From vxwexplo-errs@csg.lbl.gov Mon Jan 3 12:50:01 2000 From: Steve Bauer Date: Mon Jan 3 12:50:05 PST 2000 Subject: File transfer speed Using FTP to a VxWorks server, we're getting only about 0.4 megabytes of data per second on file transfers over a 100BaseTX network (with no other traffic), with either a PC or an Alphastation as the client. Has anyone done significantly better than this? With what hardware and software? Will a custom sockets-based program do much better? Steve Bauer bauer@kscorp.com From vxwexplo-errs@csg.lbl.gov Mon Jan 3 13:15:36 2000 From: "Robert A. Lerche" Date: Mon Jan 3 13:15:39 PST 2000 Subject: Re: File transfer speed I have not tested the VxWorks FTP server myself; I used the "blaster/blastee" demo programs for my performance testing. However, I would expect the VxWorks FTP server to give close to full TCP performance. (I think someone else at my client _did_ run an FTP test; I'll check.) We got over 10 Mbytes/second on 100baseTX using a PPC603 (Artesyn's Baja750 board). Are you testing on an isolated LAN? That can make a big difference. Also, try running the browser and see what the CPU usage on the target is. From vxwexplo-errs@csg.lbl.gov Tue Jan 4 04:00:30 2000 From: daemon@csg.lbl.gov Date: Tue Jan 4 04:00:34 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jan 4 04:00:27 PST 2000 Subject: Re: Debugging under VxSim Lite Subject: Re: embedded web browser Subject: DHCP.. Subject: Re: Keyboard on a target PC Subject: Re: embedded web browser Subject: BSP Subcontractor (Israel) Subject: Re: ATM PMC Adapter Device Driver for BSP MV2700 Subject: Re: Cost Effective Training? Subject: Re: GUI for VxWorks (Zinc or Qt) Subject: Re: __udiv3 etc unresolved ? Subject: 'SNMP daemon' and 'DHCPv4 runtime client' related? Subject: taskInit and taskActivate vs taskSpawn Subject: Re: taskInit and taskActivate vs taskSpawn Subject: Embedded developers using VxWorks, Linux, NT or CE wanted Subject: Re: File transfer speed ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Debugging under VxSim Lite Date: Tue, 28 Dec 1999 20:30:30 GMT From: ajay Organization: EarthLink Network, Inc. Message-ID: <38691DFC.88EC908B@wellknit.com> References: <946316070.16088.0.nnrp-13.c1ed9bc0@news.demon.co.uk> Reply-To: ajay@wellknit.com One technique that works is waiting on some variable like int someVar = 0; while (someVar != 1) taskDelay(WAIT_100MS) Then you could attach to the the task from Debug menu option and set the someVar to one. Also, you might want to use global toggle. This works although its tiring;-). I hope this might be useful... - -Ajay Jim Moores wrote: > I've been developping a small kernel that sits atop VxWorks, and I'm having > huge problems with the debugging facilities supplied by Tornado/VxSim Lite. > Its the first time I've developped under VxWorks, and I'm finding I'm having > to go back to printf's to find bugs. It seems that WRS don't bother fixing > bugs in VxSim or somthing. Nearly all of my problems have eventually been > diagnosed as bugs in VxSim - like signals don't work at all if you use more > than one task! No timers - I had to write a timer simulator using a timeout > semaphore! > > What I really need to know is how to use the debugger properly - I'm used to > using gdb under linux. There doesn't seem to be any way to restart the > program, or to set breakpoints before the program is run. The `run' option > asks for a function name - calling something like `sysInit' doesn't work > (can't find the symbol) - neither does calling usrInit or my main routine > (not called main() obviously). > > I've tried building my application as both a downloadable app or as a > bootable project. > > Can anyone help? I just need breakpoints and watch points... Give me > strength. > > Jim Moores. > Quickstone Systems. jimmoores@hotmail.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: embedded web browser Date: Tue, 28 Dec 1999 21:28:14 GMT From: "George W. Freas II" Organization: The Boeing Company Message-ID: <38692B6E.7040ED31@hsv.boeing.com> References: <385714A6.E835E9EC@lucent.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Not sure if you are aware of this, but the source to Lynx is available. You can find info on this at: http://lynx.browser.org/ This is advertised to work for U*NX and VMS, etc, and would take some work to port from U*NX to VxWorks, I am sure. You might have luck posting to email lists, etc relating to lynx. George Freas. - ------------------------------------------------------------ Yusuf Motiwala wrote: > > Hi. > > Is there any small embedded WEB browser (text based like lynx) available > for > vxWorks? Only requirement is, it should be able to display simple HTML > pages > and send CGI queries. It is not necessary to have standard browser > features like > ftp, frame, javascript, bookmarks, mailing capability etc. > > I will appreciate if somebody can point me to useful resource on the > same. > > Regards, > Yusuf --------------------------- Newsgroups: comp.os.vxworks Subject: DHCP.. Date: Wed, 29 Dec 1999 05:36:56 -0500 From: VijayKumar P Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: 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_000_01BF5251.D64E259E Content-Type: text/plain Hi All, I am trying to boot up a PC based target with a fixed IP address and then initialize as a DHCP-client Now: 1. Is it advisable to use a Windows-NT server as DHCP-server?. 2. How can I make the Lease renewal automatic, or can I make the lease duration very big?. Thanks in advance, Vijay - ------ =_NextPart_000_01BF5251.D64E259E Content-Type: text/x-vcard; name="vpeshkar.vcf" Content-Disposition: attachment; filename="vpeshkar.vcf" Content-Description: Card for VijayKumar P begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India fn:Vijay Kumar.P end:vcard - ------ =_NextPart_000_01BF5251.D64E259E-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Keyboard on a target PC Date: Fri, 24 Dec 1999 16:13:24 GMT From: snaphook@NOPOTTEDMEAT.swbell.net (Bruce) Organization: SBC Internet Services Message-ID: <38639b81.3352905@news.swbell.net> References: <1UM84.905$Yr1.5460915@bunson.tor.sfl.net> In comp.os.vxworks "Ilya Klebanov" wrote: >does anybody know, how to make keyboard working on a target PC. Anything >special should be done to initialize stdin device ( which I assume is >keyboard)? Do I have to define something special, when I build OS image? >Right now, If I try either > getchar(); > or > read( 0, &input, 2); >the functions do not come back at all. Look at the functions, globalStdIOSet() and ...Get() Bruce in Katy Texas --------------------------- Newsgroups: comp.os.vxworks Subject: Re: embedded web browser Date: Tue, 28 Dec 1999 21:23:00 GMT From: "George W. Freas II" Organization: The Boeing Company Message-ID: <38692A34.C0C757C@hsv.boeing.com> References: <385714A6.E835E9EC@lucent.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Not sure if you are aware of this, but the source to Lynx is available. You can find info on this at: http://lynx.browser.org/ This is advertised to work for U*NX and VMS, etc, and would take some work to port from U*NX to VxWorks, I am sure. You might have luck posting to email lists, etc relating to lynx. George Freas. - ------------------------------------------------------------ Yusuf Motiwala wrote: > > Hi. > > Is there any small embedded WEB browser (text based like lynx) available > for > vxWorks? Only requirement is, it should be able to display simple HTML > pages > and send CGI queries. It is not necessary to have standard browser > features like > ftp, frame, javascript, bookmarks, mailing capability etc. > > I will appreciate if somebody can point me to useful resource on the > same. > > Regards, > Yusuf --------------------------- Newsgroups: comp.os.vxworks Subject: BSP Subcontractor (Israel) Date: Thu, 30 Dec 1999 14:47:28 +0200 From: "Yael Abudi" Message-ID: <386b53c6.0@news.barak.net.il> Hello all, I'm looking for a subcontractor to develop a BSP in Israel. The CPU is MPC860T. Anyone who is interested (or has suggestions), please reply. Yael Abudi Shiron Satellite Communications yaela@shiron.com Tel: +972-3-9210305 ext. 212 Fax: +972-3-9214547 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ATM PMC Adapter Device Driver for BSP MV2700 Date: Mon, 20 Dec 1999 13:38:59 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <385E85D3.B05AF87B@aps.anl.gov> References: <38596A38.F422F858@spawar.navy.mil> Ron Webb wrote: > > We looking for ATM PMC Adapter Device Driver for BSP MV2700. We have > two cards from two manufacturers, Interphase 4575 and Cetia. However, > the softwares that go with both card were written (or configured) for > BSP MV2600, not MV2700 as we have. ... > 3. Can the ATM PMC driver which is built for MV2600 work with the card > and the VxWorks OS built for MV2700 ? I can't speak specifically about the cards you are asking about, but we have successfully used some video software that was only available for the MV2604 on an MV2700 with no problems. IIRC the gcc compiler flags are identical and in fact the same BSP is used for both with switches between the two models, so I suspect you'll find both those ATM cards will both work. However you should get one on trial and prove it to your (or your companies) satisfaction before paying for it... - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cost Effective Training? Date: Thu, 30 Dec 1999 16:53:00 GMT From: "George W. Freas II" Organization: Boeing Message-ID: <386B8DEC.5B9C1BA5@hsv.boeing.com> References: <38566f37.90786350@news.eznet.net> <8OY54.5232$ZT5.2767135@news1.rdc1.on.wave.home.com> <385f39f2.1452899@news.fe.internet.bosch.de> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Try out Embry Riddle Univ.: http://www.embryriddle.edu I know they have a course that covers VxWorks, but not sure if it is offered as a distance education class (they are in Daytona, FL). Another place has a class: http://www.menet.umn.edu This is Univ. of Minnesota, the course number is ME5271. Not sure of the status of these courses, and if they are still offered. Hope this helps. George Freas. - -------------------------------------------------------------- Jan Koeppen wrote: > > On Thu, 16 Dec 1999 03:14:12 GMT, "Peter on Rogers" > wrote: > > >You might want to contact your Wind River local representative. Many of the > >regions offer local on-site VxWorks Quickstart programs. While not > >formalized training, the program might meet your needs. > > > Hi Peter, > what would these Quickstart programs comprise? > We have some workers (good and experienced in every aspect of embedded > software but Tornado) in need of some quick getting to work and the > Tornado Trainig offered by WRS seems rather expensive. And I doubt > that our WRS representative would offer anything we do not explicidly > ask for :-( > > regards > > Jan Koeppen > -- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GUI for VxWorks (Zinc or Qt) Date: Sun, 02 Jan 2000 19:52:59 GMT From: Ken Hayber Organization: - Message-ID: <946842468.120587028@news-server.socal.rr.com> References: <37FDB080.9DEB0FFB@rz.tu-ilmenau.de> Followup-To: comp.os.vxworks On Fri, 08 Oct 1999, aleXXX wrote: >Hi, > >has anybody experiences with working with Zinc or with Qt under VxWorks >? >At the moment we have to decide which one to use. >Would be nice if anybody could give me some hints. > >Thanks in advance >Alex My company has begun development of a GUI application on vxWorks/Zinc on a PC system. The nice parts are being able to develop on Win32 and then just compile for vxWorks with very little visual diference; and it looks/feels like Windows (good for customer). We have had very few problems so far; but it's still a bit early in development. We did not look at Qt; as it was not available on vxWorks at that time. (I assume from your question that it is). My understanding of Qt makes me think that you might get the same benefits. Good Luck! K. Hayber --------------------------- Newsgroups: comp.os.vxworks Subject: Re: __udiv3 etc unresolved ? Date: Sun, 2 Jan 2000 20:17:56 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <84pdvi$cd6$1@newsmaster01.magnet.at> References: <83od8n$e5l$1@newsmaster01.magnet.at> <385FBDDA.390FB4DF@lucent.com> <83opip$rjf$1@newsmaster01.magnet.at> <3860D74A.BAB5B04F@lucent.com> our GNU expert looked at the problem and solved it. the problem is, that these functions are part of the C-Std lib, but this one is not linked into VxWorks and into the user App. he extracted the function from the compiler source, added it as user function and now the unresolved external is gone. While analyzing this problem, he run into some other problems. maybe some one can help here. =========== I had a look at the makefiles generated by Tornado and those used in our environment to build the OPLC Layer and I realized, that -nostdlib is specified as a compile option. Of course, this causes libgcc.a not bo be linked with the applications and I wonder why. Are there any problems linking libgcc.a with your own objects (or simply building an object just containing libgcc.a and loading this one with ld <)?. I tried to link with libgcc.a and at least the resulting object could be loaded on VxWorks. I have not yet tested if it works. Also I am a bit curious why the Tornado generated makefiles pass some options to the compiler or linker. Maybe you can enlighten me a bit about this: 1. -nostdlib. See above. Are there any problems linking with libgcc.a. I thought of using something like -print-file-name=libgcc.a in addition to -nostdlib so just libgcc.a is linked in. Of course, this would result in multiple definitions of the same symbol (by name) when loading more than one object built with this option. 2. -fno-builtin. Is there any reason why the GCC built-in functions (which I assume to be normally faster than the library functions) should not be used. According to the docs, this affects the use/linking of abort, abs, alloca, cos, exit, fabs, ffs, labs, memcmp, memcpy, sin, sqrt, strcmp, strcpy and strlen. 3. -fno-defer-pop is just another issue to optimization, but I think just a minor one. But I don't understand why this option is specified and I didn't find any hint in the Tornado documentation. 4. -fvolatile. Ouch, this really hurts. If it isn't required (and I found some comment like "optional" in the Tornado docs), I think it should not be used (just for performance reasons). Am I right in this case? 5. -r is specified in a "pre-linking" stage, and I could not find this option in any GCC documentation (nor when I quickly checked the egcs 2.90 source code, too). Do you know, what this is for? - -- regards, Mario Semo. http://www.kirchnersoft.com Douglas Fraser wrote in message <3860D74A.BAB5B04F@lucent.com>... > > >Mario Semo wrote: >> >> Doug, >> >> thx for your help. i have now added such a file + function to my VxWorks >> simulation kernel > >I haven't tried this on the vxWorks simulator, just on >a real ARM (SA110) target. > > > >I did something similar to that. It is in a C file that is >added to the kernel using the MACH_EXTRA definition in the >Makefile, for example... > >MACH_EXTRA = llUnused.o > >The build for vxWorks completes fine as before. >Then when I load my application image using 'ld' >everything is okay. > >> >> and then rebuild all. >> >> compile result: >> >> partialImage.o: In function `_tff': >> \wpwr\host\src\gnu.simpc.cpp\gcc\cp\tinfo2.cc(.text+0x1ed0): undefined >> reference to `__div >> di3' >> partialImage.o: In function `_tfUi': >> \wpwr\host\src\gnu.simpc.cpp\gcc\cp\tinfo2.cc(.text+0x1f70): undefined >> reference to `__udi >> vdi3' > >It is odd that it is failing in the compiler stage. >Again, this may be a simulator issue. > >Anybody else out there had this issue with the simulator??? > >Doug > >> make: *** [vxWorks] Error 0x1 --------------------------- Newsgroups: comp.os.vxworks Subject: 'SNMP daemon' and 'DHCPv4 runtime client' related? Date: Mon, 03 Jan 2000 14:35:44 +0530 From: VijayKumar P Organization: Cisco Systems Message-ID: <38706667.4FC3AA47@cisco.com> This is a multi-part message in MIME format. - --------------494C1A4EA1060A4EF5047CB7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, I am unable to make out how 'DHCPv4 runtime client' component is dependent on 'SNMP daemon '. Any Idea?. I am using ver 5.4 and Tornado 2.0 The 'SNMP daemon' component gets included when ever I include 'DHCPv4 runtime client'. And when I try to build image, I get ....\prj\prjConfig.c:76: snmp\snmpdInit.h: No such file or directory make: *** [prjConfig.o] Error 0x1. I believe this is because the SNMP being an optional component and not available with me is giving the trouble. But WHY should it include 'SNMP daemon' in the first place????. Thanks, Vijay - --------------494C1A4EA1060A4EF5047CB7 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for VijayKumar P Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India fn:Vijay Kumar.P end:vcard - --------------494C1A4EA1060A4EF5047CB7-- --------------------------- Newsgroups: comp.os.vxworks Subject: taskInit and taskActivate vs taskSpawn Date: Mon, 03 Jan 2000 13:38:48 -0500 From: steven.offenbacher@jhuapl.edu Organization: Johns Hopkins University Applied Physics Lab, Laurel, MD, USA Message-ID: <3870ECB8.2AFB7A18@jhuapl.edu> I'm attempting to replace a taskSpawn with the taskInit/taskActivate function call and have not been successful. I'm looking for a C code example of working taskInit/taskActivate functions. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskInit and taskActivate vs taskSpawn Date: Mon, 03 Jan 2000 21:53:30 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <84r5oj$rrs$1@nnrp1.deja.com> References: <3870ECB8.2AFB7A18@jhuapl.edu> Steven, In article <3870ECB8.2AFB7A18@jhuapl.edu>, steven.offenbacher@jhuapl.edu wrote: > I'm attempting to replace a taskSpawn with the taskInit/taskActivate > function call and have not been successful. I'm looking for a C code > example of working taskInit/taskActivate functions. Is there a good reason you are doing this? It does complicate things somewhat since you are responsible for all the allocations of memory (stack and TCB)... The most common reason for doing this is to create dormant tasks at initialisation time which can then be woken up later on. If that is what you are trying to do there are simpler ways: you could get the tasks to start and block on a semaphore, or even suspend itself immediately. Hope that helps a little, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Embedded developers using VxWorks, Linux, NT or CE wanted Date: Mon, 27 Dec 1999 08:52:36 -0800 From: "Nick Berliner" Message-ID: <4LM94.3$cT.86@news.ais.net> GoAhead Software is offering a chance to beta test remote upgrading capability in your networked appliance or device. GoAhead FieldUpgraderT provides manufacturers of networked devices with a comprehensive solution for automatically upgrading their products in the field. GoAhead FieldUpgrader allows you to upgrade software, firmware, O/Ss or data using a device-initiated approach. These upgrades can provide products with new features, reliability enhancements, performance improvements, updated applications, security patches and bug fixes. GoAhead FieldUpgrader removes the end-customer burden of manually upgrading their product, and reduces the manufacturer's expenditures for costly support visits and product recalls. If you are developing in an NT or Linux environment and your device runs Embedded NT, Embedded Linux, VxWorks, CE, NT, or Linux, you may qualify for participation. If you are interested in trying GoAhead FieldUpgrader out in your device, visit: http://www.goahead.com/fieldupgrader/fubeta.htm or email beta@goahead.com. GoAhead FieldUpgrader beta program will run through February 2000. Please sign up right away if you are interested in participating!! Best regards, Nick Berliner ____________________________________________________________________________ ____ Nick Berliner Internet Marketing Coordinator/GoAhead Software nickb@goahead.com www.goahead.com phone 425-468-5404 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: File transfer speed Date: Tue, 04 Jan 2000 08:51:05 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38719859.EDB8C89B@bitband.com> References: <84r3fn$g5m2@overload.lbl.gov> Reply-To: leonid@bitband.com Steve, you must decide what is it that you need: FTP file transfer or application specific socket-based data trqansfer, and in the later case TCP and UDP greatly vary. If you are interested in FTP and data transfer, it is known for a some time that the FTP server which is shipped with Tornado is very very slow, and has other downsides to it too. The DOS file system is not perfect either. Upgrading to DosFs 2.0 will give you an improved file system, and an enhanced (much faster too) FTP server albeit in the "unsupported" directory. Perhaps I could be more helpful if the questions where more specific. Leonid Steve Bauer wrote: > > Using FTP to a VxWorks server, we're getting only about 0.4 megabytes of > data per second on file transfers over a 100BaseTX network (with no other > traffic), with either a PC or an Alphastation as the client. Has anyone > done significantly better than this? With what hardware and software? > Will a custom sockets-based program do much better? > > Steve Bauer > bauer@kscorp.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jan 4 07:26:55 2000 From: VEASIE SHARON Date: Tue Jan 4 07:27:00 PST 2000 Subject: Graphics Driver I am using Tornado2 on a Windows platform and the target is a PentiumII. Does anyone know if a graphics driver is available for 1024x786? Thanks, Sharon Veasie Software Engineer Trex Medical, XRE Division 300 Foster Street Littleton, MA 01460 978-486-9681 ext. 8283 sveasie@xre.com From vxwexplo-errs@csg.lbl.gov Tue Jan 4 08:14:53 2000 From: "Darren Carlson" Date: Tue Jan 4 08:14:56 PST 2000 Subject: DOS-FS 2 Hello all, I'm using Tornado 2 with the DOSFS 2 patch. Since installing the patch I am unable to load our software modules. I'm getting an error saying that the system time is earlier than a predefined date. I have set the date to today using DOS's date function. The error is still there. I have also set the date to 1999, the error is still there. Is there something I can undefine or do, that will stop this Y2K test from being performed? We are using a Pentium BSP and our system has no problem going into and out of the year 2000. Thank You Darren Carlson Software Development Engineer Applied Biometrics Inc 612-890-1123 From vxwexplo-errs@csg.lbl.gov Wed Jan 5 07:08:40 2000 From: "Dipl.-Phys. S. Richter, IKP, Tel. 16-5152" Date: Wed Jan 5 07:08:44 PST 2000 Subject: Ethernetinterface - tuFilterSetup Hi! I need some clarification concerning the use of the tuFilterSetup routine. That's the routine that handels the address-filtering of the DECchip 21040. We're using an AXPvme233 (Alpha-Processor) and VxWorks 5.31. So far as I understand this routine it allows to reset the filter for the tu-interface. What I want to do now is to change the filter to accept the original MAC- address and in addition to that another MAC-address and one multicast- address. But when I try to use this function I encounter some severe problems. If I follow the example from the documentation by typing in those lines from the console -> q = malloc (1600) -> bzero (q, 1600) -> q[0]=0x0403020100000001 -> q[8]=0x00000605 -> tuFilterSetup(0,0x1|0x20,q) I receive an OK that the filter change was successful. If I try the same inside a program or logged onto the system via rlogin I lose the contact to my system. Even if I set the filter-address to original MAC-address I see the same behavior. And that's something I really won't expect. Any hints how to use this function correctly would be very much appreciated! Regards Simone Dipl.-Phys. Simone Richter simone@ikp.tu-darmstadt.de Institut f. Kernphysik Schlossgartenstr. 9 TU Darmstadt 64289 Darmstadt Tel: +49 (0) 6151 / 16-5152, 16-3323 Fax: +49 (0) 6151 / 16-4321 From vxwexplo-errs@csg.lbl.gov Wed Jan 5 11:51:48 2000 From: mdelgior@TIGER.rst.com (Mark Del Giorno) Date: Wed Jan 5 11:51:52 PST 2000 Subject: Problem elf'ing vxWorks_rom for PPC Heyyyyyyyyyyyyyy, I'm trying to burn my application into the mv2700 flash so it will run on power up. I added the patch to allow > 1 MB flash applications to run (untested). I've added my application to MACH_EXTRA. I've edited usrAppInit.c to call it. I've confirmed that if I make vxWorks and have my board boot, my application runs. I changed my project to make vxWorks_rom. but: When I run elfToBin on it, I get the following output: > elfToBin < vxWorks_rom > test.bin seg 1: Expected load address 0xfff00100 but file address is 0x00586a88 test.bin is still generated (and looks to be about the right size). Has anyone seen this? Is this an error or just a warning that can be ignored? (I know the code will be copied to RAM, then run....???) Thanks in advance for any help! - Mark _/_/_/_/ _/_/_/_/ _/_/_/_/_/ Mark Del Giorno (del@rst.com) _/ _/ _/ _/ General Dynamics Robotic Systems _/_/_/_/ _/_/_/_/ _/ 1234 Tech Court _/ _/ _/ _/ Westminster, MD 21157 _/ _/ _/_/_/_/ _/ (410) 876-9200 Fax: 876-9470 http://www.rst.com From vxwexplo-errs@csg.lbl.gov Wed Jan 5 12:05:21 2000 From: Alan Biocca Date: Wed Jan 5 12:05:25 PST 2000 Subject: Exploder Problems The VxWorks User Group Exploder has had some difficulty, not due to Y2K, but due to unix support security improvements to reduce a particular type of relay that was occurring. This new sendmail didn't work quite right with the exploder, and it was not properly tested. Sorry for the difficulties. Thanks, Alan K Biocca From vxwexplo-errs@csg.lbl.gov Thu Jan 6 08:47:08 2000 From: "Marty Christophe" Date: Thu Jan 6 08:47:12 PST 2000 This is a multi-part message in MIME format. ------=_NextPart_000_0004_01BF586E.33472290 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I would like to be on the mailing list of vsWorks. It seems that if I send you this mail it will be OK USENET comp.os.vxworks Sorry If it is not good. Best Regards ... end ------=_NextPart_000_0004_01BF586E.33472290 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I would like to be on the mailing list of=20 vsWorks.
It seems that if I send you this mail it will be=20 OK
 
USENET comp.os.vxworks
 
Sorry If it is not good.
 
 
Best Regards ...
 
