700 From: saurabh@timetra.com (Saurabh Shrivastava) Organization: http://groups.google.com/ Message-ID: Hi!, is there a way to wake up only 1 task, identified by its TaskId which is blocked on a semaphore ? i couldnt find anything related to this. the one which comes closest is semFlush. thanks! - -saurabh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about bootable application project Date: 19 Jul 2001 18:11:35 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107191711.473a7d68@posting.google.com> References: <31f1a7e2.0107182311.603ec815@posting.google.com> Just add 'extern "C"' before the definition of your functions in your C++ files, and try again, I think it will be OK. ellin gcai@netease.com (cai) wrote in message news:<31f1a7e2.0107182311.603ec815@posting.google.com>... > hi, > I have just finished my downloadable application modules(*.o).Can you > tell me how to link them to the bootable module? > I can't compile my C++ files with those bootable module files,it seems > the compiler does not support C++ in that project. (I can compiler > these files in a downloadable project.) how can I set it up to use > c++? > > thanks in advance. > > cai --------------------------- Newsgroups: comp.os.vxworks Subject: semaphore in vxworks Date: 19 Jul 2001 19:51:10 -0700 From: saurabh@timetra.com (Saurabh Shrivastava) Organization: http://groups.google.com/ Message-ID: Hi!, is there a way to unblock task selectively (based on their task ids), tasks which are blocked on a mutex ? thanks! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Date: 20 Jul 2001 04:53:02 -0700 From: gvarndell@hotmail.com (George Varndell) Organization: http://groups.google.com/ Message-ID: References: <3B57257A.FDEEC54B@ll.mit.edu> It works, but it's not something you simply enable by turning it on -- it needs your help. There's a structure to build and initialize, for PCI memory and I/O space mapping allocation, and then you must call it. If your BSP doesn't currently call it, then you'll need to add a call to it. Autoconfig doesn't know how to allocate PCI interrupts, so you must supply a user hook to do it. Given that you're using a COTS board, all the info you need is already available in the BSP. Be aware though, if you leave any hard-coded (find and config) PCI configurations intact while trying to add autoconfig for specific devices, you'll need to be very careful in configuring autoconfig. Otherwise, you'll likely run into resource conflicts that will be difficult to debug. Regards, George Varndell - -- Varndell Engineering, LLC. Embedded Systems Expertise 81 Baltimore St. Suite 206 http://www.varndellengineering.com Cumberland MD. 21502 Michael Reed wrote in message news:<3B57257A.FDEEC54B@ll.mit.edu>... > Hello, > > TORNADO = Tor2.0.2 > KERNEL = vxWorks5.4 > HOST = Solaris2.6 > BSP = mv2400 > TARGET = mvme2400 > CPU ARCH = ppc750 > > I'm porting a PCI/PMC driver which was written for MVME2300 board to > MVME2400 board. It was written with the standard form of: > > pciFindDevice(...) > > pciDevConfig(...) > > set up int levels > > But I keep reading snippets about INCLUDE_PCI_AUTOCONFIG. Can I use > this instead of pciFindDevice & pciDevConfig? I tried removing > pciFindDevice & pciDevConfig and running pciHeaderShow, and it appeared > that the Config Header has some valid BAR addresses and int lines in > it. Does this mean that pciAutoConfig ran? > > Thanks, > Mike R. --------------------------- Newsgroups: comp.os.vxworks Subject: Boot sequence problem Date: 19 Jul 2001 14:18:39 -0700 From: ahmadrana@usa.net (ahmadrana) Organization: http://groups.google.com/ Message-ID: <22b28edd.0107191318.7e258cee@posting.google.com> I want to block-align the _binArrayStart (in the data segment of the bootstrap code for bootable applications). For this, I added the following target to the makefile: vxWorks.st_rom.bin : depend.$(BSP_NAME) bootInit_st.o romInit.o \ vxWorks.st.Z.o $(LDDEPS) $(LIBS) $(BOOT_EXTRA) - @ $(RM) $(IMI) - @ $(RM) $(IBR) - @ $(RM) version.o $(CC) -c $(CFLAGS) -o version.o $(CONFIG_ALL)/version.c $(LD) $(LDFLAGS) -e $(ROM_ENTRY) -Tdata 0060ffff $(LD_HIGH_FLAGS) \ -Map tmp1.map -o tmp_image vxWorks.st.Z.o romInit.o bootInit_st.o version.o $(BOOT_EXTRA) \ $(LIBS) $(ROMSIZEPROG) -b $(ROM_SIZE) tmp_image $(LDOUT_CONV) tmp_image coffArmToBin < tmp_image > $@ As you can see, it is the same as the vxWorks.st_rom target defined in the target/h/make/rules.bsp makefile. The only difference is that when tmp_image (in place of vxWorks.st_rom image in the rules.bsp file) is generated by the linker ($(LD) ), I add the flag -Tdata 0060ffff. Also, I have added the flag -Map tmp1.map. The mapfile generated shows that the binArrayStart symbol now starts from 0x00610000 (0x0060ffff is filled with 0x1). binArrayStart, by the way, is the start of the compressed image. But this doesn't work. Seems like the download to RAM_LOW_ADRS is not happening. Any ideas what I can do to make it work? - -Ahmad --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Date: 20 Jul 2001 06:26:59 -0700 From: spk@uswest.net (Scott Kamara) Organization: http://groups.google.com/ Message-ID: <73e78f4e.0107200526.194d3522@posting.google.com> References: <3B57257A.FDEEC54B@ll.mit.edu> Yes, pci autoconfig probably ran. The pci autoconfig support has worked fine for us and its great to have. Too bad this wasnt here a long time ago for the PowerPC boards. You should not have to remove any existing logic to have it run, just enable INCLUDE_PCI_AUTOCONFIG in 'config.h' and recompile. You should also build a new vxWorks bootrom file and reflash it. The idea is that autoconfig runs once via the vxWorks bootrom program. This has worked fine for us with the Moto PMC-span expansion board as well. It is possible for autoconfig to run out of pool space as it auto-allocates memory space for each pmc/pci device that requires it. For instance, we have two same-type cards installed that want to use about 4MB each of non-prefetch memory space. The default size for this memory pool was 8MB. When autoconfig could not allocate any more for a given card it left the pci header BARx values for this and subsequent cards as 0 (or all ffs, i forget which). There is no warning message displayed that there was not enough pool space. Yes, one can run the autoconfig support with its debug flag on and see these types of warning messages, but it would be nice if 'out-of-resource' messages were displayed automatically at boot time to make it clear. Scott Kamara Michael Reed wrote in message news:<3B57257A.FDEEC54B@ll.mit.edu>... > Hello, > > TORNADO = Tor2.0.2 > KERNEL = vxWorks5.4 > HOST = Solaris2.6 > BSP = mv2400 > TARGET = mvme2400 > CPU ARCH = ppc750 > > I'm porting a PCI/PMC driver which was written for MVME2300 board to > MVME2400 board. It was written with the standard form of: > > pciFindDevice(...) > > pciDevConfig(...) > > set up int levels > > But I keep reading snippets about INCLUDE_PCI_AUTOCONFIG. Can I use > this instead of pciFindDevice & pciDevConfig? I tried removing > pciFindDevice & pciDevConfig and running pciHeaderShow, and it appeared > that the Config Header has some valid BAR addresses and int lines in > it. Does this mean that pciAutoConfig ran? > > Thanks, > Mike R. --------------------------- Newsgroups: comp.os.vxworks Subject: Is VxWorks a RTOS? Date: 20 Jul 2001 06:50:11 -0700 From: mnylund@mail.com (Martin Nylund) Organization: http://groups.google.com/ Message-ID: <4113af3.0107200550.72823e0d@posting.google.com> Hello, I read some arguments that VxWorks would not be a real RTOS, but real- time tasking library. What is the impact of this, if so? I also read a comment that VxWorks started as a bug fixing consultant company for VRTX. Is VxWorks similiar to VRTX? Thanks Martin --------------------------- Newsgroups: comp.os.vxworks Subject: Multiple Watchdog timeouts Date: 20 Jul 2001 09:41:25 -0700 From: pepsi_000@hotmail.com (vxJunkie) Organization: http://groups.google.com/ Message-ID: <676eb679.0107200841.d0ff405@posting.google.com> hi folks, I had a question releated to the OS.. Suppose i start 20-25 different wdogs wdStart(...) wdStart(...) . . . . . 20 times. I do not have any blocking calls in my timeout handler.the timeout is about 1 second. How does the OS handle the timeouts. Does it automatically stagger the timeouts? Will it effect my other tasks...for eg: is there a chance of getting a "workQpanic...error etc.."?? Thanks, --------------------------- Newsgroups: comp.os.vxworks Subject: ccppc problem Date: 20 Jul 2001 12:10:26 -0700 From: forddavi@msu.edu (Davis Ford) Organization: http://groups.google.com/ Message-ID: Hi, I'm having the following make error when trying to compile an SDK: ccppc: installation problem, cannot exec `cpp': No such file or directory make: *** [socdiag.o] Error 1 ccppc and cpp are both installed and in the beginning of my $PATH like so: $which ccppc /home/dford/TMS_BETA/t2tms/tms20beta4/host/sun4-solaris2/bin/ccppc $which cpp /home/dford/TMS_BETA/t2tms/tms20beta4/host/sun4-solaris2/lib/gcc-lib/powerpc-wrs-vxworks/cygnus-2.7.2-960126/cpp Any ideas what the problem is? Thanks in advance, Davis --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How can I block VxWorks default shell (->)? Date: Sat, 21 Jul 2001 01:20:26 GMT From: "Stephen Banville" Organization: Road Runner Message-ID: References: <9hjt5a$ee2$1@news1.kornet.net> You can also write your own using pty's and changing the global std file descriptors. "Johan Borkhuis" wrote in message news:Xns90D198216D0FDborkhuisagerecom@135.7.153.102... > "±èÀÀ½Ä" wrote: > > As far as I know windriver don't show the source of ``telnet'' and > > ``shell''. > > However I am looking forward to enter my own shell command interpreter > > when I connect target throught telnet. > > There is an unsupported Telnet server available from WindRiver. Ask your > FAE for it. > > Groeten, > Johan > > -- > o o o o o o o . . . _____________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | borkhuis@agere.com | > >(________|__|_[_________]_|__________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ccppc problem Date: 21 Jul 2001 08:39:52 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107210739.238b045b@posting.google.com> References: Hello, Have you sourced the torVars.csh or torVars.sh file that is in the sun4-solaris2/bin directory? Select the one that matches your shell, source it, and then try the compiler again. HTH, John... forddavi@msu.edu (Davis Ford) wrote in message news:... > Hi, I'm having the following make error when trying to compile an SDK: > > ccppc: installation problem, cannot exec `cpp': No such file or > directory > make: *** [socdiag.o] Error 1 > > ccppc and cpp are both installed and in the beginning of my $PATH like > so: > > $which ccppc > /home/dford/TMS_BETA/t2tms/tms20beta4/host/sun4-solaris2/bin/ccppc > > $which cpp > /home/dford/TMS_BETA/t2tms/tms20beta4/host/sun4-solaris2/lib/gcc-lib/powerpc-wrs-vxworks/cygnus-2.7.2-960126/cpp > > Any ideas what the problem is? > > Thanks in advance, > > Davis --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Sat, 21 Jul 2001 10:41:54 -0600 From: "Joe Woodbury" Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> Normally I wouldn't reply to such a childish comment, however I was intrigued by the complete absence of thought in making it. I gather the logic is; GCC supports processors that Diab does not, ergo GCC *MUST* be used. Perhaps the intellectual giant who makes this assertion is unaware that there is nothing to preclude Diab from supporting additional processors except that it hasn't been done. Apparently, I gather, only developers who work on GCC are capable of writing compilers for the processors in question. "Bill Pringlemeir" wrote in message news:uae1zk4hb.fsf@yahoo.com... > >>>>> "Joe" == Joe Woodbury writes: > > Joe> I think you misunderstand. WindRiver ships GCC with Tornado. To > Joe> do so, they make, and are continuing to make, modifications to > Joe> it. All those changes fall under the GPL which means WindRiver > Joe> is giving away it's labor and any trade secrets and/or > Joe> proprietary algorithms it may develop. This makes no business > Joe> sense, especially when they have an excellent product in the > Joe> Diab compiler. Why not concentrate on making Diab even better? > > What processors does the Diab compiler support? What processors does > GCC support? What processors does vxWorks support? Please shut-up. > > regards, > Bill Pringlemeir. > > -- > Useful vxWorks URLS, > "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" > "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Date: Sat, 21 Jul 2001 10:51:07 -0600 From: "Joe Woodbury" Message-ID: References: <3B564B7E.BDB4C67A@masuma.com> <7820a34b.0107190755.43b03102@posting.google.com> I've been looking at Green Hills MULTI , but it wasn't clear how well it worked with VxWorks and the Tornado tools (esp. WindView and PerformancePak.) It says it does, but I'm curious at how seamless and effective that integration is. Joe Woodbury "Sandra Nielsen" wrote in message news:7820a34b.0107190755.43b03102@posting.google.com... > Try Green Hills MULTI. We've been using it for over a year to to > PowerPC (inc. Altivec) C++ code. It's more than just a compiler, it's > an entire Integrated Development Environment (IDE) for real-time > embedded development. Works great. Goto http://www.ghs.com/. > > Ian Collins wrote in message news:<3B564B7E.BDB4C67A@masuma.com>... > > I have been attempting to port an application to VxWorks on PowerPC and > > I have been derailed by the lack of a decent C++ compiler for this > > platform. > > > > The version of gcc supplied (2.7) is very old, Diab 4.4b has some nasty > > template bugs (and a non-standard STL). > > > > I have been told there is a new release of Tornado with gcc 2.95, but I > > have been unable to get a release date. > > > > So, does any one know of a gcc port that works with Tornado?? > > > > Ian --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Date: 21 Jul 2001 10:59:12 -0700 From: spk@uswest.net (Scott Kamara) Organization: http://groups.google.com/ Message-ID: <73e78f4e.0107210959.40b2eac4@posting.google.com> References: <3B57257A.FDEEC54B@ll.mit.edu> gvarndell@hotmail.com (George Varndell) wrote in message news:... > It works, but it's not something you simply enable by > turning it on -- it needs your help. Not in the case of the MVME2400 BSP which is what Michael is using. Its ready to go. Just enable INCLUDE_PCI_AUTOCONFIG in config.h by however means you choose (T2 project facility or other) and rebuild the kernel and reflash the MVME2400 bootrom. Note that this is the case for the MVME2400 BSP as purchased from WRS. If you are using an 'old', 'early access' version of the BSP from Motorola, then perhaps it is not ready to go (i havent seen it) The BSP version WRS delivers is: 1.2/0 The Early Access BSP version from Motorola should be this or lower: 1.1/0.3 - -Scott Kamara --------------------------- Newsgroups: comp.os.vxworks Subject: Re: port to VxWorks Date: 21 Jul 2001 12:17:36 -0700 From: h0ust0n@my-deja.com (bob h) Organization: http://groups.google.com/ Message-ID: <2dab5758.0107211117.19f2462a@posting.google.com> References: <3B56CB83.4070803@sybernet.com> Frederic Lebouc wrote in message news:<3B56CB83.4070803@sybernet.com>... > Hi all, > > Is there a newsgroup that talks about porting stuff to VxWorks ? > > Thanks. > > Fred. In general, think of vxworks as BSD unix with many 'extensions'. So it's easiest to port bsd apps to vxworks. Again this is offered as a very broad statement, but it's served me pretty well. - bob --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to communicate using Serial Port in Vxworks Date: 21 Jul 2001 12:21:16 -0700 From: h0ust0n@my-deja.com (bob h) Organization: http://groups.google.com/ Message-ID: <2dab5758.0107211121.6800c2f7@posting.google.com> References: <92dbccc1.0107190024.42f7c61f@posting.google.com> chwin@china.com (Adom) wrote in message news:<92dbccc1.0107190024.42f7c61f@posting.google.com>... > Hi, > > In my project ,i must use Terminal to config my system parameters > (PPC860,Vxwors),but i don't know how to communicate using Serial Port. > Can you give a suggest? > Thanks! > Adom > 2001-7-19 Look into the 'target' shell which gives you terminal access (default 9600 baud no parity) to the command line interpreter. look at tyLib and tyColib - bob --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Recommend a quad serial I/O PMC Date: 21 Jul 2001 12:30:08 -0700 From: h0ust0n@my-deja.com (bob h) Organization: http://groups.google.com/ Message-ID: <2dab5758.0107211130.3376fc4@posting.google.com> References: <9j102l$h72$1@trog.dera.gov.uk> "R Grommet" wrote in message news:<9j102l$h72$1@trog.dera.gov.uk>... > Hi > > Can anyone recommend a good quad RS232/RS422 serial I/O PMC with driver > support for T2 and the Motorola MVME5100? I'm having amazing problems with > my current card and I'm wondering if there is a better one out there. > > TIA > > Ralph G Try General Standards, I think its the PMC-SIO4. It offers very large Tx and RX fifos on each of the 4 serial ports. You can make very large message transfers (8 Kb) without any interrupts at all. Async, Hdlc, and other modes. go to www.google.com and do a search on +"General Standards" +SIO4 to locate the vendor. if u want more driver support, write me at bobi@sis-tek.com - bob --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Sat, 21 Jul 2001 12:30:45 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B59D865.107CB71E@covad.net> References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> Reply-To: drdiags@covad.net Joe, I think the real point is that up until the purchase of Diab by Wind River, the GNU Compiler Collection was bundled as the main compiler used for vxWorks development. So you now have the expectation that as new versions of GCC are made available, Wind River is expected to provide the updated features. There are probably even contractual obligations which require WRS to continue to support the GCC version compilers (though several folks could speak to how expectation and reality may not coincide, i.e. where is that 2.95.3 release for PPC?). So until a roadmap, put together by those business folks you feel WRS lacks, can be delivered to the developers, the expectation is that GCC will be supported out into some TBD future date. The same can be said about the debugger, is it SingleStep for the IDE or Look or one of many others (Crosswind/GDB may not extend past the Tornado 2.x versions)? With all due respect, I must say that I do agree with Bill on this issue. I do hope that you are able to get to the point where you can move past the tools and get to developing your product. But as someone pointed out here, if everything was perfect, we wouldn't need this newsgroup. Joe Woodbury wrote: > > Normally I wouldn't reply to such a childish comment, however I was > intrigued by the complete absence of thought in making it. I gather the > logic is; GCC supports processors that Diab does not, ergo GCC *MUST* be > used. Perhaps the intellectual giant who makes this assertion is unaware > that there is nothing to preclude Diab from supporting additional processors > except that it hasn't been done. Apparently, I gather, only developers who > work on GCC are capable of writing compilers for the processors in question. > > "Bill Pringlemeir" wrote in message > news:uae1zk4hb.fsf@yahoo.com... > > >>>>> "Joe" == Joe Woodbury writes: > > > > Joe> I think you misunderstand. WindRiver ships GCC with Tornado. To > > Joe> do so, they make, and are continuing to make, modifications to > > Joe> it. All those changes fall under the GPL which means WindRiver > > Joe> is giving away it's labor and any trade secrets and/or > > Joe> proprietary algorithms it may develop. This makes no business > > Joe> sense, especially when they have an excellent product in the > > Joe> Diab compiler. Why not concentrate on making Diab even better? > > > > What processors does the Diab compiler support? What processors does > > GCC support? What processors does vxWorks support? Please shut-up. > > > > regards, > > Bill Pringlemeir. > > > > -- > > Useful vxWorks URLS, > > "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" > > "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: waking only 1 task out o many blocked on a semaphore Date: Sat, 21 Jul 2001 20:41:58 GMT From: Pierre Frigon Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: <3B59E963.114B88A6@yahoo.com> References: You can create a semaphore specifically for that task. Saurabh Shrivastava wrote: > Hi!, > is there a way to wake up only 1 task, identified by its TaskId > which is blocked on a semaphore ? i couldnt find anything related to > this. the one which comes closest is semFlush. > > thanks! > -saurabh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: semaphore in vxworks Date: Sat, 21 Jul 2001 20:45:34 GMT From: Pierre Frigon Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: <3B59EA3A.70A66C47@yahoo.com> References: I assume that your tasks is blocked on a Mutex because it is waiting for a resource. Why would you want to unblock a task in this situation? It would be helpful to have more details. Saurabh Shrivastava wrote: > Hi!, > is there a way to unblock task selectively (based on their task > ids), tasks which are blocked on a mutex ? > > thanks! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PCI_AUTOCONFIG, does it work? Date: 21 Jul 2001 14:27:02 -0700 From: gvarndell@hotmail.com (George Varndell) Organization: http://groups.google.com/ Message-ID: References: <3B57257A.FDEEC54B@ll.mit.edu> <73e78f4e.0107210959.40b2eac4@posting.google.com> spk@uswest.net (Scott Kamara) wrote in message news:<73e78f4e.0107210959.40b2eac4@posting.google.com>... > gvarndell@hotmail.com (George Varndell) wrote in message news:... > > > It works, but it's not something you simply enable by > > turning it on -- it needs your help. > > Not in the case of the MVME2400 BSP which is what Michael > is using. Its ready to go. Just enable INCLUDE_PCI_AUTOCONFIG > in config.h by however means you choose (T2 project facility or other) > and rebuild the kernel and reflash the MVME2400 bootrom. Right. I should have said that it's not *necessarily* something you simply enable. I am not familiar at all with that particular BSP. It's most likely, as you say, ready to go. Regards, George Varndell - - Varndell Engineering, LLC. Embedded Systems Expertise 81 Baltimore St. Suite 206 http://www.varndellengineering.com Cumberland MD. 21502 --------------------------- Newsgroups: comp.os.vxworks Subject: NFS issue between VxWorks 5.4 and Linux 2.4 Date: 21 Jul 2001 21:38:06 -0700 From: brendan@sighup.net (Brendan Bouffler) Organization: http://groups.google.com/ Message-ID: <8c33e989.0107212038.db1f80b@posting.google.com> Hi all, This has been driving me nuts for days. I'm trying to mount an NFS-exported volume from a VxWorks box (a PixelPower Collage 2 in this instance) onto a Linux Box running RedHat 7.1. I get the mount in place, no problems. But if I do an 'ls' on the mount point, it returns "0 files". If I reach out and try to read (or write) a file that I know is there, it works. I just can't get a listing of the files that _are_ there. For the record: o VxWorks (for H050 (603)) version 5.4. Kernel: WIND version 2.5. o RedHat 7.1. Kernel is 2.4.2 Anyone have any ideas why this is the case and if so, how to fix it? - -- brendan bouffler http://www.sighup.net/ New York, NY --------------------------- Newsgroups: comp.os.vxworks Subject: DHCPc Date: Sun, 22 Jul 2001 11:41:51 +0200 From: "Pavius" Organization: Internet Gold, ISRAEL Message-ID: <9je3kb$e2e$1@news.inter.net.il> Hi. I'm trying to implement a dhcp client for a target running vxworks. The code is extremely simple, compiles and links fine. The problem is it doesnt work... Now, since there are a huge amount of "black boxes" in getting config'ed by a remote server (do i send the request packets, are they valid, do i get a response from the server etc) the only way i thought of debugging it was by implementing a simple sniffer which would at least show me the outgoing / incoming packets so i might understand the problem. I wrote something simple implemented as a SNARF protocol bound to the mux which runs fine if i use a target with no DHCPc but doesnt run on those who do. Since i am running Tornado 2 i cannot bind more than one SNARF protocol so i immediately suspected the DHCP client has attached its own filter. calling muxShow(0, 0) displays all the bound protocols, and as suspected there was a snarf protocol (type 257) named "etherInputHook". The manual says something about DHCP requiring the berkly packet filter library but i didnt find any help on that (ill look more into that now). 1) how would you go about debugging such a problem? 2) is there any hidden help resource i am missing? i am used to having tons of information being a win32 programmer (such as the MSDN) and its quite hard diving into such uncharted (?) territories. 3) Why would DHCP need a packet filter? RFC 2131 (page 10) hints that the DHCP client is implemented above the IP stack... 4) how can i access incoming / outgoing packets without binding a new protocol to the mux ? Thanks in advance --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 23 04:03:20 2001 From: Vxworks Exploder Date: Mon Jul 23 04:03:22 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 23 04:03:05 PDT 2001 Subject: Re: Recommend a quad serial I/O PMC Subject: Re: Is Wind River really this pathetic Subject: Re: Is Wind River really this pathetic Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Subject: Re: LoadLib and unldLib Subject: [Q] Linking... Subject: gui library Subject: 64-bit access and Exception Error Subject: Re: gui library Subject: I have the Ethernet problem..... Subject: Re: I have the Ethernet problem..... Subject: DOS FTP ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Recommend a quad serial I/O PMC Date: 22 Jul 2001 10:11:40 -0700 From: spk@uswest.net (Scott Kamara) Organization: http://groups.google.com/ Message-ID: <73e78f4e.0107220911.4aad055b@posting.google.com> References: <9j102l$h72$1@trog.dera.gov.uk> I am using T2/MVME5101 with the SBS Octal-Serial (async) PMC card as well as T2/MVME2400 with the SBS Maxim634 (formerly SciTech Com360) which is a quad serial (sync or async) PMC card; both successfully. I had problems regarding driver/BSP/PCI-config integration but once I was past those hurdles operation was ok in that our 'loopback' testing works fine. At this point, I would recommend one of the cards above (depending on sync or async requirement) with the understanding that our testing has been of the 'external-loopback' type. One needs to be careful when choosing a card in the case of sync/framing protocols becuase the datasheet for the card may list protocols supported ,such as HDLC, but the vxWorks driver for the card does not drive these protocols specifically. - -Scott Kamara "R Grommet" wrote in message news:<9j102l$h72$1@trog.dera.gov.uk>... > Hi > > Can anyone recommend a good quad RS232/RS422 serial I/O PMC with driver > support for T2 and the Motorola MVME5100? I'm having amazing problems with > my current card and I'm wondering if there is a better one out there. > > TIA > > Ralph G --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Sun, 22 Jul 2001 18:05:28 -0600 From: "Joe Woodbury" Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> <3B59D865.107CB71E@covad.net> > I do hope that you are able to get to the point where > you can move past the tools and get to developing your product. That is precisely the problem. I had to spend four hours on Friday, rewriting ANSI C++ compliant code that GCC wouldn't compile but which did compile in both Visual C++ and Diab (the Diab compiler was being used with pSOS by a subsidiary before being shut down a few months back--we don't have a license for it.) It doesn't compile under GCC 2.95 either. Joe Woodbury "DrDiags" wrote in message news:3B59D865.107CB71E@covad.net... > Joe, > > I think the real point is that up until the purchase of Diab > by Wind River, the GNU Compiler Collection was bundled as the > main compiler used for vxWorks development. So you now have the > expectation that as new versions of GCC are made available, Wind River > is expected to provide the updated features. There are probably even > contractual obligations which require WRS to continue to support > the GCC version compilers (though several folks could speak to how > expectation and reality may not coincide, i.e. where is that 2.95.3 > release for PPC?). So until a roadmap, put together by those business > folks you feel WRS lacks, can be delivered to the developers, the > expectation is that GCC will be supported out into some TBD future > date. The same can be said about the debugger, is it SingleStep for > the IDE or Look or one of many others (Crosswind/GDB may not extend > past the Tornado 2.x versions)? > With all due respect, I must say that I do agree with Bill on > this issue. I do hope that you are able to get to the point where > you can move past the tools and get to developing your product. But > as someone pointed out here, if everything was perfect, we wouldn't > need this newsgroup. > > Joe Woodbury wrote: > > > > Normally I wouldn't reply to such a childish comment, however I was > > intrigued by the complete absence of thought in making it. I gather the > > logic is; GCC supports processors that Diab does not, ergo GCC *MUST* be > > used. Perhaps the intellectual giant who makes this assertion is unaware > > that there is nothing to preclude Diab from supporting additional processors > > except that it hasn't been done. Apparently, I gather, only developers who > > work on GCC are capable of writing compilers for the processors in question. > > > > "Bill Pringlemeir" wrote in message > > news:uae1zk4hb.fsf@yahoo.com... > > > >>>>> "Joe" == Joe Woodbury writes: > > > > > > Joe> I think you misunderstand. WindRiver ships GCC with Tornado. To > > > Joe> do so, they make, and are continuing to make, modifications to > > > Joe> it. All those changes fall under the GPL which means WindRiver > > > Joe> is giving away it's labor and any trade secrets and/or > > > Joe> proprietary algorithms it may develop. This makes no business > > > Joe> sense, especially when they have an excellent product in the > > > Joe> Diab compiler. Why not concentrate on making Diab even better? > > > > > > What processors does the Diab compiler support? What processors does > > > GCC support? What processors does vxWorks support? Please shut-up. > > > > > > regards, > > > Bill Pringlemeir. > > > > > > -- > > > Useful vxWorks URLS, > > > "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" > > > "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" > > -- > http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home > page) > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: Mon, 23 Jul 2001 03:36:41 GMT From: lucius1@telo_large_urban_area.com (Lucius Chiaraviglio) Organization: Somethingorother Message-ID: <3b5b9ad1.4699637@news.telocity.com> References: <6qH17.79$5f6.133962@news.uswest.net> Reply-To: lucius1@telo_large_urban_area.com "Joe Kerkes" wrote: >"Joe Woodbury" wrote in message >news:6qH17.79$5f6.133962@news.uswest.net... >> Since we ordered VxWorks/Tornado II, we've had nothing but problems with >> Wind River. They filled the order incorrectly and I had to resort to a lot >> of yelling and threats to not pay our purchase order to get their attention. >> Yet, two days later, they still have yet to do anything but email the >> correct BSP. On top of all this, the highest level manager at Wind River >> kept blaming us for some of the problems. >> >> So far, I've concluded that Wind River employees are given an IQ test. If >> they fail, they are given a personal relationship test. If they fail that >> also, they are hired. >> >> And heaven forbid they actually do something so obvious like creating a >> compiled help (.chm) file for Windows and fixing obvious bugs in Tornado. >> >> The shame is that VxWorks proper is such an excellent product. >> >> The questions are: Are our experiences unique? Did our company just happen >> to run into the one group of bastards working at Wind River? > >They are all bastards if you ask me! We also have had a lot of problems. >Order mistakes, license keys, bsp and eval boards we have had problems with >them all! Your experience is not unique. I work for Ziatech Corporation, an Intel Company, and they do pretty much the same things to us (we don't get evaluation boards from them, but anything we are supposed to get from them they demonstrate total bureaucratic incompetence with)! - -- Lucius Chiaraviglio E-mail address is approximately: lucius1@telo_large_urban_area.com To get the exact address: ^^^^^^^^^^^^^^^^^ Replace indicated characters with common 4-letter word meaning the same thing and remove underscores (Spambots of Doom, take that!). --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Decent C++ compiler for Tornado2/VxWorks? Date: Mon, 23 Jul 2001 17:20:39 +1200 From: Ian Collins Organization: Masuma Message-ID: <3B5BB427.A9F69F19@masuma.com> References: <3B564B7E.BDB4C67A@masuma.com> Phew! Thanks for the input. Ian Johan Borkhuis wrote: > > Ian Collins wrote: > > > So, does any one know of a gcc port that works with Tornado?? > > http://www.newgcc4vxworks4ppc.cjb.net/ > > Groeten, > Johan > > -- > o o o o o o o . . . _____________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | borkhuis@agere.com | > >(________|__|_[_________]_|__________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === - -- Ian Collins Masuma Ltd, Christchurch New Zealand. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: LoadLib and unldLib Date: Fri, 20 Jul 2001 08:20:18 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <9jgcbs$81b$1@at-vie-newsmaster01.nextra.at> References: <3B5690DD.FE84ED2E@sepro-robotique.com> Hello, i load my complete with loadlib. (i am loading about 16MB of binaries (30 binaries) dynamically. no problem. unld does NOT work - it hangs!!, does not return - when the loaded binary uses STL. i received a private fix from WindRiver for this problem. (new ldlib objects which i have to insert into the CRT libraries). (SPR30803). And i didnt manage to get it working under vxSim (ldlib traps). - -- regards, Mario Semo. http://www.kirchnersoft.com Emmanuel Herbreteau wrote in message news:3B5690DD.FE84ED2E@sepro-robotique.com... > Hi, > > Is there someone here who had already used the > loadLib ("loadModule()") and unldLib ("unld()") > to load and unload program (from ATA disk) > when other real time task are running ? > > What are the pitfalls ? > > Is it reliable ? > > -- > Regards > Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: [Q] Linking... Date: Mon, 23 Jul 2001 14:30:48 +0900 From: "Ogre" Organization: Korea Telecom Message-ID: <9jgcqr$f1g$1@news1.kornet.net> I want to link my image and library.a. but, it's not static linking.. it's important dynamic linking. First, I download libraray using FTP. Next, I link my main image and library. MUST I contain "link module" in main image? How can I do for it ? [ I don't have SYM_TAB. I don't have SHELL ] --------------------------- Newsgroups: comp.os.vxworks Subject: gui library Date: Mon, 23 Jul 2001 09:11:49 +0200 From: "Mehmet Demirel" Organization: Swisscom IP+ (post doesn't reflect views of Swisscom) Message-ID: <9jginj$201$1@rex.ip-plus.net> Reply-To: "Mehmet Demirel" Hi, I am looking for gui libraries in c / c++ for embedded systems, especially for vxworks. If someone knows where I can find such libraries please contact me. Its better for me if the libraries are free, but it isn`t a must. Thanks mehmet.demirel@aon.at --------------------------- Newsgroups: comp.os.vxworks Subject: 64-bit access and Exception Error Date: 23 Jul 2001 01:41:32 -0700 From: netproa@hanmail.net (IL HWAN, KIM) Organization: http://groups.google.com/ Message-ID: <29f27f2c.0107230041.773b01df@posting.google.com> Hello. I'm beginner at vxWorks/MPC. My system configuration about below problem: CPU: MPC8240 RTOS: vxWorks5.4 Flash Memory: Intel 28F320J3A I made the routine that download the vxWorks file and write to Flash, and boot from Flash. 64-bit data bus used to Flash interface(it's RCS1, not RCS0(Boot Flash)). Writing to Flash should be the data-path width since there(MPC8240) is only a single Write Enable(WE) strobe available. So I used FPR(floating point regiser). /*************************************************************** void write64(uns32 *src_addr, uns32 *dst_addr) ****************************************************************/ write64: lfd f0,0(p0) sync stfd f0,0(p1) sync bclr 20,0 I didn't insert error check routine yet.(like crc32) But when I printed and compared the data of the memory(SDRAM) and the Flash, it's same.(I checked 64bytes of the first and middle) So I think the data written is correct. But after "Starting at 0x10000..." exception error printed. Starting at 0x10000... program Exception current instruction address: 0x00010000 Machine Status Register: 0x0008b000 Condition Register: 0x24000082 Task: 0x7b493c8 "tBoot" r0 = 402254 sp = 7b49078 r2 = 0 r3 = 0 r4 = 8000000 r5 = 0 r6 = 0 r7 = 0 r8 = 0 r9 = 460000 r10 = 0 r11 = 46275c r12 = 0 r13 = 0 r14 = 0 r15 = 0 r16 = 0 r17 = 0 r18 = 0 r19 = 0 r20 = 0 r21 = 0 r22 = 0 r23 = 0 r24 = 0 r25 = 0 r26 = 0 r27 = 400e28 r28 = 0 r29 = 10000 r30 = a r31 = 7b49298 msr = 8b000 lr = 40225c ctr = 0 pc = 10000 cr = 24000082 xer = 0 I don't know how to analyze this message. Is it that some code in vxWorks file is incorrect? or System register setting is need to use FPU(floating point register)? or anything else? Please let me know how to fix it. Best regards. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: gui library Date: Mon, 23 Jul 2001 10:46:31 +0200 From: Joern Tietjen Organization: =?iso-8859-1?Q?Dr=E4ger?= Medical AG & Co. KGaA Message-ID: <3B5BE467.8186F5B1@draeger.com> References: <9jginj$201$1@rex.ip-plus.net> contact Reccoware (www.reccoware.de). they ported MGR (free software window-manager) for VxWorks. WRS sells WindML (C) and Zinc (C++) libraries for GUI support. HTH, Joern Mehmet Demirel wrote: > > Hi, > > I am looking for gui libraries in c / c++ for embedded systems, especially > for vxworks. > > If someone knows where I can find such libraries please contact me. > Its better for me if the libraries are free, but it isn`t a must. > > Thanks > mehmet.demirel@aon.at --------------------------- Newsgroups: comp.os.vxworks Subject: I have the Ethernet problem..... Date: Mon, 23 Jul 2001 19:13:56 +0900 From: "ThisCase" Organization: A poorly-installed InterNetNews site Message-ID: <9jgsvq$mqf$1@news.kornet.net> Hi. We are using powerPC860, vxWorks5.4, with END, ADS860 BSP. target is connected to host directly (using cross ethernet cable) target 192.168.0.85(our board), host 192.168.0.143(Windows2000 PC). Here is my problem. After booting my target system. - -> ifShow cpm (unit number 0): Flags: (0x8063) UP BROADCAST MULTICAST ARP RUNNING Type: ETHERNET_CSMACD Internet address: 192.168.0.86 Broadcast address: 192.168.0.255 Netmask 0xffffff00 Subnetmask 0xffffff00 Ethernet address is 08:00:3e:03:02:34 Metric is 0 Maximum Transfer Unit size is 1500 0 packets received; 1 packets sent 0 multicast packets received 0 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped lo (unit number 0): Flags: (0x8069) UP LOOPBACK MULTICAST ARP RUNNING Type: SOFTWARE_LOOPBACK Internet address: 127.0.0.1 Netmask 0xff000000 Subnetmask 0xff000000 Metric is 0 Maximum Transfer Unit size is 32768 0 packets received; 0 packets sent 0 multicast packets received 0 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped value = 29 = 0x1d - -> arpShow LINK LEVEL ARP TABLE destination gateway flags Refcnt Use Interface - -------------------------------------------------------------------------- - -------------------------------------------------------------------------- value = 75 = 0x4b = 'K' But Pinging dose not work. After pinging from target to host. - -> ping("192.168.0.143") PING 192.168.0.143: 56 data bytes no answer from 192.168.0.143 value = -1 = 0xffffffff = udbinfo + 0xfff172cf - -> arpShow LINK LEVEL ARP TABLE destination gateway flags Refcnt Use Interface - -------------------------------------------------------------------------- 192.168.0.143 00:01:02:4b:c3:bd 405 0 5 cpm0 - -------------------------------------------------------------------------- value = 75 = 0x4b = 'K' As shown above, arp table updated correctly. we exam packet that flow from target to host, and find any problem. ethernet destination address : ff:ff:ff:ff:ff:ff ethernet source address : 08:00:3e:03:02:34 Is it correct? I think that ethernet destination address is incorrect. I don't know why this happen. when pinging from host to target, request time out has occurred. ICMP packet from host to target is perfect.