end
------=_NextPart_000_0004_01BF586E.33472290-- From vxwexplo-errs@csg.lbl.gov Thu Jan 6 09:55:22 2000 From: Christopher Manders Date: Thu Jan 6 09:55:25 PST 2000 Subject: [Fwd: VXWorks] Christopher Manders wrote: > VXworks vx From vxwexplo-errs@csg.lbl.gov Thu Jan 6 14:47:30 2000 From: Christopher Manders Date: Thu Jan 6 14:47:34 PST 2000 Subject: vxworks vxworks From vxwexplo-errs@csg.lbl.gov Thu Jan 6 14:54:04 2000 From: Christopher Manders Date: Thu Jan 6 14:54:08 PST 2000 Subject: vxworks vxworks From vxwexplo-errs@csg.lbl.gov Fri Jan 7 07:06:32 2000 From: Christopher Manders Date: Fri Jan 7 07:06:36 PST 2000 Subject: [Fwd: vxworks] Christopher Manders wrote: > vxworks From vxwexplo-errs@csg.lbl.gov Fri Jan 7 10:08:32 2000 From: Joe Schlesselman Date: Fri Jan 7 10:08:36 PST 2000 Subject: Re: [Fwd: Re: Error from memPartFree] Hi all, > In article <84ufdo$39e$1@nnrp1.deja.com>, john_94501@my- > deja.com wrote: > > > > No simple tools, but I can tell you a little about the error which might > > help you track it down. This error is usually the result of corrupting > > the block tracking information that is held in the 8 bytes just before > > the pointer that is returned by malloc(). > > I believe MemScope will do what he wants. Although I don't > think it has been released for Tornado 1.0.1. Slight correction: MemScope will do what you want, and it is in fact available for both Tornado 1.0.1 and T2. Joe joe@rti.com From vxwexplo-errs@csg.lbl.gov Fri Jan 7 15:13:40 2000 From: Alan Biocca Date: Fri Jan 7 15:13:43 PST 2000 Subject: VxUG Exploder Problems The VxWorks User's Group Exploder has had some (non-Y2K) problems since 12/20/99 when the sendmail was upgraded to protect against spam relaying. Apparently the new sendmail considers our list to be spam. In any case, it has been rearranged somewhat to comply with the extensive constraints that are now in place. Hopefully we are back now. During most of the outage the articles were processed normally to the newsgroup and archives. For the last few days those processes were shut down in order to test the changes with a small recipient list. This mailing is the first with a full recipient list - we'll soon see if there are remaining problems. There may be some other consequences of these changes. Welcome and Thanks also go to Chris Manders, CJManders@lbl.gov, our new unix sysadmin and sendmail expert. Address problem reports to vxwexplo-request@lbl.gov or akbiocca@lbl.gov -- Alan Biocca From vxwexplo-errs@csg.lbl.gov Fri Jan 7 17:11:11 2000 From: Alan Biocca Date: Fri Jan 7 17:11:14 PST 2000 Subject: VxUG Exploder Problems The VxWorks User's Group Exploder has had some (non-Y2K) problems since 12/20/99 when the sendmail was upgraded to protect against spam relaying. Apparently the new sendmail considers our list to be spam. In any case, it has been rearranged somewhat to comply with the extensive constraints that are now in place. Hopefully we are back now. During most of the outage the articles were processed normally to the newsgroup and archives. For the last few days those processes were shut down in order to test the changes with a small recipient list. This mailing is the first with a full recipient list - we'll soon see if there are remaining problems. There may be some other consequences of these changes. Welcome and Thanks also go to Chris Manders, CJManders@lbl.gov, our new unix sysadmin and sendmail expert. Address problem reports to vxwexplo-request@lbl.gov or akbiocca@lbl.gov -- Alan Biocca From vxwexplo-errs@csg.lbl.gov Sat Jan 8 09:20:22 2000 From: Vxworks Exploder Date: Sat Jan 8 09:20:24 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jan 8 04:03:04 PST 2000 Subject: CTL_NET in TII ? Subject: Test Subject: Time() representation in VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: CTL_NET in TII ? Date: Fri, 07 Jan 2000 13:57:25 +0100 From: Antoine Lacroix Organization: ericsson Message-ID: <3875E2B5.C67AD169@esf.ericsson.se> Hi everybody, I am currently trying to use the ip_sysctl function at the network level but I just can't find the required definition of the CTL_NET macro (it appears in a comment of socket.h but is not defined) which is the first argument. Does anybody know its location or at least the appropriate value that it should take ? Thanks in advance, Antoine Lacroix. --------------------------- Newsgroups: comp.os.vxworks Subject: Test Date: Fri, 7 Jan 2000 20:30:45 +0800 From: bula@guoguang.com.cn (Zhanh Zhiqun) Organization: ¹ú¹âµç×Ó×ܹ«Ë¾ Message-ID: Test --------------------------- Newsgroups: comp.os.vxworks Subject: Time() representation in VxWorks Date: Fri, 07 Jan 2000 10:02:58 -0500 From: John Sestak Organization: Raytheon, Falls Church Message-ID: <38760022.53AB93CB@fallschurch.esys.com> When you set the current time in VxWorks, what starting date and time is it supposed to be relative to? Is it the number of days since year 0, year 1900 AD, 1970 AD, etc.? Thanks, John Sestak --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jan 8 17:45:14 2000 From: Operator Date: Sat Jan 8 17:45:16 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jan 4 04:00:27 PST 2000 Subject: Re: Debugging under VxSim Lite Subject: Re: embedded web browser Subject: DHCP.. Subject: Re: Keyboard on a target PC Subject: Re: embedded web browser Subject: BSP Subcontractor (Israel) Subject: Re: ATM PMC Adapter Device Driver for BSP MV2700 Subject: Re: Cost Effective Training? Subject: Re: GUI for VxWorks (Zinc or Qt) Subject: Re: __udiv3 etc unresolved ? Subject: 'SNMP daemon' and 'DHCPv4 runtime client' related? Subject: taskInit and taskActivate vs taskSpawn Subject: Re: taskInit and taskActivate vs taskSpawn Subject: Embedded developers using VxWorks, Linux, NT or CE wanted Subject: Re: File transfer speed ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Debugging under VxSim Lite Date: Tue, 28 Dec 1999 20:30:30 GMT From: ajay Organization: EarthLink Network, Inc. Message-ID: <38691DFC.88EC908B@wellknit.com> References: <946316070.16088.0.nnrp-13.c1ed9bc0@news.demon.co.uk> Reply-To: ajay@wellknit.com One technique that works is waiting on some variable like int someVar = 0; while (someVar != 1) taskDelay(WAIT_100MS) Then you could attach to the the task from Debug menu option and set the someVar to one. Also, you might want to use global toggle. This works although its tiring;-). I hope this might be useful... - -Ajay Jim Moores wrote: > I've been developping a small kernel that sits atop VxWorks, and I'm having > huge problems with the debugging facilities supplied by Tornado/VxSim Lite. > Its the first time I've developped under VxWorks, and I'm finding I'm having > to go back to printf's to find bugs. It seems that WRS don't bother fixing > bugs in VxSim or somthing. Nearly all of my problems have eventually been > diagnosed as bugs in VxSim - like signals don't work at all if you use more > than one task! No timers - I had to write a timer simulator using a timeout > semaphore! > > What I really need to know is how to use the debugger properly - I'm used to > using gdb under linux. There doesn't seem to be any way to restart the > program, or to set breakpoints before the program is run. The `run' option > asks for a function name - calling something like `sysInit' doesn't work > (can't find the symbol) - neither does calling usrInit or my main routine > (not called main() obviously). > > I've tried building my application as both a downloadable app or as a > bootable project. > > Can anyone help? I just need breakpoints and watch points... Give me > strength. > > Jim Moores. > Quickstone Systems. jimmoores@hotmail.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: embedded web browser Date: Tue, 28 Dec 1999 21:28:14 GMT From: "George W. Freas II" Organization: The Boeing Company Message-ID: <38692B6E.7040ED31@hsv.boeing.com> References: <385714A6.E835E9EC@lucent.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Not sure if you are aware of this, but the source to Lynx is available. You can find info on this at: http://lynx.browser.org/ This is advertised to work for U*NX and VMS, etc, and would take some work to port from U*NX to VxWorks, I am sure. You might have luck posting to email lists, etc relating to lynx. George Freas. - ------------------------------------------------------------ Yusuf Motiwala wrote: > > Hi. > > Is there any small embedded WEB browser (text based like lynx) available > for > vxWorks? Only requirement is, it should be able to display simple HTML > pages > and send CGI queries. It is not necessary to have standard browser > features like > ftp, frame, javascript, bookmarks, mailing capability etc. > > I will appreciate if somebody can point me to useful resource on the > same. > > Regards, > Yusuf --------------------------- Newsgroups: comp.os.vxworks Subject: DHCP.. Date: Wed, 29 Dec 1999 05:36:56 -0500 From: VijayKumar P Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: 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_000_01BF5251.D64E259E Content-Type: text/plain Hi All, I am trying to boot up a PC based target with a fixed IP address and then initialize as a DHCP-client Now: 1. Is it advisable to use a Windows-NT server as DHCP-server?. 2. How can I make the Lease renewal automatic, or can I make the lease duration very big?. Thanks in advance, Vijay - ------ =_NextPart_000_01BF5251.D64E259E Content-Type: text/x-vcard; name="vpeshkar.vcf" Content-Disposition: attachment; filename="vpeshkar.vcf" Content-Description: Card for VijayKumar P begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India fn:Vijay Kumar.P end:vcard - ------ =_NextPart_000_01BF5251.D64E259E-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Keyboard on a target PC Date: Fri, 24 Dec 1999 16:13:24 GMT From: snaphook@NOPOTTEDMEAT.swbell.net (Bruce) Organization: SBC Internet Services Message-ID: <38639b81.3352905@news.swbell.net> References: <1UM84.905$Yr1.5460915@bunson.tor.sfl.net> In comp.os.vxworks "Ilya Klebanov" wrote: >does anybody know, how to make keyboard working on a target PC. Anything >special should be done to initialize stdin device ( which I assume is >keyboard)? Do I have to define something special, when I build OS image? >Right now, If I try either > getchar(); > or > read( 0, &input, 2); >the functions do not come back at all. Look at the functions, globalStdIOSet() and ...Get() Bruce in Katy Texas --------------------------- Newsgroups: comp.os.vxworks Subject: Re: embedded web browser Date: Tue, 28 Dec 1999 21:23:00 GMT From: "George W. Freas II" Organization: The Boeing Company Message-ID: <38692A34.C0C757C@hsv.boeing.com> References: <385714A6.E835E9EC@lucent.com> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Not sure if you are aware of this, but the source to Lynx is available. You can find info on this at: http://lynx.browser.org/ This is advertised to work for U*NX and VMS, etc, and would take some work to port from U*NX to VxWorks, I am sure. You might have luck posting to email lists, etc relating to lynx. George Freas. - ------------------------------------------------------------ Yusuf Motiwala wrote: > > Hi. > > Is there any small embedded WEB browser (text based like lynx) available > for > vxWorks? Only requirement is, it should be able to display simple HTML > pages > and send CGI queries. It is not necessary to have standard browser > features like > ftp, frame, javascript, bookmarks, mailing capability etc. > > I will appreciate if somebody can point me to useful resource on the > same. > > Regards, > Yusuf --------------------------- Newsgroups: comp.os.vxworks Subject: BSP Subcontractor (Israel) Date: Thu, 30 Dec 1999 14:47:28 +0200 From: "Yael Abudi" Message-ID: <386b53c6.0@news.barak.net.il> Hello all, I'm looking for a subcontractor to develop a BSP in Israel. The CPU is MPC860T. Anyone who is interested (or has suggestions), please reply. Yael Abudi Shiron Satellite Communications yaela@shiron.com Tel: +972-3-9210305 ext. 212 Fax: +972-3-9214547 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ATM PMC Adapter Device Driver for BSP MV2700 Date: Mon, 20 Dec 1999 13:38:59 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <385E85D3.B05AF87B@aps.anl.gov> References: <38596A38.F422F858@spawar.navy.mil> Ron Webb wrote: > > We looking for ATM PMC Adapter Device Driver for BSP MV2700. We have > two cards from two manufacturers, Interphase 4575 and Cetia. However, > the softwares that go with both card were written (or configured) for > BSP MV2600, not MV2700 as we have. ... > 3. Can the ATM PMC driver which is built for MV2600 work with the card > and the VxWorks OS built for MV2700 ? I can't speak specifically about the cards you are asking about, but we have successfully used some video software that was only available for the MV2604 on an MV2700 with no problems. IIRC the gcc compiler flags are identical and in fact the same BSP is used for both with switches between the two models, so I suspect you'll find both those ATM cards will both work. However you should get one on trial and prove it to your (or your companies) satisfaction before paying for it... - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cost Effective Training? Date: Thu, 30 Dec 1999 16:53:00 GMT From: "George W. Freas II" Organization: Boeing Message-ID: <386B8DEC.5B9C1BA5@hsv.boeing.com> References: <38566f37.90786350@news.eznet.net> <8OY54.5232$ZT5.2767135@news1.rdc1.on.wave.home.com> <385f39f2.1452899@news.fe.internet.bosch.de> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Try out Embry Riddle Univ.: http://www.embryriddle.edu I know they have a course that covers VxWorks, but not sure if it is offered as a distance education class (they are in Daytona, FL). Another place has a class: http://www.menet.umn.edu This is Univ. of Minnesota, the course number is ME5271. Not sure of the status of these courses, and if they are still offered. Hope this helps. George Freas. - -------------------------------------------------------------- Jan Koeppen wrote: > > On Thu, 16 Dec 1999 03:14:12 GMT, "Peter on Rogers" > wrote: > > >You might want to contact your Wind River local representative. Many of the > >regions offer local on-site VxWorks Quickstart programs. While not > >formalized training, the program might meet your needs. > > > Hi Peter, > what would these Quickstart programs comprise? > We have some workers (good and experienced in every aspect of embedded > software but Tornado) in need of some quick getting to work and the > Tornado Trainig offered by WRS seems rather expensive. And I doubt > that our WRS representative would offer anything we do not explicidly > ask for :-( > > regards > > Jan Koeppen > -- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GUI for VxWorks (Zinc or Qt) Date: Sun, 02 Jan 2000 19:52:59 GMT From: Ken Hayber Organization: - Message-ID: <946842468.120587028@news-server.socal.rr.com> References: <37FDB080.9DEB0FFB@rz.tu-ilmenau.de> Followup-To: comp.os.vxworks On Fri, 08 Oct 1999, aleXXX wrote: >Hi, > >has anybody experiences with working with Zinc or with Qt under VxWorks >? >At the moment we have to decide which one to use. >Would be nice if anybody could give me some hints. > >Thanks in advance >Alex My company has begun development of a GUI application on vxWorks/Zinc on a PC system. The nice parts are being able to develop on Win32 and then just compile for vxWorks with very little visual diference; and it looks/feels like Windows (good for customer). We have had very few problems so far; but it's still a bit early in development. We did not look at Qt; as it was not available on vxWorks at that time. (I assume from your question that it is). My understanding of Qt makes me think that you might get the same benefits. Good Luck! K. Hayber --------------------------- Newsgroups: comp.os.vxworks Subject: Re: __udiv3 etc unresolved ? Date: Sun, 2 Jan 2000 20:17:56 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <84pdvi$cd6$1@newsmaster01.magnet.at> References: <83od8n$e5l$1@newsmaster01.magnet.at> <385FBDDA.390FB4DF@lucent.com> <83opip$rjf$1@newsmaster01.magnet.at> <3860D74A.BAB5B04F@lucent.com> our GNU expert looked at the problem and solved it. the problem is, that these functions are part of the C-Std lib, but this one is not linked into VxWorks and into the user App. he extracted the function from the compiler source, added it as user function and now the unresolved external is gone. While analyzing this problem, he run into some other problems. maybe some one can help here. =========== I had a look at the makefiles generated by Tornado and those used in our environment to build the OPLC Layer and I realized, that -nostdlib is specified as a compile option. Of course, this causes libgcc.a not bo be linked with the applications and I wonder why. Are there any problems linking libgcc.a with your own objects (or simply building an object just containing libgcc.a and loading this one with ld <)?. I tried to link with libgcc.a and at least the resulting object could be loaded on VxWorks. I have not yet tested if it works. Also I am a bit curious why the Tornado generated makefiles pass some options to the compiler or linker. Maybe you can enlighten me a bit about this: 1. -nostdlib. See above. Are there any problems linking with libgcc.a. I thought of using something like -print-file-name=libgcc.a in addition to -nostdlib so just libgcc.a is linked in. Of course, this would result in multiple definitions of the same symbol (by name) when loading more than one object built with this option. 2. -fno-builtin. Is there any reason why the GCC built-in functions (which I assume to be normally faster than the library functions) should not be used. According to the docs, this affects the use/linking of abort, abs, alloca, cos, exit, fabs, ffs, labs, memcmp, memcpy, sin, sqrt, strcmp, strcpy and strlen. 3. -fno-defer-pop is just another issue to optimization, but I think just a minor one. But I don't understand why this option is specified and I didn't find any hint in the Tornado documentation. 4. -fvolatile. Ouch, this really hurts. If it isn't required (and I found some comment like "optional" in the Tornado docs), I think it should not be used (just for performance reasons). Am I right in this case? 5. -r is specified in a "pre-linking" stage, and I could not find this option in any GCC documentation (nor when I quickly checked the egcs 2.90 source code, too). Do you know, what this is for? - -- regards, Mario Semo. http://www.kirchnersoft.com Douglas Fraser wrote in message <3860D74A.BAB5B04F@lucent.com>... > > >Mario Semo wrote: >> >> Doug, >> >> thx for your help. i have now added such a file + function to my VxWorks >> simulation kernel > >I haven't tried this on the vxWorks simulator, just on >a real ARM (SA110) target. > > > >I did something similar to that. It is in a C file that is >added to the kernel using the MACH_EXTRA definition in the >Makefile, for example... > >MACH_EXTRA = llUnused.o > >The build for vxWorks completes fine as before. >Then when I load my application image using 'ld' >everything is okay. > >> >> and then rebuild all. >> >> compile result: >> >> partialImage.o: In function `_tff': >> \wpwr\host\src\gnu.simpc.cpp\gcc\cp\tinfo2.cc(.text+0x1ed0): undefined >> reference to `__div >> di3' >> partialImage.o: In function `_tfUi': >> \wpwr\host\src\gnu.simpc.cpp\gcc\cp\tinfo2.cc(.text+0x1f70): undefined >> reference to `__udi >> vdi3' > >It is odd that it is failing in the compiler stage. >Again, this may be a simulator issue. > >Anybody else out there had this issue with the simulator??? > >Doug > >> make: *** [vxWorks] Error 0x1 --------------------------- Newsgroups: comp.os.vxworks Subject: 'SNMP daemon' and 'DHCPv4 runtime client' related? Date: Mon, 03 Jan 2000 14:35:44 +0530 From: VijayKumar P Organization: Cisco Systems Message-ID: <38706667.4FC3AA47@cisco.com> This is a multi-part message in MIME format. - --------------494C1A4EA1060A4EF5047CB7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, I am unable to make out how 'DHCPv4 runtime client' component is dependent on 'SNMP daemon '. Any Idea?. I am using ver 5.4 and Tornado 2.0 The 'SNMP daemon' component gets included when ever I include 'DHCPv4 runtime client'. And when I try to build image, I get ....\prj\prjConfig.c:76: snmp\snmpdInit.h: No such file or directory make: *** [prjConfig.o] Error 0x1. I believe this is because the SNMP being an optional component and not available with me is giving the trouble. But WHY should it include 'SNMP daemon' in the first place????. Thanks, Vijay - --------------494C1A4EA1060A4EF5047CB7 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for VijayKumar P Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India fn:Vijay Kumar.P end:vcard - --------------494C1A4EA1060A4EF5047CB7-- --------------------------- Newsgroups: comp.os.vxworks Subject: taskInit and taskActivate vs taskSpawn Date: Mon, 03 Jan 2000 13:38:48 -0500 From: steven.offenbacher@jhuapl.edu Organization: Johns Hopkins University Applied Physics Lab, Laurel, MD, USA Message-ID: <3870ECB8.2AFB7A18@jhuapl.edu> I'm attempting to replace a taskSpawn with the taskInit/taskActivate function call and have not been successful. I'm looking for a C code example of working taskInit/taskActivate functions. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskInit and taskActivate vs taskSpawn Date: Mon, 03 Jan 2000 21:53:30 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <84r5oj$rrs$1@nnrp1.deja.com> References: <3870ECB8.2AFB7A18@jhuapl.edu> Steven, In article <3870ECB8.2AFB7A18@jhuapl.edu>, steven.offenbacher@jhuapl.edu wrote: > I'm attempting to replace a taskSpawn with the taskInit/taskActivate > function call and have not been successful. I'm looking for a C code > example of working taskInit/taskActivate functions. Is there a good reason you are doing this? It does complicate things somewhat since you are responsible for all the allocations of memory (stack and TCB)... The most common reason for doing this is to create dormant tasks at initialisation time which can then be woken up later on. If that is what you are trying to do there are simpler ways: you could get the tasks to start and block on a semaphore, or even suspend itself immediately. Hope that helps a little, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Embedded developers using VxWorks, Linux, NT or CE wanted Date: Mon, 27 Dec 1999 08:52:36 -0800 From: "Nick Berliner" Message-ID: <4LM94.3$cT.86@news.ais.net> GoAhead Software is offering a chance to beta test remote upgrading capability in your networked appliance or device. GoAhead FieldUpgraderT provides manufacturers of networked devices with a comprehensive solution for automatically upgrading their products in the field. GoAhead FieldUpgrader allows you to upgrade software, firmware, O/Ss or data using a device-initiated approach. These upgrades can provide products with new features, reliability enhancements, performance improvements, updated applications, security patches and bug fixes. GoAhead FieldUpgrader removes the end-customer burden of manually upgrading their product, and reduces the manufacturer's expenditures for costly support visits and product recalls. If you are developing in an NT or Linux environment and your device runs Embedded NT, Embedded Linux, VxWorks, CE, NT, or Linux, you may qualify for participation. If you are interested in trying GoAhead FieldUpgrader out in your device, visit: http://www.goahead.com/fieldupgrader/fubeta.htm or email beta@goahead.com. GoAhead FieldUpgrader beta program will run through February 2000. Please sign up right away if you are interested in participating!! Best regards, Nick Berliner ____________________________________________________________________________ ____ Nick Berliner Internet Marketing Coordinator/GoAhead Software nickb@goahead.com www.goahead.com phone 425-468-5404 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: File transfer speed Date: Tue, 04 Jan 2000 08:51:05 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38719859.EDB8C89B@bitband.com> References: <84r3fn$g5m2@overload.lbl.gov> Reply-To: leonid@bitband.com Steve, you must decide what is it that you need: FTP file transfer or application specific socket-based data trqansfer, and in the later case TCP and UDP greatly vary. If you are interested in FTP and data transfer, it is known for a some time that the FTP server which is shipped with Tornado is very very slow, and has other downsides to it too. The DOS file system is not perfect either. Upgrading to DosFs 2.0 will give you an improved file system, and an enhanced (much faster too) FTP server albeit in the "unsupported" directory. Perhaps I could be more helpful if the questions where more specific. Leonid Steve Bauer wrote: > > Using FTP to a VxWorks server, we're getting only about 0.4 megabytes of > data per second on file transfers over a 100BaseTX network (with no other > traffic), with either a PC or an Alphastation as the client. Has anyone > done significantly better than this? With what hardware and software? > Will a custom sockets-based program do much better? > > Steve Bauer > bauer@kscorp.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jan 8 17:56:54 2000 From: Operator Date: Sat Jan 8 17:56:56 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jan 6 04:00:35 PST 2000 Subject: Re: SENS - IP bind Subject: Re: SNMP m2lib sysUpTime error Subject: Error in loadCoff.h ? Subject: getch() alike? Subject: !!!!!!1500 SEXY PICS FREE TO DOWNLOAD NOW!!!!! Subject: Re: __udiv3 etc unresolved ? Subject: Re: DHCP.. Subject: Stability of dosFS2.0 ?? Subject: Re: Reading from stdin? Subject: ldppc for bootrom_res Subject: Re: Problems with fei ethernet driver ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: SENS - IP bind Date: Tue, 04 Jan 2000 12:21:03 +0000 From: K-Boy Organization: Spider Software Limited Message-ID: <3871E5AF.4ADF@tardis_hates_spam.ed.ac.uk> References: <385FB6D5.BC4C0855@emw.ericsson.se> <3860EC1A.5DA3@tardis_hates_spam.ed.ac.uk> <38708B8D.5B6A8B36@emw.ericsson.se> Hi, Perhaps you could do a ifunit() and see if it returns non-NULL. Similarly muxFindByName() will tell you if the device is loaded into the MUX. - -- K. Tord Andersson wrote: > > K-Boy wrote: > > > > Tord Andersson wrote: > > > > > > Is there a simple way except muxShow to find out if a loaded device > > > has been attached to IP and ARP? > > > > > > Regards, > > > > > > Tord Andersson > > > > ifShow() will work too. > > Yes, true but need interpretation of output string. > Is there some other way to get a more boolean like result? > MIB access? > > Regards, > > Tord Andersson > > > > > -- > > K. > > > > > -- > > > ----------------------------------------------------------- > > > EMW/SB/TN Tord Andersson, QMWTORD > > > Phone: +46 31 74 70 486 Mobile: +46 708 89 50 85 > > > Email: tord.andersson@emw.ericsson.se > > > Ericsson Microwave Systems AB, S-431 84 Molndal, Sweden > > -- > ----------------------------------------------------------- > EMW/SB/TN Tord Andersson, QMWTORD > Phone: +46 31 74 70 486 Mobile: +46 708 89 50 85 > Email: tord.andersson@emw.ericsson.se > Ericsson Microwave Systems AB, S-431 84 Molndal, Sweden - -- Khalid Rafiq Spider Software Limited Tel: +44 131 475 7009 8 John's Place Fax: +44 131 475 7001 Edinburgh mailto:khalid@spider.com EH6 7EL http://www.spider.com/ Scotland --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SNMP m2lib sysUpTime error Date: Tue, 4 Jan 2000 14:55:09 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <84t03e$6h2$1@ndsukns01.ndsuk.com> References: <84su8r$5io$1@ndsukns01.ndsuk.com> Will Fookes wrote in message <84su8r$5io$1@ndsukns01.ndsuk.com>... >If the calculation were performed without losing the top bits, I would >expect sysUpTime to wrap after 828 days, 12h:6m:28.25s with the default 60Hz >tick. ...though of course, the TIMETICKS object is restricted to 32 bits and will therefore wrap after 497 days, 2h:27m:52.95s. Since both the returned sysUpTime object and the tickGet() result are wrapping independently, the actual wrapping points after the initial 497 days will strobe! (But we will have leaked away all our memory by then, so this becomes academic ;-) Will --------------------------- Newsgroups: comp.os.vxworks Subject: Error in loadCoff.h ? Date: Tue, 4 Jan 2000 10:32:59 -0500 From: "Dave Hill" Message-ID: Hello All, Whenever I include the file "\Tornado\target\h\loadCoff.h" I receive the following error message during project creation: can't read "macros(WDB_MODE)": no such element in array I have isolated the problem to the following 2 lines of code in loadCoff.h: #define n_zeroes _n._n_n._n_zeroes #define n_offset _n._n_n._n_offset If I comment out these lines of code the error disappears. I am using Tornado II for the PowerPC. I am developing on a Win98 platform. Please let me know if you have any info. Thank you very much, David Hill Safetran Systems --------------------------- Newsgroups: comp.os.vxworks Subject: getch() alike? Date: Tue, 4 Jan 2000 17:20:21 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <84t6js$sl$1@newsmaster01.magnet.at> Hello, I want to port my testenviron (my apps test envir) to vxworks. but the test program uses low level io requests like: getch() (read char with echo from keyboard, do not wait for CRLF). is there something available for VxWorks? getchar() does not echo and waits for endofline. - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: !!!!!!1500 SEXY PICS FREE TO DOWNLOAD NOW!!!!! Date: Tue, 04 Jan 2000 19:48:52 GMT From: eqeq@lll.com Message-ID: http://fly-in.to/youngteens --------------------------- Newsgroups: comp.os.vxworks Subject: Re: __udiv3 etc unresolved ? Date: 4 Jan 2000 21:49:00 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <84tpsc$555$1@news1.xs4all.nl> References: <83od8n$e5l$1@newsmaster01.magnet.at> <385FBDDA.390FB4DF@lucent.com> <83opip$rjf$1@newsmaster01.magnet.at> <3860D74A.BAB5B04F@lucent.com> <84pdvi$cd6$1@newsmaster01.magnet.at> Mario Semo wrote: : 1. -nostdlib. See above. Are there any problems linking with libgcc.a. I : thought of using something like -print-file-name=libgcc.a in addition : to -nostdlib so just libgcc.a is linked in. Of course, this would result in : multiple definitions of the same symbol (by name) when loading more than one : object built with this option. I think this is caused by the fact that the functions that are in the stdlib are replaced by the functions in the VxWorks libraries. : 5. -r is specified in a "pre-linking" stage, and I could not find this : option in any GCC documentation (nor when I quickly checked the egcs 2.90 : source code, too). Do you know, what this is for? This is used to generate relocatable code. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: DHCP.. Date: Wed, 05 Jan 2000 14:56:50 +0300 From: "Arthur V. Grigorian" Organization: Auriga, Inc. Message-ID: <38733182.3FC30E13@auriga.ru> References: <3869E447.68CD6536@cisco.com> <3869F017.6F48D50@auriga.ru> <386C3426.F628A571@cisco.com> <38721211.53FB067F@auriga.ru> <38730E67.89AE1C96@cisco.com> Reply-To: archy@auriga.ru Hello Vijay, > Thanks Arthur. I should not have any such prob. as I have set my own private > net. > > If you are not bored, I have some more queries......;) > > After booting over the net, I am successfully able to initialize the dhcplib > using > > dhcpcLibInit( 67, 68, 4, 5, 2760, 30 ); FYI: If you have 'INCLUDE_DHCPC' macro defined vxWorks performs dhcpcLibInit(DHCPC_SPORT, DHCPC_CPORT, ...etc...) call setting the default DHCP parameters, so you don't have to double this call if these default settings are enough for you. (see .../target/config/comps/src/net/usrNetDhcpcCfg.c) > Next I am calling the following code to set the options, bind and then get > params... > > STATUS My_dhcpc_start() > { > struct ifnet *pIf; > STATUS result; > struct dhcp_param pParamList; > void * pLeaseCookie; > > pIf = ifunit ("ene0"); /* Access network device. */ > > /* Initialize lease variables for automatic configuration. */ > > pLeaseCookie = dhcpcInit (pIf, TRUE); > if (pLeaseCookie == NULL) > { > printf( "dhcpcInit Failed \n"); > return (ERROR); > } > > /* Set any lease options here. */ > > result = dhcpcOptionSet (pLeaseCookie, _DHCP_LEASE_TIME_TAG, 2760, 0, > NULL); > if (result != OK) > { > printf( "dhcpcOptionSet Failed \n"); > return (ERROR); > } > > result = dhcpcBind (pLeaseCookie, FALSE); /* Asynchronous execution. > */ > if (result != OK) > { > printf( "dhcpcBind Failed \n"); > return (ERROR); > } > > result = dhcpcParamsGet( pLeaseCookie, &pParamList ); > if (result != OK) > { > printf( "dhcpcParamsGet Failed \n"); > return (ERROR); > } > } > > Here in the function 'dhcpcParamsGet' , I am getting 'ERROR'. > Any clue???. That's the result of your /* Asynchronous execution. */ Read manual. You request parameters while the negotiation hasn't completed. You have two variants: - Start your function as a separate task and use dhcpcBind (pLeaseCookie, TRUE); - Or install dhcp events hook by dhcpcEventHookAdd() and do dhcpcParamsGet() there. Anyway you'd better RTFM at last :} > Also something funny I am observing is that, I am 'unable' to > 'Launch' when I am in my private network (which I have set up > using a hub). So I need to connect host and target PCs to the > LAN, then Launch and then continue my debugging on private net!!!. > Again any clues??? Not enough information. What do ya mean saying "Launch"..? Regards, __ R2 --------------------------- Newsgroups: comp.os.vxworks Subject: Stability of dosFS2.0 ?? Date: 5 Jan 2000 23:31:00 GMT From: cjohnson@oresis.com Organization: NewsOne.Net - Free Usenet News via the Web - http://newsone.net/ Message-ID: <850k7k$ilt$1@news.netmar.com> VxWorks has a new release of the dosFS file system as an update to Tornado 2.0 which has many new features. (FAT32, VFAT, chkdsk, etc.) Does anyone have any experience with this in terms of stability & reliability ?? Thanks. Charles Johnson Oresis Communcations, Inc. Beaverton, OR cjohnson@oresis.com ----- Posted via NewsOne.Net: Free Usenet News via the Web ----- ----- http://newsone.net/ -- Discussions on every subject. ----- 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: Re: Reading from stdin? Date: Thu, 06 Jan 2000 00:17:03 GMT From: Jeremey Barrett Organization: SPYRUS Message-ID: <3873DE97.7FDB1CB5@terisa.com> References: <387228AE.4081010B@terisa.com> Reply-To: jeremey@terisa.com Jeremey Barrett wrote: > > I'm having some trouble reading from stdin (using fgets()). I'm > connecting to the target via windsh, loading the app, and running it... > it displays output on stdout and stderr just fine. When it calls > fgets(), it just sits, and nothing I type will get it to return. > > Is there something related to windsh that prevents this from > working properly, or something I need to set up differently? That's > the only thing I can figure. An addendum to this: stdio is being redirected correctly (via /vio/0), but reading from stdin still hangs, as if what is typed into windsh is never received by the running app. Any pointers/ideas would be helpful. Thanks, Jeremey. - -- Jeremey Barrett GPG fingerprint = 7BB2 E1F1 5559 3718 CE25 565A 8455 D60B 8FE8 B38F --------------------------- Newsgroups: comp.os.vxworks Subject: ldppc for bootrom_res Date: Thu, 06 Jan 2000 06:45:27 GMT From: findik@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <851dm5$8lf$1@nnrp1.deja.com> I am trying to build vxWorks.res_rom.bin. I used Tornado II, standard BSP builds: vxWorks.res_rom I couldn't succeded to run it on my target, MBX860 board. I also tried to build bootrom_res and the following are some trials to build bootrom_res: ldppc -X -N -e _romInit -Ttext FE000100 -Tdata 00200000 -o bootrom_res romInit_res.o bootInit_res.o version.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a E:\Tornado\host\x86-win32\bin\romsize PPC -b 00200000 bootrom_res E:\Tornado\host\x86-win32\bin\elfToBin bootrom_res.bin I get the following error after elfToBin: seg1: Expected load address 0xfe000100 but file address is 0x002078D0 (This is the standard link and I see romInit is placed at 0x78D0) ldppc -X -N -e _romInit -Ttext FE000100 -Tdata 00200000 -o bootrom_res romInit_res.o bootInit_res.o version.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a SILrmtDevCtl.o (I add an extra object SILrmtDecCtl.o, I see romInit is advanced to 0x78F4-I expect that data part of this extra object results the difference btwn 78F4 and 78D0) ldppc -X -N -e _romInit -Ttext FE000100 -Tdata 00200000 -o bootrom_res bootInit_res.o version.o romInit_res.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a (I place romInit_res.o after version.o, I see romInit is advanced to 0x79c4) ldppc -X -N -e _romInit -Ttext FE000100 -Tdata 00200000 -o bootrom_res bootInit_res.o romInit_res.o version.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a (I place romInit_res.o before version.o, I see romInit is advanced to 0x79c4 - to be before or after version.o doesn't make any difference) ldppc -X -N -e _romInit -Ttext FE000100 -Tdata 00200000 -o bootrom_res bootInit_res.o version.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a romInit_res.o (I place romInit.o to the end, I see romInit is advanced to 0x5a540) ldppc -X -N -e _romInit -Ttext FE000100 -o bootrom_res romInit_res.o bootInit_res.o version.o bootConfig.o sysALib.o sysLib.o E:\Tornado\target\lib\libPPC860gnuvx.a (I removed -Tdata 00200000, I see romInit is placed at 0x100 - This position will be the right one, isn't it? But I want the text in ROM and the data part in RAM, shouldn't I define data location seperately for rom resident bootrom or vxWorks builds? ) Best regards, reyhan Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with fei ethernet driver Date: Thu, 06 Jan 2000 11:04:50 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38745AB2.701BB2BB@bitband.com> References: <38738E04.CD6DF15@t-online.de> Reply-To: leonid@bitband.com I am not sure if my scenario was similar enough to yours, but yes, I did encounter the fact that the BSD fei driver as delivered with Tornado 2 would freeze under certain conditions. Which version of the driver do you ise, BSD or END ? The drivers are totally separate. Leonid Heribert Volkhausen wrote: > > Hi, > > I discovered some strange behaviour using the fei driver for a > Intel 82559 fast ethernet controller. > I'm using the pentium BSP. > > I started a flood ping two times on a LINUX operated PC (call: ping > -f ). After a > second, the target didn't answer anymore. I stopped the pinging and > started a simple ping and > got no answer. I had to reset my target to get a new connection. > > Did anyone discover this effect, too? > > Regards > Heribert Volkhausen --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Jan 8 18:08:35 2000 From: Operator Date: Sat Jan 8 18:08:37 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jan 7 04:00:25 PST 2000 Subject: Re: MAC address from IP address Subject: What is the equvalent of a Unix call ftok? Subject: TELNET CLIENT Subject: Re: HIADJ() and the gnu compiler ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: MAC address from IP address Date: 6 Jan 2000 20:17:44 GMT From: Troy Cauble Organization: Bell Labs, Holmdel, NJ, USA Message-ID: <852t98$469@nntpb.cb.lucent.com> References: <852e7v$1flh$1@buggy.easynet.fr> Yves Nortier wrote: : Hello, : How can a Vxworks program retrieve the MAC address associated to an IP : address ? etherAddrResolve() --------------------------- Newsgroups: comp.os.vxworks Subject: What is the equvalent of a Unix call ftok? Date: Thu, 06 Jan 2000 23:07:20 GMT From: caltuskim@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <85376v$jj3$1@nnrp1.deja.com> Hi, I am porting a UNIX program onto VxWorks. The Unix program uses ftok function to generate a key when creating or accessing the IPC structures such as message queue, semaphore, and shared memory. I tried to find a similar function call as ftok, but I couldn't find one. I think there is no such function in Vxworks. Has anybody done this call mapping from Unix to Vxworks? Do I have to use predefined keys instead of generated keys by ftok function? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: TELNET CLIENT Date: Fri, 07 Jan 2000 02:31:02 GMT From: tornado2000@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <853j56$rtn$1@nnrp1.deja.com> I am in need of a telnet client for vxWorks. Anybody has tried porting BSD telnet client to vxworks ? I need help in porting it to vxworks. The vxworks tty driver doesn't seem to support many of the unix like features ? Anybody has tried porting BSD code to vxworks ? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: HIADJ() and the gnu compiler Date: Fri, 07 Jan 2000 09:00:02 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8549ud$bdn$1@nnrp1.deja.com> References: <__6d4.1385$9F3.108986@news.corecomm.net> Hello Dave, To understand this you need to understand the instruction set a little... PPC 860 is a RISC CPU that has one 32 bit word per instruction; the largest constant value you can load into a register is 16 bits because the remainder of the bits are needed to encode the instruction and the destination register etc. So, to load a 32 bit constant requires two instructions: a load+shift and an add operation. In the PPC world the add is a signed operation, with the 16 bit value being sign extended to 32 bits first. HIADJ() obtains the upper half of the 32 bit constant, but it also compensates for the sign of the lower half so the addition that is expected to follow (an addi instruction, or equivalent) will result in the correct value being used. There is another macro, HI(), which doesn't do the compensation and is used with an OR operation following. LO() is used in both cases. So, why not use the OR option always? Well, the add can be implied as part of an indexed load to access external memory through a pointer. For example, the following two instructions load the value pointed to by myPointer into r30: lis r31, HIADJ(myPointer) lwz r30,LO(myPointer)(r31) Had you used HI(myPointer), the operation would only work if LO(myPointer) was between 0x0000 and 0x7fff; outside that range you would be loading from the wrong location. Hope that helps, John... In article <__6d4.1385$9F3.108986@news.corecomm.net>, "Dave Hill" wrote: > Hello All, > > I notice in some of Wind's code for the PowerPC 860 that the > macro HIADJ() is used. Looking at assembler output, I see that > HIADJ(arg) translates to arg@ha . Unfortunately, I can't find out what the > "@" or the "ha" are supposed to be doing. It looks like a 16-bit shift > left, but I am confused on this syntax. Can anyone enlighten me? > > Thanks, > Dave Hill > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jan 9 23:39:26 2000 From: DOUET Raymond Date: Sun Jan 9 23:39:28 PST 2000 Subject: Cross-compiler for 68k & wrs on OSF Hi, I'm looking for someone who has succeed (I did'nt !)to build a cross compiler for 68040 and vxworks on OSF1 V4.0. Thanks in advance. Raymond Douet ++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++ Raymond Douet S2I - I.P.N Orsay bt 102 91406 ORSAY Cedex FRANCE email: douet@ipno.in2p3.fr phone: +33 (0)1 69 15 64 41 fax: +33 (0)1 69 15 45 03 ++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++-++ From vxwexplo-errs@csg.lbl.gov Mon Jan 10 01:42:29 2000 From: "Michael Lawnick" Date: Mon Jan 10 01:42:31 PST 2000 Subject: Re: Problem elf'ing vxWorks_rom for PPC > From: "Mark Del Giorno" > Subject: Problem elf'ing vxWorks_rom for PPC > Heyyyyyyyyyyyyyy, > > I'm trying to burn my application into the mv2700 flash so it will run on > power up. I added the patch to allow > 1 MB flash applications to run > (untested). I've added my application to MACH_EXTRA. I've edited > usrAppInit.c to call it. I've confirmed that if I make vxWorks and have my > board boot, my application runs. I changed my project to make vxWorks_rom. > > but: > When I run elfToBin on it, I get the following output: > > > elfToBin < vxWorks_rom > test.bin > seg 1: Expected load address 0xfff00100 but file address is 0x00586a88 > > test.bin is still generated (and looks to be about the right size). > Has anyone seen this? > Is this an error or just a warning that can be ignored? > (I know the code will be copied to RAM, then run....???) > > Thanks in advance for any help! > > - Mark > Hi Mark, this is a bug in TII for PPC. I have mailed to WRS in december last year and after a while I got a new elfToBin, only to be used for romable vxWorks-Images ( not bootrom, this needs the original ) on all PPC. The SPR# is 8845 (supporter said he took the file from there), but when I looked for it on WindSurf (for this reply), I could not access it. I'm sorry not beeing able to send you the file as an attachment, as I don't know about license/copyright problems on this file. So please contact your sales representative and tell him the SPR- number. MfG Lawnick, SOFTEC GmbH ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== From vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:28:04 2000 From: "Glaubitz, Paul" Date: Mon Jan 10 05:28:07 PST 2000 Subject: Tornado "standard" C header files A coworker handed me part of a discussion on problems on varargs under Tornado 2. The discussion was centered on PPC's, based on a problem I ran into recently when "upgrading" from T1 to T2, several functions which used varargs stopped working (Our platform is x86). We tracked it down to the new needing CPU to be defined with values from which is part of . I was told by customer support that making sure was the first include was the way to "fix" the problem. They put in a enhancement request that be put at the beginning of the standard headers, the SPR is 29706. If more people start requesting it hopefully it will get moved up in priority. Paul Glaubitz Senior Software Engineer Virtual Instruments Group Areoflex Lintek From vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:32:26 2000 From: =?iso-8859-1?q?Eric=20Lucazeau?= Date: Mon Jan 10 05:32:28 PST 2000 Subject: Secure File System for Flash Memory Hello happy vxWorks users, I am a software engineer and I am in charge of developping a secure File System for a Flash Memory device. The file system must resist to loss of power : in case of a suddent power breakdown, the file system must stay consistent and loose nothing but the last initiated transaction. It appears to me that implementing that type of service with DosFs + TrueFFS will be quite difficult. Though TrueFFS is safe, DosFs is not. Does anyone know about another file system - a bit richer than rawFs - that would satisfy my needs ? I am sure somebody has already had to face this problem. Any help is welcome. ===== Eric Lucazeau elucazeau@yahoo.fr ___________________________________________________________ Do You Yahoo!? Achetez, vendez! À votre prix! Sur http://encheres.yahoo.fr From vxwexplo-errs@csg.lbl.gov Mon Jan 10 06:58:37 2000 From: "Robert L. Perry" Date: Mon Jan 10 06:58:39 PST 2000 Subject: RE: vxWorks dosFs 2 stability & reliability Charles, We've worked w/ the new file system on our (fibre channel) raid disks since the early access release in Tornado 1. It is a big improvement over the older file system, not to mention the large disk support. Best of luck, Bob Perry bobp@excaliburs.com VxWorks has a new release of the dosFS file system as an update to Tornado 2.0 which has many new features. (FAT32, VFAT, chkdsk, etc.) Does anyone have any experience with this in terms of stability & reliability ?? Thanks. Charles Johnson Oresis Communcations, Inc. Beaverton, OR cjohnson@oresis.com From vxwexplo-errs@csg.lbl.gov Mon Jan 10 07:03:57 2000 From: "Freeland, Mike" Date: Mon Jan 10 07:03:59 PST 2000 Subject: building vxWorks.st_rom with T2 Hi, I am new to Tornado 2 for Windows and would like to build vxWorks.st_rom. Can anyone help me tell me how to use the IDE to do that? Thanks, Mike Freeland Software Engineer Lockheed Martin Space Electronics and Communications 9500 Godwin Dr. Mail Stop 110/018 Manassas, VA 20110 office 703-367-5328 fax 703-367-2229 email - mike.freeland@lmco.com From vxwexplo-errs@csg.lbl.gov Mon Jan 10 07:21:06 2000 From: "Robert L. Perry" Date: Mon Jan 10 07:21:09 PST 2000 Subject: RE: vxWorks dosFs 2 stability & reliability Charles, We've worked w/ the new file system on our (fibre channel) raid disks since the early access release in Tornado 1. It is a big improvement over the older file system, not to mention the large disk support. Best of luck, Bob Perry bobp@excaliburs.com VxWorks has a new release of the dosFS file system as an update to Tornado 2.0 which has many new features. (FAT32, VFAT, chkdsk, etc.) Does anyone have any experience with this in terms of stability & reliability ?? Thanks. Charles Johnson Oresis Communcations, Inc. Beaverton, OR cjohnson@oresis.com From vxwexplo-errs@csg.lbl.gov Mon Jan 10 10:38:18 2000 From: David Chen Date: Mon Jan 10 10:38:21 PST 2000 Subject: Startup Job Opportunity Hi All: Clarinet Systems is a dynamic new startup company in the heart of the Silicon Valley. The company designs, manufactures and markets high-speed Infrared and Bluetooth LAN/WAN and Internet access systems. The projects are exciting and challenging in a productive and comfortable office environment. If you're looking for a challenge and an opportunity to work with cutting edge products look no further. We offer excellent benefits and compensation. For more details, please visit our web site @ www.clarinetsys.com. Sr. Software Eng: BS CS/EE or higher degree with a minimum of 5+ yrs SW experience using C, real time OS and knowledge of networking protocols. Exp. in at least one of following: vxworks, network switching, bridge or router, infrared protocol, Bluetooth or device driver. Participate in all phases of SW development. Ability to take responsibility, ownership and initiate action. Must be a team player and a self-starter. Please send your resume to HR, Clarinet Systems, 1415 Koll Circle, #101, San Jose, CA 95112; fax 408-501-0255; email dchen@clarinetsys.com From vxwexplo-errs@csg.lbl.gov Mon Jan 10 11:05:34 2000 From: "Blake, Don" Date: Mon Jan 10 11:05:36 PST 2000 Subject: RE: Secure File System for Flash Memory I can't answer you specific questions but there is something else you should be aware of. Flash memory consistency during power loss is going to impose two additional requirements on your power system: 1. Power must be held up long enough to complete the current write operation. Depending on the implementation, this may need to include the time it takes to perform a bank erase followed by a number of write operations. 2. The software must get notification (e.g., an interrupt) that power is possibly going down so that it can complete any in process write operation and wait for power to go down or to come back before initiating any subsequent writes. Don Blake Sr. Programmer Lockheed Martin Federal Systems Owego, NY -----Original Message----- Submitted-by vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:32:26 2000 Submitted-by: =?iso-8859-1?q?Eric=20Lucazeau?= Hello happy vxWorks users, I am a software engineer and I am in charge of developping a secure File System for a Flash Memory device. The file system must resist to loss of power : in case of a suddent power breakdown, the file system must stay consistent and loose nothing but the last initiated transaction. It appears to me that implementing that type of service with DosFs + TrueFFS will be quite difficult. Though TrueFFS is safe, DosFs is not. Does anyone know about another file system - a bit richer than rawFs - that would satisfy my needs ? I am sure somebody has already had to face this problem. Any help is welcome. From vxwexplo-errs@csg.lbl.gov Mon Jan 10 11:16:01 2000 From: "Darren Carlson" Date: Mon Jan 10 11:16:04 PST 2000 Subject: Tornado DOSFS 2 warning message Hello all, How can you get rid of this warning message: WARNING : dosChkLib : system clock is being set to WED JAN 05 01:13:54 2000 Value obtained from file system referenced by volume descriptor pointer: 0x7aa898 The old setting was THU JAN 01 00:00:00 1970 Accepted system dates are greater than FRI JAN 01 00:00:00 1999 The RTC has been set using DOS's date function, to the current date, however this message will not go away. Does anyone know a flag or #define I can change to get rid of this Y2K check. Thank You Darren Carlson Software Development Engineer Applied Biometrics Inc 612-890-1123 From vxwexplo-errs@csg.lbl.gov Mon Jan 10 15:16:42 2000 From: "Vu, Jodat" Date: Mon Jan 10 15:16:44 PST 2000 Subject: scsi driver for Adaptec 2920 Hi, This is my first time in soliciting advises and assistance from this vxWorks working group, so please forgive me if this E-mail is not being sent to the proper channel. I have a Pentium PC with a Maxtor ATA hard drive, a NEC CD-ROM ATA drive, and an ADAPTEC 2920 SCSI drive. My goal is to accessing and transferring files between those 3 storage devices under vxWorks. My first task is to establish the communication with the SCSI disk, and I could not find a scsi driver for the Adaptec 2920 in T1 and T2. I had tried to use the AIC7880 driver in T2, and it does not work with the Adaptec 2920. WRS Support advised me in using the Adaptec 2940 for their AIC7880 driver. I'll follow WRS' advise, but in the mean time, I'm requesting assistance on any information about the availability of a vxWorks SCSI driver for the Adaptec 2920. Your helps are very appreciated. l Jodat Vu (972) 603-7730 M/S : MC 00 From vxwexplo-errs@csg.lbl.gov Mon Jan 10 16:13:03 2000 From: "Ferrara, Bob" Date: Mon Jan 10 16:13:05 PST 2000 Subject: netIf vs. endDevTbl in Tornado 2 of VxWorks Can someone tell me the how netIf table is used, as compared the endDevTbl table? I am specifically confused about the IF_USR_xxx definitions; in the code I am looking at, these IF_USR_xxx definitions match up with my CPM Boot String? Any help is appreciated! Thanks, Bob Ferrara bferrara@unispheresolutions.com vxworks From vxwexplo-errs@csg.lbl.gov Tue Jan 11 02:07:00 2000 From: David Laight Date: Tue Jan 11 02:07:02 PST 2000 Subject: RE: Secure File System for Flash Memory IMHO it ought to be possible - with a little thought... The flash devices I have allow you to convert a 1 to a 0 at any time, the 0 to 1 transition requires that a large (128k in my case) area be erased. This means that you can modify the logical to physical map by changing the next map entry from ~0 to the desired value. Use two maps (for when the map area is full) with a count at the start (so you know which is current). Paranoia about partial word writes (which are reasonably quick anyway) could be handled by writing the sector map entries in two stages - leaving the top bit set (and hence an illegal value) on the first write. System startup would need to check whether there was a partially written sector, and validate the last entry in the sector map (rewrite (to ensure fully written) if valid and zero (not a sector) if not valid. File system code - especially the order of writes is left as an exercise... David > Date: Mon, 10 Jan 2000 11:12:02 -0800 (PST) > Subject: RE: Secure File System for Flash Memory > Submitted-by: "Blake, Don" > > > I can't answer you specific questions but there is something else you should > be aware of. Flash memory consistency during power loss is going to impose > two additional requirements on your power system: > > 1. Power must be held up long enough to complete the current write > operation. Depending on the implementation, this may need to include the > time it takes to perform a bank erase followed by a number of write > operations. > > 2. The software must get notification (e.g., an interrupt) that power > is possibly going down so that it can complete any in process write > operation and wait for power to go down or to come back before initiating > any subsequent writes. > > > Don Blake > Sr. Programmer > Lockheed Martin Federal Systems > Owego, NY > > -----Original Message----- > > Submitted-by: =?iso-8859-1?q?Eric=20Lucazeau?= > > Hello happy vxWorks users, > > I am a software engineer and I am in charge of developping a secure > File > System for a Flash Memory device. The file system must resist to > loss of power > : in case of a suddent power breakdown, the file system must stay > consistent > and loose nothing but the last initiated transaction. > > It appears to me that implementing that type of service with DosFs + > TrueFFS > will be quite difficult. Though TrueFFS is safe, DosFs is not. Does > anyone > know about another file system - a bit richer than rawFs - that > would satisfy > my needs ? > > I am sure somebody has already had to face this problem. Any help is > welcome. > ---------------------------------------------------------------- David Laight email: dsl@tadpole.co.uk Tadpole Technology plc phone: +44 1223 428 232 Cambridge, UK fax: +44 1223 428 201 From vxwexplo-errs@csg.lbl.gov Tue Jan 11 04:04:56 2000 From: Vxworks Exploder Date: Tue Jan 11 04:04:58 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jan 11 04:04:53 PST 2000 Subject: DOS Fs ?? VxWOrks newbie Subject: Re: SNMP m2lib sysUpTime error Subject: C`A,B.L`E DE-SCRAMBLER INSTRUCTIONS......FREE C`A-B,L.E`................................. 425 Subject: Re: tgtServer Subject: Re: Problem elf'ing vxWorks_rom for PPC Subject: Re: CPU util measurement? Subject: Using Tornado II and SNiFF+ Subject: Re: PPC performance decrease after "ROMMING" appl. ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: DOS Fs ?? VxWOrks newbie Date: Sat, 08 Jan 2000 14:15:50 -0500 From: tim keck Message-ID: <38778CE6.2D27@tqci.net> Reply-To: kecks@tqci.net I just got involved with a project using VxWorks 5.4, Tornado 2 running on a Synergy PowerPC board. My only previos VxWorks experience was one project running 5.2 on a FORCE SParc 2CE years ago. The design engineer wants to use DOSFs to record data on a large SCSI hard disk so it can be read back on a PC. I finally came to the conclusion that VxWorks doesn't support the parition table on the hard disk, so I wrote the following code to handle partition table entries, figuring I'd fdisk and format on a PC, and then mount the drive on VxWorks and record data: /* start paste of code file ------------------------------- */ #include "stdio.h" #include "dosFsLib.h" #include "scsiLib.h" /* DosPart.c - DOS FS partition handler */ /* modification history - -------------------- 2000-01-06 First code t keck J.F. Taylor Inc. Copyleft J. F. Taylor Inc. under GNU Public License */ /* DESCRIPTION This file is used to query all available SCSI devices for partition tables, and initialize DOS file systems accordingly. They are named dosIP, wher I = ID number, and P = partition location in table */ #define MINDRIVE 0 /* minimum SCSI device number to poll */ #define MAXDRIVE 7 /* maximum+1 SCSI device number to poll */ #define MAXPART 4 /* maximum number partitions per device */ #define PARTSIZE 1 /* size of partition table in sectors */ #define PARTOFFSET 0 /* offset of first sector of partition table from start of disk, in sectors */ #define VALIDPART 0x55AA /* tag to indicate valid parition, in PC byte order */ /******************************************************************************* * * DosPart - configure SCSI peripherals w/ DOS partitions * * This code configures the SCSI disks whch have valid DOS parition tables, and * assigns DOS Fs structure to each valid partition * */ void DosPart (void) { int drivecount; int partcount; char diskname[8]; SCSI_PHYS_DEV * pSpd; /* SCSI_PHYS_DEV ptr */ BLK_DEV * pSbd; /* SCSI_BLK_DEV ptr */ BLK_DEV * pDbd; /* DOS BLK_DEV ptr*/ struct OnePart { char boot __attribute__ ((packed)); char startCHS[3] __attribute__ ((packed)); char type __attribute__ ((packed)); char endCHS[3] __attribute__ ((packed)); unsigned int offset __attribute__ ((packed)); unsigned int size __attribute__ ((packed)); } ; struct PartionTable { char bootcode[446]; /* the boot code, which we will skip */ struct OnePart partentry[MAXPART]; /* the partition table entries */ unsigned short int tag ; /* the tag as valid parition table */ } parttable; /* * HARD DRIVE CONFIGURATION * * In order to configure the hard disks and initialise dosFs we must find * each valid SCSI device, determine it has a valid partition table, * create block devices for the partitions in the table, * and create the file systems for the devices */ /* now cycle through, looking for valid DOS partitions */ for ( drivecount = MINDRIVE; drivecount < MAXDRIVE; drivecount++) { if ( (pSpd = scsiPhysDevIdGet (pSysScsiCtrl, drivecount, 0)) != NULL ) { if ( (pSbd = scsiBlkDevCreate (pSpd, PARTSIZE, PARTOFFSET)) != NULL) { scsiRdSecs((SCSI_BLK_DEV *)pSbd, PARTOFFSET, PARTSIZE, (char *)&parttable ); if ( parttable.tag == VALIDPART) { for ( partcount=0; partcount < MAXPART; partcount++) { binvert((char *)&parttable.partentry[partcount].offset, 4); binvert((char *)&parttable.partentry[partcount].size, 4); if (parttable.partentry[partcount].size != 0) { if ( (pDbd = scsiBlkDevCreate (pSpd, parttable.partentry[partcount].size, parttable.partentry[partcount].offset) ) != NULL) { sprintf (diskname, "dos%x%x", drivecount, partcount); if( dosFsDevInit (diskname, pDbd, NULL) == NULL ) perror ("dosFsDevInit"); } } } } } } } } /* end paste of code file ----------------- *? BTW, I can't guarantee that this code will actually compile and run, since I took my code and chopped out all the debugging messages to post it here, to save space. But it's close. If anyone is really interested, I'll repost the final version when it works. Anyway, the problem is that I get an "Error 380001" or some such at the perror. It I use a dosMkFs instead of the dos FsDevInit call, I get a device of the correct name in my system device table, but then when I put the disk pack on the PC, it no longer recognizes it, though the partition seems OK under fdisk; so I assume the formatting is FUBAR. The PC is running Win98 (shudder) by the way. So my question is, when they say 'dos file system', just what do they mean? FAT12? FAT16? How can I read & write hard disk partitions that are interchangeable w/ my PC, that is REALLY a DOS file system? I'd appreciate any help anyone has out there. Thanks. Tim Keck Senior Engineer J.F. Taylor Inc. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SNMP m2lib sysUpTime error Date: Sat, 08 Jan 2000 23:53:19 GMT From: pwhicker@notthis.home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <84su8r$5io$1@ndsukns01.ndsuk.com> <84t03e$6h2$1@ndsukns01.ndsuk.com> <858g5d$8kq$1@newsmaster01.magnet.at> For the use of WindView there is generally a timestamp driver, this will give the best resolution of whatever hardware was available to the BSP. This is, necessarily, a BSP issue; if you are on a target with a good timer then you get good resolution (small part of a microsecond on my StrongARM board) but if you have poor hardware or nobody bothered to write the driver for your hardware then you're SOL.I think the macro INCLUDE_TIMESTAMP is used to include the driver, perhaps someone knows the right call to retrieve the current value? Other than that most targets have two clocks implemented, the system clock and the auxiliary clock. The 60Hz rate is an anachronism stemming from the US mains frequency, if you want to change it then use sysClkRateSet(RATE) where RATE is the new number. Bear in mind that if you increase the system clock rate you also increase the proportion of total resources expended on rescheduling and whatever else the system does in the tick. Alternatively you can connect a function to the auxiliary clock then set that to a higher rate, if I remember correctly it goes something like:- sysAuxClkConnect(clockFunction); sysAuxClkRateSet(RATE) There might be an enable call too, I don't remember. This has the advantage that it's a simple call. I hope that is some help, Paul In article <858g5d$8kq$1@newsmaster01.magnet.at>, "Mario Semo" wrote: >Hello, > >is there no 64bit timer available in VxWorks? My application uses 64bit >timers (highlevel) (either with long long ints, or with QuadWord structures) >(and under OS2, NT, and WinCE i get 64bit high resolution timers with >frequencies of up to 1MHZ). >I have not found anything usefull under VxWorks. But i assume i missed >something important. timeget() cant be what i need: > a) it just returns values with 60HZ > b) it overflows too early as each other 32bit timer would overflow. >so, what high resolution (1MHZ), not overflowing timer can i use under >VxWorks? > --------------------------- Newsgroups: comp.os.vxworks Subject: C`A,B.L`E DE-SCRAMBLER INSTRUCTIONS......FREE C`A-B,L.E`................................. 425 Date: Sun, 09 Jan 2000 16:17:55 GMT From: kswgzx@sdfsdf.com Message-ID: LEGAL C`A`B`L`E TV D`E-S`C`R`A`M`B`L`E`R Want to watch Sporting Events?--Movies?--Pay-Per-View??....FREE!!!! *This is the Famous R-O Shack TV D-e-s-c-r-a-m-b-l-e-r You can assemble it from Radio Shack parts for about $12 to $15. We Send You: 1, E-Z To follow Assembly Instructions. 2. E-Z To read Original Drawings. 3. Total Parts List. **** PLUS SOMETHING NEW YOU MUST HAVE! **** Something you can't do without. THE UP-TO-DATE 6 PAGE REPORT: USING A D-E-S-C-R-A-M-B-L-E-R LEGALLY Warning: You should not build a TV D-e-s-c-r-a-m-b-l-e-r without reading this report first. You get the complete 6 page report and instruction package including the instruction plans, the easy to follow diagram, and most important of all the "Using a D`e`s`c`r`a`m`b`l`e`r LEGALLY Report all for just--$10.00 Fill out form below and send it, along with your $10.00 payment to: C.a.b.l.e.t.r.o.n FREE-TV 12187 S. Orange Blossom Trail #116 Orlando Fl 32837 (Cash, Check or Money Order.) (Florida residents include 7% Florida State Sales Tax) (All orders outside the U.S.A. add $5.00) PRINT YOUR: NAME______________________________________________________ ADDRESS___________________________________________________ CITY/STATE/ZIP____________________________________________ E-MAIl ADDRESS____________________________________________ rtg45t34t43t34t34t43t34t34t34t34t34t34t34 nsrciyylcnduonybcsqodytrokjgulnjfjrnbxtzlljiquvdjdjixjkhwefongxnjumrucgvfmjsftzueiclzddbbnlvolmp --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tgtServer Date: 9 Jan 2000 22:47:41 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <85b36d$p86$1@news1.xs4all.nl> References: <8554vk$jod$1@newsmaster01.magnet.at> <85587m$mn7$2@news1.xs4all.nl> <855anp$pki$1@newsmaster01.magnet.at> Mario Semo wrote: : got it! : thx a lot. : is there a way to turn this feature off? in which cases does it make sense : to run a command on the host, when there is a real target connected? : should i add the @ to ALL commands? There are some commands that are quicker using the host. The target- server keeps a local symbol list in memory, so when you access the symbol table there is no communication with the target, but it is handled locally on the host. This is the case for symbol table access, for directories (using ls) and also for the help command. : e.g. now i find the modules, and i can locate an entrypoint offtst with : @lkup, but i cannot start it: I think you should enable the Symbol table synchronisation in the configuration menu of the target server. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem elf'ing vxWorks_rom for PPC Date: Mon, 10 Jan 2000 02:03:51 GMT From: paulvroberts@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <85bem0$72s$1@nnrp1.deja.com> References: <200001051957.OAA25070@TIGER.rst.com> In article <200001051957.OAA25070@TIGER.rst.com>, mdelgior@TIGER.rst.com (Mark Del Giorno) wrote: > Heyyyyyyyyyyyyyy, > > I'm trying to burn my application into the mv2700 flash so it will run on > power up. > I added the patch to allow > 1 MB flash applications to run (untested). Which patch is that out of interest? I haven't done this yet, but I might need it soon... > I've added my application to MACH_EXTRA. > I've edited usrAppInit.c to call it. > I've confirmed that if I make vxWorks and have my board boot, my > application runs. > I changed my project to make vxWorks_rom. > > but: > When I run elfToBin on it, I get the following output: > > > elfToBin < vxWorks_rom > test.bin > seg 1: Expected load address 0xfff00100 but file address is 0x00586a88 > > test.bin is still generated (and looks to be about the right size). > Has anyone seen this? > Is this an error or just a warning that can be ignored? > (I know the code will be copied to RAM, then run....???) > > Thanks in advance for any help! > > - Mark > If this is Tornado 1.0.1, then when I burnt the flash, it didn't work. Ask WRS, there is a patch available (SPR 8845, I think). Paul Roberts. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CPU util measurement? Date: Mon, 10 Jan 2000 08:08:27 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <3879775B.D33052ED@bitband.com> References: <3874C795.732E9B0C@email.mot.com> <38782C28.E11BD702@bitband.com> <3878ac4d.367860@news.swbell.net> Reply-To: leonid@bitband.com Bruce wrote: [snip] > > If you need something like this, it is trivial to create a task with a > priority of one greater than the idle task and put a counter in it. > > Bruce in Katy Texas Bruce, VxWorks does not have an IDLE task! Instead it has got a special kernel state for "Idle" so that when the first task becomes ready, it only has to do "half" a context switch. Still, to measure CPU utilization is possible with a task at the lowest priority 255, which needs to first calibrate itself to measure how fast is the CPU and then measure and report it's relative speed and thus availability. I have done this several times, and it is not trivial at times, because of all those caches changing the CPU performance in the most peculiar ways. Take care, Leonid --------------------------- Newsgroups: comp.os.vxworks Subject: Using Tornado II and SNiFF+ Date: Mon, 10 Jan 2000 17:26:20 +0100 From: "Martin Raabe" Organization: TakeFive Message-ID: <85d11k$rv1$1@news05.btx.dtag.de> Reply-To: "Martin Raabe" Hello, I wonder if there are users of Tornado II using SNiFF+ SourceCode Navigation and Management Tool. I am interrested in hearing from any experience! - -- Martin Raabe --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPC performance decrease after "ROMMING" appl. Date: Tue, 11 Jan 2000 08:43:55 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <387AD12B.73E27A11@bitband.com> References: <387A00AC.C9CF6C67@nlr.nl> Reply-To: leonid@bitband.com Frankly, it would be a surprise if the results where different. ROM is allways slower then DRAM, because its data bus is typically narrow while DRAM data bus is typically full 32 bit, and ROM does not have any burst or fast-page capability which is used with DRAMs to fill cache lines. I am not intimately familiar with the particula rboard, but I do know that PowerPC is particularly sensitive to Cache and memory performance. Leonid Bert Pleijsier wrote: > > Hi, all, > > I (still) use VxWorks 5.3.1 and Tornado 1.0.1, > and have a MVME2305 PPC board (with VxWorks BSP). > A simple testprog writes a fixed number of times to DRAM on PPC board. > > Two cases: > 1. Bootrom in (flash)ROM, > VxWorks download in RAM over ethernet + testprog execution via shell > in RAM. > 2. VxWorks + testprog in (flash)ROM (not ROM resident, but copied to > RAM) > using the "vxWorks_rom" making (..\h\make\rules.bsp) as template. > > Testprog executes slower in case 2 than in case 1. > --------------------------------------------------- > > Does anyone have an idea about this ??? > Thanks in advance > > Bert Pleijsier --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jan 11 08:42:11 2000 From: "Robert L. Perry" Date: Tue Jan 11 08:42:13 PST 2000 Subject: RE: comp.os.vxworks newsdigest vxWorks / Tornado Tim, There is a new dosFs available for Tornado that you should check out. This new version pretty much supports everything you would expect. When you create the file system, it will use whatever FAT type is appropriate for the disk size. The code is available via FTP or a new CD, however, you'll still need an install key -check w/ your local WRS FAE. Good Luck, Bob Perry Tim Keck Wrote: I finally came to the conclusion that VxWorks doesn't support the parition table on the hard disk, so I wrote the following code to handle partition table entries, figuring I'd fdisk and format on a PC, and then mount the drive on VxWorks and record data: ... So my question is, when they say 'dos file system', just what do they mean? FAT12? FAT16? How can I read & write hard disk partitions that are interchangeable w/ my PC, that is REALLY a DOS file system? I'd appreciate any help anyone has out there. Thanks. Tim Keck Senior Engineer J.F. Taylor Inc. From vxwexplo-errs@csg.lbl.gov Tue Jan 11 08:45:43 2000 From: enrico.ferro@it.abb.com Date: Tue Jan 11 08:45:45 PST 2000 Subject: DHCP client Hi guys, is there anybody here who has developed a DHCP client on a VxWorks target? I am trying to make one, but I have some trouble in getting my client to start. Thanks in advance, Enrico From vxwexplo-errs@csg.lbl.gov Tue Jan 11 09:09:01 2000 From: Andrew Booth Date: Tue Jan 11 09:09:03 PST 2000 Subject: anyone used VxWorks NTP? Hi, I'm using T2 and 5.4 on a PPC MVME2307. I need to set the time on some VME h/w to the correct absolute time to better than 0.1s accuracy. I see VxWorks provides an NTP facility with sntpcLib, and I'd like to use this as I'm familiar with obtaining the time for UNIX systems with NTP. Has anyone used the VxWorks version? I would be interested in any information on timing reliability. I'm a bit suspicious that WR calls their implementation a "simple" NTP implementation. There doesnt seem to be any info in the manuals about how much of the NTP functionality WR has provided, nor how accurate the timing can be expected to be. Thanks, Andrew Booth From vxwexplo-errs@csg.lbl.gov Tue Jan 11 18:39:42 2000 From: Nagaraj G Date: Tue Jan 11 18:39:44 PST 2000 Subject: VxWorks: VxLoad.com problem Hi, I want to load VxWorks from MS DOS 6.22 using the "vxload.com" utility supplied with Tornado II kit. I installed MS DOS 6.22 freshly and tried running vxload.com. I got the error message "Insufficient memory to load image file..". Then I referred to Appendix D and made the recommended modifications to CONFIG.SYS file. Then I got the error "Not Enough Memory to Load File". Has anyone tried this option? I am trying on a Pentium 120 system. Thanks in advance G. Nagaraj Senior Software Engineer EnThink Inc. 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 211 Fax: +91-80-557-6032 www.EnThink.com From vxwexplo-errs@csg.lbl.gov Tue Jan 11 18:39:51 2000 From: Nagaraj G Date: Tue Jan 11 18:39:53 PST 2000 Subject: VxWorks: aic7880 library Hi, I want to use a scsi card with a HDD/printer. I was in search of the SCSI cards supported by VxWorks (I mean with built-in drivers). I found the aic7880 library for Adaptec's AIC 7880 based controllers. I would like to know which PCI card should I buy for using this library. Thanks in advance. G. Nagaraj Senior Software Engineer EnThink Inc. 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 211 Fax: +91-80-557-6032 www.EnThink.com From vxwexplo-errs@csg.lbl.gov Wed Jan 12 01:40:11 2000 From: David Laight Date: Wed Jan 12 01:40:14 PST 2000 Subject: Re: DHCP client The DHCP client code shipped with T1 and T2 is somewhat buggy. I did get it working - but only after raising so many TSRs that WRS let me have the source. The biggest stumbling block is likely to be that the client code rejects all responses from servers that conform to the later RFC (I can't remember the numbers of hand) because it requires that the packet length always be padded to the size of a BOOTP message. I had the additional requirement to use DHCP for a pcmcia network card that might be inserted/removed while the system is running. This exposed the limited testing that WRS had done on the dhcp code. I was told that WRS are implementing a new version of DHCP that conforms to the latest RFCs. I don't know if/when this will ship. WRS support (europe) have my fixed code, I don't know if they have done anything with it... David > Date: Tue, 11 Jan 2000 08:55:55 -0800 (PST) > Subject: DHCP client > X-UIDL: 92d6c197a50d8c1e26acf0f19fd7e7b1 > > Submitted-by: enrico.ferro@it.abb.com > > is there anybody here who has developed a DHCP client on a VxWorks target? > I am trying to make one, but I have some trouble in getting my client to > start. > > Enrico ---------------------------------------------------------------- David Laight email: dsl@tadpole.co.uk Tadpole Technology plc phone: +44 1223 428 232 Cambridge, UK fax: +44 1223 428 201 From vxwexplo-errs@csg.lbl.gov Wed Jan 12 04:03:58 2000 From: Vxworks Exploder Date: Wed Jan 12 04:04:01 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jan 12 04:03:56 PST 2000 Subject: taskInit pb Subject: Tornado : header files? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: taskInit pb Date: Wed, 12 Jan 2000 11:48:50 +0100 From: "Pascal LEJEUNE" Organization: Matra Grolier Network Message-ID: <85hm90$fip$1@news.mgn.net> Hello, I don't understand how to use taskInit. Does anyone has a small example ? Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado : header files? Date: Wed, 12 Jan 2000 08:41:14 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <85hq2v$9d1$1@newsmaster01.magnet.at> Hello, this is definitly a simple question, but i have too low experience and cant find a solution: primary i develop and compile my sources with the compilers / asm / binder etc. from Tornado, but outside of Tornado. i prefer to use my own environment. sometime i try to create a WSP / Project. sometime means, till now i just tried very very simple projects, as in the getting started. just a single .c / .cpp file. today i created a project, then i copied all .cpp and .hpp files in the project directory. ADDED (within) tornado the .cpp files and tried rebuild all. error: dont know how to make xxx.hpp. (this is the first .hpp file). ok. i tried to add the .hpp files, i tried to rename them to lowercase letters, rename them to .h. whatever i do: dont know how to make xxx.hpp. the dependencies of xxx.cpp shows all 5 .hpp files and they are listed in EXTERNAL depenencies. and when i double click on one of the external dependencies files, it is opened in my editor. i have no idea what i can do to make Tornado happy. what i am doing wrong? - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jan 12 06:42:17 2000 From: "Michael Lawnick" Date: Wed Jan 12 06:42:19 PST 2000 Subject: Re: Tornado : header files? > Hello, > > this is definitly a simple question, but i have too low experience and > cant find a solution: > > primary i develop and compile my sources with the compilers / asm / binder > etc. from Tornado, but outside of Tornado. i prefer to use my own > environment. > > sometime i try to create a WSP / Project. sometime means, till now i just > tried very very simple projects, as in the getting started. just a single > .c / .cpp file. > > today i created a project, then i copied all .cpp and .hpp files in the > project directory. ADDED (within) tornado the .cpp files and tried rebuild > all. > > error: dont know how to make xxx.hpp. (this is the first .hpp file). > > ok. i tried to add the .hpp files, i tried to rename them to lowercase > letters, rename them to .h. > > whatever i do: dont know how to make xxx.hpp. Hi Mario, when you add a file to your project it isn't copied. The TCL-script that make's your dependency file is a bit stupid and does worry about files that are outside the projects dir. So: Add VPATH=xxx:yyy:zzz to Macro folder and -Ixxx -Iyyy -Izzz to the c/c++ compiler parameters folder in your build properties, with xxx,yyy,zzz as an example for three external directories. HTH MfG Lawnick, SOFTEC GmbH ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== From vxwexplo-errs@csg.lbl.gov Wed Jan 12 08:21:02 2000 From: stephen.thompson@tde.alstom.com Date: Wed Jan 12 08:21:04 PST 2000 Subject: Re: Tornado : header files? vxwexplo@lbl.gov (the vxWorks Users Group Exploder) on 12/01/2000 14:52:01 To: vxworks_users@csg.lbl.gov cc: (bcc: Stephen THOMPSON/PCS/PCG/TDE/GECALSTHOM) Subject: Re: Tornado : header files? Submitted-by vxwexplo-errs@csg.lbl.gov Wed Jan 12 06:42:17 2000 Submitted-by: "Michael Lawnick" > Hello, > > this is definitly a simple question, but i have too low experience and > cant find a solution: > > primary i develop and compile my sources with the compilers / asm / binder > etc. from Tornado, but outside of Tornado. i prefer to use my own > environment. > > sometime i try to create a WSP / Project. sometime means, till now i just > tried very very simple projects, as in the getting started. just a single > .c / .cpp file. > > today i created a project, then i copied all .cpp and .hpp files in the > project directory. ADDED (within) tornado the .cpp files and tried rebuild > all. > > error: dont know how to make xxx.hpp. (this is the first .hpp file). > > ok. i tried to add the .hpp files, i tried to rename them to lowercase > letters, rename them to .h. > > whatever i do: dont know how to make xxx.hpp. Hi Mario, This is a problem with include search paths and dependancies. Say your .cpp files all include a file called fred.hpp. Tornado will have worked out that fred.hpp is a dependancy of all the .cpp files (i.e. it must be present in order to compile the .cpp files). But if it doesn't know where to find fred.hpp it will assume that it must make it first, but it doesn't know how to. What you really need to do is tell Tornado where to find fred.hpp (and any other .hpp files) using the Include search path specifiers for the Compiler. To do this click on the Builds tab at the bottom of the workspaces window, then select your project in the workspace list and then click on the + sign which will open up the list of build environments for your project. You will probably only have one so select it then from the main toolbar at the top select View-> Properties. This will bring up the buld properties window for your project. Select the C/C++ compiler tag and you will see the compile line used to compile your source files. This will probably be something like -g -w -O0 -mcpu=860 -ansi -nostdinc -fvolatile -fno-builtin you will need to add to this some include file search path specifiers, something like -IC:\YourProject\Includes You will need one of these for each seperate directory where you .hpp files live. Once you've added these then press the Apply button on the properties window. Then from the main Tornado tool bar select Build->Dependencies and start off a rescan of the project dependencies. Hopefully Tornado will now find where all your .hpp files live so that when you do a build it can find them. Happy Building Steve Thompson Alstom England. From vxwexplo-errs@csg.lbl.gov Wed Jan 12 08:51:52 2000 From: enrico.ferro@it.abb.com Date: Wed Jan 12 08:51:54 PST 2000 Subject: still on the DHCP server Hello, I have a DHCP client running vxWorks that receives the IP address from a server that runs on an NT machine. The address is assigned, but when the dhcpConfigSet() routine returns, the Ead value is: Ead : 10.39.89.6:fffffc00:d2f00:a the first two items are OK (IP address and mask) but I can't understand the other two, that in my opinion should not be there; besides, I can't connect the Target Server to my target (I get a time-out) even if the target is alive (I can ping it). Does anybody out there have any idea about this issue? Thanks in advance Enrico From vxwexplo-errs@csg.lbl.gov Thu Jan 13 04:03:14 2000 From: Vxworks Exploder Date: Thu Jan 13 04:03:17 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jan 13 04:03:12 PST 2000 Subject: Re: DHCP client Subject: Re: TELNET CLIENT Subject: Re: VxWorks: VxLoad.com problem Subject: Re: Tornado : header files? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: DHCP client Date: Wed, 12 Jan 2000 10:41:09 -0500 From: Tim Sohacki Organization: Nortel Networks Message-ID: <387CA095.F63287AC@americasm01.nt.com> References: <200001120940.JAA22877@dsl-2.tadpole.co.uk> Have you encountered any failure to rebind problems? If you set time of day (beyond the default January 1970), T1 and T2 are expired. This is documented on WindSurf. So then instead of a renewal, the next lease action is a rebind. Often, this rebind fails on my target, and fails repeatedly, though sometimes it works. Different (application linked) loads act differently, even when the BSP stays the same. In retaliation, I set the lease to infinite ... but T1 and T2 (with dhcpcTimersShow()) are displayed negative when an all F's lease length is used. Whoops ... is this a print error only or does the target not really understand an infinite lease? I decided to use a long lease that is still positive, something on the order of a few years. (Yeah, yeah, yeah, a 2003 problem.) Cheers, Tim. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TELNET CLIENT Date: Wed, 12 Jan 2000 11:57:55 -0500 From: Yibing Fan Organization: UUNET Canada News Reader Service Message-ID: <387CB293.890F20C3@lynx.tsb.ca> References: <853j56$rtn$1@nnrp1.deja.com> Hi vxworkers, I need telnet client too. I believe this is not a first time raised question. I thought the port is in archive somewhere but I searched vxworks archives and could not found it. Anyone know where I can get it? either commertial or freeware, we willing to pay. Many Thanks. tornado2000@my-deja.com wrote: > I am in need of a telnet client for vxWorks. Anybody has tried porting > BSD telnet client to vxworks ? I need help in porting it to vxworks. > The vxworks tty driver doesn't seem to support many of the unix like > features ? Anybody has tried porting BSD code to vxworks ? > > Sent via Deja.com http://www.deja.com/ > Before you buy. - -- Yibing Fan, Software Engineer TSB International Inc. 5399 Eglinton Ave. West, Suite 115 Etobicoke, Ontario, Canada M9C 5K6 Tel:(416)622-7010 x501 Fax:(416)622-3540 email:yibing_fan@email.tsb.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks: VxLoad.com problem Date: Thu, 13 Jan 2000 09:06:16 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <387D7968.B177CEA5@bitband.com> References: <85grkf$i3j11@overload.lbl.gov> Reply-To: leonid@bitband.com The vxWorks image by default expects to start when the processor is already in "protected mode", while "bootrom_uncmp" expects to be started in "real mode" and thus contain code to switch into protected mode first thing it does. It is also possible to build the vxWork simage with the real mode initial condition, by building vxWorks_rom_low instead, and make sure it does fit the 640KB less its starting address of 0x8000. Its been years since I tried this myself, but I hope this does give a general sense of direction ... Leonid Nagaraj G wrote: > > Hi, > > I want to load VxWorks from MS DOS 6.22 using the "vxload.com" utility > supplied with Tornado II kit. > > I installed MS DOS 6.22 freshly and tried running vxload.com. I got the > error message "Insufficient memory to load image file..". Then I > referred to Appendix D and made the recommended modifications to > CONFIG.SYS file. Then I got the error "Not Enough Memory to Load File". > > Has anyone tried this option? I am trying on a Pentium 120 system. > > Thanks in advance > > G. Nagaraj > Senior Software Engineer > EnThink Inc. > 37, Castle Street, Ashok Nagar, > Bangalore - 560 025, India > Phone: +91-80-536-7297 Ext: 211 > Fax: +91-80-557-6032 > www.EnThink.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado : header files? Date: Thu, 13 Jan 2000 08:50:18 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <85k0qv$qh0$1@newsmaster01.magnet.at> References: <85icrj$i3k7@overload.lbl.gov> Hello, as i wrote i already copied (on a command line) the .cpp and the .hpp file into the project dir. and tornado already have -I. as compiler search path. and the error is not a compiler, but a "make" error (dont know how to make), so, it cant be fixed by added my include pathes to the compilers command line (since the compiler is in fact never called. i can even remove the compiler from the harddisk) and Tornado knows about the source file. it shows the "fred.hpp" in the list of external dependencies and i can double click on it and it is opened. anyway, i tried to change the compile options, and i am not able to do this 1st try: selectr Build page in notebook, simmntgnu, C/C++ compiler. here i see in a TextBox a currect compile flags. but i cannot edit here. And here the apply button is greyed. 2nd try : you suggested MainToolbar/tools/properties. but in the tools pull-down there is no properties. there is a "options". but here i cannot see any way to change compiler command line options. i have created a mini wsp (zipped, with absolute pathes : \vxw\.... i had installed it on n:\ ). when i try to compile, here is the log: vxrm *.o *.rpo ctdt.c symTbl.c vxApp* vxrm n:\vxw\WithHpp\prjComps.h n:\vxw\WithHpp\prjParams.h n:\vxw\WithHpp\prjConfig.c n:\vx w\WithHpp\linkSyms.c vxrm n:\vxw\WithHpp\libs.nm n:\vxw\WithHpp\libs.size make: *** No rule to make target `foo.hpp', needed by `bar.o'. Stop. Done. - -- regards, Mario Semo. http://www.kirchnersoft.com stephen.thompson@tde.alstom.com wrote in message <85icrj$i3k7@overload.lbl.gov>... > >What you really need to do is tell Tornado where to find fred.hpp (and any other >.hpp files) >using the Include search path specifiers for the Compiler. To do this click on >the Builds tab >at the bottom of the workspaces window, then select your project in the >workspace list > >-IC:\YourProject\Includes > >You will need one of these for each seperate directory where you .hpp files >live. >Once you've added these then press the Apply button on the properties window. >Then from the main Tornado tool bar select Build->Dependencies and start off >a rescan of the project dependencies. Hopefully Tornado will now find where all >your .hpp files live so that when you do a build it can find them. > begin 666 WithHpp.zip M4$L#! H``````,9$+2@````````````````,`&,`=GAW+U=I=&A(<' O4T1. M`' `````" #E5T1[8V1@:1!A8&!088 `!R!F9 (S616 A *0S<@*X8L"B8DZ M'D'*T=LE>AE=JD'JF!@,&"#*11C^,\HS,#*".1"2&U,,`%54#0`'Y(!]..2 M?3AF@'TX4$L#!!0````(`(1$+2BXTY_8? ```*0````/`& `=GAW+U=I=&A( M<' N=W-P4T1+`%P`````" !05(&S8V1@:1!A8&!088 `!R!F9 (S616 A *0 MS<@*X8L"B8DZ'D'*T=LE>AE=JD'JF!AD&!C!TB(,_QGE&1@A'# )`%54#0`' M9X!].&> ?3AG@'TX<\E/+LU-S2M12,O,257057#Q5_#S#U%P=?$,X>*R<7)U M]_2S4W#V#W*-]_1S\X\/B0QPY0K/+\HN+DA,3N6R*1D"05[R+9Y"F?GAF289'08%^.836*R_(@ND$ M`%!+`P04````" "W1"TH=ACWS4P```!5````$P!@`'9X=R]7:71H2'!P+V)A M0J8&&08&,'2(@S_&>49&"$<, D`550-``?* M@'TXRH!].,J ?3A3SLQ+SBE-256P20J8&&08&,'2(@S_&>49&"$<, D`550-``?D@'TXY(!].&> ?3C=5>XV-&L(Q_D9P6!+:$$1$W)(/D!9H=@4KP1Y(V4,8LWA*!1@BAT0A2 M+@A0EG-1Q@WE#"$_>, +Q[5!73/8TV9W7U73??78.RTG@%C8&\VK^$+Z:DUA,.9JO'=?"H6^;@SETEEZT92V*[J+S.E?CC>-9 M>&Z$]D1O8K$E#4)JA*2E1::EG.5TVXI^BGXBA&A.?L/X:GRL,_E^*#%!IK^& MLVNHCZ+5RKUP=#U9]L)8NQ$.U@,X1VAXVR#",IK+FBPMVHS(AH<9)OI.+R4M M>D;R>CK0\*7&C0`N&X]%3/=;6ZN>RL^"SU]4^Z#M:P(:VA;@A:SFH+&>-UD$CG0ML-' M01/0!;5C2@@/;]J+ Z-IS MIO)>W.H1%RS.^+!Q])T,D^S/%-UJ>GR$;O9/&O1!>X:XKDF9%-+6;14MK:K/ M-N7['[ %FJT63Y%KO:&TR!2EKG7!J1Q*J]ND?JD;4L[VE&5\7\LP[!M!Y!@N M5O$R3"!O^69=5@[%MVCE1\[*PU(.CF?CI6$&JR[-.C@N,P]/$@[R&@%_(D)0J;!.5" !IXQVTJF! ML^(%$E+PO3SQ: T%901](B5.9,!)RL<2%FU!3C3\)17W1^ M^G8U-LW)_Z2W M=X[J\\T-6BJQ'7XE$[T#0@KCP'"':49DPQEA*24U&I!Z)P5RSJ<[F=HGM341 MBESYU^N90N@O4$L#! H``````,9$+2@````````````````5`&,`=GAW+U=I M=&A(<' O4TE-3E1G;G4O4T1.`' `````" #E5T1[8V1@:1!A8&!088 `!R!F M9 (S616 A *0S<@*X8L"B8DZ'D'*T=LE>AE=JD'JF!@,&"#*11C^,\HS,#*" M.1"2&U,,`%54#0`'Y(!]..2 ?3CD@'TX4$L#! H``````,1$+2@````````` M```````2`& `=GAW+U=I=&A(<' OAE=JD'JF!AD M&!C!TB(,_QGE&1@A'# )`%54#0`'X(!].." ?3C@@'TX4$L#!!0````(`,9$ M+2C-M*&16 (``&P'```7`& `=GAW+U=I=&A(<' O=VET:$AP<"YW<&I31$L` M7 `````(`%!4@;-C9&!I$&%@8%!A@ `'(&9D`C-9%8"$`I#-R KABP*)B3H> M0J8&&08&,'2(@S_&>49&"$<, D`550-``?D@'TXY(!].&: M?3BU54MOHS 0/J]_!8<>6E6$57NKVDH4.XTK`LB81KM:R2+@)*[ 1IALTUWU MOZ\ACZW21B6'7F >WS?VC&=LJ+)ER65CS43!+=N"H16$U$(04P"N[] ]#FZM MNP3[D,5X'-"Y7+*QZY&0N02DM19EE8%K%,#;3^'>"#\B<'(:D0<&,3ESM@CG M632+454-TGZ18I!E_1?VO./@0]^]CX&=,#;!04"M7^";G1CQ\F(M;NR,[3Q; M$&-C'-Q/+B^V/K9/VU-VN#>^LC*G(99EIZ12BTZ02C>YD%FGS/^;"C'ME)E4 M]G0IBD;(G9[S&:_M2E6=99(612= ,F'CQ*>8T1%!+EP;&4$HH,3=Y(('Z]_) MJ=D99'=NC,Z<)JWGO'$6:X87)3==[8XH:WMV7Y9AU'U7[3?5FI?3PCC;UK*S MJOJBK*)HVUR6C2RS!7O5K\U\"(J\?U?Z<-.6MEY.]8MN>'GS+&2NGG5//HM< M0K'KLTV@NA92U(6+P/L8^?JEY[:FR?4';090- MKZV ?3A02P$"%@L4```` M" "W1"TH=ACWS4P```!5````$P`1```````!`" `MH&6`0``=GAW+U=I=&A( M<' O8F%R+F-P<%-$! !<````550%``?*@'TX4$L!`A8+"@``````G40M*#\G M-:L4````% ```!,`$0```````0`@`+:!' Date: Thu Jan 13 10:16:34 PST 2000 Subject: Setting RAM_HIGH_ADRS above 0x02000000 (VxWorks) Hi - Has anyone successfully set RAM_HIGH_ADRS to 0x02000000 or higher? I am working on a PPC8260 and whenever I set RAM_HIGH_ADRS to 0x02000000 or higher, the PC gets whacked at various times. I have successfully set it to 0x300000, 0x800000 and 0x01A00000. Thanks! Bob Ferrara bferrara@unispheresolutions.com vxworks/Tornado From vxwexplo-errs@csg.lbl.gov Thu Jan 13 23:14:43 2000 From: Roger Holmquist Date: Thu Jan 13 23:14:46 PST 2000 Subject: Re: Setting RAM_HIGH_ADRS above 0x02000000 (VxWorks) >Has anyone successfully set RAM_HIGH_ADRS to 0x02000000 or higher? > >I am working on a PPC8260 and whenever I set RAM_HIGH_ADRS to 0x02000000 >or higher, the PC gets whacked at various times. > >I have successfully set it to 0x300000, 0x800000 and 0x01A00000. I remember that there is a (or have been a) problem using more than 32 MB (0x02000000) local RAM on a MVME2700-card. As far as I can remember WindRiver have solved that problem. Maybe it has something to do with that (former?) bug? Greetings / Roger - # --------------------------------------------------------# # Roger Holmquist, FOA, Defence Research Establishment # # sect 750, Olaus Magnus vag 1, Box 1165 ----------------# # ------------- 581 11 LINKOPING SWEDEN ------------------# # Phone: +46 13-378382 -----------------------------------# # Fax: +46 13-378009 -------------------------------------# # Home: +46-142-20542 Mobile +46-706-250123 -------------# # Email: roghol@lin.foa.se -------------------------------# # Home address: Hovslagargrand 5, 590 18 MANTORP ---------# # --------------------------------------------------------# | This message is MIME-coded in order to avoid | | problems with national characters. MIME coding | | should be enabled when replying in Swedish. | xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx From vxwexplo-errs@csg.lbl.gov Fri Jan 14 03:46:22 2000 From: Reyhan ERGUN Date: Fri Jan 14 03:46:25 PST 2000 Subject: how can we debug by using ROM based vxWorks (st_rom or res_rom)? This is a multi-part message in MIME format. --------------641B3652DED091D497D22CC2 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit After a long trial and got new elfToBin.exe from WindRiver we have managed to build both vxWorks.res_rom_res_low and it worked. Now, we have problems with debugging, we could debug neither with vxWorks.st_rom nor with vxWorks.res_rom_res_low. As opposite isn't stated we assume that we can debug by using vxWorks.res_rom_res_low - we are using this built at that moment. We are using Tornado II (WinNT), configuring target server by using core file vxWorks.res_rom_res_low. Shell did not complain about the debug commands such as "b", "c", "s", but never obey this commands. For example, from the shell which we create from Tornado II ->b
->sp program did not break at that address, and no error messages were display. While using vxWorks.st_rom version we have faced with the same situation. Please let me know how can we debug by using ROM based vxWorks (st_rom or res_rom). We are using MBX860 boards. Best regards, Reyhan --------------641B3652DED091D497D22CC2 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Reyhan (FINDIK) ERGUN Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Reyhan (FINDIK) ERGUN n: ;Reyhan (FINDIK) ERGUN org: ASELSAN INC. adr: Microwave and System Technologies Division ;;Software Engineering Dept.;ANKARA;;;TURKEY email;internet: findik@mst.aselsan.com.tr title: Principal Engineer tel;work: (90-312) 385 19 00 (Ext. 2668) tel;fax: (90-312) 354 52 05 note: http://www.aselsan.com.tr x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------641B3652DED091D497D22CC2-- From vxwexplo-errs@csg.lbl.gov Fri Jan 14 06:22:57 2000 From: David Laight Date: Fri Jan 14 06:22:59 PST 2000 Subject: Re: how can we debug by using ROM based vxWorks (st_rom or res_rom)? The short answer is no. In order to be (largly) platform independant, the vxworks breakpoint/single step work by modifying the code to contain a trap instruction - so cannot breakpoint rom resident code. Breakpointing (and single stepping) rom resident code would involve using facilities provided by the cpu and/or system hardware, or using a rom emulater. Other options exist - eg using the mmu to place a page of RAM at the virtual address where the breakpoint is... David > Date: Fri, 14 Jan 2000 03:52:00 -0800 (PST) > Subject: how can we debug by using ROM based vxWorks (st_rom or res_rom)? > Submitted-by: Reyhan ERGUN > > After a long trial and got new elfToBin.exe from WindRiver > we have managed to build both vxWorks.res_rom_res_low and it worked. > Now, we have problems with debugging, we could debug neither with > vxWorks.st_rom nor with vxWorks.res_rom_res_low. > As opposite isn't stated we assume that we can debug by using > vxWorks.res_rom_res_low - we are using this built at that moment. > We are using Tornado II (WinNT), configuring target server by using core > file vxWorks.res_rom_res_low. > Shell did not complain about the debug commands such as "b", "c", "s", > but never obey this commands. > For example, from the shell which we create from Tornado II > ->b
> ->sp > program did not break at that address, and no error messages were > display. While using vxWorks.st_rom version we have faced with the same > situation. > > Please let me know how can we debug by using ROM based vxWorks (st_rom > or res_rom). > > We are using MBX860 boards. > > Best regards, > Reyhan > ---------------------------------------------------------------- David Laight email: dsl@tadpole.co.uk Tadpole Technology plc phone: +44 1223 428 232 Cambridge, UK fax: +44 1223 428 201 From vxwexplo-errs@csg.lbl.gov Sat Jan 15 04:03:16 2000 From: Vxworks Exploder Date: Sat Jan 15 04:03:19 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jan 15 04:03:13 PST 2000 Subject: Cross compiling linux->powerpc-vxworks Subject: Re: Tornado v2 - first bootable VxWorks image Subject: Re: tffsBootImagePut failing, TFFS related? Subject: Re: Tornado v2 - first bootable VxWorks image Subject: Safe Configuration with VxWorks and NT? Subject: Re: VxWorks: VxLoad.com problem ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Cross compiling linux->powerpc-vxworks Date: Thu, 13 Jan 2000 13:44:08 +0000 From: Adrian Cox Message-ID: <20000113134411.2EE73CBC.NFL@trout.agsite2.demon.co.uk> Reply-To: Adrian Cox I'm trying to build object modules on Linux/x86 for a VxWorks/PPC target. My existing powerpc-eabi compiler generates object files that can be loaded by VxWorks without problems. If I link these modules into an application under the Tornado IDE under NT (by adding them to the PRJ_LIBS macro) I get errors like this when I attempt to run under the debugger: `F:/Tornado/test/PPC604gnu/test.out' has changed; re-reading symbols. Premature end of file reading symbol table Has anybody managed to link an archive generated under a Linux hosted cross-compiler into an application under Tornado 2? If so, what compiler version and configuration options did you use? - - Adrian Cox, AG Electronics --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado v2 - first bootable VxWorks image Date: Fri, 14 Jan 2000 17:31:01 +0300 From: "Arthur V. Grigorian" Organization: Auriga, Inc. Message-ID: <387F3325.A7711EE9@auriga.ru> References: <85lr0c$de3$1@soap.pipex.net> Reply-To: archy@auriga.ru Hello, > This is my first Tornado / VxWorks project so apologies for the lengthy > posting ... > > I've installed Tornado v2 on NT 4, have added a new BSP by creating a new > directory \tornado\target\config\newbsp, and have copied all BSP source > files to that location. > > Now I'm trying to create a basic bootable image with VxWorks and the BSP by > using the Tornado project wizard. The wizard seems to create the project > fine and auto-generates a number of files in \tornado\target\proj\newproj. > > The problem is that the project does not build correctly as below. > > Firstly, romStart.c fails to compile with: > > \tornado\target\config\comps\src\romStart.c: In function `bootClear': > \tornado\target\config\comps\src\romStart.c:424: `RAM_DATA_ADRS' undeclared > (first use this function) > > It's conceivable that this (and perhaps other) symbols need to be manually > defined though I would hate to guess at 'reasonable' values. > > Other targets fail to build with numerous compilation errors, the first of > which is: > > \tornado\target\proj\Blah\default\D:\tornado\target\config\comps\src\usrWind > view.c:84: undefined reference to `wvLibInit2' > > I can see a function prototype for wvLibInit2 in \tornado\target\h\wvLib.h > and I assume that the auto-generated code has, for some reason, not > generated the correct list of dependent header files. > > Rather than battle to correct these compilation problems, I am assuming that > I have made some elementary mistake. Perhaps I have: > > a) integrated the new BSP into Tornado v2 incorrectly > b) misused the project wizard > c) failed to read some very significant document regarding building > bootable images > d) done something else incorrectly. I think this problem is caused by (a). AFAIK the integration of a new BSP process is not just cp -r ${NEW_BSP_DIR} ${WIND_BASE}/target/config/ Actually we were always receiving new BSPs for NT as InstallShield images, so I can't help you with some instructions but I think you've got something like an install script or some installation notes in your BSP directory. Regards, __ R2 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tffsBootImagePut failing, TFFS related? Date: Fri, 14 Jan 2000 09:58:07 -0500 From: Burt Bicksler Organization: BBB Message-ID: References: Answering myself.. Verified that the erase failure was caused by having TFFS active, even though one wouldn't expect that to be the case since the bootrom area is supposed to be blocked from access by TFFS. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado v2 - first bootable VxWorks image Date: Fri, 14 Jan 2000 11:15:34 -0500 From: "JohnDoc" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <85ni1v$h4i$1@soap.pipex.net> References: <85lr0c$de3$1@soap.pipex.net> <387F3325.A7711EE9@auriga.ru> Arthur V. Grigorian wrote in message news:387F3325.A7711EE9@auriga.ru... > Hello, > > > [cut] > > > > Rather than battle to correct these compilation problems, I am assuming that > > I have made some elementary mistake. Perhaps I have: > > > > a) integrated the new BSP into Tornado v2 incorrectly > > b) misused the project wizard > > c) failed to read some very significant document regarding building > > bootable images > > d) done something else incorrectly. > > I think this problem is caused by (a). AFAIK the integration of a new BSP > process is not just cp -r ${NEW_BSP_DIR} ${WIND_BASE}/target/config/ > Actually we were always receiving new BSPs for NT as InstallShield images, > so I can't help you with some instructions but I think you've got > something like an install script or some installation notes in your BSP > directory. > > Regards, > __ > R2 Thanks for the input Arthur. I shall look in to it. --------------------------- Newsgroups: comp.os.vxworks Subject: Safe Configuration with VxWorks and NT? Date: Fri, 14 Jan 2000 18:05:40 +0100 From: touchard Organization: Laboratoire Ganil (CEA/CNRS) Message-ID: <387F5764.EDE40D12@ganil.fr> Hi, I am writing from France. Excuse me for my poor english... We are going to deploy several VME crate with VxWorks which are loading there configuration from a Windows NT Server. We are looking for a safe solution concerning the NT server. May we install two servers, or is there any cheeper solution than buying a second expensive WindRiver license? May we buy spare cards and disks etc... to repair an out of function Windows NT Server PC? (does the license work with an other CPU or Network Card?) Thank's for your help. TOUCHARD Dominique. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks: VxLoad.com problem Date: Fri, 14 Jan 2000 10:17:17 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <85ntqk$pka$1@newsmaster01.magnet.at> References: <85grkf$i3j11@overload.lbl.gov> <85i18d$vqi$1@nnrp1.deja.com> Hello, i too have problems with vxLoad. i boot DOS 5.0 , without any config.sys, autoexec.bat (so no protect mode driver is loaded). the HW is a SCSI CDROM. i have c:\vxworks and there vxload.com and my bootrom.sys. (i have a mkboot generated floppy with which i can boot. i load vxworks itself with FTP) vxLoad --> Image memory is occopied. try to reduce system space vxload bootrom.sys --> Not enough memory to load. and the (C) msg is: VxLoad version 1.1 (c) WindRiver Systems 1993 when i try your vxsize stmt against my a:\bootrom.sys i get : F:\Tornado\host\x86-win32\bin\size386.exe: a:\bootrom.sys: File format not recognized (the target HW is a 486 DX4, 100) when i use vxsize i386 a:\bootrom.sys i see: usage: vxsize [-v h_adr l_adr]|[-k nnn]|[-b xxx] ??? PS: what are the items you mention below for? [menu],[bootrom],... do i have to write a special file and put all of them into to make vxload running? - -- regards, Mario Semo. http://www.kirchnersoft.com elucazeau@my-deja.com wrote in message <85i18d$vqi$1@nnrp1.deja.com>... >I use vxload.com with DOS 6.22 and have no problem. I join below a copy of my >`config.sys' - where "a:" is not the floppy but a Flash disk. > >Maybe your VxWorks bootrom image is really too big ! Check it with the >`vxsize' utility. Here is the output for my own kernel : > >` vxsize i386 -v 0010000 002000 a:\vxworks\bootrom >` a:\vxworks\bootrom: 333144(t) + 8048(d) + 16676(b) = 357868 > >Remember that the bootrom kernel that you try to load must fit in the 640 kB >low memory of your PC. > >HTH > >====================== > >[menu] >menuitem=BOOTROM >menuitem=DOS >menudefault=BOOTROM,2 > >[BOOTROM] >device=a:\himem.sys /verbose /testmem:off >dos=high, umb >shell=a:\vxworks\vxload.com a:\vxworks\bootrom > >[DOS] >DEVICE=a:\SETVER.EXE >device=a:\himem.sys /verbose /testmem:off >DOS=HIGH >COUNTRY=033,850,a:\COUNTRY.SYS >DEVICE=a:\DISPLAY.SYS CON=(EGA,,1) >FILES=30 > > >Sent via Deja.com http://www.deja.com/ >Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jan 16 04:03:07 2000 From: Vxworks Exploder Date: Sun Jan 16 04:03:09 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jan 16 04:03:04 PST 2000 Subject: Re: fatal WTX error 0x10136, 0x1012f Subject: Re: Safe Configuration with VxWorks and NT? Subject: Re: Secure File System for Flash Memory ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: fatal WTX error 0x10136, 0x1012f Date: Sat, 15 Jan 2000 14:27:09 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <85psjc$p97$1@newsmaster01.magnet.at> References: <85ha0m$com$1@newsmaster01.magnet.at> Hello, you are right. this is a bug in VxWorks and it fixes my problem too!!, setup REMOTE REGISTRY with LOCAL IP address instead of LOCAL REGISTRY fixes the sync problem! thx for help. - -- regards, Mario Semo. http://www.kirchnersoft.com Markus MItterer wrote in message ... >Hello Mario, >I suppose that you use Tornado2. There you must set up your registry (last >tab in the option-box). You have to specify whether the registry is running >on your local or on a remote host. My problem was, that I had two network >cards at host side and the registry could not work with the local entry >(127.0.0.1). The workaround was to set up remote registry with your local >host ip address (10.12.17.223). >At last I had to set a route path for vxworks: routeAdd("10.12.17.223", >"90.0.0.1") >where 90.0.0.1 is my target and 10.12.17.223 is my host. >I don`t know if you need all of this when you only have one network card. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Safe Configuration with VxWorks and NT? Date: Sat, 15 Jan 2000 22:09:48 GMT From: pwhicker@notthis.home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <387F5764.EDE40D12@ganil.fr> You don't need a development license for a boot host, you don't even need a Tornado installation. You just need to have the vxWorks object and any application object files. Anything with an FTP or NFS server will do. The Windows NT FTP server will do, it's not very good but it will work. The simplest, though not the most flexible, method is to install the FTP server and select a directory to export. Configure the target to use that host and give the relative path of the vxworks file. Reboot and enjoy. HTH Paul In article <387F5764.EDE40D12@ganil.fr>, touchard wrote: >Hi, > >I am writing from France. Excuse me for my poor english... > >We are going to deploy several VME crate with VxWorks which are loading >there configuration from a Windows NT Server. > >We are looking for a safe solution concerning the NT server. May we >install two servers, or is there any cheeper solution than buying a >second expensive WindRiver license? May we buy spare cards and disks >etc... to repair an out of function Windows NT Server PC? (does the >license work with an other CPU or Network Card?) > >Thank's for your help. > >TOUCHARD Dominique. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Secure File System for Flash Memory Date: Tue, 11 Jan 2000 09:00:54 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <387AD526.2533F87C@bitband.com> References: <85db4e$i3k4@overload.lbl.gov> Reply-To: leonid@bitband.com The additioanl power requirements may be helpful but are not necesarily needed to develop a safe Flash filing system. If the Flash filing system is aware of a possible interruption at all times, it will carefully ORDER Flash write operations in such a manner that during next startup it will be able to "reverse" a previously unfinished modification sequence (i.e. roll-back). There is much base theory for these issues from the world of distributed Data Bases which one could adopt for a Flash File System development. Leonid "Blake, Don" wrote: > > I can't answer you specific questions but there is something else you should > be aware of. Flash memory consistency during power loss is going to impose > two additional requirements on your power system: > > 1. Power must be held up long enough to complete the current write > operation. Depending on the implementation, this may need to include the > time it takes to perform a bank erase followed by a number of write > operations. > > 2. The software must get notification (e.g., an interrupt) that power > is possibly going down so that it can complete any in process write > operation and wait for power to go down or to come back before initiating > any subsequent writes. > > Don Blake > Sr. Programmer > Lockheed Martin Federal Systems > Owego, NY > > -----Original Message----- > > Submitted-by vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:32:26 2000 > Submitted-by: =?iso-8859-1?q?Eric=20Lucazeau?= > > Hello happy vxWorks users, > > I am a software engineer and I am in charge of developping a secure > File > System for a Flash Memory device. The file system must resist to > loss of power > : in case of a suddent power breakdown, the file system must stay > consistent > and loose nothing but the last initiated transaction. > > It appears to me that implementing that type of service with DosFs + > TrueFFS > will be quite difficult. Though TrueFFS is safe, DosFs is not. Does > anyone > know about another file system - a bit richer than rawFs - that > would satisfy > my needs ? > > I am sure somebody has already had to face this problem. Any help is > welcome. > --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jan 17 01:13:09 2000 From: "Michael Lawnick" Date: Mon Jan 17 01:13:11 PST 2000 Subject: Re: Tornado : header files? Just overwrite an existing makro with VPATH. Set the value and click apply. VPATH will be added this way, the old one still remains. Take care: MultiDir is seperated by : not ; as expected. From: "Mario Semo" > Hello, > > I had already copied the files on a command line into the project > directory. what i had done: > > open tornado > create new project/workspace > n:\vxw\WithHpp > > open a commandline outside of tornado > copy the files into the project directory (n:\vxw\WithHpp) > > back to Tornado > Add Files > ---> i see foo.cpp immediate in the current directory > add foo.cpp > rebuild all > > error: > > .... > vxrm n:\vxw\WithHpp\prjComps.h n:\vxw\WithHpp\prjParams.h > n:\vxw\WithHpp\prjConfig.c n:\vx > w\WithHpp\linkSyms.c > vxrm n:\vxw\WithHpp\libs.nm n:\vxw\WithHpp\libs.size > make: *** No rule to make target `foo.hpp', needed by `bar.o'. Stop. > > now i tried to modify the MACROS. > > where have i tried this: > i have a windows inside tornado (after opening my WSP) called > WorkSpace:WithHpp > here i have 3 notebook pages > i selected page 3, BUILDS > here i opened (click on the +) for WithHpp Builds, > and then double clicked on simntGnu > > now i see a dialog "properties: build specification ...." > > here i have subpages for all thre tools, and i see a page called MACROS. > > here i can select 1 of 3 macros: > LD_PARTIAL_FLAGS > POST_BUILD_RULE > PRJ_LIBS > > i cannot add other macros nor do anything else on this window. > > also i tried the C/C++ compiler notebook page, and i see the compiler > flags, but i cannot edit anythinh. > > i zipped my mini project (3K) and added it to the 2nd mail i wrote > (reply to stephan.tompson). > > I am working with Tornado 2.0, FP3 applied. > > Thanks for any other ideas what i am doing wrong. > > -- > regards, Mario Semo. http://www.kirchnersoft.com > Michael Lawnick wrote in message > <200001121434.PAA00645@net.softec.de>... > >> Hello, > > >when you add a file to your project it isn't copied. > >The TCL-script that make's your dependency file is a bit stupid and > >does worry about files that are outside the projects dir. So: Add > >VPATH=xxx:yyy:zzz to Macro folder and -Ixxx -Iyyy -Izzz to the c/c++ > >compiler parameters folder in your build properties, with xxx,yyy,zzz > >as an example for three external directories. > > MfG Lawnick, SOFTEC GmbH ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== From vxwexplo-errs@csg.lbl.gov Mon Jan 17 04:03:08 2000 From: Vxworks Exploder Date: Mon Jan 17 04:03:10 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jan 17 04:03:05 PST 2000 Subject: Re: How to add the new END driver to Tornado Project. Subject: Re: when not compiled works fine. Subject: Re: VxWorks Equivalent of ScanDisk for fsDOS? Subject: performance measurement APIs on VxWorks? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to add the new END driver to Tornado Project. Date: Sun, 16 Jan 2000 16:00:31 GMT From: mrvalli@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <85spup$jii$1@nnrp1.deja.com> References: <85n8e6$hdv$1@news.cz.js.cn> You have to modify the component descriptor files and add the configlettes. This file is the one like this : \tornado\target\config\\00bsp.cdf. Modify the file suitably taking examples as the existing ones. By the way, can you share your driver with me. I can also test and give you feed back. Thanks. Valli. In article <85n8e6$hdv$1@news.cz.js.cn>, "Zhang Zhiqun" wrote: > Hi, VxWrokers > > I am making the Ethernet END Driver for Realtek RTL8139A Fast Ethernet NIC. > > I have created the drvier's object, rtl8139a.o. > > But, I dont't know how to add the rtl8139a.o to the Tornado II's project. > I try to modify the BSP's config.h and configNet.h, there is not effect. > > Can anyone tell me the method? > > -- > > Thanks and best regards. > > Zhang Zhiqun (Bula Cheung) > bula@guoguang.com.cn > > Guoguang Electronic Group Co. > http://www.guoguang.com.cn > 93 Qingtan Road, Changzhou, Jiangsu, 213015, China > Tel: +86 519 696 7935 ext. 2372 > Fax: +86 519 696 9104 > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: when not compiled works fine. Date: Mon, 17 Jan 2000 07:09:21 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <85uf6q$npa$1@nnrp1.deja.com> References: <85s3du$604$1@nnrp1.deja.com> <20000116160312.01114.00000389@ng-cl1.aol.com> > Try moving the taskDelay and printf statement around to say > after lock = intLock(); to help isolate the cause.The taskDelay would only be > for diagnostics, and should later be removed later. FYI: taskDelay() after intLock() could unlock the interrupts (on most architectures, perhaps all, any rescheduling activity will cause interrupts to be re-enabled. John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Equivalent of ScanDisk for fsDOS? Date: 05 Jan 2000 02:15:59 GMT From: sbobcontractor@aol.com (Sbobcontractor) Organization: AOL http://www.aol.com Message-ID: <20000104211559.02176.00000460@ng-fd1.aol.com> References: <84u1pa$fkc$1@news.netmar.com> My WRS FAE got us chkdsk, that we embedded into our kernel. --------------------------- Newsgroups: comp.os.vxworks Subject: performance measurement APIs on VxWorks? Date: Mon, 17 Jan 2000 17:58:33 +0900 From: "J.H. Liu" Organization: Inet Internet Services Message-ID: <85ulvr$fgc$1@news.nuri.net> I'm searching for API suite for performance measurement. Is there a way to know the following 2 things on VxWorks? 1. the CPU time a process consumes in a given interval 2. the amount of dynamically allocated memory by a process in a given interval. Thanx for advance --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jan 17 13:29:25 2000 From: john.c.calvin@exgate.tek.com Date: Mon Jan 17 13:29:27 PST 2000 Subject: RE: Secure File System for Flash Memory Eric, Apart from assuring that the FFS mechanics are in place to ensure that a new block of data is properly written before assigning the previous "stale" block to the reclamation pool, there are a number of good starting points for what your looking for. I'm rather surprised Leonid did not speak up more regarding this, since this was his creation. We use RST's FlashLib, which provides similar high level functionality as TrueFFS (it looks like a Dos file system, but it's really not)with a better and more flexible integration into a variety of Flash parts. One of the additional provisions FlashLib provides is hooks for a compression utility. I thought one of the main goals of Eric's inquiry was to incorporate a security mechanism. My cut on this would be to re-do the compression mechanism to incorporate DES or 3XDES, and your off to the races. So to a certain extent everything Eric's looking for has already been done, and its a matter of integration not engineering. Now obtaining FlashLib may be more problematic, but I'm sure Leonid can help out here;-) "Leonid, wrote:" >The additioanl power requirements may be helpful but are not >necesarily needed to develop a safe Flash filing system. >If the Flash filing system is aware of a possible interruption >at all times, it will carefully ORDER Flash write operations >in such a manner that during next startup it will be able to >"reverse" a previously unfinished modification sequence (i.e. >roll-back). > >There is much base theory for these issues from the world of distributed >Data Bases which one could adopt for a Flash File System development. > >Leonid > >"Blake, Don" wrote: >> >> I can't answer you specific questions but there is something else you should >> be aware of. Flash memory consistency during power loss is going to impose >> two additional requirements on your power system: >> >> 1. Power must be held up long enough to complete the current write >> operation. Depending on the implementation, this may need to include the >> time it takes to perform a bank erase followed by a number of write >> operations. >> >> 2. The software must get notification (e.g., an interrupt) that power >> is possibly going down so that it can complete any in process write >> operation and wait for power to go down or to come back before initiating >> any subsequent writes. >> >> Don Blake >> Sr. Programmer >> Lockheed Martin Federal Systems >> Owego, NY >> >> -----Original Message----- >> >> Submitted-by vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:32:26 2000 >> Submitted-by: =?iso-8859-1?q?Eric=20Lucazeau?= >> >> Hello happy vxWorks users, >> >> I am a software engineer and I am in charge of developping a secure >> File >> System for a Flash Memory device. The file system must resist to >> loss of power >> : in case of a suddent power breakdown, the file system must stay >> consistent >> and loose nothing but the last initiated transaction. >> >> It appears to me that implementing that type of service with DosFs + >> TrueFFS >> will be quite difficult. Though TrueFFS is safe, DosFs is not. Does >> anyone >> know about another file system - a bit richer than rawFs - that >> would satisfy >> my needs ? >> >> I am sure somebody has already had to face this problem. Any help is >> welcome. >> Cheers, John Calvin, MBD, Oscilloscope Application Engineering, Tektronix Inc. 14150 S.W. Karl Braun Drive P.O. Box 500, M.S. 39-732 Beaverton OR 97077 U.S.A. John.C.Calvin@tek.com, Tel:503-627-4465,Page:503-497-1729,Fax:503-627-5548 From vxwexplo-errs@csg.lbl.gov Tue Jan 18 04:03:10 2000 From: Vxworks Exploder Date: Tue Jan 18 04:03:13 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jan 18 04:03:07 PST 2000 Subject: Re: RTOS comparisons?? Subject: Re: PPC performance decrease after "ROMMING" appl. Subject: Re: Problems with Booting from Network Subject: Re: Connecting GDB and Tornado 1.0.1 using RDB Subject: Re: Problems with Booting from Network Subject: Re: How do I continue building with errors? Subject: Re: How to get the size of struct exactly? Subject: Re: Secure File System for Flash Memory ------------------------------------------------------- Newsgroups: comp.os.qnx,comp.os.vxworks Subject: Re: RTOS comparisons?? Date: Mon, 17 Jan 2000 17:02:31 GMT From: pwhicker@notthis.home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <387F7C4D.CD5FFE15@juno.com> <948105929.19918.0.nnrp-14.d4e53aea@news.demon.co.uk> Enea, now which one are they? OSE? We looked at that but the idea that execution was all on the back of message passing didn't fit in with our mindset and the price wasn't marvelous at the end of the day. They sell message passing as if it were a new invention. Personally I'd prefer not to have the funstional elements of my software so strongly related. http://www.enea.com/PRODUCT/i_prod.htm I like their little animated GIF with the little remote control vehicle from Mars Pathfinder. Was that theirs then? I thought that was a vxWorks job according to NASA? Since that more than doubled it's target life I don't see their point. I thought the little six wheel thing used an 8051 though, does OSE work on the 8051? I guess this was an OSE advert then? Paul In article <948105929.19918.0.nnrp-14.d4e53aea@news.demon.co.uk>, "Richard Copeman" wrote: >Take a look at www.enea.com for embedded RTOS support for PowerPC. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPC performance decrease after "ROMMING" appl. Date: Mon, 17 Jan 2000 04:46:28 -0500 From: Bert Pleijsier Organization: National Aerospace Laboratory Message-ID: References: <387A00AC.C9CF6C67@nlr.nl> <85irut$lcb$1@nnrp1.deja.com> Reply-To: pleysier@nlr.nl 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_000_01BF60D3.F2D3626E Content-Type: text/plain Thank you, John, You put me on the right "track", with your idea to check the HID0 register. The "rom" version does (later on) enable ICACHE and DCACHE, but examining the HID0 register, I saw some bit differences eventually (when my appl. runs): "ROM version" (using romInit.s and bootInit.c): 1. EMCP bit is set (DRAM ECC detection) 2. SIED bit cleared (serial instr exec disable) 3. BHTE bit cleared (branch history table enable) "RAM version" (using sysAlib.s): 1. EMCP bit cleared 2. SIED bit set (in sysInit() ) 3. BHTE bit set (in sysInit() ) The SIED and BHTE bit being set strongly improve the PPC performance (of course). Setting these bits in ROM version resulted in same performance figures as for RAM version. Best regards, Bert John Fusco wrote: > > Interesting. > > Here's an dumb question. In test case 1 you say you are running the > test program from the shell. Do you invoke the test program from the > shell in case 2 also? I ask because the shell defaults to high > priority, so maybe you are running at two different priorities in each > case. That shouldn't matter much if your CPU is idle. > > Next I would suspect that maybe the data cache is disabled in the > vxWorks_rom target, but I doubt it. > > I noticed that romInit.s disables the instruction and data caches (ICE > and DCE bits in HID0), but doesn't appear to re-enable the data cache. > Maybe it does, but it's just not obvious to me from just reading the > source. Or, maybe this occurs later. > > You could check the HID0 register yourself to verify. From C use [int > vxHid0Get(void)] and mask the return with 0xc000. A value of 0xc000 > means that both the instuction and data caches are enabled. A value of > 0x8000 means that only the instruction cache is enabled and a value of > 0x4000 means that only the data cache is enabled. > > Let me know what you discover, I may want to ROM my kernel someday. > > Regards, > John > > In article <387A00AC.C9CF6C67@nlr.nl>, > pleysier@nlr.nl wrote: > > This is a multi-part message in MIME format. > > --------------328BDACEF96C1B22428C1854 > > Content-Type: text/plain; charset=us-ascii > > Content-Transfer-Encoding: 7bit > > > > Hi, all, > > > > I (still) use VxWorks 5.3.1 and Tornado 1.0.1, > > and have a MVME2305 PPC board (with VxWorks BSP). > > A simple testprog writes a fixed number of times to DRAM on PPC board. > > > > Two cases: > > 1. Bootrom in (flash)ROM, > > VxWorks download in RAM over ethernet + testprog execution via > shell > > in RAM. > > 2. VxWorks + testprog in (flash)ROM (not ROM resident, but copied to > > RAM) > > using the "vxWorks_rom" making (..\h\make\rules.bsp) as template. > > > > Testprog executes slower in case 2 than in case 1. > > --------------------------------------------------- > > > > Does anyone have an idea about this ??? > > Thanks in advance > > - ------ =_NextPart_000_01BF60D3.F2D3626E Content-Type: text/x-vcard; name="pleysier.vcf" Content-Disposition: attachment; filename="pleysier.vcf" Content-Description: Card for Bert Pleijsier begin:vcard n:Pleijsier;Bert tel;fax:+31 727 248210 tel;work:+31 527 248328 x-mozilla-html:FALSE url:www.nlr.nl org:National Aerospace Laboratory (NLR);Avionics Department, Electronics Division (EE) adr:;;Po.Box 31;Emmeloord;;8300 AD;Netherlands version:2.1 email;internet:pleysier@nlr.nl title:Senior Design Engineer x-mozilla-cpt:;2144 fn:A. Pleijsier end:vcard - ------ =_NextPart_000_01BF60D3.F2D3626E-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with Booting from Network Date: Mon, 17 Jan 2000 12:13:49 -0500 From: bruce@figureitout.com (Bruce) Message-ID: <38834df1.6435337@199.93.124.85> References: <85vh0h$g17$1@nnrp1.deja.com> On Mon, 17 Jan 2000 11:46:15 -0500, bfeng@noran.com wrote: >messages. I appreciate anyone's hint. What does error code "0xd0003" >actually mean? Look up printErrno(). 0xd0003 means that you have passed an invalid file descriptor to a function in the I/O System library. Bruce in Katy, TX --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Connecting GDB and Tornado 1.0.1 using RDB Date: Mon, 17 Jan 2000 16:55:23 GMT From: realtime_guru@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <85vhhh$g9t$1@nnrp1.deja.com> References: <8504o1$ahg$1@nnrp1.deja.com> Also, don't forget to define INCLUDE_CONFIGURATION_5_2 when compiling your kernel. In article , "Don Bowman" wrote: > > wrote in message news:8504o1$ahg$1@nnrp1.deja.com... > > I am using GDB and trying to connect to Tornado 1.0.1 using RDB. My > > copy of Tornado is for a PowerPC and the BSP is for a Motorola MBX860. > > > I have the following in usrNetwork.c > If its been removed from yours you could try adding it back. > > #ifdef INCLUDE_TELNET > telnetInit (); > #endif /* INCLUDE_TELNET */ > > /* initialize rpc daemon if specified */ > > #ifdef INCLUDE_RPC > rpcInit (); > #endif /* INCLUDE_RPC */ > > #ifdef INCLUDE_RDB > rdbInit (); /* initialize remote rpc debugging */ > #endif /* INCLUDE_RDB */ > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with Booting from Network Date: 17 Jan 2000 19:46:30 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <85vrim$pb7$1@news1.xs4all.nl> References: <85vh0h$g17$1@nnrp1.deja.com> bfeng@noran.com wrote: : I am trying to boot a Pentium PC via a network using a default factory : Pentium BSP boot disk. I managed to change the booting parameters after : "[boot rom]" prompt showed up. A ftp server "wftpd.exe" from : "Tornado2.0" CD is always launched on my host NT computer. The network : connection was tested OK while booting the Pentium PC into Win95. The : FTP login user name and password were also tested OK in Win95. However : everytime I try to boot it with Vxworks, I have the following error : messages. I appreciate anyone's hint. What does error code "0xd0003" : actually mean? Instead of using wftpd we are using tfptd, also supplied with Tornado. We had no problems with this at all. : file nmae : /tornado/target/config/bobPentium/vxworks.st Have you checked the logfile of wftpd (I don't know if there are any available). One thing I could think of is that the '/' is a problem for NT. Try placing the vxworks.st file in the root. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I continue building with errors? Date: 18 Jan 2000 00:01:16 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <860agc$jpp$2@news1.xs4all.nl> References: <8604qo$ve4$1@nnrp1.deja.com> zeaflal@my-deja.com wrote: : I have just started using T2 with the GNU compiler and inherited a : large number of almost correct C++ files. In doing a BUILD-ALL, I would : like to continue compiling all remaining files after the first error is : detected. For GNU, this means passing the -k option to the GNU MAKE : utility. It appears that this is specified in BUILDS, PROPERTIES, : MACROS dialog box. However, I cannot find the list of supported : Tornado macros. I don't think there is a macro for this. What you can do is start make from the command line. You have to set some variables before this works (at least on a pc, there you have to execute the batch-file torVars.bat in the bin directory of your Tornado distribution). Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get the size of struct exactly? Date: Tue, 18 Jan 2000 13:10:24 +0800 From: John Pollard Organization: Curtin University of Technology Message-ID: <3883F5C0.40A6FCA3@atcrc.com> References: <860r1b$f9s$1@nnrp1.deja.com> Hi, You will need to make sure that the structure is "packed", then you will get the true size of the combined elements in the structure. Your problem arises because the compiler has padded out the structure so that it is aligned correctly on a word boundary. That's about the "long" and "short" of it !! Regards, John P. zhouxiaochen@my-deja.com wrote: > Hi everybody, > > My environment is Tornado 1.0.1.(Windows NT 4.0),Vxworks 5.3, target is > PPC860. > > Please see this program. > > typedef struct aaa > { > long b; > }aaa; > > typedef struct bbb > { > short c; > }bbb; > > typedef struct ccc > { > long b; > short c; > }ccc; > > i = sizeof(struct aaa); > printf("%d",i); ------------i=4 > i = sizeof(struct bbb); > printf("%d",i); ------------i=2 > i = sizeof(struct ccc); > printf("%d",i); ------------i=8 > > How can I get the sizeof(struct ccc) exactly? > > Thanks in advance, > > Zhou Xiaochen > > Sent via Deja.com http://www.deja.com/ > Before you buy. - -- ______________________________________________________________________ John Pollard Australian Telecommunications Cooperative Research Centre (AT-CRC) Phone: +61 8 9266 3581 Curtin Uni of Technology .-_|\ Fax: +61 8 9266 3244 Bentley WA, 6102. / \ e-mail: johnp@atcrc.com Australia P_.-._/ _____________________________________________________________________o _// _/_/_/_/_/ _/_/_/ _/_/_/_/ _/_/_/ _/ _/ _/ _/ _/ _/ _/ _/____/ _/ _/_/_/ _/ /_/_/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/_/ _/ _/ _/_/_/ ____________________________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Secure File System for Flash Memory Date: Tue, 18 Jan 2000 08:45:59 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38840C27.5428B55B@bitband.com> References: <86049q$9gf2@overload.lbl.gov> Reply-To: leonid@bitband.com John, I am extremely pleased to see your posting, as I was wondering how many customers are still using FLashLib (as opposed to the ones who have switche to TrueFFS). Although I dont own FlashLib, I might be able to provide pointers inside Wind River to poeple who insist to have FlashLib, despite its discontinued status. FlashLib does do everything Eric is looking for, and much more. FlashLib offers mroe functionality then TrueFFS can offer too, and compression is a good example. FlashLib uses Linear File System technology, based on Microsoft FFS specifications, which arrange the data on Flash memories in variable-sized records (extents). Since the extents are of variable size, each one of these extents is optionally compressed (deflate/inflate is integral with FlashLib) and protected with a CRC32. The CRC32 in the last versions was applied to all extents, wheather compressed or not to assure data integrity. Additional data encoding engine(s) can be configured into FlashLib which could encrypt the data if such security is needed. TrueFFS does not protect data integrity (except when its used on NAND based M-Systems hardware which got hardware ECC builtin), nor does it offer a transparent way to compress the data, except at application level. For some clients this means they have to double Flash memory size when switching from FlashLib to TrueFFS to accomodate all of their storage needs. I think this is all for this posting. Leonid john.c.calvin@exgate.tek.com wrote: > > Eric, > > Apart from assuring that the FFS mechanics are in place to ensure > that a new block of data is properly written before assigning the > previous "stale" block to the reclamation pool, there are a number > of good starting points for what your looking for. I'm rather > surprised Leonid did not speak up more regarding this, since this was > his creation. We use RST's FlashLib, which provides similar high > level functionality as TrueFFS (it looks like a Dos file system, > but it's really not)with a better and more flexible integration into > a variety of Flash parts. One of the additional provisions FlashLib > provides is hooks for a compression utility. I thought one of the > main goals of Eric's inquiry was to incorporate a security mechanism. > My cut on this would be to re-do the compression mechanism to incorporate > DES or 3XDES, and your off to the races. So to a certain extent > everything Eric's looking for has already been done, and its a matter > of integration not engineering. > > Now obtaining FlashLib may be more problematic, but I'm sure Leonid > can help out here;-) > > "Leonid, wrote:" > >The additioanl power requirements may be helpful but are not > >necesarily needed to develop a safe Flash filing system. > >If the Flash filing system is aware of a possible interruption > >at all times, it will carefully ORDER Flash write operations > >in such a manner that during next startup it will be able to > >"reverse" a previously unfinished modification sequence (i.e. > >roll-back). > > > >There is much base theory for these issues from the world of distributed > >Data Bases which one could adopt for a Flash File System development. > > > >Leonid > > > >"Blake, Don" wrote: > >> > >> I can't answer you specific questions but there is something else you > should > >> be aware of. Flash memory consistency during power loss is going to > impose > >> two additional requirements on your power system: > >> > >> 1. Power must be held up long enough to complete the current write > >> operation. Depending on the implementation, this may need to include the > >> time it takes to perform a bank erase followed by a number of write > >> operations. > >> > >> 2. The software must get notification (e.g., an interrupt) that > power > >> is possibly going down so that it can complete any in process write > >> operation and wait for power to go down or to come back before initiating > >> any subsequent writes. > >> > >> Don Blake > >> Sr. Programmer > >> Lockheed Martin Federal Systems > >> Owego, NY > >> > >> -----Original Message----- > >> > >> Submitted-by vxwexplo-errs@csg.lbl.gov Mon Jan 10 05:32:26 2000 > >> Submitted-by: =?iso-8859-1?q?Eric=20Lucazeau?= > > >> > >> Hello happy vxWorks users, > >> > >> I am a software engineer and I am in charge of developping a > secure > >> File > >> System for a Flash Memory device. The file system must resist to > >> loss of power > >> : in case of a suddent power breakdown, the file system must stay > >> consistent > >> and loose nothing but the last initiated transaction. > >> > >> It appears to me that implementing that type of service with > DosFs + > >> TrueFFS > >> will be quite difficult. Though TrueFFS is safe, DosFs is not. > Does > >> anyone > >> know about another file system - a bit richer than rawFs - that > >> would satisfy > >> my needs ? > >> > >> I am sure somebody has already had to face this problem. Any help > is > >> welcome. > >> > > Cheers, > > John Calvin, MBD, Oscilloscope Application Engineering, Tektronix Inc. > 14150 S.W. Karl Braun Drive P.O. Box 500, M.S. 39-732 > Beaverton OR 97077 U.S.A. > John.C.Calvin@tek.com, Tel:503-627-4465,Page:503-497-1729,Fax:503-627-5548 --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jan 18 04:22:25 2000 From: "Dipl.-Phys. S. Richter, IKP, Tel. 16-5152" Date: Tue Jan 18 04:22:27 PST 2000 Subject: Problems with tu-driver Hi! I need some clarification concerning the use of the tuFilterSetup routine. That's the routine that handels the address-filtering of the DECchip 21040. We're using an AXPvme233 (Alpha-Processor) and VxWorks 5.31. So far as I understand this routine it allows to reset the filter for the tu-interface. What I want to do now is to change the filter to accept the original MAC- address and in addition to that another MAC-address and one multicast- address. But when I try to use this function I encounter some severe problems. If I follow the example from the documentation by typing in those lines from the console -> q = malloc (1600) -> bzero (q, 1600) -> q[0]=0x0403020100000001 -> q[8]=0x00000605 -> tuFilterSetup(0,0x1|0x20,q) I receive an OK that the filter change was successful. If I try the same inside a program or logged onto the system via rlogin I lose the contact to my system. Even if I set the filter-address to original MAC-address I see the same behavior. And that's something I really won't expect. Any hints how to use this function correctly would be very much appreciated! Regards Simone Dipl.-Phys. Simone Richter simone@ikp.tu-darmstadt.de Institut f. Kernphysik Schlossgartenstr. 9 TU Darmstadt 64289 Darmstadt Tel: +49 (0) 6151 / 16-5152, 16-3323 Fax: +49 (0) 6151 / 16-4321 From vxwexplo-errs@csg.lbl.gov Tue Jan 18 09:11:31 2000 From: Erin Swieringa Date: Tue Jan 18 09:11:34 PST 2000 Subject: DOS file size of 0 Hello, Using Tornado2 on a Pentium platform using a DOS file system on a hard disk, if you create an empty file from VxWorks then go to DOS and run the DOS tool Scandisk on the hard disk, Scandisk complains that the file is misreporting its size. If you allow Scandisk to fix it, it grabs data from an adjacent file and marks the size up. There is no problem with files with at least 1 byte of data. Has anybody run into this and/or resolved this? To recreate do a simple open and close from VxWorks to create a new file. Boot into DOS and run Scandisk. I've also tried similar code using "creat" with the same results. I've also truncated the file to 0 using ioctl with FIOTRUNC and I get the same error in Scandisk. fp = fopen("c:\\test.jnk", "w"); fclose(fp); Thanks, Erin From vxwexplo-errs@csg.lbl.gov Wed Jan 19 04:03:08 2000 From: Vxworks Exploder Date: Wed Jan 19 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jan 19 04:03:05 PST 2000 Subject: Re: Max socket allowed Subject: Re: VxWorks: VxLoad.com problem Subject: SCC serial devices on MBX860 Subject: Do I need a BSP ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max socket allowed Date: Tue, 18 Jan 2000 08:04:05 -0700 From: "Russell A. Johnson" Message-ID: <388480E5.69A05525@bigfoot.com> References: <85vk4i$qpg@journal.concentric.net> If I remember correctly, the max value is tied to the number of file handles defined. We override that in config.h for our purposes. Russ Yi Qin wrote: > > A quick one, what's the default number of sockets can a VxWorks task > open? And where is this default MAX value defined? If I want to change > this default MAX value, is it just a matter of changing the default MAX > or there are other values I have to change? > > TIA. > > Yi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks: VxLoad.com problem Date: Tue, 18 Jan 2000 22:50:34 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <862nau$a83$1@newsmaster01.magnet.at> References: <85grkf$i3j11@overload.lbl.gov> <85i18d$vqi$1@nnrp1.deja.com> <85ntqk$pka$1@newsmaster01.magnet.at> <85ul6o$sit$1@nnrp1.deja.com> Hello, > >The line >`shell=a:\vxworks\vxload.com a:\vxworks\bootrom >in "config.sys" saves free memory because "command.com" will not be loaded. good idea. i have to test this one! > >> when i try your vxsize stmt against my a:\bootrom.sys i get : >> F:\Tornado\host\x86-win32\bin\size386.exe: a:\bootrom.sys: File format not recognized > >"bootrom.sys" ? This is a confusing name. The file you must *vxload* is >"bootrom_uncmp" or "vxworks.st". Rename it whatever you like but do not >*vxcopy* the a.out formated file to produce "bootrom.sys". ok. i go to the target directory (tornado\....\.486) there i call make clear make bootrom_uncmp and then i call mkboot bootrom_uncmp a: and this (TORNADO!! mkboot.bat file does): vxcopy %2 %1bootrom.sys so, the file is copies as bootrom.sys to my floppy disk. so, the default name generated by tornado is bootrom.sys. so, what is confusing with this name? >When the format of the "bootrom_uncmp" file is correct both utilities >*vxsize* and *size386* should work fine. interesting. i now tried to use vxSize BEFORE and AFTER i copied it to the floppy. (with mkboot): F:\Tornado\target\config\pc486>vxsize i386 -v 0010000 002000 bootrom_uncmp bootrom_uncmp: 332928(t) + 7472(d) + 16676(b) = 357076 F:\Tornado\target\config\pc486>vxsize i386 -v 0010000 002000 a:\bootrom.sys F:\Tornado\host\x86-win32\bin\size386.exe: a:\bootrom.sys: File format not recognized >> usage: vxsize [-v h_adr l_adr]|[-k nnn]|[-b xxx] > >This is not clear reading this help line, but you must provide at least one >of the 3 options (-v -k -b) in order to have *vxsize* working. ah. i see. > >> PS: what are the items you mention below for? >> [menu],[bootrom],... > >This is a DOS 6.2 feature for multi DOS configurations. If you don't have >this feature, I suggest you have a floppy rescue disk before you modify the >"config.sys" file on hard disk. i have OS2 and NT installed on othe partitions of this machine. so i can boot the system when i crash my config.sys via boot managers and other OSes and repy my config.sys... and i have a DOS 5.0 boot floppy by hand, and i have a NT boot floppy, and i have my 3 OS2 install disks which allow me to open a command prompt.... --------------------------- Newsgroups: comp.os.vxworks Subject: SCC serial devices on MBX860 Date: Tue, 18 Jan 2000 15:15:32 -0800 From: Matthew Volpe Organization: Radix Technologies Message-ID: <3884F414.F96D3@radixtek.com> I received some unsupported code from Wind River to initialize the SCC serial devices com2, 3 & 4 for the standard MBX860 BSP. I believe that the integration of this code was successful as I see all 4 serial interfaces (/tyCo/0, /tyCo/1, /tyCo/2, /tyCo/3) listed from the shell. However, I'm wondering if I'm missing a step in making use of the new interface. /tyCo/0 is initialized as the console and I can still read/write this interface via STD_IN/STD_OUT. This /tyCo/0 console device interfaces through the "COM1" connector (J18) on MBX860. I now want to make use of one of the new serial devices (/tyCo/1) and interface with it via the "Utility #2" connector (J19) on the board. I open the device via: mFd = open("/tyCo/1",O_RDWR,0); (void) ioctl (mFd, FIOBAUDRATE, CONSOLE_BAUD_RATE); (void) ioctl (mFd, FIOSETOPTIONS, OPT_TERMINAL); This is the same way that the console gets initialized in usrConfig.c (usrRoot() routine). When this new code gets executed, I see a file descriptor = 8 returned. If I issue a write via: if (write(mFd, "\rATS0=0\r",8) == ERROR) /* ultimately I'm trying to connect a modem and read the modem strings from this device */ there is no error returned, but I see no characters displayed on the console that I have attached to Utility connector #2 and when I attempt a read, the attached console does not allow input. Is there anything missing to bind or initialize the /tyCo/1 device to this connector? Is there any additional initialization to allow console input/output from this device? How do I know that /tyCo/1 is initialized for COM2/UTILITY connector #2 in the code? I feel that I'm so close, but missing something simple... Any comments would be greatly appreciated. - -- Matt Volpe- Sr. Software Engineer | mvolpe@radixtek.com | (650)944-1239 Radix Technologies 329 N Bernardo Ave. Mountain View, CA 94043 - -- --------------------------- Newsgroups: comp.os.vxworks Subject: Do I need a BSP Date: Tue, 18 Jan 2000 17:43:19 -0800 From: Chuck Hardy Organization: Pacific Bell Internet Services Message-ID: <388516B7.75E0784F@pacbell.net> Reply-To: crhardy@pacbell.net My previous experience with vxWorks has been limited to projects where I used a Wind River BSP, or a BSP supplied by the vendor of a particular processor board. We are now considering a project that will involve a new processor board (CPU, RAM, flash, etc.) to be built by us. My question is this: do I need to create a BSP from scratch to support our new hardware, or can I configure an existing one to work with our system configuration? Also, we are considering the Intel 80186 processor for this project. I have never seen any reference to a BSP for this 16-bit processor. Is there such a thing? I would appreciate any help I can get. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jan 19 05:11:30 2000 From: David Anderson Date: Wed Jan 19 05:11:33 PST 2000 Subject: RE: DOS file size of 0 On 18 January 2000 17:14, Erin Swieringa [SMTP:ESwieringa@etcconnect.com] wrote: > Hello, > > Using Tornado2 on a Pentium platform using a DOS file system on a hard disk, > if you create an empty file from VxWorks then go to DOS and run the DOS tool > Scandisk on the hard disk, Scandisk complains that the file is misreporting > its size. If you allow Scandisk to fix it, it grabs data from an adjacent > file and marks the size up. There is no problem with files with at least 1 > byte of data. Has anybody run into this and/or resolved this? > > To recreate do a simple open and close from VxWorks to create a new file. > Boot into DOS and run Scandisk. I've also tried similar code using "creat" > with the same results. I've also truncated the file to 0 using ioctl with > FIOTRUNC and I get the same error in Scandisk. > > fp = fopen("c:\\test.jnk", "w"); > fclose(fp); > > Thanks, > > Erin I'm no expert on Tornado 2 but this seems to point to the following happening: The open creates the directory entry and (possibly) allocates the first cluster of the file. The first cluster number is held in the directory entry. The directory entry also contains the file length. I suspect this is not correctly set to zero. The close file does what it should... but as there is no file write the file length is not updated ( nothing cached to flush to disk ). I suspect that you would need to write to the file to ensure the file length is correctly set. You may well find that writing zero bytes could work... if this marks the cache entry as dirty. Writing one byte, flushing to disk, and then truncating the file may also work. As an alternative suggestion try using open/close/creat rather than the buffered fopen/fclose. This does not have any buffering and would perhaps prove the point above. e.g. // Tornado 2 bug work around ?? void CreateZeroLengthFile(char *name) { int fp; fp = creat(name); close(fp); } Hope that helps David A. From vxwexplo-errs@csg.lbl.gov Wed Jan 19 06:42:13 2000 From: Reyhan ERGUN Date: Wed Jan 19 06:42:15 PST 2000 Subject: adding files to Tornado 2 project? This is a multi-part message in MIME format. --------------E76C4FE05D09754347AAC386 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have some number of files which are generated by another tool (Telelogic SDT), and I want to link them all to my vxWorks build. I will compile and generate object files for these extra codes within T2 (UNIX and WinNT) and add all these files to my Tornado II project, but I am looking for a nicer way to do this. There is a MACRO as EXTRA_MODULES and I added a definition as /*.c for this macro but it doesn't work. Your comments will be so helpful for me in this issue. Best regards, Reyhan --------------E76C4FE05D09754347AAC386 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Reyhan (FINDIK) ERGUN Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Reyhan (FINDIK) ERGUN n: ;Reyhan (FINDIK) ERGUN org: ASELSAN INC. adr: Microwave and System Technologies Division ;;Software Engineering Dept.;ANKARA;;;TURKEY email;internet: findik@mst.aselsan.com.tr title: Principal Engineer tel;work: (90-312) 385 19 00 (Ext. 2668) tel;fax: (90-312) 354 52 05 note: http://www.aselsan.com.tr x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------E76C4FE05D09754347AAC386-- From vxwexplo-errs@csg.lbl.gov Wed Jan 19 09:27:03 2000 From: "Minbo Shim" Date: Wed Jan 19 09:27:06 PST 2000 Subject: vxworks: Visicom RtX-Window text widget class support Hello vxWorks X users, I have a Visicom VigraVision PMC graphic card in a Motorola MV2400 PowerPC board. The RtX-Windows running on this card serves both X server and client. I am trying to run a X application that was successfully ran on Solaris/Linux boxes and display it into a miniature display that is directly connected to the VGA out of the graphic card. With all that efforts I was not able to run the application through the RtX-Window server. It seems to me that the X server somehow can't handle the text widget classes. To verify this, I ran the following very simple code: //==================================================================== #include #include int show_text(char *mytext) { Widget toplevel; XtAppContext app; int argc; char *argv[3] = {0}; argc = 2; argv[0] = "show_text"; argv[1] = mytext; rtxLibInit ( X11LIB | XMULIB | XEXTLIB | XTLIB | XMLIB | XAWLIB ); XtSetLanguageProc(NULL,NULL,NULL); toplevel = XtVaAppInitialize(&app,"Demos",NULL, 0,&argc,argv,NULL,NULL); XtVaCreateManagedWidget("text", xmTextWidgetClass,toplevel, XmNvalue,mytext, NULL); XtRealizeWidget(toplevel); XtAppMainLoop(app); } //==================================================================== As I expected, it failed. All other X clients like button, label, list, etc. were created and handled properly by the RtX-Window Xserver. I am wondering this is sort of a known problem, or I am missing something here. Thanks, M.Shim GDRS From vxwexplo-errs@csg.lbl.gov Wed Jan 19 17:02:34 2000 From: michael Date: Wed Jan 19 17:02:36 PST 2000 Subject: Question about VxSim Hi, I want to use a PC with Tornado2.0 and VxSim running to simulate the hardware board. Now there are 2 such PCs on an Ethernet, can a task on one VxSim(PC1) communicate with another task on another VxSim(PC2) using the socket? Thanks. From vxwexplo-errs@csg.lbl.gov Thu Jan 20 04:03:07 2000 From: Vxworks Exploder Date: Thu Jan 20 04:03:09 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jan 20 04:03:04 PST 2000 Subject: Re: [URGENT] Does InetGateway works ? Subject: Re: Inet Config Mask Subject: Re: Question about memory confilict Subject: SNTP v. 4 Subject: fppSave() on MC68LC040 CPU ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: [URGENT] Does InetGateway works ? Date: Wed, 19 Jan 2000 07:53:20 -0600 From: Frank Minich Organization: NCP Message-ID: <3885C1CF.206A@ti.com> References: <3885BADC.A5D6654F@ganil.fr> Reply-To: fminich@ti.com touchard wrote: > > In fact, we don't succeed to communicate throw a gateway with VxWorks? > > Can a VxWorks OS open a TCP connection throw a gateway ? You need an entry for the gateway in your routing table. Check out routeAdd(). Frank. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Inet Config Mask Date: Wed, 19 Jan 2000 07:58:51 -0600 From: Frank Minich Organization: NCP Message-ID: <3885C31B.59BD@ti.com> References: <3885B1F6.D021B07F@ganil.fr> Reply-To: fminich@ti.com touchard wrote: > > How do we configure the Network Mask with the change command on the > VxWorks kernel ? > > Dominique. For example, when you set your host inet: 128.247.104.32:fffffe00 Frank. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Question about memory confilict Date: Wed, 19 Jan 2000 16:17:08 +0100 From: "Yves Nortier" Organization: [Posted via] Easynet France Message-ID: <864ki5$ut4$1@buggy.easynet.fr> References: <85eccb$cdh$1@nnrp1.deja.com> Personnaly, I allocate my buffers in the user reserved memory (between sysMemTop and SysPhysMemTop). In order to do that, define USER_RESERVED_MEM in config.h. Yves NORTIER zhouxiaochen@my-deja.com a écrit dans le message <85eccb$cdh$1@nnrp1.deja.com>... >Hi everybody, > >I am writing a program of AAL5 SAR driver for PPC860. My eviroment is >Tornado 1.01.(Windows NT 4.0) > >At first I should allocate some memory for buffer pool. So I can >allocate buffer to BD Link from buffer pool and release buffer of BD >Link to buffer pool. > >But when Tornado download object modules to target and the modules is >relocatable in memory and I can't control the address of bss, data and >the text segment in memory. This will induce memory conflict between >buffer pool and object files in memory. > >How can I do? > >Thanks in advance, > >Zhou Xiaochen > > >Sent via Deja.com http://www.deja.com/ >Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: SNTP v. 4 Date: Wed, 19 Jan 2000 10:04:41 -0600 From: Daniel Stewart Organization: Motorola Message-ID: <3885E099.5A12EF48@email.mot.com> Does anyone know if VxWorks supports version 4 of SNTP? I am getting a VERSION_UNSUPPORTED error during the call to sntpcTimeGet(). Thanks. Daniel --------------------------- Newsgroups: comp.os.vxworks Subject: fppSave() on MC68LC040 CPU Date: Wed, 19 Jan 2000 13:29:33 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <3886109D.60CBA327@aps.anl.gov> We have an application that runs on both mv162 and mv162lc boards which needs to do floating-point calculations in an ISR. It follows the rules and calls fppSave beforehand as is required. The problem we have is that it fails on the mv162lc because the vxWorks routine fppSave() for the 68LC040 is identical to the version for the 68040 and has an FSAVE instruction in it -- naughty! It seems reasonable that when running on an LC040 there should be no need to call fppSave, but we don't want to change the code (although we obviously do build it for both architectures, using the relevent gcc flags). Has anybody seen a patch for this for Tornado 1.0.1? - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jan 20 07:45:16 2000 From: David Laight Date: Thu Jan 20 07:45:18 PST 2000 Subject: re: Question about memory conflict If you need fixed addresses, another alternative is to increase RAM_LOW_ADRS and use the low memory addresses. This does, however, need a new BSP build. (Beware that the initial stack is just below RAM_LOW_ADRS, and various bits of vxWorks use other parts of [0..RAM_LOW_ADRS).) A big advantage of leaving USER_RESERVED_MEM zero is that the same vxWorks image can be run on systems with different amounts of memory. Build for the minimum memory size, but modify the code in sysLib.c to do a grope (I do it in sysMmuInit().) As an aside, note that the ARM Brutus BSP breaks if USER_RESERVED_MEM is changed - all to do with inconsistencies in the way the address of the temporary page table is determined in different bits of code. David > > Personnaly, I allocate my buffers in the user reserved memory (between > sysMemTop and SysPhysMemTop). > In order to do that, define USER_RESERVED_MEM in config.h. > > Yves NORTIER > > zhouxiaochen@my-deja.com a écrit dans le message > > >Hi everybody, > > > >I am writing a program of AAL5 SAR driver for PPC860. My eviroment is > >Tornado 1.01.(Windows NT 4.0) > > > >At first I should allocate some memory for buffer pool. So I can > >allocate buffer to BD Link from buffer pool and release buffer of BD > >Link to buffer pool. > > > >But when Tornado download object modules to target and the modules is > >relocatable in memory and I can't control the address of bss, data and > >the text segment in memory. This will induce memory conflict between > >buffer pool and object files in memory. > > > >How can I do? > > > >Thanks in advance, > > > >Zhou Xiaochen ---------------------------------------------------------------- David Laight email: dsl@tadpole.co.uk Tadpole Technology plc phone: +44 1223 428 232 Cambridge, UK fax: +44 1223 428 201 From vxwexplo-errs@csg.lbl.gov Thu Jan 20 10:04:07 2000 From: "Morris, John" Date: Thu Jan 20 10:04:09 PST 2000 Subject: Multiple Mailbox Interrupts On MVME167 Hi, I'm brand-spanking new to VxWorks, so I hope this isn't a stupid question. I'm in need of two mailbox interrupts on a MVME167. (That is, the 167 will be interrupted by other cards via mailbox.) I am using VxWorks 5.1. I've seen one defined and I can make it function perfectly. Is there a way to easily set up another (preferably of the same type : a one-byte write)? I know the card itself has the capability of 4 mailbox interrupts. Any help will be GREATLY appreciated. Thanks, John From vxwexplo-errs@csg.lbl.gov Thu Jan 20 10:42:05 2000 From: Alan Biocca Date: Thu Jan 20 10:42:07 PST 2000 Subject: Re: exploder problem - duplicate messages There is an occasional duplication of outgoing messages in the VxWorks Users Group exploder. We are investigating. -- Alan K Biocca At 09:39 AM 1/20/00 -0800, you wrote: >Alan, > >I've noticed that maybe 40% of the time, I'm receiving the same message >twice via the exploder. Am I the only one with this problem? > >It does not happen all the time. It just happened again with the most >recent post. I have copied in, below; not that that will help any. > >Ron > >--------------------------------------------- >Ronald Wagner >Principal Engineer >Manager, DRI Driving Simulator > >Dynamic Research, Inc. +1 310 212 5211 (v) >355 Van Ness Avenue +1 310 212 5046 (f) >Torrance, CA USA 90501 rwagner@dynres.com ... From vxwexplo-errs@csg.lbl.gov Thu Jan 20 13:57:24 2000 From: David Chen Date: Thu Jan 20 13:57:26 PST 2000 Subject: Startup Job Opening Hi, Please forward to your friend. Clarinet Systems will offer $500 as a referral fee for any engineering hires. Best Regards, David Chen Clarinet Systems is a dynamic new startup company in the heart of the Silicon Valley. The company designs, manufactures and markets high-speed Infrared and Bluetooth LAN/WAN and Internet access systems. The projects are exciting and challenging in a productive and comfortable office environment. If you're looking for a challenge and an opportunity to work with cutting edge products look no further. We offer excellent benefits and compensation. For more details, please visit our web site @ www.clarinetsys.com. Clarinet Systems will offer $500 as a referral fee for any engineering hires. Sr. Software Engineer: BS CS/EE or higher degree with a minimum of 5+ yrs SW experience using C, real time OS and knowledge of networking protocols. Exp. in at least one of following: VxWroks, network switching, bridge or router, infrared protocol, Bluetooth or device driver. Participate in all phases of SW development. Ability to take responsibility, ownership and initiate action. Must be a team player and a self-starter. Sr. QA Engineer: Responsible for developing test scenarios, test plan for wireless mobile networking product. Then execute testing according to that plan, and document and report problems to the product engineering team. Qualifications: Requires a BS/BA and 3+ years of QA experience in networking product. Prior experience working with networking protocols, VxWorks, TCP/IP, PPP, SNMP or infrared-protocol is a plus. Strong troubleshooting and problem-solving skills. Team player, quick learner, self-starter. Please send your resume to HR, Clarinet Systems, 1415 Koll Circle, #101, San Jose, CA 95112; fax 408-501-0255; email dchen@clarinetsys.com From vxwexplo-errs@csg.lbl.gov Thu Jan 20 16:39:15 2000 From: Jim Way Date: Thu Jan 20 16:39:17 PST 2000 Subject: Tornado 2.0 Post Mortem Analysis 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_01BF63A7.89AA7080 Content-Type: text/plain; charset="iso-8859-1" VxWorkers, I am a neophyte VxWorks user. My current effort is to setup my application to support post-mortem debugging through Tornado 2.0 / WindView. I've been using the WindView User's Guide and the on-line wvLib documentation. I've been trying to compile the sample program provided (see below) but I get several compile errors: D:\Work\Tomahawk\wvPostMortemUtilities.cpp:32: syntax error before `;' D:\Work\Tomahawk\wvPostMortemUtilities.cpp:34: syntax error before `;' D:\Work\Tomahawk\wvPostMortemUtilities.cpp: In function `bool PostMortem_Start(unsigned int)': D:\Work\Tomahawk\wvPostMortemUtilities.cpp:53: `taskBufId' undeclared (first use this function) The first three messages seem to indicate that WV_UPLOAD_TASK_ID and WV_TASKBUF_ID are unrecognized typedefs. A keyword search through WindSurf and a grep through the vxWorks installation directory seem to confirm this theory. Can anyone offer some guidance on where these symbols might be found? And if you're feeling particularly generous, any advice on setting up for a post-mortem analysis would be greatly appreciated . I'm trying to develop Start(), Stop(), and Upload() routines that can be reused on future projects. I'll be happy to share the code if anyone is interested. Jim Way jway@austron.com ==================================== EXAMPLE CODE FROM WINDRIVER FOLLOWS. Keyword search on WV_UPLOAD_TASK_ID will return this code as the only reference. /* Includes, as in the examples above. */ BUFFER_ID bufId; UPLOAD_ID pathId; WV_UPLOAD_TASK_ID upTaskId; // this is "line 32." WV_LOG_HEADER_ID hdrId; WV_TASKBUF_ID taskBufId; // this is "line 34." /* @ To prepare the event log and start logging: */ /* @ Create event buffer in non-zeroed memory, allowing overwrite, @ yielding bufId. */ wvEvtLogInit (bufId); taskBufId = wvTaskNamesPreserve (postMortemPartId, 32); hdrId = wvLogHeaderCreate (postMortemPartId); wvEvtClassSet (WV_CLASS_1); /* set to log class 1 events */ wvEvtLogStart (); /* @ System fails and reboots. Note that taskBufId, bufId and @ hdrId must be preserved through the reboot so they can be @ used to upload the data. */ /* Create an uplaod path, yielding pathId. */ wvLogHeaderUpload (hdrId, pathId); upTaskId = wvUploadStart (bufId, pathId, TRUE); wvUploadStop (upTaskId); wvTaskNamesUpload (taskBufId, pathId); /* Close the upload path and destroy the event buffer */ ==================== END OF EXAMPLE CODE. ------_=_NextPart_001_01BF63A7.89AA7080 Content-Type: text/html; charset="iso-8859-1"
VxWorkers,
 
I am a neophyte VxWorks user. My current effort is to setup my application to support post-mortem debugging through Tornado 2.0 / WindView. I've been using the WindView User's Guide and the on-line wvLib documentation. I've been trying to compile the sample program provided (see below) but I get several compile errors:
 
D:\Work\Tomahawk\wvPostMortemUtilities.cpp:32: syntax error before `;'
D:\Work\Tomahawk\wvPostMortemUtilities.cpp:34: syntax error before `;'
D:\Work\Tomahawk\wvPostMortemUtilities.cpp: In function `bool PostMortem_Start(unsigned int)':
D:\Work\Tomahawk\wvPostMortemUtilities.cpp:53: `taskBufId' undeclared (first use this function)
<etc>
 
The first three messages seem to indicate that WV_UPLOAD_TASK_ID and WV_TASKBUF_ID are unrecognized typedefs. A keyword search through WindSurf and a grep through the vxWorks installation directory seem to confirm this theory.
 
Can anyone offer some guidance on where these symbols might be found? And if you're feeling particularly generous, any advice on setting up for a post-mortem analysis would be greatly appreciated <g>. I'm trying to develop Start(), Stop(), and Upload() routines that can be reused on future projects. I'll be happy to share the code if anyone is interested.
 
Jim Way
 
====================================
EXAMPLE CODE FROM WINDRIVER FOLLOWS. Keyword search on WV_UPLOAD_TASK_ID will return this code as the only reference.
 
   /* Includes, as in the examples above. */
 
          BUFFER_ID           bufId;
          UPLOAD_ID           pathId;
          WV_UPLOAD_TASK_ID   upTaskId;        // this is "line 32."
          WV_LOG_HEADER_ID    hdrId;
          WV_TASKBUF_ID       taskBufId;            // this is "line 34."
 
          /*
           @ To prepare the event log and start logging:
           */
 
          /*
           @ Create event buffer in non-zeroed memory, allowing overwrite,
           @ yielding bufId.
           */
 
          wvEvtLogInit (bufId);
          taskBufId = wvTaskNamesPreserve (postMortemPartId, 32);
          hdrId = wvLogHeaderCreate (postMortemPartId);
          wvEvtClassSet (WV_CLASS_1);         /* set to log class 1 events */
          wvEvtLogStart ();
 
          /*
           @ System fails and reboots.  Note that taskBufId, bufId and
           @ hdrId must be preserved through the reboot so they can be
           @ used to upload the data.
           */
 
          /* Create an uplaod path, yielding pathId. */
 
          wvLogHeaderUpload (hdrId, pathId);
          upTaskId = wvUploadStart (bufId, pathId, TRUE);
          wvUploadStop (upTaskId);
          wvTaskNamesUpload (taskBufId, pathId);
 
          /* Close the upload path and destroy the event buffer */
 
====================
END OF EXAMPLE CODE.
------_=_NextPart_001_01BF63A7.89AA7080-- From vxwexplo-errs@csg.lbl.gov Thu Jan 20 17:50:32 2000 From: Fred Roeber Date: Thu Jan 20 17:50:34 PST 2000 Subject: Re: Tornado 2.0 Post Mortem Analysis On Thu, 20 Jan 2000, Jim Way asked: > The first three messages seem to indicate that WV_UPLOAD_TASK_ID and > WV_TASKBUF_ID are unrecognized typedefs. A keyword search through WindSurf > and a grep through the vxWorks installation directory seem to confirm this > theory. I think you want WV_UPLOADTASK_ID and TASKBUF_ID. These are both found in wvLib.h. As a general suggestion, when example code doesn't work, I normally double check it. In the case of the WindView sample code, the problems were in declaring the return values from certain functions. Finding the function declarations and checking their prototypes showed the correct return type declarations. Good luck. Fred | Fred J Roeber, BBN Systems & Technologies | | 4 John Clarke Road Middletown, RI 02842-5202 | | froeber@bbn.com 401-848-3548 | | TraceMaker product manager | | http://www.tracemaker.bbn.com | From vxwexplo-errs@csg.lbl.gov Thu Jan 20 23:50:17 2000 From: Reyhan ERGUN Date: Thu Jan 20 23:50:20 PST 2000 Subject: External Dependencies for T2 under UNIX This is a multi-part message in MIME format. --------------42CF4880C222A3B3CE5AD5E1 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am using T2 under UNIX and making a vxWorks build. I have 2 problems while generating external dependency: 1-) I am doing a simple trial and including some dummy header files in usrAppInit.c, after generating dependecy for all project files or for this file only, it doesn't make any change, I couldn't see those dummy header files under external dependency tap in the workspace and I can't see those header files in Makefile ad dependency for usrAppInit.o, under vxWorks properties application initialization macro is included! 2-) I add a new file to the project, I can see the header files under ext. dependecies. I added path for this file to C/C++ Compiler options, but I see that "file not found" under the property for this header file. Is there any immaturity for T2 under UNIX? Has anybody encountered such problems? Thanks for your valuable comments on the subject. Best regards, reyhan --------------42CF4880C222A3B3CE5AD5E1 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Reyhan (FINDIK) ERGUN Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Reyhan (FINDIK) ERGUN n: ;Reyhan (FINDIK) ERGUN org: ASELSAN INC. adr: Microwave and System Technologies Division ;;Software Engineering Dept.;ANKARA;;;TURKEY email;internet: findik@mst.aselsan.com.tr title: Principal Engineer tel;work: (90-312) 385 19 00 (Ext. 2668) tel;fax: (90-312) 354 52 05 note: http://www.aselsan.com.tr x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------42CF4880C222A3B3CE5AD5E1-- From vxwexplo-errs@csg.lbl.gov Fri Jan 21 03:55:25 2000 From: Abhijit Lahiri Date: Fri Jan 21 03:55:27 PST 2000 Subject: dialing in VxWorks Hi VxWorks users, Greetings. Is there a free modem dialing software available for VxWorks ? If it is available, please send me the URL. Thanks in Advance, Abhijit Lahiri From vxwexplo-errs@csg.lbl.gov Sun Jan 23 13:12:05 2000 From: Sara@openhere.com Date: Sun Jan 23 13:12:08 PST 2000 Subject: Your site has been included on OpenHere Hi, Your site has been included in the OpenHere.com index and search engine. OpenHere is one of the 10 largest index and search sites on the Internet. Your site listing is: Link: http://www.pso.com/vxfaq.html Title: VxWorks FAQ Description: OpenHere Category: http://www.openhere.com/tech1/software/operating-systems/realtime/vxworks/ You might have already received a message similar to this one, as we send a note when ever we add a link to your site in a different category on OpenHere.com. At OpenHere you can dynamically modify your site's listing at any time. You can also include your site's listing in other categories on OpenHere.com. When you modify your site's listing, it is automatically placed at the top of the category in which it is included, and is placed first in the search engine results for the keywords relating to your site. To modify, add or delete your listing: 1. Go to the page on OpenHere where your site is presently listed or you would like it listed. 2. Select "Suggest a Site". 3. Follow the instructions for changing your listing. All of the modifications you submit to OpenHere.com are processed in real time. As soon as you see the response to your submission, your site listing should be updated. www.OpenHere.com is frequented by both children and families. As a result, www.OpenHere.com does not include links to material which is illegal to display to minors. Sara www.OpenHere.com Your key to the Net! From vxwexplo-errs@csg.lbl.gov Sun Jan 23 21:48:57 2000 From: Nagaveni Date: Sun Jan 23 21:48:59 PST 2000 Subject: System time in vxWorks This is a multi-part message in MIME format. --------------389D924480CFFEFADAC1687A Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, how can I get the system time and date in vxWorks on a PC-486 target? Thanks in advance, Veni. --------------389D924480CFFEFADAC1687A Content-Type: text/x-vcard; charset=us-ascii; name="Nagaveni.Srikanth.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Nagaveni Content-Disposition: attachment; filename="Nagaveni.Srikanth.vcf" begin:vcard n:Srikanth;Nagaveni tel;work:0091-80-5722296 x-mozilla-html:FALSE adr:;;;;;; version:2.1 email;internet:Nagaveni.Srikanth@wipro.com ; ksnveni@wipinfo.soft.net fn:Nagaveni Srikanth end:vcard --------------389D924480CFFEFADAC1687A-- From vxwexplo-errs@csg.lbl.gov Mon Jan 24 00:09:38 2000 From: Abhijit Lahiri Date: Mon Jan 24 00:09:46 PST 2000 Subject: Re: System time in vxWorks > > Hi All, > how can I get the system time and date in vxWorks on a PC-486 target? All the time routines for VxWorks are provided by the ansiTime Lib. Subroutines like time(), localtime etc. will give what you want. Have a look at ansiTime library in VxWorks reference Guide. regards, abhijit From vxwexplo-errs@csg.lbl.gov Mon Jan 24 00:59:37 2000 From: "Girish V Gulawani" Date: Mon Jan 24 00:59:39 PST 2000 Subject: Re: System time in vxWorks > how can I get the system time and date in vxWorks on a PC-486 target? pl refer ansiTime or search for a specific rtc chip driver. From vxwexplo-errs@csg.lbl.gov Mon Jan 24 01:29:09 2000 From: Jaya Prakash M Date: Mon Jan 24 01:29:12 PST 2000 Subject: RE: System time in vxWorks Refer clockLib and time related functions of VxWorks first... Time ??? RTC or dedicated Chip ???.... Any way go thru the piece of code below: /* Check out the structures for timespec and tm*/ ............... ............. printf("\n\nEnter the Date and Time (24Hr Format) :mm/dd/yyyy hr::min:sec ->"); scanf("\n\n%d%d%d%d%d%d",&mm,&dd,&yyyy,&hr,&min,&sec); /* initialize the tm structures */ ...... .......... t1.tv_sec = mktime(&myStructTime); /* pointer to broken-down structure */ t1.tv_nsec = 0L; tRes = clock_settime(CLOCK_REALTIME,&t1); printf("\n\n Clock Set Ok/Failure: %d",tRes); if(clock_gettime(CLOCK_REALTIME,&t1)==OK) { printf("\n\n%ld",t1.tv_sec); printf("\n\n%s",ctime (&(t1.tv_sec))); } Good Luck...??? Regards, JP & ARP From vxwexplo-errs@csg.lbl.gov Mon Jan 24 09:32:57 2000 From: "Darnold, Joel R" Date: Mon Jan 24 09:33:00 PST 2000 Subject: RE: adding files to Tornado 2 project? Hello, I recently went through an exercise to do what I believe you are trying to do. I wanted my system to boot from the server without having to start the Tornado tools. If you add a definition to the VxWorks Makefile called MACH_EXTRA identifying the objects that you would like to link into the kernel Tornado will include them into the link. Following example: MY_ADD_DIR=c:\project\src MACH_EXTRA = $(MY_ADD_DIR)\file1.o $(MY_ADD_DIR)\file2.o Beware that this could increase the size of the VxWorks image dramatically depending on how you are organizing you data. The VxWorks kernel is loaded in the memory region defined between RAM_LOW_ADRS and RAM_HI_ADRS (see config.h)on your target. I've found that this area must be at least twice the size of your VxWorks image. I haven't discovered the reason for this but I found that if my VxWorks image is one byte larger than half the program memory region, my target will crash during bootup. I'm speculating that when the bootloader is FTPing the image, it copies it to the program area and then recopies it to it's appropriate memory locations afterwards, thus requiring twice the size. To increase the size of the program area, modify RAM_HI_ADRS in config.h and RAM_HIGH_ADRS in Makefile. You can also have the kernel call your user initialization function at bootup by adding a line to config.h defining USER_APPL_INIT. Following example: #define USER_APPL_INIT CallMyUserInit() Another thing I would like to do is add my files to the VxWorks project so that if I modify one of my files, the Makefiles for VxWorks will recognize that a dependency has changed and will recompile my files and relink the VxWorks image. I haven't had time to work on this yet as the Makefile(s) become very convoluted. If anyone has any insight...... Joel Darnold Member of Technical Staff Sandia National Laboratories Control Subsystems Dept. 2338 Voice: 505-844-2487 Fax: 505-844-2925 I have some number of files which are generated by another tool (Telelogic SDT), and I want to link them all to my vxWorks build. I will compile and generate object files for these extra codes within T2 (UNIX and WinNT) and add all these files to my Tornado II project, but I am looking for a nicer way to do this. There is a MACRO as EXTRA_MODULES and I added a definition as /*.c for this macro but it doesn't work. Your comments will be so helpful for me in this issue. Best regards, Reyhan From vxwexplo-errs@csg.lbl.gov Mon Jan 24 12:59:41 2000 From: "Minbo Shim" Date: Mon Jan 24 12:59:43 PST 2000 Subject: cksum: out of data vxWorks users: I am getting the message "cksum: out of data" from the vxWorks built for the SM network targeting the Motorola mv2604 board. All I did on top of the working vxWorks configuration was to include INCLUDE_PROXY_SERVER. Is there a fix or a way to turn this off? Minbo GDRS From vxwexplo-errs@csg.lbl.gov Tue Jan 25 04:04:39 2000 From: Vxworks Exploder Date: Tue Jan 25 04:04:42 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jan 25 04:04:36 PST 2000 Subject: Re: posix semaphore Subject: Re: iosDrvInstall() execs all methods Subject: Re: Max socket allowed Subject: DCHP only for IP address Subject: Re: VxSim on Solaris-7 (5.7) Subject: How does async notify work? Re: CrossWind breakpoints not working Subject: stack trace Subject: Using system-debug mode with VxSIM Subject: Structure size different between modules Subject: Re: Multiple Mailbox Interrupts On MVME167 Subject: Re: How to add the new END driver to Tornado Project. Subject: rpc library Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Subject: Loading symbols ONLY Subject: mvme1603 boot problem Subject: \r\n in .wsp and .wpj files Subject: VxWorks in Japanese Subject: Re: mvme1603 boot problem Subject: How venerable is VxWorks?? Subject: Re: DCHP only for IP address Subject: About "select" question? Subject: Re: adding files to Tornado 2 project? Subject: Re: adding files to Tornado 2 project? ------------------------------------------------------- Newsgroups: comp.unix.programmer,comp.os.vxworks Subject: Re: posix semaphore Date: Thu, 20 Jan 2000 18:32:34 GMT From: LiJia Jin Message-ID: <388754EC.1BC80E48@yahoo.com> References: <38852958.9126D1F0@yahoo.com> <866h6g$ki0$1@info.service.rug.nl> Fred Zwarts wrote: > "LiJia Jin" wrote in message news:38852958.9126D1F0@yahoo.com... > > I am trying to simulate VxWorks semLib by using POSIX semaphore > > calls. If anyone has done this before, please give me some hints here. > > > > Is there a standard way of doing this? For example, the POSIX > > semaphore calls don't include a function of waiting on time-out. The > > solutions I got are using usleep or nanosleep to wait. Is this the right > > way to handle this? > > > > Thanks > > > > I used a combination of POSIX semaphores and POSIX conditions. > POSIX has a wait-for-condition with a time-out. would you mind point out what's the POSIX conditions calls? Thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: iosDrvInstall() execs all methods Date: Thu, 20 Jan 2000 13:06:17 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <38875CA9.73FEA43F@aps.anl.gov> References: <867bju$fai$1@snipp.uninett.no> Geir Frode Raanes wrote: > > With the following install routine: > > iosDrvInstall( (FUNCPTR) NULL, > (FUNCPTR) NULL, > (FUNCPTR) testDrv_open(), (FUNCPTR) NULL, > (FUNCPTR) testDrv_read(), > (FUNCPTR) testDrv_write(), > (FUNCPTR) NULL); > > I get the following log from the printf() > in my dummy methods: > > ********************************************** > Entering testDrv_Init... > > Entering testDrv_open() with argument: > funcArg_DevDescr: 25 > funcArg_DevName: 4146, (null) > funcArg_DevMode: 25 > > Entering testDrv_read() with argument: > funcArg_DevDescr: 25 > user_buffer: 4146, (null) > StrLen: 1 > > Entering testDrv_read() with argument: > funcArg_DevDescr: 129 > user_buffer: 4146, (null) > StrLen: 1 Syntax misunderstanding in your 'C' code: Your install routine above is *not* passing the address of your testDrv_open routine to iosDrvInstall(), it's passing the return value it gets from calling that routine - ie it's your code that's calling these functions, not vxWorks. Take the () off the ends of the three routine names to make them into pointers to the functions. - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max socket allowed Date: Mon, 24 Jan 2000 12:11:18 -0700 From: "Russell A. Johnson" Message-ID: <388CA3D6.A96D24B0@bigfoot.com> References: <85vk4i$qpg@journal.concentric.net> <388480E5.69A05525@bigfoot.com> <388cc5d0.0@dnews.sbell.com.cn> Put the following in config.h and rebuild your BSP. #ifdef NUM_FILES /* Redefine number of files to 256 */ #undef NUM_FILES #endif #define NUM_FILES 256 /* max 256 files open simultaneously */ Russ Zha Junhai wrote: > > Hi, > Could you tell me how you override it? > > > If I remember correctly, the max value is tied to the number of file > > handles defined. We override that in config.h for our purposes. --------------------------- Newsgroups: comp.os.vxworks Subject: DCHP only for IP address Date: Thu, 20 Jan 2000 19:27:07 GMT From: gonflyn@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <867nhk$hkc$1@nnrp1.deja.com> Hi, I am trying to get a PC based target to boot from its flash disk (both the bootrom.sys and vxworks image are in flash) but I want to have the target IP address assigned from a DHCP server. If I configure the boot line to boot from the network device, the target fails to load because it didn't receive a valid boot host and file name. If I configure the boot line to boot from the flash disk, it never sends the DHCP request to discover its ip address and vxworks fails to attach the network device due to no ip address. How can I get it to boot from flash and only request an IP address from the DHCP server? Am I going to have to write some custom code for this? Thanks for any help. - -Randy Smith Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxSim on Solaris-7 (5.7) Date: 20 Jan 2000 19:29:09 GMT From: toni@devsoft.devsoft.com (Anton Hartl) Organization: Customer of SpaceNet GmbH Message-ID: References: <865h9f$u64$1@nnrp1.deja.com> In article <865h9f$u64$1@nnrp1.deja.com>, tttt9999@my-deja.com wrote: >Does anyone successfully use VxSim (Tornado 2.0) on Solaris-7 ? > >I have not been able to get it to work at all. WRS' technical support >was not helpful at all, and a lot of days has been wasted. I am not >sure if it is ever test on Solaris-7 ... > >To make a long story short, here is a summary what I tried : [...] >2. SO they suggested me to try using PPP. It turned out that their > supplied instructions and config files was for pre-Solaris-7, > which uses "asppp" instead of PPP 3.0 (which is the prefered > PPP version Sun suggested to use). In fact, in /etc/rc2.d, they > intentionally disabled the starting of the "asppp" services. > So I have to maually start "asppp" and it seems to be up and running. Are you trying to use the bundled PPP stuff that comes with Solaris 2.7? The bundled version does not work if you are booting the 64-bit Solaris kernel; you have to boot the 32 bit kernel in order for the *bundled* version of PPP to work. You can change the boot kernel in the boot monitor of your Sparc although I don't remember what the command is. Regards, -Toni - -- /dev Software Gmbh | Anton Hartl | Programming is like sex: Gabelsbergerstr. 51 | Phone: +49 (0)89-28659843 | one mistake and you have 80333 Muenchen | Fax: +49 (0)89-28659845 | to support for a lifetime. Germany | Email: toni@devsoft.com | --------------------------- Newsgroups: comp.os.vxworks Subject: How does async notify work? Re: CrossWind breakpoints not working Date: Thu, 20 Jan 2000 20:22:08 GMT From: marky_mark@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <867qor$keu$1@nnrp1.deja.com> References: <85vqmt$nmu$1@nnrp1.deja.com> Thought I would lend more food for thought here in hopes that somebody can help me in understanding this problem. The problem simply is that the CrossWind debugger tool is not being notified of the breakpoint by the target agent when it occurs. I see from the target shell that the task I'm debugging is suspended. Only when I send some RPC request to the agent (when starting a tool like WindSh or Browser, or executing some command from one of these tools), does the debugger update itself with the breakpoint event. According the the Tornado 2.0 API manual, page 38 "Normally the agent only sends data to the host in response to RPC requests, The convention is that if the host receives data when it is not waiting for a reply, it means that an event is pending on the target". Well it doesn't appear that my host is receiving data other than RPC replies. And if it is it is not occuring asynchronously. Any help would be appreciated. Of course I've contacted customer support at WRS. Thanks! Mark D. In article <85vqmt$nmu$1@nnrp1.deja.com>, marky_mark@my-deja.com wrote: > Using Tornado 2.0/SunOs 5.6 host, VxWorks 5.4/PPC603E target over > network connection. > > In CrossWind, I can attach to a task and can set breakpoints in the task > function. Looking at Back End log, I can see the event points being > added. I "continue" task expecting to see breakpoint in debugger > window, but only see the debugger's task indicator running. In the > shell, I see that the task has stopped. When I "interrupt" task, > debugger program counter is at my breakpoint. When I continue at this > point I hit the breakpoint (verified by WDB_EVENT_GET in Back End log). > I can single step at this point, though "stepping over" function calls > will 9 times out of 10 cause task indicator to show "running" task again > with no update to debugger window. > > I have simplified application and increased priority of task I wish to > debug. This has made it better, but I still have to interrupt task to > see first breakpoint, and single step is still unreliable. Task is not > being created with "VX_UNBREAKABLE" option. > > Reading the Tornado API Programmer Guide, I know that the target queues > asynchronous events like breakpoints until the host grabs them with the > WDB_EVENT_GET. I want to verify that: A.) the target is queueing these > events, and B:) the host is receiving something from the target > indicating that events are pending. > > Any ideas on how to do this, or advice on this problem would be greatly > appreciated. Thanks! > > Mark D. > > Sent via Deja.com http://www.deja.com/ > Before you buy. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: stack trace Date: Mon, 24 Jan 2000 11:28:16 -0800 From: Mike Johnston Organization: TRW, Inc. Message-ID: <388CA7D0.5B04AA7B@trw.com> is there a vxworks call I can make to get this programmatically? --------------------------- Newsgroups: comp.os.vxworks Subject: Using system-debug mode with VxSIM Date: Thu, 20 Jan 2000 15:27:01 -0500 From: steven.offenbacher@jhuapl.edu Organization: Johns Hopkins University Applied Physics Lab, Laurel, MD, USA Message-ID: <38876F94.33D27407@jhuapl.edu> How does one use the system-debug mode with VxSIM (bundled product) ? All attempts have failed to halt/suspend all tasks. --------------------------- Newsgroups: comp.os.vxworks Subject: Structure size different between modules Date: Mon, 24 Jan 2000 14:03:25 -0600 From: "Daniel Scheurell" Organization: Chorus Netwoks Message-ID: <86ib6e$njo$1@news.chorus.net> References: <86hqh4$qr1$1@news.netvision.net.il> Sender: reguser@etc-proxy.etcconnect.com We are getting a different size for one of our classes when it is compiled in a module that #includes . Two different modules in the same program then give a different size. We are using Tornado 1.0.1 for i960jx. We have a custom BSP, based on the Cyclone EP960JX. The good part is that we shipped an example to Wind River, they were able to build it with the ep960jx BSP and run it in the debugger with no problem. Yet we can do the same thing, build it with the ep960jx BSP, then run it in the debugger and get the error (on our target hardware.) The same code builds and gives consistent results under Tornado 2, though we've only tested that with the x86 BSP. We would cretainly appreciate any insight into this problem. Even shared commiseration would be a comfort. Thank you, Dan - -- - ----------------------- Daniel Scheurell dscheurell@etcconnect.com Electronic Theater Controls http://www.etcconnect.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Multiple Mailbox Interrupts On MVME167 Date: Mon, 24 Jan 2000 14:02:03 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <388CAFBB.66B509C8@aps.anl.gov> References: <86cahk$6sf4@overload.lbl.gov> "Morris, John" wrote: > > I'm in need of two mailbox interrupts on a MVME167. (That is, the 167 will > be interrupted by other cards via mailbox.) I am using VxWorks 5.1. I've > seen one defined and I can make it function perfectly. Is there a way to > easily set up another (preferably of the same type : a one-byte write)? I > know the card itself has the capability of 4 mailbox interrupts. Any help > will be GREATLY appreciated. If you're not using the backplane/shared memory networking then you can use the interface that the mv167 BSP defines for if_sm to use as your first mailbox (from your message I suspect you may already be doing this). This sysMailbox code (found in /src/drv/vme/pcc2Vme.c) can also be used as a template to extend to the other mailbox interrupts that the VMEchip2 includes. If you haven't already got it then also look at the Programmers Reference Guide for the MVME167 (available on the MCG web-site), the VMEchip2 is described in detail therein. - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to add the new END driver to Tornado Project. Date: Thu, 20 Jan 2000 18:10:04 +0800 From: "Zhang Zhiqun" Organization: The news server of Changzhou TeleCom Message-ID: <866mlv$2b76$1@news.cz.js.cn> References: <85n8e6$hdv$1@news.cz.js.cn> <85spup$jii$1@nnrp1.deja.com> <85svhp$hbs$1@news1.xs4all.nl> <85u049$dop$1@nnrp1.deja.com> Ah Sorry, I forget to tell you that I just made a virtual RTL drivers. But I have a plan that is: 1. Use the virtual driver to let the Tornado IDE has the RTL END Driver component. 2. Use the virtual driver to let the Tornado IDE has the RTL BSD Driver component. Of course, these driver can not work correctly. 3. Port the RTL's FreeBSD driver (BSD Compatible) to the VxWorks BSD Driver. 4. Upgrade the successful RTL VxWorks BSD Driver to END. - -- Thanks and best regards. Zhang Zhiqun (Bula Cheung) bula@guoguang.com.cn Guoguang Electronic Group Co. http://www.guoguang.com.cn 93 Qingtan Road, Changzhou, Jiangsu, 213015, China Tel: +86 519 696 7935 ext. 2372 Fax: +86 519 696 9104 wrote in message news:85u049$dop$1@nnrp1.deja.com... > By the way if any one of you have VxWorks drivers for the Realtek > Network chipsets (RTL8019AS, RTL8029, RTL8139, etc which uses registers > to set the media type rather than jumpers), could you please provide me > and I will be much thankful. I had lot of pain making the NE2000 > drivers built in Tornado II distribution but all in vain. > > Regards, > > Valli. > > In article <85svhp$hbs$1@news1.xs4all.nl>, > johan.borkhuis@peektraffic.nl wrote: > > mrvalli@my-deja.com wrote: > > : You have to modify the component descriptor files and add the > > : configlettes. This file is the one like this : > > : \tornado\target\config\\00bsp.cdf. Modify the file suitably > > : taking examples as the existing ones. > > > > If you take a look at my page, there is an example on how to > implement new > > devices in the configlettes. > > > > Groeten, > > Johan > > > > -- > > o o o o o o o . . . ___________________________________ > > o _____ || Johan Borkhuis | > > .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | > > >(________|__|_[_________]_|________________________________| > > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > > === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html > === > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.linux.misc,comp.os.linux.networking,comp.os.vxworks Subject: rpc library Date: Thu, 20 Jan 2000 17:00:03 -0500 From: Sakir Yucel Organization: University of Delaware, EE/CIS Lab Message-ID: Hello all, I am looking for a lightweight rpc library that runs on link layer protocols such as hdlc (for example in a local ethernet network) without requiring tcp/ip stack. Basically, I would like to eliminate tcp/ip stack and work closely on the ethernet interface for performance reasons. Any pointers will greatly be appreciated. Thanks, Sakir --------------------------- Newsgroups: comp.sys.powerpc.tech,comp.os.vxworks Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Date: Mon, 24 Jan 2000 14:25:03 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <388CB51F.80BD80F3@aps.anl.gov> References: <86i0pb$m0s$1@nnrp1.deja.com> John Fusco wrote: > > What I have found is that if I don't clear the FPSCR (and disable > floating point exceptions), I can't return to my code. That is, the > exception handler is called continuously once an exception occurs, > unless I disable all floating point exceptions from the exception > handler. > > I tried to clear the FPSCR status bits to clear the exception, but this > does not work. I only return to my code my exception handler sets the > FPSCR to zero. > > Am I missing something? My guess (I'm not an expert on PPC) is that when it returns from your exception handler, the CPU retries the FP instruction that caused it to fail. Unless you've changed the conditions that caused the exception, this retry is just going to make it happen again. If I'm right, then you need to get that FP instruction executed with exceptions turned off, then turn them on again for the next instruction. Easier said than done, but possible. - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Loading symbols ONLY Date: Thu, 20 Jan 2000 17:23:20 -0500 From: "Gutknecht, Bill [NC1:3T62:EXCH]" Organization: Nortel Networks Message-ID: <38878AD7.FEF7ED7F@nortelnetworks.com> I've got a vxWorks.st image loaded on my board, I connect with Tornado/target manager and then want to run either windsh or crosswind, but am having a hard time figuring how to load the "symbols only" to the target server from my object files. Is there a way? Thx, Bill - -- Bill Gutknecht "If I die, I must go before [Crom] and he billag@nortelnetworks.com will ask me 'What is the Riddle of Steel?' Embedded Software Engineer If I do not know it, he will cast me out Nortel Networks of Valhalla and laugh at me ..." - Conan --------------------------- Newsgroups: comp.os.vxworks Subject: mvme1603 boot problem Date: Thu, 20 Jan 2000 20:05:55 -0500 From: "Frank Martinez" Organization: AT&T WorldNet Services Message-ID: <868bf3$pu0$1@bgtnsc01.worldnet.att.net> I am having trouble getting my MVME1603 module to boot up VxWorks. I have followed the directions in the target.txt file and the module will boot up with the VxWorks banner over the serial port but I cannot connect with my shell. I get a timeout error when I try to connect the target server. I have gotten a MVME2604 module up and running with no problem but the process for the 1603 is different and I am at a loss what to try next. I am running Tornado and VxWorks 5.3.1. thanks, Frank --------------------------- Newsgroups: comp.os.vxworks Subject: \r\n in .wsp and .wpj files Date: Thu, 20 Jan 2000 20:22:00 -0800 From: Stas Maximov Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3887DEE8.DDC50D1B@pigeonpoint.com> Hi! I am using WinCVS and I want to put my TII project under CVS control. Tornado workspace (.wsp) and project (.wpj) files are plain text files, but use single LF character as a line separator instead of CRLF. If we replace LF separators with CRLF (it usually happens with WinCVS), Tornado complains that it has "Problem opening C:/Tornado/target/proj/vxplay/Workspace.wsp: Workspace open failed: This project workspace is an older format". - - I want to treat .wsp and .wpj files as text files. At least for storing just minimal changes and being able to perform diffs on them; - - I want to check out text files as appropriate for the platform I am using, i.e. CRLF for Windows; Question: Is there a Tornado option that makes Tornado happy about CRLN in .wsp and .wpj? I have searched in WRS known problems support area and could not find anything related. Thanks, Stas. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks in Japanese Date: Tue, 25 Jan 2000 00:09:50 GMT From: "S. Virgo" Organization: MetroNet Communications Group Inc. Message-ID: Does anyone know if the VxWorks OS comes in the Japanese language? Thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: mvme1603 boot problem Date: Fri, 21 Jan 2000 05:28:17 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <868qp4$b95$1@nnrp1.deja.com> References: <868bf3$pu0$1@bgtnsc01.worldnet.att.net> Having the VxWorks banner (as opposed to the smaller WDB ready banner) would suggest that you might not have WDB configured into your kernel. Are you building a kernel with INCLUDE_CONFIGURATION_5_2 defined? If so, you either need to undefine that macro, or add back the INCLUDE_WDB_xxx macros that it causes to be undefined (see target/src/config/usrDepend.c for details). Hope that helps, John... In article <868bf3$pu0$1@bgtnsc01.worldnet.att.net>, "Frank Martinez" wrote: > I am having trouble getting my MVME1603 module to boot up VxWorks. I have > followed the directions in the target.txt file and the module will boot up > with the VxWorks banner over the serial port but I cannot connect with my > shell. I get a timeout error when I try to connect the target server. > > I have gotten a MVME2604 module up and running with no problem but the > process for the 1603 is different and I am at a loss what to try next. > > I am running Tornado and VxWorks 5.3.1. > > thanks, > Frank > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: How venerable is VxWorks?? Date: Fri, 21 Jan 2000 12:04:19 +0530 From: VijayKumar P Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <3887FDEB.DA0A6545@cisco.com> This is a multi-part message in MIME format. - --------------0FC9DF581F245D0041422CF6 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi All, I was trying to compare different RTOSs. Have a doubt... With all the non-real-time services like TCP/IP, graphics display and windowing systems, file and data base systems etc., added to the kernel, does it make vxWorks venerable?. Or do you think a completely pre-emptable kernel (as in RT-IX(Modcomp)) is a better design?. Or how about a simple real-time executive running on a non-real-time kernel as its lowest priority task, using a virtual machine layer to make the standard kernel fully pre-emptable as in RT-Linux?? Thanks, Vijay - --------------0FC9DF581F245D0041422CF6 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for VijayKumar P Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India note;quoted-printable:Seldom does an individual exceed his own expectations=0D=0A=0D=0AWishing you a most beautiful day wherever you may be! fn:Vijay Kumar.P end:vcard - --------------0FC9DF581F245D0041422CF6-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: DCHP only for IP address Date: Fri, 21 Jan 2000 12:18:46 +0530 From: VijayKumar P Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <3888014E.984EBEE8@cisco.com> References: <867nhk$hkc$1@nnrp1.deja.com> This is a multi-part message in MIME format. - --------------A2A5152BB744C33568DA6612 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, You can do it in two ways.. 1. Include the DHCP runtime client option in your image. Tornado>Workspace>vxWorks>network components> network applications>DHCPv4 runtime client Here when you boot from flashdisk, the DHCP component will be included and 'everything' will be taken care of by vxWorks. 2.If you really want to start DHCP at a later stage, say some time in your application, then just use the DHCP calls provided by vxWorks. The complete procedure has been explained in detail ( with an example!) in the Network Programmer's Guide, Chap 5. Good luck, Vijay gonflyn@my-deja.com wrote: > Hi, > > I am trying to get a PC based target to boot > from its flash disk (both the bootrom.sys and > vxworks image are in flash) but I want to have > the target IP address assigned from a DHCP server. > If I configure the boot line to boot from the > network device, the target fails to load because > it didn't receive a valid boot host and file > name. If I configure the boot line to boot from > the flash disk, it never sends the DHCP request > to discover its ip address and vxworks fails to > attach the network device due to no ip address. > > How can I get it to boot from flash and only > request an IP address from the DHCP server? Am I > going to have to write some custom code for this? > > Thanks for any help. > > -Randy Smith > > Sent via Deja.com http://www.deja.com/ > Before you buy. - --------------A2A5152BB744C33568DA6612 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for VijayKumar P Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India note;quoted-printable:Seldom does an individual exceed his own expectations=0D=0A=0D=0AWishing you a most beautiful day wherever you may be! fn:Vijay Kumar.P end:vcard - --------------A2A5152BB744C33568DA6612-- --------------------------- Newsgroups: comp.os.vxworks Subject: About "select" question? Date: Fri, 21 Jan 2000 13:12:46 +0800 From: "lips" Organization: The news server of Changzhou TeleCom Message-ID: <8693rr$1fml$1@news.cz.js.cn> Reply-To: "lips" Hi all: I works Server/Client use select,but I come into contact with quaint question: (initialize a socket sith) #define BUFFER_SIZE 1024 ... fd_set rfds; fd_set wfds; while(TRUE) { FD_ZERO(&rfds); FD_ZERO(&wfds); FD_SET(socket,&rfds); FD_SET(socket,&wfds); if(select(socket+1,&rfds,&wfds,NULL,NULL)<0) { printf("select failed!\n"); } else { if(FD_ISSET(socket,&rfds)) { printf("RecvData runing!\n"); if(recv(socket,recvBuffer,BUFFER_SIZE,0)== ERROR) { printf("The ReadHost failed!\n"); } } if(FD_ISSET(socket,&wfds)) { printf("SendData runing!\n"); if(send (socket,sendBuffer,BUFFER_SIZE,0) ==ERROR) { printf("The WriteHost failed!\n"); } } } } ... ÎÒÒѾ­¶¨ÒåÁËrecvBufferºÍsendBuffer.µ«ÊÇÔËÐеĽá¹ûÊÇÁ¬ÐøÑ­»·ÔËÐÐÁË·¢Ëͺ¯Êý£¬ ¶ø½ÓÊÕº¯ÊýûÓÐÔËÐУ¡¾ÍÁ¬SendData runing!¶¼Ã»ÓÐÏÔʾ£¡µ«ÊÇÈç¹ûÖ»¼ÓÈëÒ»¸ö¾ä±ú £¨rfds»òwfds)£¬¶ÔÓÚrfds£¬½ÓÊܺ¯ÊýÔËÐÐÁË£¬¶ÔÓÚwfds£¬·¢Ëͺ¯ÊýÒ²ÔËÐÐÁË£¬µ«ÊÇÁ½ ¸ö¾ä±ú¶¼¼ÓÈëʱ£¬ÎªÊ²Ã´Ö»ÔËÐз¢Ëͺ¯Êý£¿£¿£¿ лл£¡ lips@guoguang.com.cn --------------------------- Newsgroups: comp.os.vxworks Subject: Re: adding files to Tornado 2 project? Date: Fri, 21 Jan 2000 13:31:02 +0530 From: VijayKumar P Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <3888123E.85A47935@cisco.com> References: <864ke8$9gf4@overload.lbl.gov> <867bfh$3s9$1@news1.xs4all.nl> This is a multi-part message in MIME format. - --------------E836E99F876448467B6E2878 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, johan.borkhuis@peektraffic.nl wrote: > Reyhan ERGUN wrote: > > : I have some number of files which are generated by another tool > : (Telelogic SDT), > : and I want to link them all to my vxWorks build. > : I will compile and generate object files for these extra codes within T2 > : (UNIX and WinNT) and add all these files to my Tornado II project, but I > : am looking for a nicer way to do this. > : There is a MACRO as EXTRA_MODULES Better still if you open your project's makefile itself and assign the path of these object files to EXTRA_MODULES = You can build a separate project for the tool generated files and compile them there. And then use those object files generated here as above. > > : and I added a definition as /*.c for this macro > : but it doesn't work. > > Have you tried adding them to your project? If not, why? If so, what were > the problems you had? > Yes this should also work. Vijay - --------------E836E99F876448467B6E2878 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for VijayKumar P Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Peshkar;Vijay tel;fax:+91-80-5537663 tel;work:+91-80-5530035,34,36 Extn: 3451 x-mozilla-html:FALSE org:WIPRO Technologies;Global R&D Solutions version:2.1 email;internet:vijay.peshkar@wipro.com title:Senior Engineer - Software adr;quoted-printable:;;Vijay Kumar .P, WIPRO Technologies, #8, 7th Main, 1st Block, Koramangala=0D=0A;Bangalore;Karnataka;560 034;India note;quoted-printable:Seldom does an individual exceed his own expectations=0D=0A=0D=0AWishing you a most beautiful day wherever you may be! fn:Vijay Kumar.P end:vcard - --------------E836E99F876448467B6E2878-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: adding files to Tornado 2 project? Date: Fri, 21 Jan 2000 08:05:34 GMT From: findik@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <869406$hef$1@nnrp1.deja.com> References: <864ke8$9gf4@overload.lbl.gov> <867bfh$3s9$1@news1.xs4all.nl> I tried adding them all to my project, it worked but I was wondering an easier way to add them all to my project i.e. I may direct a new directy only and tell T2 that you may get all *.c files under this directory so that if I generate a new c file from SDT I won't bother about adding this file to T2 project, it will be added automatically since it is already under that source file directory. As I said I added all these file to my project by using add files utility, I encountered new problems as follows: (I am using T2 under winNT and I haven't come accross with such problems) under UNIX, with T2 and while making a vxWorks build, I have 2 problems while generating external dependency: 1-) I am doing a simple trial and including some dummy header files in usrAppInit.c, after generating dependecy for all project files or for this file only, it doesn't make any change, I couldn't see those dummy header files under external dependency tap in the workspace and I can't see those header files in Makefile ad dependency for usrAppInit.o, under vxWorks properties application initialization macro is included! 2-) I add a new file to the project, I can see the header files under ext. dependecies. I added path for this file to C/C++ Compiler options, but I see that "file not found" under the property for this header file. Is there any immaturity for T2 under UNIX? Has anybody encountered such problems? Thanks for your valuable comments on the subject. Best regards, reyhan > Have you tried adding them to your project? If not, why? If so, what were > the problems you had? > > Groeten, > Johan > > -- > o o o o o o o . . . ___________________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | > >(________|__|_[_________]_|________________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jan 25 08:05:36 2000 From: "Tyler, Paul A (DBY)" Date: Tue Jan 25 08:05:39 PST 2000 Subject: Re: Getting TaskID in a TaskSwitchHook The task ID is the address of the Task Control Block (WIND_TCB). The clue is in the macros TASK_ID_VERIFY, TASK_LOCK, and so on defined in taskLib.h VxWorks, Tornado Paul Tyler Measurement Systems Specialist Electronics & Measurement Systems Dept. Rolls-Royce plc. e-mail Paul.A.Tyler@Rolls-Royce.com Tel. +44 (1332) 247808 Fax. +44 (1332) 247928 John Beauford wrote in message news:<388C7A69.1E6C33EB@email.mot.com>... > Has anyone used a switch hook and gotten the taskID(of either old task > or new task) inside the switch hook? > If you could enlighten me I would appreciate it. > > > -- > John Beauford From vxwexplo-errs@csg.lbl.gov Tue Jan 25 08:13:26 2000 From: Doug Owens Date: Tue Jan 25 08:13:28 PST 2000 Subject: Re: comp.os.vxworks newsdigest Seven copies - I didn't subscribe to be spammed. From vxwexplo-errs@csg.lbl.gov Tue Jan 25 08:37:36 2000 From: Jim Way Date: Tue Jan 25 08:37:38 PST 2000 Subject: Looking for information on Ethernet with VxWorks VxWorkers, I'm using VxWorks for a PowerPc (PPC) with Ethernet. We're seeing some interesting behavior (see below) with the units when they are connected directly to our office network (Microsoft NT servers, email servers, a Linux box or two, typical office environment stuff). When we connected our target directly to a pc through a hub, bypassing the network, it ran ok. We set up another test yesterday to connect our target to the office network through a hub (rather than a direct-connect) and as of this morning it is still running strong. Does anyone have any ideas, advice, or suggestions for what might be going on and how we might be able to fix it? ALL comments will be greatly appreciated. Thanks! Jim Way ------------------------- The Interesting Behavior: 1. Monitoring two units via Telnet sessions, both died at the same time (observed one time). 2. Monitoring a unit over the weekend, beginning about 4pm on Friday, the unit ran fine until about 8:30am Monday morning. 3. Units connected directly to the network fail, average time is less than four hours. 4. The software continues to run, but anything involving Ethernet, such as Telnet sessions, goes away. ------------------------- From vxwexplo-errs@csg.lbl.gov Tue Jan 25 10:25:07 2000 From: Mitch Hendrickson Date: Tue Jan 25 10:25:10 PST 2000 Subject: RE: comp.os.vxworks newsdigest 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_01BF6761.606557EA Content-Type: text/plain; charset="iso-8859-1" > We are getting a different size for one of our classes when it is > compiled in a module that #includes . Two > different modules in > the same program then give a different size. > We are using Tornado 1.0.1 for i960jx. We have a custom > BSP, based on > the Cyclone EP960JX. > The good part is that we shipped an example to Wind > River, they were > able to build it with the ep960jx BSP and run it in the > debugger with no > problem. Yet we can do the same thing, build it with the > ep960jx BSP, then > run it in the debugger and get the error (on our target hardware.) > The same code builds and gives consistent results under Tornado 2, > though we've only tested that with the x86 BSP. They probably didn't duplicate it under Tornado 1. That would be expected for WRS support, in my experience. In any case, the problem is very likely the fact that radix.h came with SENS, and is included eventually when you include inetLib.h, and it (radix.h) has a #pragma align *before* it includes mbuf.h. This means that when mbuf.h also does a #pragma align, the original alignment value (16) is lost forever, and anything after that point will be aligned differently. Where I first saw it was in the shared memory interface. I reported this to WRS about 14 months ago (TSR 120283) and although we got a fix for our particular case, it appeared to disappear into a black hole. About a year later, they apparently listened to me or someone else, though, as a patch was released under SPR 24065. It's on their website and goes under the title "SENS1.1 i960 Shared Memory Network Fails." I seem to recall that it's fixed in T2 (we Beta'ed it), but I can't check for sure, as we're not running T2 yet. This is a 960-only problem. I hope this helps. -Mitch -- Mitch Hendrickson Sr. Engineer Telestream, Inc. http://www.telestream.net ------_=_NextPart_001_01BF6761.606557EA Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable RE: comp.os.vxworks newsdigest