( i think so...) ICMP packet(Echo request, from host to target) has ethernet destination address : 08:00:3e:03:02:34 ethernet source address : 00:01:02:4b:c3:bd But. ICMP packet(Echo reply, from target to host) has ethernet destination address : ff:ff:ff:ff:ff:ff ethernet source address : 08:00:3e:03:02:34 I think, every packet that initiated at target have wrong ethernet destination address(ff:ff:ff:ff:ff:ff). (address resolution with arp dosn't work...) Can anybody tell me what could be the problem? Thanks,. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I have the Ethernet problem..... Date: Mon, 23 Jul 2001 19:34:33 +0900 From: "Ogre" Organization: A poorly-installed InterNetNews site Message-ID: <9jgu8h$ov7$1@news.kornet.net> References: <9jgsvq$mqf$1@news.kornet.net> you checked routing table? and checked ping reply packet using packet capture program exactly in host computer ? and checked ping reply packet using etherhooker in your board? "ThisCase" wrote in message news:<9jgsvq$mqf$1@news.kornet.net>... > > Hi. > We are using powerPC860, vxWorks5.4, with END, ADS860 BSP. > target is connected to host directly (using cross ethernet cable) > target 192.168.0.85(our board), host 192.168.0.143(Windows2000 PC). > Here is my problem. > After booting my target system. > > -> ifShow > cpm (unit number 0): > Flags: (0x8063) UP BROADCAST MULTICAST ARP RUNNING > Type: ETHERNET_CSMACD > Internet address: 192.168.0.86 > Broadcast address: 192.168.0.255 > Netmask 0xffffff00 Subnetmask 0xffffff00 > Ethernet address is 08:00:3e:03:02:34 > Metric is 0 > Maximum Transfer Unit size is 1500 > 0 packets received; 1 packets sent > 0 multicast packets received > 0 multicast packets sent > 0 input errors; 0 output errors > 0 collisions; 0 dropped > lo (unit number 0): > Flags: (0x8069) UP LOOPBACK MULTICAST ARP RUNNING > Type: SOFTWARE_LOOPBACK > Internet address: 127.0.0.1 > Netmask 0xff000000 Subnetmask 0xff000000 > Metric is 0 > Maximum Transfer Unit size is 32768 > 0 packets received; 0 packets sent > 0 multicast packets received > 0 multicast packets sent > 0 input errors; 0 output errors > 0 collisions; 0 dropped > value = 29 = 0x1d > -> arpShow > > LINK LEVEL ARP TABLE > destination gateway flags Refcnt Use Interface > -------------------------------------------------------------------------- > -------------------------------------------------------------------------- > value = 75 = 0x4b = 'K' > > > But Pinging dose not work. > After pinging from target to host. > > -> ping("192.168.0.143") > PING 192.168.0.143: 56 data bytes > no answer from 192.168.0.143 > value = -1 = 0xffffffff = udbinfo + 0xfff172cf > -> arpShow > > LINK LEVEL ARP TABLE > destination gateway flags Refcnt Use Interface > -------------------------------------------------------------------------- > 192.168.0.143 00:01:02:4b:c3:bd 405 0 5 cpm0 > -------------------------------------------------------------------------- > value = 75 = 0x4b = 'K' > > > As shown above, arp table updated correctly. > we exam packet that flow from target to host, and find any problem. > > ethernet destination address : ff:ff:ff:ff:ff:ff > ethernet source address : 08:00:3e:03:02:34 > > Is it correct? > > I think that ethernet destination address is incorrect. > I don't know why this happen. > > when pinging from host to target, request time out has occurred. > ICMP packet from host to target is perfect.( i think so...) > > ICMP packet(Echo request, from host to target) has > ethernet destination address : 08:00:3e:03:02:34 > ethernet source address : 00:01:02:4b:c3:bd > > But. > ICMP packet(Echo reply, from target to host) has > ethernet destination address : ff:ff:ff:ff:ff:ff > ethernet source address : 08:00:3e:03:02:34 > > I think, every packet that initiated at target have > wrong ethernet destination address(ff:ff:ff:ff:ff:ff). > (address resolution with arp dosn't work...) > > Can anybody tell me what could be the problem? > > Thanks,. > > --------------------------- Newsgroups: comp.os.vxworks Subject: DOS FTP Date: Mon, 23 Jul 2001 10:39:37 +0000 (UTC) From: frankrae@post.com (frank rae) Organization: Mailgate.ORG Server - http://www.Mailgate.ORG Message-ID: <20010723103929.25982.qmail@mail.com> Can I use the DOS FTP commands MPUT *.*, MGET *.*, MDELETE *.* with an FTP server running on VxWorks? Regards, Frank Rae - -- _______________________________________________ FREE Personalized E-mail at Mail.com http://www.mail.com/?sr=signup Talk More, Pay Less with Net2Phone Direct(R), up to 1500 minutes free! http://www.net2phone.com/cgi-bin/link.cgi?143 - -- Posted from 205-158-62-49.outblaze.com [205.158.62.49] via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 23 05:22:04 2001 From: "chandrak" Date: Mon Jul 23 05:22:06 PDT 2001 Subject: code coverage tool can u tell some site for code coverage tool for vxworks with regards From vxwexplo-errs@csg.lbl.gov Tue Jul 24 04:03:32 2001 From: Vxworks Exploder Date: Tue Jul 24 04:03:34 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 24 04:03:09 PDT 2001 Subject: Re: Is Wind River really this pathetic Subject: Re: Recommend a quad serial I/O PMC Subject: Re: I have the Ethernet problem..... Subject: Re: I have the Ethernet problem..... Subject: .ZIP Utilities And VxWorks Subject: Re: semaphore in vxworks Subject: Re: .ZIP Utilities And VxWorks Subject: CrossWind, system mode, API_NOT_CONNECTED Subject: FTP Server Problems Subject: Re: .ZIP Utilities And VxWorks Subject: fork porting Subject: Re: .ZIP Utilities And VxWorks Subject: Tornado installation logs me off Subject: Re: Tornado installation logs me off Subject: dividing time between infinite loops Subject: Target Server Subject: Re: Timezone calc utility Subject: Re: I have the Ethernet problem..... Subject: Re: FTP Server Problems Subject: [Q] Codebase address in Flash.. Subject: Re: dividing time between infinite loops Subject: Re: semaphore in vxworks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is Wind River really this pathetic Date: 23 Jul 2001 09:01:37 -0400 From: Bill Pringlemeir Organization: Bell Nexxia Message-ID: References: <6qH17.79$5f6.133962@news.uswest.net> <3B540F05.BE9B26EB@sepro-robotique.com> <3B59D865.107CB71E@covad.net> Sender: bpringlemeir@DeadDuck >>>>> "Joe" == Joe Woodbury writes: Joe> That is precisely the problem. I had to spend four hours on Joe> Friday, rewriting ANSI C++ compliant code that GCC wouldn't Joe> compile but which did compile in both Visual C++ and Diab (the Joe> Diab compiler was being used with pSOS by a subsidiary before Joe> being shut down a few months back--we don't have a license for Joe> it.) It doesn't compile under GCC 2.95 either. I do empathize here. The Gnu C++ support that ships with Tornado is not good at all. Hopefully the GCC 3.0 C++ support will make it into a WRS product soon. WRS originally had support from Cygnus, but they stopped paying this. The version of GCC that has been shipping with the Tornado tools is rather archaic! My perspective is usually from pure `C' with some assembler (as I often write custom BSP's). In this case, gcc is much better than Diab or Green Hill's as many CPUs are supported and the linker and assembler are similar for all. This is very nice as it lets you change your processor to fit your application, which is often beneifical for price or power sensitive applications. I believe that DIAB compilers have been written with specific processors in mind. Compiler written like typically have generate great code for one processor and not so great code for others. GCC is fairly consistent in generating ok code across several processor families. If your target was the ARM, you probably wouldn't be interested in Diab. The ARM SDT provides the best code for that processor. Anyways, I hope that explains things a little better a long with the other posts... regards, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Recommend a quad serial I/O PMC Date: 23 Jul 2001 08:00:44 -0700 From: spk@uswest.net (Scott Kamara) Organization: http://groups.google.com/ Message-ID: <73e78f4e.0107230700.7e797594@posting.google.com> References: <9j102l$h72$1@trog.dera.gov.uk> <73e78f4e.0107220911.4aad055b@posting.google.com> spk@uswest.net (Scott Kamara) wrote in message news:<73e78f4e.0107220911.4aad055b@posting.google.com>... > > One needs to be careful when choosing a card in the case of > sync/framing protocols becuase the datasheet for the card may list protocols > supported,such as HDLC, but the vxWorks driver for the card does not drive > these protocols specifically. > The above needs clarification. I should have said that the driver *may* not drive these protocols or provide an interface to these protocols specifically. Often, datasheets will list on-card chip features but that doesnt mean the vxWorks driver for the card supports that feature. - -Scott --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I have the Ethernet problem..... Date: 23 Jul 2001 09:56:18 -0700 From: tlksk@hotmail.com (Kumar) Organization: http://groups.google.com/ Message-ID: References: <9jgsvq$mqf$1@news.kornet.net> <9jgu8h$ov7$1@news.kornet.net> Hi Ogre The arp packet are perfectly correct. There is no need to panic about that the destination address is "0xffffffff" the reason being that is the ethernet broadcast address. So when some on arps for other guys MAC address they send an broadcast to the whole network/subnet they are in. Kumar "Ogre" wrote in message news:<9jgu8h$ov7$1@news.kornet.net>... > you checked routing table? > and checked ping reply packet using packet capture program exactly in host > computer ? > and checked ping reply packet using etherhooker in your board? > > "ThisCase" wrote in message > news:<9jgsvq$mqf$1@news.kornet.net>... > > > > Hi. > > We are using powerPC860, vxWorks5.4, with END, ADS860 BSP. > > target is connected to host directly (using cross ethernet cable) > > target 192.168.0.85(our board), host 192.168.0.143(Windows2000 PC). > > Here is my problem. > > After booting my target system. > > > > -> ifShow > > cpm (unit number 0): > > Flags: (0x8063) UP BROADCAST MULTICAST ARP RUNNING > > Type: ETHERNET_CSMACD > > Internet address: 192.168.0.86 > > Broadcast address: 192.168.0.255 > > Netmask 0xffffff00 Subnetmask 0xffffff00 > > Ethernet address is 08:00:3e:03:02:34 > > Metric is 0 > > Maximum Transfer Unit size is 1500 > > 0 packets received; 1 packets sent > > 0 multicast packets received > > 0 multicast packets sent > > 0 input errors; 0 output errors > > 0 collisions; 0 dropped > > lo (unit number 0): > > Flags: (0x8069) UP LOOPBACK MULTICAST ARP RUNNING > > Type: SOFTWARE_LOOPBACK > > Internet address: 127.0.0.1 > > Netmask 0xff000000 Subnetmask 0xff000000 > > Metric is 0 > > Maximum Transfer Unit size is 32768 > > 0 packets received; 0 packets sent > > 0 multicast packets received > > 0 multicast packets sent > > 0 input errors; 0 output errors > > 0 collisions; 0 dropped > > value = 29 = 0x1d > > -> arpShow > > > > LINK LEVEL ARP TABLE > > destination gateway flags Refcnt Use Interface > > -------------------------------------------------------------------------- > > -------------------------------------------------------------------------- > > value = 75 = 0x4b = 'K' > > > > > > But Pinging dose not work. > > After pinging from target to host. > > > > -> ping("192.168.0.143") > > PING 192.168.0.143: 56 data bytes > > no answer from 192.168.0.143 > > value = -1 = 0xffffffff = udbinfo + 0xfff172cf > > -> arpShow > > > > LINK LEVEL ARP TABLE > > destination gateway flags Refcnt Use Interface > > -------------------------------------------------------------------------- > > 192.168.0.143 00:01:02:4b:c3:bd 405 0 5 cpm0 > > -------------------------------------------------------------------------- > > value = 75 = 0x4b = 'K' > > > > > > As shown above, arp table updated correctly. > > we exam packet that flow from target to host, and find any problem. > > > > ethernet destination address : ff:ff:ff:ff:ff:ff > > ethernet source address : 08:00:3e:03:02:34 > > > > Is it correct? > > > > I think that ethernet destination address is incorrect. > > I don't know why this happen. > > > > when pinging from host to target, request time out has occurred. > > ICMP packet from host to target is perfect.( i think so...) > > > > ICMP packet(Echo request, from host to target) has > > ethernet destination address : 08:00:3e:03:02:34 > > ethernet source address : 00:01:02:4b:c3:bd > > > > But. > > ICMP packet(Echo reply, from target to host) has > > ethernet destination address : ff:ff:ff:ff:ff:ff > > ethernet source address : 08:00:3e:03:02:34 > > > > I think, every packet that initiated at target have > > wrong ethernet destination address(ff:ff:ff:ff:ff:ff). > > (address resolution with arp dosn't work...) > > > > Can anybody tell me what could be the problem? > > > > Thanks,. > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I have the Ethernet problem..... Date: 23 Jul 2001 13:07:34 -0400 From: Bill Pringlemeir Organization: Bell Nexxia Message-ID: References: <9jgsvq$mqf$1@news.kornet.net> <9jgu8h$ov7$1@news.kornet.net> Sender: bpringlemeir@DeadDuck >> As shown above, arp table updated >> correctly. we exam packet that flow from target to host, and >> find any problem. ethernet destination address : >> ff:ff:ff:ff:ff:ff ethernet source address : 08:00:3e:03:02:34 > >> Is it correct? I think that ethernet destination address is >> incorrect. I don't know why this happen. >>>>> "Kumar" == Kumar writes: Kumar> Hi Ogre The arp packet are perfectly correct. There is no need Kumar> to panic about that the destination address is "0xffffffff" Kumar> the reason being that is the ethernet broadcast address. So Kumar> when some on arps for other guys MAC address they send an Kumar> broadcast to the whole network/subnet they are in. This is the correct behavior for an `ARP' packets. However, he has said that the ICMP packets from the target to the host contained all `FF's. This is not normal AFAIK. After the initial ARP request, both ends should use the MAC addresses of the source and destination. It is actually bad to use it for more than this as it will cause an interrupt on all nodes of the network to process the packet... just to see that it is not used. Also, a switch hub will not be able to `switch' these packets and every one will get them. It is definitely a problem... but I don't know what that problem is. regards, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: .ZIP Utilities And VxWorks Date: Mon, 23 Jul 2001 10:16:09 -0700 From: "AG" Organization: SBC Internet Services Message-ID: <__Y67.183$sX6.30506@news.pacbell.net> I would like to unzip a "img.zip" file on a vxworks target after zipping a bunch of files on my PC say, using WINZIP. I know that VxWorks has the "inflate()" utility but I don't know if it will inflate (or unzip) a .zip file. Does anybody know how to do this? I appreciate your comments. - -ag --------------------------- Newsgroups: comp.os.vxworks Subject: Re: semaphore in vxworks Date: 23 Jul 2001 10:28:01 -0700 From: tlksk@hotmail.com (Kumar) Organization: http://groups.google.com/ Message-ID: References: <3B59EA3A.70A66C47@yahoo.com> Well you should have a timeout mechanism. And if you are blocked on an Mutex you would have to wait till either you get it or timeout occurs. And I dont think there is any way of unblocking the tasks like you are talking about. Kumar Pierre Frigon wrote in message news:<3B59EA3A.70A66C47@yahoo.com>... > I assume that your tasks is blocked on a Mutex because it is waiting for > a resource. Why would you want to unblock a task in this situation? It > would be helpful to have more details. > > Saurabh Shrivastava wrote: > > > Hi!, > > is there a way to unblock task selectively (based on their task > > ids), tasks which are blocked on a mutex ? > > > > thanks! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: .ZIP Utilities And VxWorks Date: Mon, 23 Jul 2001 13:13:43 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B5C6957.AD8D8FF7@fnal.gov> References: <__Y67.183$sX6.30506@news.pacbell.net> It inflates a zlib compressed file ... you can have more info onzlib files at http://quest.jpl.nasa.gov/zlib/ use this compression it instead of winzip and inflate() should work ... hth AG wrote: > > I would like to unzip a "img.zip" file on a vxworks target after zipping a > bunch of files on my PC say, using WINZIP. I know that VxWorks has the > "inflate()" utility but I don't know if it will inflate (or unzip) a .zip > file. > > Does anybody know how to do this? I appreciate your comments. > > -ag --------------------------- Newsgroups: comp.os.vxworks Subject: CrossWind, system mode, API_NOT_CONNECTED Date: Mon, 23 Jul 2001 14:37:35 -0400 From: Michael Reed Message-ID: <3B5C6EEF.25B841D5@ll.mit.edu> Hello, TORNADO = Tor2.0.2 KERNEL = vxWorks5.4 HOST = Solaris2.6 BSP = mv2400 TARGET = mvme2400 CPU ARCH = ppc750 Seems I need to kill and relaunch CrossWind everytime before I download my app to the target (every time I reboot the target). If CrossWind was just launched, it'll download the app without problems. If CrossWind has been up for a previous download, and I rebooted the target, and then try to download the app again, CrossWind gives the following error: Target server load of a.out failed: API_NOT_CONNECTED Has anyone else encountered this problem? Thanks, Mike --------------------------- Newsgroups: comp.os.vxworks Subject: FTP Server Problems Date: Mon, 23 Jul 2001 15:04:19 -0500 From: "Joel Hannah" Organization: WebUseNet Corp. - "ReInventing The UseNet" Message-ID: I have developed a VxWorks application that requires the native VxWorks 5.4 FTP Server. We send and receive alot of large files (larger than 1MB) to and from the VxWorks target from external PCs. The VxWorks platforms is a Pentium II Class Single Board PC.After a few transfers of files like these, there is a REALLY LARGE slowdown in transfers both to and from the VxWorks target and sometimes transfers never complete thus hanging the FTP server.. In addition to FTP transactions, our application also has a custom server task that performs a TCP/IP data link operation to external PCs. I have tried quadrupling the NUM_SYS_64, 128, 256 and 512 block sizes thinking that the problem is a network buffer allocation problem but this doesn't seem to fix it. I would appreciate any help you guys could provide.... Joel - -- S. Joel Hannah Senior Engineer Alpha Beta Technologies, Inc 3112 12th Ave SW Huntsville, AL 35805 Phone: (256) 534-9067 Fax: (256) 534-9069 Pager: (256) 580-8947 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: .ZIP Utilities And VxWorks Date: Mon, 23 Jul 2001 13:06:13 -0700 From: "AG" Organization: SBC Internet Services Message-ID: References: <__Y67.183$sX6.30506@news.pacbell.net> <3B5C6957.AD8D8FF7@fnal.gov> This link comes from the VxWorsk manual but it doesn't work... "Dinker Charak" wrote in message news:3B5C6957.AD8D8FF7@fnal.gov... > It inflates a zlib compressed file ... you can have more info onzlib > files at > > http://quest.jpl.nasa.gov/zlib/ > > use this compression it instead of winzip and inflate() should work ... > > hth > > AG wrote: > > > > I would like to unzip a "img.zip" file on a vxworks target after zipping a > > bunch of files on my PC say, using WINZIP. I know that VxWorks has the > > "inflate()" utility but I don't know if it will inflate (or unzip) a .zip > > file. > > > > Does anybody know how to do this? I appreciate your comments. > > > > -ag > --------------------------- Newsgroups: comp.os.vxworks Subject: fork porting Date: Mon, 23 Jul 2001 23:15:35 +0100 From: DiStress Organization: FORTHnet S.A., Thetidos 6, GR-11528 Athens, Greece, Tel: +30 (1) 7295100, Fax: +30 (1) 7258520, url: http://www.forthnet.gr Message-ID: <3B5CA207.16757431@yahoo.com> Doed anyone know any nice way for porting applications that use fork.... into applications using taskSpawn or something familiar? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: .ZIP Utilities And VxWorks Date: Mon, 23 Jul 2001 15:23:30 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B5C87C2.E774C9AB@fnal.gov> References: <__Y67.183$sX6.30506@news.pacbell.net> <3B5C6957.AD8D8FF7@fnal.gov> well i had got it from google ... has done a search before answering you ... do a search for zlib on jpl.nasa.gov website and check if they have changed the link or something ... AG wrote: > > This link comes from the VxWorsk manual but it doesn't work... > > "Dinker Charak" wrote in message > news:3B5C6957.AD8D8FF7@fnal.gov... > > It inflates a zlib compressed file ... you can have more info onzlib > > files at > > > > http://quest.jpl.nasa.gov/zlib/ > > > > use this compression it instead of winzip and inflate() should work ... > > > > hth > > > > AG wrote: > > > > > > I would like to unzip a "img.zip" file on a vxworks target after zipping > a > > > bunch of files on my PC say, using WINZIP. I know that VxWorks has the > > > "inflate()" utility but I don't know if it will inflate (or unzip) a > .zip > > > file. > > > > > > Does anybody know how to do this? I appreciate your comments. > > > > > > -ag > > - -- Dinker Charak Computing Division/Online and Database Systems 630-840-8604 (office) Fermi National Accelerator Laboratory http://www.fnal.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado installation logs me off Date: Mon, 23 Jul 2001 15:29:20 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B5C8920.12015D6@fnal.gov> Hello, Am trying to install Tornado II on solaris server using my linux desk top ... but it log me off as soon as it shows the setup welcome dialog box ... it is still constructing the dialog box as the buttons are still not visible ... and being a windows user i find reboot a "normal behaviour" not logging off ... any ideas? - -- Dinker Charak Computing Division/Online and Database Systems 630-840-8604 (office) Fermi National Accelerator Laboratory http://www.fnal.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado installation logs me off Date: Mon, 23 Jul 2001 15:50:03 -0500 From: Dinker Charak Organization: Fermi National Accelerator Labortary Message-ID: <3B5C8DFB.30F02181@fnal.gov> References: <3B5C8920.12015D6@fnal.gov> I do not know what did it ... but I did a reboot ... used GNOME instead of KDE ... added another machine whose exported directory is being used ... but now the installation work ok ... Dinker Charak wrote: > > Hello, > > Am trying to install Tornado II on solaris server using my linux desk > top ... > > but it log me off as soon as it shows the setup welcome dialog box ... > it is still constructing the dialog box as the buttons are still not > visible ... > > and being a windows user i find reboot a "normal behaviour" not logging > off ... > > any ideas? > > -- > Dinker Charak > Computing Division/Online and Database Systems 630-840-8604 (office) > Fermi National Accelerator Laboratory http://www.fnal.gov - -- Dinker Charak Computing Division/Online and Database Systems 630-840-8604 (office) Fermi National Accelerator Laboratory http://www.fnal.gov --------------------------- Newsgroups: comp.os.vxworks Subject: dividing time between infinite loops Date: Sun, 22 Jul 2001 23:57:11 +0200 From: "dboy" Organization: NetVision Israel Message-ID: <9ji0e7$nu0$1@news.netvision.net.il> Hi, how can i divide the time between two infinite loops in vx when loop1 ignites loop2? after i call loop2 from loop1, i would never return to loop1 again. i would like to use the loops in a function as a task. 10X --------------------------- Newsgroups: comp.os.vxworks Subject: Target Server Date: Mon, 23 Jul 2001 23:17:29 +0200 From: "dboy" Organization: NetVision Israel Message-ID: <9ji0nn$nve$1@news.netvision.net.il> I ma running Vx on my win 2k machine with VxSim. I would like to configure a target server for my needs. But why can't i run the server and always get a BackEnd error? What do i have todo in order to run the target safely on my simulator? 10X --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Timezone calc utility Date: 23 Jul 2001 14:39:25 -0700 From: a1kapoor@yahoo.com (Ashish Kapoor) Organization: http://groups.google.com/ Message-ID: References: <9j9t0q$bro$1@news.storm.ca> Hi Yi, I was faced with the same problem a few months back. I was not able to come up with a solution and had to use a hack. Please let me knw if someone has mailed a solution to you. Thanks, Ashish ____________________________________________________________ "Yi Qin" wrote in message news:<9j9t0q$bro$1@news.storm.ca>... > Hi. > > Is there any timezone calculation utility I can find somewhere? > I am trying to make my software work with all the standard > timezones. The current difficulty is the calculation of the start > and end of the daylight saving time for each timezone. > > Any help will be rgeatly appreciated. > > Yi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I have the Ethernet problem..... Date: Mon, 23 Jul 2001 23:08:26 GMT From: "Girish V. Gulawani" Organization: Global Online Japan // Exodus Communications KK. Message-ID: References: <9jgsvq$mqf$1@news.kornet.net> i also had the same problem, to my surprise the ifShow was showing all the parameters correct except ethernet address. then when i did ifShow from the target shell and also the host shell. i found target shell was showing correct ethernet address and the host shell wasnt. dont know where is the problem. the ethernet driver as such works correctly. arp entries at the hosts are okay. "ThisCase" wrote in message news:9jgsvq$mqf$1@news.kornet.net... > > Hi. > We are using powerPC860, vxWorks5.4, with END, ADS860 BSP. > target is connected to host directly (using cross ethernet cable) > target 192.168.0.85(our board), host 192.168.0.143(Windows2000 PC). > Here is my problem. > After booting my target system. > > -> ifShow > cpm (unit number 0): > Flags: (0x8063) UP BROADCAST MULTICAST ARP RUNNING > Type: ETHERNET_CSMACD > Internet address: 192.168.0.86 > Broadcast address: 192.168.0.255 > Netmask 0xffffff00 Subnetmask 0xffffff00 > Ethernet address is 08:00:3e:03:02:34 > Metric is 0 > Maximum Transfer Unit size is 1500 > 0 packets received; 1 packets sent > 0 multicast packets received > 0 multicast packets sent > 0 input errors; 0 output errors > 0 collisions; 0 dropped > lo (unit number 0): > Flags: (0x8069) UP LOOPBACK MULTICAST ARP RUNNING > Type: SOFTWARE_LOOPBACK > Internet address: 127.0.0.1 > Netmask 0xff000000 Subnetmask 0xff000000 > Metric is 0 > Maximum Transfer Unit size is 32768 > 0 packets received; 0 packets sent > 0 multicast packets received > 0 multicast packets sent > 0 input errors; 0 output errors > 0 collisions; 0 dropped > value = 29 = 0x1d > -> arpShow > > LINK LEVEL ARP TABLE > destination gateway flags Refcnt Use Interface > -------------------------------------------------------------------------- > -------------------------------------------------------------------------- > value = 75 = 0x4b = 'K' > > > But Pinging dose not work. > After pinging from target to host. > > -> ping("192.168.0.143") > PING 192.168.0.143: 56 data bytes > no answer from 192.168.0.143 > value = -1 = 0xffffffff = udbinfo + 0xfff172cf > -> arpShow > > LINK LEVEL ARP TABLE > destination gateway flags Refcnt Use Interface > -------------------------------------------------------------------------- > 192.168.0.143 00:01:02:4b:c3:bd 405 0 5 cpm0 > -------------------------------------------------------------------------- > value = 75 = 0x4b = 'K' > > > As shown above, arp table updated correctly. > we exam packet that flow from target to host, and find any problem. > > ethernet destination address : ff:ff:ff:ff:ff:ff > ethernet source address : 08:00:3e:03:02:34 > > Is it correct? > > I think that ethernet destination address is incorrect. > I don't know why this happen. > > when pinging from host to target, request time out has occurred. > ICMP packet from host to target is perfect.( i think so...) > > ICMP packet(Echo request, from host to target) has > ethernet destination address : 08:00:3e:03:02:34 > ethernet source address : 00:01:02:4b:c3:bd > > But. > ICMP packet(Echo reply, from target to host) has > ethernet destination address : ff:ff:ff:ff:ff:ff > ethernet source address : 08:00:3e:03:02:34 > > I think, every packet that initiated at target have > wrong ethernet destination address(ff:ff:ff:ff:ff:ff). > (address resolution with arp dosn't work...) > > Can anybody tell me what could be the problem? > > Thanks,. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FTP Server Problems Date: Mon, 23 Jul 2001 23:25:20 GMT From: Harvey Taylor Organization: MTS Internet Message-ID: <3B5CCDB9.67FE@despam.autobahn.mb.ca> References: In article , Joel Hannah wrote: > > I have developed a VxWorks application that requires the native VxWorks 5.4 > FTP Server. We send and receive alot of large files (larger than 1MB) to and > from the VxWorks target from external PCs. The VxWorks platforms is a > Pentium II Class Single Board PC.After a few transfers of files like these, > there is a REALLY LARGE slowdown in transfers both to and from the VxWorks > target and sometimes transfers never complete thus hanging the FTP server.. > I do not currently have access to a VxWorks system, so this is going to be a little vague. I have seen a problem similar to this in the past on an x86 (5.3.1) system. The problem in our case was a reuse flag which is set in the net device definition string. Take a look through the network modules for that string and check for a reuse buffers flag. - -het - -- All human beings are born free and equal in dignity and rights. They are endowed with reason and conscience and should act towards one another in a spirit of brotherhood. UDHR:Article 1: http://www.autobahn.mb.ca/~het/udhr.html H.E. Taylor http://www.autobahn.mb.ca/~het/ --------------------------- Newsgroups: comp.os.vxworks Subject: [Q] Codebase address in Flash.. Date: Tue, 24 Jul 2001 13:06:05 +0900 From: "Ogre" Organization: A poorly-installed InterNetNews site Message-ID: <9jirs5$1t3$1@news.kornet.net> romStart() calculates Flash Address. I want to know why calculate Flash Address in that manner. Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dividing time between infinite loops Date: 24 Jul 2001 06:18:31 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9ji0e7$nu0$1@news.netvision.net.il> "dboy" wrote: > Hi, > > how can i divide the time between two infinite loops in vx when loop1 > ignites loop2? > > after i call loop2 from loop1, i would never return to loop1 again. > > i would like to use the loops in a function as a task. Add the statement taskDelay(0) to both loops. A taskDelay(0) does not create a delay, but the scheduler is activated. So on each call to taskDelay processing is switched from one loop to the other loop, or other processes at the same priority, if there is not higher priority task that needs to run. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: semaphore in vxworks Date: Tue, 24 Jul 2001 02:45:56 -0700 From: "Denis" Organization: University of California, Los Angeles Message-ID: <9jjg2b$hta$1@siamese.noc.ucla.edu> References: "Saurabh Shrivastava" wrote in message news:b194825.0107191851.421e967f@posting.google.com... > Hi!, > is there a way to unblock task selectively (based on their task > ids), tasks which are blocked on a mutex ? i seem to remember there was a way to change task's priority. however,i dont know if there was a way to set the priority of the task blocked on the semaphore, but if there was, you could just change that task's priority to something really high, and then semGive(), then change the priority back to something 'normal' for that task. this does not sound like a clean solution, but if this is the only one available, what the heck :) this is just a guess.... denis --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 24 07:16:15 2001 From: NAVEEN BHATIA Date: Tue Jul 24 07:16:17 PDT 2001 Subject: problem with setjmp and longjmp hi, im creating a downloadable project in VxWorks, which uses calls to setjmp and longjmp. i have included setjmp.h in my file. the project compiles and builds fine but when i download the project i get an error which reads:- "error while downloading - unresolved references _setjmp _longjmp". can anyone help? i need a solution urgently. im including the source as well: #include "vxworks.h" #include #include #include jmp_buf main_loop; void abort_to_main_loop (int status) { longjmp (main_loop, status); } int main (void) { while (1) if (setjmp (main_loop)) puts ("Back at main loop...."); else do_command (); } void do_command (void) { char buffer[128]; if (fgets (buffer, 128, stdin) == NULL) abort_to_main_loop (-1); else exit (EXIT_SUCCESS); } __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/ From vxwexplo-errs@csg.lbl.gov Tue Jul 24 08:34:45 2001 From: g parab Date: Tue Jul 24 08:34:47 PDT 2001 Subject: c++query... #include "iostream.h" class myclass { public: x() { cout<<"calling myclass constructor\n";} ~x() { cout<<"calling myclass destructor\n";} void print(){ cout<<"printing …printing…\n";} }; int run() { myclass mc1; mc1->print(); return 0; } I have downloaded the vxworks simulator version from net. While executing above code segment it doesn’t show any o/p on the terminal( target window ). It also doesn’t display any o/p on the windshell . what could be the problem . Is there any setting need to be done in vxworks for executing the c++ code. Note.that when the windshell start it says C++ constructor/destructor strategy is AUTOMATIC . Which is correct as I am creating a downloadable application. Thanks, parab From vxwexplo-errs@csg.lbl.gov Tue Jul 24 09:20:11 2001 From: Shlomo Gutman Date: Tue Jul 24 09:20:14 PDT 2001 Subject: Preformance Problem Hi , We use Tornado 2 / SENS 1.1 and we have a performance problem and we suspect on IP Stack / Net Task ( we use MIPS processor) . Do you know about patch for Net task which solved performance issue ? Any idea about Net Task / IP Stack performance Best regards, Gutman Shlomo. ****************************************************** Shlomo Gutman, V.P. of Engineering & CTO E_Mail : sgutman@globaloop.com Phone : +972-9-741-4441 Ext. 301 Fax : +972-9-741-4445 Mobile : +972-54-343-119 http://www.globaloop.com Address : GlobaLoop Ltd. 1 Hanagar st., Industrial zone Neve Ne'eman B, Hod-Hasharon, 45800 , ISRAEL *************************************************** From vxwexplo-errs@csg.lbl.gov Tue Jul 24 13:40:21 2001 From: Rajkumar Malli Date: Tue Jul 24 13:40:23 PDT 2001 Subject: Per machine tgtsvr Hi all, How do I configure a tgtsvr on a per machine basis (all logins on a PC) as opposed to per-login basis? I use Tornado 2.0 and VxWorks 5.4 Any input is appreciated. Thanks Raj From vxwexplo-errs@csg.lbl.gov Tue Jul 24 15:28:33 2001 From: "Ben Kelly" Date: Tue Jul 24 15:28:35 PDT 2001 Subject: RE: c++query... First off, I assume that the constructor and destructor being named "x" is a typo. Constructors and destructors must have the same name as the class. Second, where is stdout directed? I've never use the simulator before, but I believe if you attach to a normal target with windsh you need to setup a virtual console. I usually just telnet in and the outputs are redirected for me. In any case, cout should be going to stdout. Do printf's show up? If your goal is simply to know that the methods have run you can try updating a variable instead. For example: class myclass { public: myclass(int &i) : _i = i { _i += 1; } ~myclass() { _i += 1; } void print() { _i += 1; } private: int &_i; } int run() { int i = 0; { myclass m(i); // i + 1 = 1 m.print(); // i + 1 = 2 } // i + 1 = 3 return i; } If you execute run() from the windsh or target shell you should see the return value as 3. Finally, I believe that the AUTOMATIC constructor/destructor strategy really only refers to static objects. AUTOMATIC means to run the constructor automatically when the object file is loaded onto the target. Many times this is not a good idea since you don't actually want to start any processing until everything is loaded. I generally try to avoid static objects. Static references to objects are ok, but I have seen problems with static objects themselves. Hope that helps. - Ben vxworks > #include "iostream.h" > > class myclass > { public: x() { cout<<"calling myclass constructor\n";} > ~x() { cout<<"calling myclass destructor\n";} > void print(){ cout<<"printing ...printing...