>     We are getting a = different size for one of our classes when it is
> compiled in a module that #includes = <inetLib.h>.  Two
> different modules in
> the same program then give a different = size.
>     We are using Tornado = 1.0.1 for i960jx.  We have a custom
> BSP, based on
> the Cyclone EP960JX.
>     The good part is that = we shipped an example to Wind
> River, they were
> able to build it with the ep960jx BSP and run = it in the
> debugger with no
> problem.  Yet we can do the same thing, = build it with the
> ep960jx BSP, then
> run it in the debugger and get the error (on = our target hardware.)
>     The same code builds = and gives consistent results under Tornado 2,
> though we've only tested that with the x86 = BSP.

They probably didn't duplicate it under Tornado = 1.  That would be expected for WRS support, in my experience. =

In any case, the problem is very likely the fact that = radix.h came with SENS, and is included eventually when you include = inetLib.h, and it (radix.h) has a #pragma align *before* it includes = mbuf.h.  This means that when mbuf.h also does a #pragma align, = the original alignment value (16) is lost forever, and anything after = that point will be aligned differently.  Where I first saw it was = in the shared memory interface.

I reported this to WRS about 14 months ago (TSR = 120283) and although we got a fix for our particular case, it appeared = to disappear into a black hole.  About a year later, they = apparently listened to me or someone else, though, as a patch was = released under SPR 24065.  It's on their website and goes under = the title "SENS1.1 i960 Shared Memory Network = Fails."

I seem to recall that it's fixed in T2 (we Beta'ed = it), but I can't check for sure, as we're not running T2 yet.

This is a 960-only problem.

I hope this helps.

-Mitch

--
Mitch Hendrickson
Sr. Engineer
Telestream, Inc.
http://www.telestream.net
 

------_=_NextPart_001_01BF6761.606557EA-- From vxwexplo-errs@csg.lbl.gov Tue Jan 25 10:44:48 2000 From: matownse@collins.rockwell.com Date: Tue Jan 25 10:44:50 PST 2000 Subject: Re: comp.os.vxworks newsdigest Seven? Consider yourself lucky, I got 11!!!!! (so far) From vxwexplo-errs@csg.lbl.gov Tue Jan 25 15:32:19 2000 From: Mario Gallotta Date: Tue Jan 25 15:32:21 PST 2000 Subject: Vxworks printErrno I am using library function "printErrno". I have included the following libraries: errnoLib.h usrLib.h when I download my code to the target I get unresolved symbol, printErrno. Here are the list of vxWorks include files: #include "vxWorks.h" #include "taskLib.h" #include "fioLib.h" #include "blkIo.h" #include "ioLib.h" #include "iosLib.h" #include "memLib.h" #include "semLib.h" #include "msgQLib.h" #include "stdio.h" #include "msgQLib.h" #include "logLib.h" #include "string.h" #include "errnoLib.h" #include "usrLib.h" Regards, ************************************************ Mario Gallotta Centerpoint Broadband Technologies 1731 Technology Drive Suite 610 San Jose, CA 95110 Phone - 408-452-2918 Cellular - 408-391-7479 Fax - 408-436-0982 mailto:mgallotta@cptbt.com ************************************************ From vxwexplo-errs@csg.lbl.gov Wed Jan 26 04:03:17 2000 From: Vxworks Exploder Date: Wed Jan 26 04:03:20 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jan 26 04:03:13 PST 2000 Subject: Re: SNMP setting variables Subject: Re: How venerable is VxWorks?? Subject: Re: SNMP setting variables Subject: *** No rule to make target Subject: Motorola eval boards- Subject: Re: \r\n in .wsp and .wpj files Subject: Re: *** No rule to make target Subject: CrossWind question Subject: Re: \r\n in .wsp and .wpj files Subject: National Instruments VISA Subject: Mouse input Subject: Re: How do I boot x86 without BIOS Subject: Re: SDS Single Step vs. EST VisionProbe debug tools Subject: FTP Server for vxWorks Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Subject: Re: *** No rule to make target Subject: Re: How do I boot x86 without BIOS Subject: Re: comp.os.vxworks newsdigest Subject: Re: when not compiled works fine. Subject: Re: *** No rule to make target Subject: excVecInit()? Subject: Re: taskLock & taskUnlock Subject: Re: Questions about TFFS Subject: Re: send() doesn't fail? Subject: Re: CrossWind question Subject: Display vt100 escape codes from VxSim Lite? Subject: Re: Looking for information on Ethernet with VxWorks Subject: Re: when not compiled works fine. Subject: A Question about BSP Subject: Re: Builds in T2 Subject: Re: SNMP : Entries in Interfaces MIB Subject: Re: File dependencies Subject: pipes Subject: Getting TaskID in a TaskSwitchHook Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Subject: Re: excVecInit()? Subject: Re: CrossWind question Subject: Re: System time in vxWorks Subject: Re: System time in vxWorks Subject: Re: Getting TaskID in a TaskSwitchHook Subject: Re: Mouse input Subject: Boot problem with VxWorks 5.2 and ADS8260 board Subject: Re: Network Booting on a Pentium PC104+ board Subject: Display drivers Subject: Re: A Question about BSP Subject: Re: Boot problem with VxWorks 5.2 and ADS8260 board Subject: Re: How do I boot x86 without BIOS Subject: Re: posix semaphore Subject: Tornado1.0.1 and Solaris 2.7 Subject: Re: Vxworks printErrno ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: SNMP setting variables Date: Fri, 21 Jan 2000 13:28:35 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <869l9h$8u5$1@ndsukns01.ndsuk.com> References: <948452669.386608@big.hacom.nl> Hi Frank, Do you know what community name you are using in your setRequests ? Note that the default implementation in VxWorks requires "private" if you are to be given write access. As far as I remember, the WindNet agent doesn't respond if the community name is invalid, so it is quite possible that the community name being sent by OpenView is not listed in your Agent. I would have thought OV a bit top-heavy for use while developing your agent, I find MG-Soft MIB Browser a useful tool for this (it also has a good MIB editor). HTH Will Frank van Eijkelenburg wrote in message <948452669.386608@big.hacom.nl>... >Hi, > >I'm experimenting with snmp. The agent is running on a VxWorks OS. In >VxWorks I have implemented my own MIB, and the get, set, etc. routines to >operate on the variables. I'm using HP OpenView to manage my variables. > >Everything is working fine in the sense that I can read the variables in HP >OpenView. But I can't modify the variables which I set as "read-write" in >the MIB. Even the standard variables that are "read-write" can't modified >(sysName, sysLocation). The error message I get in OpenView is: "No response >from device". Since I even can't modify the standard variables I supose that >the problem is not in the set routine but somewhere else. > >If you have any suggestions......, please let me know. > >Thanks, >Frank > >-- >Please don't sent email to me, but post your answer in the newsgroup > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How venerable is VxWorks?? Date: Fri, 21 Jan 2000 08:08:56 -0600 From: Frank Minich Organization: NCP Message-ID: <38886878.4582@ti.com> References: <3887FDEB.DA0A6545@cisco.com> Reply-To: fminich@ti.com VijayKumar P wrote: > > ... does it make vxWorks venerable?. From my dictionary: "... calling forth respect through age, character, and attainments; broadly : conveying an impression of aged goodness and benevolence ..." Well, yes, I guess so. Frank. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SNMP setting variables Date: Fri, 21 Jan 2000 15:48:35 +0100 From: "Frank van Eijkelenburg" Organization: HACOM Internet - Amersfoort - The Netherlands. (033-455611) Message-ID: <948465946.665029@big.hacom.nl> References: <948452669.386608@big.hacom.nl> <869l9h$8u5$1@ndsukns01.ndsuk.com> Thanks Will, the community name for get was set to "public". I changed to "private" and it works fine now. Frank Will Fookes wrote in message <869l9h$8u5$1@ndsukns01.ndsuk.com>... >Hi Frank, > >Do you know what community name you are using in your setRequests ? Note >that the default implementation in VxWorks requires "private" if you are to >be given write access. > >As far as I remember, the WindNet agent doesn't respond if the community >name is invalid, so it is quite possible that the community name being sent >by OpenView is not listed in your Agent. > >I would have thought OV a bit top-heavy for use while developing your agent, >I find MG-Soft MIB Browser a useful tool for this (it also has a good MIB >editor). > >HTH >Will > >Frank van Eijkelenburg wrote in message <948452669.386608@big.hacom.nl>... >>Hi, >> >>I'm experimenting with snmp. The agent is running on a VxWorks OS. In >>VxWorks I have implemented my own MIB, and the get, set, etc. routines to >>operate on the variables. I'm using HP OpenView to manage my variables. >> >>Everything is working fine in the sense that I can read the variables in HP >>OpenView. But I can't modify the variables which I set as "read-write" in >>the MIB. Even the standard variables that are "read-write" can't modified >>(sysName, sysLocation). The error message I get in OpenView is: "No >response >>from device". Since I even can't modify the standard variables I supose >that >>the problem is not in the set routine but somewhere else. >> >>If you have any suggestions......, please let me know. >> >>Thanks, >>Frank >> >>-- >>Please don't sent email to me, but post your answer in the newsgroup >> >> > > --------------------------- Newsgroups: comp.os.vxworks Subject: *** No rule to make target Date: Fri, 21 Jan 2000 11:47:03 -0500 From: "John Turner" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86a2ge$601$1@soap.pipex.net> I'm sure that this is everyone's favorite problem when they first encounter Tornado v2. Even though I have provided the appropriate list of include directories (by specifying the compiler options -I/myinc1 -I/myinc2) and manually re-run the dependency generator, the Tornado IDE persists in generating an incorrect project and an incorrect makefile. Specifically it lists certain header files as dependencies but it gets their paths wrong (actually it gives no path) and consequently any attempt to build the project fails with "*** No rule to make target 'xyz.h'". Is there any way to correct this behavior other than manually editing the project file and makefile? Thanks very much, John --------------------------- Newsgroups: comp.os.vxworks Subject: Motorola eval boards- Date: Fri, 21 Jan 2000 16:36:02 GMT From: rfowler@dese.com Organization: Deja.com - Before you buy. Message-ID: <86a1th$7hk$1@nnrp1.deja.com> Wanted to find Motorola Eval board Yellowknife PPcEval-Yk-EX4 we are currently using one with PPC740 and need another unit. please reply to rfowler@dese.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: \r\n in .wsp and .wpj files Date: Fri, 21 Jan 2000 12:02:33 -0500 From: "John Turner" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86a3df$6dq$1@soap.pipex.net> References: <3887DEE8.DDC50D1B@pigeonpoint.com> Stas Maximov wrote in message news:3887DEE8.DDC50D1B@pigeonpoint.com... > Hi! > > I am using WinCVS and I want to put my TII project under CVS control. > > Tornado workspace (.wsp) and project (.wpj) files are plain text files, > but use single LF character as a line separator instead of CRLF. If we > replace LF separators with CRLF (it usually happens with WinCVS), > Tornado complains that it has "Problem opening > C:/Tornado/target/proj/vxplay/Workspace.wsp: Workspace open failed: This > project workspace is an older format". > > - I want to treat .wsp and .wpj files as text files. At least for > storing just minimal changes and being able to perform diffs on them; > > - I want to check out text files as appropriate for the platform I am > using, i.e. CRLF for Windows; > > Question: Is there a Tornado option that makes Tornado happy about CRLN > in .wsp and .wpj? > > I have searched in WRS known problems support area and could not find > anything related. > > Thanks, > Stas. Perhaps you should write a small utility to convert back and forth between LF and CRLF? Use it before checking in and after checking out. Maybe it could even become an automated part of your check-in and check-out scripts? It's not pretty, I know. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: Fri, 21 Jan 2000 13:17:28 -0500 From: "John Turner" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86a7pv$8gl$1@soap.pipex.net> References: <86a2ge$601$1@soap.pipex.net> <38889536.484F6670@t-online.de> Heribert Volkhausen wrote in message news:38889536.484F6670@t-online.de... > Hi John, > > I contacted the support hotline and they told me the following: > > There is no 'clean' way to insert the path to the include directories. > The only chance is, to use the macro VPATH. In addition I have to set > the compiler option -I for every include directory, that is > used. I have to make the settings of include directories at two > different > locations (VPATH and compiler options). There is no way to use a browser > to get these diretories. > > Regards > Heribert > > > John Turner wrote: > > > > I'm sure that this is everyone's favorite problem when they first encounter > > Tornado v2. > > > > Even though I have provided the appropriate list of include directories (by > > specifying the compiler options -I/myinc1 -I/myinc2) and manually re-run the > > dependency generator, the Tornado IDE persists in generating an incorrect > > project and an incorrect makefile. Specifically it lists certain header > > files as dependencies but it gets their paths wrong (actually it gives no > > path) and consequently any attempt to build the project fails with "*** No > > rule to make target 'xyz.h'". > > > > Is there any way to correct this behavior other than manually editing the > > project file and makefile? > > > > Thanks very much, > > John Thanks Heribert but by VPATH, do you mean an environment variable or a macro manually inserted into the Makefile? Where did you set VPATH? There is no online documentation for this (gosh, I'm surprised). --------------------------- Newsgroups: comp.os.vxworks Subject: CrossWind question Date: Fri, 21 Jan 2000 18:14:50 GMT From: marky_mark@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86a7ma$c77$1@nnrp1.deja.com> I'm running CrossWind 4.0 (GDB 4.17) on Solaris 2.6. The startup message for CrossWind says: This GDB was configured as "--host=sparc-sun-solaris2.5.1 - --target=powerpc-wrs-vxworks". Does running GDB configured for Solaris 2.5.1 on Solaris 2.6 cause problems? The reason I ask is because I am having problems with the debugger not updating when breakpoints occur (see my previous posts from 1/20). A shot in the dark, but I'm at a loss for why it's doing this. Thanks! Mark D. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: \r\n in .wsp and .wpj files Date: Fri, 21 Jan 2000 17:08:57 GMT From: bruce@figureitout.com (Bruce) Organization: COM.BOX Winet Message-ID: <38889255.2932251@212.124.35.7> References: <3887DEE8.DDC50D1B@pigeonpoint.com> On Thu, 20 Jan 2000 20:22:00 -0800, Stas Maximov wrote: >Tornado workspace (.wsp) and project (.wpj) files are plain text files, >but use single LF character as a line separator instead of CRLF. If we >replace LF separators with CRLF (it usually happens with WinCVS), >Tornado complains that it has "Problem opening >C:/Tornado/target/proj/vxplay/Workspace.wsp: Workspace open failed: This >project workspace is an older format". > >- I want to treat .wsp and .wpj files as text files. At least for >storing just minimal changes and being able to perform diffs on them; > >- I want to check out text files as appropriate for the platform I am >using, i.e. CRLF for Windows; I don't know if this is any help, but CodeWright (www.premia.com) will view, edit, and diff the files in Windows, without changing the LF format. Bruce in Katy, TX --------------------------- Newsgroups: comp.os.vxworks Subject: National Instruments VISA Date: Fri, 21 Jan 2000 19:02:32 GMT From: ksbarr@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86aafs$ed9$1@nnrp1.deja.com> Hi all, I am in the process of picking out a VME ARINC 429 board and one of the boards I am looking at (Excalibur) says it needs the National Instruments VISA. I am running VxWorks on a Motorolla 2604 board, does anyone know what VISA is and how it works? Will the addition of that screw up our other boards in the VME backplane? The company sent me their software and I don't see any mention of this VISA anywhere. I looked at the National Instruments website and it looks like VISA is a way of interfacing to Labview, but we aren't using that. Any help would be greatly appreciated. - -- Keith Barr Senior Software Engineer Coherent Technologies, Inc. keithb@ctilidar.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Mouse input Date: Tue, 25 Jan 2000 15:54:34 +0000 From: Paul Brewer Message-ID: <388DC73A.B8E3CF86@compd.com> Reply-To: pbrewer-removethis-@compd.com I have a PowerPC Single board computer running vxWorks. The application I'm running on the board takes input from a keyboard but I also want it to take input from a serial mouse. Does anyone have a driver/source code for doing this with vxWorks. Any help would be greatly appreciated. Paul Brewer. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I boot x86 without BIOS Date: Tue, 25 Jan 2000 18:55:05 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <388DD569.DBD806C7@bitband.com> References: <388D9CF7.8CB7F8B0@pacbell.net> Reply-To: leonid@bitband.com Years ago I did a BSP for a custom board with a i486 CPU, no BIOS and entirely non-PC-compatible hardware design. Was practically the same as any BSP for any architecture other ten x86. You can use the ns486 BSP delivered with Tornado as an example of a true BSP for an x86 architecture board. Leonid Chuck Hardy wrote: > > The Wind River documentation for booting Intel x86 systems seems to > imply that there has to be a BIOS to boot vxWorks. They only refer to > PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed > an x86-based system that boots vxWorks directly from ROM? If so, how do > you start on power-up? As I understand it, the x86 starts executing at > F000:FFF0. That doesn't leave much room for code, so you have to jump > somewhere else to start the boot process. Do the Wind River tools > support putting this jmp instruction into ROM, or do I have to somehow > patch it in by hand? > > Any help would be greatly appreciated. > > Thanks, > Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SDS Single Step vs. EST VisionProbe debug tools Date: Tue, 25 Jan 2000 19:35:03 +0200 From: "Shuky Persky" Organization: Bezeq International Ltd. Message-ID: <388ddec7@news.bezeqint.net> References: - -- Hi, For the early stage of the bootROM development, I recommend the AMC's CodeTAP or SuperTAP emulator (connects the MPC8xx via BDM port). Try @ amc.com Hope it helps, Shuky - ------------------------------ voice : 972-50-432266 Dave Hill wrote in message news:S7Hc4.942$9F3.66719@news.corecomm.net... > Hello All, > We are currently developing a board based on the PowerPC 860 chip. We are > using the Tornado II dev. suite. We need to purchase > a debugger. We are currently looking at the SDS Single Step product and at > EST VisionProbe. Does anyone out there have experience with either (or > both) of these products and could give me some advice?? > Thank you in advance, > Dave Hill > Safetran Systems > > --------------------------- Newsgroups: comp.os.vxworks Subject: FTP Server for vxWorks Date: Tue, 25 Jan 2000 18:08:12 GMT From: "Steeve Gagne" Organization: MetroNet Communications Group Inc. Message-ID: Hi I search for a ported FTP daemon on vxWorks to replace the default vxWorks package ... Anyone have some helpful comments for me ? thanks Steeve Gagne steeve.gagne@simplernetworks.com --------------------------- Newsgroups: comp.sys.powerpc.tech,comp.os.vxworks Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Date: Tue, 25 Jan 2000 18:53:45 GMT From: John Fusco Organization: Deja.com - Before you buy. Message-ID: <86krfo$pc6$1@nnrp1.deja.com> References: <86i0pb$m0s$1@nnrp1.deja.com> <86ii7a$qqc$1@user2.teleport.com> I tried clearing out all the sticky bits and that didn't help. It sounds like the program counter is the problem; i.e. the CPU is trying to execute the instruction again. I seem to remember reading about this so I'll dig into the manuals again. Maybe I am going about this the wrong way. I just want to isolate the cause of the exception to a particular data set, but the results in most cases are still usable (e.g. inexact or underflow). The sticky bits tell me what I need to know so maybe I don't even need to enable the exception in the MSR. Maybe I'll just let the application clear the exceptions in the FPSCR before it starts, and check for exceptions after it completes. Thanks for the info. In article <86ii7a$qqc$1@user2.teleport.com>, andrew@user2.teleport.com (Andrew Klossner) wrote: > I wrote an FP exception handler for an 8240 running VxWorks. I found > that it's necessary to AND the FPSCR register contents with 0x0007f0ff > so that the FP exception doesn't fire all over again when the next FP > instruction is executed. You don't have to disable exceptions, just > clear the outstanding reports (the status and "sticky" bits.) > > Also, when you enter the exception handler, SRR0 (the exception-time > program counter) will point to the instruction that caused the > exception. You need to add 4 to it before returning, unless your > version of VxWorks does that for you. > > -=- Andrew Klossner (andrew@teleport.com) > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: Fri, 21 Jan 2000 15:03:19 -0500 From: "John Turner" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86ae0e$b7s$1@soap.pipex.net> References: <86a2ge$601$1@soap.pipex.net> <38889536.484F6670@t-online.de> <86a7pv$8gl$1@soap.pipex.net> John Turner wrote in message news:86a7pv$8gl$1@soap.pipex.net... > > Thanks Heribert but by VPATH, do you mean an environment variable or a macro > manually inserted into the Makefile? Where did you set VPATH? There is no > online documentation for this (gosh, I'm surprised). > Please forget that, I see that vpath is a GNU make directive rather than a Tornado IDE facility. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I boot x86 without BIOS Date: Tue, 25 Jan 2000 12:35:07 -0600 From: "Dennis Gabler" Message-ID: References: <388D9CF7.8CB7F8B0@pacbell.net> <388DD569.DBD806C7@bitband.com> This does sound like a good starting point. One caveat the NS486SXF comes up in protected mode, no need for all the hoops to get there from real mode. Consequently, you may have to add in code to do that. Dennis Leonid Rosenboim wrote in message <388DD569.DBD806C7@bitband.com>... >Years ago I did a BSP for a custom board with a i486 CPU, >no BIOS and entirely non-PC-compatible hardware design. >Was practically the same as any BSP for any architecture other ten x86. >You can use the ns486 BSP delivered with Tornado as an example >of a true BSP for an x86 architecture board. > >Leonid > >Chuck Hardy wrote: >> >> The Wind River documentation for booting Intel x86 systems seems to >> imply that there has to be a BIOS to boot vxWorks. They only refer to >> PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed >> an x86-based system that boots vxWorks directly from ROM? If so, how do >> you start on power-up? As I understand it, the x86 starts executing at >> F000:FFF0. That doesn't leave much room for code, so you have to jump >> somewhere else to start the boot process. Do the Wind River tools >> support putting this jmp instruction into ROM, or do I have to somehow >> patch it in by hand? >> >> Any help would be greatly appreciated. >> >> Thanks, >> Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: Re: comp.os.vxworks newsdigest Date: Tue, 25 Jan 2000 14:54:35 -0600 From: "Daniel Scheurell" Organization: Chorus Netwoks Message-ID: <86l2ij$qb5$1@news.chorus.net> References: <86kso5$frg3@overload.lbl.gov> Sender: reguser@etc-proxy.etcconnect.com Mitch, You hit the nail on the head. I went and editted the radix.h file, moving the mbuf.h include line out of the pragma align block, and voila, consistent structure sizes. I move it back in and the problem re-occurs. Thank you very much. I also have Tornado 2 loaded here. We have a custom BSP that was quite difficult to port to T2, so we are staying with T1.0.1 for now, until we have more time to work the port. In any case, I looked at the radix.h file in the T2 distribution and, indeed, the only difference from the T1 file was moving the #include "net/mbuf.h" line above pragma align. Thanks again, Dan Mitch Hendrickson wrote in message news:86kso5$frg3@overload.lbl.gov... > [ major snippage due to finicky news server ] > > In any case, the problem is very likely the fact that radix.h came with > SENS, and is included eventually when you include inetLib.h, and it > (radix.h) has a #pragma align *before* it includes mbuf.h. This means that > when mbuf.h also does a #pragma align, the original alignment value (16) is > lost forever, and anything after that point will be aligned differently. > Where I first saw it was in the shared memory interface. > > I reported this to WRS about 14 months ago (TSR 120283) and although we got > a fix for our particular case, it appeared to disappear into a black hole. > About a year later, they apparently listened to me or someone else, though, > as a patch was released under SPR 24065. It's on their website and goes > under the title "SENS1.1 i960 Shared Memory Network Fails." > [ more snippage ] --------------------------- Newsgroups: comp.os.vxworks Subject: Re: when not compiled works fine. Date: 21 Jan 2000 21:31:10 GMT From: "Klaus Popp" Message-ID: <01bf6457$079b16c0$140101c0@popp> References: <85s3du$604$1@nnrp1.deja.com> Declare your reg pointer as volatile: volatile unsigned int *reg; Otherwise the optimizer will hide the actual hardware access - -- Klaus Popp men GmbH Klaus.Popp@men.de www.men.de girishvg@yahoo.com schrieb im Beitrag <85s3du$604$1@nnrp1.deja.com>... > hello, > following code tries to write value to some register at some address. > if a particular part of code is not compiled the function executes > (prints both entry...exit messages) properly. but when the part of code > is added, the function stops in the start itself. it does not print > "entry" message also. what/where could be the problem? > any help in this regard is greatly appreciated. > thanks in advance and regards, > girish. > > *** source code *** > > #include "vxworks.h" > #include "intLib.h" > #include "stdio.h" > > /*#define NOWCOMPILE*/ > #define SOME_ADDRESS ((unsigned int*)(0x82000714)) > > void test(int mask) > { > int lock; > unsigned int val, *reg= SOME_ADDRESS; > > printf("enter\n"); > lock = intLock (); > > val = *reg; > val &= ~(1< #ifdef NOWCOMPILE > *reg = val; > #endif > > intUnlock (lock); > printf("return\n"); > } > > > Sent via Deja.com http://www.deja.com/ > Before you buy. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: Fri, 21 Jan 2000 16:49:28 -0500 From: "John Turner" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86ak7l$dp6$1@soap.pipex.net> References: <86a2ge$601$1@soap.pipex.net> <38889536.484F6670@t-online.de> <86a7pv$8gl$1@soap.pipex.net> <86ae0e$b7s$1@soap.pipex.net> John Turner wrote in message news:86ae0e$b7s$1@soap.pipex.net... > John Turner wrote in message > news:86a7pv$8gl$1@soap.pipex.net... > > > > Thanks Heribert but by VPATH, do you mean an environment variable or a > macro > > manually inserted into the Makefile? Where did you set VPATH? There is > no > > online documentation for this (gosh, I'm surprised). > > > Please forget that, I see that vpath is a GNU make directive rather than a > Tornado IDE facility. > The complete solution, as prescribed by WRS, is as follows: 1. Add C/C++ compiler option "-I" in the build specification To display the build specification, double-click on the build name in the "Build" view of the workspace to display the property sheet for the build. Then, select the "C/C++ compiler" tab of the build specification property sheet to view the compiler options. Add another option in the text box. For example: -IE:\header -IE:\header2 Each directory path requires a separated "-I" option. 2. Add makefile macro "VPATH" in the build specification Select the "Macros" tab of the build specification property sheet to view the makefile macros associated with the current project. Add the new macro VPATH by changing the name of an existing macro. Replace the value of the existing macro by the search path(s) of the header files, separated by spaces. For example: E:\header1 E:\header2 Then click the "Add/Set" button and the "OK" button. 3. Rebuild the dependencies Select "Dependencies" from the workspace context menu and click the OK button to rebuild the dependencies. 4. Rebuild the object modules/application You can now click "ReBuild All" button in the tool bar to rebuild the object modules/application. The new compiler option "-I" will be pulled into the dependencies and the compiler will search the specified directories for the header files. --------------------------- Newsgroups: comp.os.vxworks Subject: excVecInit()? Date: 21 Jan 2000 17:52:51 -0600 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: Sender: bpringle@DeadDuck I am porting VxWorks to the SA1110 eval board from Intel. My code is getting to usrInit() and making a call to excVecInit(). I have verified that the init data is correct and the bss is zeroed. I have my memory set to 0-16M is data, with the first 0x10000 reserved for system stack and interrupts. Hmm, I just looked at the ARM section of the VxWorks 5.4 docs and it says that LOCAL_MEM... should be at 0x0? That is where my vectors will go. I also have my TEXT after data. 0x1000-0x1000000 is data. 0x1000000->(0x2000000-0x4000) is code. I have put the MMU table at the top of the code section. Maybe I should revert to the model in the ARM doc. I am loading with CE tools and this seemed the best order at the time. Any ideas? tia, Bill - -- Saskatchewan is much like Texas; only friendlier to the United States - Adlai Stevenson --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskLock & taskUnlock Date: Fri, 21 Jan 2000 21:39:52 GMT From: bruce@figureitout.com (Bruce) Organization: COM.BOX Winet Message-ID: <3888d161.19075370@212.124.35.7> References: On Fri, 21 Jan 2000 16:10:13 -0500, Dheerendra Talur wrote: >- Does anybody know what exactly happens, when taskUnlock() is called without >- ever called taskLock? >- The man pages only say that the taskLock-counter is decremented by >- taskUnlock, but maybe there are some undocumented traps. > >If the "taskLock-counter" is zero or less, taskUnlock() returns OK. >Otherwise, taskUnlock() decrements the "taskLock-counter" by 1 and returns >OK. Don't any about any traps. You can disassemble it at the shell prompt and see what it does. try l(taskUnlock,1) Then just keep hitting l to see the entire function. Bruce Bruce in Katy, TX --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Questions about TFFS Date: Fri, 21 Jan 2000 23:44:14 GMT From: bob_42690@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86ar0a$r46$1@nnrp1.deja.com> References: <3888797C.35D8924D@worldnet.att.net> In article <3888797C.35D8924D@worldnet.att.net>, Tom Weber wrote: > We are considereing using TFFS in an upcoming project and I have a > couple of questions for anyone that has used TFFS or is familiar with > it. We currently use an 8-bit AMD flash part. > > 1. How would one rate the performance and stability of TFFS? I used a Samsung (NAND FLASH). The performance was not fast for reads and writes when the Flash disk was mounted remotely over NFS. Alot of time was spent speeding the system up, (DMA, PCI Bursting, using interrupts to signal flash chip completion of Erase and Write operations, setting the block size, optimizing the number of sectors/cluster etc... There has been a lot of discussion in this newsgroup with disk corruption if the system is not powered down gracefully. I suggest you scan these postings. > > 2. For our requirements dosFs is a bit of overkill. Can rawFs be used > with TFFS instead? > > 3. In the code and doco I keep coming across the term "interleave" , > but I cannot find any description of this term and what the consequences > are using different interleave values. I believe that the interleave represents the number of flash devices in parallel. So if one accessed two 8-bit chips simultaneously using a 16 bit word, it would have an interleave of 2. > > 4. Has anyone written an MTD? There is a WRS provided MTD for our > current 8-bit AMD flash part. We will be moving to a larger and wider > Intel part. I don't believe this part is supported by any of the > provided Intel MTD files, so I am looking at writing/porting an MTD for > this part. Does anyone have any caveats they wish to share? Porting from a similar MTD will help alot. The documentation for an MTD and its interface to TFFS is sparse (or at least it used to be). > > Thanks > Tom > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: send() doesn't fail? Date: Tue, 25 Jan 2000 08:29:26 -0500 From: Charles H. Chapman Organization: Dynacs Engineering Company Message-ID: <86k8fs$iql@niven.ksc.nasa.gov> References: <38915a46.588878202@news.ibm.net.il> Reply-To: chc@nasa2.ksc.nasa.gov Amit Raam wrote: > Hi, > Is send() supposed to fail (return ERROR) when the remote machine has crashed? The default parameters for KEEP ALIVE are set to detect a lost connection after some ridiculously long time like two hours. You can change this time by modifying the values of the variables tcp_keepintvl and tcp_keepidle. See the manual page for setsockopt under the SO_KEEPALIVE option. Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CrossWind question Date: Sat, 22 Jan 2000 02:07:03 GMT From: John or Jenn Organization: Who, us organized??? Message-ID: References: <86a7ma$c77$1@nnrp1.deja.com> In article <86a7ma$c77$1@nnrp1.deja.com> , marky_mark@my-deja.com writes: >Does running GDB configured for Solaris 2.5.1 on Solaris 2.6 cause >problems? The reason I ask is because I am having problems with the >debugger not updating when breakpoints occur (see my previous posts from >1/20). A shot in the dark, but I'm at a loss for why it's doing this. There used to be a T1.0.1 problem with CrossWind and Solaris (though I think it caused things to hang, rather then not updating breakpoints). Does your host system meet all the criteria in the Tornado Release Notes? --------------------------- Newsgroups: comp.os.vxworks Subject: Display vt100 escape codes from VxSim Lite? Date: Tue, 25 Jan 2000 20:12:44 -0500 From: Graham Benton Organization: Verio Message-ID: <388E4A0C.CCA7CBE5@magicnet.net> This is a multi-part message in MIME format. - --------------9B941E5F202310BF7737AB5B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I have had to port some software to VxWorks that presents a nice display on a vt100 terminal, using the DEC escape codes. The codes are not understood by the VxSim Lite console or the windsh tool. Is there a way to connect a vt100 emulator to VxSim Lite? Alternatively, does the simulator console understand any similar codes? Are they documented anywhere? Graham Benton - --------------9B941E5F202310BF7737AB5B Content-Type: text/x-vcard; charset=us-ascii; name="gpbenton.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Graham Benton Content-Disposition: attachment; filename="gpbenton.vcf" begin:vcard n:Benton;Graham x-mozilla-html:TRUE adr:;;;;;; version:2.1 email;internet:gpbenton@magicnet.net x-mozilla-cpt:;-4608 fn:Graham Benton end:vcard - --------------9B941E5F202310BF7737AB5B-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Looking for information on Ethernet with VxWorks Date: Tue, 25 Jan 2000 20:07:12 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <388E72F0.5F64@despam.pangea.ca> References: <86kln8$frf3@overload.lbl.gov> Jim Way wrote: > > I'm using VxWorks for a PowerPc (PPC) with Ethernet. We're seeing some > interesting behavior (see below) > [...] > ------------------------- > The Interesting Behavior: > 1. Monitoring two units via Telnet sessions, both died at the same time > (observed one time). > 2. Monitoring a unit over the weekend, beginning about 4pm on Friday, the > unit ran fine until about 8:30am Monday morning. > 3. Units connected directly to the network fail, average time is less than > four hours. > 4. The software continues to run, but anything involving Ethernet, such as > Telnet sessions, goes away. > You don't mention what driver you are using. Is it an end driver? The only time I have seen this kind of behaviour, we had mbuf shortages coupled with a netmask problem which caused a lot of ICMP resends. Become friends with the netshow utilities. - -het - --- "I don't care about what any thing was designed to do. I care about what it can do." - Gene Krantz Harvey Taylor Internet: het@despam.portal.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Re: when not compiled works fine. Date: Mon, 24 Jan 2000 15:01:56 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <388C4D44.436F9354@bitband.com> References: <85s3du$604$1@nnrp1.deja.com> <01bf6457$079b16c0$140101c0@popp> <388cf16d.0@dnews.sbell.com.cn> Reply-To: leonid@bitband.com There is a general assumption that anyone who programs on VxWorks, knows at least the "C" programming language. "vilatile" is an ANSI C directive, and is documented in ANY C programming language text book or manual. Leonid Zha Junhai wrote: > > Hi > > Could you tell me the "volatile"'s means or where can I > find it, thanks > > Best Regards > > Tenshou Zha > Klaus Popp wrote in message > news:01bf6457$079b16c0$140101c0@popp... > > Declare your reg pointer as volatile: > > > > volatile unsigned int *reg; > > > > Otherwise the optimizer will hide the actual hardware access > > > > > > -- > > Klaus Popp > > > > men GmbH > > Klaus.Popp@men.de > > www.men.de > > > > > > girishvg@yahoo.com schrieb im Beitrag <85s3du$604$1@nnrp1.deja.com>... > > > hello, > > > following code tries to write value to some register at some address. > > > if a particular part of code is not compiled the function executes > > > (prints both entry...exit messages) properly. but when the part of code > > > is added, the function stops in the start itself. it does not print > > > "entry" message also. what/where could be the problem? > > > any help in this regard is greatly appreciated. > > > thanks in advance and regards, > > > girish. > > > > > > *** source code *** > > > > > > #include "vxworks.h" > > > #include "intLib.h" > > > #include "stdio.h" > > > > > > /*#define NOWCOMPILE*/ > > > #define SOME_ADDRESS ((unsigned int*)(0x82000714)) > > > > > > void test(int mask) > > > { > > > int lock; > > > unsigned int val, *reg= SOME_ADDRESS; > > > > > > printf("enter\n"); > > > lock = intLock (); > > > > > > val = *reg; > > > val &= ~(1< > > #ifdef NOWCOMPILE > > > *reg = val; > > > #endif > > > > > > intUnlock (lock); > > > printf("return\n"); > > > } > > > > > > > > > Sent via Deja.com http://www.deja.com/ > > > Before you buy. > > > --------------------------- Newsgroups: comp.os.vxworks Subject: A Question about BSP Date: Mon, 24 Jan 2000 14:28:51 GMT From: zhouxiaochen@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86hnit$eq6$1@nnrp1.deja.com> Hi everybody, My environment is Tornado 1.0.1(Windows NT 4.0),Vxworks 5.3, target is PPC860. The PPC860 has 32-bit address and the space of address is 0x00000000- 0xffffffff. But when I add a device which use the address 0x10000000- 0x80000000, write the ORx and BRx registers, and add the space to struct PHYS_MEM_DESC sysPhysMemDesc[], and burn into flash memory. The PPC860 is reset again and again automatically. But when I use the space 0x10000000-0x3fffffff, the BSP works very well. Is 0x40000000-0xffffffff is reserved for vxWorks? or I do something wrong? Thanks in advance, Zhou Xiaochen Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Builds in T2 Date: Tue, 25 Jan 2000 19:17:26 -0800 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <388E6746.10D6@gv.net> References: <86jm0p$ds1$1@crusher.de.colt.net> Reply-To: mkesti@gv.net Achim Zimmer wrote: >How can we advise the builds to place the outputs in a different directory ? See section 4.5.1 of the Tornado User's Guide for information concerning build specifications' macros. - -- ======================================================================== Michael Kesti | "And like, one and one don't make | two, one and one make one." mkesti@gv.net | - The Who, Bargain --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SNMP : Entries in Interfaces MIB Date: Wed, 19 Jan 2000 22:59:32 -0800 From: yuval_lorian Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <065f8722.35f6d461@usw-ex0105-038.remarq.com> References: <1R4h4.529$Vz6.601@198.235.216.4> Yes, you can do it. Take a look at the library m2IfLib. Yuval. * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: File dependencies Date: Tue, 25 Jan 2000 19:50:47 -0800 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <388E6F17.4AEF@gv.net> References: <86ldi1$7c4$1@nnrp1.deja.com> Reply-To: mkesti@gv.net zwliu99@my-deja.com wrote: > One the other hand, if I remove the .c file which >includes the .h file, then add it back in, the dependencies are >generated correctly! > >Does any body have an answer or have similar problems? I've experienced it, and discovered the workaround by searching DejaNews (http://www.deja.com) and finding article <37B87129.D07888EC@vixel.com> in which mkrueger@vixel.com describes using it to solve a similar problem. Deja really is handy. I've used it to help resolve several problems. - -- ======================================================================== Michael Kesti | "And like, one and one don't make | two, one and one make one." mkesti@gv.net | - The Who, Bargain --------------------------- Newsgroups: comp.os.vxworks Subject: pipes Date: Tue, 25 Jan 2000 12:10:08 +0100 From: "Frank van Eijkelenburg" Organization: HACOM Internet - Amersfoort - The Netherlands. (033-455611) Message-ID: <948798434.514614@big.hacom.nl> Is it possible to delete a pipe after you created it with pipeDevCreate ()? If not, what is happening with the reserved memory? Frank - -- Please don't sent email to me, but post your answer in the newsgroup --------------------------- Newsgroups: comp.os.vxworks Subject: Getting TaskID in a TaskSwitchHook Date: Mon, 24 Jan 2000 10:14:33 -0600 From: John Beauford Organization: Motorola Message-ID: <388C7A69.1E6C33EB@email.mot.com> Reply-To: qjb005@email.mot.com Has anyone used a switch hook and gotten the taskID(of either old task or new task) inside the switch hook? If you could enlighten me I would appreciate it. - -- John Beauford --------------------------- Newsgroups: comp.sys.powerpc.tech,comp.os.vxworks Subject: Re: How to handle FP exceptions on PowerPC/VxWorks Date: 24 Jan 2000 14:03:22 -0800 From: andrew@user2.teleport.com (Andrew Klossner) Organization: Teleport Inc. Message-ID: <86ii7a$qqc$1@user2.teleport.com> References: <86i0pb$m0s$1@nnrp1.deja.com> I wrote an FP exception handler for an 8240 running VxWorks. I found that it's necessary to AND the FPSCR register contents with 0x0007f0ff so that the FP exception doesn't fire all over again when the next FP instruction is executed. You don't have to disable exceptions, just clear the outstanding reports (the status and "sticky" bits.) Also, when you enter the exception handler, SRR0 (the exception-time program counter) will point to the instruction that caused the exception. You need to add 4 to it before returning, unless your version of VxWorks does that for you. -=- Andrew Klossner (andrew@teleport.com) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: excVecInit()? Date: Sat, 22 Jan 2000 10:06:17 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86bvem$jnk$1@nnrp1.deja.com> References: Assuming that you started with the SA-1100/brutus BSP, you would be very well advised to stick with the memory layout that it uses... During the initialisation a section mapping table is created which maps the 4 banks of memory into one contiguous block starting at zero. excVecInit() is trying to install the exception vectors at zero since that is where the CPU expects them to be (unless you are in the CE model, where things are a little different). Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CrossWind question Date: Sat, 22 Jan 2000 10:25:13 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86c0i4$kdb$1@nnrp1.deja.com> References: <86a7ma$c77$1@nnrp1.deja.com> Hi, There are a couple of things you could try here to narrow down the search a little: 1) Have you tried setting bps from the WindSh? Does the shell get notified correctly when the breakpoint is hit? 2) You could try a serial backend to remove the network from the equation (I seem to remember there being network related problems with Solaris 2.6, but I could be mistaken there). Let me know what you discover... In the mean time I can try to find a 2.6 box to try it on. TTFN, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: System time in vxWorks Date: Wed, 26 Jan 2000 08:29:59 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86mbcc$dkf$1@newsmaster01.magnet.at> References: <86h8qq$6sf7@overload.lbl.gov> Hello, PMFJI, but what means "enter date/time" ? He asked for a 486 PC (and thats what i have to) and this PC has a real time clock. Doesnt VxWorks support this 486 PC HW? i want to be able to boot my PC and then asked the PC for the current date/time. not he should ask ME. i am confused. - -- regards, Mario Semo. http://www.kirchnersoft.com Jaya Prakash M wrote in message <86h8qq$6sf7@overload.lbl.gov>... > printf("\n\nEnter the Date and Time (24Hr Format) :mm/dd/yyyy hr::min:sec ->"); > scanf("\n\n%d%d%d%d%d%d",&mm,&dd,&yyyy,&hr,&min,&sec); > > /* initialize the tm structures */ >...... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: System time in vxWorks Date: Wed, 26 Jan 2000 08:29:59 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86mb8q$dfv$1@newsmaster01.magnet.at> References: <86h8qq$6sf7@overload.lbl.gov> Hello, PMFJI, but what means "enter date/time" ? He asked for a 486 PC (and thats what i have to) and this PC has a real time clock. Doesnt VxWorks support this 486 PC HW? i want to be able to boot my PC and then asked the PC for the current date/time. not he should ask ME. i am confused. - -- regards, Mario Semo. http://www.kirchnersoft.com Jaya Prakash M wrote in message <86h8qq$6sf7@overload.lbl.gov>... > printf("\n\nEnter the Date and Time (24Hr Format) :mm/dd/yyyy hr::min:sec ->"); > scanf("\n\n%d%d%d%d%d%d",&mm,&dd,&yyyy,&hr,&min,&sec); > > /* initialize the tm structures */ >...... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Getting TaskID in a TaskSwitchHook Date: Tue, 25 Jan 2000 08:39:25 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86jnfs$v8t$1@nnrp1.deja.com> References: <388C7A69.1E6C33EB@email.mot.com> In article <388C7A69.1E6C33EB@email.mot.com>, qjb005@email.mot.com wrote: > Has anyone used a switch hook and gotten the taskID(of either old task > or new task) inside the switch hook? > If you could enlighten me I would appreciate it. A little look in the manual for taskSwitchHookAdd() will enlighten you! The function that gets called by the kernel has the following prototype: void switchHook ( WIND_TCB * pOldTcb, WIND_TCB * pNewTcb ) Now, a WIND_TCB* is equivalent to a task ID, so your function will be explicitly given both the old and new task IDs. Enjoy, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Mouse input Date: Wed, 26 Jan 2000 09:32:32 GMT From: pwhicker@notthis.home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <388DC73A.B8E3CF86@compd.com> Mouse support is generally via UGL, don't know if it's available on PPC though. Sorry Paul In article <388DC73A.B8E3CF86@compd.com>, pbrewer-removethis-@compd.com wrote: >I have a PowerPC Single board computer running vxWorks. The application >I'm running on the board takes input from a keyboard but I also want it >to take input from a serial mouse. Does anyone have a driver/source >code for doing this with vxWorks. > >Any help would be greatly appreciated. > >Paul Brewer. > --------------------------- Newsgroups: comp.os.vxworks Subject: Boot problem with VxWorks 5.2 and ADS8260 board Date: Wed, 26 Jan 2000 11:59:06 +0200 From: Rafi Hacham Organization: RADGUARD Message-ID: <388EC56A.DBEBF7CE@radguard.com> Hi All !!! I am working with VxWorks 5.4 (Tornado 2) with ADS8260 board with WindRiver BSP for the ADS8260. I have a problem understanding the boot sequence of the MPC8260. From the configuration word (which is 0E228205) the processor vector table starts at address 0 hence the reset vector handler starts at 0x00000100 and might be 256 bytes long. I understand (I am not sure about it) that romInit starts executing at address 0x00000100. from the configuration word the the BR0 [BA] is 0x0000_0000. romInit initialize OR0 to 0x0836. later in the code (this is I think executed from the 256 bytes of the reset vector starting at address 0x100) there is calculation of the address space of the flash and next there is a branching to this code. My problem is that this branching made to the upper region of the memory address space to address 0xff800000+offset.I do not understand this because BR0 and OR0 which defines CS0 is configured to the lower region of memory starting at 0x00000000. My question is how the flash gets its CS when the following romInit code branches to ff800000+off. ******************************************************************** lis r4, HIADJ(start) /* load r4 with the address */ addi r4, r4, LO(start) /* of start */ lis r5, HIADJ(romInit) /* load r5 with the address */ addi r5, r5, LO(romInit) /* of romInit() */ lis r6, HIADJ(ROM_TEXT_ADRS) /* load r6 with the address */ addi r6, r6, LO(ROM_TEXT_ADRS) /* of ROM_TEXT_ADRS */ sub r4, r4, r5 /* */ add r4, r4, r6 mtspr LR, r4 /* save destination address*/ /* into LR register */ blr /* jump to flash mem address */ start: - -- Best Regards !!! - ---------------------------------------------------------------------------------------- Rafi Hacham Phone: +972-3-7658918 RADGUARD Ltd. Fax: +972-3-6480859 Atidim Tower mailto:rafi.hacham@radguard.com Tel-Aviv 61581 Israel http://www.radguard.com/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Network Booting on a Pentium PC104+ board Date: Thu, 20 Jan 2000 08:38:51 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <866f6q$c4c$2@newsmaster01.magnet.at> References: <85mant$17sq$1@news.doit.wisc.edu> i had exactly the same problems some month before and i reported a bug to WindRiver and they called me daily and tried to make it working but the TechSSupport of WindRiver fails. I solved the problem for my own. a) the factory default disk is unusable. you have to create youre own boot floppy. and your own BSP. do this at best at command line. start a Tornado session. (there is a torvars.bat inside tornado). Directory of F:\Tornado\host\x86-win32\bin 08.01.00 21:11 148 torVars.bat next change to your target directory. in my case a 486: F:\Tornado\target\config\pc486> and edit the file config.h there is a section which defines which network driver you have, and a bit later there are config infos for the network: i use a NE2000 card, and so i have the following options: /* Network driver options */ /*#undef INCLUDE_END /* Use Enhanced Network Drivers */ /*!!!MS!!!*/ #define INCLUDE_END /* Use Enhanced Network Drivers */ #undef INCLUDE_ULTRA /* include SMC Elite16 Ultra interface */ #define INCLUDE_ENE /* include Eagle/Novell NE2000 interface */ /*!!!!MS!!!*/ #undef INCLUDE_ELT /* include 3COM EtherLink III interface */ #undef INCLUDE_ESMC /* include SMC 91c9x Ethernet interface */ #undef INCLUDE_FEI /* include Intel Ether Express PRO100B PCI */ #define INCLUDE_SLIP */ /* include serial line interface */ /*#define SLIP_TTY 1*/ /* serial line IP channel COM2 */ /* !!! MS WAS SOLL DAs DA ????***/ #undef INCLUDE_ELC /* include SMC Elite16 interface */ #undef INCLUDE_EEX /* include INTEL EtherExpress interface */ #undef INCLUDE_EEX32 /* include INTEL EtherExpress flash 32 */ #undef INCLUDE_EX /* include Excelan Ethernet interface */ #undef INCLUDE_ENP /* include CMC Ethernet interface*/ #undef INCLUDE_SM_NET /* include backplane net interface */ #undef INCLUDE_SM_SEQ_ADDR /* shared memory network auto address setup */ #undef INCLUDE_EL_3C90X_END /* 3com fast etherLink XL PCI */ #undef INCLUDE_LN_97X_END /* AMD 79C972 END DRIVER * and later: #define IO_ADRS_ENE 0x300 /*!!!MS!!!!*/ #define INT_LVL_ENE 0x05 /*!!!MS!!!!*/ /* Hardware jumper is used to set */ /* RJ45(Twisted Pair) AUI(Thick) BNC(Thin) */ finally correct the boot parameters in this file (you find it near the begin of the file: /* !!!MS!!!*/ #elif (CPU == I80486) #define DEFAULT_BOOT_LINE "ene(0,0)MARIOSEMO:o:/vxWorks/x486/vxWorks h=192.168.3.220 e=192.168.3.221 u=vxworks2 pw=vxworks2 tn=vxWorks o=ene s=o:/vxWorks/x486/startup" /* */ the syntax of this line is explained when you boot and press help. here i define ene as network drive (what i have below), the host name is MARIOSEMO. i want to load my VxWorks image from FTP Host (you must start the FTP server on the host, and must allow everything (address directories outside,....) (o:\...), h=hostio, t=targetip, users/pw for ftp, name of target, and VERY VERY important, in the OTHER section i have to add "ene" again. (this was my problem). finally i want to use a startup script, which is executed after vxworks is loaded, and it contains simple statement (like ld < a.out). ok, now build your own BSP. make clear make bootrom_uncmp new build your first bootrom floopy. you have to take a fresh formated floppy. the next WindRiver program has about 10 bugs (and i have SPR open against all of them). (and it is just a 5 line DOS batch file!). (e.g. when there is not enough space on the floopy, the program write "copy and verify of files to a: successfull and ok" and when you try to boot from that floppy, the system crashes ,...) mkboot a: bootrom_uncmp you should be able to boot from that floppy and network support should be loaded. the next step is, that you should built your own vxworks ... PS: report the problems you have to WindRiver. i told them, that the docu regarding starting with a real target is worst case. they should write a "getting starting with real target" docu. when i look into the msg below i would check / change (but remember: i started with Tornado/VxWorks in November last year). .) add others=ene .) give him a FULL path for the file name .) on the host FTP server turn messages on .) allow him access to any location on your host system. (e.g. i have n:\temp\ms as homedir for the user vyxworks2, but he access the file on o:\vxworks\....) additional: everything is case sensitive. when the vxworks file is named vxWorks, he will NOT find a vxworks. hope it helps a bit. - -- regards, Mario Semo. http://www.kirchnersoft.com Bob Feng wrote in message <85mant$17sq$1@news.doit.wisc.edu>... >I am trying to boot a Vxworks image onto my Pentium target from a NT host >via an ethernet connection. I used the factory default boot floppy disk and >edited booting parameters manualy. No matter what I try, error messages >below never stop showing up. Can anybody give me a light? I would highly >appreciate your help. > >boot device : ene >unit number : 0 >processor number : 0 >host name : bfeng2.YP.solsys >file nmae : /tornado/target/config/bobPentium/vxworks.st >inet on ethernet (e) : 90.0.0.50:ffffff00 >host inet (h) : 90.0.0.3 >user(u) : bfeng >ftp password(pw) : xxxxx >flags (f) : 0x0 > >Attached TCP/IP interface to ene0. >Attaching network interface lo0... done. >Loading... >Error loading file: errno = 0xd0003. >dosFsDevInit failed. >userFdConfig failed. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Display drivers Date: Wed, 26 Jan 2000 12:10:18 +0200 From: "Steven Venter" Message-ID: <388ec793$0$35291@helios.is.co.za> Hi Can any one help me with the following display drivers that will work under VxWorks and Zinc? ELO Touch Conrac 5118 The screen resolution must not be less that 800 x 600. Thanks Steven - -- sventer@grintek.com sventer@icon.co.za --------------------------- Newsgroups: comp.os.vxworks Subject: Re: A Question about BSP Date: Wed, 26 Jan 2000 11:11:05 +0100 From: "Richard Marsh" Organization: Siemens Inc. Message-ID: <86mh7n$3g7$1@news.mch.sbs.de> References: <86hnit$eq6$1@nnrp1.deja.com> Reply-To: "Richard Marsh" I can think of 2 possible problems: 1. If you have the hardware watchdog enabled then adding that amount of memory takes longer than the timeout period or 2. You don't have enough memory for the MMU page descriptors (the region is broken down into 4Kbyte pages and your system memory used for storing these tables) HTH Richard --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Boot problem with VxWorks 5.2 and ADS8260 board Date: Wed, 26 Jan 2000 11:20:46 +0100 From: "Richard Marsh" Organization: Siemens Inc. Message-ID: <86mhpr$3uh$1@news.mch.sbs.de> References: <388EC56A.DBEBF7CE@radguard.com> Reply-To: "Richard Marsh" I think you'll find that your flash is reperated 'n' times through memory from 0x00000000 upto 0xffffffff. Therefore 0x00000100 is the same as 0xff800100 if your flash size is 8M bytes. In the romInit code which then runs the final settings for BRx and ORx registers are wirtten. Hence, OR0 should be written with the location of your flash i.e. 0xff800000 and the code continues to run as prescribed. HTH Richard Marsh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I boot x86 without BIOS Date: Wed, 26 Jan 2000 11:25:05 +0100 From: Gerald van Kampen Organization: =?iso-8859-1?Q?Oc=E9=2DTechnologies?= B.V. Message-ID: <388ECB81.187BFA21@oce.nl> References: <388D9CF7.8CB7F8B0@pacbell.net> Have a look at: ftp://ftp.amd.com/pub/epd/e86/SC4x0/vxworks/elan.tar Chuck Hardy wrote: > > The Wind River documentation for booting Intel x86 systems seems to > imply that there has to be a BIOS to boot vxWorks. They only refer to > PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed > an x86-based system that boots vxWorks directly from ROM? If so, how do > you start on power-up? As I understand it, the x86 starts executing at > F000:FFF0. That doesn't leave much room for code, so you have to jump > somewhere else to start the boot process. Do the Wind River tools > support putting this jmp instruction into ROM, or do I have to somehow > patch it in by hand? > > Any help would be greatly appreciated. > > Thanks, > Chuck - -- Gerald van Kampen --------------------------- Newsgroups: comp.unix.programmer,comp.os.vxworks Subject: Re: posix semaphore Date: Wed, 26 Jan 2000 11:24:35 +0100 From: "Fred Zwarts" Organization: Kernfysisch Versneller Instituut Message-ID: <86mi13$rkg$1@info.service.rug.nl> References: <38852958.9126D1F0@yahoo.com> <866h6g$ki0$1@info.service.rug.nl> <388754EC.1BC80E48@yahoo.com> "LiJia Jin" wrote in message = news:388754EC.1BC80E48@yahoo.com... > Fred Zwarts wrote: >=20 > > "LiJia Jin" wrote in message = news:38852958.9126D1F0@yahoo.com... > > > I am trying to simulate VxWorks semLib by using POSIX = semaphore > > > calls. If anyone has done this before, please give me some hints = here. > > > > > > Is there a standard way of doing this? For example, the POSIX > > > semaphore calls don't include a function of waiting on time-out. = The > > > solutions I got are using usleep or nanosleep to wait. Is this the = right > > > way to handle this? > > > > > > Thanks > > > > > > > I used a combination of POSIX semaphores and POSIX conditions. > > POSIX has a wait-for-condition with a time-out. >=20 > would you mind point out what's the POSIX conditions calls? Thanks >=20 That would be too much for a news article. In addition I don't have the time at this moment too look up all the = details. Check the POSIX documentation. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado1.0.1 and Solaris 2.7 Date: Tue, 18 Jan 2000 11:01:40 -0500 From: David.Kuechenmeister@gtri.gatech.edu (David R. Kuechenmeister) Organization: Georgia Tech Research Institute Message-ID: I've installed our Tornado 1.0.1 on a Ultra Sparc 5 with Solaris 2.7 and have noticed very odd behavior from the launcher window. The normal graphics have been replaced by some non-descriptive text, e.g. tObjectButton, or such. Is this a problem with the OS or is it more likely that some components just didn't get installed correctly from the CD. Thanks, Dave - -- David R. Kuechenmeister Atlanta, GA mailto:david.kuechenmeister@gtri.gatech.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxworks printErrno Date: 26 Jan 2000 11:01:05 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <86mk5h$b6d$1@news1.xs4all.nl> References: <86leal$if72@overload.lbl.gov> Mario Gallotta wrote: : I am using library function "printErrno". I have included the following : libraries: : errnoLib.h : usrLib.h : : when I download my code to the target I get unresolved symbol, : printErrno. Is printErrno included in your kernel? You can try to find out using the lkup command in the shell. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jan 26 07:07:52 2000 From: "Slutsker, Rasty" Date: Wed Jan 26 07:07:54 PST 2000 Subject: RE: comp.os.vxworks newsdigest 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_01BF680E.E05A2C20 Content-Type: text/plain Hi Chuck, Some time ago I wrote a custom loader for VxWorks. You can start with PC486 BSP, generate bootrom_uncmp as it would be boot from Floppy or HD. The loader should do the following: 1.Initialize CPU and chip-set if needed. 2.Initialize DRAM controller 3.Copy bootrom_uncmp from ROM to 0x8000 4.Jump to 0x8000 Few notes: *Usually PC does not see more than 256K of boot ROM in real mode, hence bootrom_uncmp + loader should be less than 256K. From my experience bootrom_uncmp with network support is about 170K, DOS FS + DiskOnChip support w/o network consumes about 250K *DO NOT modify romInit.s it consists of hardcoded jumps and references to the Descriptor table (pretty wild!) Rasty --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I boot x86 without BIOS Date: Tue, 25 Jan 2000 18:55:05 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <388DD569.DBD806C7@bitband.com> References: <388D9CF7.8CB7F8B0@pacbell.net> Reply-To: leonid@bitband.com Years ago I did a BSP for a custom board with a i486 CPU, no BIOS and entirely non-PC-compatible hardware design. Was practically the same as any BSP for any architecture other ten x86. You can use the ns486 BSP delivered with Tornado as an example of a true BSP for an x86 architecture board. Leonid Chuck Hardy wrote: > > The Wind River documentation for booting Intel x86 systems seems to > imply that there has to be a BIOS to boot vxWorks. They only refer to > PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed > an x86-based system that boots vxWorks directly from ROM? If so, how do > you start on power-up? As I understand it, the x86 starts executing at > F000:FFF0. That doesn't leave much room for code, so you have to jump > somewhere else to start the boot process. Do the Wind River tools > support putting this jmp instruction into ROM, or do I have to somehow > patch it in by hand? > > Any help would be greatly appreciated. > > Thanks, > Chuck --------------------------- ------_=_NextPart_001_01BF680E.E05A2C20 Content-Type: text/html Content-Transfer-Encoding: quoted-printable RE: comp.os.vxworks newsdigest

Hi Chuck,

Some time ago I wrote = a custom loader for = VxWorks. You can start with PC486 BSP, generate bootrom_uncmp as it = would be boot from Floppy or HD. The loader should do the following:

1.Initialize CPU and = chip-set if needed.

2.Initialize DRAM controller

3.Copy bootrom_uncmp = from ROM to 0x8000

4.Jump to 0x8000

Few notes:

*Usually PC does not see more than 256K of boot ROM in = real mode, hence bootrom_uncmp + loader should be less than 256K. From = my experience bootrom_uncmp with network support is about = 170K, DOS FS + DiskOnChip = support w/o network consumes about 250K

*DO NOT modify romInit.s it consists of hardcoded jumps and references to the = Descriptor table (pretty wild!)

Rasty

    ---------------------------

    Newsgroups: = comp.os.vxworks

    Subject: Re: How do I = boot x86 without BIOS

    Date: Tue, 25 Jan 2000 = 18:55:05 +0200

    From: Leonid Rosenboim = <leonid@bitband.com>

    Organization: BitBand = Technologies Ltd. http://www.bitband.com

    Message-ID: = <388DD569.DBD806C7@bitband.com>

    References: = <388D9CF7.8CB7F8B0@pacbell.net>

    Reply-To: = leonid@bitband.com

    Years ago I did a BSP for = a custom board with a i486 = CPU,

    no BIOS and entirely = non-PC-compatible hardware design.

    Was practically the same = as any BSP for any architecture other ten x86.

    You can use the ns486 BSP = delivered with Tornado as an example

    of a true BSP for an x86 = architecture board.

    Leonid

    Chuck Hardy = wrote:

    >

    > The Wind River = documentation for booting Intel x86 systems seems to

    > imply that there has = to be a BIOS to boot vxWorks.  They only  refer to

    > PC-AT systems that = boot via hard-disk, tffs, etc.  Has anyone developed

    > an x86-based system = that boots vxWorks directly from ROM?  If so, how do

    > you start on = power-up?  As I understand it, the x86 starts executing = at

    > F000:FFF0.  = That doesn't leave much room for code, so you have to jump

    > somewhere else to = start the boot process.  Do = the Wind River tools

    > support putting this = jmp instruction into ROM, or do I have to somehow

    > patch it in by = hand?

    >

    > Any help would be = greatly appreciated.

    >

    > Thanks,

    > Chuck

    ---------------------------

------_=_NextPart_001_01BF680E.E05A2C20-- From vxwexplo-errs@csg.lbl.gov Wed Jan 26 08:53:55 2000 From: "Thomas Keith Buchanan" Date: Wed Jan 26 08:53:57 PST 2000 Subject: IPSEC Has anyone has any good/bad experiences using the IPSEC stacks from RouterWare or Intoto with VxWorks? TIA, ------------------------------------ /\ Thomas Keith Buchanan /**\ Principal Engineer /****\ SPARTA, Inc. /****/.. 205 East IH 30 /****/.... Rockwall, TX 75087 .\****\__... ...\**\ /.... Voice: 972 772 4487 ....\**\/\.... FAX: 214 824 3670 ....\****\.. PCS: 972 672 9657 ....\****/ WWW: www.sparta.com .../***/ ./***/ SPARTA \**/ ====== \/ Pride In Performance From vxwexplo-errs@csg.lbl.gov Wed Jan 26 13:21:50 2000 From: Steve Bauer Date: Wed Jan 26 13:21:52 PST 2000 Subject: Re: comp.os.vxworks newsdigest Keith- VISA is a specification for an open I/O library for the VXI bus, published by the VXIplug&play Systems Alliance. The alliance has a web site at www.vxipnp.org. The "Specifications" page will swamp you with the technical details. Steve Bauer KineticSystems Company, LLC From vxwexplo-errs@csg.lbl.gov Wed Jan 26 21:21:28 2000 From: Jaya Prakash M Date: Wed Jan 26 21:21:30 PST 2000 Subject: RE: Vxworks printErrno I am using library function "printErrno". I have included the following libraries: errnoLib.h usrLib.h when I download my code to the target I get unresolved symbol, printErrno. Here are the list of vxWorks include files: Reply : Include developement tool component's ----> target shell Components in the VxWorks boot image rebuild the vxWorks.exe and download the boot image and try to download U R application. Regards, JP From vxwexplo-errs@csg.lbl.gov Wed Jan 26 21:46:39 2000 From: Jaya Prakash M Date: Wed Jan 26 21:46:42 PST 2000 Subject: Re :Vxworks printErrno ---------- From: Jaya Prakash M Sent: Thursday, January 27, 2000 10:50 AM To: vxworks_users@csg.lbl.gov; 'the vxWorks Users Group Exploder' Subject: RE: Vxworks printErrno I am using library function "printErrno". I have included the following libraries: errnoLib.h usrLib.h when I download my code to the target I get unresolved symbol, printErrno. Here are the list of vxWorks include files: Reply : Include developement tool component's ----> target shell Components in the VxWorks boot image rebuild the vxWorks.exe and download the boot image and try to download U R application. Regards, JP From vxwexplo-errs@csg.lbl.gov Thu Jan 27 13:41:04 2000 From: Rick Davis Date: Thu Jan 27 13:41:06 PST 2000 Subject: RapidLogic experiences We are developing a datacom product based on VxWorks. We're considering using the RapidLogic tools for development of the command line interface and SNMP control. Has anyone had any major problems with this tool? It looks like a good framework for interface development, especially when one is given the precise task description of building something "Cisco-like" ;) - Rick Davis rdavis@zhone.com Zhone Technologies From vxwexplo-errs@csg.lbl.gov Thu Jan 27 19:09:08 2000 From: "Ferrara, Bob" Date: Thu Jan 27 19:09:11 PST 2000 Subject: excHookAdd 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_01BF693C.BDEEB91A Content-Type: text/plain VxWorks/Tornado users - Can someone tell me if excHookAdd() can be called more than once, and have the routines specified on the excHookAdd calls ALL be invoked when an exception occurs? It's behaving like it can only be called one time. It is called in bootConfig.c to add a boot exception handler (standard VxWorks) and I have called it a 2nd time to add my own specific things to do when an exception occurs. My exception routine only gets invoked when I comment out the first call to excHookAdd in bootConfig.c Thanks, Bob Ferrara bferrara@unispheresolutions.com ------_=_NextPart_001_01BF693C.BDEEB91A Content-Type: text/html excHookAdd

VxWorks/Tornado users -


Can someone tell me if excHookAdd() can be called more than once, and have
the routines specified on the excHookAdd calls ALL be invoked when an exception
occurs? 

It's behaving like it can only be called one time.  It is called in bootConfig.c to add
a boot exception handler (standard VxWorks) and I have called it a 2nd time to add
my own specific things to do when an exception occurs.  My exception routine only
gets invoked when I comment out the first call to excHookAdd in bootConfig.c

Thanks,
Bob Ferrara
bferrara@unispheresolutions.com

------_=_NextPart_001_01BF693C.BDEEB91A-- From vxwexplo-errs@csg.lbl.gov Fri Jan 28 00:38:59 2000 From: "QinTing" Date: Fri Jan 28 00:39:02 PST 2000 Subject: question This is a multi-part message in MIME format. ------=_NextPart_000_000E_01BF69AE.97F36540 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: quoted-printable Hi,everyone Recently a question confusing me is that: What's the difference or the association between the two functions = muxBind( ) and ipAttach( )?=20 When we have prepared a network device, which one above can connect = my device with the vxWorks's network protocal. =20 Best regard queent ------=_NextPart_000_000E_01BF69AE.97F36540 Content-Type: text/html; charset="gb2312" Content-Transfer-Encoding: quoted-printable
Hi,everyone
 
   Recently a question confusing me is = that:
   What's the difference or the = association=20 between the two functions muxBind( ) and=20 ipAttach( )?
   When we have prepared a network = device,=20 which one above can connect my device with = the vxWorks's network=20 protocal.
 