\n";} > }; > > int run() > { > myclass mc1; > mc1->print(); > return 0; > } > > I have downloaded the vxworks simulator version from net. While executing > above code segment it doesn't show any o/p on the terminal( target window ). > It also doesn't display any o/p on the windshell . what could be the problem . > Is there any setting need to be done in vxworks for executing the c++ code. > Note.that when the windshell start it says > C++ constructor/destructor strategy is AUTOMATIC . > Which is correct as I am creating a downloadable application. > > Thanks, > parab From vxwexplo-errs@csg.lbl.gov Wed Jul 25 03:41:20 2001 From: "Burkin, Valery" Date: Wed Jul 25 03:41:22 PDT 2001 Subject: 3C90x PCI on vxWorks Hi, I've configured vxworks for elPci (3C90x PCI NIC) interface, BUT can't ping anything.... except for itself. -> ifShow elPci (unit number 0): Flags: (0x8863) UP BROADCAST RUNNING ARP MULTICAST Type: ETHERNET_CSMACD Internet address: 172.16.98.26 Broadcast address: 172.16.98.255 Netmask 0xffff0000 Subnetmask 0xffffff00 Ethernet address is f0:02:1a:00:c7:d4 Metric is 0 Maximum Transfer Unit size is 1500 0 packets received; 6 packets sent 0 multicast packets received 0 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped -> routeShow ROUTE NET TABLE destination gateway flags Refcnt Use Interface ---------------------------------------------------------------------------- 172.16.98.0 172.16.98.26 101 0 0 elPci0 ---------------------------------------------------------------------------- -> muxShow Device: elPci Unit: 0 Description: 3COM 3c90X Fast Etherlink Endhanced Network Driver. Protocol: IP 4.4 ARP Type: 2054 Recv 0x134b9c Shutdown 0x134e0c Protocol: IP 4.4 TCP/IP Type: 2048 Recv 0x134b9c Shutdown 0x134d1c BUT!!!!!!!!! -> arpShow LINK LEVEL ARP TABLE destination gateway flags Refcnt Use Interface -------------------------------------------------------------------------- -------------------------------------------------------------------------- Anything wrong??? v. From vxwexplo-errs@csg.lbl.gov Fri Jul 20 01:25:17 2001 From: "Burkin, Valery" Date: Wed Jul 25 04:03:58 PDT 2001 Subject: EtherLink III ISA/PCI setup Hi, I've got problem configuring EtherLink III ISA or PCI cards under vxWorks. The basic target is Pentium PC... Actually I'm trying to get them up during boot stage in order to download kernel image from host via TFTP. config.h for ISA card: #define INCLUDE_NETWORK #define INCLUDE_ELT #undef IO_ADRS_ELT #define IO_ADRS_ELT 0x300 #undef INT_LVL_ELT #define INT_LVL_ELT 0x07 #undef CONFIG_ELT #define CONFIG_ELT 1 /* 0=EEPROM 1=AUI 2=BNC 3=RJ45 */ I/O base, IRQ level, transmiter type configured with DOS 3c5x9cfg. and tested under Win95/NT with those parameters. at the vxworks boot prompt: I try to get up elt0 interface with: "n elt" here's the output: Attaching network interface elt0... done. Error setting inet address of elt0 to 0, errno = 0 Cannot initialize interface named "elt0" in the boot line inet address: e=172.16.98.26, i.e. not null....... And...Oh, of course I cannot trace down this problem because of known reasons the only way is printf... Any help is welcome, valery. From vxwexplo-errs@csg.lbl.gov Wed Jul 25 04:03:56 2001 Received: from postal1.lbl.gov (postal1.lbl.gov [128.3.7.82]) by csg.lbl.gov (8.9.3/8.9.3) with ESMTP id EAA02904 for ; Wed, 25 Jul 2001 04:03:55 -0700 (PDT) Received: from SpamWall.lbl.gov (localhost [127.0.0.1]) by postal1.lbl.gov (8.11.2/8.11.2) with ESMTP id f6PB3hS15293 for ; Wed, 25 Jul 2001 04:03:45 -0700 (PDT) Received: from csg.lbl.gov (vx_explo@csg.lbl.gov [128.3.7.35]) by SpamWall.lbl.gov (8.11.2/8.11.2) with ESMTP id f6PB3FS15099 for ; Wed, 25 Jul 2001 04:03:21 -0700 (PDT) Received: (from vx_explo@localhost) by csg.lbl.gov (8.9.3/8.9.3) id EAA02898 for vxwexplo@lbl.gov; Wed, 25 Jul 2001 04:03:14 -0700 (PDT) Date: Wed, 25 Jul 2001 04:03:14 -0700 (PDT) From: Vxworks Exploder Message-Id: <200107251103.EAA02898@csg.lbl.gov> Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 25 04:03:13 PDT 2001 Subject: Re: Timezone calc utility Subject: Re: Timezone calc utility Subject: Re: 64-bit access and Exception Error Subject: Re: fork porting Subject: Re: [Q] Codebase address in Flash.. Subject: Anyone have experience with Galileo MPSC? Subject: Cache problems Subject: Re: Cache problems Subject: VxWorks on PPC-750 with cache / FAA Subject: Tcl problem in VxSim launch? Subject: select dsp for proiect multiprocessor Subject: simple tutorial? where Subject: Re: Tcl problem in VxSim launch? Subject: Serial Port Interface for Unix (Solaris) Subject: Re: Timezone calc utility Subject: Re: 64-bit access and Exception Error Subject: Re: Serial Port Interface for Unix (Solaris) Subject: wdbConfig error Subject: CS8900 driver - multiple units Subject: script file error in Tornado 1 Subject: Re: select dsp for proiect multiprocessor Subject: Re: script file error in Tornado 1 Subject: Re: simple tutorial? where Subject: Re: script file error in Tornado 1 Subject: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Subject: get_object_identifier_async() Subject: Re: VxWorks on PPC-750 with cache / FAA Subject: problem with booting vxworks as standalone Subject: Re: Cache problems ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Timezone calc utility Date: Tue, 24 Jul 2001 12:20:12 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: <0Kd77.559$bJ6.47436040@news1.mtl.metronet.ca> References: <9j9t0q$bro$1@news.storm.ca> enjoy !! { char charBuf[256]; ... getUTCOffset ( &offsetUTC ) ; /* get utc offset from nvram ... */ /* * If the environment variable TIMEZONE is set, the information is retrieved * from this variable, otherwise from the locale information. TIMEZONE is of the form: * * name_of_zone:<(unused)>:time_in_minutes_from_UTC:daylight_start:daylight_end * * To calculate local time, the value of time_in_minutes_from_UTC is subtracted from UTC; time_in_minutes_from_UTC * must be positive. * * Daylight information is expressed as mmddhh (month-day-hour), for example: * * UTC::0:040102:100102 * * Reference : vxWorks Reference Manual -- AnsiTime Library */ sprintf ( charBuf, "TIMEZONE=UTC::%d::", (int)(offsetUTC * 60.0) ); putenv ( charBuf ) ; } "Ashish Kapoor" wrote in message news:e3c6dafb.0107231339.d97a8cc@posting.google.com... > Hi Yi, > > I was faced with the same problem a few months back. I was > not able to come up with a solution and had to use a hack. > > Please let me knw if someone has mailed a solution to you. > > Thanks, > > Ashish > ____________________________________________________________ > > "Yi Qin" wrote in message news:<9j9t0q$bro$1@news.storm.ca>... > > Hi. > > > > Is there any timezone calculation utility I can find somewhere? > > I am trying to make my software work with all the standard > > timezones. The current difficulty is the calculation of the start > > and end of the daylight saving time for each timezone. > > > > Any help will be rgeatly appreciated. > > > > Yi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Timezone calc utility Date: Tue, 24 Jul 2001 12:22:09 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: References: <9j9t0q$bro$1@news.storm.ca> <0Kd77.559$bJ6.47436040@news1.mtl.metronet.ca> Be aware that getUTCOffset is a function that we wrote to save some data in the NVRam ... (We use a MBX860) "Steeve Gagne" wrote in message news:0Kd77.559$bJ6.47436040@news1.mtl.metronet.ca... > enjoy !! > > { > > char charBuf[256]; > > ... > > getUTCOffset ( &offsetUTC ) ; /* get utc offset from nvram ... */ > > /* > * If the environment variable TIMEZONE is set, the information is > retrieved > * from this variable, otherwise from the locale information. > TIMEZONE is of the form: > * > * > name_of_zone:<(unused)>:time_in_minutes_from_UTC:daylight_start:daylight_end > * > * To calculate local time, the value of > time_in_minutes_from_UTC is subtracted from UTC; time_in_minutes_from_UTC > * must be positive. > * > * Daylight information is expressed as mmddhh (month-day-hour), > for example: > * > * UTC::0:040102:100102 > * > * Reference : vxWorks Reference Manual -- AnsiTime Library > */ > > sprintf ( charBuf, "TIMEZONE=UTC::%d::", (int)(offsetUTC * > 60.0) ); > putenv ( charBuf ) ; > > } > > "Ashish Kapoor" wrote in message > news:e3c6dafb.0107231339.d97a8cc@posting.google.com... > > Hi Yi, > > > > I was faced with the same problem a few months back. I was > > not able to come up with a solution and had to use a hack. > > > > Please let me knw if someone has mailed a solution to you. > > > > Thanks, > > > > Ashish > > ____________________________________________________________ > > > > "Yi Qin" wrote in message > news:<9j9t0q$bro$1@news.storm.ca>... > > > Hi. > > > > > > Is there any timezone calculation utility I can find somewhere? > > > I am trying to make my software work with all the standard > > > timezones. The current difficulty is the calculation of the start > > > and end of the daylight saving time for each timezone. > > > > > > Any help will be rgeatly appreciated. > > > > > > Yi > > Steeve Gagne steeve.gagne@videotron.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 64-bit access and Exception Error Date: 24 Jul 2001 07:56:20 -0700 From: eric.e.lesch@lmco.com (Eric Lesch) Organization: http://groups.google.com/ Message-ID: References: <29f27f2c.0107230041.773b01df@posting.google.com> Hi, Is the instruction at 0x10000 a valid instruction? I had a similar problem the other day when I copied an image, but forgot to convert to a binary image. The ELF image I loaded was correct and it loaded into memory with no errors, but when trying to start it, I got the same error you got. My target was a PPC604 though. Make sure what you are loading is the correct format. Hope this helps, Eric Lesch --------------------------- Newsgroups: comp.os.vxworks Subject: Re: fork porting Date: Tue, 24 Jul 2001 12:41:17 +0100 From: vxworksintent@tao-group.com Organization: Tao Systems Ltd. Message-ID: References: <3B5CA207.16757431@yahoo.com> Sender: dg@pearl.tao.co.uk In article <3B5CA207.16757431@yahoo.com>, DiStress writes: > Doed anyone know any nice way for porting applications that use > fork.... into applications using taskSpawn or something familiar? No. We have the same problem with our own OS, intent. Basically, fork()'s semantics require duplicating the address space of the process that called fork(). On non-protected mode systems, like intent and VxWorks, this isn't possible, so you just have to work round it somehow. An example of the problem: say you have a program that uses a global variable. It forks. Each child should have its own *distinct* copy of that variable. On a protected-mode system, that's easy --- the address of the variable won't change; the MMU translates the two process' address space to different areas of physical RAM. On a non-protected system, you either have to update all references of the variable inside the programs code to point to a different memory location for one of the child processes, or jump through hoops. One technique the 68000 used to get around this was to indirect all memory accesses through a pointer register. So you could change the process' address space by just modifying this pointer. This was slow. Another technique was to copy the process' address space to a fixed location in RAM when scheduling it; this meant all processes ran at the same address. This was very slow. The conclusion we've come to is that there's no easy way to simulate fork(). You just have to rewrite your program logic to compensate. - -- David Given Tao Systems Ltd. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: [Q] Codebase address in Flash.. Date: 24 Jul 2001 11:07:33 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <9jirs5$1t3$1@news.kornet.net> Sender: bpringlemeir@DeadDuck >>>>> "Ogre" == Ogre writes: Ogre> romStart() calculates Flash Address. I want to know why Ogre> calculate Flash Address in that manner. "http://groups.google.com/groups?selm=u1yogtqwe.fsf%40yahoo.com" "http://groups.google.com/groups?selm=ulnfohf5v.fsf%40teklogix.com" "http://groups.google.com/groups?selm=uhf14dww5.fsf%40yahoo.com" "http://groups.google.com/groups?selm=ulnj887j0.fsf%40teklogix.com" Basically the fact is that the "compile/link address" and the "load/physical address" are at different places. This can further be complicated if the image that you are examining is a `ROM' loader that is moving code to RAM. Typical CPUs boot at either zero or 0xffff0000. Some code must execute here and transfer control to the address where the compiler expects the code to be. regards, Bill Pringlemeir. ps, Probably others have written better explainations at google. Please search around that sight. - -- Communication is not a bozo, I smell! Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Anyone have experience with Galileo MPSC? Date: Tue, 24 Jul 2001 11:46:39 -0400 From: Doug Dotson Organization: Verio Message-ID: <3B5D985E.4E4C0D3C@digidata.com> I'm working with a MIPS R7000 + Galileo GT-64240, trying to configure it to do a loopback test on the MSPC configured as a UART. Can't get a peep out of the thing. Looking at the BSP code it looks like I am initializing everything properly. Thanks, Doug --------------------------- Newsgroups: comp.os.vxworks Subject: Cache problems Date: Tue, 24 Jul 2001 16:46:40 GMT From: cgan76@nospam4me.com (cgan76) Organization: MTS Internet Message-ID: <3b5da4f4.956879249@200.123.101.100> Hello folks, We're currently having problems with turning on data cache on our boards (custom BSP). Looking over the BAT and memDesc, things seem to be okay. The problem is when we turn on data caching, the board hangs during the boot of the kernel. Anyone have any helpful hints on what could be causing this? Thanks! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cache problems Date: Tue, 24 Jul 2001 16:48:51 GMT From: cgan76@nospam4me.com (cgan76) Organization: MTS Internet Message-ID: <3b5da59d.957047711@200.123.101.100> References: <3b5da4f4.956879249@200.123.101.100> I forgot to mention, We are currently running an MPC8240 with 64MB of SDRAM. The board is a VME board (via Universe II). --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks on PPC-750 with cache / FAA Date: Tue, 24 Jul 2001 19:34:20 +0200 From: "Haim Kuper" Organization: NetVision Israel Message-ID: <9jk8l2$f6n$1@news.netvision.net.il> Gentlewomen & Gentlemen, I need to certify a PowerPC-750 platform, including internal & external CACHE as a LEVEL-A software for an FAA program. I am using VxWorks Tornado framework for the development. I've heard that because of the un-deterministic behaviour of a processor using a cache, I'll face problems in certifing this project. How can I tackle this situation ? --------------------------- Newsgroups: comp.os.vxworks Subject: Tcl problem in VxSim launch? Date: Tue, 24 Jul 2001 14:42:24 -0500 From: Dinker Charak Organization: Fermi National Accelerator Laboratory Message-ID: <3B5DCFA0.6FD77412@fnal.gov> Platform Solaris, Tornado II I start VxSim from launcher. On clicking on confirm, I get this error: Tcl Error unmatched open quote in list while executing "lsearch $line *vxWorks" (procedure "launchVxSimConfirm" line 39) invoked from within "launchVxSimConfirm" I opened the file $WIND_BASE/host/resource/app-config/launch/01VxSim.tcl and commented out this piece of code and it worked with out any trouble. this is standard WRS code freshly installed. since this piece checks if there is an another instance of VxSim running ... could do easily with out it. any tcl guru knowing what went wrong here. has someone else also faced/reported this problem? ################################################################################ # # launchVxSimConfirm - get options and execute the simulator. # # GLOBALS: (R) env, simPath # RETURN: n/a # ... while {![eof $fd]} { set line [gets $fd] # commented out this line onwards till ... # # if {[lsearch $line *vxWorks] != -1} { # set ix [lsearch $line "-p"] # if {$ix != -1} { # if {[lindex $line [incr ix]] == $procNum} { # set reply [noticePost question \ # "A VxWorks simulator $procNum is already running on this machine. #Connect to this simulator?" Yes No] # # if {$reply} { # set vxsimRunning 1 # break # } else { # close $fd # exec rm $tmpFile # return # } # } # } # } # # till here } Dinker Charak - --- Computing Division/Online and Database Systems 630-840-8604 (office) Fermi National Accelerator Laboratory http://www.fnal.gov --------------------------- Newsgroups: comp.os.vxworks Subject: select dsp for proiect multiprocessor Date: Tue, 24 Jul 2001 22:11:43 +0200 From: "federico torretta" Organization: Tiscali Spa Message-ID: <9jkknn$s5j$6@pegasus.tiscalinet.it> I have to plan a multiprocessor dsp and I wanted to know if you know some dsps floating point to make a will and with how much dsp can be interfacciati and as. thanks P.S. you excuse my very bad English natorret@tiscalinet.it --------------------------- Newsgroups: comp.os.vxworks Subject: simple tutorial? where Date: Tue, 24 Jul 2001 16:32:02 -0400 From: "newuser" Organization: Bell Sympatico Message-ID: Where can I find easy and simple tutorial for beginner. Please advice, I am looking for very simple introduction and I prefer with many visual example that might help to get the idea behind the VxWork. I did not find vxworks_guide is usuful... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tcl problem in VxSim launch? Date: Tue, 24 Jul 2001 20:47:22 GMT From: "Peter @ Home" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <3B5DCFA0.6FD77412@fnal.gov> Hi, Try this: Go to file host/resource/tcl/app-config/all/sun4-solaris2.tcl, you'll see two lines of code like set remoteShellCmd remsh set processListCmd "/usr/bin/ps -ef" change the second line to set processListCmd "/usr/bin/ps -ef | grep vxWorks". Instead of searching for all directories and files, now just look for vxWorks'. Peter. "Dinker Charak" wrote in message news:3B5DCFA0.6FD77412@fnal.gov... > Platform Solaris, Tornado II > > I start VxSim from launcher. On clicking on confirm, I get this error: > > Tcl Error > > unmatched open quote in list > while executing > "lsearch $line *vxWorks" > (procedure "launchVxSimConfirm" line 39) > invoked from within > "launchVxSimConfirm" > > I opened the file > > $WIND_BASE/host/resource/app-config/launch/01VxSim.tcl and commented out > this piece of code and it worked with out any trouble. this is standard > WRS code freshly installed. since this piece checks if there is an > another instance of VxSim running ... could do easily with out it. any > tcl guru knowing what went wrong here. > > has someone else also faced/reported this problem? > > > ############################################################################ #### > # > # launchVxSimConfirm - get options and execute the simulator. > # > # GLOBALS: (R) env, simPath > # RETURN: n/a > # > > ... > > while {![eof $fd]} { > > set line [gets $fd] > > # commented out this line onwards till ... > # > # if {[lsearch $line *vxWorks] != -1} { > # set ix [lsearch $line "-p"] > # if {$ix != -1} { > # if {[lindex $line [incr ix]] == $procNum} { > # set reply [noticePost question \ > # "A VxWorks simulator $procNum is already running on this > machine. #Connect to this simulator?" Yes No] > # > # if {$reply} { > # set vxsimRunning 1 > # break > # } else { > # close $fd > # exec rm $tmpFile > # return > # } > # } > # } > # } > # > # till here > > } > > > Dinker Charak > --- > Computing Division/Online and Database Systems 630-840-8604 (office) > Fermi National Accelerator Laboratory http://www.fnal.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Serial Port Interface for Unix (Solaris) Date: Tue, 24 Jul 2001 18:46:15 -0500 From: "Kevin Kulhanek" Organization: Lucent Technologies, Columbus, Ohio Message-ID: <9jl1lf$bej@nntpb.cb.lucent.com> Hi, We are being moved to Sun Boxes from our friendly Windows PC environment. I've installed Tornado on the Sun ok, but how do you configure the serial port to talk to my board and what terminal emulator should be used? In Windows, Hyperterminal was used. I know I can use the shell to talk to the board once vxWorks is up and running, but I really need the serial port to do the intial configuration of my hardware. Any info will be appreciated, Kevin Kulhanek kkulhanek@lucent.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Timezone calc utility Date: Wed, 25 Jul 2001 02:09:36 GMT From: yqin Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: <3B5E2A6E.6A22BCBE@home.com> References: <9j9t0q$bro$1@news.storm.ca> <0Kd77.559$bJ6.47436040@news1.mtl.metronet.ca> Thanks Steeve for the reply. I was able to use putenv() to set the local time from UTC. But I can only set it based on the current year if you look at the TIMEZONE string in the example. As the daylight saving time doesn't start or end on the calendar date, but rather starts and ends on the Mondays. So, the start and end values cannot always set using the same mmddhh values if a system is supposed to run for years. I think there might be a way to calculate this for different regions in the world. Any pointers or suggestion will be greatly appreciated. BTW, time_to_UTC_in_minutes can be negative if TZ is ahead of UTC. For example, -480 is Beijing, China. Yi Steeve Gagne wrote: > Be aware that getUTCOffset is a function that we wrote to save some > data in the NVRam ... (We use a MBX860) > > "Steeve Gagne" wrote in message > news:0Kd77.559$bJ6.47436040@news1.mtl.metronet.ca... > > enjoy !! > > > > { > > > > char charBuf[256]; > > > > ... > > > > getUTCOffset ( &offsetUTC ) ; /* get utc offset from nvram ... */ > > > > /* > > * If the environment variable TIMEZONE is set, the information is > > retrieved > > * from this variable, otherwise from the locale information. > > TIMEZONE is of the form: > > * > > * > > > name_of_zone:<(unused)>:time_in_minutes_from_UTC:daylight_start:daylight_end > > * > > * To calculate local time, the value of > > time_in_minutes_from_UTC is subtracted from UTC; time_in_minutes_from_UTC > > * must be positive. > > * > > * Daylight information is expressed as mmddhh > (month-day-hour), > > for example: > > * > > * UTC::0:040102:100102 > > * > > * Reference : vxWorks Reference Manual -- AnsiTime Library > > */ > > > > sprintf ( charBuf, "TIMEZONE=UTC::%d::", (int)(offsetUTC * > > 60.0) ); > > putenv ( charBuf ) ; > > > > } > > > > "Ashish Kapoor" wrote in message > > news:e3c6dafb.0107231339.d97a8cc@posting.google.com... > > > Hi Yi, > > > > > > I was faced with the same problem a few months back. I was > > > not able to come up with a solution and had to use a hack. > > > > > > Please let me knw if someone has mailed a solution to you. > > > > > > Thanks, > > > > > > Ashish > > > ____________________________________________________________ > > > > > > "Yi Qin" wrote in message > > news:<9j9t0q$bro$1@news.storm.ca>... > > > > Hi. > > > > > > > > Is there any timezone calculation utility I can find somewhere? > > > > I am trying to make my software work with all the standard > > > > timezones. The current difficulty is the calculation of the start > > > > and end of the daylight saving time for each timezone. > > > > > > > > Any help will be rgeatly appreciated. > > > > > > > > Yi > > > > > > Steeve Gagne > steeve.gagne@videotron.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 64-bit access and Exception Error Date: 24 Jul 2001 19:29:25 -0700 From: netproa@hanmail.net (IL HWAN, KIM) Organization: http://groups.google.com/ Message-ID: <29f27f2c.0107241829.1fc0aedd@posting.google.com> References: <29f27f2c.0107230041.773b01df@posting.google.com> eric.e.lesch@lmco.com (Eric Lesch) wrote in message news:... > Hi, > > Is the instruction at 0x10000 a valid instruction? I had a similar > problem the other day when I copied an image, but forgot to convert to > a binary image. The ELF image I loaded was correct and it loaded into > memory with no errors, but when trying to start it, I got the same > error you got. My target was a PPC604 though. Make sure what you are > loading is the correct format. > > Hope this helps, > Eric Lesch Thanks for your reply. I don't understand your advice. I use the vxWorks file made by Makefile or Tonado2.0 I don't know why convert the vxWorks file to a binary image. Please tell me about this. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Serial Port Interface for Unix (Solaris) Date: Tue, 24 Jul 2001 19:53:54 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B5E34C2.523863A3@covad.net> References: <9jl1lf$bej@nntpb.cb.lucent.com> Reply-To: drdiags@covad.net Kevin, I am no Sun guru, but I think there is an application called "tip", so maybe you can do a man on tip or an apropos (man -k) on "Serial". I also know there is a program like "cu" (connect unix?) but not sure about Sun. See what the FAQ has to offer (bottom of this post). Kevin Kulhanek wrote: > > Hi, > > We are being moved to Sun Boxes from our friendly Windows PC environment. > > I've installed Tornado on the Sun ok, but how do you configure the serial > port to > talk to my board and what terminal emulator should be used? > > In Windows, Hyperterminal was used. I know I can use the shell to talk to > the board > once vxWorks is up and running, but I really need the serial port to do the > intial configuration > of my hardware. > > Any info will be appreciated, > Kevin Kulhanek > kkulhanek@lucent.com - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: wdbConfig error Date: 24 Jul 2001 19:57:56 -0700 From: forddavi@msu.edu (Davis Ford) Organization: http://groups.google.com/ Message-ID: Hello, Other people seem to have had similar problems on this group. I'm curious if anyone has any helpful experience fixing the following wdbConfig and tRootTask errors. This is PPC603E, network driver is dec21x4xEnd.o provided by WRS - via download from windsurf. Bootrom code works ok, just when I load the image over the network, it fails to attach dc0. Any ideas? Thanks in advance, davis. Attached TCP/IP interface to dc0. Attaching network interface lo0... done. Loading... 1683840 Starting at 0x10000... Target Name: kahlua muxDevLoad failed for device entry 0! Attaching network interface dc0... wdbConfig: error configuring WDB communication interface Timebase: 8.250477 MHz, PCI: 16.500954 MHz, CPU: 99.5724 MHz SOC BIOS (VxWorks) MPC8240 - KAHLUA Board: Mousse, PLD rev: 1, Map B (CHRP) v5.4. Kernel: WIND version 2.5. Made on Jul 24 2001, 18:14:27. 0x3fffdf8 (tRootTask): dc0: not enough memory provided 0x3fffdf8 (tRootTask): dc0: not enough memory provided 0x3fffdf8 (tRootTask): muxLoad failed! Firmware Base: FIRMWARE1.1 Built Tue Jul 24 18:14:13 PDT 2001 Built by dford on devil1 (/home/sdk/work/sdk) Compiled for chip: BCM5680_B0 Found unit 0: BCM5680_B0 on bus 0, dev 0xa, func 0 Found unit 1: BCM5680_B0 on bus 0, dev 0x1c, func 0 Found unit 2: BCM5680_B0 on bus 0, dev 0x1d, func 0 Found unit 3: BCM5680_B0 on bus 0, dev 0x1e, func 0 Attaching default SOC unit 0... BCM5680_B0 PCI device 0 attached. BCM5680_B0 PCI device 0 initialized. GBP auto-sized to 0 MB, 0 banks MMU initialized ... General port parameters initialized ... Fast Ethernet MACs initialized ... Gigabit Ethernet MACs initialized ... MACs enabled. BCM driver initialized. VLAN 1 created. Set PVLAN for ports 0x80000ff to 1. PVLAN: All ports in VLAN 1 Port scan modes initialized ... --------------------------- Newsgroups: comp.os.vxworks Subject: CS8900 driver - multiple units Date: Wed, 25 Jul 2001 04:23:53 GMT From: "Shimon Gersten" Organization: Time Warner Message-ID: The CS8900 driver distributed with Tornado 2 (if_cs.c) for x86 CPU claims to only support a single unit. Is there a way to support more than one CS8900 device? - -- ~~~~~~~~~~~~~~~~~~~~~ Shimon Gersten shimon@san.rr.com ~~~~~~~~~~~~~~~~~~~~~ --------------------------- Newsgroups: comp.os.vxworks Subject: script file error in Tornado 1 Date: Wed, 25 Jul 2001 15:12:18 +0900 From: "Dae-il Kang" Organization: ONSE Telecom Message-ID: <9jlo4t$1vo$1@hiline.shinbiro.com> How are you, folks? I have a problem about script file in Tornado 1. I want to run a script file in startup script (s) of boot configuration. For example, - ------- test.txt ---------------- cd "d:/script" ls ld < hello.o hello - ------------------------------- When I run each line in the command line, It worked well. But when i run the test.txt in the startup script, It does not work. Test result is as follows: - --------------------------------------- Executing startup script bigone:d:/script/test.txt ... cd "d:/script" value = 0 = 0x0 ls hello.c hello.o rokmc.txt script-1.txt script.txt scsi.txt scsi_sd0.txt sub_test.txt test test.txt value = 0 = 0x0 ld < hello.o Error: File D:\script\hello.o does not exist 'an't open input 'hello.o errno = 0x226 - -------------------------------------- Please give me some advices and information. Thanks in advance. - ------------------------------ Dae-il, Kang Technical Support Dept. Tel)82-2-420-2343 (ext. 313) Fax)82-2-420-2212 support@acetronix.co.kr kangdaeil@acetronix.co.kr - ------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: select dsp for proiect multiprocessor Date: Wed, 25 Jul 2001 09:02:33 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <9jlqu4$9as$01$1@news.t-online.com> References: <9jkknn$s5j$6@pegasus.tiscalinet.it> Hello Frederico, try: http://www.windriver.com/products/html/virtuoso.html This is a multiprocessor RTOS for DSPs. Ciao Martin PS: Speaking for myself not for my employer "federico torretta" schrieb im Newsbeitrag news:9jkknn$s5j$6@pegasus.tiscalinet.it... > I have to plan a multiprocessor dsp and I wanted to know if you know some > dsps floating point to make a will and with how much dsp can be > interfacciati and as. > thanks > P.S. > you excuse my very bad English natorret@tiscalinet.it > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: script file error in Tornado 1 Date: Wed, 25 Jul 2001 09:14:28 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <3B5E71D4.633B92D4@ks.ericsson.se> References: <9jlo4t$1vo$1@hiline.shinbiro.com> Hi, Startup scripts run in the target shell and not in the host shell. Try changing 'cd "d:/script"' to 'cd "bigone:d:/script"'. I'm using the Solaris version of Tornado, so I don't have to care about drive letters and the '\' problems :-) HTH Urban Dae-il Kang wrote: [snip] > Executing startup script bigone:d:/script/test.txt ... > cd "d:/script" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: simple tutorial? where Date: 25 Jul 2001 07:13:23 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: "newuser" wrote: > Where can I find easy and simple tutorial for beginner. Please advice, > I am looking for very simple introduction and I prefer with many visual > example that might help to get the idea behind the VxWork. I did not > find vxworks_guide is usuful... Try these links: http://www.smartmodulartech.com/systems/library/docs/vxworks/vxintro/VXINTR O.HTM http://www.eventhelix.com/RealtimeMantra/default.htm Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: script file error in Tornado 1 Date: 25 Jul 2001 07:16:58 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9jlo4t$1vo$1@hiline.shinbiro.com> "Dae-il Kang" wrote: > ld < hello.o > Error: File D:\script\hello.o does not exist > 'an't open input 'hello.o > errno = 0x226 Did you try to use the full path with "/" instead of "\"? (so the command would be "ld < D:/script/hello.o" Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Date: 25 Jul 2001 01:26:17 -0700 From: fjpan@cmmail.com (Baumann) Organization: http://groups.google.com/ Message-ID: <6603907b.0107250026.718425d@posting.google.com> hi, i need the instructions sets of sa1100, and anyone know how to debug our programm via TCP under vxworks ? thanks in advance for any suggestions --------------------------- Newsgroups: comp.protocols.snmp,comp.os.psos,comp.os.vxworks Subject: get_object_identifier_async() Date: 25 Jul 2001 01:43:01 -0700 From: alexvn@bigfoot.com (Alex Vinokur) Organization: http://groups.google.com/ Message-ID: <15eacd17.0107250043.343b208c@posting.google.com> ============================ pRISM+ for pSOSystem/PowerPC pRISM+ 1.2.3 . pSOSystem 2.2.2 Diab Data D-CC Optimizing C compiler SNMP Release 7.4 (SNMPv1) Epilogue Technology Emissary SNMP MIB Compiler, version 6.3 Host : Windows98 Target : PowerPC (MPC860) MG-SOFT MIB Browser, Professional Edition 7.0.0.3730 ============================ Using the get_object_identifier_async() routine I have got the following results via MG-SOFT MIB Browser //------------------ 1 ------------------ OIDC_T snmp_product_id[] = { 1, // ISO 3, // ORG 6, // DOD 1, // INTERNET 4, // PRIVATE 1, // ENTERPRISES 12, // Epilogue Technology 7, // Products 1, // Unix Agent 3, // Version 3 0 // Subversion }; GetRequset : sysObjectID GetResponse is : sysObjectID.0 (object identifier) enterprises.12.7.1.3.0 //--------------------------------------- //------------------ 2 ------------------ scopusRoot OBJECT IDENTIFIER ::= { enterprises 4466 } scopusProducts OBJECT IDENTIFIER ::= { scopusRoot 1 } scopusEncoder OBJECT IDENTIFIER ::= { scopusProducts 5 } scopusE1000 OBJECT IDENTIFIER ::= { scopusEncoder 1 } OIDC_T snmp_product_id[] = { 1, // ISO 3, // ORG 6, // DOD 1, // INTERNET 4, // PRIVATE 1, // ENTERPRISES 4466, // Scopus Root 1, // Scopus Products 5, // Scopus Encoder 1, // Encoder E1000 0 // Subversion }; GetRequset : sysObjectID GetResponse is : sysObjectID.0 (object identifier) scopusE1000.0 //--------------------------------------- In the first case we have relative OID starting from enterprises : enterprises.12.7.1.3.0 In the second case we have relative the last node only : scopusE1000.0 Any suggestions? Thanks, ===================== Alex Vinokur mailto:alexvn@bigfoot.com http://up.to/alexvn http://go.to/alexv_math ===================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks on PPC-750 with cache / FAA Date: Wed, 25 Jul 2001 09:56:49 +0100 From: "Martin Dowie" Message-ID: <3b5e8725$1@pull.gecm.com> References: <9jk8l2$f6n$1@news.netvision.net.il> You will also face problems getting certification using the 'bog standard' version of VxWorks! You should talk to you're local Wind River rep about using 'VxWorks Cert' (a cut down VxWorks 5.4 for use in Level A systems) or think about changing your OS all together (e.g. Green Hills 'Integrity', LynxOS). Haim Kuper wrote in message news:9jk8l2$f6n$1@news.netvision.net.il... > Gentlewomen & Gentlemen, > > I need to certify a PowerPC-750 platform, including internal & external > CACHE > as a LEVEL-A software for an FAA program. I am using VxWorks Tornado > framework > for the development. > I've heard that because of the un-deterministic behaviour of a processor > using a cache, > I'll face problems in certifing this project. > How can I tackle this situation ? > > --------------------------- Newsgroups: comp.os.vxworks Subject: problem with booting vxworks as standalone Date: 25 Jul 2001 02:00:31 -0700 From: nav@sasken.com (naveen) Organization: http://groups.google.com/ Message-ID: <95ccca9f.0107250100.27328092@posting.google.com> Hi, I am developing a standalone vxworks application. Vxworks along with the application start automatically on power on from flash. I have selected vxworks_rom when building the image. I have the following problems: 1. While compiling the image i get the following warning "excess elements in symbol table". I have made symbol table as part of the image. The size of the symbol table is 3180 symbols. Is there any way by which i can increase the size of the symbol table array or is the size of the symbol table calculated at runtime by vxworks itself. Is there any upper limit on the size of symbol table. 2.After the vxworks image is built i burn the image into flash memory using jflash utility. vxworks then gets booted. During this stage the vxworks banner comes on the hyperterminal and there after it hangs. Control has not come uptil usrAppInit function where my application is spawned. Any pointers or solutions to the problem. Thanks B.Naveen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cache problems Date: Wed, 25 Jul 2001 09:06:39 GMT From: "dgold1" Organization: Road Runner Message-ID: References: <3b5da4f4.956879249@200.123.101.100> cgan76 wrote in message <3b5da4f4.956879249@200.123.101.100>... >Hello folks, > >We're currently having problems with turning on data cache on our >boards (custom BSP). Looking over the BAT and memDesc, things seem to >be okay. > >The problem is when we turn on data caching, the board hangs during >the boot of the kernel. > >Anyone have any helpful hints on what could be causing this? There could be many reasons for this, some of the more obvious things might be: 1. Have you tried just turning on I-cache first and see if that works ? 2. Are you sure you don't have any BATs overlapping each other or conflicting with addresses setup in memDesc ? 3. Are you sure your SDRAM setup (in the 8240 and the chips themselves) are correctly setup for burst ? Have you looked at a logic analyzer trace of a burst cycle yet ? 4. Do you know exactly where during the boot process your kernel hangs ? 5. Did you remember to allocate memory using cacheDmaMalloc() or use non-cacheable memory where appropriate (i.e. for tty buffers and network receive buffers if you have snooping turned off). 6. Is your SDRAM mapped with a BAT or memDesc ? If it's with a BAT, you can't use cacheDmaMalloc(). 7. Are you using external memory (instead of DPRAM) for your BDs, and it's cacheable ? (w/o snooping of course). hth, Dan Gold gold@ensemble.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 25 06:57:21 2001 From: Ilia Bosis Date: Wed Jul 25 06:57:23 PDT 2001 Subject: RE: 3C90x PCI on vxWorks From the ifShow, You can see that the packets reach the driver, (so the routing table is OK) but no receive packets exits from the driver. Looks as Your interface failed to transmit anything on the LAN or to receive from it. Check the cables, connections, hub, etc. Connect some network analyzer to see what's going on the LAN. If You can't, try at least to see the hub LED blinks when You ping from the target. Look at the ARP table on the machine You ping, or on the closest router if it's behind a router. If Your target is listed there with the correct MAC & IP addresses ? If yes - You succeeded to transmit correctly, but failed to receive. If not - Your target is failed to transmit or/and You possibly have some hardware (may be hub/cables/connectors) problem. Do Your target IP address (172.16.98.26) may be found in the target machine's ARP table BEFORE You ping it ? If yes - You have this address already used by someone else. All this in assumption that the driver in question is the off-the-shelf, reasonably debugged SW. If You is into the driver debugging - it's another, longer story. At least, check the WRS website for SPR's on this driver. I never used this specific piece of hardware. Hope, this helps Ilia -------------------------------------------- Ilia Bosis Software Engineer PacketLight Networks E-mail: Ilia_Bosis@packetlight.com Web: www.packetlight.com > -----Original Message----- > From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] > Sent: Wednesday, July 25, 2001 12:52 PM > To: vxworks_users@csg.lbl.gov > Subject: 3C90x PCI on vxWorks > > > Submitted-by vxwexplo-errs@csg.lbl.gov Wed Jul 25 03:41:20 2001 > Submitted-by: "Burkin, Valery" > > Hi, > > I've configured vxworks for elPci (3C90x PCI NIC) interface, > BUT can't ping anything.... except for itself. > > -> ifShow > elPci (unit number 0): > Flags: (0x8863) UP BROADCAST RUNNING ARP MULTICAST > Type: ETHERNET_CSMACD > Internet address: 172.16.98.26 > Broadcast address: 172.16.98.255 > Netmask 0xffff0000 Subnetmask 0xffffff00 > Ethernet address is f0:02:1a:00:c7:d4 > Metric is 0 > Maximum Transfer Unit size is 1500 > 0 packets received; 6 packets sent > 0 multicast packets received > 0 multicast packets sent > 0 input errors; 0 output errors > 0 collisions; 0 dropped > > -> routeShow > ROUTE NET TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > -------------- > 172.16.98.0 172.16.98.26 101 0 0 > elPci0 > -------------------------------------------------------------- > -------------- > > -> muxShow > Device: elPci Unit: 0 > Description: 3COM 3c90X Fast Etherlink Endhanced Network Driver. > Protocol: IP 4.4 ARP Type: 2054 Recv 0x134b9c > Shutdown 0x134e0c > Protocol: IP 4.4 TCP/IP Type: 2048 Recv 0x134b9c > Shutdown 0x134d1c > > BUT!!!!!!!!! > -> arpShow > LINK LEVEL ARP TABLE > destination gateway flags Refcnt Use > Interface > -------------------------------------------------------------- > ------------ > -------------------------------------------------------------- > ------------ > > Anything wrong??? > > v. From vxwexplo-errs@csg.lbl.gov Wed Jul 25 07:49:26 2001 From: "AMASS" Date: Wed Jul 25 07:49:28 PDT 2001 Subject: Testing Tools This is a multi-part message in MIME format. ------=_NextPart_000_000B_01C11520.AC7D15A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi All We are looking for a test tool to aid with as many facets of the testing = process as possible - script control, test case generation, code = coverage, etc. Can anyone recommend a product? We operate NT host and PowerPC target. Regards Andy vxworks ------=_NextPart_000_000B_01C11520.AC7D15A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi All
 
We are looking for a test tool to aid = with as many=20 facets of the testing process as possible - script control, test case=20 generation, code coverage, etc.
 
Can anyone recommend a product? We = operate NT host=20 and PowerPC target.
 
Regards
 
Andy
 
vxworks
------=_NextPart_000_000B_01C11520.AC7D15A0-- From vxwexplo-errs@csg.lbl.gov Thu Jul 26 02:58:17 2001 From: "Burkin, Valery" Date: Thu Jul 26 02:58:19 PDT 2001 Subject: RE: 3C90x PCI on vxWorks Hi Ilia, >From the ifShow, You can see that the packets reach the driver, >(so the routing table is OK) >but no receive packets exits from the driver. >Looks as Your interface failed to transmit anything on the LAN or to receive >from it. >Check the cables, connections, hub, etc. >Connect some network analyzer to see what's going on the LAN. >If You can't, try at least to see the hub LED blinks when You ping from the >target. First, I've tested this 3c905C-TX card/and cable connection with NT system installed on the same PC, it's WORKING fine... Pinging this PC from my Linux host and watching ARP table on Linux I've detected correct MAC address of this card 00:01:02:9D:D5:9F, but when I start vxWorks on that PC with el3c90xEnd driver I see it seemed to find this card, IP succesfully attached to MUX, but from ifShow output I see INCORRECT MAC address, AND moreover(!!) MAC address is CHANGED ALL THE TIME WHEN VXWORKS IS RESTARTED!!! I've taken that as the INCOMPATIBILITY of el3c90xEnd driver with 3c905C-TX board. That's clear the driver reads MAC ID from incorrect registers(!!!) Am I mistaken??? valery. From vxwexplo-errs@csg.lbl.gov Thu Jul 26 04:03:29 2001 From: Vxworks Exploder Date: Thu Jul 26 04:03:31 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 26 04:03:11 PDT 2001 Subject: Re: problem with booting vxworks as standalone Subject: Some VxWorks OS Capability Questions for those experts.... Subject: Re: simple tutorial? where Subject: Re: VxWorks on PPC-750 with cache / FAA Subject: Re: dividing time between infinite loops Subject: Re: Serial Port Interface for Unix (Solaris) Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Subject: Re: Cache problems Subject: Re: Tcl problem in VxSim launch? Subject: Re: Tornado installation logs me off Subject: Re: Tcl problem in VxSim launch? Subject: How to measure CPU utilization of a task at run time Subject: Re: Some VxWorks OS Capability Questions for those experts.... Subject: Route created through ICMP Redirect Subject: tasks switch timing Subject: clear command through 'pty' such as in unix ? Subject: Re: Tornado installation logs me off Subject: Re: DHCPc Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Subject: Re: fork porting Subject: MPC7450 support ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem with booting vxworks as standalone Date: 25 Jul 2001 10:09:36 -0400 From: Bill Pringlemeir Organization: Bell Nexxia Message-ID: References: <95ccca9f.0107250100.27328092@posting.google.com> Sender: bpringlemeir@DeadDuck >>>>> "naveen" == naveen writes: [snip] naveen> I have the following problems: naveen> 1. While compiling the image i get the following warning naveen> "excess elements in symbol table". naveen> I have made symbol table as part of the image. The size of naveen> the symbol table is 3180 symbols. Is there any way by which i naveen> can increase the size of the symbol table array or is the naveen> size of the symbol table calculated at runtime by vxworks naveen> itself. Is there any upper limit on the size of symbol table. [snip] naveen> Any pointers or solutions to the problem. $(WIND_BASE)/target/all/configAll.h #define SYM_TBL_HASH_SIZE_LOG2 8 #define STAT_TBL_HASH_SIZE_LOG2 6 $(WIND_BASE)/target/all/usrConfig.c #ifdef INCLUDE_STANDALONE_SYM_TBL sysSymTbl = symTblCreate (SYM_TBL_HASH_SIZE_LOG2, TRUE, memSysPartId); symTblCreate() is documented in symLib module (vxWorks Reference manual). In my image I have, ULONG standTblSize = 9796; Your problem could be that the symbol table array size and standTblSize do not match. Check out your symTbl.c file. This is a generated file. You might do a clean build and retry. hth, Bill Pringlemeir. - -- The reason why and sex are not food. Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Some VxWorks OS Capability Questions for those experts.... Date: 25 Jul 2001 14:32:56 GMT From: PC@User.com (PC User) Organization: ExecPC Internet - Milwaukee, WI Message-ID: <3b5ed897$0$93661$272ea4a1@news.execpc.com> Thanks for looking at these questions for me. I am trying to make a decision for selecting an RTOS and wanted to know the followings: 1. Can VxWorks allow the user to handle any and all hardware interrupts, including those reporting the detection of an error. 2. Can VxWorks provide services to install and remove user defined interrupt handlers. 3. Can VxWorks allow development of interrupt handlers that handle nested interrupts on the same or different levels. (How?) 4. Can VxWorks provide services to mask interrupts on a level by level basis. (An example would be nice). 5. Can VxWorks provide deterministic services to interrupts of different levels. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: simple tutorial? where Date: 25 Jul 2001 08:01:19 -0700 From: christian.delez@ascom.ch (Christian Delez) Organization: http://groups.google.com/ Message-ID: References: "newuser" wrote in message news:... > Where can I find easy and simple tutorial for beginner. Please advice, I am > looking for very simple introduction and I prefer with many visual example > that might help to get the idea behind the VxWork. I did not find > vxworks_guide is usuful... You can make the tutorial given with the Tornado Prototyper. It's a step by step on the Tools (IDE, debugger, ..) and the kernel features (message queues, tasks control, ...). go to http://www.wrs.com and link on Download Prototyper. It's simple and you see exactly what's appening. Christian. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks on PPC-750 with cache / FAA Date: Wed, 25 Jul 2001 16:03:19 +0100 From: Jason Garnham Organization: BAE SYSTEMS Message-ID: <3B5EDFB7.5B9D579@baesystems.com> References: <9jk8l2$f6n$1@news.netvision.net.il> I don't believe the cache exhibits non deterministic behaviour. It is just variable between two limits. Turning the cache off will force everything to take the longest time possible (or the logest path can be calculated). I don't know much about level-A cert, but it is my understanding that you have no chance with standard Tornado. Haim Kuper wrote: > Gentlewomen & Gentlemen, > > I need to certify a PowerPC-750 platform, including internal & external > CACHE > as a LEVEL-A software for an FAA program. I am using VxWorks Tornado > framework > for the development. > I've heard that because of the un-deterministic behaviour of a processor > using a cache, > I'll face problems in certifing this project. > How can I tackle this situation ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dividing time between infinite loops Date: Wed, 25 Jul 2001 15:37:00 +0000 (UTC) From: wbrown@brownie.als.lbl.gov (Bill Brown) Organization: Ernest Orlando Lawrence Berkeley National Laboratory Message-ID: <9jmp2s$ias$1@overload.lbl.gov> References: <9ji0e7$nu0$1@news.netvision.net.il> I'm not sure I understand the problem, but I'll give it a go! In article , Johan Borkhuis wrote: >"dboy" wrote: > >> Hi, >> >> how can i divide the time between two infinite loops in vx when loop1 >> ignites loop2? >> The implication here is that each loop would be a thread or "task" as it's called in vxWorks. Loop 1 can run as one task, and spawn loop 2 as another task OF THE SAME PRIORITY. This would usually happen at init time, but it wouldn't have to; it could happen whenever loop 1 is called for the first time. >> i would like to use the loops in a function as a task. I don't understand what this means. Johan has explained how the tasks can be made to alternate. >> after i call loop2 from loop1, i would never return to loop1 again. >> Is this a statement or a question? Loop 1 and loop 2 cannot really "call" each other, but could pass data back and forth by the use of some mechanism like message queues or pipes. Other tasks could also send/receive data to/from task1 and/or task 2 via similar mechanisms. > >Add the statement taskDelay(0) to both loops. A taskDelay(0) does not >create a delay, but the scheduler is activated. So on each call to >taskDelay processing is switched from one loop to the other loop, or other >processes at the same priority, if there is not higher priority task that >needs to run. > Clearly I don't understand the problem as described; I hope this is helpful nevertheless. - -bill, speaking only for myself. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Serial Port Interface for Unix (Solaris) Date: Wed, 25 Jul 2001 10:16:22 -0600 From: Don Small Organization: Sandia National Laboratories Message-ID: <3B5EF0D6.2CA9C440@sandia.gov> References: <9jl1lf$bej@nntpb.cb.lucent.com> <3B5E34C2.523863A3@covad.net> DrDiags is right. I use tip until I get it to boot and then proceed with rlogin to the target board. There is a file called /etc/remote which configures the serial port. There is a line (really a description name) in this file called hardwire. Hardwire defaults to /dev/term/b which is the port b serial port. It also defaults to 9600 baud (see br#9600). You can modify this file all you want. I normally copy this line to another so I don't screw up the original and then make my changes. See the man pages on tip for what all of the codes mean but I just use the ones for hardwire and modify the dv=/dev/term/? and baud rate if necessary. Heres a tip on tip. To get out of tip use the key strokes "~." (thats only tilde followed by period). It took a while to find that the first time. Here is a copy of hardwire and my new line called lgc for an example. hardwire:\ :dv=/dev/term/b:br#9600:el=^C^S^Q^U^D:ie=%$:oe=^D: lgc:\ :dv=/dev/term/a:br#9600:el=^C^S^Q^U^D:ie=%$:oe=^D: You will need to have root priviledge to modify /etc/remote, or get someone who does to allow you to modify the file (or change the priviledge). Don DrDiags wrote: > > Kevin, > > I am no Sun guru, but I think there is an application called > "tip", so maybe you can do a man on tip or an apropos (man -k) on > "Serial". > I also know there is a program like "cu" (connect unix?) but not > sure about Sun. See what the FAQ has to offer (bottom of this post). > > Kevin Kulhanek wrote: > > > > Hi, > > > > We are being moved to Sun Boxes from our friendly Windows PC environment. > > > > I've installed Tornado on the Sun ok, but how do you configure the serial > > port to > > talk to my board and what terminal emulator should be used? > > > > In Windows, Hyperterminal was used. I know I can use the shell to talk to > > the board > > once vxWorks is up and running, but I really need the serial port to do the > > intial configuration > > of my hardware. > > > > Any info will be appreciated, > > Kevin Kulhanek > > kkulhanek@lucent.com > > -- > http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home > page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Date: 25 Jul 2001 09:44:52 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107250844.50863a56@posting.google.com> References: <6603907b.0107250026.718425d@posting.google.com> fjpan@cmmail.com (Baumann) wrote in message news:<6603907b.0107250026.718425d@posting.google.com>... > hi, > i need the instructions sets of sa1100, and anyone know how to debug > our programm via TCP under vxworks ? > thanks in advance for any suggestions The instruction set of the SA-1100 is the same as all StrongARM chips, and some ARM chips - probably the best place to look is in the ARM Architectural Reference Manual, published by Prentice Hall, edited by Dave Jagger (ISBN 0 13 736299 4). The SA-1100 uses the version 4 instruction set, and does not support Thumb (the book I have contains v3, v4 and Thumb; newer ones might also include v5). As for debugging over a network, the Tornado debugger (Crosswind/gdb) will do this. Read the documentation that came with your Tornado package for information on setting up the target, a target server and starting the debugger. HTH, John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cache problems Date: Wed, 25 Jul 2001 17:44:40 GMT From: cgan76@nospam4me.com (cgan76) Organization: MTS Internet Message-ID: <3b5f02f7.1046477755@200.123.101.100> References: <3b5da4f4.956879249@200.123.101.100> Actually we've somewhat got around the problem. I don't quite understand it yet but maybe you can help. When I was doing my tests, to simplify things I had all portions of my memory was set to VMSTATE_CACHEABLE_NOT. And typically (from the template) the first portion of RAM from 0x0000_0000 to RAM_LOW_ADRS is always set to non-cacheable. But when I was debugging with an ICE, I noticed that after a crash there were valid pages in the data cache even in the portions of memory that I had marked non-cacheable (from 0x0000_0000 to LOCAL_MEM_SIZE. So just to test things I set all of this memory to cacheable and lo and behold everything works. Why VxWorks was trying to do cache processes on memory space I had marked non-cacheable I don't know yet and I will have to investigate later. On Wed, 25 Jul 2001 09:06:39 GMT, "dgold1" wrote: > >cgan76 wrote in message <3b5da4f4.956879249@200.123.101.100>... >>Hello folks, >> >>We're currently having problems with turning on data cache on our >>boards (custom BSP). Looking over the BAT and memDesc, things seem to >>be okay. >> >>The problem is when we turn on data caching, the board hangs during >>the boot of the kernel. >> >>Anyone have any helpful hints on what could be causing this? > > >There could be many reasons for this, some of the more obvious things might >be: >1. Have you tried just turning on I-cache first and see if that works ? >2. Are you sure you don't have any BATs overlapping each other or >conflicting with addresses setup in memDesc ? >3. Are you sure your SDRAM setup (in the 8240 and the chips themselves) are >correctly setup for burst ? Have you looked at a logic analyzer trace of a >burst cycle yet ? >4. Do you know exactly where during the boot process your kernel hangs ? >5. Did you remember to allocate memory using cacheDmaMalloc() or use >non-cacheable memory where appropriate (i.e. for tty buffers and network >receive buffers if you have snooping turned off). >6. Is your SDRAM mapped with a BAT or memDesc ? If it's with a BAT, you >can't use cacheDmaMalloc(). >7. Are you using external memory (instead of DPRAM) for your BDs, and it's >cacheable ? (w/o snooping of course). > > hth, > Dan Gold gold@ensemble.com > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tcl problem in VxSim launch? Date: Wed, 25 Jul 2001 13:34:20 -0500 From: Dinker Charak Organization: Fermi National Accelerator Laboratory Message-ID: <3B5F112C.F18F8F6A@fnal.gov> References: <3B5DCFA0.6FD77412@fnal.gov> It does work. But am wondering is the problem due to something in my set up do all face this problem? Did you too face this problem? Did you report it? Dinker "Peter @ Home" wrote: > > Hi, > Try this: > > Go to file host/resource/tcl/app-config/all/sun4-solaris2.tcl, you'll > see two lines of code like > > set remoteShellCmd remsh > set processListCmd "/usr/bin/ps -ef" > > change the second line to > > set processListCmd "/usr/bin/ps -ef | grep vxWorks". > > Instead of searching for all directories and files, now just look for > vxWorks'. > > Peter. > > "Dinker Charak" wrote in message > news:3B5DCFA0.6FD77412@fnal.gov... > > Platform Solaris, Tornado II > > > > I start VxSim from launcher. On clicking on confirm, I get this error: > > > > Tcl Error > > > > unmatched open quote in list > > while executing > > "lsearch $line *vxWorks" > > (procedure "launchVxSimConfirm" line 39) > > invoked from within > > "launchVxSimConfirm" > > > > I opened the file > > > > $WIND_BASE/host/resource/app-config/launch/01VxSim.tcl and commented out > > this piece of code and it worked with out any trouble. this is standard > > WRS code freshly installed. since this piece checks if there is an > > another instance of VxSim running ... could do easily with out it. any > > tcl guru knowing what went wrong here. > > > > has someone else also faced/reported this problem? > > > > > > > ############################################################################ > #### > > # > > # launchVxSimConfirm - get options and execute the simulator. > > # > > # GLOBALS: (R) env, simPath > > # RETURN: n/a > > # > > > > ... > > > > while {![eof $fd]} { > > > > set line [gets $fd] > > > > # commented out this line onwards till ... > > # > > # if {[lsearch $line *vxWorks] != -1} { > > # set ix [lsearch $line "-p"] > > # if {$ix != -1} { > > # if {[lindex $line [incr ix]] == $procNum} { > > # set reply [noticePost question \ > > # "A VxWorks simulator $procNum is already running on this > > machine. #Connect to this simulator?" Yes No] > > # > > # if {$reply} { > > # set vxsimRunning 1 > > # break > > # } else { > > # close $fd > > # exec rm $tmpFile > > # return > > # } > > # } > > # } > > # } > > # > > # till here > > > > } > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado installation logs me off Date: Wed, 25 Jul 2001 13:37:56 -0500 From: Dinker Charak Organization: Fermi National Accelerator Laboratory Message-ID: <3B5F1204.1AD165DE@fnal.gov> References: <3B5C8920.12015D6@fnal.gov> <3B5C8DFB.30F02181@fnal.gov> I narrowed it down to this ... if we run the installation (linux xserver, am using my linux box to run install Tornado II on solaris) in GNOME, all is OK ... but it logs off the user in KDE ... Anyone any idea what is happening? maybe it is a x problem and linux is a better newsgroup to ask this Q ... Dinker Dinker Charak wrote: > > I do not know what did it ... but I did a reboot ... used GNOME instead > of KDE ... added another machine whose exported directory is being used > ... > > but now the installation work ok ... > > Dinker Charak wrote: > > > > Hello, > > > > Am trying to install Tornado II on solaris server using my linux desk > > top ... > > > > but it log me off as soon as it shows the setup welcome dialog box ... > > it is still constructing the dialog box as the buttons are still not > > visible ... > > > > and being a windows user i find reboot a "normal behaviour" not logging > > off ... > > > > any ideas? > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tcl problem in VxSim launch? Date: Wed, 25 Jul 2001 19:38:09 GMT From: "Peter @ Home" Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <3B5DCFA0.6FD77412@fnal.gov> <3B5F112C.F18F8F6A@fnal.gov> Hi, This was a known problem in the script file. The solution below is the one published by Wind River Tech Support. Regards, P. "Dinker Charak" wrote in message news:3B5F112C.F18F8F6A@fnal.gov... > It does work. But am wondering is the problem due to something in my set > up do all face this problem? Did you too face this problem? Did you > report it? > > Dinker > > "Peter @ Home" wrote: > > > > Hi, > > Try this: > > > > Go to file host/resource/tcl/app-config/all/sun4-solaris2.tcl, you'll > > see two lines of code like > > > > set remoteShellCmd remsh > > set processListCmd "/usr/bin/ps -ef" > > > > change the second line to > > > > set processListCmd "/usr/bin/ps -ef | grep vxWorks". > > > > Instead of searching for all directories and files, now just look for > > vxWorks'. > > > > Peter. > > > > "Dinker Charak" wrote in message > > news:3B5DCFA0.6FD77412@fnal.gov... > > > Platform Solaris, Tornado II > > > > > > I start VxSim from launcher. On clicking on confirm, I get this error: > > > > > > Tcl Error > > > > > > unmatched open quote in list > > > while executing > > > "lsearch $line *vxWorks" > > > (procedure "launchVxSimConfirm" line 39) > > > invoked from within > > > "launchVxSimConfirm" > > > > > > I opened the file > > > > > > $WIND_BASE/host/resource/app-config/launch/01VxSim.tcl and commented out > > > this piece of code and it worked with out any trouble. this is standard > > > WRS code freshly installed. since this piece checks if there is an > > > another instance of VxSim running ... could do easily with out it. any > > > tcl guru knowing what went wrong here. > > > > > > has someone else also faced/reported this problem? > > > > > > > > > > > ############################################################################ > > #### > > > # > > > # launchVxSimConfirm - get options and execute the simulator. > > > # > > > # GLOBALS: (R) env, simPath > > > # RETURN: n/a > > > # > > > > > > ... > > > > > > while {![eof $fd]} { > > > > > > set line [gets $fd] > > > > > > # commented out this line onwards till ... > > > # > > > # if {[lsearch $line *vxWorks] != -1} { > > > # set ix [lsearch $line "-p"] > > > # if {$ix != -1} { > > > # if {[lindex $line [incr ix]] == $procNum} { > > > # set reply [noticePost question \ > > > # "A VxWorks simulator $procNum is already running on this > > > machine. #Connect to this simulator?" Yes No] > > > # > > > # if {$reply} { > > > # set vxsimRunning 1 > > > # break > > > # } else { > > > # close $fd > > > # exec rm $tmpFile > > > # return > > > # } > > > # } > > > # } > > > # } > > > # > > > # till here > > > > > > } > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: How to measure CPU utilization of a task at run time Date: Wed, 25 Jul 2001 12:49:03 -0700 From: Sunil Kulkarni Organization: Infosys Technologies Ltd. Message-ID: <3B5F22AE.70C159C1@ntl.inf.com> How to measure the CPU utilization of a task in runtime so that some action can be taken based on the % CPU utilization of the task. Thanks, Sunil --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Some VxWorks OS Capability Questions for those experts.... Date: Wed, 25 Jul 2001 12:24:07 -0700 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <3B5F1CD7.D6D9C634@lmco.com> References: <3b5ed897$0$93661$272ea4a1@news.execpc.com> Most of the answers would be platform dependent. I'm not all that familiar with X86, which you may be using. 1. Not all. For example, I don't think that you could capture machine check without corrupting the kernel. 2. Install, yes. Remove, not directly. Workarounds exist. There is normally no more than 10 lines of code in a properly written int handler for vxworks. 3. Not on PowerPC, since the hardware doesn't allow it. Other platforms, such as 68K, yes. I don't know about X86. 4. Platform and BSP dependent. Some BSPs support sysIntEnable, which masks interrupts by line (IRQ) number. 5. You can get the system set up so that a specific maximum delay exists, if that's what you mean. There can always be some variability, although it doesn't have to be larger than 10s of microseconds. In short, you need to talk to a specific machine and a BSP to get fairly usefull answers. Speaking only for myself, Joe Durusau PC User wrote: > > Thanks for looking at these questions for me. I am trying to make a decision > for selecting an RTOS and wanted to know the followings: > > 1. Can VxWorks allow the user to handle any and all hardware interrupts, > including those reporting the detection of an error. > > 2. Can VxWorks provide services to install and remove user defined interrupt > handlers. > > 3. Can VxWorks allow development of interrupt handlers that handle nested > interrupts on the same or different levels. (How?) > > 4. Can VxWorks provide services to mask interrupts on a level by level basis. > (An example would be nice). > > 5. Can VxWorks provide deterministic services to interrupts of different > levels. --------------------------- Newsgroups: comp.os.vxworks Subject: Route created through ICMP Redirect Date: Wed, 25 Jul 2001 17:06:33 -0400 From: "Michael Yeung" Organization: AT&T Canada IES Message-ID: Hi all, Does VxWorks associate timer to route that is created through ICMP redirect? If not, the system will be hit with the routing black hole situation upon a dead gateway. Based on RFC1122, there are few remedies suggested but somehow I don't think VxWorks has done any of them. Please help!! thanx, Michael --------------------------- Newsgroups: comp.os.vxworks Subject: tasks switch timing Date: 25 Jul 2001 14:20:22 -0700 From: jpriss@hotmail.com (jean pascal riss) Organization: http://groups.google.com/ Message-ID: <1ba7ad74.0107251320.676cb668@posting.google.com> hi, I'have done some tasks swithing timing with the following code. 2 process are created they are synchronized on 2 semaphores (one process give a sem that allows the other process to continu and so on). one process start a LED the other one Stop it and that's all. I measured that a sem operation(take/give)is about 25us. on a scope it gives me a signal with a period of 300us. minus 4 sem operations and divide by 2, it gives 100us for task switch. no others tasks are running. so first, is this methods good? secondly, 100us seems to be huge to compare with some other timing I found on the news group. the system is a ARM7 30Mhz, without cache and running in a thumb mode. Is this time normal? what can be the source of such a delay? (how to improve?) thanks, jean pascal riss --------------------------- Newsgroups: comp.os.vxworks Subject: clear command through 'pty' such as in unix ? Date: Thu, 26 Jul 2001 09:45:58 +0900 From: "ozlab" Organization: A poorly-installed InterNetNews site Message-ID: <9jnouu$c94$1@news.kornet.net> Is there 'clear' command in vxworks library such as in any other unix? How do I start if it doesn't exist ? ====================================== Eung-Shik Kim (Ed) ed@ebridgecom.com PowerVoIP Gateway Hardware Devl. Dept. 467-12 Dogok-Dong Kangnam-Gu MMAA Bldg 29th Fl. Seoul (zip 135-270) KOREA, Republic of Tel+82 2 575 6476 ext 517, 3713 Fax+82 2 575 6478 C.P. +82 11 9704 0840 ICQ#61658172 / Send me plain text please... ======================================= --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado installation logs me off Date: 25 Jul 2001 20:43:34 -0700 From: aiox@my-deja.com (WakeUp) Organization: http://groups.google.com/ Message-ID: <157a795d.0107251943.4bbec50@posting.google.com> References: <3B5C8920.12015D6@fnal.gov> I ran into similar problems. Install on Solaris through Exceed xterm on my NT box - crash Exceed Install on Solaris through Gnome/Sawmill xterm on my RedHat 6.2 Linux box - installer errors out. Install on Solaris 8 via its KDE console - installer errors out. Finally I logged onto the console of a Solaris 2.7 box, running KDE, and THAT installed, but by then however, I had already pursued planB: install and develop on Winblows NT. I spent very little time trying to track down why installation was so difficult. I found some commands in the installation scripts that attempted to manipulate the Xserver or tried to get colormap or something. Didn't dig too deep. Everything was in TCL, afterall. :P I can't believe that a simple installer would be so X specific. Give me the good-ole days when everything was run from the command line! Since that episode, I've learned set my expectations lower as far as the Tornado toolset is concerned, and so far, I haven't been disappointed. AIOX Dinker Charak wrote in message news:<3B5C8920.12015D6@fnal.gov>... > Hello, > > Am trying to install Tornado II on solaris server using my linux desk > top ... > > but it log me off as soon as it shows the setup welcome dialog box ... > it is still constructing the dialog box as the buttons are still not > visible ... > > and being a windows user i find reboot a "normal behaviour" not logging > off ... > > any ideas? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: DHCPc Date: 25 Jul 2001 21:58:00 -0700 From: thedoctor68467@yahoo.com (DoctorDoctor) Organization: http://groups.google.com/ Message-ID: References: <9je3kb$e2e$1@news.inter.net.il> Pavius... To answer your questions: 1) Get a packet sniffer......like ethereal.......for your host machine..... 2) The help resource is WindSurf.....you can find it on the support page at windriver.com. You'll need your license number to register. 3) T2 DHCP uses a Snarf protocol for packets, T2.0.2 uses Berkley Packet Filters. 4) You can also use etherInputHookAdd, which is what DHCP uses, but it allows multiple entries as opposed to adding a Snarf protocol. But I think they plan on making the etherHooks obsolete. Just out of curiousity...why are you writing your own DHCP client, when one is included with vxWorks? "Pavius" wrote in message news:<9je3kb$e2e$1@news.inter.net.il>... > Hi. > > I'm trying to implement a dhcp client for a target running vxworks. The code > is extremely simple, compiles and links fine. The problem is it doesnt > work... Now, since there are a huge amount of "black boxes" in getting > config'ed by a remote server (do i send the request packets, are they valid, > do i get a response from the server etc) the only way i thought of debugging > it was by implementing a simple sniffer which would at least show me the > outgoing / incoming packets so i might understand the problem. I wrote > something simple implemented as a SNARF protocol bound to the mux which runs > fine if i use a target with no DHCPc but doesnt run on those who do. Since i > am running Tornado 2 i cannot bind more than one SNARF protocol so i > immediately suspected the DHCP client has attached its own filter. calling > muxShow(0, 0) displays all the bound protocols, and as suspected there was a > snarf protocol (type 257) named "etherInputHook". The manual says something > about DHCP requiring the berkly packet filter library but i didnt find any > help on that (ill look more into that now). > > 1) how would you go about debugging such a problem? > 2) is there any hidden help resource i am missing? i am used to having tons > of information being a win32 programmer (such as the MSDN) and its quite > hard diving into such uncharted (?) territories. > 3) Why would DHCP need a packet filter? RFC 2131 (page 10) hints that the > DHCP client is implemented above the IP stack... > 4) how can i access incoming / outgoing packets without binding a new > protocol to the mux ? > > Thanks in advance --------------------------- Newsgroups: comp.os.vxworks Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Date: 25 Jul 2001 23:30:03 -0700 From: fjpan@cmmail.com (Baumann) Organization: http://groups.google.com/ Message-ID: <6603907b.0107252230.72ebe770@posting.google.com> References: <6603907b.0107250026.718425d@posting.google.com> <488e459a.0107250844.50863a56@posting.google.com> thanks ya,i will buy the Manual . i have the GNUPRO TOOLKiT , strongarm-elf-gdb seems only supports the serial debug, anyone knows how to connect the target via the TCP ? and how can i open a PORT for debugging? i have scanned the target when it launched , but i can not find it opens any port, how the TORNADO connect the target server? john_94501@yahoo.com (John) wrote in message news:<488e459a.0107250844.50863a56@posting.google.com>... > As for debugging over a network, the Tornado debugger (Crosswind/gdb) > will do this. Read the documentation that came with your Tornado > package for information on setting up the target, a target server and > starting the debugger. > > HTH, > > John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: fork porting Date: Thu, 26 Jul 2001 01:00:51 -0700 From: "Joe Z Chung" Organization: SBC Internet Services Message-ID: References: <3B5CA207.16757431@yahoo.com> I'm assuming you're taking about porting a Unix(like) application to VxWorks 5.x. First convert all processes into threads. That should iron out a lot of issues, such as: 1. Global namespace collision - global variable names and function names now all exist in one address space. 2. Process reentrancy - the same function that is called in two separate processes may not be reentrant at all when called by two threads within the same process. 3. Interprocess communication - shared memory and other cumbersome methods of cross-process communication and synchronization may become unnecessary. These links are good introductions. http://www.windriver.com/windword/html/8.rehosting.html http://www.windriver.com/windword/html/10.rehosting.html Think of VxWorks as just a single process - the kernel tasks and your application tasks are just threads in that process. After you have converted your multi-process application to a single-process application running multiple-threads, then it should be easier to port to VxWorks. - -jc "DiStress" wrote in message news:3B5CA207.16757431@yahoo.com... > Doed anyone know any nice way for porting applications that use > fork.... into applications using taskSpawn or something familiar? > > --------------------------- Newsgroups: comp.os.vxworks Subject: MPC7450 support Date: Fri, 27 Jul 2001 00:46:07 +0300 From: Assaf Hoffman Organization: Galileo Technology Message-ID: <3B608F9F.C1537A37@galileo.co.il> Hi all !! Has anyone have any experience with the MPC7450 ?? - -- The only true freedom is freedom from your heart's desires.... ======================================================================== _/_/_/_/_/ _/_/_/_/_/_/_/ _/_/_/_/_/_/_/_/ _/_/_/_/ _/_/_/ _/_/_/ _/_/_/_/ _/_/_/ _/_/_/ _/_/_/ _/_/_/ Assaf Hoffman _/_/_/_/_/_/_/_/ Galileo Technology Ltd. _/_/_/_/_/_/ E-mail - hoffman@galileo.co.il _/_/_/_/ Tel - 972-4-9999555 ext. 1203 Check our Web site: http://www.galileoT.com ========================================================================= --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jul 26 05:50:44 2001 From: "Burkin, Valery" Date: Thu Jul 26 05:50:46 PDT 2001 Subject: RE: 3C90x PCI on vxWorks I've managed to setup ISA 3c509 Ethernet card, with el3c509End driver... #define IO_ADRS_ELT 0x300 #define INT_LVL_ELT 0x07 #define INT_VEC_ELT (INT_VEC_GET (INT_LVL_ELT)) #define ELT_ATTACHMENT_TYPE 0x1 #define ELT_RX_FRAMES 0x0 "Attached TCP/IP interface to elt0 unit0" "Attached interface lo0...done" Everything have been hopefully Attached ;))) ifShow: elt (unit number 0): Flags: (0x8063) UP BROADCAST RUNNING ARP MULTICAST Type: ETHERNET_CSMACD Internet address: 172.16.98.26 Broadcast address: 172.16.98.255 Netmask 0xffff0000 Subnetmask 0xffffff00 Ethernet address is 00:3c:19:00:00:20 Metric is 0 Maximum Transfer Unit size is 1500 42 packets received; 1 packets sent 42 multicast packets received 0 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped lo (unit number 0): Flags: (0x8069) UP LOOPBACK RUNNING ARP MULTICAST Type: SOFTWARE_LOOPBACK Internet address: 127.0.0.1 Netmask 0xff000000 Subnetmask 0xff000000 Metric is 0 Maximum Transfer Unit size is 32768 0 packets received; 0 packets sent 0 multicast packets received 0 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped BUT the situation is exactly like with PCI 3c905c-TX, INCORRECT Ethernet address(!!!!!!!!!) the correct MAC address of this ISA board is: 00:20:AF:2F:50:63 verified with NT and Linux(!!!!!!!!!!!!!) It's getting fascinating, you know... ;), I've been noticed smiling last 3 hours ;) Anybody got stuck in anything similar? Or is that just my unique PC playing mad games.....? valery., From vxwexplo-errs@csg.lbl.gov Thu Jul 26 07:04:34 2001 From: "Burkin, Valery" Date: Thu Jul 26 07:04:36 PDT 2001 Subject: RE: 3C90x PCI on vxWorks I DID IT!!!!!!!!!!!!!!!!!!!!!!!!! I've got network under vxworks, just unbelievable........ I can load vxworks kernel image over FTP, ping, NFS and whatever, But can anybody explain me that: ifShow: elt (unit number 0): Flags: (0x8063) UP BROADCAST RUNNING ARP MULTICAST Type: ETHERNET_CSMACD Internet address: 172.16.98.26 Broadcast address: 172.16.98.255 Netmask 0xffff0000 Subnetmask 0xffffff00 Ethernet address is 70:1d:19:00:00:20 Metric is 0 Maximum Transfer Unit size is 1500 531 packets received; 112 packets sent 429 multicast packets received 2 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped Ethernet addr: 70:1d:19:00:00:20 (INCORRECT!!!!!!!) PING tmman (172.16.98.26): 56 data bytes 64 bytes from tmman (172.16.98.26): icmp_seq=0. time=0. ms 64 bytes from tmman (172.16.98.26): icmp_seq=1. time=0. ms 64 bytes from tmman (172.16.98.26): icmp_seq=2. time=0. ms ----tmman PING Statistics---- 3 packets transmitted, 3 packets received, 0% packet loss round-trip (ms) min/avg/max = 0/0/0 value = 0 = 0x0 -> arpShow LINK LEVEL ARP TABLE destination gateway flags Refcnt Use Interface -------------------------------------------------------------------------- 172.16.98.25 00:20:af:ed:9e:bc 405 0 9 elt0 172.16.98.26 00:20:af:2f:50:63 405 0 6 lo0 172.16.98.109 00:b0:d0:5c:29:c2 405 1 92 elt0 -------------------------------------------------------------------------- Ethernet addr: 00:20:af:2f:50:63 (CORRECT!!!!!!!!) Why does ifShow displays incorrect Ethernet address(?????????????????) I killed a week of time, because of what I saw in ifShow..... valery. From vxwexplo-errs@csg.lbl.gov Fri Jul 27 04:03:51 2001 From: Vxworks Exploder Date: Fri Jul 27 04:03:54 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 27 04:03:14 PDT 2001 Subject: char and INT8 are unsigned types ? Subject: Not able to ping to the board Subject: Making bootrom with Kernel to load in ROM Subject: Re: Not able to ping to the board Subject: Re: Making bootrom with Kernel to load in ROM Subject: CGI Subject: Re: How to measure CPU utilization of a task at run time Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Subject: Re: char and INT8 are unsigned types ? Subject: Re: char and INT8 are unsigned types ? Subject: QMC problems Subject: Re: Not able to ping to the board Subject: Re: CGI Subject: avlLib Missing Subject: "lkup" functionality through C API Subject: Re: "lkup" functionality through C API Subject: Man pages for Tornado Release 2.0v4 Subject: RealTek 8139A Driver Subject: VxWorks and IGMP messages Subject: Re: avlLib Missing Subject: question about IP over ATM Subject: question about IP over ATM Subject: Re: Serial Port Interface for Unix (Solaris) Subject: VME access with MVME230x ? Subject: Question about array definition Subject: Re: Question about array definition Subject: STLport and VxWorks Subject: Re: question about IP over ATM Subject: Re: Question about array definition Subject: MVME230x: How to setup mailbox in A24:D16 slave adress window Subject: Re: Cache problems Subject: option to zinc Subject: IRC Subject: tcl/tk ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: char and INT8 are unsigned types ? Date: Thu, 26 Jul 2001 13:55:33 +0200 From: Rainer Stelzer Organization: Stoye GmbH, =?iso-8859-1?Q?K=F6ln?= Message-ID: <3B600535.158F6609@stoye.de> - --------------F19607206CB2178C15825852 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi, we have some funny behaivour casting char to short. testit() { char test1; short test2; test1 = -1; test2 = test1; } results in test2 to be 255 instead of -1, as expected. and even the hardware independant INT8 type is unsigned ?? cause testit() { INT8 test1; short test2; test1 = -1; test2 = test1; } results in test2 to be 255 instead of -1, too. We can see this on the the powerPC (MPC6xx and MPC8xx) architecture as well as in the simulator. Is this a known feature of the compiler or a bug ? ( If it's a feature I can't understand why INT8 is defined as char ) Any hint is welcome. Have a nice day ! R.Stelzer ===================================================== Gebrüder Stoye GmbH department: traffic control systems name: Rainer Stelzer adr: Longericher Strasse 177 city: Cologne, Germany zip: 50739 phone: ++49/221/1707019 fax: ++49/221/1707010 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== vxworks, tornado - --------------F19607206CB2178C15825852 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi,