Best regard
queent
 ------=_NextPart_000_000E_01BF69AE.97F36540-- From vxwexplo-errs@csg.lbl.gov Fri Jan 28 01:52:49 2000 From: "Tyler, Paul A (DBY)" Date: Fri Jan 28 01:52:52 PST 2000 Subject: Re: System time in vxWorks VxWorks does not not automatically read the battery backed real time clock. After all, in many embedded applications time of year does not matter. I found a WindSurf article (I have lost the reference) which included a description of the registers in the PC's CMOS (Motorola MC146818 or equivalent) I have an application which will read the clock based on that article I can share. VxWorks, Tornado Paul Tyler Measurement Systems Specialist Electronics & Measurement Systems Dept. Rolls-Royce plc. e-mail Paul.A.Tyler@Rolls-Royce.com Tel. +44 (1332) 247808 Fax. +44 (1332) 247928 Frank Minich wrote in message news:<388F4B00.2136@ti.com>... > Mario Semo wrote: > > i am confused. > > The PC has a time of day clock, which maintains and continues > to update its 'time of day' even across PC power cycles. As > VxWorks boots, it should obtain the current time from the time > of day clock and jam it into VxWorks' clock. I don't think > this happens automatically in VxWorks 5.3. > > Further, if you never boot the PC to DOS or Windows, you will > probably want some way to update the PC's time of day clock > from the VxWorks shell. > > If I remember correctly (it's been a while), the TOD clock is > accessed in a manner similar to that of the PC's CMOS. > > Regards, > Frank. From vxwexplo-errs@csg.lbl.gov Fri Jan 28 04:03:16 2000 From: Vxworks Exploder Date: Fri Jan 28 04:03:19 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jan 28 04:03:12 PST 2000 Subject: Re: DOS Fs ?? VxWOrks newbie Subject: sockets and c++ Subject: Re: sockets and c++ Subject: Re: CORBA on VxWorks Subject: Re: Structure Alignment Problem - PLEASE HELP ... Subject: ethernet-driver for pc486 Subject: Re: Structure Alignment Problem - PLEASE HELP ... Subject: Re: pipes Subject: STL for Tornado 1.0.1 Subject: Re: \r\n in .wsp and .wpj files Subject: Re: Structure Alignment Problem - PLEASE HELP ... Subject: ARE YOU A YOUNG TEEN MODEL WHO WANTS TO MAKE IT BIG? 89918 Subject: Re: System time in vxWorks Subject: get money for browsin web sites! 34802 Subject: Re: VxWorks: VxLoad.com problem Subject: Is there still an comp.os.psos ? Subject: send() doesn't fail? Subject: Re: Is there still an comp.os.psos ? Subject: VERSATILE MESSAGE TRANSACTION PROTOCOL(VMTP) Subject: Re: Is there still an comp.os.psos ? Subject: templates Subject: Re: *** No rule to make target Subject: Re: adding files to Tornado 2 project? Subject: crash when linking un-called function Subject: Re: Different SCSI access time between MVME2700 and MVME2400 Subject: Re: MPC860-Ethernet-vxWorks Subject: Start your Own Business 77033 Subject: Re: posix semaphore Subject: Special Workshop on Intelligence at the Network Edge Subject: Re: DCOM on VxWorks? Subject: Re: STL for Tornado 1.0.1 Subject: Re: dialing in VxWorks Subject: Re: when not compiled works fine. Subject: QED RM7000 PROM Monitor Source Package Subject: Re: crash when linking un-called function Subject: Booting from a SCSI HD Subject: How do I boot x86 without BIOS Subject: Re: sockets and c++ Subject: MibWay for VxWorks Subject: Re: Booting from a SCSI HD Subject: Re: *** No rule to make target Subject: Re: External Dependencies for T2 under UNIX Subject: Re: System time in vxWorks Subject: Re: *** No rule to make target - more questions Subject: Re: thumb mode on arm7 TDMI processor ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: DOS Fs ?? VxWOrks newbie Date: Sat, 15 Jan 2000 22:52:20 -0800 From: "Cameron Fowler" Organization: AT&T WorldNet Services Message-ID: <85rplh$giu$1@bgtnsc02.worldnet.att.net> References: <38778CE6.2D27@tqci.net> You don't mention whether you are using one partition or multiple. Basically vxWorks dosFs 1.0 does not support partitions. (DosFs 2.0 does). It is relatively simple to work around this though in dosFs 1.0 Assuming you have formatted the disk on your pc using FAT16. When setting up your file system/scsi device in vxWorks you will make three calls scsiPhysDevCreate() scsiBlkDevCreate() dosFsDevInit() When calling scsiBlkDevCreate, the last parameter is the blockOffset. You need to specify this to basically step over the partition table that was put on the disk by MSDOS. Often the value passed in their is 0x3f. so you will see something like the following (working from memory at home in case of any discrepancies) assuming Bus 0 LUN 0 pPhys = scsiPhysDevCreate (pSysScsiCtrl, 0, 0, 0, -1, 0, 0, 0); pBlk = scsiBlkDevCreate (pPhys, 0, 0x3f); dosFsDevInit ("/sd0", pBlk, 0); tim keck wrote in message <38778CE6.2D27@tqci.net>... >I just got involved with a project using VxWorks 5.4, Tornado 2 running >on a Synergy PowerPC board. My only previos VxWorks experience was one >project running 5.2 on a FORCE SParc 2CE years ago. > >The design engineer wants to use DOSFs to record data on a large SCSI >hard disk so it can be read back on a PC. >I finally came to the conclusion that VxWorks doesn't support the >parition table on the hard disk, so I wrote the following code to handle >partition table entries, figuring I'd fdisk and format on a PC, and then >mount the drive on VxWorks and record data: > > >/* start paste of code file ------------------------------- */ > >#include "stdio.h" >#include "dosFsLib.h" >#include "scsiLib.h" > >/* DosPart.c - DOS FS partition handler */ > >/* modification history >-------------------- >2000-01-06 First code t keck J.F. Taylor Inc. >Copyleft J. F. Taylor Inc. under GNU Public License >*/ > >/* >DESCRIPTION >This file is used to query all available SCSI devices for partition >tables, >and initialize DOS file systems accordingly. >They are named dosIP, wher I = ID number, and P = partition location in >table >*/ > >#define MINDRIVE 0 /* minimum SCSI device number to poll */ >#define MAXDRIVE 7 /* maximum+1 SCSI device number to poll */ >#define MAXPART 4 /* maximum number partitions per device */ > >#define PARTSIZE 1 /* size of partition table in sectors */ >#define PARTOFFSET 0 /* offset of first sector of partition table from >start of disk, in sectors */ >#define VALIDPART 0x55AA /* tag to indicate valid parition, in PC byte >order */ >/************************************************************************** ***** >* >* DosPart - configure SCSI peripherals w/ DOS partitions >* >* This code configures the SCSI disks whch have valid DOS parition >tables, and >* assigns DOS Fs structure to each valid partition >* >*/ > >void DosPart (void) { > int drivecount; > int partcount; > > char diskname[8]; > > SCSI_PHYS_DEV * pSpd; /* SCSI_PHYS_DEV ptr */ > BLK_DEV * pSbd; /* SCSI_BLK_DEV ptr */ > BLK_DEV * pDbd; /* DOS BLK_DEV ptr*/ > > struct OnePart { > char boot __attribute__ ((packed)); > char startCHS[3] __attribute__ ((packed)); > char type __attribute__ ((packed)); > char endCHS[3] __attribute__ ((packed)); > unsigned int offset __attribute__ ((packed)); > unsigned int size __attribute__ ((packed)); > } ; > > struct PartionTable { > char bootcode[446]; /* the boot code, which we will skip */ > struct OnePart partentry[MAXPART]; /* the partition table entries */ > unsigned short int tag ; /* the tag as valid parition table */ > } parttable; > > > /* > * HARD DRIVE CONFIGURATION > * > * In order to configure the hard disks and initialise dosFs we must >find > * each valid SCSI device, determine it has a valid partition table, > * create block devices for the partitions in the table, > * and create the file systems for the devices > */ > > /* now cycle through, looking for valid DOS partitions */ > for ( drivecount = MINDRIVE; drivecount < MAXDRIVE; drivecount++) { > if ( (pSpd = scsiPhysDevIdGet (pSysScsiCtrl, drivecount, 0)) != NULL >) { > if ( (pSbd = scsiBlkDevCreate (pSpd, PARTSIZE, PARTOFFSET)) != >NULL) { > scsiRdSecs((SCSI_BLK_DEV *)pSbd, PARTOFFSET, PARTSIZE, (char >*)&parttable ); > if ( parttable.tag == VALIDPART) { > for ( partcount=0; partcount < MAXPART; partcount++) { > binvert((char *)&parttable.partentry[partcount].offset, 4); > binvert((char *)&parttable.partentry[partcount].size, 4); > if (parttable.partentry[partcount].size != 0) { > if ( (pDbd = scsiBlkDevCreate (pSpd, > parttable.partentry[partcount].size, > parttable.partentry[partcount].offset) ) != NULL) >{ > sprintf (diskname, "dos%x%x", drivecount, >partcount); > if( dosFsDevInit (diskname, pDbd, NULL) == NULL ) > perror ("dosFsDevInit"); > } > } > } > } > } > } > } >} > >/* end paste of code file ----------------- *? > >BTW, I can't guarantee that this code will actually compile and run, >since I took my code and chopped out all the debugging messages to post >it here, to save space. But it's close. If anyone is really >interested, I'll repost the final version when it works. > >Anyway, the problem is that I get an "Error 380001" or some such at the >perror. It I use a dosMkFs instead of the dos FsDevInit call, I get a >device of the correct name in my system device table, but then when I >put the disk pack on the PC, it no longer recognizes it, though the >partition seems OK under fdisk; so I assume the formatting is FUBAR. >The PC is running Win98 (shudder) by the way. > >So my question is, when they say 'dos file system', just what do they >mean? FAT12? FAT16? How can I read & write hard disk partitions that >are interchangeable w/ my PC, that is REALLY a DOS file system? > >I'd appreciate any help anyone has out there. >Thanks. > >Tim Keck >Senior Engineer >J.F. Taylor Inc. --------------------------- Newsgroups: comp.os.vxworks Subject: sockets and c++ Date: Wed, 26 Jan 2000 14:21:09 +0100 From: "Frank van Eijkelenburg" Organization: HACOM Internet - Amersfoort - The Netherlands. (033-455611) Message-ID: <948892706.324636@big.hacom.nl> I'm trying to make a class in C++ with socket functionallity. When I use the function socket (with 3 integers as input) I get the following error message: socks.cpp:21: no matching function for call to `socket::socket (int, int, int)' D:\\TORNADO/target/h/net/socketvar.h:135: candidates are: socket::socket(const socket &) D:\\TORNADO/target/h/net/socketvar.h:135: socket::socket() If I make a "normal" C-file I can use this function without any problems. Can anybody help me? Thanks, Frank - -- Please don't sent email to me, but post your answer in the newsgroup --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sockets and c++ Date: Wed, 26 Jan 2000 14:26:12 +0100 From: "Proxy" Message-ID: <388ef5b6$0$20390@SSP1NO17.highway.telekom.at> References: <948892706.324636@big.hacom.nl> I assume, your class is called socket, and just calling socket() tries to call a member function of this class. Try ::socket() to call the socket() API function HTH, Thomas Frank van Eijkelenburg wrote in message <948892706.324636@big.hacom.nl>... >I'm trying to make a class in C++ with socket functionallity. When I use the >function socket (with 3 integers as input) I get the following error >message: > >socks.cpp:21: no matching function for call to `socket::socket (int, int, >int)' >D:\\TORNADO/target/h/net/socketvar.h:135: candidates are: >socket::socket(const socket &) >D:\\TORNADO/target/h/net/socketvar.h:135: socket::socket() > >If I make a "normal" C-file I can use this function without any problems. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CORBA on VxWorks Date: Tue, 25 Jan 2000 05:58:06 GMT From: bill.beckwith@realtime-corba.com (Bill Beckwith) Organization: Objective Interface Message-ID: <8EC5F047Fbeckwbnewscom@sigma.ois.com> References: <85u217$f63$1@nnrp1.deja.com> st1734 wrote in <85u217$f63$1@nnrp1.deja.com>: > > >I'm looking for CORBA vendors on VXWORKS. > >Any pointers ? ORBexpress by Objective Interface runs on VxWorks. There are versions of ORBexpress for three different C++ compilers (gcc, egcs, and Green Hills C++) and two Ada compilers (Rational Apex and Green Hills Ada). For more info on ORBexpress see: http://www.ois.com/ - -- Bill --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structure Alignment Problem - PLEASE HELP ... Date: Tue, 25 Jan 2000 08:43:42 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <388D461E.5C1ABA1E@bitband.com> References: <86hqh4$qr1$1@news.netvision.net.il> Reply-To: leonid@bitband.com Here is my advise: Within the end.h file, where the END_OBJ type is defined, make sure the type definition is enclosed in extern C { ... }, declaring it as a plain C structure as opposed to a C++ object. Hope this helps, Leonid Edris Abzakh wrote: > > Hi, > > Please follow the current scenario and tell me what is wrong !! > > I opened a new project under Tornado 2 and added 2 simple files with the > same content > The first file suffix is .cc (C++ file ) and the other is .c (C file) > > Each file contains one function as follows : > > cTest.c: > ---------- > #include "end.h" > > void cTestMain () > { > printf("size of END_OBJ = %d \n",sizeof(END_OBJ)); > } > > result : > -------- > size of END_OBJ = 720 > > ccTest.cc > ------------ > #include "end.h" > > void cTestMain () > { > printf("size of END_OBJ = %d \n",sizeof(END_OBJ)); > } > > result: > ------- > size of END_OBJ = 752 > > Does anyone has any idea why this should happen ? my target is 960jx and I > tried to use #pragma align 1 > statement at the beginning of each file .. still the same result ??? > > Please Help .... > > Edris --------------------------- Newsgroups: comp.os.vxworks Subject: ethernet-driver for pc486 Date: Wed, 26 Jan 2000 16:06:41 +0100 From: Gerhard Juen Organization: 'FH Gelsenkirchen, Abt. Bocholt' Message-ID: <388F0D81.2C9EB5F6@bocholt.fh-gelsenkirchen.de> hello i am searching for an ethernet driver for the AMD 79C961 ethernet-chip. (AMD sometimes call it "PCnet-ISA") my hardware platform is pc486 with ISA bus. can anybody help? thank you gerhard juen - ----------------------------------------------------------------------- Prof. Dr. Gerhard Juen FH Gelsenkirchen, Abt. Bocholt Münsterstr. 265 D-46397 Bocholt email: juen@bocholt.fh-gelsenkirchen.de web: http://www-et.bocholt.fh-gelsenkirchen.de/~juen Fon: 02871-2155 830 Fax: 02871-2155 800 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structure Alignment Problem - PLEASE HELP ... Date: Wed, 26 Jan 2000 15:41:52 GMT From: "Barry Kelly" Organization: NBTel Internet Message-ID: <4BEj4.1421$pa6.45180@sodalite.nbnet.nb.ca> References: <86hqh4$qr1$1@news.netvision.net.il> GCC switch -fpack-struct on the CPP project --------------------------- Newsgroups: comp.os.vxworks Subject: Re: pipes Date: Wed, 26 Jan 2000 11:10:59 -0500 From: "Terrence Meiczinger" Organization: 3Com Corp. Marlborough MA, USA Message-ID: <86n6g9$mht$1@news.ne.3com.com> References: <948798434.514614@big.hacom.nl> Try something like this: pipeDevCreate ( "/pipeDevice", 1, 1000 ); if (iosDevFind("/pipeDevice",0)) iosDevDelete(iosDevFind("/pipeDevice",0)); - -Terrence Frank van Eijkelenburg wrote in message news:948798434.514614@big.hacom.nl... > Is it possible to delete a pipe after you created it with pipeDevCreate ()? > If not, what is happening with the reserved memory? > > Frank > -- > Please don't sent email to me, but post your answer in the newsgroup > > --------------------------- Newsgroups: comp.os.vxworks,gnu.gcc.help Subject: STL for Tornado 1.0.1 Date: Wed, 26 Jan 2000 17:08:02 GMT From: Kirk Ellett Organization: The Boeing Co. Message-ID: <388F29F2.22C28E18@boeing.com> Reply-To: kirk.d.ellett@boeing.com Sender: nntp@news.boeing.com (Boeing NNTP News Access) Hello, Can anyone point me to an STL that will work well with the GNU 2.7.2 compiler provided in Tornado 1.0.1? I'm not doing many tricky things, just lists and vectors, iterating through them. Kirk --------------------------- Newsgroups: comp.os.vxworks Subject: Re: \r\n in .wsp and .wpj files Date: Wed, 26 Jan 2000 10:17:42 -0800 From: Stas Maximov Organization: Posted via Supernews, http://www.supernews.com Message-ID: <388F3A46.F0DC89AE@pigeonpoint.com> References: <3887DEE8.DDC50D1B@pigeonpoint.com> Hi Martin, You are damn right! That was quite a surprise for me, since when I store a word file in CVS, diff does not work. Great news! CVS stores only deltas for binary files (-kb) and also allows me to perform diff on them. I will simply use -kb switch with Tornado .wsp and .wpj files. I think it is the best way to go when CVS is used. Thanks a lot! Stas. Martin Oberhuber wrote: > > Dear Sir, > > I think that if you define your .wpj files as "binary" for CVS, > it should just disable keyword substitution and disable the > CRLF translation, but AFAIK it should still save minimal > differences (please correct me if I'm wrong). > > Regards, > Martin Oberhuber > > -- > DI Martin Oberhuber, Technical Support ||| > mailto:sniff-support@takefive.co.at (o o) > -------------------------------------------+----oOO--( )--OOo----+ > http://www.TakeFive.com/ | The Source Code | > Phone: +43 (662) 457915 ext 85 (UTC +1h) | Engineering Company | > Fax: +43 (662) 457915-6 +======oo=====oo======+ > > ----- Original Message ----- > From: Stas Maximov > Newsgroups: comp.os.vxworks > Sent: Friday, January 21, 2000 5:22 AM > Subject: \r\n in .wsp and .wpj files > > > Hi! > > > > I am using WinCVS and I want to put my TII project under CVS control. > > > > Tornado workspace (.wsp) and project (.wpj) files are plain text files, > > but use single LF character as a line separator instead of CRLF. If we > > replace LF separators with CRLF (it usually happens with WinCVS), > > Tornado complains that it has "Problem opening > > C:/Tornado/target/proj/vxplay/Workspace.wsp: Workspace open failed: This > > project workspace is an older format". > > > > - I want to treat .wsp and .wpj files as text files. At least for > > storing just minimal changes and being able to perform diffs on them; > > > > - I want to check out text files as appropriate for the platform I am > > using, i.e. CRLF for Windows; > > > > Question: Is there a Tornado option that makes Tornado happy about CRLN > > in .wsp and .wpj? > > > > I have searched in WRS known problems support area and could not find > > anything related. > > > > Thanks, > > Stas. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structure Alignment Problem - PLEASE HELP ... Date: Wed, 26 Jan 2000 09:33:07 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86ng8e$jpn$2@newsmaster01.magnet.at> References: <86hqh4$qr1$1@news.netvision.net.il> can you post your ÉND_OBJ structure? - -- regards, Mario Semo. http://www.kirchnersoft.com Edris Abzakh wrote in message <86hqh4$qr1$1@news.netvision.net.il>... >Hi, >size of END_OBJ = 752 > > >Does anyone has any idea why this should happen ? my target is 960jx and I >tried to use #pragma align 1 >statement at the beginning of each file .. still the same result ??? > --------------------------- Newsgroups: comp.os.vms,comp.os.vxworks,comp.parallel.pvm Subject: ARE YOU A YOUNG TEEN MODEL WHO WANTS TO MAKE IT BIG? 89918 Date: Sunday, 23 Jan 2000 14:47:53 -0600 From: alt.make.money.fast@nntp.8m.com Organization: nanas@nntp.8m.com Message-ID: <23010014.4753@nntp.8m.com> Reply-To: usenet@adsl-63-197-143-124.dsl.scrm01.pacbell.net My name is Darren Mulligan and I am an aspiring photographer from California. I will get right to the meat and gravy of this post, so as not to waste your time! +++++WHAT AM I SEEKING?+++++ I am looking for young models (prefer early teen 14-16 year old female) for nude and semi-nude photography. This is a great way for me to update my portfolio, and also a great way for you to get a free portfolio, too! This could lead to employment in the Adult XXX entertainment business. What a great way to make money! Please respond to one of the following e-mail addresses. I run my very own mail server on an ADSL connection so you can be assured to complete discretion and privacy. If one e-mail address does not work, simply try the next! Sorry for the long list! My ADSL connection goes down from time to time, so I posted all the e-mail addresses I can be reached at. If one does not work, just move on to the next! nanae@adsl-63-197-143-124.dsl.scrm01.pacbell.net usenet@adsl-63-197-143-124.dsl.scrm01.pacbell.net 24hoursupport.helpdesk@adsl-63-197-143-124.dsl.scrm01.pacbell.net news.admin.net-abuse.email@adsl-63-197-143-124.dsl.scrm01.pacbell.net spamlinkpage@adsl-63-197-143-124.dsl.scrm01.pacbell.net alt.make.money.fast@nntp.8m.com nanas@nntp.8m.com hostmaster@FREESERVERS.COM THIS IS NOT A SCAM! IT IS NOT CHILD PORN!! There is nothing wrong or illegal about the artistic representation of a young body. If you have a problem with this post, please e-mail me and I will not post to your newsgroup again! Sincerely, Darren Mulligan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: System time in vxWorks Date: Wed, 26 Jan 2000 13:29:05 -0600 From: Frank Minich Organization: NCP Message-ID: <388F4B00.2136@ti.com> References: <86h8qq$6sf7@overload.lbl.gov> <86ng88$jpn$1@newsmaster01.magnet.at> Reply-To: fminich@ti.com Mario Semo wrote: > i am confused. The PC has a time of day clock, which maintains and continues to update its 'time of day' even across PC power cycles. As VxWorks boots, it should obtain the current time from the time of day clock and jam it into VxWorks' clock. I don't think this happens automatically in VxWorks 5.3. Further, if you never boot the PC to DOS or Windows, you will probably want some way to update the PC's time of day clock from the VxWorks shell. If I remember correctly (it's been a while), the TOD clock is accessed in a manner similar to that of the PC's CMOS. Regards, Frank. --------------------------- Newsgroups: comp.os.vxworks,comp.periphs.printers,comp.periphs.scanners,comp.periphs.scsi,comp.programming Subject: get money for browsin web sites! 34802 Date: Sat, 22 Jan 2000 01:30:45 -0600 From: frree@money.com Organization: esgjerglkesrjgler Message-ID: <220100013045@money.com> Everyone likes making easy money. This place pays you 50 cents for every hour you browse the web!!! Get even more if you refer people! I just got a check for $36.00!!! Sign up here: http://www.alladvantage.com/go.asp?refid=IMY627 cGYd]q57[^ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks: VxLoad.com problem Date: Thu, 20 Jan 2000 08:15:53 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <866f6j$c4c$1@newsmaster01.magnet.at> References: <85grkf$i3j11@overload.lbl.gov> <85i18d$vqi$1@nnrp1.deja.com> <85ntqk$pka$1@newsmaster01.magnet.at> <85ul6o$sit$1@nnrp1.deja.com> <862nau$a83$1@newsmaster01.magnet.at> <863u0p$n8m$1@nnrp1.deja.com> Hello, ok, now i see my problem. i was aware of all the reasons and problems with differrent solutions. So, as a first step i created the bootable floopy via mkboot (this is not a misuse, is it?). Later i found vxload and so i booted DOS from my Hardrive and tried vxload + the file i found on the floppy, which was bootrom.sys. additional, i expected that bootrom.sys and the bootrom_uncmp are identical. the reason was, that i looked into mkboot and i saw vxcopy %1 %2:bootrom.sys and i thought this is a COPY, but it is not a COPY. it is a copy + a modification. Later i saw the files are not equal. next i assumed, that vxload definitly will load modified file. so i tried vxload with the bootrom.sys. as i now understand, i have to use the original bootrom_uncmp with vxLoad. - -- regards, Mario Semo. http://www.kirchnersoft.com >*vxload.com* expects a a.out file as an input - such as "bootrom_uncmp" or >"vxworks.st" and not "bootrom.sys". > --------------------------- Newsgroups: comp.os.vxworks Subject: Is there still an comp.os.psos ? Date: Wed, 26 Jan 2000 14:36:36 -0600 From: John Bundy Inet Message-ID: <388F5AD4.6B8BA43B@megsinet.net> To all, Is there still a comp.os.psos news group? I have tried two different ISPs and don't see it. John Bundy --------------------------- Newsgroups: comp.os.vxworks Subject: send() doesn't fail? Date: Thu, 20 Jan 2000 15:04:07 GMT From: real@email.in.sig (Amit Raam) Organization: Unspecified Organization Message-ID: <388a1efa.180482019@news.ibm.net.il> Hi, Is send() supposed to fail (return ERROR) when the remote machine has crashed? I'm trying to communicate between two vxWorks computers; I've opened a socket, and sent data successfully. However, when the receiving computer resets, the sending computer can continue to send() without any error indication. What am I doing wrong? the code looks something like this: ... int optOn = 1; socketFd = socket(AF_INET, SOCK_STREAM, 0); setsockopt(socketFd, IPPROTO_TCP, TCP_NODELAY, (char *) &optOn, sizeof(optOn)); setsockopt(socketFd, SOL_SOCKET, SO_KEEPALIVE, (char *) &optOn, sizeof(optOn)); serverAddr.sin_family = AF_INET; serverAddr.sin_port = htons(pConn->connectionPort); serverAddr.sin_addr.s_addr = pConn->connectionAddress; connect(socketFd, (struct sockaddr *) &serverAddr, sockAddrSize); .. Later, I send data: write(socketFd, (char *)&connType, sizeof(connType)); tried also: send(socketFd, (char *)&connType, sizeof(connType), 0); Thanks for any idea, Amit Raam Email address: raama at indigo.co.il - -------------------------------------------------- ".sigless and proud of it" (R) (C) (tm) (look-and-feel) --------------------------- Newsgroups: comp.os.psos,comp.os.vxworks Subject: Re: Is there still an comp.os.psos ? Date: Wed, 26 Jan 2000 23:27:57 +0100 From: "Martin Raabe" Organization: T-Online Message-ID: <86nsf4$m70$1@news02.btx.dtag.de> References: <388F5AD4.6B8BA43B@megsinet.net> Hello Joh, yes there is but it is very quiet here in comp.os.psos. __ Martin Raabe Project Manager SNiFF+/pRISM+ Integration TakeFive Software Im Leuschnerpark 4 D-64347 Griesheim T: (+49) 6155 / 878912 F: (+49) 6155 / 605-257 M: (+49) 170 / 1838405 E: MRaabe@TakeFive.de T A K E F I V E S O F T W A R E "the SOURCE CODE engineering company" John Bundy Inet schrieb in im Newsbeitrag: 388F5AD4.6B8BA43B@megsinet.net... > To all, > Is there still a comp.os.psos news group? I have tried two different > ISPs and don't see it. > John Bundy > --------------------------- Newsgroups: comp.os.vxworks Subject: VERSATILE MESSAGE TRANSACTION PROTOCOL(VMTP) Date: Wed, 26 Jan 2000 17:33:58 -0500 From: Sakir Yucel Organization: University of Delaware, EE/CIS Lab Message-ID: Hello all, Does anyone know if the VERSATILE MESSAGE TRANSACTION PROTOCOL (VMTP) specified in rfc1045 ever has been implemented in a VxWorks release? Best regards, Sakir Yucel CIS Graduate Student University of Delaware http://www.cis.udel.edu/~yucel/ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is there still an comp.os.psos ? Date: 26 Jan 2000 17:49:48 -0600 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <388F5AD4.6B8BA43B@megsinet.net> <86nsf4$m70$1@news02.btx.dtag.de> Sender: bpringle@DeadDuck My ISP isn't listing it as available. They are PSI.net which is rather large. I don't think that this ng has made it into wide distribution. >>>>> "Martin" == Martin Raabe writes: Martin> Hello Joh, yes there is but it is very quiet here in Martin> comp.os.psos. - -- We're off to see the wizard. Because because because.... of all the wonderful things he does. --------------------------- Newsgroups: comp.os.vxworks Subject: templates Date: Thu, 27 Jan 2000 00:48:42 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86o13r$8ak$1@newsmaster01.magnet.at> Hello, whats the simplest way to handle templates in GNU/Vxworks? is -frepo a way to go? how does it work? i looked into GNU toolkit user guide (Toprnado 2.0) page 26, there i can read "-frepo is for templates. see page 156". and on page 156 i see nothing about -frepo. there (p153ff) i can read, that there is no automatioc way in template code generation. mh. i thought -frepo is doing this. i would, if possible, try to avoid these GNU specific #pragmas (pragma interface...) and i would like to avoid instantiation by hand. (#include templ.hpp #include templ.c template class Foo; ... ). currently i have my templ. code in xx.c and the header is xx.hpp and the 2 are in the same directory. - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: 23 Jan 2000 23:13:06 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <86g1u2$s9s$1@news1.xs4all.nl> References: <86a2ge$601$1@soap.pipex.net> John Turner wrote: : I'm sure that this is everyone's favorite problem when they first encounter : Tornado v2. : Even though I have provided the appropriate list of include directories (by : specifying the compiler options -I/myinc1 -I/myinc2) and manually re-run the : dependency generator, the Tornado IDE persists in generating an incorrect : project and an incorrect makefile. Specifically it lists certain header : files as dependencies but it gets their paths wrong (actually it gives no : path) and consequently any attempt to build the project fails with "*** No : rule to make target 'xyz.h'". The use of the VPATH macro will correct this problem. Take a look at my VxWorks page on a detailed description of this problem. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: adding files to Tornado 2 project? Date: 20 Jan 2000 16:00:49 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <867bfh$3s9$1@news1.xs4all.nl> References: <864ke8$9gf4@overload.lbl.gov> Reyhan ERGUN wrote: : I have some number of files which are generated by another tool : (Telelogic SDT), : and I want to link them all to my vxWorks build. : I will compile and generate object files for these extra codes within T2 : (UNIX and WinNT) and add all these files to my Tornado II project, but I : am looking for a nicer way to do this. : There is a MACRO as EXTRA_MODULES : and I added a definition as /*.c for this macro : but it doesn't work. Have you tried adding them to your project? If not, why? If so, what were the problems you had? Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: crash when linking un-called function Date: Thu, 27 Jan 2000 18:38:56 +0200 From: Ofer Goren Organization: nortelnetworks Israel Message-ID: <3890749F.E28F3FE8@baynetworks.com> Hi We encountered a very strange problem. This problem happened for CPU = PPC740 under Tornado-1.0.1 and also for CPU = PPC405 under Tornado-2.0 Problem description: - -------------------- 1. We have a function: int test () { UINT16 testInt; printf ("This is test number = %d", testInt); } 2. This function test() is compiled and linked but never called. 3. When the executable runs it crashes - you get the registers dump. 4. If you put the printf line in a comment and re-compile and re-link the executable will not crash. 5. Again I'm reminding you that the function test() is never called from the executable Notes - ----- 1. If we use logMsg() instead of printf() the problem disappears. But we have printf() in other functions in our executable. 2. If we add some code lines to the function test() the problem sometimes disappears. But we can't find the "rule" that causes the problem to disappear or re-appear. Anybody? - -- Embedded SW Team Leader Email : ogoren@nortelnetworks.com Phone number: +9723-6456023 Fax number : +9723-6479579 ESN : 826-6023 http://www.nortelnetworks.com - ---------------------------------------------------- Give someone a fish and you feed them for a day. Teach them to use the Net and they won't bother you for weeks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Different SCSI access time between MVME2700 and MVME2400 Date: Thu, 27 Jan 2000 12:16:44 -0500 From: Steve Alexander <salexand@fore.com> Organization: FORE Systems, Inc. Message-ID: <38907D7C.FB5BB9DB@fore.com> References: <2cIj4.660$m5.4956044@news-read1.qis.net> Minbo Shim wrote: > Hello, > > I have a SCSI disk in a VMEbus form factor that is connected to a Motorola > MVME2700 board. The disks are ATA hard drives and the company provides a > SCSI-2 interface for them. The theoretical data transferring rate is about > 2.5MB/sec based on the spec sheet from the company, but I am only gettting > 1.1MB/sec which I can still endure. But the worst part is that the rate > severely drops to about 350KB/sec when I try to get access to the SCSI disk > from another board (MVME2400) in the same VMEbus. Two boards (MVME2700 and > MVME2400) are connected via a 100BaseT hub. I really don't know why the > access > times are so different between two boards, and like to know who's causing > the > rate drop. > > Any help regarding this issue would be greatly appreciated. > > M.Shim > GDRS Not quite sure, but assume the slow transfer is from 2700 SCSI to 2700 PCI to VME to 2400 PCI to 2400 DRAM. The bandwidth killer is probably the VMEbus rearbitration for every word transfer. Several changes to the Universe 2.0 (see www.tundra.com for docs) chip should help. Set the master (2700) Univese MAST_CTL register bit VRM to enable release-on-request(ROR aka lazy) VMEbus arbitration. This should help a lot. If there is other VMEbus traffic, then setting the system controller to priority arbitration and setting the 2700 to high priority may help. If your application permits, enable posted writes (PWEN bit of the Universe VSIx_CTL registers) for each map after setting PWON of MAST_CTL to some reasonable value (not 0b1111). The VME BLT transfers are much faster but require software changes. Same is true of your 100Mhz enets !! hope that helps, Stephen Alexander - consultant --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC860-Ethernet-vxWorks Date: Thu, 27 Jan 2000 12:48:37 -0500 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <389084F5.26BA66B@lucent.com> References: <86pk2r$bl8$1@crusher.de.colt.net> <389055D4.7019B297@lucent.com> <86pq0b$e7v$1@crusher.de.colt.net> Achim Zimmer wrote: > > the target is an ads860 > If we change in hwInit this variable then it is fine. You cannot change the value once you have run ifAttach(). I assumed that you wanted to permanently change the enet address for this board, not at every boot, and not while it is running. The sysEnetAddr Get and Set routines are used to make a persistent change to the values stored in non-volatile memory (FLASH SEEPROM BBRAM). Doug > > The problem is to change it later in runtime. > > Bye and thanks Achim > --------------------------- Newsgroups: comp.os.vxworks,comp.sys.acorn.networking,comp.sys.amiga.networking,comp.sys.ibm.pc.hardware.networking,conn.jobs.offered Subject: Start your Own Business 77033 Date: Mon, 24 Jan 2000 04:41:59 GMT From: emailpro@primus.com.au Organization: Business Opportunity's Message-ID: <24010015.4323@primus.com.au> Reply-To: Business Opportunity Looking for honest, dependable associates who are interested in making a good income from the comfort of their own home. With proven methods to generate a good income, this is a business that you will want to be a part of. There are Absolutely No Upfront Fees. And all you need is a computer with Internet capabilities and a bit of spare time. For more information Please contact me on : emailpro@primus.com.au?subject=EmailProcessing 1 --------------------------- Newsgroups: comp.unix.programmer,comp.os.vxworks Subject: Re: posix semaphore Date: Thu, 27 Jan 2000 13:23:15 -0500 From: Steve Coleman Organization: Johns Hopkins Applied Physics Laboratory Message-ID: <38908D13.8A73C371@jhuapl.edu> References: <38852958.9126D1F0@yahoo.com> <866h6g$ki0$1@info.service.rug.nl> <388754EC.1BC80E48@yahoo.com> LiJia Jin wrote: > would you mind point out what's the POSIX conditions calls? Thanks pthread_cond_broadcast condition (3t) - condition variables pthread_cond_destroy condition (3t) - condition variables pthread_cond_init condition (3t) - condition variables pthread_cond_signal condition (3t) - condition variables pthread_cond_timedwait condition (3t) - condition variables pthread_cond_wait condition (3t) - condition variables pthread_condattr_destroy pthread_condattr_init (3t) - condition variable initialization attributes pthread_condattr_getpshared pthread_condattr_init (3t) - condition variable initialization attributes pthread_condattr_init pthread_condattr_init (3t) - condition variable initialization attributes pthread_condattr_setpshared pthread_condattr_init (3t) - condition variable initialization attributes - -- Steve Coleman http://www.jhuapl.edu/ High Performance, fault tolerant, distributed, real-time computing <<-------->> Johns Hopkins Applied Physics Laboratory <<--------->> Balt:443-778-6330 Fax:443-778-5597 Wash:240-228-6330 Fax:240-228-5597 --------------------------- Newsgroups: comp.arch.embedded,comp.realtime,comp.robotics,sci.engr.control,comp.home.automation,comp.os.vxworks,com.os.qnx,comp.dcom.cell-relay,comp.dcom.fax Subject: Special Workshop on Intelligence at the Network Edge Date: Thu, 27 Jan 2000 18:48:18 GMT From: moun@usenix.org (Moun Chau) Organization: USENIX Association Keywords: Embedded Systems, QoS, SmartBox, Security, Networks, Networking, USENIX, workshops, seminar, panel, addressing, routing, infrastructure, applications, software, internet, middleware, caches, USENIX, conference, workshop, invited talks, presentation, 3Com, RSIP, servers, systems Message-ID: SPECIAL WORKSHOP ON INTELLIGENCE AT THE NETWORK EDGE Sponsored by USENIX, The Advanced Computing Systems Association with support from 3Com. March 20, 2000 Holiday Inn Fisherman's Wharf San Francisco, California, USA This event is intended to be a workshop, not a short conference or symposium. Session and panel chairs will engage the audience as active participants in the discussions. Those who are interested in identifying new opportunities in networking technology would benefit most from attending this workshop. Regardless of your level of expertise, you'll come away from this workshop with new information and practical solutions. Our topics include: *SmartBox: An Add-on solution for Guaranteed QoS *Integrating Active Networking and Commercial Grade Routing *QoS provisioning at the network edge *RSIP: Address Sharing with End-to-End Security *Internet Addressing at the Edge *Proxy Caches Move to Middleware Pillars *Trends and Forecasts in Internet Publishing Infrastructure *Open Session on Intelligence at the Network Edge ======================================================== Register online at: http://www.usenix.org/events/ine2000 Hotel Reservation special rates end February 25, 2000 ======================================================== Bookmark the USENIX website for details about our Workshop on Applications of Embedded Systems in San Jose, California in September 2000. USENIX is the Advanced Computing Systems Association, an international not-for-profit society of scientists, engineers, and system administrators working on the cutting edge of systems and software. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: DCOM on VxWorks? Date: 27 Jan 2000 20:16:10 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <86q92a$e6b$1@news1.xs4all.nl> References: Carsten Spiess wrote: : Hello all, : AFAIK WindRiver has anounced a DCOM implementation for VxWorks. : Has someone already seen a version and made experince with it? The manual is available on WindSurf, so if you want to read about it, that is possible. I don't have any experience with it. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STL for Tornado 1.0.1 Date: Thu, 27 Jan 2000 15:14:25 -0600 From: "Gregory E. Allen" Organization: Applied Research Labs - The University of Texas at Austin Message-ID: <3890B531.1197FB8C@arlut.utexas.edu> References: <388F29F2.22C28E18@boeing.com> <86pie2$cff$1@news.mch.sbs.de> Richard Marsh wrote: > We used STLport from http://www.stlport.org/, for gcc 2.7.2, it works OK... You make it sound so easy! I downloaded STLport-3.2.1 and messed with it for an hour before giving up. I can't find anything on the stlport site about VxWorks. Do you know of any net resources which can help to point me in the right direction? Thanks, - -Greg --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dialing in VxWorks Date: Sat, 22 Jan 2000 18:01:31 GMT From: snaphook@southwesternbell.net (Bruce) Organization: SBC Internet Services Message-ID: <3889f038.2039881@news.swbell.net> References: <86cahf$6se3@overload.lbl.gov> In comp.os.vxworks Abhijit Lahiri wrote: >Is there a free modem >dialing software available for VxWorks ? Just open the serial port and write() "ATDT1800CALLME" Bruce in Katy Texas --------------------------- Newsgroups: comp.os.vxworks Subject: Re: when not compiled works fine. Date: Mon, 24 Jan 2000 16:36:23 +0800 From: "Zha Junhai" Organization: Alcatel Message-ID: <388cf16d.0@dnews.sbell.com.cn> References: <85s3du$604$1@nnrp1.deja.com> <01bf6457$079b16c0$140101c0@popp> Hi Could you tell me the "volatile"'s means or where can I find it, thanks Best Regards Tenshou Zha Klaus Popp wrote in message news:01bf6457$079b16c0$140101c0@popp... > Declare your reg pointer as volatile: > > volatile unsigned int *reg; > > Otherwise the optimizer will hide the actual hardware access > > > -- > Klaus Popp > > men GmbH > Klaus.Popp@men.de > www.men.de > > > girishvg@yahoo.com schrieb im Beitrag <85s3du$604$1@nnrp1.deja.com>... > > hello, > > following code tries to write value to some register at some address. > > if a particular part of code is not compiled the function executes > > (prints both entry...exit messages) properly. but when the part of code > > is added, the function stops in the start itself. it does not print > > "entry" message also. what/where could be the problem? > > any help in this regard is greatly appreciated. > > thanks in advance and regards, > > girish. > > > > *** source code *** > > > > #include "vxworks.h" > > #include "intLib.h" > > #include "stdio.h" > > > > /*#define NOWCOMPILE*/ > > #define SOME_ADDRESS ((unsigned int*)(0x82000714)) > > > > void test(int mask) > > { > > int lock; > > unsigned int val, *reg= SOME_ADDRESS; > > > > printf("enter\n"); > > lock = intLock (); > > > > val = *reg; > > val &= ~(1< > #ifdef NOWCOMPILE > > *reg = val; > > #endif > > > > intUnlock (lock); > > printf("return\n"); > > } > > > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > > --------------------------- Newsgroups: comp.os.vxworks Subject: QED RM7000 PROM Monitor Source Package Date: Thu, 27 Jan 2000 17:58:02 -0800 From: Stefan Dinev Organization: ACC Message-ID: <3890F7AA.27692BE1@acc.com> Hi all: I am looking for RM7000 PROM Monitor Source Package with functions for memory/register access, Ethernet/RS-232 file download, flash memory support. Any information? Thanks. Stefan Dinev --------------------------- Newsgroups: comp.os.vxworks Subject: Re: crash when linking un-called function Date: Thu, 27 Jan 2000 20:26:29 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38911A75.4808@despam.pangea.ca> References: <3890749F.E28F3FE8@baynetworks.com> Hi Ofer, Sounds like there is some other source of corruption which is aggravated by your function. What do moduleCheck() & checkStack() tell you? What happens if you change the name of "test"? eg. UsedToBeTest() - -het PS. Ofer Goren wrote: > Embedded SW Team Leader > Email : ogoren@nortelnetworks.com Hmmm. I am soon to be ex-Nortel [NBWA-Winnipeg] - -- "The road to wisdom? Well it’s plain and simple to express: Err and err and err again, but less and less and less." -Piet Hein Harvey Taylor Internet: het@despam.portal.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Booting from a SCSI HD Date: Mon, 24 Jan 2000 06:57:51 +0700 From: "Deni D. Muslim" Organization: IDOLANet Message-ID: <1t7j4.3298$15.17480@news> Hi, does anybody have information on how to boot VxWorks from a SCSI HD ? Particularly in creating partition on HD, formatting, and finally boot from it. Regarda and Thank you. --------------------------- Newsgroups: comp.os.vxworks Subject: How do I boot x86 without BIOS Date: Tue, 25 Jan 2000 04:54:15 -0800 From: Chuck Hardy Organization: Pacific Bell Internet Services Message-ID: <388D9CF7.8CB7F8B0@pacbell.net> Reply-To: crhardy@pacbell.net The Wind River documentation for booting Intel x86 systems seems to imply that there has to be a BIOS to boot vxWorks. They only refer to PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed an x86-based system that boots vxWorks directly from ROM? If so, how do you start on power-up? As I understand it, the x86 starts executing at F000:FFF0. That doesn't leave much room for code, so you have to jump somewhere else to start the boot process. Do the Wind River tools support putting this jmp instruction into ROM, or do I have to somehow patch it in by hand? Any help would be greatly appreciated. Thanks, Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sockets and c++ Date: Thu, 27 Jan 2000 20:52:57 -0800 From: engineer_scotty Organization: Aracnet Internet Message-ID: <389120A9.1823573B@aracnet.com> References: <948892706.324636@big.hacom.nl> <388ef5b6$0$20390@SSP1NO17.highway.telekom.at> Proxy wrote: > > I assume, your class is called socket, and just calling socket() tries to > call a member function of this class. Try ::socket() to call the socket() > API function > > HTH, > Thomas > > Frank van Eijkelenburg wrote in message <948892706.324636@big.hacom.nl>... > >I'm trying to make a class in C++ with socket functionallity. When I use > the > >function socket (with 3 integers as input) I get the following error > >message: > > > >socks.cpp:21: no matching function for call to `socket::socket (int, int, > >int)' > >D:\\TORNADO/target/h/net/socketvar.h:135: candidates are: > >socket::socket(const socket &) > >D:\\TORNADO/target/h/net/socketvar.h:135: socket::socket() > > > >If I make a "normal" C-file I can use this function without any problems. More specifically, calling "socket" from within "class socket" tries to invoke a constructor. As "Proxy" writes, use ::socket. AND.....rename your class to Socket or MySocket or something like that. Having a class with the same name as an API function is generally not a good idea..... - -- - ------------------------------------------------------------------------ engineer_scotty (no, not that one) -- consumer of fine ales and lagers some days you're the fire hydrant, some days you're the dog | go blazers no small furry creatures were harmed in the creation of this .signature - ------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: MibWay for VxWorks Date: Fri, 28 Jan 2000 10:42:48 +0530 From: Yusuf Motiwala Organization: Lucent Technologies, Columbus, Ohio Message-ID: <38912550.4F448385@lucent.com> Hi, Is anybody has used the RapidLogic product, "MIBWay for Wind Web Server" Product URL: http://www.rapidlogic.com/products/mibway_wws.html We want to use this product to implement the WEB based management (and possibly CLI also). We are using WindNET SNMP and would not like to change our SNMP implementation. I would like to get any feedback on this product, about technical issues, performance, stability, documentation etc. I will very much appreciate your feedback. Thanks in Advance, Yusuf --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Booting from a SCSI HD Date: Mon, 24 Jan 2000 14:29:35 GMT From: xrad@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86hnk9$eql$1@nnrp1.deja.com> References: <1t7j4.3298$15.17480@news> Okay here goes. Boot a vxWorks kernel that has support for your SCSI hardware. In the shell type the following commands: pSpd=scsiPhysDevCreate(pSysScsiCtrl,SCSIID,LUN,0,-1,0,0,0) This creates a physical device pointer to the drive at SCSIID and LUN. pSbd=scsiBlkDevCreate(pSpd,SIZE,0) This creates a block device pointer of SIZE (generally 512 bytes) blocks (use 0 for the entire disk). pDosVolDesc=dosFsMkfs("/sd0/",pSbd) This creates and formats the filesystem and returns a dos volume descriptor pointer. At this point the disk is set up with a DOS partition at the beginning of the disk that is able to hold the boot files. Now you can copy the vxWorks kernel, symbol, and application files to the DOS volume. This can be done via FTP or NFS. Once the files are present you need to have bootrom code that has SCSI boot enabled. This can be checked by booting to the parameters part of booting and type '?'. The bootable devices will be listed at the end of the help. If SCSI is listed you are ready to go, otherwise the bootrom needs to be rebuilt (see your BSP docs). The parameters to boot the scsi are: Boot Device: scsi=SCSIID,LUN Kernel location: /sd0/vxWorks The only thing left that I am not sure if it is necessary is a routine that is present in my BSP. The routine is called sysScsiConfig. This routine is called from usrScsiConfig in src/config/usrScsi.c if SYS_SCSI_CONFIG is defined. I initialize my DOS volume in this routine (same steps as above except use dosFsDevInit instead of dosFsMkfs). I do this because I think that it needs to be initialized to load the symbol table and application from the DOS volume. I have not tested the booting without implementing the sysScsiConfig routine. Hope this helps, Ron In article <1t7j4.3298$15.17480@news>, "Deni D. Muslim" wrote: > Hi, does anybody have information on how to boot VxWorks from a SCSI HD ? > Particularly in creating partition on HD, formatting, and finally boot from > it. > > Regarda and Thank you. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: Mon, 24 Jan 2000 11:08:34 -0500 From: "JohnDoc" Organization: UUNET WorldCom server (post doesn't reflect views of UUNET WorldCom Message-ID: <86htc2$inn$1@soap.pipex.net> References: <86a2ge$601$1@soap.pipex.net> <86g1u2$s9s$1@news1.xs4all.nl> wrote in message news:86g1u2$s9s$1@news1.xs4all.nl... > John Turner wrote: > : I'm sure that this is everyone's favorite problem when they first encounter > : Tornado v2. > > : Even though I have provided the appropriate list of include directories (by > : specifying the compiler options -I/myinc1 -I/myinc2) and manually re-run the > : dependency generator, the Tornado IDE persists in generating an incorrect > : project and an incorrect makefile. Specifically it lists certain header > : files as dependencies but it gets their paths wrong (actually it gives no > : path) and consequently any attempt to build the project fails with "*** No > : rule to make target 'xyz.h'". > > The use of the VPATH macro will correct this problem. Take a look at my > VxWorks page on a detailed description of this problem. > > Groeten, > Johan > Thanks Johan. Interestingly, I found your FAQ at the weekend. Good work. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: External Dependencies for T2 under UNIX Date: Thu, 27 Jan 2000 08:40:31 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86rej5$7qb$1@newsmaster01.magnet.at> References: <86cahe$6sf2@overload.lbl.gov> I had a similar prioblem about 1.5 weeks ago and discussed it here, but under NT. on the MACROS page i had to add a VPATH which points to the directories where my headers are. the problem is, that tornado starts the compile / make somewhere in the tornado directory. this was the reply i received: Add VPATH=xxx:yyy:zzz to Macro folder and -Ixxx -Iyyy -Izzz to the c/c++ compiler parameters folder in your build properties, with xxx,yyy,zzz as an example for three external directories. - -- regards, Mario Semo. http://www.kirchnersoft.com Reyhan ERGUN wrote in message <86cahe$6sf2@overload.lbl.gov>... >This is a multi-part message in MIME format. >--------------42CF4880C222A3B3CE5AD5E1 >Content-Type: text/plain; charset=us-ascii >Content-Transfer-Encoding: 7bit > >I am using T2 under UNIX and making a vxWorks build. >I have 2 problems while generating external dependency: >1-) I am doing a simple trial and including some dummy header files in >usrAppInit.c, after generating dependecy for all project files or for >this file only, it doesn't make any change, I couldn't see those dummy >header files under external dependency tap in the workspace and I can't >see those header files in Makefile ad dependency for usrAppInit.o, under >vxWorks properties application initialization macro is included! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: System time in vxWorks Date: Thu, 27 Jan 2000 08:44:31 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86rejb$7qb$2@newsmaster01.magnet.at> References: <86h8qq$6sf7@overload.lbl.gov> <86ng88$jpn$1@newsmaster01.magnet.at> <388F4B00.2136@ti.com> - -- regards, Mario Semo. http://www.kirchnersoft.com Frank Minich wrote in message <388F4B00.2136@ti.com>... >Mario Semo wrote: >> i am confused. > >The PC has a time of day clock, which maintains and continues >to update its 'time of day' even across PC power cycles. As >VxWorks boots, it should obtain the current time from the time >of day clock and jam it into VxWorks' clock. I don't think >this happens automatically in VxWorks 5.3. i use T2, so i have 5.4. but as it looks to me, it doesnt happen here too. o, i really have to tell vxWorks the current TimeAndDate? is there no way to automatically initialize the TAD from the RTC in the CMOS ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target - more questions Date: Thu, 27 Jan 2000 08:52:09 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86reje$7qb$3@newsmaster01.magnet.at> References: <86a2ge$601$1@soap.pipex.net> <38889536.484F6670@t-online.de> <86a7pv$8gl$1@soap.pipex.net> <86ae0e$b7s$1@soap.pipex.net> <86ak7l$dp6$1@soap.pipex.net> <86njh5$qgb$1@nnrp1.deja.com> we have this situation always. since we have a LOCAL source tree on all workstations (thats where developers are working with, and this tree just contains the files the developer needs), and we have a GLOBAL TopVersion source tree, on a server, shared, readonly, which contains ALL files. now we want to compile in a way, that always a local .cpp file is used to compile when it is available, and a global else. the same for .hpp files. therefore we have our own build environment which handles this perfectly. (e.g. we disable all include search pathes, we use only <...> for include to disable local searches (as happens with "...". its very easy to have 2 different include file versions in one compile when you use "..." includes) and explicitly define the inlcude pathes on the command line. For .cpp compiles we use inference rules (insread of dependency) and we use nmake from VC++ 6.0. But, as you, i have not found a single way to do something similar inside tornado. so we just use the tools, but not the IDE. if you find a way to solve your problem from inside ide, i would be happy to read your solution. - -- regards, Mario Semo. http://www.kirchnersoft.com zeaflal@my-deja.com wrote in message <86njh5$qgb$1@nnrp1.deja.com>... > >If the file is included in a local CPP file, the local HPP file will be >used for dependency checking. If the HPP file is also included in >either a master directory CPP or another HPP that is in the master >directory only, then the master directory HPP will be used for >dependency checking. This checking of the local directory first can >occur at any level of #include nesting. This means that two different >versions of the HPP will be used in one build for dependency checking. >This is easily confirmed by finding the same file name twice in the T2 >dependency list with properties showing different file locations or by >examining the MAKE file and finding different CPP files referencing >different locations of the HPP file. This is very definitely NOT what >we want. > . --------------------------- Newsgroups: comp.os.vxworks Subject: Re: thumb mode on arm7 TDMI processor Date: Fri, 28 Jan 2000 08:49:50 +0000 From: john Rose Organization: Cadence Design Systems, Inc. Message-ID: <3891582E.C7E1C0A5@cadence.com> References: <389041C6.D166A6A1@cadence.com> <86r50g$dcn$1@nnrp1.deja.com> john_94501@my-deja.com wrote: > > Silly question, but you are porting the ARM version of the 7TDMI bootrom > and not the Thumb one aren't you? If you are using the standard WRS BSPs > then the pid7t is ARM mode and the pid7t_t is Thumb mode. The archive > files follow a similar convention: libARM7TDMIgnuvx.a is the ARM code > and libARM7TDMI_Tgnuvx.a is Thumb mode. > > Just a thought, > > John... > Not a silly question, I wasn't sure what the t on the other version stood for. The answer is no, I am using the arm version. John --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Fri Jan 28 06:04:47 2000 From: Abhijit Lahiri Date: Fri Jan 28 06:04:51 PST 2000 Subject: Re: question > muxBind() is for any protocol to bind to the mux given a driver name. ipAttach is a generic attach routine for the TCP/IP stack. regards, Abhijit > > Hi,everyone > > Recently a question confusing me is that: > What's the difference or the association between the two functions = > muxBind( ) and ipAttach( )?