we have some funny behaivour casting char to short.

testit()
{
    char   test1;
    short  test2;

    test1 = -1;
    test2 = test1;
}

results in test2 to be 255 instead of -1, as expected.

and even the hardware independant INT8 type is unsigned ??
cause
testit()
{
    INT8  test1;
    short  test2;

    test1 = -1;
    test2 = test1;
}

results in test2 to be 255 instead of -1, too.

We can see this on the the powerPC  (MPC6xx and MPC8xx) architecture as
well as in the simulator.

Is this a known feature of the compiler or a bug ?
( If it's a feature I can't understand why INT8 is defined as char )

Any hint is welcome.

Have a nice day !

R.Stelzer

=====================================================
            Gebrüder Stoye GmbH
department: traffic control systems
      name: Rainer Stelzer
       adr: Longericher Strasse 177
      city: Cologne, Germany
       zip: 50739
     phone: ++49/221/1707019
       fax: ++49/221/1707010
    mailto: r.stelzer@stoye.de
       www: http://www.stoye.de
=====================================================
vxworks, tornado - --------------F19607206CB2178C15825852-- --------------------------- Newsgroups: comp.os.vxworks Subject: Not able to ping to the board Date: 26 Jul 2001 05:02:54 -0700 From: nav@sasken.com (naveen) Organization: http://groups.google.com/ Message-ID: <95ccca9f.0107260402.66a211c5@posting.google.com> Hi, I have created a standalone vxworks image. This image is burnt into the flash. A valid ipaddress has been provided to the board. The problem is a)i am not able to ping to this board from a pc in the same network. b) loopback with 127.0.0.1 or with the specified ipaddress of the board happen fine. I am using tornado II. Is there any settings problem. Any solutions to the problem. Thanks B.Naveen --------------------------- Newsgroups: comp.os.vxworks Subject: Making bootrom with Kernel to load in ROM Date: Thu, 26 Jul 2001 15:56:16 +0200 From: "Jonathan ILIAS" Organization: University of Angers, France. Message-ID: <9jp7fm$3e2$1@news.univ-angers.fr> Hi all, I want to build a bootrom that is able to load VxWorks system from ROM. I read the documentation about it and I didn't found anything about it, all solutions needs to download the VxWorks image file from the host system (from network or serial line). Is it possible to do this without writing my own bootrom ? Thanks, Jonathan ILIAS --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Not able to ping to the board Date: Thu, 26 Jul 2001 13:54:25 GMT From: "Ken Kline" Organization: Time Warner Road Runner - Binghamton NY Message-ID: References: <95ccca9f.0107260402.66a211c5@posting.google.com> Have you checked your routes with the "routeShow" command. May need to establish a default route using "routeAdd". Also, verify your PC has a route to the target with a "route print" command. Network and target both have the same netmask settings? Ken "naveen" wrote in message news:95ccca9f.0107260402.66a211c5@posting.google.com... > Hi, > > I have created a standalone vxworks image. This image is > burnt into the flash. A valid ipaddress has been provided > to the board. The problem is > > a)i am not able to ping to this board from a pc in the > same network. > b) loopback with 127.0.0.1 or with the specified ipaddress of > the board happen fine. > > > I am using tornado II. Is there any settings problem. > > Any solutions to the problem. > > Thanks > > B.Naveen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Making bootrom with Kernel to load in ROM Date: Thu, 26 Jul 2001 16:17:18 +0200 From: "Werner Schiendl" Message-ID: <3b602711$0$22316$6e365a64@newsreader02.highway.telekom.at> References: <9jp7fm$3e2$1@news.univ-angers.fr> Hi, You can use a RAM disk and map a RAW filesystem on it. With this combination you get a file descriptor that can be feed to the usual load process. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: CGI Date: Thu, 26 Jul 2001 17:37:10 +0300 From: Nick Tripanis Organization: National Technical University of Athens, Greece Message-ID: <3B602B15.11882A1B@bu.edu> Does anyone have an idea of how to redirect the standard input from the web browser instead of the keyboard? i want to build a simple web server using cgi. Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to measure CPU utilization of a task at run time Date: Thu, 26 Jul 2001 14:49:28 GMT From: "Ken Kline" Organization: Time Warner Road Runner - Binghamton NY Message-ID: References: <3B5F22AE.70C159C1@ntl.inf.com> Sunil - Done this kind of thing in the past by using "taskSwitchHookAdd". By storing the system time and task name when a task switch occurs, you can keep a cumulative count of the time used by a particular task. In our case, we did this for every task to get an overall CPU util% by task. Of course, taskSwitchHookAdd takes extra clock cycles on EVERY task switch. This worked for us, since we only enabled this when we wanted to take a snapshot of system performance. The time taken by the OS was still a relatively small amount of total CPU - maybe around 2%, even with monitoring turned on. This was on a PPC603 200MHz. (OS time = 100% - sum(all tasks time)). Don't forget, "all tasks" includes the 'idle' task, so we considered OS time to be anything like task switch time, interrupt time, etc. There might be an easier way to do this, but I'm not aware of one. Ken "Sunil Kulkarni" wrote in message news:3B5F22AE.70C159C1@ntl.inf.com... > How to measure the CPU utilization of a task in runtime > > so that some action can be taken based on the % CPU > > utilization of the task. > > Thanks, > > Sunil > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: hi, I need the SA1100 's instructions set ,anyone know where i can get it? thanks Date: 26 Jul 2001 08:23:13 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107260723.a0bc546@posting.google.com> References: <6603907b.0107250026.718425d@posting.google.com> <488e459a.0107250844.50863a56@posting.google.com> <6603907b.0107252230.72ebe770@posting.google.com> Hello, fjpan@cmmail.com (Baumann) wrote in message news:<6603907b.0107252230.72ebe770@posting.google.com>... > thanks ya,i will buy the Manual . i have the GNUPRO TOOLKiT , > strongarm-elf-gdb seems only supports the serial debug, anyone knows > how to connect the target via the TCP ? and how can i open a PORT for > debugging? i have scanned the target when it launched , but i can not > find it opens any port, how the TORNADO connect the target server? For new versions of VxWorks/Tornado, a proprietary protocol and a target server. Do you have Tornado installed? If so, which version. If not, which version of VxWorks are you using and how are you developing for it? Regards, John... > > john_94501@yahoo.com (John) wrote in message news:<488e459a.0107250844.50863a56@posting.google.com>... > > > > As for debugging over a network, the Tornado debugger (Crosswind/gdb) > > will do this. Read the documentation that came with your Tornado > > package for information on setting up the target, a target server and > > starting the debugger. > > > > HTH, > > > > John... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: char and INT8 are unsigned types ? Date: 26 Jul 2001 11:31:24 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <3B600535.158F6609@stoye.de> Sender: bpringlemeir@DeadDuck >>>>> "Rainer" == Rainer Stelzer writes: Rainer> Hi, we have some funny behaivour casting char to short. Rainer> testit() { char test1; short test2; test1 = -1; test2 = test1; } Rainer> results in test2 to be 255 instead of -1, as expected. Rainer> and even the hardware independant INT8 type is unsigned ?? Ansi `C' does not define the sign of unadorned `char'. This allows the implementation to pick the most efficient form of char (signed or unsigned). This code should work as expected, testit() { signed char test1; short test2; test1 = -1; test2 = test1; } On a x86 architecture, the char value is usually signed. On the ARM (and PowerPC) it seems to be unsigned. Check out this URL, "http://www.arm.linux.org.uk/docs/faqs/signedchar.shtml" hth, Bill Pringlemeir. - -- How can you say that trees are unpleasant? So, the Martians are coming? Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: char and INT8 are unsigned types ? Date: Thu, 26 Jul 2001 09:24:33 -0600 From: Don Small Organization: Sandia National Laboratories Message-ID: <3B603631.42D9A699@sandia.gov> References: <3B600535.158F6609@stoye.de> See the GCC manual for "Options Controling C Dialect" except from manual "-funsigned-char Let the type char be unsigned, like unsigned char. Each kind of machine has a default for what char should be. It is either like unsigned char by default or like signed char by default. Ideally, a portable program should always use signed char or unsigned char when it depends on the signedness of an object. But many programs have been written to use plain char and expect it to be signed, or expect it to be unsigned, depending on the machines they were written for. This option, and its inverse, let you make such a program work with the opposite default. The type char is always a distinct type from each of singed char or unsigned char, even though its behaviour is always just like one of those two." I think that this describes your situation. If your writing new code I would definitely use the above advice on always defining as unsigned char or signed char. If this is old code then -funsigned-char or its opposite -fsigned-char should do the trick. Don Small Rainer Stelzer wrote: > > Hi, > > we have some funny behaivour casting char to short. > > testit() > { > char test1; > short test2; > > test1 = -1; > test2 = test1; > } > > results in test2 to be 255 instead of -1, as expected. > > and even the hardware independant INT8 type is unsigned ?? > cause > testit() > { > INT8 test1; > short test2; > > test1 = -1; > test2 = test1; > } > > results in test2 to be 255 instead of -1, too. > < snip > > R.Stelzer > --------------------------- Newsgroups: comp.os.vxworks Subject: QMC problems Date: 26 Jul 2001 08:46:42 -0700 From: huy_vu@mitel.com (Huy Vu) Organization: http://groups.google.com/ Message-ID: Hello, This posting is not directly related to vxworks but I couldn't find a newsgroup dedicated to HDLC/QMC on the 68360. So, here goes: I found that if an error bit (such as NO, LG or CRC) in the rx status register of a buffer descriptor (BD) is not cleared, the next time that BD is used, the data in the buffer is not correct (corrupted maybe?). Now, I thought that the CP only sets these bits and not looks at them when receiving data. Can someone out there corroborate this finding? Thanks a bunch. BTW, if I clear the error bit after reading off the data in the buffer then everything is fine. I just want to know if there is a bug in the microcode of the CP. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Not able to ping to the board Date: Thu, 26 Jul 2001 16:57:26 +0100 From: "timallen" Organization: ntlworld News Service Message-ID: <2aX77.27079$SK6.3178874@news6-win.server.ntlworld.com> References: <95ccca9f.0107260402.66a211c5@posting.google.com> Reply-To: "timallen" > I have created a standalone vxworks image. This image is > burnt into the flash. A valid ipaddress has been provided > to the board. The problem is > > a)i am not able to ping to this board from a pc in the > same network. Standalone vxWorks (vxWorks.st/vxWorks.st_rom) doesn't initialise the network driver by default. Are you initialising? If you have network show routines included, you can type ifShow at the prompt and it should show the status of the network drivers. Regards, Tim. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CGI Date: Thu, 26 Jul 2001 18:11:34 +0200 From: "Werner Schiendl" Message-ID: <3b6041d9$0$4100$6e365a64@newsreader02.highway.telekom.at> References: <3B602B15.11882A1B@bu.edu> Hi, you just need to accept incomming TCP connections to port 80 (the default HTTP port). The file descriptor resulting can be used to feed your CGI. If you need more web server features you can consider getting a 'real' web server. Wind River has its own Wind Web Server. You can also use GoAhead Web Server (see http://www.goahead.com/) there may be other Web Servers for VxWorks out there. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: avlLib Missing Date: Thu, 26 Jul 2001 10:11:54 -0700 From: "Davis Ford" Organization: Michigan State University Message-ID: <9jpilb$2hpd$1@msunews.cl.msu.edu> Hi, trying to build a BSP for PPC603, I get the following undefined references. Seems avlLib is not to be found. Any idea where avlLib is so that I can load it properly? Thanks in advance. dford /home/dford/Tornado2//host/sun4-solaris2/bin/ldppc -X -N -e _sysInit -Ttext 00010000 \ -o vxWorks.st tmp.2 symTbl.o ctdt.o tmp.2: In function `m2IfTableUpdate': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1063: undefined reference to `avlInsert' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1086: undefined reference to `avlDelete' tmp.2: In function `m2IfTblEntryGet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1327: undefined reference to `avlSearch' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1332: undefined reference to `avlSuccessorGet' tmp.2: In function `m2IfTblEntrySet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1708: undefined reference to `avlSearch' tmp.2: In function `m2IfStackTblUpdate': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1905: undefined reference to `avlSearch' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1907: undefined reference to `avlSearch' tmp.2: In function `m2IfStackEntryGet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1986: undefined reference to `avlSuccessorGet' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2003: undefined reference to `avlSearch' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2036: undefined reference to `avlSuccessorGet' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2050: undefined reference to `avlSuccessorGet' tmp.2: In function `m2IfStackEntrySet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2144: undefined reference to `avlSearch' tmp.2: In function `m2IfRcvAddrEntryGet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2242: undefined reference to `avlSuccessorGet' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2254: undefined reference to `avlSearch' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2299: undefined reference to `avlSuccessorGet' tmp.2: In function `m2IfRcvAddrEntrySet': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2387: undefined reference to `avlSearch' tmp.2: In function `m2IfDelete': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2499: undefined reference to `avlTreeErase' tmp.2: In function `ifWalk': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2591: undefined reference to `avlSuccessorGet' tmp.2: In function `m2IfShow': /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2618: undefined reference to `avlMinimumGet' /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2623: undefined reference to `avlSuccessorGet' make[1]: *** [vxWorks.st] Error 1 make[1]: Leaving directory `/home/dford/SDK-3.0.0/firmware/systems/mousse/vxworks/target/config/Kahlua' make: *** [bcm.mousse] Error 2 --------------------------- Newsgroups: comp.os.vxworks Subject: "lkup" functionality through C API Date: Thu, 26 Jul 2001 13:58:39 -0400 From: "Fangzhe Chang" Organization: Bell Labs Message-ID: <9jpl97$1671@news.research.bell-labs.com> Reply-To: "Fangzhe Chang" Hi, I would like to have function address lookup using C API, just like "lkup" from the VxWorks shell. I tried symFindByName; however, it did not seem to work with downloaded modules. Does anyone happen to know whether VxWorks exports such a functionality or not? Any alternate way for achieving this is also greatly appreciated. Thanks a lot. Fangzhe --------------------------- Newsgroups: comp.os.vxworks Subject: Re: "lkup" functionality through C API Date: 26 Jul 2001 14:19:59 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <9jpl97$1671@news.research.bell-labs.com> Sender: bpringlemeir@DeadDuck >>>>> "Fangzhe" == Fangzhe Chang writes: Fangzhe> Hi, I would like to have function address lookup using C Fangzhe> API, just like "lkup" from the VxWorks shell. I tried Fangzhe> symFindByName; however, it did not seem to work with Fangzhe> downloaded modules. Does anyone happen to know whether Fangzhe> VxWorks exports such a functionality or not? Any alternate Fangzhe> way for achieving this is also greatly appreciated. Thanks a Fangzhe> lot. You are experiencing something that many people have asked about in this newsgroup before. The Shell has some function on the target and some on the host. If you preceed your shell command with an `at sign' like this, - -> @lkup "Show" You might not get any response. You must build a standalone image in order to link in the symbol table. If the symbol table has not been linked with the image and loaded on the target, then the function `symFindByName()' will not work. hth, Bill Pringlemeir - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Man pages for Tornado Release 2.0v4 Date: Thu, 26 Jul 2001 19:19:51 GMT From: "William H. McCloskey" Organization: The Boeing Company Message-ID: <3B606D56.185DD954@roses.bna.boeing.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) The following man locations no longer exist in Rel 2.0v4, alias tman 'man -M \ $WIND_BASE/host/man \ $WIND_BASE/host/sun4-solaris2/man \ $WIND_BASE/target/man ` I'm moving on from 5.3.1, which has a more commmand line "friendly" look and feel. Please advise as to where I can pick up the Rel 2.0v4 man pages. Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: RealTek 8139A Driver Date: Thu, 26 Jul 2001 14:46:22 -0500 From: "Joel Hannah" Organization: WebUseNet Corp. - "ReInventing The UseNet" Message-ID: I am using a single board computer to host my VxWorks application on. I have been having TCP/IP problems and I have narrowed the problem down to ReakTek 8139A ethernet controler driver. I have been using a driver that a Windriver FAE supplied to me with the stipulation that Windriver does not support this driver. He promised me that Windriver was going to support the 8139A but I haven't seen it yet. I was wondering if someone had a later version of this driver (I am running driver version: Release 1.0/EA3 I would really appreciate any help anyone could give me.... Joel - -- S. Joel Hannah Senior Engineer Alpha Beta Technologies, Inc 3112 12th Ave SW Huntsville, AL 35805 Phone: (256) 534-9067 Fax: (256) 534-9069 Pager: (256) 580-8947 --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks and IGMP messages Date: 26 Jul 2001 13:04:43 -0700 From: sachin@blrmail.com (Sachin) Organization: http://groups.google.com/ Message-ID: Hello VxWorkers, I need some assistance here. I wanted to know if the IP stack of VxWorks (Tornado 2.0) handles IGMP Join, Query, and Report messages. If so, how does it handle them?? Does it: Configure its ethernet interface to start listening for the Group Address mentioned in the IGMP Join Message?? Configure any other network interfaces to forward multicast packets destined for the Group Address?? Is there a way for an application to be notified when an IGMP message is received and check its contents?? I would really appreciate any help from you knowledgeable folks out there. Thanks, Sachin. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: avlLib Missing Date: 26 Jul 2001 14:27:11 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107261327.7f0304c9@posting.google.com> References: <9jpilb$2hpd$1@msunews.cl.msu.edu> Hello, It looks as though you have a mismatch between the version of VxWorks you have installed and whatever networking product it was that you installed over it to get m2IfLib.c. The avlLib stuff was added in a recent version of the SNMP code, and only works with the latest versions of the basic VxWorks product (since they are the ones that include the avlLib code). Check with your local FAE and/or sales guy to see that you have the correct versions of all the products installed. HTH, John... "Davis Ford" wrote in message news:<9jpilb$2hpd$1@msunews.cl.msu.edu>... > Hi, trying to build a BSP for PPC603, I get the following undefined > references. Seems avlLib is not to be found. Any idea where avlLib is so > that I can load it properly? Thanks in advance. > > dford > > > /home/dford/Tornado2//host/sun4-solaris2/bin/ldppc -X -N -e _sysInit -Ttext > 00010000 \ > -o vxWorks.st tmp.2 symTbl.o ctdt.o > tmp.2: In function `m2IfTableUpdate': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1063: undefined reference > to `avlInsert' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1086: undefined reference > to `avlDelete' > tmp.2: In function `m2IfTblEntryGet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1327: undefined reference > to `avlSearch' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1332: undefined reference > to `avlSuccessorGet' > tmp.2: In function `m2IfTblEntrySet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1708: undefined reference > to `avlSearch' > tmp.2: In function `m2IfStackTblUpdate': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1905: undefined reference > to `avlSearch' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1907: undefined reference > to `avlSearch' > tmp.2: In function `m2IfStackEntryGet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:1986: undefined reference > to `avlSuccessorGet' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2003: undefined reference > to `avlSearch' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2036: undefined reference > to `avlSuccessorGet' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2050: undefined reference > to `avlSuccessorGet' > tmp.2: In function `m2IfStackEntrySet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2144: undefined reference > to `avlSearch' > tmp.2: In function `m2IfRcvAddrEntryGet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2242: undefined reference > to `avlSuccessorGet' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2254: undefined reference > to `avlSearch' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2299: undefined reference > to `avlSuccessorGet' > tmp.2: In function `m2IfRcvAddrEntrySet': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2387: undefined reference > to `avlSearch' > tmp.2: In function `m2IfDelete': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2499: undefined reference > to `avlTreeErase' > tmp.2: In function `ifWalk': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2591: undefined reference > to `avlSuccessorGet' > tmp.2: In function `m2IfShow': > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2618: undefined reference > to `avlMinimumGet' > /home/dford/Tornado2/target/src/netwrs/m2IfLib.c:2623: undefined reference > to `avlSuccessorGet' > make[1]: *** [vxWorks.st] Error 1 > make[1]: Leaving directory > `/home/dford/SDK-3.0.0/firmware/systems/mousse/vxworks/target/config/Kahlua' > make: *** [bcm.mousse] Error 2 --------------------------- Newsgroups: comp.os.vxworks Subject: question about IP over ATM Date: Thu, 26 Jul 2001 11:02:16 +0800 From: "Bruce Shi" Organization: The news server of Changzhou TeleCom Message-ID: <9jo1jk$e0k$1@news.cz.js.cn> Hi,all: I am doing a END driver to implement IP over ATM in VxWorks 5.4. I will use PVC(permanent virtual circuit). According to RFC1577, PVC(VPI,VCI) is correspondind to IP address, so if I have several PVC, we need IP address of the same number. but we only have an END driver, does END can be binded to several IP address? Thank you for any answer! best regards bruce --------------------------- Newsgroups: comp.os.vxworks Subject: question about IP over ATM Date: Thu, 26 Jul 2001 11:04:37 +0800 From: "Bruce Shi" Organization: The news server of Changzhou TeleCom Message-ID: <9jo1m3$eht$1@news.cz.js.cn> Hi,all: I am doing a END driver to implement IP over ATM in VxWorks 5.4. I will use PVC(permanent virtual circuit). According to RFC1577, PVC(VPI,VCI) is correspondind to IP address, so if I have several PVC, we need IP address of the same number. but we only have an END driver, does END can be binded to several IP address? Thank you for any answer! best regards bruce --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Serial Port Interface for Unix (Solaris) Date: Thu, 26 Jul 2001 20:14:23 -0500 From: "Kevin Kulhanek" Organization: Lucent Technologies, Columbus, Ohio Message-ID: <9jqfio$50@nntpa.cb.lucent.com> References: <9jl1lf$bej@nntpb.cb.lucent.com> <3B5E34C2.523863A3@covad.net> <3B5EF0D6.2CA9C440@sandia.gov> Thanks guys, It worked as you both said. I really appreciate it. Kev --------------------------- Newsgroups: comp.os.vxworks Subject: VME access with MVME230x ? Date: 27 Jul 2001 01:34:47 GMT From: csmith@nikonrca.com Organization: Concentric Internet Services Message-ID: Hello, I am having a problem accessing VME addresses using TornadoII and VxWorks5.4. I use sysBusToLocalAdrs to obtain the local address and the function returns true. when I try to reference the obtained address, I get an AGENT_MEM_ACCESS_ERROR. I've looked through the BSP master window settings and everything seems to be in order, so I have no idea why I'm getting this error. Why would sysBusToLocalAdrs return true if the address can't be read? I was able to access a board in the chassis using the A32 window, the above worked fine, but I'm trying to access a different board within the A16 window with no luck. Any ideas? --------------------------- Newsgroups: comp.os.vxworks Subject: Question about array definition Date: 26 Jul 2001 19:19:12 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107261819.39d95ba1@posting.google.com> How can I define an array which starts at a fixed address in vxWorks? As below: UCHAR a[1000000]; And the address of a is always 0Xaabbccdd. ellin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Question about array definition Date: Fri, 27 Jul 2001 02:52:42 GMT From: pkockritz@home.com (Pete Kockritz) Organization: Excite@Home - The Leader in Broadband http://home.com/faster Message-ID: References: <44a01eba.0107261819.39d95ba1@posting.google.com> In article <44a01eba.0107261819.39d95ba1@posting.google.com>, ellin_lin@263.net (ellin) wrote: > How can I define an array which starts at a fixed address in vxWorks? > As below: > UCHAR a[1000000]; > And the address of a is always 0Xaabbccdd. > > ellin You could use the linker to place the array at a specific address. But the simplest way it do just declare a pointer to the type and set the pointer to the address: #define FIXED_ADDRESS (0xaabbccdd) UCHAR *a = (UCHAR *)FIXED_ADDRESS; Now dereferencing the the pointer will either read or write to the address. If you are using this to access a physical piece of hardware, you should declare the ptr as volatile. You can use array indexing to access whatever is at that address: UCHAR x, y; x = *a; /* read address 0xaabbccdd */ x = a[0]; /* same as above */ y = a[100]; /* reads address 0xaabbccdd + 0x100 */ a[200] = x+1; /* writes to address 0xaabbccdd + 0x200 */ If you are using this to control a hardware device, often a struct will be defined that has fields corresponding to the registers of the device. typedef struct { UCHAR regA; UCHAR regB; UCHAR regC; } myHwDevice_t; volatile myHwDevice_t *hw = (myHwDevice_t *)FIXED_ADDRESS; x = hw->regA; /* read reg A */ hw->regB = 0x12 /* write reg B */ Pete - -- Pete Kockritz --------------------------- Newsgroups: comp.os.vxworks Subject: STLport and VxWorks Date: 26 Jul 2001 23:26:57 -0700 From: maximb@shiron.com (Maxim) Organization: http://groups.google.com/ Message-ID: <329d00d6.0107262226.e42d895@posting.google.com> Hello, programmers. Im trying to evaluate the STLport 4.0 suite for our development needs. After i'd placed the unpacked STLport-4.0 to my /Tornado/target directory, all the test programs continue to use the Tornado's native std. If anybody has an expirience in adaptation of STLport, pls, help me. Development environment: Platform: VxWorks 5.4 Compiler: GCC 2.7.2 Dev. tool: Tornado 2.0 10X. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: question about IP over ATM Date: 27 Jul 2001 06:48:18 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <9jo1jk$e0k$1@news.cz.js.cn> "Bruce Shi" wrote: > Hi,all: > I am doing a END driver to implement IP over ATM in VxWorks 5.4. I will > use PVC(permanent virtual circuit). According to RFC1577, PVC(VPI,VCI) > is correspondind to IP address, so if I have several PVC, we need IP > address of the same number. but we only have an END driver, does END > can be binded to several IP address? Thank you for any answer! This should not be a real problem. You only have to make sure that your END driver is reentrant, and that you use a seperate data structure for each of your PVC's. This mechanism is available in several other END drivers, as you need the same mechanism is you attach 2 identical ethernet controllers to your processor. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Question about array definition Date: Fri, 27 Jul 2001 10:13:15 +0200 From: "Werner Schiendl" Message-ID: <3b61233f$0$4100$6e365a64@newsreader02.highway.telekom.at> References: <44a01eba.0107261819.39d95ba1@posting.google.com> Hi, just a note. If you use this mechanism to access a device you must consider caching issues. Lookup cacheLib in the manuals for details. regards Werner --------------------------- Newsgroups: comp.os.vxworks Subject: MVME230x: How to setup mailbox in A24:D16 slave adress window Date: Fri, 27 Jul 2001 10:23:52 +0200 From: Rainer Stelzer Organization: Stoye GmbH, =?iso-8859-1?Q?K=F6ln?= Message-ID: <3B612518.C31A7EC2@stoye.de> Hi, I configured a VME shared memory network with sixteen A24:D16 CPU boards and one MVME230x acting as master. In general it works, but the MVME runs in poll mode instead of the prefered interrupt mode. Until now I wasn't able to move the mailbox area into the A24 slave area. Any hint how to do that ? thanx in advance regards Rainer ===================================================== Gebrüder Stoye GmbH department: traffic control systems name: Rainer Stelzer adr: Longericher Strasse 177 city: Cologne, Germany zip: 50739 phone: ++49/221/1707019 fax: ++49/221/1707010 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cache problems Date: Thu, 26 Jul 2001 14:27:46 +0100 From: "Dave Korn" Organization: Lumber Cartel (tinlc) Members #2234-2237 (owing to browser refresh) Message-ID: References: <3b5da4f4.956879249@200.123.101.100> <3b5f02f7.1046477755@200.123.101.100> cgan76 wrote in message <3b5f02f7.1046477755@200.123.101.100>... > >When I was doing my tests, to simplify things I had all portions of my >memory was set to VMSTATE_CACHEABLE_NOT. And typically (from the >template) the first portion of RAM from 0x0000_0000 to RAM_LOW_ADRS is >always set to non-cacheable. > >But when I was debugging with an ICE, I noticed that after a crash >there were valid pages in the data cache even in the portions of >memory that I had marked non-cacheable (from 0x0000_0000 to >LOCAL_MEM_SIZE. So just to test things I set all of this memory to >cacheable and lo and behold everything works. > >Why VxWorks was trying to do cache processes on memory space I had >marked non-cacheable I don't know yet and I will have to investigate >later. This suggests that your first setup had the same physical memory mapped twice, once cacheable and once non-cacheable: as you can imagine, they won't play nicely together with regard to coherency. Was there a BAT entry overlapping the range you mapped with PTEs, perhaps? DaveK - -- Last time I checked, the UN Convention of Human Rights didn't say anything about the right to run a bot on Usenet. --------------------------- Newsgroups: comp.os.vxworks Subject: option to zinc Date: Fri, 27 Jul 2001 11:22:19 +0200 From: "Mehmet Demirel" Organization: Swisscom IP+ (post doesn't reflect views of Swisscom) Message-ID: <9jrbs8$f55$1@rex.ip-plus.net> Reply-To: "Mehmet Demirel" Hi, I am looking for other user interfaces than Zinc from WindRiver. I would need the source code for user interfaces regardless free or commercial. The main requirement is, that it is tiny and and easy scalable. I am also happy, when someone could tell me where I can find more information or link list about this subject. Thanxs Mehmet mehmet.demirel@aon.at --------------------------- Newsgroups: comp.os.vxworks Subject: IRC Date: Fri, 27 Jul 2001 12:25:28 +0300 From: Nick Tripanis Organization: National Technical University of Athens, Greece Message-ID: <3B613387.785FD65B@bu.edu> Is there any channel in IRC for VxWorks, and if yes, under which server? --------------------------- Newsgroups: comp.os.vxworks Subject: tcl/tk Date: Fri, 27 Jul 2001 12:20:53 +0200 From: "Mehmet Demirel" Organization: Swisscom IP+ (post doesn't reflect views of Swisscom) Message-ID: <9jrfa6$fkb$1@rex.ip-plus.net> Reply-To: "Mehmet Demirel" Hi, is there a version of tcl/tk for vxworks ? Where can I download it and has someone experiences in it ? Thanx Mehmet --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jul 28 04:03:38 2001 From: Vxworks Exploder Date: Sat Jul 28 04:03:41 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 28 04:03:11 PDT 2001 Subject: Re: option to zinc Subject: Re: question about IP over ATM Subject: Ocelot cPCI MIPS board Subject: Re: option to zinc Subject: Re: RealTek 8139A Driver Subject: .o file format? Subject: Re: Cache problems Subject: Re: RealTek 8139A Driver Subject: Re: VxWorks Developpment on Linux Subject: END Driver Woes (LONG) Subject: Re: .o file format? Subject: Re: Not able to ping to the board Subject: Re: How to measure CPU utilization of a task at run time Subject: Re: How to communicate using Serial Port in Vxworks Subject: Re: Making bootrom with Kernel to load in ROM Subject: vxWorks boot image Subject: Re: vxWorks boot image Subject: nfsMount error resolution needed Subject: web server instead of windweb? Subject: Re: Question about array definition Subject: Re: Man pages for Tornado Release 2.0v4 Subject: Re: question about IP over ATM Subject: Re: END Driver Woes (LONG) Subject: error of the task Subject: download image to vxsim ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: option to zinc Date: Fri, 27 Jul 2001 13:39:25 +0100 From: "John Beale" Organization: MINX Message-ID: <9jrnla$d4t$1@localhost.localdomain> References: <9jrbs8$f55$1@rex.ip-plus.net> If you want to go Java then you can use AWT or a lightweight GUI toolkit such as Espresso from Espial. regards, John "Mehmet Demirel" wrote in message news:9jrbs8$f55$1@rex.ip-plus.net... > Hi, > > I am looking for other user interfaces than Zinc from WindRiver. > I would need the source code for user interfaces regardless free or > commercial. > > The main requirement is, that it is tiny and and easy scalable. > > I am also happy, when someone could tell me where I can find more > information > or link list about this subject. > > Thanxs > Mehmet > > mehmet.demirel@aon.at > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: question about IP over ATM Date: Fri, 27 Jul 2001 15:57:46 +0200 From: John Mackenzie Organization: Siemens AG Message-ID: <3B61735A.B9F8F873@icn.siemens.de> References: <9jo1jk$e0k$1@news.cz.js.cn> Reply-To: John.Mackenzie.extern@icn.siemens.de Hi, I am also working on the very same. As far as my understanding of Classical IP over ATM, you can have several PVCs terminating on the same IP address i.e. only one IP address for several PVCs . In this way you define a LIS (a logical IP subnetwork). You then need a static ATMARP table so that IP can lookup which PVC it needs to send data on for a given remote IP address. I have implemented this in my driver and provide a special address resolution function which I install on the MUX with muxAddrFuncAdd() - I am using the Network Protocol Toolkit - - but the driver is almost identical to an END driver. NPT makes it easier to do non.Ethernet drivers. IP then passes this proprietry hardware address (could be simply a VPI/VCI) to your driver and you can send the data on the right PVC. When you receive data you just pass it up IP as the source IP address is in the IP header. Of course all remote sides must use an IP address on the LIS and it could be you implement several driver instances (with different IP addresses) to support more than one LIS. When SVCs are uses and ATMARP server is needed on the LIS. But for PVCs you can just do it all statically. Regards, John Mackenzie Bruce Shi wrote: > Hi,all: > I am doing a END driver to implement IP over ATM in VxWorks 5.4. I will > use PVC(permanent virtual circuit). According to RFC1577, PVC(VPI,VCI) is > correspondind to IP address, so if I have several PVC, we need IP address of > the same number. but we only have an END driver, does END can be binded to > several IP address? Thank you for any answer! > > best regards > bruce --------------------------- Newsgroups: comp.os.vxworks Subject: Ocelot cPCI MIPS board Date: Fri, 27 Jul 2001 14:27:20 GMT From: Andray Kaganovsky Organization: @Home Message-ID: If anybody has been using BSP for Ocelot cPCI board MIPS board for MIPS software development, could you please let me know how stable it's BSP is. Does it work with T2 and T3, or with T2 only ? Does it install into Tornado tree and builds cleanly ? Thanks in advance, Andray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: option to zinc Date: 27 Jul 2001 10:49:44 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <9jrbs8$f55$1@rex.ip-plus.net> Sender: bpringlemeir@DeadDuck Mehmet> Hi, I am looking for other user interfaces than Zinc from Mehmet> WindRiver. I would need the source code for user interfaces Mehmet> regardless free or commercial. There were several discussion a few months back about this. You could try "groups.google.com" to look for them. Here are some links, "http://www.microwindows.org/" "http://www.panelsoft.com/uihwsw.htm" "http://www.swellsoftware.com/index.shtml" If you want a small gui that works out of the box with vxWorks, then you can code with WindML itself. This use to be called UGL. It is certainly capable of making a useful small screen GUI. Many features of Zinc really aren't needed for many products IMHO. Zinc is nice in allowing GUI development natively on a workstation or PC. WindML is your best choice if you don't wish to port things. Microwindows might be a good choice if you have the time to port things. hth, Bill Pringlemeir. - -- Keep things as simple as possible, but no simpler. - A. Einstein Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RealTek 8139A Driver Date: Thu, 26 Jul 2001 23:53:57 +0800 From: "susanwan" Organization: GCNet(Reach and Range Inc.) Message-ID: <9js2n3$35ss$1@member.gcn.net.tw> References: Dear Sir: I am victor a Realtek's system design engineer. Could you told me more detail about your problem. Althrough I am debuging some bug of Vxworks driver for 8139 series but I have too much experence in this Realtime of. Most of case I write DOS or Novell's server driver. Hope I could solve your problem. victor Joel Hannah wrote in message news:rs_77.3$Og1.2431@e420r-atl1.usenetserver.com... > I am using a single board computer to host my VxWorks application on. I have > been having TCP/IP problems and I have narrowed the problem down to ReakTek > 8139A ethernet controler driver. I have been using a driver that a Windriver > FAE supplied to me with the stipulation that Windriver does not support this > driver. He promised me that Windriver was going to support the 8139A but I > haven't seen it yet. > > I was wondering if someone had a later version of this driver (I am running > driver version: Release 1.0/EA3 > > I would really appreciate any help anyone could give me.... > > Joel > > -- > S. Joel Hannah > Senior Engineer > > Alpha Beta Technologies, Inc > 3112 12th Ave SW > Huntsville, AL 35805 > > Phone: (256) 534-9067 > Fax: (256) 534-9069 > Pager: (256) 580-8947 > > > --------------------------- Newsgroups: comp.os.vxworks Subject: .o file format? Date: Fri, 27 Jul 2001 22:20:57 +0400 From: "Dmitry" Organization: posted via PTT-Teleport ISP, AS6795 Message-ID: <3b61b222$1@news.ptt.ru> Where can I get .o object file format description of for Tornado? Thanks, Dmitry --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cache problems Date: Fri, 27 Jul 2001 18:33:25 GMT From: cgan76@nospam4me.com (cgan76) Organization: MTS Internet Message-ID: <3b61b239.1222407118@200.123.101.100> References: <3b5da4f4.956879249@200.123.101.100> <3b5f02f7.1046477755@200.123.101.100> Actually, I did suspect this, but it wasn't the case. The way I have things set up is that only my RAM is mapped using the sysmemdesc, the rest (IO, PCI, flash, etc) is using the BATs. When I was doing testing I did a number of combinations where, for example, everything was mapped to BATs, nothing was mapped to BATs, etc, yet it didn't seem to effect the end result. On Thu, 26 Jul 2001 14:27:46 +0100, "Dave Korn" wrote: >cgan76 wrote in message <3b5f02f7.1046477755@200.123.101.100>... >> >>When I was doing my tests, to simplify things I had all portions of my >>memory was set to VMSTATE_CACHEABLE_NOT. And typically (from the >>template) the first portion of RAM from 0x0000_0000 to RAM_LOW_ADRS is >>always set to non-cacheable. >> >>But when I was debugging with an ICE, I noticed that after a crash >>there were valid pages in the data cache even in the portions of >>memory that I had marked non-cacheable (from 0x0000_0000 to >>LOCAL_MEM_SIZE. So just to test things I set all of this memory to >>cacheable and lo and behold everything works. >> >>Why VxWorks was trying to do cache processes on memory space I had >>marked non-cacheable I don't know yet and I will have to investigate >>later. > > This suggests that your first setup had the same physical memory mapped >twice, once cacheable and once non-cacheable: as you can imagine, they won't >play nicely together with regard to coherency. Was there a BAT entry >overlapping the range you mapped with PTEs, perhaps? > > DaveK >-- > Last time I checked, the UN Convention of Human Rights didn't say anything >about the right to run a bot on Usenet. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RealTek 8139A Driver Date: Fri, 27 Jul 2001 14:16:55 -0500 From: "Joel Hannah" Organization: WebUseNet Corp. - "ReInventing The UseNet" Message-ID: References: <9js2n3$35ss$1@member.gcn.net.tw> Thanks Victor, Below is the problem as I described it to the tech support at Windriver: My name is Joel Hannah and I work for Alpha Beta Technolgies in Huntsville. We purchased Tornado 2.0/VxWorks 5.4 almost a year ago and started developing our real time application for video processing. Our VxWorks target is an Advantech 5820 which has a RealTek 8139A ethernet controller on board. When we purchased the Windriver tools, a local Windriver FAE installed the latest driver for the RealTek part and told us a fully support Windriver driver for the RealTek 8139A was imminent. Our application utilizes FTP for transfers between our VxWorks target and other PCs. Initially our files were relatively small (<10K) but that has since changed. In addition we have developed custom TCP/IP datalink protocols (both TCP and UDP). Well as this app has matured, we have had increasing problems with our ethernet transfers. We have timeouts, slow downs and downright lockups. We can't even transfer a file at all if it over 2MB in size. We have adjusted the TCP/IP buffer values but will no noticeable improvements. That has left us with thinking that the driver may have errors. I hope this describes the problems for you Victor.... Joel susanwan wrote in message news:9js2n3$35ss$1@member.gcn.net.tw... > Dear Sir: > I am victor a Realtek's system design engineer. Could you told me more > detail about your problem. Althrough I am debuging some bug of Vxworks > driver for 8139 series but I have too much experence in this Realtime of. > Most of case I write DOS or Novell's server driver. Hope I could solve your > problem. > > victor > > Joel Hannah wrote in message > news:rs_77.3$Og1.2431@e420r-atl1.usenetserver.com... > > I am using a single board computer to host my VxWorks application on. I > have > > been having TCP/IP problems and I have narrowed the problem down to > ReakTek > > 8139A ethernet controler driver. I have been using a driver that a > Windriver > > FAE supplied to me with the stipulation that Windriver does not support > this > > driver. He promised me that Windriver was going to support the 8139A but I > > haven't seen it yet. > > > > I was wondering if someone had a later version of this driver (I am > running > > driver version: Release 1.0/EA3 > > > > I would really appreciate any help anyone could give me.... > > > > Joel > > > > -- > > S. Joel Hannah > > Senior Engineer > > > > Alpha Beta Technologies, Inc > > 3112 12th Ave SW > > Huntsville, AL 35805 > > > > Phone: (256) 534-9067 > > Fax: (256) 534-9069 > > Pager: (256) 580-8947 > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Developpment on Linux Date: Fri, 27 Jul 2001 14:41:01 -0500 From: Luc Message-ID: <3B61C3CD.9C39450E@qwest.net> References: <3B31EB91.56B0BAF9@gaci.fr> Not from WindRiver but you can cook your own GNU tools. I suggest you use Linux instead off vxWorks you'll save a lot off time AND money! Christophe Rabourg wrote: > Hello, > > I am wondering whether such tools are available on Linux > (ie: Cross Compiler, remode debugger, remote shell ...) ? > > If so, where may I download these materials ? > Thanks. > > -- > Christophe Rabourg _/_/_/ _/_/_/ _/_/_/ _/ 6 Av. de Norvege BP46 > tel: +33 1 69071686 _/ _/ _/ _/ _/ 91942 Courtaboeuf > fax: +33 1 69071358 _/ _/ _/_/_/_/ _/ _/ France > crabourg@gaci.fr _/_/_/ _/ _/ _/_/_/ _/ http://www.gaci.fr --------------------------- Newsgroups: comp.os.vxworks Subject: END Driver Woes (LONG) Date: Fri, 27 Jul 2001 12:55:16 -0700 From: "Davis Ford" Organization: Michigan State University Message-ID: <9jsgjk$2qtm$1@msunews.cl.msu.edu> Hi, I have an SDK package from Broadcom for their BCM5680 Strataswitch. The SDK contains basically all the source code necessary for building the bootrom and VxWorks image, except it does grab some headers and libraries from a full-blown Tornado installation. WindRiver also supplies their own BSP for the BCM5680. The two companies thus have two separate source trees for the same thing. I am working on getting the Broadcom SDK to work. The problem I have is with the END driver for the Fast Ethernet management interface. The onboard Ethernet controller uses the DEC21143 END Driver (dec21x4xEnd.c). Broadcom only ships the object file for the driver...not the source. I have no idea why. In the WindRiver BSP for BCM5680, the source and header are included. However, there have been enough changes between the two BSPs (Broadcom Vs. WindRiver) that there are numerous incompatibilities between the two. You're stuck with trying to get one or the other to work. My problem is that when muxDevLoad is called, it calls the ec21x4xEndLoad( ) function and passes it an init string. From here, what happens is anyone's guess. Like I said, I only have the object file. The init string passed to the EndLoad function is as follows: /* * Default SENS device: DEC-21143-XX (onboard ethernet) * Int Str= unit:MBAR:MBAR:ivec:ilvl:mem:memsize:usrFlags:offset */ #ifdef INCLUDE_END_DEC_21X4X #define DEC_21X4X_FUNC dec21x4xEndLoad #define DEC_21X4X_STRING "0x80000000:0:1:1:-1:-1:0:0x40004000" mbar : mbar : ivec : ilvl : mem : memsz : flags : offset 0x8 1 1 -1 -1 0 0x4 IMPORT END_OBJ* DEC_21X4X_FUNC(char*, void*); #endif /* INCLUDE_END_DEC21X4X */ I have checked the parameters for the string, and they seem to be correct. When the switch boots up, I print out what is being passed to the EndLoad function: pDevTbl->unit = 0, pDevTbl->endLoadFunc = 14dc4, pDevTbl->endLoadString = 0x00800000:1:1:-1:-1:0:0x40004000 muxDevLoad failed for device entry 0! All I get is that muxDevLoad failed. I have little idea why because I can't get into the driver, and see what it is doing. I tried compiling the VxWorks version of the driver and replacing the Broadcom driver object file with the new one I just created. Needless to say, there seem to be multiple incompatibilities, and booting with this driver halts the system. My debugging is very limited in this setup. I can't use gdb because I can't run this code on Solaris. I'm not that familiar with Tornado. I have looked through it and tried to play with it, but it seems every time I try to create a project based alone on the Broadcom SDK, it rewrites all the Makefiles, and becomes a total nightmare. I guess if I could get it to build in Tornado, I'd have better tools available to analyze the problem, but I just don't have time to fight with that. After spending half a day with Tornado, it doesn't seem very likely to me that this is the way the software wants to work. It wants only to base a BSP on one you have purchased and installed. Well, we've done that for the BCM5680 but I don't want to start with WindRiver code, because Broadcom is continually updating the SDK for fixes, and I want to build my code based on that. I have checked and re-checked the string passed to the EndLoad function. I have tried multiple different parameters, and none of them work. muxDevLoad always fails. In addition, further on down I get the following errors from the rootTask: 0x3fffdf8 (tRootTask): dc0: not enough memory provided Firmware Release: 0x3.0.03fffdf8 (): task dead Built dc0: not enough memory provided Fri Jul 27 000x:05:20 PDT 20013fffdf8 (): task deadmuxLoad failed! There are two messages here interleaved with the time they made it to the console - the firmware release message, and the error message. I believe it should look like this: 0x3fffdf8 (tRootTask): dc0: not enough memory provided 0x3.0.03fffdf8 (): task dead dc0: not enough memory provided 0x3fffdf8 (): task deadmuxLoad failed! Firmware Release: Built Fri Jul 27:05:20 PDT 2001 Here it says not enough memory provided. I've checked the memory configuration, and it appears to be ok. I'm not sure exactly why I get this error. From the DEC driver code supplied by WindRiver, I see the following definition for the parsed string: The driver provides one standard external interface, dec21x4xEndLoad(), which takes a string of colon separated parameters. The parameters should be specified as hexadecimal strings, optionally preceded by "0x" or a minus sign "-". Although the parameter string is parsed using strtok_r(), each parameter is converted from string to binary by a call to strtoul(parameter, NULL, 16). The format of the parameter string is: "::::::::" Here I see 9 parameters. Broadcom code only passed 8. The missing parameter was the unit number (which is 0). Adding that proved to be useless in solving the problem. I'm stuck here. I don't know exactly what the dec21x4xEndLoad( ) function is doing and where it is failing. Unit number in this case is 0 because it is the first and only entry in the endDevTbl [ ] table. Device address is defined as "..the base address at which the hardware device registers are located". This corresponds to #define ENET_DEV_BASE 0x80000000 in BSP.h file. I also have this defined as #define PCI_ENET_MEMADDR 0x80000000 in config.h in addtion I have #define PCI_ENET_IOADDR 0x00800000 IVEC is one indicated by #define MOUSSE_IRQ_ENET 1 ILEVEL is 1 Membase and Memsize are both -1 here. The WindRiver driver code says about membase: "This parameter specifies the base address of a DMA-able, cache-free, pre-allocated memory region for use as a memory pool for transmit / receive buffers. If there is not pre-allocated memory available for the driver, this parameter should be -1. In which case, the driver allocates cache safe memory for its use using cacheDmaAlloc( ). And the memsize variable corresponds: "The memory size parameter specifies the size of the pre-allocated memory region. If memory base is specified as -1, the driver ignores this parameter. User flags are 0 for none. Offset is 0x40004000 So, the string I'm trying to pass is #define DEC_21X4X_STRING "0x80000000:0:1:1:-1:-1:0:0x40004000" #define DEC_21X4X_STRING "0:0x80000000:0x80000000:1:1:-1:-1:0:0x40004000" I've also tried: #define DEC_21X4X_STRING "0:0x00800000:0x80000000:1:1:-1:-1:0:0x40004000" and some other variants of this to no avail. I have no idea where the driver fails because I don't have the source for the object code I'm using. I have no debugging facility available once it enters that object code. I can't step through it, etc. Can anyone provide any clues as to how I can go about resolving this problem? I have been working on VxWorks for one week now so be gentle. Thanks in advance, Davis --------------------------- Newsgroups: comp.os.vxworks Subject: Re: .o file format? Date: 27 Jul 2001 16:14:35 -0400 From: Bill Pringlemeir Organization: Bell Sympatico Message-ID: References: <3b61b222$1@news.ptt.ru> Sender: bpringlemeir@DeadDuck Dmitry> Where can I get .o object file format description of for Dmitry> Tornado? "Wisdom is like the dew on the grass before the first bird sings grasshopper." Doesn't make sense? Neither does your question. Have you read the documentation? NO. Hmmm, you should look at the vxWorks documentation and learn about the architecture that you are using. I assume that you are using the Gnu tools. Depending on your architecture, you will have either `ELF' or `COFF' object files. These are the `.o' files. Try this really cewl web sight, "www.google.com". Then type in either ELF or COFF and read and read and read (insert `learn' somewhere). hth, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Not able to ping to the board Date: Fri, 27 Jul 2001 17:12:53 -0400 From: Chandwani Organization: Lucent Technologies Message-ID: <3B61D955.F9F2277D@ubbsrv3.ho.lucent.com> References: <95ccca9f.0107260402.66a211c5@posting.google.com> Check the MAC address on ur. board for duplication in the network. Best is if u are using this board in ur private subnet, give it a new MAC address which you are sure is not being used. We had this problem and we took half day to realize this. Use ifShow on the windshell to see MAC address. Other things to try : Bring interface UP/DOWN to force it to transmit the ARP packets. Put a Sniffer on the subnet to see what ARP packets the board is sending out. Rakesh - --- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to measure CPU utilization of a task at run time Date: Fri, 27 Jul 2001 17:14:12 -0400 From: Chandwani Organization: Lucent Technologies Message-ID: <3B61D9A4.24823DCE@ubbsrv3.ho.lucent.com> References: <3B5F22AE.70C159C1@ntl.inf.com> Use Profilescope tool if you have purchased it with ur. Tornado Suite. Rakesh - --- Sunil Kulkarni wrote: > > How to measure the CPU utilization of a task in runtime > > so that some action can be taken based on the % CPU > > utilization of the task. > > Thanks, > > Sunil --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to communicate using Serial Port in Vxworks Date: Fri, 27 Jul 2001 17:19:49 -0400 From: Chandwani Organization: Lucent Technologies Message-ID: <3B61DAF5.62E3D58C@ubbsrv3.ho.lucent.com> References: <92dbccc1.0107190024.42f7c61f@posting.google.com> You need to include the #define CONSOLE_TTY and define it to 0 or 1 based on which port you want to use as Console. Look into the config.h of ur. BSP target/config//config.h for this define. Then rebuild the boot rom with this change and load it into the board. Rakesh - --- Adom wrote: > > Hi, > > In my project ,i must use Terminal to config my system parameters > (PPC860,Vxwors),but i don't know how to communicate using Serial Port. > Can you give a suggest? > Thanks! > Adom > 2001-7-19 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Making bootrom with Kernel to load in ROM Date: Fri, 27 Jul 2001 22:13:45 GMT From: Tim Sohacki Organization: Road Runner - NC Message-ID: <3B61E8A0.4E44D408@nc.rr.com> References: <9jp7fm$3e2$1@news.univ-angers.fr> Jonathan ILIAS wrote: > > Hi all, > > I want to build a bootrom that is able to load VxWorks system from ROM. I > read the documentation about it and I didn't found anything about it, all > solutions needs to download the VxWorks image file from the host system > (from network or serial line). > > Is it possible to do this without writing my own bootrom ? > > Thanks, > Jonathan ILIAS Jonathan, I'm not sure if you are asking the right question. Using Wind River terminology, a "bootrom" image loads a "vxWorks" image, as you say, typically over a serial or network link. There are ways to hack up a Wind River boot ROM to load a VxWorks image from ROM if you really want to do this. You can use a file system on your ROM, for example, and have the boot ROM load vxWorks from a file in ROM. TFFS (True Flash File System) can help you do this. However, what are you trying to accomplish? If you just want to boot a VxWorks load directly from ROM, you can use the image type "vxWorks_rom." This image begins execution in ROM, and bootstraps itself to RAM. (But on the PowerPC, you quickly run into the 1 MB size limitation, from 0xfff0_0100 to 0xffff_ffff.) Cheers, Tim. --------------------------- Newsgroups: comp.os.vxworks Subject: vxWorks boot image Date: Fri, 27 Jul 2001 22:39:01 +0000 (UTC) From: ydai2000@yahoo.com (Yi Dai) Organization: Mailgate.ORG Server - http://www.Mailgate.ORG Message-ID: <20010727223858.23365.qmail@web3406.mail.yahoo.com> Hi, I'm puzzled by the difference of the vxWorks size value as below. Can = someone here give me the clue?=20 Thanks, /Frank - ---------------------------------------- The vxWorks image size indicated by the Tornado compiler when it = finished is as: vxWorks: 2114432(t) + 874184(d) + 328566(b) =3D 3317182 (811586 unused) However, when I check the vxWorks image on my disk, I find the actual = size is 4,187,484 bytes which is about 870k larger.=20 Is this difference nornal? btw, I use Tornado 2.0 __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/ - -- Posted from web3406.mail.yahoo.com [216.115.111.60] via Mailgate.ORG Server - http://www.Mailgate.ORG --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks boot image Date: 27 Jul 2001 18:57:59 -0400 From: Bill Pringlemeir Organization: Bell Nexxia Message-ID: References: <20010727223858.23365.qmail@web3406.mail.yahoo.com> Sender: bpringlemeir@DeadDuck >>>>> "Yi" == Yi Dai writes: [snip] Yi> vxWorks: 2114432(t) + 874184(d) + 328566(b) = 3317182 (811586 unused) Yi> However, when I check the vxWorks image on my disk, I find the Yi> actual = size is 4,187,484 bytes which is about 870k larger.=20 Yi> Is this difference nornal? The file `vxWorks' is not a binary image. [goto vxWorks 101 class] It is either an `ELF' or `Coff' format file. Also, the BSS number would not be included in your image. I think that you can run, 'make vxWorks.bin' This should give a file size of 2114432(t) + 874184(d). The command is something like `elftoBin.exe' or `coffArmToBin.exe or something like that. You can look at $(WIND_BASE)/target/h/make/rules.bsp to see what targets you can build. Look at the `vxWorks' file with a hex editor to see all the symbols (or dump it with "strings vxWorks"). hth, Bill Pringlemeir. - -- Useful vxWorks URLS, "http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html" "http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt9.html" --------------------------- Newsgroups: comp.os.vxworks Subject: nfsMount error resolution needed Date: Fri, 27 Jul 2001 23:52:29 GMT From: Gururaj patkar Organization: Optimum Online Message-ID: <3B61FE3B.B1FAEC0D@optonline.net> Hi, I have nfsServer on a slave card and nfsClient on a master card. I have need to do download some files from master to the slave under a particular CLI command. I am doing this using NFS. The slave card informs the master card of its "hostname","ipaddress","mount point dir" during its initial interaction with the master card (prior to this, the slave card has nfsExported the corresponding "mount point dir"). When the master card gets this info it tries to nfsMount the "mount point dir" to a local mount point. However, nfsMount fails with the error - RPC_PROGUNAVAIL. why am I gettin this error? The config.h for slave and master card has - INCLUDE_NFS INCLUDE_NFS_SERVER INCLUDE_RPC. your is appreciated. gururaj --------------------------- Newsgroups: comp.os.vxworks Subject: web server instead of windweb? Date: Fri, 27 Jul 2001 16:07:50 +0900 From: "ozlab" Organization: A poorly-installed InterNetNews site Message-ID: <9jr5qr$37j$1@news.kornet.net> Hi all, Is there any web server for vxworks instead of windweb package? I just wonder. ====================================== Eung-Shik Kim (Ed) ed@ebridgecom.com PowerVoIP Gateway Hardware Devl. Dept. 467-12 Dogok-Dong Kangnam-Gu MMAA Bldg 29th Fl. Seoul (zip 135-270) KOREA, Republic of Tel+82 2 575 6476 ext 517, 3713 Fax+82 2 575 6478 C.P. +82 11 9704 0840 ICQ#61658172 / Send me plain text please... ======================================= --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Question about array definition Date: 27 Jul 2001 18:38:39 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107271738.5ac8d5e7@posting.google.com> References: <44a01eba.0107261819.39d95ba1@posting.google.com> <3b61233f$0$4100$6e365a64@newsreader02.highway.telekom.at> Thanks, Maybe I didn't express my thought expressly. In one word, I want to alloc a memory block at a fixed address and reserve it from the system. ellin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Man pages for Tornado Release 2.0v4 Date: Fri, 27 Jul 2001 21:51:54 -0400 From: joe durusau Organization: BELLSOUTH.net & WebUseNet Corp - "ReInventing the UseNet" Message-ID: <3B621AB9.804AA951@bellsosuth.net> References: <3B606D56.185DD954@roses.bna.boeing.com> For some ignorant reason, they don't exist on tornado 2. There are a bunch of html files strewn around instead. Someone posted a script some time ago that does a pretty good job. You might find it on Google. Speaking only for myself, Joe Durusau "William H. McCloskey" wrote: > The following man locations no longer exist in Rel 2.0v4, > > alias tman 'man -M \ > $WIND_BASE/host/man \ > $WIND_BASE/host/sun4-solaris2/man \ > $WIND_BASE/target/man ` > > I'm moving on from 5.3.1, which has a more commmand line "friendly" look > and feel. > > Please advise as to where I can pick up the Rel 2.0v4 man pages. > > Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: question about IP over ATM Date: Sat, 28 Jul 2001 05:26:40 GMT From: Sperry Family Message-ID: <3B624D24.F4B76BA7@gte.net> References: <9jo1jk$e0k$1@news.cz.js.cn> <3B61735A.B9F8F873@icn.siemens.de> I have implemented several ENDs for ATM and Frame Relay. It is possible to set up one END as a point-to-point link using interface aliasing such that each PVC uses its own IP-pair. The IP address assigned to the interface can be the same for each PVC, or different for each PVC, or a mix. The far end IP addresses can be anything (as long as they're all different). This causes a large route table full of host routes, and it's hard to set up as the vxWorks APIs don't comprehend such a thing (you have to use the Berkeley APIs), but it offers more flexibility for your network. Vic Sperry Paradyne Networks John Mackenzie wrote: > > Hi, > > I am also working on the very same. As far as my understanding > of Classical IP over ATM, you can have several PVCs terminating > on the same IP address i.e. only one IP address for several PVCs > . > In this way you define a LIS (a logical IP subnetwork). You then > need a static ATMARP table so that IP can lookup which PVC > it needs to send data on for a given remote IP address. I have > implemented this in my driver and provide a special address > resolution function which I install on the MUX with > muxAddrFuncAdd() - I am using the Network Protocol Toolkit > - but the driver is almost identical to an END driver. NPT makes > it easier to do non.Ethernet drivers. > > IP then passes this proprietry hardware address (could be simply > a VPI/VCI) to your driver and you can send the data on the right > PVC. When you receive data you just pass it up IP as the source > IP address is in the IP header. > > Of course all remote sides must use an IP address on the LIS and > it could be you implement several driver instances (with different > IP addresses) to support more than one LIS. > > When SVCs are uses and ATMARP server is needed on the LIS. > But for PVCs you can just do it all statically. > > Regards, > > John Mackenzie > > Bruce Shi wrote: > > > Hi,all: > > I am doing a END driver to implement IP over ATM in VxWorks 5.4. I will > > use PVC(permanent virtual circuit). According to RFC1577, PVC(VPI,VCI) is > > correspondind to IP address, so if I have several PVC, we need IP address of > > the same number. but we only have an END driver, does END can be binded to > > several IP address? Thank you for any answer! > > > > best regards > > bruce --------------------------- Newsgroups: comp.os.vxworks Subject: Re: END Driver Woes (LONG) Date: Sat, 28 Jul 2001 05:42:31 GMT From: Sperry Family Message-ID: <3B6250DA.71EF4CE3@gte.net> References: <9jsgjk$2qtm$1@msunews.cl.msu.edu> This is a guess. Maybe you don't have enough network buffer pools available? ENDs allocate Mblks, ClBlks and Mbufs from special memory pools. The total amount of memory available to all ENDs (and to the route tables and protocols) is defined by some global constants in a header file. I don't have my vxWorks source tree available to me now, but I think the header file is mbuf.h. The network developer's guide talks about these constants. Hope this helps. Vic Sperry Davis Ford wrote: > > Hi, > > I have an SDK package from Broadcom for their BCM5680 Strataswitch. The SDK > contains basically all the source code necessary for building the bootrom > and VxWorks image, except it does grab some headers and libraries from a > full-blown Tornado installation. WindRiver also supplies their own BSP for > the BCM5680. The two companies thus have two separate source trees for the > same thing. I am working on getting the Broadcom SDK to work. > > The problem I have is with the END driver for the Fast Ethernet management > interface. The onboard Ethernet controller uses the DEC21143 END Driver > (dec21x4xEnd.c). Broadcom only ships the object file for the driver...not > the source. I have no idea why. In the WindRiver BSP for BCM5680, the > source and header are included. However, there have been enough changes > between the two BSPs (Broadcom Vs. WindRiver) that there are numerous > incompatibilities between the two. You're stuck with trying to get one or > the other to work. > > My problem is that when muxDevLoad is called, it calls the > ec21x4xEndLoad( ) function and passes it an init string. From here, what > happens is anyone's guess. Like I said, I only have the object file. The > init string passed to the EndLoad function is as follows: > /* > * Default SENS device: DEC-21143-XX (onboard ethernet) > * Int Str= unit:MBAR:MBAR:ivec:ilvl:mem:memsize:usrFlags:offset > */ > #ifdef INCLUDE_END_DEC_21X4X > #define DEC_21X4X_FUNC dec21x4xEndLoad > #define DEC_21X4X_STRING "0x80000000:0:1:1:-1:-1:0:0x40004000" > mbar : mbar : ivec : ilvl : mem : memsz : flags : offset > 0x8 1 1 -1 -1 0 0x4 > IMPORT END_OBJ* DEC_21X4X_FUNC(char*, void*); > #endif /* INCLUDE_END_DEC21X4X */ > > I have checked the parameters for the string, and they seem to be correct. > When the switch boots up, I print out what is being passed to the EndLoad > function: > > pDevTbl->unit = 0, pDevTbl->endLoadFunc = 14dc4, pDevTbl->endLoadString = > 0x00800000:1:1:-1:-1:0:0x40004000 > muxDevLoad failed for device entry 0! > > All I get is that muxDevLoad failed. I have little idea why because I can't > get into the driver, and see what it is doing. I tried compiling the > VxWorks version of the driver and replacing the Broadcom driver object file > with the new one I just created. Needless to say, there seem to be multiple > incompatibilities, and booting with this driver halts the system. > > My debugging is very limited in this setup. I can't use gdb because I can't > run this code on Solaris. I'm not that familiar with Tornado. I have > looked through it and tried to play with it, but it seems every time I try > to create a project based alone on the Broadcom SDK, it rewrites all the > Makefiles, and becomes a total nightmare. I guess if I could get it to > build in Tornado, I'd have better tools available to analyze the problem, > but I just don't have time to fight with that. After spending half a day > with Tornado, it doesn't seem very likely to me that this is the way the > software wants to work. It wants only to base a BSP on one you have > purchased and installed. Well, we've done that for the BCM5680 but I don't > want to start with WindRiver code, because Broadcom is continually updating > the SDK for fixes, and I want to build my code based on that. > > I have checked and re-checked the string passed to the EndLoad function. I > have tried multiple different parameters, and none of them work. muxDevLoad > always fails. In addition, further on down I get the following errors from > the rootTask: > > 0x3fffdf8 (tRootTask): dc0: not enough memory provided > Firmware Release: 0x3.0.03fffdf8 (): task dead Built dc0: not enough memory > provided > Fri Jul 27 000x:05:20 PDT 20013fffdf8 > (): task deadmuxLoad failed! > > There are two messages here interleaved with the time they made it to the > console - the firmware release message, and the error message. I believe it > should look like this: > > 0x3fffdf8 (tRootTask): dc0: not enough memory provided > 0x3.0.03fffdf8 (): task dead dc0: not enough memory provided > 0x3fffdf8 (): task deadmuxLoad failed! > Firmware Release: Built Fri Jul 27:05:20 PDT 2001 > > Here it says not enough memory provided. I've checked the memory > configuration, and it appears to be ok. I'm not sure exactly why I get this > error. From the DEC driver code supplied by WindRiver, I see the following > definition for the parsed string: > > The driver provides one standard external interface, dec21x4xEndLoad(), > which takes a string of colon separated parameters. The parameters should > be specified as hexadecimal strings, optionally preceded by "0x" or a minus > sign "-". Although the parameter string is parsed using strtok_r(), each > parameter is converted from string to binary by a call to strtoul(parameter, > NULL, 16). > > The format of the parameter string is: > > ":::::: size>::" > > Here I see 9 parameters. Broadcom code only passed 8. The missing parameter > was the unit number (which is 0). Adding that proved to be useless in > solving the problem. I'm stuck here. I don't know exactly what the > dec21x4xEndLoad( ) function is doing and where it is failing. > > Unit number in this case is 0 because it is the first and only entry in the > endDevTbl [ ] table. Device address is defined as "..the base address at > which the hardware device registers are located". This corresponds to > > #define ENET_DEV_BASE 0x80000000 > > in BSP.h file. > > I also have this defined as > > #define PCI_ENET_MEMADDR 0x80000000 > > in config.h > > in addtion I have > > #define PCI_ENET_IOADDR 0x00800000 > > IVEC is one indicated by > #define MOUSSE_IRQ_ENET 1 > > ILEVEL is 1 > > Membase and Memsize are both -1 here. The WindRiver driver code says about > membase: "This parameter specifies the base address of a DMA-able, > cache-free, pre-allocated memory region for use as a memory pool for > transmit / receive buffers. If there is not pre-allocated memory available > for the driver, this parameter should be -1. In which case, the driver > allocates cache safe memory for its use using cacheDmaAlloc( ). And the > memsize variable corresponds: "The memory size parameter specifies the size > of the pre-allocated memory region. If memory base is specified as -1, the > driver ignores this parameter. > > User flags are 0 for none. > Offset is 0x40004000 > > So, the string I'm trying to pass is > > > #define DEC_21X4X_STRING "0x80000000:0:1:1:-1:-1:0:0x40004000" > > > #define DEC_21X4X_STRING "0:0x80000000:0x80000000:1:1:-1:-1:0:0x40004000" > > I've also tried: > > #define DEC_21X4X_STRING "0:0x00800000:0x80000000:1:1:-1:-1:0:0x40004000" > > and some other variants of this to no avail. > > I have no idea where the driver fails because I don't have the source for > the object code I'm using. I have no debugging facility available once it > enters that object code. I can't step through it, etc. Can anyone provide > any clues as to how I can go about resolving this problem? I have been > working on VxWorks for one week now so be gentle. > > Thanks in advance, > > Davis --------------------------- Newsgroups: comp.os.vxworks Subject: error of the task Date: 27 Jul 2001 23:50:39 -0700 From: feng_w@hotmail.com (wang feng) Organization: http://groups.google.com/ Message-ID: <6ccda887.0107272250.95aea24@posting.google.com> hi, guys, I am a newie to vxworks, when I use tonato2.0 shell to inspect my task's status, the meaning of 'errno' column bother me a lot, just like '20100004','3006c', how to check the error? thnks in advance, Wangfeng ps: here is what the screen looks like, - -> i NAME ENTRY TID PRI STATUS PC SP ERRNO DELAY - ---------- ------------ -------- --- ---------- -------- -------- - ------- ----- tExcTask _excTask 7fb5134 0 PEND 18923a 7fb50a4 0 0 tLogTask _logTask 7fb2830 0 PEND 18923a 7fb279c 0 0 tShell _shell 7efa11c 1 PEND 115424 7ef9df0 0 0 tWdbTask 0x139c80 7efb2c0 3 READY 115424 7efb21c 0 0 TimerTask _TimerTask 7c36c40 20 DELAY 11479e 7c36bf0 0 1 tNetTask _netTask 7f0f0f8 50 READY 11538b 7f0efd8 0 0 tFtpdTask 0x13ff00 7efea18 55 PEND 115424 7efe94c 0 0 tMsgRecv0 _MsgRecvTask 7c224b8 75 PEND+T 115424 7c223a0 0 804 tConnctChec_ConnectChec 7c1e324 85 DELAY 11479e 7c1e290 0 349 SYSCTRL _TaskEntry 7c2ba34 200 PEND 18923a 7c2b9ac 0 0 BRDCTRL _TaskEntry 7c27840 200 PEND 18923a 7c277b8 0 0 BRDMONI _TaskEntry 7c2364c 200 PEND 18923a 7c235c4 0 0 DB_APP _TaskEntry 7bab0f8 200 PEND 18923a 7bab070 0 0 AMAMS _TaskEntry 7ba6f04 200 PEND 18923a 7ba6e7c 20100004 0 AMACL _TaskEntry 7ba2d10 200 PEND 18923a 7ba2c88 20100004 0 CLIS _TaskEntry 7b9eb1c 200 PEND 18923a 7b9ea94 0 0 MDIA _TaskEntry 7b9a928 200 PEND 18923a 7b9a8a0 20100004 0 AMCNF _TaskEntry 7b96734 200 PEND 18923a 7b966ac 20100004 0 tDcacheUpd _dcacheUpd 7f91148 250 DELAY 11479e 7f910f0 3006c 21 value = 0 = 0x0 - -> --------------------------- Newsgroups: comp.os.vxworks Subject: download image to vxsim Date: Sat, 28 Jul 2001 00:16:01 -0700 From: "Jin Li" Organization: Giganews.Com - Premium News Outsourcing Message-ID: Dear Gurus: I am a beginer to vxwork. I downloaded a trial version of Tornado II and wanted to go through the tutorial. After I compiled and wanted to download to vxsim, here is the error message I have. Could anyone tell me what it is and how to fix the problem? Thanks a lot. tgtsvr (vxsim@hotsky): Sat Jul 28 00:08:10 2001 Checking License ...FAILED Cannot get a valid license. Failed to receive a Target Server license. Either your network or license server is down or your LM_LICENSE_FILE variable is not set to a valid license server. Cannot find license file The license files (or server network addresses) attempted are listed below. Use LM_LICENSE_FILE to use a different license file, or contact your software provider for a license file. Feature: tornado Filename: C:/Tornado\.wind\license License path: C:/Tornado\.wind\license FLEXlm error: -1,359. System Error: 2 "No such file or directory" For further information, refer to the FLEXlm End User Manual, available at "www.globetrotter.com". Target Server will exit --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jul 28 05:59:49 2001 From: "Ravi Malhotra" Date: Sat Jul 28 05:59:52 PDT 2001 Subject: Re: Download image to Vxworks This is a multi-part message in MIME format. ------=_NextPart_000_000A_01C11793.3C9343C0 Content-Type: multipart/alternative; boundary="----=_NextPart_001_000B_01C11793.3C9343C0" ------=_NextPart_001_000B_01C11793.3C9343C0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Official StationeryHi Jin, You probably need a FlexLM license. The best way to obtain this is via the windriver website. Login to windsurf using the login and passwd you registered with. Go to "Manage" - there is a section on FlexLM license. the required steps for obtaining the license are given there Hope this helps. Regards Ravi Malhotra Newsgroups: comp.os.vxworks Subject: download image to vxsim Date: Sat, 28 Jul 2001 00:16:01 -0700 From: "Jin Li" Organization: Giganews.Com - Premium News Outsourcing Message-ID: GAt87.2890$%7.54399@news6.giganews.com Dear Gurus: I am a beginer to vxwork. I downloaded a trial version of Tornado II and wanted to go through the tutorial. After I compiled and wanted to download to vxsim, here is the error message I have. Could anyone tell me what it is and how to fix the problem? Thanks a lot. tgtsvr (vxsim@hotsky): Sat Jul 28 00:08:10 2001 Checking License ...FAILED Cannot get a valid license. Failed to receive a Target Server license. Either your network or license server is down or your LM_LICENSE_FILE variable is not set to a valid license server. Cannot find license file The license files (or server network addresses) attempted are listed below. Use LM_LICENSE_FILE to use a different license file, or contact your software provider for a license file. Feature: tornado Filename: C:/Tornado\.wind\license License path: C:/Tornado\.wind\license FLEXlm error: -1,359. System Error: 2 "No such file or directory" For further information, refer to the FLEXlm End User Manual, available at "www.globetrotter.com". Target Server will exit ------=_NextPart_001_000B_01C11793.3C9343C0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Official Stationery

Hi Jin,
 
You probably need a FlexLM = license.=20
The best way to obtain this is = via the=20 windriver website.
Login to windsurf using the = login and=20 passwd you registered with.
Go to "Manage" - there is a = section on=20 FlexLM license.
the required = steps for=20 obtaining the license are given there
 
Hope this = helps.
 
Regards
Ravi Malhotra
 
 
Newsgroups: comp.os.vxworks
Subject: download image to vxsim
Date: Sat, 28 Jul 2001 00:16:01 -0700
From: "Jin Li" <jcli007@hotmail.com>
Organization: Giganews.Com - Premium News Outsourcing
Message-ID: GAt87.2890$%7.5439= 9@news6.giganews.com
 
 
Dear Gurus:
I am a beginer to vxwork. I downloaded a trial version of = Tornado II=20 and
wanted to go through the tutorial. After I compiled and wanted = to=20 download
to vxsim, here is the error message I have. Could anyone tell = me what=20 it is
and how to fix the problem? Thanks a lot.
tgtsvr (vxsim@hotsky): Sat Jul 28 00:08:10 2001
Checking License ...FAILED
Cannot get a valid license.
Failed to receive a Target Server license.
Either your network or license server is down or your=20 LM_LICENSE_FILE
variable is not set to a valid license server.
Cannot find license file
The license files (or server network addresses) attempted = are
listed below. Use LM_LICENSE_FILE to use a different license=20 file,
or contact your software provider for a license file.
Feature: tornado
Filename: C:/Tornado\.wind\license
License path: C:/Tornado\.wind\license
FLEXlm error: -1,359. System Error: 2 "No such file or=20 directory"
For further information, refer to the FLEXlm End User = Manual,
available at "www.globetrotter.com".
 
Target Server will exit
 
 
------=_NextPart_001_000B_01C11793.3C9343C0-- ------=_NextPart_000_000A_01C11793.3C9343C0 Content-Type: text/x-vcard; name="Ravi Kumar Malhotra (E-mail).vcf" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="Ravi Kumar Malhotra (E-mail).vcf" ------=_NextPart_000_000A_01C11793.3C9343C0-- From vxwexplo-errs@csg.lbl.gov Sun Jul 29 04:03:21 2001 From: Vxworks Exploder Date: Sun Jul 29 04:03:23 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 29 04:03:05 PDT 2001 Subject: wanted: chkdsk patch or alternative Subject: Re: error of the task Subject: Re: Making bootrom with Kernel to load in ROM Subject: Re: download image to vxsim Subject: Re: char and INT8 are unsigned types ? Subject: re: Tell Frederick it's messy kicking under a thought. Subject: re: Gawd Geoffrey will annoy the newbie, and if Ronald happily interfaces it too, the CDROM will interface outside the overloaded window. Subject: re: Lots of sticky postmasters are tall and other abysmal ROMs are fast, but will Sheri relay that? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: wanted: chkdsk patch or alternative Date: Sat, 28 Jul 2001 09:19:56 -0400 From: Klein family Message-ID: <3B62BBF8.665CC159@erols.com> Does anyone know of a patch to chkdsk (DOSFS2) or an alternative that works? We are opening files on the disk, allocating contiguous disk space for them and then cycling power to the CPU without closing the files. After rebooting, we run chkdsk & then delete those files & try again. Chkdsk displays a few thousand error messages about files being cross-linked before reporting that errors have been fixed. Running dosFsShow gives very inconsistant results: running this process on a newly formatted disk, dosFsShow says that the available disk space is much less than the largest contigous block available from the chkdsk report. After going through this process a number of times without re-formatting the disk, dosFsShow says that my available disk space is greater than the size of my disk. (At one point, I had 140,000GB free on a 73GB disk.) So, I've determined that chkdsk does not fix the "available disk space" table and has difficulities with "cross linked" files, what can I do about it? (My WRS tech support guy hasn't gotten me an answer yet.) Thanks Scott Klein Northrop Grumman Baltimore, MD 410-765-0910 scott_b_klein@md.northgrum.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: error of the task Date: Sat, 28 Jul 2001 20:30:18 GMT From: Harvey Taylor Organization: MTS Internet Message-ID: <3B633C33.742E@despam.autobahn.mb.ca> References: <6ccda887.0107272250.95aea24@posting.google.com> In article <6ccda887.0107272250.95aea24@posting.google.com>, wang feng wrote: > > I am a newie to vxworks, when I use tonato2.0 shell to inspect my > task's status, the meaning of 'errno' column bother me a lot, just > like '20100004','3006c', how to check the error? > This is a FAQ. See: http://www.xs4all.nl/~borkhuis/vxworks/vxfaq.html#3.2. - -het - -- Everyone has the right to life, liberty and security of person. - -UDHR:Article 3: http://www.autobahn.mb.ca/~het/udhr.html H.E. Taylor http://www.autobahn.mb.ca/~het/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Making bootrom with Kernel to load in ROM Date: Sat, 28 Jul 2001 15:35:15 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B633E23.49D57938@covad.net> References: <9jp7fm$3e2$1@news.univ-angers.fr> Reply-To: drdiags@covad.net Jonathan, You can also build a vxWorks ROM-Resident bootrom or as suggested earlier, a vxWorks Standalone image. HTH, but you could also visit the vxWorks FAQ listed at the end of this post. Jonathan ILIAS wrote: > > Hi all, > > I want to build a bootrom that is able to load VxWorks system from ROM. I > read the documentation about it and I didn't found anything about it, all > solutions needs to download the VxWorks image file from the host system > (from network or serial line). > > Is it possible to do this without writing my own bootrom ? > > Thanks, > Jonathan ILIAS - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: download image to vxsim Date: Sat, 28 Jul 2001 16:08:15 -0700 From: DrDiags Organization: laserlink.net Message-ID: <3B6345DF.F15EF630@covad.net> References: Reply-To: drdiags@covad.net Jin Li, I haven't downloaded this and tried it myself, but when you got the download and installed, was there any mention of a temporary license file. Since this product is probably time-bombed, it will use a license.dat file or license file and FlexLM to control the use of the target server (and other tools, maybe?). Since you are running some version of Windows, I would do a search for this file, and set my environment variable LM_LICENSE_FILE to the full path and name of the file. As you see from the error message, the tool believes that the FlexLM license file is located in C:/Tornado/.wind and is named "license". See if you find a file in this location called license.dat or license.txt or just search the entire installation directory for something like these names. (Review the release notes or instruction, because there should be some mention of how to setup this feature). Good Luck! Jin Li wrote: > > Dear Gurus: > > I am a beginer to vxwork. I downloaded a trial version of Tornado II and > wanted to go through the tutorial. After I compiled and wanted to download > to vxsim, here is the error message I have. Could anyone tell me what it is > and how to fix the problem? Thanks a lot. > > tgtsvr (vxsim@hotsky): Sat Jul 28 00:08:10 2001 > Checking License ...FAILED > Cannot get a valid license. > Failed to receive a Target Server license. > Either your network or license server is down or your LM_LICENSE_FILE > variable is not set to a valid license server. > > Cannot find license file > The license files (or server network addresses) attempted are > listed below. Use LM_LICENSE_FILE to use a different license file, > or contact your software provider for a license file. > Feature: tornado > Filename: C:/Tornado\.wind\license > License path: C:/Tornado\.wind\license > FLEXlm error: -1,359. System Error: 2 "No such file or directory" > For further information, refer to the FLEXlm End User Manual, > available at "www.globetrotter.com". > > Target Server will exit - -- http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html (vxWorks FAQ home page) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: char and INT8 are unsigned types ? Date: 28 Jul 2001 18:44:01 -0700 From: Chris Torek Organization: None of The Above Message-ID: <9jvpp1$k0f$1@elf.eng.bsdi.com> References: <3B600535.158F6609@stoye.de> Sender: torek@bsdi.com In article <3B600535.158F6609@stoye.de> Rainer Stelzer wrote: >We can see [unsigned plain "char"] on the the powerPC (MPC6xx and >MPC8xx) architecture as well as in the simulator. > >Is this a known feature of the compiler or a bug ? It is a feature. The PowerPC architecture is such that unsigned "char"s tend to be slightly more efficient, so that is the default. Use explicit "signed char" if you want signed 8-bit objects. >( If it's a feature I can't understand why INT8 is defined as char ) That could be considered to be a bug. :-) (Personally I detest most of these typedefs: people tend to use them as a substitute for thinking about the problem. But one must work within whatever system one has...) - -- In-Real-Life: Chris Torek, Wind River Systems El Cerrito, CA, USA Domain: torek@bsdi.com +1 510 234 3167 http://claw.eng.bsdi.com/torek/ (not always up) I report spam to abuse@. --------------------------- Newsgroups: news.admin.net-abuse.email,comp.os.vxworks Subject: re: Tell Frederick it's messy kicking under a thought. Date: Sun, 29 Jul 2001 17:54:32 GMT From: "Andrew Borgs" Organization: Bell Sympatico Message-ID: <826EC6E6.CA4B10AB@vasowyc.net> Will you cry alongside the node, if Edwin steadily beats the ISDN? Robette will inadvertently disconnect with Charlie when the blank cryptographers bind around the root CIA. While Javas seemingly contradict, the subroutines often smile within the dry modems. The postmaster generally trains the haphazard buffer. Hey George will create the ROM, and if Yvette wanly connects it too, the programmer will dump throughout the strong field. The hard trackball rarely interfaces Francis, it proliferates Jon instead. Why did Andrew propagate to all the analysts? We can't engulf robots unless Stephanie will annually start afterwards. The ugly new zipdisks stupidly distribute as the unique webmasters exclude. It's very junk today, I'll disrupt admiringly or Alexandra will learn the warnings. What will you save the out-of-date solid tablets before Penny does? The messy stupid TCP/IP produces over Jason's surreptitious core. Try inflateing the CERT's useless proxy and Raoul will prioritize you! Nowadays, Anne never reloads until Genevieve puts the vulnerable router badly. He will vexate daily if Jeremy's Pascal isn't lower. If the shiny discs can persevere biweekly, the bizarre IPaddr may compile more Sub Sevens. I restore chaotic chatrooms beneath the filthy dense DEA, whilst Simone unbelievably tolerates them too. I'd rather give lovingly than type with Madeleine's retarded noise. William knows, then Will smartly recycles a usable procedure outside Zamfir's module. Robert, have a important plotter. You won't consume it. Darcy will sell the lost error and moan it in back of its structure. As believably as Dolf disappears, you can insulate the hacker much more lazily. --------------------------- Newsgroups: news.admin.net-abuse.email,comp.os.vxworks Subject: re: Gawd Geoffrey will annoy the newbie, and if Ronald happily interfaces it too, the CDROM will interface outside the overloaded window. Date: Sun, 29 Jul 2001 19:23:13 GMT From: "Pearl Aquino" Organization: Bell Sympatico Message-ID: As lazily as James transports, you can push the network much more loudly. Occasionally, webmasters filter throughout cold rooms, unless they're inner. If you'll compile Varla's node with JPEGs, it'll seemingly outwit the machine. If you will propagate Yvette's website inside investigators, it will globally fetch the newsgroup. When doesn't Karl tolerate usably? Go cry a engineer! Nowadays, Edith never shoots until Susanne defiles the weak backdoor truly. Where will we preserve after Walter stops the moronic mail server's client? --------------------------- Newsgroups: news.admin.net-abuse.email,comp.os.vxworks Subject: re: Lots of sticky postmasters are tall and other abysmal ROMs are fast, but will Sheri relay that? Date: Sun, 29 Jul 2001 21:20:48 GMT From: "Jethro Chattan" Organization: Bell Sympatico Message-ID: Sometimes, procedures vend outside minor tapes, unless they're loud. Don't even try to type eerily while you're relaying to a worthwhile webmaster. One more bright function or scanner, and she'll weekly load everybody. It's very dry today, I'll get daily or Jezebel will fetch the pointers. Go take a engineer! Who did Anastasia disconnect the programmer over the usable cryptographer? Ella wants to corrupt lovingly, unless Pete stops librarians without Dilbert's keypad. The output lazily pulls the surreptitious newsspool. Cristof, have a filthy thought. You won't cry it. Timothy restores, then Frederic steadily collaborates a new bug beneath Geoff's Back Orifice. Alvin will hatefully prioritize outside Ann when the soft JPEGs defile about the cosmetic SOCKS. If the stupid plotters can negotiate generally, the unique ethernet may buy more infernos. Better obscure workstations now or Owen will locally delete them against you. I float upper newsgroups near the violent dense monument, whilst George sneakily kicks them too. Try not to prepare the servers crudely, preserve them admiringly. Just saveing about a client in front of the data center is too sly for Jessica to format it. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 30 00:14:22 2001 From: Toni Romero Date: Mon Jul 30 00:14:25 PDT 2001 Subject: RE:nfsMount error resolution needed >Newsgroups: comp.os.vxworks >Subject: nfsMount error resolution needed >Date: Fri, 27 Jul 2001 23:52:29 GMT >From: Gururaj patkar >Organization: Optimum Online >Message-ID: <3B61FE3B.B1FAEC0D@optonline.net> >Hi, >I have nfsServer on a slave card and nfsClient on a master card. I have >need to do download some files from master to the slave under a >particular CLI command. I am doing this using NFS. The slave card >informs the master card of its "hostname","ipaddress","mount point dir" >during its initial interaction with the master card (prior to this, the >slave card has nfsExported the corresponding "mount point dir"). When >the master card gets this info it tries to nfsMount the "mount point >dir" to a local mount point. However, nfsMount fails with the error - >RPC_PROGUNAVAIL. >why am I gettin this error? >The config.h for slave and master card has - >INCLUDE_NFS >INCLUDE_NFS_SERVER >INCLUDE_RPC. >your is appreciated. gururaj I think for master card you don't need to include the NFS_SERVER and also for the slave don't need to include the NFS. Why don't you try to execute prior the nfsmount in the master the nfsExportShow with the name of the slave? It should return a print of your "mount point dir" as exported. By other hand you should add in your usrNetwork.c of the slave nfsstat autentification(int progNum, int versNum, int procNum, struct sockaddr_in clientAddr, NFSD_ARGUMENT * nfsdArg) { return(NFS_OK); } and init the NFS server as #ifdef INCLUDE_NFS_SERVER if (nfsdInit (xx,x,xx,(FUNCPTR) autentification,0,0) == ERROR) return (ERROR); that is with an autentification Toni Romero --------------------------- From vxwexplo-errs@csg.lbl.gov Mon Jul 30 00:49:26 2001 From: Toni Romero Date: Mon Jul 30 00:49:29 PDT 2001 Subject: RV: RE:nfsMount error resolution needed > >Newsgroups: comp.os.vxworks > >Subject: nfsMount error resolution needed > >Date: Fri, 27 Jul 2001 23:52:29 GMT > >From: Gururaj patkar > >Organization: Optimum Online > >Message-ID: <3B61FE3B.B1FAEC0D@optonline.net> > > >Hi, > > >I have nfsServer on a slave card and nfsClient on a master card. I > have > >need to do download some files from master to the slave under a > >particular CLI command. I am doing this using NFS. The slave card > >informs the master card of its "hostname","ipaddress","mount point > dir" > >during its initial interaction with the master card (prior to this, > the > >slave card has nfsExported the corresponding "mount point dir"). When > >the master card gets this info it tries to nfsMount the "mount point > >dir" to a local mount point. However, nfsMount fails with the error - > >RPC_PROGUNAVAIL. > >why am I gettin this error? > > >The config.h for slave and master card has - > >INCLUDE_NFS > >INCLUDE_NFS_SERVER > >INCLUDE_RPC. > > >your is appreciated. > > gururaj > > I think for master card you don't need to include the NFS_SERVER and > also for the slave don't need to include the NFS. > Why don't you try to execute prior the nfsmount in the master the > nfsExportShow with the name of the slave? It should return a print of > your "mount point dir" as exported. > By other hand you should add in your usrNetwork.c of the slave > > nfsstat autentification(int progNum, int versNum, int procNum, struct > sockaddr_in clientAddr, NFSD_ARGUMENT * nfsdArg) > > { > return(NFS_OK); > } > > and init the NFS server as > #ifdef INCLUDE_NFS_SERVER > if (nfsdInit (xx,x,xx,(FUNCPTR) autentification,0,0) == ERROR) > return (ERROR); > that is with an autentification > > [Toni Romero] also you should add in usrNetwork.c of the master the > following code or execute it prior your TCP/IP command if your slave > programmable > > #ifdef INCLUDE_NFS > > hostAdd("slavename in the your network or cpu name as in boot > parameters","ipaddresslave"); > nfsAuthUnixSet ("slavename in the your network or cpu name as in > boot parameters", NFS_USER_ID, NFS_GROUP_ID, 0, (int *) 0); > > > Toni Romero > > --------------------------- > From vxwexplo-errs@csg.lbl.gov Mon Jul 30 04:04:20 2001 From: Vxworks Exploder Date: Mon Jul 30 04:04:22 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 30 04:03:09 PDT 2001 Subject: workQPanic:kernel work queue overflow Subject: Can netClPoolIdGet, netClusterGet and netClFree be called from ISR routine? Subject: Re: vxworks vs. threadx ? Subject: Re: download image to vxsim Subject: Re: error of the task Subject: How to get map file of a project .out? Subject: Re: How to get map file of a project .out? Subject: Q: End driver, Network Interface and etherHooks Subject: Selective Linking Subject: Re: Q: End driver, Network Interface and etherHooks Subject: Re: Selective Linking Subject: Re: web server instead of windweb? Subject: Re: Question about array definition Subject: Re: .o file format? Subject: Re: error of the task Subject: Re: workQPanic:kernel work queue overflow ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: workQPanic:kernel work queue overflow Date: 29 Jul 2001 17:29:00 -0700 From: gujc@huawei.com (Gu, Jiangchun) Organization: http://groups.google.com/ Message-ID: Hi, all! On my target board, the message "workQPanic:kernel work queue overflow" is printed on console, then the board reset. Mpc8260 is used, and version of vxWorks is 5.4. ATM driver(AAL5) is running. I guess the error is relevant with AAL5 receive interrupt, but I don't know what the real cause is. Any suggestion will be appreciated. Thanks in advance! Gu, Jiangchun --------------------------- Newsgroups: comp.os.vxworks Subject: Can netClPoolIdGet, netClusterGet and netClFree be called from ISR routine? Date: 29 Jul 2001 18:54:28 -0700 From: gujc@huawei.com (Gu, Jiangchun) Organization: http://groups.google.com/ Message-ID: Can netClPoolIdGet, netClusterGet and netClFree be called from ISR routine? Thanks in advance! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks vs. threadx ? Date: Mon, 30 Jul 2001 02:33:05 GMT From: "Vividh Siddha" Organization: AT&T Worldnet Message-ID: References: <3B27232F.55E4803@wipro.com> Try http://www.dedicated-systems.com/encyc/buyersguide/rtos/evaluations you can find come comparisons of RTOSes. Probably you will find something you need. All the best. Vividh "Vijay Kumar Peshkar" wrote in message news:3B27232F.55E4803@wipro.com... > Hi Friends, > > Any URLs/documents comparing the two OSs?. > > Thanks and Regards, > Vijay > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: download image to vxsim Date: 29 Jul 2001 22:21:53 -0700 From: u97124@yahoo.com (Pawan Joshi) Organization: http://groups.google.com/ Message-ID: References: Jin, We faced the same problem few days back. We wrote to license@windriver.com And this is what we got in reply. - ------------------------------------------------------------------------------- Hello, To run the target server, you need to obtain a license key. This key is not included in the setup as it is host specific. You can obtain the license key on-line via the following link, please follow the instructions provided: http://www.wrs.com/cgi-bin/csdocs/private/mksurfLogin.cgi?2 You should then place this file under $Wind_BASE/.wind/license where $Wind_BASE is where you have installed Tornado. If this directory does not exist, create one. Regards, - ------------------------------------------------------------------------------- Go to the link. Generate License key depending on your environment. Follow the instructions. Hope this will help. Regards - -Pawan "Jin Li" wrote in message news:... > Dear Gurus: > > I am a beginer to vxwork. I downloaded a trial version of Tornado II and > wanted to go through the tutorial. After I compiled and wanted to download > to vxsim, here is the error message I have. Could anyone tell me what it is > and how to fix the problem? Thanks a lot. > > tgtsvr (vxsim@hotsky): Sat Jul 28 00:08:10 2001 > Checking License ...FAILED > Cannot get a valid license. > Failed to receive a Target Server license. > Either your network or license server is down or your LM_LICENSE_FILE > variable is not set to a valid license server. > > Cannot find license file > The license files (or server network addresses) attempted are > listed below. Use LM_LICENSE_FILE to use a different license file, > or contact your software provider for a license file. > Feature: tornado > Filename: C:/Tornado\.wind\license > License path: C:/Tornado\.wind\license > FLEXlm error: -1,359. System Error: 2 "No such file or directory" > For further information, refer to the FLEXlm End User Manual, > available at "www.globetrotter.com". > > > Target Server will exit --------------------------- Newsgroups: comp.os.vxworks Subject: Re: error of the task Date: 29 Jul 2001 23:14:58 -0700 From: feng_w@hotmail.com (wang feng) Organization: http://groups.google.com/ Message-ID: <6ccda887.0107292214.5f0483e6@posting.google.com> References: <6ccda887.0107272250.95aea24@posting.google.com> <3B633C33.742E@despam.autobahn.mb.ca> hi, Many thanks for your help. I see errono '300bc' means '_TASK_UNDELAYED',but what about '20100004', if it is in the format of 'prefix + errorno', '2010' far over the range Best Regards, Wangfeng Harvey Taylor wrote in message news:<3B633C33.742E@despam.autobahn.mb.ca>... > In article <6ccda887.0107272250.95aea24@posting.google.com>, > wang feng wrote: > > > > I am a newie to vxworks, when I use tonato2.0 shell to inspect my > > task's status, the meaning of 'errno' column bother me a lot, just > > like '20100004','3006c', how to check the error? > > > > This is a FAQ. > See: http://www.xs4all.nl/~borkhuis/vxworks/vxfaq.html#3.2. > > > -het --------------------------- Newsgroups: comp.os.vxworks Subject: How to get map file of a project .out? Date: 29 Jul 2001 23:54:31 -0700 From: ellin_lin@263.net (ellin) Organization: http://groups.google.com/ Message-ID: <44a01eba.0107292254.2cde96fe@posting.google.com> I want to get the map file of a project created with T2. I knew that T2's linker utility 'ldppc' has a parameter of '-Map mapfile'. I add this parameter in the project build propertis dialogue form. But I get nothing except the .out file. Anyone has the experience of getting the map file of a project's .out? Thx. ellin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get map file of a project .out? Date: 30 Jul 2001 07:03:37 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <44a01eba.0107292254.2cde96fe@posting.google.com> ellin_lin@263.net (ellin) wrote: > I want to get the map file of a project created with T2. > I knew that T2's linker utility 'ldppc' has a parameter of '-Map > mapfile'. > I add this parameter in the project build propertis dialogue form. But > I get nothing except the .out file. Take a look here: http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt1.html#1.6-D Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Q: End driver, Network Interface and etherHooks Date: 30 Jul 2001 00:50:33 -0700 From: bmurr@my-deja.com (Brian Murphy) Organization: http://groups.google.com/ Message-ID: Can someone explain to be the relationship between an End driver and network interface. For example there are 2 files associated with the AMD 7990 LANCE, these are:- target/src/drv/netif/if_ln.c target/src/drv/end/ln7990End.c If I want to develop a new ethernet driver, do I have to develop both the End driver and the network interface? Is it essential that I write both? The Network Protocol Toolkit User Guide gives good instruction in Chapter 2 to write an End driver. However, it does not explain much about the network interface - how it should be built, which routines have to be created, how it is called within the TCP/IP stack. There is a call to ether_output () in if_ln.c, this does not appear to be documentation. Can anyone explain what is happening here? After writing the End driver my main is to be able to intercept messages as they are received from the driver before they are passed to the network layer. I want link layer to link layer traffic. I was hoping to do this using etherInputHook () and etherOutptuHook () however, these will not be supported in the future. How can I achieve the inteception of traffic without using these functions. How will the functionality of the etherxxxHook functions be replaced. Thanks in advance. Brian --------------------------- Newsgroups: comp.os.vxworks Subject: Selective Linking Date: 30 Jul 2001 01:01:50 -0700 From: pritamganguly@hotmail.com (Pritam) Organization: http://groups.google.com/ Message-ID: <166770ed.0107300001.5cbe48e1@posting.google.com> I am trying to find out a way to link only those modules that are referenced in a program so as to reduce the driver size. Is there any linker option for it in VxWorks. Something similar to /OPT:REF in VC++. Thanking in advance. Pritam --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Q: End driver, Network Interface and etherHooks Date: 30 Jul 2001 08:30:19 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: bmurr@my-deja.com (Brian Murphy) wrote: > Can someone explain to be the relationship between an End driver and > network interface. > > For example there are 2 files associated with the AMD 7990 LANCE, > these are:- > > target/src/drv/netif/if_ln.c > target/src/drv/end/ln7990End.c > > If I want to develop a new ethernet driver, do I have to develop both > the End driver and the network interface? > > Is it essential that I write both? No, just go for the END driver, and leave the netif file. The netif file is an older type of network interface, and is replaced by the END style driver. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Selective Linking Date: 30 Jul 2001 08:31:21 GMT From: "Johan Borkhuis" Organization: Agere Systems Message-ID: References: <166770ed.0107300001.5cbe48e1@posting.google.com> pritamganguly@hotmail.com (Pritam) wrote: > I am trying to find out a way to link only those modules that are > referenced in a program so as to reduce the driver size. Is there any > linker option for it in VxWorks. Something similar to /OPT:REF in > VC++. Build a library of your modules, and then link against this library. Now only the referenced modules will be inlcuded in the build. Groeten, Johan - -- o o o o o o o . . . _____________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | borkhuis@agere.com | >(________|__|_[_________]_|__________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks FAQ: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: web server instead of windweb? Date: Mon, 30 Jul 2001 11:47:32 +0200 From: "Werner Schiendl" Message-ID: <3b652dda$0$22346$6e365a64@newsreader02.highway.telekom.at> References: <9jr5qr$37j$1@news.kornet.net> Hi, Maybe GoAhead Webserver See http://www.goahead.com/ There is a web server that exists in a VxWorks port. regards Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Question about array definition Date: Mon, 30 Jul 2001 11:46:11 +0200 From: "Werner Schiendl" Message-ID: <3b652dd9$0$22346$6e365a64@newsreader02.highway.telekom.at> References: <44a01eba.0107261819.39d95ba1@posting.google.com> <3b61233f$0$4100$6e365a64@newsreader02.highway.telekom.at> <44a01eba.0107271738.5ac8d5e7@posting.google.com> To achive this, your best bet is probably to adapt sysLib.c of your BSP to your needs. This is where the amount of memory is detected and set up (usually). In the VxWorks users guide, there is a pretty complete decription of the boot process. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: .o file format? Date: Mon, 30 Jul 2001 11:50:07 +0200 From: "Werner Schiendl" Message-ID: <3b652e74$0$26752$6e365a64@newsreader02.highway.telekom.at> References: <3b61b222$1@news.ptt.ru> Hi, > ... Depending on your architecture, you will > have either `ELF' or `COFF' object files. These are the `.o' files. > one other possibility is A.OUT (e. g. i386 platform, IIRC) hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: error of the task Date: Mon, 30 Jul 2001 11:55:30 +0200 From: "Werner Schiendl" Message-ID: <3b652fb9$0$27366$6e365a64@newsreader02.highway.telekom.at> References: <6ccda887.0107272250.95aea24@posting.google.com> <3B633C33.742E@despam.autobahn.mb.ca> <6ccda887.0107292214.5f0483e6@posting.google.com> Hi, Module codes (the high-word of the error number) higher that 500 are free for application use. So your error codes stem from some of your (or a third party library's) code. See their documentation about the error value. hth Werner --------------------------- Newsgroups: comp.os.vxworks Subject: Re: workQPanic:kernel work queue overflow Date: Mon, 30 Jul 2001 12:01:21 +0200 From: "Werner Schiendl" Message-ID: <3b653117$0$28924$6e365a64@newsreader02.highway.telekom.at> References: Hi, There is a list of VxWorks system calls, that may be called from ISR context. These calls are put on the workQueue and processed later by the kernel in task context. If you call too many such functions in your ISR routine, you run out of space in that queue and this is what the error means. Try to eliminate as much VxWorks system calls as possible in your ISR routine(s). Your best bet is to save the data and notify a task that does the rest of the work. hth Werner --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 30 18:54:51 2001 From: karunakaran nair Date: Mon Jul 30 18:54:53 PDT 2001 Subject: startup scripts please throw more light on the startup scripts given with bootChange. what are those names separated by commas ? i see some ip address, some directory etc. i have found vxworks crazy here. there are no sample shell scripts, nothing. it is crazy documentation - no logic in the information furnished. __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/ From vxwexplo-errs@csg.lbl.gov Tue Jul 31 04:03:25 2001 From: Vxworks Exploder Date: Tue Jul 31 04:03:28 PDT 2001 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 31 04:03:10 PDT 2001 Subject: socket failing connect: persistent SYN_SENT state Subject: Re: RealTek 8139A Driver Subject: Re: RealTek 8139A Driver Subject: Booting using TSFS Subject: Re: workQPanic:kernel work queue overflow Subject: Re: Selective Linking Subject: Error while downloading __eabi Subject: Re: Selective Linking Subject: Re: PowerPC VME read problem Subject: Re: VxWorks Developpment on Linux Subject: Re: error of the task Subject: What happened to vxWorks archives on ftp.utd.ucar.edu ?? Subject: prjConfig.c in BSP for Power PC Subject: Re: VxWorks Developpment on Linux Subject: Re: get_object_identifier_async() Subject: Re: Error while downloading __eabi Subject: Where Is the FAQ and Archive of Code? Subject: intLock() or intDisable() ? Subject: excVecGet Subject: Problem in scaning user input ! Subject: Re: Problem in scaning user input ! Subject: Re: Problem in scaning user input ! Subject: Re: Error while downloading __eabi ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: socket failing connect: persistent SYN_SENT state Date: Mon, 30 Jul 2001 10:35:43 +0200 From: "Andrea Bigagli" Organization: Simtel Message-ID: <3b651c58.0@news.dada.it> Reply-To: "Andrea Bigagli" Hi vxworkers, I use vxWorks 5.4 on a i960 based board which controls 2 Motorola GPRS mobile phones. I start up a ppp link with each of the mobiles and use it to transfer files (HTTP and FTP) using sockets. Even if the sockets are all on the client side, i bind them to their respective ppp IP address to make sure that data flows in the 2 ppp links (I must do some measurements on each mobile so I need to have data flowing in each of them). Sometimes a file transfer goes in timeout and the software aborts the download with a shutdown(sock,2) followed by a close (sock). With inetstatShow I've seen this leaves the socket in the FIN_WAIT_1 state. Now comes the problem: when this happens (timeout in transfer) the next file transfers fail in the connect phase and the sockets remain in the SYN_SENT phase. Snooping on the linux box which acts as the server for the file transfers (HTTP and FTP) reveals that the vxWorks target has sent the SYN packets and that the linux box has ACKed them and resent its SYN, but the third phase of the handshake never takes place... I've noticed this happens only with the ppp link which is not associated with the default route, but I really can't understand what makes the socket unable to answer to the SYN sent by the linux box. Actually this happens also using a PC as the server. I've also tried to change the linger time on the closing of the socket, to use only shutdown or only close, but nothing changes... Please if someone can point me in the right direction, I really don't know what else I can do... Thank you in advance... Andrea Bigagli Electronics Engineer Simtel, Advanced Telecommunications Technologies --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RealTek 8139A Driver Date: 30 Jul 2001 05:27:31 -0700 From: fred@faraday.com.tw (fred) Organization: http://groups.google.com/ Message-ID: References: <9js2n3$35ss$1@member.gcn.net.tw> "Joel Hannah" wrote in message news:... > Thanks Victor, > > Below is the problem as I described it to the tech support at Windriver: > > My name is Joel Hannah and I work for Alpha Beta Technolgies in Huntsville. > We purchased Tornado 2.0/VxWorks 5.4 almost a year ago and started > developing our real time application for video processing. Our VxWorks > target is an Advantech 5820 which has a RealTek 8139A ethernet controller on > board. When we purchased the Windriver tools, a local Windriver FAE > installed the latest driver for the RealTek part and told us a fully support > Windriver driver for the RealTek 8139A was imminent. > > Our application utilizes FTP for transfers between our VxWorks target and > other PCs. Initially our files were relatively small (<10K) but that has > since changed. In addition we have developed custom TCP/IP datalink > protocols (both TCP and UDP). Well as this app has matured, we have had > increasing problems with our ethernet transfers. We have timeouts, slow > downs and downright lockups. We can't even transfer a file at all if it over > 2MB in size. We have adjusted the TCP/IP buffer values but will no > noticeable improvements. That has left us with thinking that the driver may > have errors. > > > I hope this describes the problems for you Victor.... > Joel > > susanwan wrote in message > news:9js2n3$35ss$1@member.gcn.net.tw... > > Dear Sir: > > I am victor a Realtek's system design engineer. Could you told me > more > > detail about your problem. Althrough I am debuging some bug of Vxworks > > driver for 8139 series but I have too much experence in this Realtime of. > > Most of case I write DOS or Novell's server driver. Hope I could solve > your > > problem. > > > > victor > > > > Joel Hannah wrote in message > > news:rs_77.3$Og1.2431@e420r-atl1.usenetserver.com... > > > I am using a single board computer to host my VxWorks application on. I > have > > > been having TCP/IP problems and I have narrowed the problem down to > ReakTek > > > 8139A ethernet controler driver. I have been using a driver that a > Windriver > > > FAE supplied to me with the stipulation that Windriver does not support > this > > > driver. He promised me that Windriver was going to support the 8139A but > I > > > haven't seen it yet. > > > > > > I was wondering if someone had a later version of this driver (I am > running > > > driver version: Release 1.0/EA3 > > > > > > I would really appreciate any help anyone could give me.... > > > > > > Joel > > > > > > -- > > > S. Joel Hannah > > > Senior Engineer > > > > > > Alpha Beta Technologies, Inc > > > 3112 12th Ave SW > > > Huntsville, AL 35805 > > > > > > Phone: (256) 534-9067 > > > Fax: (256) 534-9069 > > > Pager: (256) 580-8947 > > > > > > > > > > > > > I am Developing TCP/IP Application on VxWorks.And I also use RealTek RTL8139C Etherent Controller.And When I Transmit a packet,it will hold on the FIFO,instead of send to the line immediately.When I reset the system,the packet transmit.BTW,I check the packet size is over 60bytes,which is not Runt packet. So,I wonder if the driver is something wrong or the chip? Someone ahve the experience?.Pls let me Know. Tks a lot. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RealTek 8139A Driver Date: Sun, 29 Jul 2001 20:39:56 +0800 From: "susanwan" Organization: GCNet(Reach and Range Inc.) Message-ID: <9k3kf5$dmhq$1@member.gcn.net.tw> References: <9js2n3$35ss$1@member.gcn.net.tw> fred wrote in message news:fc23fac1.0107300427.370ff25d@posting.google.com... > "Joel Hannah" wrote in message news:... > > Thanks Victor, > > > > Below is the problem as I described it to the tech support at Windriver: > > > > My name is Joel Hannah and I work for Alpha Beta Technolgies in Huntsville. > > We purchased Tornado 2.0/VxWorks 5.4 almost a year ago and started > > developing our real time application for video processing. Our VxWorks > > target is an Advantech 5820 which has a RealTek 8139A ethernet controller on > > board. When we purchased the Windriver tools, a local Windriver FAE > > installed the latest driver for the RealTek part and told us a fully support > > Windriver driver for the RealTek 8139A was imminent. > > > > Our application utilizes FTP for transfers between our VxWorks target and > > other PCs. Initially our files were relatively small (<10K) but that has > > since changed. In addition we have developed custom TCP/IP datalink > > protocols (both TCP and UDP). Well as this app has matured, we have had > > increasing problems with our ethernet transfers. We have timeouts, slow > > downs and downright lockups. We can't even transfer a file at all if it over > > 2MB in size. We have adjusted the TCP/IP buffer values but will no > > noticeable improvements. That has left us with thinking that the driver may > > have errors. > > > > > > I hope this describes the problems for you Victor.... > > Joel > > > > susanwan wrote in message > > news:9js2n3$35ss$1@member.gcn.net.tw... > > > Dear Sir: > > > I am victor a Realtek's system design engineer. Could you told me > > more > > > detail about your problem. Althrough I am debuging some bug of Vxworks > > > driver for 8139 series but I have too much experence in this Realtime of. > > > Most of case I write DOS or Novell's server driver. Hope I could solve > > your > > > problem. > > > > > > victor > > > > > > Joel Hannah wrote in message > > > news:rs_77.3$Og1.2431@e420r-atl1.usenetserver.com... > > > > I am using a single board computer to host my VxWorks application on. I > > have > > > > been having TCP/IP problems and I have narrowed the problem down to > > ReakTek > > > > 8139A ethernet controler driver. I have been using a driver that a > > Windriver > > > > FAE supplied to me with the stipulation that Windriver does not support > > this > > > > driver. He promised me that Windriver was going to support the 8139A but > > I > > > > haven't seen it yet. > > > > > > > > I was wondering if someone had a later version of this driver (I am > > running > > > > driver version: Release 1.0/EA3 > > > > > > > > I would really appreciate any help anyone could give me.... > > > > > > > > Joel > > > > > > > > -- > > > > S. Joel Hannah > > > > Senior Engineer > > > > > > > > Alpha Beta Technologies, Inc > > > > 3112 12th Ave SW > > > > Huntsville, AL 35805 > > > > > > > > Phone: (256) 534-9067 > > > > Fax: (256) 534-9069 > > > > Pager: (256) 580-8947 > > > > > > > > > > > > > > > > > > > > I am Developing TCP/IP Application on VxWorks.And I also use RealTek > RTL8139C Etherent Controller.And When I Transmit a packet,it will hold > on the FIFO,instead of send to the line immediately.When I reset the > system,the packet transmit.BTW,I check the packet size is over > 60bytes,which is not Runt packet. > So,I wonder if the driver is something wrong or the chip? > Someone ahve the experience?.Pls let me Know. > > Tks a lot. Do you use WindRiver's driver or develop by youself ? --------------------------- Newsgroups: comp.os.vxworks Subject: Booting using TSFS Date: Mon, 30 Jul 2001 14:00:44 +0100 From: Paul Black Organization: Oxford Semiconductor Ltd Message-ID: <3B655A7C.8CC8ADBA@oxsemi.com> I'm trying to download a VxWorks image using TSFS and I get the following: [VxWorks Boot]: @ boot device : tsfs unit number : 0 processor number : 0 host name : vienna file name : /vxWorks flags (f) : 0x0 Booting using TSFS... Make sure that your Target Server is started with -R[oot] option. Loading /tgtsvr/vxWorks... 431380 + error loading file: status = 0x502. Error loading file: errno = 0x502. tgtsvr (940@vienna): Mon Jul 30 13:56:00 2001 Wind River Systems Target Server: NT/Win95 version Connecting to target agent... succeeded. Attaching C++ interface... succeeded. Warning: No core file. Use -c option if you wish to use one. Attaching coff OMF reader for ARM CPU family... succeeded. Warning: Cannot redirect the target shell to the target server Console. This VIO might be already open and the redirection done. If the redirection does not work, restart the target and the target server. TSFS: opened file (T:\vxWorks). Error: rpccore backend client Server can't decode arguments TSFS: closed file (T:\vxWorks). Any clues as to what might be going on? - -- Paul Black Oxford Semiconductor Ltd Tel: +44 (0) 1235 824 909 25 Milton Park, Abingdon, Fax: +44 (0) 1235 821 141 Oxfordshire. OX14 4SH --------------------------- Newsgroups: comp.os.vxworks Subject: Re: workQPanic:kernel work queue overflow Date: Mon, 30 Jul 2001 15:07:02 +0200 From: Joern Tietjen Organization: =?iso-8859-1?Q?Dr=E4ger?= Medical AG & Co. KGaA Message-ID: <3B655BF6.1006FE6@draeger.com> References: "Gu, Jiangchun" wrote: > > Hi, all! > On my target board, the message "workQPanic:kernel work queue > overflow" is printed on console, then the board reset. Mpc8260 is > used, and version of vxWorks is 5.4. ATM driver(AAL5) is running. I > guess the error is relevant with AAL5 receive interrupt, but I don't > know what the real cause is. > VxWorks hangs when an ISR does not properly acknowledge its triggering interrupt condition. Continuously pending interrupt requests (which prevent the system from doing anything apart from the ISR) result in overflowing queues and system reboots. So, make sure your ISR for the AAL5 int removes the int request by acknowledging the approriate hardware and prevent the IRQ from being "overrun". HTH. Joern --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Selective Linking Date: 30 Jul 2001 07:44:40 -0700 From: pritamganguly@hotmail.com (Pritam) Organization: http://groups.google.com/ Message-ID: <166770ed.0107300644.6396fc12@posting.google.com> References: <166770ed.0107300001.5cbe48e1@posting.google.com> "Johan Borkhuis" wrote in message news:... > pritamganguly@hotmail.com (Pritam) wrote: > > > I am trying to find out a way to link only those modules that are > > referenced in a program so as to reduce the driver size. Is there any > > linker option for it in VxWorks. Something similar to /OPT:REF in > > VC++. > > Build a library of your modules, and then link against this library. Now > only the referenced modules will be inlcuded in the build. > > Groeten, > Johan Thanks. But I want a more better solution. To elaborate, I have got a class having around 500 public methods and it is required. i.e. I cannot break the class into smaller ones. In the application, normally only 10 to 20 methods are called. But as a object of that class is required all 500 methods are linked. In VC++ that /OPT:REF option caters to my need, but I am unable to find anything similar in VxWorks. Thanks Again. Pritam --------------------------- Newsgroups: comp.os.vxworks Subject: Error while downloading __eabi Date: Mon, 30 Jul 2001 10:08:24 -0500 From: Dinker Charak Organization: Fermi National Accelerator Laboratory Message-ID: <3B657868.71A55A33@fnal.gov> Hello. I created a simple Hello.c with just main() { } and compiled (right click in workspace window, then compile Hello.c) it and downloaded (using download object file icon in workspace) it, all using tornado. I get this error ... Errors while downloading /usr/products/VxWorks/TornadoII/target/proj/Project0/default/Hello.o: __eabi any idea? regards, dinker --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Selective Linking Date: Mon, 30 Jul 2001 16:27:28 +0100 From: Paul Black Organization: Oxford Semiconductor Ltd Message-ID: <3B657CE0.7E35AABF@oxsemi.com> References: <166770ed.0107300001.5cbe48e1@posting.google.com> <166770ed.0107300644.6396fc12@posting.google.com> pritamganguly@hotmail.com (Pritam) wrote: > > "Johan Borkhuis" wrote in message news:... > > pritamganguly@hotmail.com (Pritam) wrote: > > > > > I am trying to find out a way to link only those modules that are > > > referenced in a program so as to reduce the driver size. Is there any > > > linker option for it in VxWorks. Something similar to /OPT:REF in > > > VC++. > > > > Build a library of your modules, and then link against this library. Now > > only the referenced modules will be inlcuded in the build. > > > > Groeten, > > Johan > > Thanks. > But I want a more better solution. To elaborate, I have got a class > having around 500 public methods and it is required. i.e. I cannot > break the class into smaller ones. In the application, normally only > 10 to 20 methods are called. But as a object of that class is required > all 500 methods are linked. In VC++ that /OPT:REF option caters to my > need, but I am unable to find anything similar in VxWorks. > Thanks Again. > Pritam If all of the functions are not virtual, have you tried splitting the member functions in to separate files, creating a library with the various files in and linking against that? Paul --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PowerPC VME read problem Date: Mon, 30 Jul 2001 17:04:16 +0100 From: Jason Garnham Organization: BAE SYSTEMS Message-ID: <3B658580.210C2E56@baesystems.com> References: <9gkmsn$dqv$1@inn.jlab.org> <3B2E367C.D92FB2FC@sandia.gov> Cris/Don >> I am going to try to modify the VDW bit of the Universe chip's LSI2_CTL >> register to make 16-bit VME accesses the maximum. Does this solution sound >> correct to anyone? I've done this, it works fine. It does however limit the entire window to D16 access, this could be a problem if you have other hardware in the window that you wish to access using D32, but it was fine for me. >If you access byte 1, >then 0, then 3, then 2 it will make this a longword (32bit) access to >memory at address 0. You can override this behavior. An instruction >called EIEIO can be used between each access to force the PowerPC to >read/write the IO device before the next IO operation starts. I've done this too. It didn't work. Even when accessing only a single byte, by the time it hit the VME bus it was D32. I don't know why. Some UniverseII optimization no doubt. Jason --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Developpment on Linux Date: 30 Jul 2001 09:29:18 -0700 From: jtj3@hotmail.com (PC Gameplayer) Organization: http://groups.google.com/ Message-ID: References: <3B31EB91.56B0BAF9@gaci.fr> <3B61C3CD.9C39450E@qwest.net> > Hello, > > I am wondering whether such tools are available on Linux > (ie: Cross Compiler, remode debugger, remote shell ...) ? Tornado 3 is available under Linux. The downside is that T3 is tied to VxWorks AE. If you're already developing on VxWorks 5.4 you may not be able to move to this release, but if you haven't started development yet, you should be able to do what you want with T3... Jim > > If so, where may I download these materials ? > Thanks. > > -- > Christophe Rabourg _/_/_/ _/_/_/ _/_/_/ _/ 6 Av. de Norvege BP46 > tel: +33 1 69071686 _/ _/ _/ _/ _/ 91942 Courtaboeuf > fax: +33 1 69071358 _/ _/ _/_/_/_/ _/ _/ France > crabourg@gaci.fr _/_/_/ _/ _/ _/_/_/ _/ http://www.gaci.fr --------------------------- Newsgroups: comp.os.vxworks Subject: Re: error of the task Date: 30 Jul 2001 10:02:22 -0700 From: john_94501@yahoo.com (John) Organization: http://groups.google.com/ Message-ID: <488e459a.0107300902.4ef0e34f@posting.google.com> References: <6ccda887.0107272250.95aea24@posting.google.com> Hello, > I am a newie to vxworks, when I use tonato2.0 shell to inspect my > task's status, the meaning of 'errno' column bother me a lot, just > like '20100004','3006c', how to check the error? Unless a system call actually returns an error indication in its return value (normally either ERROR or NULL depending on the return type), then you should ignore the errno values. If you are not checking for error returns in your system calls already, then you should probably go through the code and add the checks. HTH, John... > thnks in advance, > > Wangfeng > > ps: here is what the screen looks like, > > > -> i > NAME ENTRY TID PRI STATUS PC SP > ERRNO DELAY > ---------- ------------ -------- --- ---------- -------- -------- > ------- ----- > tExcTask _excTask 7fb5134 0 PEND 18923a 7fb50a4 > 0 0 > tLogTask _logTask 7fb2830 0 PEND 18923a 7fb279c > 0 0 > tShell _shell 7efa11c 1 PEND 115424 7ef9df0 > 0 0 > tWdbTask 0x139c80 7efb2c0 3 READY 115424 7efb21c > 0 0 > TimerTask _TimerTask 7c36c40 20 DELAY 11479e 7c36bf0 > 0 1 > tNetTask _netTask 7f0f0f8 50 READY 11538b 7f0efd8 > 0 0 > tFtpdTask 0x13ff00 7efea18 55 PEND 115424 7efe94c > 0 0 > tMsgRecv0 _MsgRecvTask 7c224b8 75 PEND+T 115424 7c223a0 > 0 804 > tConnctChec_ConnectChec 7c1e324 85 DELAY 11479e 7c1e290 > 0 349 > SYSCTRL _TaskEntry 7c2ba34 200 PEND 18923a 7c2b9ac > 0 0 > BRDCTRL _TaskEntry 7c27840 200 PEND 18923a 7c277b8 > 0 0 > BRDMONI _TaskEntry 7c2364c 200 PEND 18923a 7c235c4 > 0 0 > DB_APP _TaskEntry 7bab0f8 200 PEND 18923a 7bab070 > 0 0 > AMAMS _TaskEntry 7ba6f04 200 PEND 18923a 7ba6e7c > 20100004 0 > AMACL _TaskEntry 7ba2d10 200 PEND 18923a 7ba2c88 > 20100004 0 > CLIS _TaskEntry 7b9eb1c 200 PEND 18923a 7b9ea94 > 0 0 > MDIA _TaskEntry 7b9a928 200 PEND 18923a 7b9a8a0 > 20100004 0 > AMCNF _TaskEntry 7b96734 200 PEND 18923a 7b966ac > 20100004 0 > tDcacheUpd _dcacheUpd 7f91148 250 DELAY 11479e 7f910f0 > 3006c 21 > value = 0 = 0x0 > -> --------------------------- Newsgroups: comp.os.vxworks Subject: What happened to vxWorks archives on ftp.utd.ucar.edu ?? Date: 30 Jul 2001 10:04:46 -0700 From: zoat2020@yahoo.com (John Murphy) Organization: http://groups.google.com/ Message-ID: Have these archives been mirrored or put up elsewhere? I'm in need of the xmodem implementation posted on these archives and am hoping that these archives have been moved. Failing that, does anybody have a kermit/ Xmodem/ Zmodem port for vxWorks? I just need to send an occasional file down to my vxWorks target over the serial shell. Thanks, John Murphy zoat2020@yahoo.com --------------------------- Newsgroups: comp.os.vxworks Subject: prjConfig.c in BSP for Power PC Date: Mon, 30 Jul 2001 17:59:52 GMT From: jan.larcombe@betaweb.demon.co.uk (Jan Larcombe) Message-ID: <3b659edb.851409982@news.demon.co.uk> I'm having problems building my BSP, I end up with an unresolved external muxMaxBinds. I've checked why this is and it seems to be because the prjConfig.c which is generated is calling the routine usrMuxLibInit. In earlier versons of my BSP prjConfig.c was calling MuxLibInit instead. Obviously there is no point in ediiting prjConfig.c but does anyone know how to change configurations so that the prjConfig.c file which is automatically generated will call the correct function. Thanks in advance --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Developpment on Linux Date: Mon, 30 Jul 2001 13:55:31 -0500 From: Dinker Charak Organization: Fermi National Accelerator Laboratory Message-ID: <3B65ADA3.D85179A5@fnal.gov> References: <3B31EB91.56B0BAF9@gaci.fr> <3B61C3CD.9C39450E@qwest.net> > > I am wondering whether such tools are available on Linux > > (ie: Cross Compiler, remode debugger, remote shell ...) ? > > Tornado 3 is available under Linux. The downside is that T3 is tied > to VxWorks AE. Is it? I guess there is an option what kind of project you can build ... it can be a old VxWorks 5.4 type kernel too ... if I remember right what I saw in a demo in EST Embedded Conference ... dinker --------------------------- Newsgroups: comp.protocols.snmp,comp.os.psos,comp.os.vxworks Subject: Re: get_object_identifier_async() Date: Mon, 30 Jul 2001 15:37:01 -0500 From: "Jim Jones" Message-ID: References: <15eacd17.0107250043.343b208c@posting.google.com> >...clip... > OIDC_T snmp_product_id[] = { > 1, // ISO > 3, // ORG > 6, // DOD > 1, // INTERNET > 4, // PRIVATE > 1, // ENTERPRISES > 12, // Epilogue Technology > 7, // Products > 1, // Unix Agent > 3, // Version 3 > 0 // Subversion > }; > > > GetRequset : sysObjectID > > GetResponse is : > sysObjectID.0 (object identifier) enterprises.12.7.1.3.0 ...clip... > > scopusRoot OBJECT IDENTIFIER ::= { enterprises 4466 } > scopusProducts OBJECT IDENTIFIER ::= { scopusRoot 1 } > scopusEncoder OBJECT IDENTIFIER ::= { scopusProducts 5 } > scopusE1000 OBJECT IDENTIFIER ::= { scopusEncoder 1 } > > > OIDC_T snmp_product_id[] = { > 1, // ISO > 3, // ORG > 6, // DOD > 1, // INTERNET > 4, // PRIVATE > 1, // ENTERPRISES > 4466, // Scopus Root > 1, // Scopus Products > 5, // Scopus Encoder > 1, // Encoder E1000 > 0 // Subversion > }; > > GetRequset : sysObjectID > > GetResponse is : > sysObjectID.0 (object identifier) scopusE1000.0 > > In the first case we have relative OID starting from enterprises : > enterprises.12.7.1.3.0 > > In the second case we have relative the last node only : > scopusE1000.0 > > Any suggestions? Your question is not clear. The reason you got the OID starting from enterprises is because you did not supply your manager with the OID name information beyond "enterprises" (which was built-in to the manager). The manager simply replaced as much of the prefix of the complete OID as possible, which turned out to be "enterprises". I'm sure you have an epilogue.mib file somewhere in your Envoy distribution, where you will find that the above OID is enterprises.epilogue.etcProduct.etcUnixAgent.3.0. Thus, if you supplied that mib (or the OID info contained within it in some form) to your manager then you should see the sysObjectID.0 reported as etcUnixAgent.3.0. Presumably, that is version 3.0 of something called etcUnixAgent. However, that is not what you really want. You see this reported simply because you are using the sample code provided by Envoy for your agent. So if you want the agent to report the OID for scopusE1000.0, then you simply use the assignment of OID values as shown in your email (above) for snmp_product_id[] . Jim Jones Icon Labs > > Thanks, > ===================== > Alex Vinokur > mailto:alexvn@bigfoot.com > http://up.to/alexvn > http://go.to/alexv_math > ===================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Error while downloading __eabi Date: Tue, 31 Jul 2001 00:31:30 +0200 From: "JJS" Organization: LT Message-ID: <9k4n1g$hmh$1@news4.isdnet.net> References: <3B657868.71A55A33@fnal.gov> >I created a simple Hello.c with just > >main() >{ >} > >and compiled (right click in workspace window, then compile Hello.c) it >and downloaded (using download object file icon in workspace) it, all >using tornado. I get this error ... > >Errors while downloading >/usr/products/VxWorks/TornadoII/target/proj/Project0/default/Hello.o: >__eabi > Change main() to Mymain() Jack --------------------------- Newsgroups: comp.os.vxworks Subject: Where Is the FAQ and Archive of Code? Date: 30 Jul 2001 16:38:01 -0700 From: clusardi2k@aol.com (Christopher M. Lusardi) Organization: http://groups.google.com/ Message-ID: Hello, Where are the archives for the source code in this newsgroup? Thanks, Christopher Lusardi --------------------------- Newsgroups: comp.os.vxworks Subject: intLock() or intDisable() ? Date: 30 Jul 2001 18:01:15 -0700 From: xgdong@sina.com (Docular Xu) Organization: http://groups.google.com/ Message-ID: hi, I'm a beginner in VxWorks. When entering an ISR, I know the interrupt should be locked first. In VxWorks' docs, there are no directions told me how to do this. There are two functions which seem approporate to this task : intLock() and intDisable(). Which one should be used? Thanks, scincerely. --------------------------- Newsgroups: comp.os.vxworks Subject: excVecGet Date: Tue, 31 Jul 2001 14:28:06 +0900 From: "Ogre" Organization: A poorly-installed InterNetNews site Message-ID: <9k5f9g$9sm$1@news2.kornet.net> I have a MPC860. and At Booting Sequence, it exist a stage that get the default vector connected to the External Interrupt (0x500) look at the following... - ----------------------------------------------- c:\Tornado\target\src\drv\intrCtl\ppc860Intr.c - ----------------------------------------------- STATUS ppc860IntrInit ( VOIDFUNCPTR * cpmIntrVec /* Intr level of the CPM Intr ctrl */ ) { VOIDFUNCPTR defaultVec; /* INTR3 default vector */ UINT32 regBase; /* device register base address */ UINT32 cicrIntLevel; /* CICR interupt level */ UINT32 simaskIntLevel; int vector; /* Get the default vector connected to the External Interrupt (0x500) */ defaultVec = (VOIDFUNCPTR) excVecGet ((FUNCPTR *) _EXC_OFF_INTR); /* Connect the interrupt demultiplexer to External Interrupt (0x500) */ excIntConnect ((VOIDFUNCPTR *) _EXC_OFF_INTR, ppc860IntrDeMux); /* Install `system' intConnect routine */ if (_func_intConnectRtn == NULL) _func_intConnectRtn = ppc860IntConnect; if (_func_intEnableRtn == NULL) _func_intEnableRtn = ppc860IntEnable; if (_func_intDisableRtn == NULL) _func_intDisableRtn = ppc860IntDisable; .................... - ---------------------------------------------------- sometimes, when excVecGet() is executed, system crashed. what's problem? thanks in advanced.. --------------------------- Newsgroups: comp.os.vxworks Subject: Problem in scaning user input ! Date: 31 Jul 2001 00:33:06 -0700 From: mridulgupta1@yahoo.com (Mridul Gupta) Organization: http://groups.google.com/ Message-ID: hi all, i am a new user to VxWorks and i wrote a interactive program which takes inputs from user. the problem is that the scanf doesnt works ... it seems like the shell is eating the input in between. I tried suspending the shell task but it didnt work. i also tried changing priorities but nothing worked ... please tell me how do i take user input ?? Thanx Mridul --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem in scaning user input ! Date: Tue, 31 Jul 2001 10:14:36 +0200 From: "Jesper Quorning" Organization: TDC Internet Message-ID: <3b6669ae$0$73420$edfadb0f@dspool01.news.tele.dk> References: "Mridul Gupta" wrote in message news:fef31d0c.0107302333.1b62f612@posting.google.com... > hi all, > i am a new user to VxWorks and i wrote a interactive program > which takes inputs from user. the problem is that the scanf doesnt > works ... it seems like the shell is eating the input in between. I > tried suspending the shell task but it didnt work. i also tried > changing priorities but nothing worked ... please tell me how do i > take user input ?? I think it is the host shell which does the trick. When you hit enter the string is sendt to the target. When I bypass the host shell by using rlogin is works as expected. jesper --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem in scaning user input ! Date: Tue, 31 Jul 2001 17:32:51 +0900 From: "Ogre" Organization: A poorly-installed InterNetNews site Message-ID: <9k5q3t$8g$1@news2.kornet.net> References: If you don't need "Shell", subtract SHELL in your configuration. And don't use scanf. in usrConfig.c, vxWorks use fioRdString. "Mridul Gupta" wrote in message news:fef31d0c.0107302333.1b62f612@posting.google.com... > hi all, > i am a new user to VxWorks and i wrote a interactive program > which takes inputs from user. the problem is that the scanf doesnt > works ... it seems like the shell is eating the input in between. I > tried suspending the shell task but it didnt work. i also tried > changing priorities but nothing worked ... please tell me how do i > take user input ?? > > Thanx > > Mridul --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Error while downloading __eabi Date: Tue, 31 Jul 2001 11:58:02 +0200 From: "Werner Schiendl" Message-ID: <3b6681d3$0$9436$6e365a64@newsreader02.highway.telekom.at> References: <3B657868.71A55A33@fnal.gov> <9k4n1g$hmh$1@news4.isdnet.net> > Change main() to Mymain() ...and be sure to invoke the function yourself, e. g. by typing Mymain in the shell. Or adding the call to usrAppInit.c file of your VxWorks Image (later). VxWorks has no concept like executables in desktop OS. hth Werner --------------------------- End of New-News digest **********************