=20 > When we have prepared a network device, which one above can connect = > my device with the vxWorks's network protocal. > =20 > Best regard > queent From vxwexplo-errs@csg.lbl.gov Sat Jan 29 04:03:11 2000 From: Vxworks Exploder Date: Sat Jan 29 04:03:14 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jan 29 04:03:08 PST 2000 Subject: send() doesn't fail? Subject: Re: Looking for information on Ethernet with VxWorks Subject: cannot boot VxWorks on MCPN750 Subject: Re: *** No rule to make target Subject: Re: excHookAdd Subject: Re: cannot boot VxWorks on MCPN750 Subject: IDE CDROM driver support on PC platform Subject: routing code Subject: Re: cannot boot VxWorks on MCPN750 Subject: Re: excVecInit()? Subject: vxWorks memory protection Subject: Re: A Question about BSP Subject: Re: send() doesn't fail? Subject: Re: CORBA on VxWorks Subject: Re: cannot boot VxWorks on MCPN750 Subject: Re: cannot boot VxWorks on MCPN750 Subject: Re: Tornado & MPC823 Subject: Re: Problems with Booting from Network Subject: test Subject: How to get routing table entries?... newbie Subject: DCOM on VxWorks? Subject: thumb mode on arm7 TDMI processor Subject: Re: STL for Tornado 1.0.1 Subject: Re: test Subject: Re: send() doesn't fail? Subject: MPC860-Ethernet-vxWorks Subject: Re: MPC860-Ethernet-vxWorks Subject: Re: How do I boot x86 without BIOS Subject: FTP Server with PPP problem Subject: Re: vxWorks memory protection ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: send() doesn't fail? Date: Tue, 25 Jan 2000 08:11:40 GMT From: real@email.in.sig (Amit Raam) Organization: Unspecified Organization Message-ID: <38915a46.588878202@news.ibm.net.il> Hi, Is send() supposed to fail (return ERROR) when the remote machine has crashed? I'm trying to communicate between two vxWorks computers; I've opened a socket, and sent data successfully. However, when the receiving computer resets, the sending computer can continue to send() without any error indication. What am I doing wrong? the code looks something like this: ... int optOn = 1; socketFd = socket(AF_INET, SOCK_STREAM, 0); setsockopt(socketFd, IPPROTO_TCP, TCP_NODELAY, (char *) &optOn, sizeof(optOn)); setsockopt(socketFd, SOL_SOCKET, SO_KEEPALIVE, (char *) &optOn, sizeof(optOn)); serverAddr.sin_family = AF_INET; serverAddr.sin_port = htons(pConn->connectionPort); serverAddr.sin_addr.s_addr = pConn->connectionAddress; connect(socketFd, (struct sockaddr *) &serverAddr, sockAddrSize); .. Later, I send data: write(socketFd, (char *)&connType, sizeof(connType)); tried also: send(socketFd, (char *)&connType, sizeof(connType), 0); Thanks for any idea, Amit Raam Email address: raama at indigo.co.il - -------------------------------------------------- ".sigless and proud of it" (R) (C) (tm) (look-and-feel) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Looking for information on Ethernet with VxWorks Date: Fri, 28 Jan 2000 08:50:44 -0600 From: "Richard B. Ralston" Organization: Rockwell Automation Message-ID: <3891af15$0$2951@news.execpc.com> References: <86kln8$frf3@overload.lbl.gov> I have not seen exactly what you have, but I have noticed problems with ENet on a PPC. I am using EST's MDPro board and an 860T. I'm using the EST BSP to drive the FEC. I'm making a standard build with only part of the EST demo code linked in. (The demo code just flashes the LED's.) I added a call to "ping" to ping my gateway after each LED test. I noticed that after a while the LED's stopped flashing. It turned out the task was stuck in some sort of mem free routine. It never came out. I then took out the call, rebuilt and started the board running. I then when to my NT machine and did a "ping -t" to the dev board. Again, "after a while" (and not very long at that), the ping stopped getting responses from the dev board. I check the net task, and it was suspended too. (I didn't check to see where the net task was stuck.) I've asked WRS and EST about this, but haven't heard anything from them yet. Hopefully, I will soon. - -Rich "Jim Way" wrote in message news:86kln8$frf3@overload.lbl.gov... > VxWorkers, > > I'm using VxWorks for a PowerPc (PPC) with Ethernet. We're seeing some > interesting behavior (see below) with the units when they are connected > directly to our office network (Microsoft NT servers, email servers, a Linux > box or two, typical office environment stuff). When we connected our target > directly to a pc through a hub, bypassing the network, it ran ok. We set up > another test yesterday to connect our target to the office network through a > hub (rather than a direct-connect) and as of this morning it is still > running strong. > > Does anyone have any ideas, advice, or suggestions for what might be going > on and how we might be able to fix it? > > ALL comments will be greatly appreciated. > > Thanks! > Jim Way > > ------------------------- > The Interesting Behavior: > 1. Monitoring two units via Telnet sessions, both died at the same time > (observed one time). > 2. Monitoring a unit over the weekend, beginning about 4pm on Friday, the > unit ran fine until about 8:30am Monday morning. > 3. Units connected directly to the network fail, average time is less than > four hours. > 4. The software continues to run, but anything involving Ethernet, such as > Telnet sessions, goes away. > ------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: cannot boot VxWorks on MCPN750 Date: Fri, 28 Jan 2000 10:54:00 -0600 From: "Tim Berger" Organization: Motorola Message-ID: <86shh2$54q3@nntp.cig.mot.com> Hello, Booting the MCPN750 board with the Tornado II boot loader fails after programming the flash memory; there's no response from the board after we reboot. However, the Tornado I boot loader starts up successfully. Then, when we try to load the VxWorks image, the board freezes after the "Starting at 0x100000" line. We've tried using version 1 and 2 VxWorks images with the same result. Here's a log of the serial port output just before the board freezes: ... boot device : dc processor number : 0 host name : Tom file name : vxWorks inet on ethernet (e) : 10.0.0.10 host inet (h) : 10.0.0.1 user (u) : guest ftp password (pw) : guest flags (f) : 0x0 target name (tn) : mcpn750 Attaching network interface dc0... done. Attaching network interface lo0... done. Loading... 705916 Starting at 0x100000... [freeze] Does anybody have any clues as to what the problem might be? Anything we should try? Thanks, bergerta@cig.mot.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: *** No rule to make target Date: Fri, 28 Jan 2000 17:53:19 GMT From: Ken Ostaszewski Organization: Boeing, Inc. Message-ID: <3891D78E.2343B9A1@boeing.com> References: <86a2ge$601$1@soap.pipex.net> Sender: nntp@news.boeing.com (Boeing NNTP News Access) Hi, I know that others have said that to get the correct include directories to work you need to include the VPATH macro. However, I added complete path names to the -I command, and it appears to compile for my MV2604 bsp. For example, if you would normally do "-I./Hdr", I do a "-I/Hdr" and the dependencies get generated as they should. However, if "" is a macro, it doesn't get expanded, and I get bad dependencies. Hope it helps, just my 2 cents. Ken Ostaszewski John Turner wrote: > I'm sure that this is everyone's favorite problem when they first encounter > Tornado v2. > > Even though I have provided the appropriate list of include directories (by > specifying the compiler options -I/myinc1 -I/myinc2) and manually re-run the > dependency generator, the Tornado IDE persists in generating an incorrect > project and an incorrect makefile. Specifically it lists certain header > files as dependencies but it gets their paths wrong (actually it gives no > path) and consequently any attempt to build the project fails with "*** No > rule to make target 'xyz.h'". > > Is there any way to correct this behavior other than manually editing the > project file and makefile? > > Thanks very much, > John --------------------------- Newsgroups: comp.os.vxworks Subject: Re: excHookAdd Date: Fri, 28 Jan 2000 13:07:55 -0500 From: Tim Sohacki Organization: Nortel Networks Message-ID: <3891DAFB.C241269E@americasm01.nt.com> References: <86r1k5$n841@overload.lbl.gov> Bob, If I remember correctly, a single hook can be added. The hook is called from the default VxWorks exception handling, I believe near its completion. The global variable excExcepHook is a pointer to a function which is set by the call to excHookAdd(). I thought that multiple calls to excHookAdd() overwrote the old excExcepHook value. (By the way, note that the exception handling code in bootConfig.c is for boot ROM images, not vxWorks images.) Cheers, Tim. "Ferrara, Bob" wrote: > > > Can someone tell me if excHookAdd() can be called more than once, and have > the routines specified on the excHookAdd calls ALL be invoked when an > exception > occurs? > > It's behaving like it can only be called one time. It is called in > bootConfig.c to add > a boot exception handler (standard VxWorks) and I have called it a 2nd time > to add > my own specific things to do when an exception occurs. My exception routine > only > gets invoked when I comment out the first call to excHookAdd in bootConfig.c > > Thanks, > Bob Ferrara > bferrara@unispheresolutions.com > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: Fri, 28 Jan 2000 11:55:18 -0600 From: Frank Minich Organization: NCP Message-ID: <3891D806.198B@ti.com> References: <86shh2$54q3@nntp.cig.mot.com> Reply-To: fminich@ti.com Tim Berger wrote: > there's no response from the board after we reboot. > Loading... 705916 > Starting at 0x100000... > [freeze] > > Does anybody have any clues as to what the problem might be? > Anything we should try? I'm really guessing (no experience with MCPN750 or T2), but on my system I get Loading... xxxxxx + yyyyyy + zzzzzz where I think the numbers represent something like code size, bss, etc. This makes me think that maybe the file you are loading ("vxWorks") is not a valid boot image. The one I load is named "vxWorks.st". Good luck, Frank. --------------------------- Newsgroups: comp.os.vxworks Subject: IDE CDROM driver support on PC platform Date: Fri, 28 Jan 2000 18:19:48 GMT From: "Ilya Klebanov" Message-ID: <85lk4.148$K5.609810@bunson.tor.sfl.net> Does anybody know, how to make CDROM drive visibale under VxWorks on a PC platform? Thanks, Ilya --------------------------- Newsgroups: comp.os.vxworks Subject: routing code Date: Fri, 28 Jan 2000 10:42:06 -0800 From: ravi Organization: InterNex Information Services 1-800-595-3333 Message-ID: <3891E2FE.512785C4@vitesse.com> Hello Everyone, We are trying to build a L2/L3 switch. I was just wondering if anyone is aware of companies that sell the code for implementation along with support of Routing protocols like RIP,OSPF,BGP and bridging over ethernet. I would appreciate if u could mail me the names of those companies along with their contacts if possible. Thanks Ravi ravi@vitesse.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: Fri, 28 Jan 2000 22:11:44 +0300 From: "Arthur V. Grigorian" Organization: Auriga, Inc. Message-ID: <3891E9F0.5C2101C4@auriga.ru> References: <86shh2$54q3@nntp.cig.mot.com> Reply-To: archy@auriga.ru Hello, > Booting the MCPN750 board with the Tornado II boot loader fails after > programming the flash memory; there's no response from the board after we > reboot. However, the Tornado I boot loader starts up successfully. Then, > when we try to load the VxWorks image, the board freezes after the "Starting > at 0x100000" line. We've tried using version 1 and 2 VxWorks images with > the same result. Here's a log of the serial port output just before the > board freezes: > > ... > boot device : dc > processor number : 0 > host name : Tom > file name : vxWorks > inet on ethernet (e) : 10.0.0.10 > host inet (h) : 10.0.0.1 > user (u) : guest > ftp password (pw) : guest > flags (f) : 0x0 > target name (tn) : mcpn750 > Attaching network interface dc0... done. > Attaching network interface lo0... done. > Loading... 705916 > Starting at 0x100000... > [freeze] > > Does anybody have any clues as to what the problem might be? Anything we > should try? - Please make sure that you build your vxWorks images with optimization level = 2 ( see properties for your project build(s) ) This problem can also be a result of your changing some header files in the BSP. - As for MCPN750 bootrom.bin in T2 -- have you tried to flash the bootrom.bin file which came with the BSP and what were the results? - If you use transition module TMCPN750 with IDE HD then such problem can be caused by your chassis (I saw it on CPX8216) Contact me if you are facing the same problem. Regards, __ R2 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: excVecInit()? Date: 28 Jan 2000 13:15:39 -0600 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <86bvem$jnk$1@nnrp1.deja.com> Sender: bpringle@DeadDuck I found the following load constant code in excVecInit(). mvn r3, #22000 sub r3, r3, #100 sub r3, r3, #18000001 I also printed the Vector table and it seems to be full of 0xe7fddefe. I put this in an assembler file (.long) and it disassembles to an illegal instruction. My kernelInit is hanging when I get to that in usrInit(). I have mapped 0xc1000000 to 0 and 0xc0000000 to 0x1000000. I don't think this is a problem. I have re-written the phys to virt (e.v.v) to account for this and the MMU table initialization in sysLib.c I have some code to emit on the console in the interrupt driver code and at the start of usrRoot(). All these lines are followed by while(1); to make sure that the console prints them. I have verified that my data and bss are correct. Here are the parameters to kernelInit() About to enter kernelInit(). ROOT_STACK_SIZE 00004000 FREE_RAM_ADRS 0007D2A8 sysMemTop () 01FFC000 ISR_STACK_SIZE 00000800 INT_LOCK_LEVEL 00000000 I use the same polling serial output routines in the interrupt driver and the usrRoot(); that was from the console. I also have a line to clear the interrupt mask (all off) before the call to kernelInit(). I also tried to set the I and F bits of the cpsr and that didn't not cause it to hang. The values in the vector table look suspicious to me. Is anything else done besides enabling I and F bits? thanks again, Bill ps, I am basing this off the brutus BSP. I am porting to the SA1110. These boards have 32 MB of SDRAM. I set the MMU table at the top of memory and the low memory is identical to the brutus BSP. >>>>> "john" == john 94501 writes: john> Assuming that you started with the SA-1100/brutus BSP, you john> would be very well advised to stick with the memory layout john> that it uses... john> During the initialisation a section mapping table is created john> which maps the 4 banks of memory into one contiguous block john> starting at zero. excVecInit() is trying to install the john> exception vectors at zero since that is where the CPU john> expects them to be (unless you are in the CE model, where john> things are a little different). - -- ..The TENSION mounts as I MASSAGE your RIGHT ANKLE according to ancient Tibetan ACCOUNTING PROCEDURES..are you NEUROTIC yet?? --------------------------- Newsgroups: comp.os.vxworks Subject: vxWorks memory protection Date: Fri, 28 Jan 2000 15:03:04 -0500 From: usirsiwal@nexabit.com Message-ID: <3891F5F8.176AEAB9@nexabit.com> Hello! I am trying to implement a simple memory protection for vxWorks (Tornado 1.0.2 PowerPC). My model is simple. I have a special taskSpawn, which allocates the stack space, calls taskInit, and taskActivate. I put a stamp in reserved2 field of TCB to indicate that their stack is safe. In taskSwitchHook, I check if this is a safe task. If so, I write protect the old task stack space in the page table, and if unprotect the stack space for new task. But, sometimes I get data access exception for resons unknown to me. I do know that the exception is data access... Any help for this will be appreciated! - -Umesh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: A Question about BSP Date: Thu, 27 Jan 2000 07:44:53 GMT From: zhouxiaochen@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86ot1k$oik$1@nnrp1.deja.com> References: <86hnit$eq6$1@nnrp1.deja.com> <86mh7n$3g7$1@news.mch.sbs.de> I only allocate 3M memory for vxWorks. Admire you! Thank you very much! In article <86mh7n$3g7$1@news.mch.sbs.de>, "Richard Marsh" wrote: > I can think of 2 possible problems: > > 1. If you have the hardware watchdog enabled then adding that amount of > memory takes longer than the timeout period > > or > > 2. You don't have enough memory for the MMU page descriptors (the region is > broken down into 4Kbyte pages and your system memory used for storing these > tables) > > HTH > > Richard > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: send() doesn't fail? Date: Thu, 27 Jan 2000 08:11:11 GMT From: real@email.in.sig (Amit Raam) Organization: Unspecified Organization Message-ID: <388ffd0b.52437651@news.ibm.net.il> References: <38915a46.588878202@news.ibm.net.il> <86k8fs$iql@niven.ksc.nasa.gov> Hi, Thanks for your reply; Do I understand you correctly, I have to rely on SO_KEEPALIVE to detect a failed send() ? this is not a dropped inactive connection, but an active TCP/IP session. Shouldn't the TCP layer detect the lack of an ACK from the far side? Thanks, Amit Charles H. Chapman wrote: >Amit Raam wrote: >> Hi, >> Is send() supposed to fail (return ERROR) when the remote machine has crashed? > >The default parameters for KEEP ALIVE are set to detect a lost >connection after some ridiculously long time like two hours. You can >change this time by modifying the values of the variables tcp_keepintvl and >tcp_keepidle. See the manual page for setsockopt under the SO_KEEPALIVE >option. > >Chuck Amit Raam Email address: raama at indigo.co.il - -------------------------------------------------- ".sigless and proud of it" (R) (C) (tm) (look-and-feel) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CORBA on VxWorks Date: Fri, 28 Jan 2000 21:32:29 GMT From: shibu_pillai@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86t1te$q9r$1@nnrp1.deja.com> References: <85u217$f63$1@nnrp1.deja.com> We are using Visibroker on VxWorks, supplied by Highlander Communications. Check their website at: http://www.highlander.com/ - -- In article <85u217$f63$1@nnrp1.deja.com>, st1734 wrote: > > > I'm looking for CORBA vendors on VXWORKS. > > Any pointers ? > > Sent via Deja.com http://www.deja.com/ > Before you buy. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: Fri, 28 Jan 2000 15:20:24 -0600 From: "Tim Berger" Organization: Motorola Message-ID: <86t14i$54t6@nntp.cig.mot.com> References: <86shh2$54q3@nntp.cig.mot.com> <3891D806.198B@ti.com> I always thought there was something fishy about the number after "Loading," because it never corresponds to anything. It always hovers around 705,000, and most of the vxWorks files we are trying to use are over a meg in size. But anyways, we tried a vxWorks.st file we found... no luck. Thanks for the suggestion, we'll keep trying... Frank Minich wrote in message news:3891D806.198B@ti.com... > I'm really guessing (no experience with MCPN750 or T2), but on > my system I get > Loading... xxxxxx + yyyyyy + zzzzzz > where I think the numbers represent something like code size, > bss, etc. > This makes me think that maybe the file you are loading ("vxWorks") > is not a valid boot image. > > The one I load is named "vxWorks.st". --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: Fri, 28 Jan 2000 15:47:29 -0600 From: "Tim Berger" Organization: Motorola Message-ID: <86t2na$54q5@nntp.cig.mot.com> References: <86shh2$54q3@nntp.cig.mot.com> <3891E9F0.5C2101C4@auriga.ru> Arthur V. Grigorian wrote in message news:3891E9F0.5C2101C4@auriga.ru... > - Please make sure that you build your vxWorks images with optimization > level = 2 ( see properties for your project build(s) ) > This problem can also be a result of your changing some header files > in the BSP. Changed it to level 2... the image still won't start. > - As for MCPN750 bootrom.bin in T2 -- > have you tried to flash the bootrom.bin file which came with the BSP > and what were the results? As far as I know, there was no pre-built bootrom.bin file in the MCPN750 BSP we got. We built a bootrom_uncmp, which we then ELFToBin'd to a bootrom.bin file. After flashing and rebooting the board, the result was nothing -- blank screen (on the serial-port terminal) and no response. We tried building the compressed bootrom also...same result. > - If you use transition module TMCPN750 with IDE HD then such problem > can be caused by your chassis (I saw it on CPX8216) Contact me if you > are facing the same problem. We're using a TMCPN710... The chassis is a CPX1204. I've looked through all the T2 release notes, BSP how-to guides, etc., and made sure that I followed all the directions correctly. Someone here suggested that it may be a problem with the Raven chipset; the newest MCPN750's have new chipsets that may not work correctly with the boot loader (some addressing issue). Anyways, we're taking this issue up with Windriver support at this point. We've tried just about everything except chanting. Thanks for your suggestions, Tim bergerta@cig.mot.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado & MPC823 Date: Thu, 27 Jan 2000 17:34:54 +0800 From: Wilson Lo Organization: Motorola Semiconductor Products Sector Message-ID: <3890113E.BF6D4D6A@email.sps.mot.com> References: <869e7u$rnq$1@minus.oleane.net> Reply-To: Wilson.Lo@motorola.com Hi, U may use MPC823ADS. It is a good reference design including many interface. BSP of MPC860 is available. MPC860 is similar to MPC823. It will reduce much time for design. regards Wilson Alex U wrote: > Hi group, > > We are to be starting a new design based around the Motorola MPC823 and powered by > VxWorks. > > Which Evaluation Board(s) is(are) compatible with VxWorks BSPs ? > > Which one would you recommend (preferably with direct connection to an LCD screen) ? > > What about those available through EST ? > > Thanks for your help, > > Alex Ungerer - -- ======================================== Wilson Lo Software Application Engineer Asia Pacific Operation Networking and Communications Systems Division Motorola Semiconductor Products Sector ======================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with Booting from Network Date: Thu, 27 Jan 2000 18:02:27 +0800 From: Wilson Lo Organization: Motorola Semiconductor Products Sector Message-ID: <389017B3.84249549@email.sps.mot.com> References: <85vh0h$g17$1@nnrp1.deja.com> Reply-To: Wilson.Lo@motorola.com Hi, The error code "0xd0003" means that the board can't from the image file. U may check if the path of FTP is correct. I use the "Mircosoft Peer web Service" to act as a FTP server. It is Ok. and I think the problem is not due to the "/". One thing to remain is that U need to setup the path name of directory in the Web Service. regards Wilson bfeng@noran.com wrote: > I am trying to boot a Pentium PC via a network using a default factory > Pentium BSP boot disk. I managed to change the booting parameters after > "[boot rom]" prompt showed up. A ftp server "wftpd.exe" from > "Tornado2.0" CD is always launched on my host NT computer. The network > connection was tested OK while booting the Pentium PC into Win95. The > FTP login user name and password were also tested OK in Win95. However > everytime I try to boot it with Vxworks, I have the following error > messages. I appreciate anyone's hint. What does error code "0xd0003" > actually mean? > > boot device : ultra > unit number : 0 > processor number : 0 > host name : bfeng2.YP.solsys > file nmae : /tornado/target/config/bobPentium/vxworks.st > inet on ethernet (e) : x.x.x.x:ffffff00 > host inet (h) : z.z.z.z > user(u) : bfeng > ftp password(pw) : xxxxx > flags (f) : 0x0 > > Attaching TCP/IP interface to ultra0. > Attaching network interface lo0... done. > Loading... > Error loading file: errno = 0xd0003. > dosFsDevInit failed. > userFdConfig failed. > > Sent via Deja.com http://www.deja.com/ > Before you buy. - -- ======================================== Wilson Lo Software Application Engineer Asia Pacific Operation Networking and Communications Systems Division Motorola Semiconductor Products Sector ======================================== --------------------------- Newsgroups: comp.os.vxworks Subject: test Date: Thu, 27 Jan 2000 02:57:16 -0800 From: vxworks-ppc Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <06949404.afcc50ea@usw-ex0106-046.remarq.com> test * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free! --------------------------- Newsgroups: comp.os.vxworks Subject: How to get routing table entries?... newbie Date: Fri, 28 Jan 2000 23:46:47 GMT From: glmohan@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86t9p6$3q$1@nnrp1.deja.com> Hi, Could anyone tell me how to retrieve the routing table entries in VxWorks?. Is routeShow source code distributed with VxWorks?. It would be sufficient if I can get the default gateway address. Does VxWorks support standard ioctl calls to get/set network related parameters?. ( like SIOCGIFADDR, SIOCGIFNETMASK, etc.) I have seen ifAddrGet, ifGetNetmask functions but they are just not portable. Are they?. TIA, Mohan. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: DCOM on VxWorks? Date: Thu, 27 Jan 2000 13:29:37 +0100 From: Carsten Spieß Message-ID: Hello all, AFAIK WindRiver has anounced a DCOM implementation for VxWorks. Has someone already seen a version and made experince with it? Thank's for your reply's Carsten Spiess --------------------------- Newsgroups: comp.os.vxworks Subject: thumb mode on arm7 TDMI processor Date: Thu, 27 Jan 2000 13:01:58 +0000 From: john Rose Organization: Cadence Design Systems, Inc. Message-ID: <389041C6.D166A6A1@cadence.com> Hi, Can anyone think of a reason why the bootrom code for VxWorks is forcing the Arm processor I am trying to port it to to run in thumb mode? Thanks John Rose --------------------------- Newsgroups: comp.os.vxworks,gnu.gcc.help Subject: Re: STL for Tornado 1.0.1 Date: Thu, 27 Jan 2000 14:50:17 +0100 From: "Richard Marsh" Organization: Siemens Inc. Message-ID: <86pie2$cff$1@news.mch.sbs.de> References: <388F29F2.22C28E18@boeing.com> Reply-To: "Richard Marsh" Hi, We used STLport from http://www.stlport.org/, for gcc 2.7.2, it works OK but watch out as you have to play games to ensure that you instantiate once, otherwise your code will bloat. Richard --------------------------- Newsgroups: comp.os.vxworks Subject: Re: test Date: Thu, 27 Jan 2000 06:03:00 -0800 From: vxworks-ppc Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <1d0398b0.e03f3e1d@usw-ex0106-046.remarq.com> References: <06949404.afcc50ea@usw-ex0106-046.remarq.com> done * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: send() doesn't fail? Date: Thu, 27 Jan 2000 08:58:18 -0500 From: Charles H. Chapman Organization: Dynacs Engineering Company Message-ID: <86piu5$5bm@niven.ksc.nasa.gov> References: <38915a46.588878202@news.ibm.net.il> <86k8fs$iql@niven.ksc.nasa.gov> <388ffd0b.52437651@news.ibm.net.il> Reply-To: chc@nasa2.ksc.nasa.gov > Thanks for your reply; Do I understand you correctly, I have to rely on > SO_KEEPALIVE to detect a failed send() ? this is not a dropped inactive > connection, but an active TCP/IP session. Well, you said in your original post that the remote machine had crashed so that probably wouldn't be an active TCP/IP session any more. > Shouldn't the TCP layer detect the > lack of an ACK from the far side? I'm not real familiar with how all that works but the send() definitely won't wait for that ACK to come back before returning. The place you're sending to could be 10,000 miles away, going through many hops so it could be a significant amount of time before that ACK makes it back. Your program probably doesn't want to wait around that long so I think send() returns as soon as the packet is accepted for transmission (probably only fails if you have a buffer overflow or something like that on the transmit side). Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: MPC860-Ethernet-vxWorks Date: Thu, 27 Jan 2000 15:18:38 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <86pk2r$bl8$1@crusher.de.colt.net> Windriver adjust the ethernet MAC address for PPC cpm's in syslib.c in a variable called "sysCpmEnetAddr". This is a hard compiled value. In runtime it should be possible to change this address. The problem is: 1. if we change this variable-> no effect to the address 2. if we change the physical address pins in the Ethernet parameter RAM of the PPC the connection to Tornado goes down. You can ping the target, but MAC address didn't change. What must we do with the Ethernet SCC do make it correct? Thanks Achim --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC860-Ethernet-vxWorks Date: Thu, 27 Jan 2000 15:24:48 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <86pke9$bsd$1@crusher.de.colt.net> References: <86pk2r$bl8$1@crusher.de.colt.net> Sorry, I dont mean address pins, I mean the address register (PADDRx_x) in problem 2. > Windriver adjust the ethernet MAC address for PPC cpm's in syslib.c in a > variable called "sysCpmEnetAddr". This is a hard compiled value. > In runtime it should be possible to change this address. > The problem is: > 1. if we change this variable-> no effect to the address > 2. if we change the physical address pins in the Ethernet parameter RAM of > the PPC > the connection to Tornado goes down. You can ping the target, but MAC > address didn't change. > > What must we do with the Ethernet SCC do make it correct? > > Thanks Achim > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I boot x86 without BIOS Date: Tue, 25 Jan 2000 08:12:11 -0600 From: "Dennis Gabler" Message-ID: <46ij4.3$6X5.551@news.uswest.net> References: <388D9CF7.8CB7F8B0@pacbell.net> I have no direct experience in doing this with vxWorks but I have done it numerous times with other RTOS's. You are in real mode at reset and as you say execution begins at F000:FFF0. Do an absolute jump back 64K or so and at that point you can place your bootstrapping code to bring up the hardware and if you want, read in the OS image from a drive. The specifics of all this depend on your design of course :-). Chuck Hardy wrote in message <388D9CF7.8CB7F8B0@pacbell.net>... >The Wind River documentation for booting Intel x86 systems seems to >imply that there has to be a BIOS to boot vxWorks. They only refer to >PC-AT systems that boot via hard-disk, tffs, etc. Has anyone developed >an x86-based system that boots vxWorks directly from ROM? If so, how do >you start on power-up? As I understand it, the x86 starts executing at >F000:FFF0. That doesn't leave much room for code, so you have to jump >somewhere else to start the boot process. Do the Wind River tools >support putting this jmp instruction into ROM, or do I have to somehow >patch it in by hand? > >Any help would be greatly appreciated. > >Thanks, >Chuck > --------------------------- Newsgroups: comp.os.vxworks Subject: FTP Server with PPP problem Date: Thu, 27 Jan 2000 15:04:19 GMT From: "Nick Gagnon" Organization: Mitel Corporation Message-ID: I am working on a board using MC68360. We modified the bootRom to allow us to start an FTP deamon and a PPP server. We are also using a TFFS file system. When I am trying to put a file larger than half a K, my system reset without any explanation. I don't have that problem when I am using the FTP deamon over the NET. It is hard to find why I am failling because I don't have any serial interface to get feedback from (It is used by PPP). I can't also used a target server because my system is not Up and Running at that time. I tried to isolate the problem and I found that it is when I am doing an "fioRead" on my socket that the system crashed. If you have any Idea about my problem or any debugging Idea about it , please let me know Regards Nick Gagnon Mitel Corporation 350 Legget Drive P.O. Box 13089 Kanata, Ontario Canada K2K 2W7 Tel: (613) 592-2122 Ext 1797 Fax: (613) 592-4784 E-mail : nick_gagnon@mitel.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks memory protection Date: Sat, 29 Jan 2000 04:15:52 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <86tphl$aoo$1@nnrp1.deja.com> References: <3891F5F8.176AEAB9@nexabit.com> > In taskSwitchHook, I check if this is a safe task. If so, I write > protect the old task stack space in the page table, > and if unprotect the stack space for new task. But, sometimes I get data > access exception for resons unknown > to me. Two things come to mind: 1) Have you taken care of invalidating any TLB entries that may exist? You don't say which PPC you are using, but even the lowliest of 8xx members have some TLB entries. If you are changing the access permissions of pages you need to flush the TLB of any affected entries to make sure that the new task doesn't see those instead of reading the translation tables in memory again. 2) If you have your cache in copyback mode, you want to ensure that you are flushing the lines containing the updated page tables back to memory since the cache is probably bypassed when a TLB miss occurs. Hope that helps, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jan 30 04:03:11 2000 From: Vxworks Exploder Date: Sun Jan 30 04:03:14 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jan 30 04:03:08 PST 2000 Subject: Technobox 2229 pmc serial, mvme2604, V5.4 Subject: Re: Technobox 2229 pmc serial, mvme2604, V5.4 Subject: Re: CORBA on VxWorks Subject: Re: Is there still an comp.os.psos ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Technobox 2229 pmc serial, mvme2604, V5.4 Date: Sat, 29 Jan 2000 12:41:08 -0800 From: Michael Imai Message-ID: <38935063.1310AA25@osl.com> I have a serial port driver for the technobox 2229 8 port serial pmc card which work on the MVME2604 under V5.3.1. We have recently switched to V5.4 (Tornado2) to take advantage of some of the other things officially supported under V5.4. The bsp changed from using pciIomapLib to pciConfigLib. Does anyone know the differences between these? Is there anything I need to initialize? Thanks. mike i. - -- Michael Imai, P.Eng. Senior Software Engineer Offshore Systems Ltd. (604)904-4651 107-930 1st Street Fax: (604)987-2555 North Vancouver, BC V7J 1R8 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Technobox 2229 pmc serial, mvme2604, V5.4 Date: Sat, 29 Jan 2000 14:29:52 -0800 From: Michael Imai Message-ID: <389369E0.B44147CB@osl.com> References: <38935063.1310AA25@osl.com> I forgot to say that the driver doesn't work under v5.4 Michael Imai wrote: > I have a serial port driver for the technobox 2229 8 port serial pmc > card which work on the MVME2604 under V5.3.1. > > We have recently switched to V5.4 (Tornado2) to take advantage of some > of the other things officially supported under V5.4. > > The bsp changed from using pciIomapLib to pciConfigLib. Does anyone know > the differences between these? > > Is there anything I need to initialize? > > Thanks. > > mike i. > > -- > Michael Imai, P.Eng. > Senior Software Engineer > Offshore Systems Ltd. (604)904-4651 > 107-930 1st Street Fax: (604)987-2555 > North Vancouver, BC V7J 1R8 - -- Michael Imai, P.Eng. Senior Software Engineer Offshore Systems Ltd. (604)904-4651 107-930 1st Street Fax: (604)987-2555 North Vancouver, BC V7J 1R8 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CORBA on VxWorks Date: Sun, 30 Jan 2000 02:57:29 GMT From: st1734@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8709an$vhc$1@nnrp1.deja.com> References: <85u217$f63$1@nnrp1.deja.com> <86t1te$q9r$1@nnrp1.deja.com> How much memory does Visi on Vxworks use ? In article <86t1te$q9r$1@nnrp1.deja.com>, shibu_pillai@my-deja.com wrote: > > > We are using Visibroker on VxWorks, supplied by > Highlander Communications. > > Check their website at: http://www.highlander.com/ > > -- > > In article <85u217$f63$1@nnrp1.deja.com>, > st1734 wrote: > > > > > > I'm looking for CORBA vendors on VXWORKS. > > > > Any pointers ? > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is there still an comp.os.psos ? Date: Sun, 30 Jan 2000 08:48:22 GMT From: pwhicker@notthis.home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <388F5AD4.6B8BA43B@megsinet.net> <86nsf4$m70$1@news02.btx.dtag.de> <31o09ssrmt1h2ta45kv2frcgf2ks9v89du@4ax.com> It's there but gets little traffic. Paul In article <31o09ssrmt1h2ta45kv2frcgf2ks9v89du@4ax.com>, Evandro Menezes wrote: >On Wed, 26 Jan 2000 23:27:57 +0100, "Martin Raabe" > wrote in <86nsf4$m70$1@news02.btx.dtag.de>: > >>yes there is but it is very quiet here in comp.os.psos. > >It's still not run by many servers. In general, ISP's only add groups >upon request by the customers. If yours doesn't run it, demand it! >;^) > >______________________________________________________ >Evandro Menezes Austin, TX ICQ:7957253 >evandro@geocities.com geocities.com/evandro --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jan 31 04:03:11 2000 From: Vxworks Exploder Date: Mon Jan 31 04:03:14 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jan 31 04:03:07 PST 2000 Subject: Re: MPC860-Ethernet-vxWorks Subject: fwrite and mempartAlloc error.... Subject: Re: looking for linux embedded system Subject: Re: Serial device Subject: Re: cannot boot VxWorks on MCPN750 Subject: Re: fwrite and mempartAlloc error.... Subject: Re: templates Subject: pci interrupts for MBX 860 BSP Subject: SNiFF+ in Tornado II Subject: Re: CPU util measurement? Subject: Re: CPU util measurement? Subject: Booting from MVME147? Subject: SNMP : Entries in Interfaces MIB Subject: Re: Losing PowerPC in PCI write Subject: Tornado II<->SourceSafe Subject: Re: cannot boot VxWorks on MCPN750 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC860-Ethernet-vxWorks Date: Sun, 30 Jan 2000 16:18:52 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <3894484C.EA75BA1C@bitband.com> References: <86pk2r$bl8$1@crusher.de.colt.net> Reply-To: leonid@bitband.com The problem is eeesntially that when you are succesful in changing the MAC address, you will disrupt the TCP/IP connections which where present at the time, because the MAC address change needs to be reflected in the ARP tables of the peering nodes. This takes some time but eventually may work, e.g. by manually flushing the other host's ARP table (e.g. "arp -d"). In your case however, you changed the MAC address at chip level without modifying the driver's notion of its MAC address and that would cause permamanet and irreprable lack of communications. There should be a way to change the hardware and driver notion of MAC address with an ioctl() to the driver, but depending on your particular driver, this may have not been implemented within the driver. Another idea, is after you changed the variable, try to re-assign the IP address to the network interface, ifAddrSet(), even if the IP address is the same, this may do what you need. HTH, Leonid Achim Zimmer wrote: > > Windriver adjust the ethernet MAC address for PPC cpm's in syslib.c in a > variable called "sysCpmEnetAddr". This is a hard compiled value. > In runtime it should be possible to change this address. > The problem is: > 1. if we change this variable-> no effect to the address > 2. if we change the physical address pins in the Ethernet parameter RAM of > the PPC > the connection to Tornado goes down. You can ping the target, but MAC > address didn't change. > > What must we do with the Ethernet SCC do make it correct? > > Thanks Achim --------------------------- Newsgroups: comp.os.vxworks Subject: fwrite and mempartAlloc error.... Date: Sun, 30 Jan 2000 10:38:20 -0500 From: Christian Williamson Message-ID: <38945AEC.BC466CF5@erols.com> I'm getting a mempartAlloc error while using the fopen-fread-fwrite series of i/o function in VxWorks 5.2 and 5.3. I get this error when I use nfsLib to a file on a Sun workstation, when I use netLib to a file on a Sun workstation, and when I use dosFsLib to a RAM disk. The sizes of the files I'm writing are many megabytes in size (I don't have trouble with 'small' files, in the several hundred kilobyte to megabyte range). I've looked for discussion of this problem and haven't come up with anything matching fopen, fwrite, and mempartAlloc. Does anyone have an idea about this? The sequence in all instances are: fopen(), loop while fread() and fwrite(). I've thought about going to open(), read(), and write(); there is some discussion on WindSurf that makes me think there may be a memory leak, too much space allocated for the f- function calls. Any ideas? Thanks. Chris Williamson --------------------------- Newsgroups: comp.os.vxworks Subject: Re: looking for linux embedded system Date: Sun, 30 Jan 2000 10:32:09 -0600 From: "Stephen B. Johnson" Organization: Newshosting.com Message-ID: <38946788.5C4CCFEC@inxpress.net> References: <856t9q$jlu$4@netnews2.csie.nctu.edu.tw> Reply-To: stevebj@inxpress.net We have a Linux house wanting to put it on one of our PMC based processor modules. PMC is basically PCI in a IEEE standardized (open standard) form factor. I use a PMC-to-PCI adapter for working with the modules in my desktop. Otherwise they go on most proprietary boards with PMC sites and any VME or cPCI boards with PMC sites. The image of the module may make things clear. www.artesyn.com/cp Steve NortonNg wrote: > hi all, > > i am a novice in embedded system. i am looking for a embedded system > os and hardware for building a firewall, and i have to porting the > applications on linux to the embedded system. > anyone can give me some suggestion? Is there have some hardware that > support real-time linux based os? > > jkng. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Serial device Date: Sun, 30 Jan 2000 11:33:57 -0600 From: "Stephen B. Johnson" Organization: Newshosting.com Message-ID: <38947605.C1D25D1D@inxpress.net> References: <10gp79m$7u$2@fe2.cs.interbusiness.it> Reply-To: stevebj@inxpress.net Usually devices like the 683xx and mpc8xx are untouched by the OS (or BSP). The OS usually only provides the two serial console ports which show up as a tyCo device with 'devs'. Devices such as the 683xx and mpc8xx have so many different modes and options that if set up to a particular mode would not be practical. In general, it would be up to the Application to set everything up for it's particular use. We (my company...Artesyn www.artesyn.com/cp) provide protocol solutions with our hardware via in house protocols based on Trillium and external protocol support from Spider Software. Steve Mauro Martorelli wrote: > What are the functions for utilize and configure the serial device (target > with mc68340)? > > Have you any examples? > > Thank You. > > my address martorelli@tnteam.it --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: 30 Jan 2000 20:52:04 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <87289k$s3h$1@news1.xs4all.nl> References: <86shh2$54q3@nntp.cig.mot.com> Tim Berger wrote: : Hello, : Booting the MCPN750 board with the Tornado II boot loader fails after : programming the flash memory; there's no response from the board after we : reboot. However, the Tornado I boot loader starts up successfully. Then, : when we try to load the VxWorks image, the board freezes after the "Starting : at 0x100000" line. We've tried using version 1 and 2 VxWorks images with : the same result. Here's a log of the serial port output just before the : board freezes: Are you building both the T1 and the T2 bootrom yourself? If so, what are the differences in the sources? Are there any differences in the memory layout (you can check these with the xxx-obj- tools)? I don't have any experience with this board, so these are just some suggenstions. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: fwrite and mempartAlloc error.... Date: 30 Jan 2000 20:58:26 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8728li$s3h$2@news1.xs4all.nl> References: <38945AEC.BC466CF5@erols.com> Christian Williamson wrote: : I'm getting a mempartAlloc error while using the fopen-fread-fwrite : series of i/o function in VxWorks 5.2 and 5.3. I get this error when I : use nfsLib to a file on a Sun workstation, when I use netLib to a file : on a Sun workstation, and when I use dosFsLib to a RAM disk. The sizes : of the files I'm writing are many megabytes in size (I don't have : trouble with 'small' files, in the several hundred kilobyte to megabyte : range). There was a discussion some time ago about the use of netLib. It seems that netLib buffers the whole file in memory, so if you try to open a large file, the memory cannot be allocated. The suggested solution for this was to use nfsLib. Please take a look at my page (part 3.4), there you can find the original answer from M. Lang. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: templates Date: Fri, 28 Jan 2000 13:28:57 +0100 From: "Mario Semo" Organization: Telenor magnet GmbH Message-ID: <86t1pr$dh4$1@newsmaster01.magnet.at> References: <86o13r$8ak$1@newsmaster01.magnet.at> Hello, i am not able to do any automatic template resolution till now except the trival inline templates. Hello, i have to resolve templates, and inline template resolutions is absolutly not acceptable for me (but it works, obvious) due to code bloat. the documentation on your side is not very deep regarding template resolution. i have searched all files in my tornado installation for the word "template" and found it in the following files: "F:\Tornado\docs\gnu\toolkit\gcc.html" "F:\Tornado\docs\vxworks\guide\c-wfc2.html" "F:\Tornado\docs\vxworks\guide\GuideIX7.html" "F:\Tornado\host\x86-win32\info\gcc.info" "F:\Tornado\target\h\make\defs.bsp" "F:\Tornado\target\h\make\defs.default" "F:\Tornado\target\h\make\defs.project" "F:\Tornado\host\x86-win32\i386-pc-mingw32\lib\libadvapi32.a" "F:\Tornado\host\x86-win32\bin\windnav\bin\windnvdb.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-pc-mingw32\egcs-2.90.29\cc1.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-pc-mingw32\egcs-2.90.29\cc1obj.e xe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-pc-mingw32\egcs-2.90.29\cc1plus. exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-pc-mingw32\egcs-2.90.29\f771.exe " "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-wrs-vxworks\cygnus-2.7.2-960126\ cc1.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-wrs-vxworks\cygnus-2.7.2-960126\ cc1obj.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\i386-wrs-vxworks\cygnus-2.7.2-960126\ cc1plus.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\powerpc-wrs-vxworks\cygnus-2.7.2-9601 26\cc1.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\powerpc-wrs-vxworks\cygnus-2.7.2-9601 26\cc1obj.exe" "F:\Tornado\host\x86-win32\lib\gcc-lib\powerpc-wrs-vxworks\cygnus-2.7.2-9601 26\cc1plus.exe" i looked into the printed docu to. but cant find anything additional. first: the GNU users guide lists -frepo, and in the printed version it refers to page 153, but on page 153, there is nothing regading -frepo. much more, page 153 says "there is no automatic template resolution ..." ???. but -frepo should do automatic template resolution. `-frepo' Enable automatic template instantiation. This option also implies `-fno-implicit-templates'. *Note Template Instantiation::, for more information. fist i created a standalone sample, as in the docs (2nd docu in the above list, c-wfc2.html , chapter 5.2.7 Template Instantiation) shown but i was not able to builde a target which has the template functions resolved. thn i tried the same with the Tornado IDE. but i had the same problems there i added VPATH macro to get the headers, and added -I $(PRJ) and -frepo to the compile options. again, no go. i tried to download, and the 2 template functions are still unresolved. so, i am out of ideas how to use -frepo for template resolution. attached is the complete docu i found regarding templated (not so much :-<<) and the standalone sample and the wsp. envir: simpc (Tornado 2.0 + current Fixpack). NT 4.05 - -- regards, Mario Semo. http://www.kirchnersoft.com Mario Semo wrote in message <86o13r$8ak$1@newsmaster01.magnet.at>... >Hello, > begin 666 doc1.dat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` end begin 666 vxwWithTemplate.zip M4$L#!!0````(`-V:.RC"!:T!?P```*X````4`& `=GAW+U=I=&A496UP;&%T M92YW0J8&&08&,'2(@S_&>49&"$<, D`550-``<" MC9 XA'Z1. *-D#ASR4\NS4W-*U%(R\Q)5=!5&9)1DAJ;D%.8DDJ"D>OO" + M9@8`4$L#!!0````(`*AB/"@AC4/<50(``.$&```:`& `=GAW+U=I=&A496UP M;&%T92]G96Y86"YC;6131$L`7 `````(`%!4@;-C9&!I$&%@8%!A@ `'(&9D M`C-9%8"$`I#-R KABP*)B3H>0J8&&08&,'2(@S_&>49&"$< M, D`550-``>L>Y$X%X&1.!>!D3CM55B(*!V#2\ M!-A.):01QN-X5& 0C-4L^O&]#-@0IZWZ`6'!?9P[]W& RR/+3@*)8;U5I;S6.X<3Q4^V8VD@L0)QYS=BLP-:Y1T2A!T` MLO?/1)_<72D$Y T2A89"KHF&[@Q'(_;SU\0Q%O/$2<(X\?THV;[M1/VC2=;H M!OS@:*,_.@/'Z+W=?39+UA<&5,79T(-AT"='7T4:(B:E2]O3P^^4:H%NASI" M9)?F.0C3-JAI;2W'#\!PC,%J0=]RJ;&)8KK4(PN1HF*EY.<"D6,IR($=64TJ M42&RH71G>UX,&LC%?/!0VOD41JEK>ZO=8JZ\]%WH6.O@BS'>EO<-(*-%MENJ]L(0VY3 M387O9%J_,GEW4CQM7W9^^!Q!!Y"^9I7HJO"BRGG&)9$,M%2R!O57>VAEA6W+ M_EQQMOSVQX>@"/^D\+\I_*3O2I\_GTZFDRQK@$,HGR%\(08CW'[YN"/1,57G M".%^,^%95@$Y`J*Z)08[2BW5JXT><+\K\:C",& -S.=?R(L2C]=8I)+V.Q-. MED5_\G7PHE'X+_13OLDL'X^=G*!*TM19'FC?"AO+FJPS+%E)'=P'@Y!EV5:G&Y*2IF,_?G(<5J_D MPM;7@86^&^?[='2LS=9K=$%L6MY9Q&,="4?A6>>52L9F.W'N`H)-7V:NR\WV MJ2QD['*AG6)1E,X_WECM`NF.I>^89_IX'H($8!K9IFZ'*I0TW^^?$P`_,862 M52#F1"[('P("SH/4CB.\HK3(%HL\66F_NGR)4#K$=CE*Y]GBMKBZW$7I'Z5C MBU((QG'MA(ERK1%F)DB$C0VQEB8U;O=.&@KAQEA4;S+\O99)IB=#F;+?+EI!\Q(VE]OBYSZPVK;1F9;T4LNU1V M<[=Z?Y_-=U4'>$D=O7 06(0/$^^8]!/"G?O/VU_S]FG@[=1?\ZUZ#$2%",&M M,,*S=)CV7"TA9'09+X3X=&[U#C_LMH]G0!#"%CGFHV0JTZP5Q$)@]!ZT?95" M9^%^HJ>XC5RA$9U=&JT,GI?=&H(H52CUVO".]&68!8V&;A_V[\WH4.EK:JL4 M\P%+@V7=*Z*1PH=_5!\$I_3W-O7UZCVNO[>IK\<^KK_?+S7>6_:\,#!NK.<" MA\BJ@+U0EK^^[ &J!QVW3**SZR14"*;6]-XO4$L#!!0````(`/AD/"B<*(D* M5P```&P````5`& `=GAW+U=I=&A496UP;&%T92]L;VAE= MJD'JF!AD&!C!TB(,_QGE&1@A'# )`%54#0`'!("1. 2 D3@$@)$X!:[66%86UO"KMT M@3CIK^_PX>]8C92>JB)9+#-OF9DW\Q;W\#3ZP58\97C-!%-1R1*\?,7EAN%< MR2<6ESB+1+1F"O400KT>CJ5BF(N55%E4X2XU9Y?MMHQL &%W[S#:YY?]#QOS+/Q8[ WYQ%7^D@V]_%( MHO'<=DP:>);1`0)[ZH9K4:'P/CP-VQ\L;->D8SVPAEH9J34K$6K+6E8\34@L MQ8JO*]54UE2)$%^QGWC0'QSB##_O0@R1X M0(]7K7MGS7(F2EYEF$2BX)@(690),(?)NGM(^1*3E9"D)JGDHGT`"IDBN='F-&G;[)BFY9VV&6QGO8="25$MB]>B9-G=EHM$;@N 44_W0UMW M:(L'F$+N]"*6R+J$UFCFA?;,I: =V[7H5#?\6;W-W#ELUW#FIM6('>PV\F9! M2-N1;X4&^=6-<>QQ<)QS8'^WSL(6_!=KB_SJZ>'#J?/0WIU(>U@^,Z4XZ+26 M)H86<<%K`198BO05+UDJMW"@\@*G7##TCBW1$@!'6ZX?!*I*V=%)\*&SH#M7 M;]&G_L PAO^2;M\X]_^L7$SBHWYK+3T@,=1]>/X3=9VH<4-4)Y%Z*!,&921, MQ)P5:#]J;_*+/5B--K"H3:,8[0E_,\HEO(Y7%4RUPH(_-(U*$/H-4$L#!!0` M```(``57.RC4>1,A9 ```(@````7`& `=GAW+U=I=&A496UP;&%T92]P86ER M+F-31$L`7 `````(`%!4@;-C9&!I$&%@8%!A@ `'(&9D`C-9%8"$`I#-R KA MBP*)B3H>0J8&&08&,'2(@S_&>49&"$<, D`550-``(A"L&EN0H:("Y01U%^26IR26J*%4P2J+H2*%4+EE;.S$O.*4T!6ED`-$TO MV8Z7"P!02P,$% ````@`#U<[*&N,TDU8````:0```!H`8 !V>'+E MRLPK47!,25'0`#$J=!1 5*4F+UT40LKS%32 *BHUK2%R M1:DEI45Y('&]X-)"TO%P!02P,$% ````@`&U<[*!^GW>A4````70`` M`!H`8 !V>'+ERLPK47#)+TW*2570`+$K-'FYJGFY0&H4;( "=@HA MY?D*&A4Z0!EK7JZBU)+2HCR0F%YP::Z"!DBLEI<+`%!+`P04````" "R9#PH M$5J^&!$````4````' !@`'9X=R]7:71H5&5M<&QA=&4O<')J3V)J0J8&&08&,'2(@S_&>49&"$<, D`550-``> ?Y$X@'^1 M.']_D3@K2,PLAE= MJD'JF!@,&"#*11C^,\HS,#*".1"2&U,,`%54#0`'@7^1.(%_D3C(?I$X4$L# M!!0````(`+)D/"@Q8,8=$ 0``%0*```A`& `=GAW+U=I=&A496UP;&%T92]3 M24U.5&=N=2]P86ER02YO4T1+`%P`````" !05(&S8V1@:1!A8&!088 `!R!F M9 (S616 A *0S<@*X8L"B8DZ'D'*T=LE>AE=JD'JF!AD&!C!TB(,_QGE&1@A M'# )`%54#0`'?W^1.']_D3A_?Y$XA59?:]M6%+]R["Z,P#JRAU(*2VD9ZW 4 MZ5JV%6DM#6L"@Z68QDMA&.YD27'$%%M8L>-NC# M/L >]K#!((/!GOJRI>?^D7P\Q]T%23^=\SM_[CE7NO<#K4+X>+9(R.L"E34] M"\<9F0P+KE_@^G)1OI?@6B'D8SWP,H^\?-PFY(G>3=/_H9'/."_-O"Z278% M%S5"OE-QRS+X/<%+LV'.\V *U87_. 3>AR>_/_IS\?'F4NOQYF+KB\WGK3^X MYM$/2PB73[X_^?FT?%HZ)3\])403QE7A\PU @7)'$6[)?,5X@/#7"#]%^#>$ M7]4F>!?A;Q'^$>&_$;Y4FN";"-]'>(CP5P@_0_@OA)<7D$^$=Q#N(?PYPO\` M?JCPY?($B_F<@^OEB>W5"L=R[5GPN"Z[2^X`_@8P;_FVPDO*YH[BY/C"N;B" M\ 5RN<"O$ .>NRKN\[.S,RY_`E?._Y7PSLMQG\Q'D^*[3:6]NMSJ^;QBF M81AZ2OI.YW!\U+D?9?OM\"")O2SL[+R_?;?=ZX\ZH%T#[1K6KB5>--S0_20A M43]SLK>-JGGCYE \W%5J6DW+KC4LVRU@TR7^OC<45"JI](9KN"8%33SH]U9R M1S6IK-["O MF;#-B=-I5^8<7^G^8)@5YK8TMZ$<-=ILV"Z_%ZR9:.N2OLXGV:C7:W5@2D91 M?M-0&1G@TZ2V[$'A:<++6\I3=VD=7.W%`T^UFQ;3LD!+@L&H&X=252M4-E>) MLF"]5>A-R@G^`%95."[F8$*G4GL8>K$C:%58#=2-#KR>>J]1+I@8HK3R%C94 M7CD%Q\\[TE0)YIR91%7MX:DR[0X&L51!G6^G-M2OX9+#013(A0V5E0_"V!$O M))-I455*:J*^,':8=6.69, Z3F1(2OG4&0O"./,:>[U/&OV,H7J## MI\5@ZRH$S$_B4<@Y:'X3!XD 2\@)02P,$% ````@`L60\*'GB*P[S M````I0$``",`8 !V>'V;!V5X5=R+2=,+-B(\(8'%)-;2 F(8W; M?OY,5K?NXR;G.<]Y0\(I([E&K^]OJ.'V1-B@!;4,:R+ M$I->CF&01G%2`S0%QB1V/-%J&5]9@-E?$@!E@;?-:GF9P*\K/WD.?4\&S:3E MX^"8RC-W5:JS[;AL'?=?0O"#XZ-4R6'DPG(YMQT[,I-HI9UHJ!"N9E4#9?U( M"B /59YFWD&5YYA4Z>?GBX4_-^L7HHRDG4*6FIY9=/+A^UU]YW?C4__LUK]O MF%8.VC@,GJ+]. #8V]VT])MG[LQ5RR?Q`5!+`P04````" "R9#PHXNM!JP<$ M``!""@``(0!@`'9X=R]7:71H5&5M<&QA=&4O4TE-3E1G;G4O<&%I9(MZ0/04$!(2? #$I^!840X@<> #<. `$E*0D#CU`N'-']MO MNUD8:>R?W_N]/_/>C.VWM0;AXYME0IX6J*[I>3C-234LF#_#O+/?[O^Q_.G6=7_GFOO?G4>X?OSM\4\G]9/:"?GA(2&: M,'Y9^'P&4*#<482[,E\Q/D3X2X0?(OPKPD]J%7X/X:\1_A[AOQ!^ME;AJPC? M1GB,\!<(_XCPGPA?6$(^$=Y!>(CP)PC_#?ACA2_6*RS6VR["[9!OP58-5RGIUS^`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`N:G\23CD["JAYR%=@9AL"V0A/P+ M4$L#!!0````(`+)D/"CZ"D9A\P```*4!```C`& `=GAW+U=I=&A496UP;&%T M92]324U.5&=N=2]P86ER0BYR<&]31$L`7 `````(`%!4@;-C9&!I$&%@8%!A M@ `'(&9D`C-9%8"$`I#-R KABP*)B3H>0J8&&08&,'2(@S_ M&>49&"$<, D`550-``=_?Y$X?W^1.']_D3AMD%UK@S 4AN\%_X-W7MFP=E>% M7;CIML!TQ4:$,3BDFMI 3$(:M_W\F:RNW<=-SG.>\X:$4T1RC=X^WE'#[8&P M00MJ&=*4F]M%JW489!%"\F]BBXN2]'(,@S2*DQJ@P65)8L<3K9;QA068_2D! M4.#RH5DM3Q/X=>4GSZ'S9-!,6CX.CJD\E.MJ.R]9Q_RT$WSG>2Y7L1BXL MEW/;L3TSB5;:B88*X6I6-5#43P0#>:SR-/,.JCPO295^?1XO_'F_?B7*2-HI M9*GIF44''[[;U#=^-S[USV[]^X9IY:"-P^ YVHX#@+W>3$N_>N'.7+1\$I]0 M2P,$% ````@`LV0\*)](/N!4!0``[A@``"H`8 !V>'0J8&&08 M&,'2(@S_&>49&"$<, D`550-``>!?Y$X@7^1.(%_D3CM6$]H'%48_W:S&T.- M(B:*A(*5EF)KLMF9G?V3&6L332)"TX8FMI NC).=S>[B9G?9G20;41#;@S$* M@AX\%L\YU(,G/936@X*"1T$/'A2$% 1%Z$&-WYOW9[[M9-%J\*#[8#*_^=[O M^][W;]YFWIE('-@8>0@@Z:-8).$5VQX$HXS7-WA=&>;/,;R.`+R0KC-?RG&7"FT37SJ"]';'NO>Q/%,[[O);79(_7 MD//+?0!O]M]A$'G/;WU_^=; ]LS@_/;,P/Q;,[?G?V SES\9)#BV=6/KR]W8 M;G07/KL)('4&[D)'#A[Z*#CH$W/9%7*=X'D>JS]>(OA=@F\2_!W!AR(!OD#P M^P1_2O!/!#\2#? I@B\2W"3X'8(_)_A'@H?ZB$V"%P@N$?P&P;\A?DW@D5B M_7CVP>E8H/M8G&'>MP;>COD=B;V'^"IBUH=S`@\*G6G!D;A_7QPGN!]&%+[' M?S^,)6"*]84#0&PSW>H./@^X- MEMM9TAML'$0_G/N;_; "LA\>!K,OZ(Z3V>'-5.G&KZ-VM,UXRLD4MEC)RE8-:"0MEI^E2=4_435M+2=)RI MUFNE(])0BL^FNAM:J[4JI5K152H&5S&8P3%-V NQTIR5[F!U+)U1,?B9"0\K MF=UO4%NA9;.!T4Y36A=;K7*]Z2GU'%?/83I2>C:3L]A?Q0JM-L'I$RS(3#J= M2B.3,U3ZM:3P*(DV-3W':Z L!3Q94N:ZI:?1U$JU[HARZRHL`V?!K:\M5XM\ M*J6F-8C?H5F75*8GGE,X$ M@2)Q2Y8P(_R2%+J^K$A6."@Y(4=%[O$N/%VNUZM\"O,\V]U;KMH-#UF;#;ZDKK/0;=LM5CW'].L^RJ*Q M;+M27R5$V5\D[N]YXD" M,,Z36)ZGS$6VC,&<;)OC&BG/IGJ4'C@MTW<.JWGT* -8F^-1B> M5;(%);L"_.RQ']Y6Z%9$HM>C$EU%]*"/=A0:B$GTGD+78OXY)J*/%=J)2Y18 MJ52+B.['ZW?\OB]%5IP7BQ :\APA`O+K?ZA#0D]3F:0OPLXCHATVZ$EJU.>$ MUZ&GJ'W=..0$->9S#D>!G'0HCC@]C?L<)QZR(V(_#'\>NQ&*/9#(V*7DWXA= M1MP7>:!K["QB'OM2U]A/_H78RQVQCW5(9.Q2LJ_/H=@'PIR[B'TT)F,?CG!V M./:/#LG8S;X[[=AVP:LW6\J?*$I<7Q)3$C9.$QTN.0=!E;GDDI@/),4.+9:; M4J&@VVS_P)2["5S,QM]&); +C>I:BUWA&;#Q!YW\RH,=_%"#+;^*F![Y1P#^ M`%!+`P0*``````!09#PH````````````````%P!@`'9X=R]7:71H5&5M<&QA M=&4OAE=JD'JF!AD&!C!TB(,_QGE&1@A'# )`%54 M#0`'QWZ1.,=^D3C'?I$X4$L#!!0````(`(5E/"@B_1["Q@(``!@*```A`& ` M=GAW+U=I=&A496UP;&%T92]7:71H5&5M<&QA=&4N=W!J4T1+`%P`````" !0 M5(&S8V1@:1!A8&!088 `!R!F9 (S616 A *0S<@*X8L"B8DZ'D'*T=LE>AE= MJD'JF!AD&!C!TB(,_QGE&1@A'# )`%54#0`'"H&1. J!D3@!C9 XM5;?;YLP M$'Z>_PH>^M"J(IG:MVJM!-AI71% QC3:-,DBQ$E<`4:8M-VF_N\S)*1=?C1D MVE[ =_=]=S[?<0;*9)'QO#*F(N6&:4#?\'QJ((@I`%]L=(N]&\..L L9B5S$ M1J*:4YX5:5SQGEQ4&H0\>+,3&Y?)7#SQCR!R_,B32NV!A'CHT5F^8(U8$\!6 M_(^)0\LA/K,(B$LELB+I"'?N\ ,")Z[>CRHK6Q3=J&L,:] MLV6%[AZQR!HASI5H%KE4U43D22/,WE2I&#?"-)?F>"'22N1K><*GO#0+632: M49RFS0*2$1M&+L6,WA%DP:62$80\2JQ5+KBW?)V<2!UU7];[D'S?.E?L9*\6R<:N.S[CPS*8JC\^V6 M51"T;6>8R-!;,%^Z-: +03KIWJ\NA"@(#WWP+7;5W*9:C-4/5?'L^EGD$_FL M.O)98!&*+9>M')6=>-X0Y-GJ+&:=&'Y L>\QB ;80TL=,.$Q5.PY;@11W9' MQ)V8@1]2]C9,NYUIW?4NMCM6(,3?$%#B)^]>X(? HG?O9NSH_JNQTZ#*9)\ID?E4S/JO!\)3!UQ=5K*/8C'Q )=7> B][G3= `Z_OSW155Q**T>GI L G7 MV]K=KO]]+>Z4O-UE#=3ZRD3('3=\[/ MC41FA?XAV?I<=Y'MOTW:_B=)VUM)V]V2MKLF'1#_'CF4U4Y"L.OHC%VN-]TL M%"\=F=5_>_7XS"M>&A.NDE(4E9"Y,>'L>Y$X4$L!`A8+% ````@`YV0\*(8W#/#?`0`` M-08``!0`$0```````0`@`+:!_@,``'9X=R]7:71H5&5M<&QA=&4O;&]G4T0$ M`%P```!55 4`!^%_D3A02P$"%@L4````" #X9#PHG"B)"E<```!L````%0`1 M```````!`" `MH%O!@``=GAW+U=I=&A496UP;&%T92]L;V''''I$X M&7J1.!EZD3B=3KL.PC 0VY'X!XL)BF@9F&!$B T6F*-K$TJD-$&YZ]"_YU7Q MJ(0$W.3SV6=G"0H)D4%>0]]1C-0P)A-(0&Y0L]'(&\3:BZT,N&$Q%9*LW\L2 MX*HJR#FPD-@"1? LL7[Y:I[[8Q[NOZ9U;[:[U1Q[-B"L-WLLL^5X?"E0G:PS M\5;LCB%'RSA<4-JZ^SVEB"NE,!RDFH0&H\4;5[J0DX-2M^+=*SE;>LPZ="N> M=^AKX]0%7V+Z,47_DJ*_23D#4$L#!!0````(`-%A/"AHDM+VF ```)$!```& M`!$`0U1$5"YO550-``<:>I$X&GJ1.!IZD3CS861F`($I0,P#9K$RZI6D5I0P MX <*# P)>BF))8E(8AQ W(.FSH&!X8!>4G$Q`>,8&H#J8&R]M,R<5"C[WW\& MAG3&M,3L5$P]+%":$8C9P"QA%!%D?X!$F!DQS4#V`Q-8#0>F&B3W,^,P)SX^ MN22_J!AN#A-0) 4LP@(74062Z;X```#:`0``"P`1`$-41%1P86ER02YC550-``>P M>Y$XL'N1."!ZD3B=3KL.PC 0VY'X!XL)BF@9F&!$B T6F*-K$TJD-$&YZ]"_ MYU7QJ(0$W.3SV6=G"0H)D4%>0]]1C-0P)A-(0&Y0L]'(&\3:BZT,N&$Q%9*L MW\L2X*HJR#FPD-@"1? LL7[Y:I[[8Q[NOZ9U;[:[U1Q[-B"L-WLLL^5X?"E0 MG:PS\5;LCB%'RSA<4-JZ^SVEB"NE,!RDFH0&H\4;5[J0DX-2M^+=*SE;>LPZ M="N>=^AKX]0%7V+Z,47_DJ*_23D#4$L#!!0````(`*MB/"AHDM+VF ```)$! M```+`!$`0U1$5'!A:7)!+F]55 T`![%[D3BQ>Y$XL7N1./-A9&8`@2E S -F ML3+JE:16E##@!PH,# EZ*8DEB4AB'$#<@Z;.@8'A@%Y2<3$!XQ@:@.I@;+VT MS)Q4*/O??P:&=,:TQ.Q43#TL4)H1B-G +&$4$61_@$28&3'-0/8#$U@-!Z8: M)/5 MP@7BDQD`4$L#!!0````(`*AB/"@AC4/<50(``.$&```)`!$`9V5N6%@N8VUD M550-``>L>Y$XK'N1..YYD3CM55B(*!V#2\!-A. M):01QN-X5& 0C-4L^O&]#-@0IZWZ`6'!?9P[]W& RR/+3@*)8;U5I;S6.X<3Q4^V8VD@L0)QYS=BLP-:Y1T2A!T`LO?/ M1)_<72D$Y T2A89"KHF&[@Q'(_;SU\0Q%O/$2<(X\?THV;[M1/VC2=;H!OS@ M:*,_.@/'Z+W=?39+UA<&5,79T(-AT"='7T4:(B:E2]O3P^^4:H%NASI"9)?F M.0C3-JAI;2W'#\!PC,%J0=]RJ;&)8KK4(PN1HF*EY.<"D6,IR($=64TJ42&R MH71G>UX,&LC%?/!0VOD41JEK>ZO=8JZ\]%WH6.O@BS'>EO<-(*-%MENJ]L(0VY3387O M9%J_,GEW4CQM7W9^^!Q!!Y"^9I7HJO"BRGG&)9$,M%2R!O57>VAEA6W+_EQQ MMOSVQX>@"/^D\+\I_*3O2I\_GTZFDRQK@$,HGR%\(08CW'[YN"/1,57G".%^ M,^%95@$Y`J*Z)08[2BW5JXT><+\K\:C",& -S.=?R(L2C]=8I)+V.Q-.ED5_ M\G7PHE'X+_13OLDL'X^=G*!*TM19O>Y$XKWN1.*][D3CS861F M`($C0,P+9K$PZI6D5I0P(( *$/<`\08HGPF(%1@8$O12$DL2&? #!P:&`WI) MQ<4$E#$T`-6%=CYM?L,1_+^4YW\I1V_LC^ 7()GF(SP01G?LEY.'A8 ,;B 6 M89 &DCQ@EEY:9DXJU)Q__QD8TAG3$K-3,>U@@=*,0,P&9@FCB"#[&R3"S*@" M]2T"(/N9":P&TQYD_S+C4 ,"NDAV06PQ1Y*%B+BAB(0!R?3D9*/XY/S<`J"G M4_08XN/CT_-*X0+QR04YI<4@S!#OF)(2'^^6F0E24V(2D)A99!B5">(&E^8B MB3 ``%!+`P04````" "J8CPHO98GV `!``"M`0``# `1`$E-1W!A:7)!+G)P M;U54#0`'KWN1.*][D3@^>I$X79!-3X0P$(;O)/P';IS81O!DXH$(KLTNL.%C M,<9DPO*A34K;0%%_OI2"KEYFGO>=F>3-1):HR.#O:B%,([ 0.B!:>RZB2+:C M1!]?GS/T0DO3\"W;J>VYE!6EJO>B99),O>*.<:=INW9P!!?** !*',?YPC-Y M[K4+L/GK!D"$XWWIN>L$_IW\Y6WI=U*QD:C.^"@;PNH?IN2RY;M,A$K"E S2 M$J+BF&/(G]+0#Q8/TC",\]37D8*'4W&?X4@K/)?'N]><#ZQJ.)+5\-9*]*[O MX/Q<)NDAT[&ZH15<0;(DP[MUZ1ALP=441WO]?6Z;1F)E4P\@;T^S=?-"E',E MR6Q\`U!+`P04````" `:8CPH&1^/-+4```#R````!P`1`&QO9RYT>'155 T` M!Z-ZD3BC>I$XHWJ1.%V-2VL",12%]X'\A[.T4)V,U@?!"EVW"T&DZV1RE4@> M8I+I"/WQC5 $W5R^>QZ<\0;.8(VSLI>/"0W$V5=4QH8C/F7CNMFT<4VFE)M^ M^*G@S__OO8!?SO;!T,$&,DA7KZ-+DK-M#:QMR!LI=\6/^FC-RX-ZPU'%5]13 MK6]U"757(NH3=1D^FN((7ET18H8FE*1T531UJB1"/ #E>7G"6:]<(;RC7;;+ MJ7A;511#*Q9B,1><_0%02P,$"@``````T6$\*$&/\Q0(````" ````,`$0!, M4U155 T`!QEZD3@9>I$X&7J1.$E-1RYO( T*4$L#! H``````*IB/"@4Y$X('J1.$E-1W!A:7)! M+F\@#0I02P,$% ````@`!5<[*-1Y$R%D````B ````8`$0!P86ER+F-55 T` M!SH6D#B[>9$XNWF1."M)S2W(22Q)5;!)SDDL+E8(L>/E"DC,++()L;.R`C$4 M-$(4XBMT%(!DI::"E4*%@D9\A::.0J4&D,_+59$X*TG-+9$X4\[,2\XI34E5L E(S"S2RR@HL./EXN7*S"M1<$Q) M4= `,2IT%$!4I28O5S4OEP(0@-0JV `%[11"RO,5-( J*C6M(7)%J26E17D@ M<;W@TEP%#9!X+2\7`%!+`P04````" "K8CPHO98#(",!``#V`@``"0`1`'!A M:7)!+F5X9554#0`'L7N1.+%[D3BQ>Y$X\V%D9@"!,T L"6:Q,.J5I%:4,"" M"A#W`/$.*)\)B!48&!+T4A)+$I'4< #Q!@94X,# <$ OJ;@8QM=AP X:@.I" M.Y\VO^$(_E_*\[^4HS?V1_ +D$SS$1X(HSOVR\G#,/5"0"P!Q"(,TD!2#,S2 M2\O,206RN(#XWW\&AG3&M,3L5$R[6* T(Q"S@5G"*"+(_@>),#.J0'V-`,A^ M9P*KP;0'V=_,N-1 W2Q"A)M54-RLBR("D"CX]/S M2N$"\7Y"AH5.D 9:UZNHM22TJ(\D)A><&FN@@9( MK):7"P!02P,$% ````@`JF(\*-=W9A+L````! (```H`$0!04E1P86ER02YO M550-``>P>Y$XL'N1.+![D3CS861F`($C0,P+9K$PZI6D5I0P(( *$/<`\08H MGPF(%1@8$O12$DL2D=1@`PX,# ?TDHJ+&0BH:P"J"^U\VOR&(_A_*<__4H[> MV!_!+T RS4=X((SNV"\G#PL!&3Q +,(@#22YP"R]M,R<5"COWW\&AG3&M,3L M5$P[6* T(Q"S@5G"*"+(_@:),#.J0'V+`,A^9@*KP;0'V;_,.-2 @"X2&V*+ M+9+M$!%W%#5A0#(].=DH/CD_MP#HZ10]AOCX^/2\4KA ?')!3FDQ"#/$!Y?F MQL>7F 0D9A891F4RQ#NFI,3'NV5F@O3 A8%<`%!+`0(6"Q0````(`-%A/"CS MYJ)YO@```-H!```&``D```````$`( "V@0````!#5$14+F-55 4`!QEZD3A0 M2P$"%@L4````" #183PH:)+2]I@```"1`0``!@`)`````````" `MH'S```` M0U1$5"YO550%``<:>I$X4$L!`A8+% ````@`JF(\*//FHGF^````V@$```L` M"0```````0`@`+:!P $``$-41%1P86ER02YC550%``>P>Y$X4$L!`A8+% `` M``@`JV(\*&B2TO:8````D0$```L`"0`````````@`+:!N (``$-41%1P86ER M02YO550%``>Q>Y$X4$L!`A8+% ````@`J&(\*"&-0]Q5`@``X08```D`"0`` M`````0`@`/^!B@,``&=E;EA8+F-M9%54!0`'K'N1.%!+`0(6"Q0````(`*IB M/"@Z%5B#Z0````0"```*``D`````````( "V@1<&``!)34=P86ER02YO550% M``>O>Y$X4$L!`A8+% ````@`JF(\*+V6)]@``0``K0$```P`"0```````0`@ M`+:!.0<``$E-1W!A:7)!+G)P;U54!0`'KWN1.%!+`0(6"Q0````(`!IB/"@9 M'X\TM0```/(````'``D```````$`( "V@70(``!L;VC>I$X M4$L!`A8+"@``````T6$\*$&/\Q0(````" ````,`"0```````0`@`+:!7PD` M`$Q35%54!0`'&7J1.%!+`0(6"PH``````*IB/"@41,A9 ```(@````&``D```````$`( "V@=T)``!P86ER+F-55 4`!SH6 MD#A02P$"%@L4````" !85SLHN4V,`VH```"&````" `)```````!`" `MH%V M"@``<&%I Organization: Subscriber of Pacific Internet Message-ID: <8730h4$d7g$1@newton3.pacific.net.sg> HI all I would like to know what function calls are used to hook a PCi interrupt to the interrupt vector table in MBX 860 processor. Apparently pciIniConnect and pciIntDisconnect are not available in MBX860 BSP . So is there any other equivalent routine. Thanks in advance regards s.shiva - -- =========================== S.Shivalingam Senior Analyst Programmer Maverick Research Inc Singapore Email: shiva@optasia-group.com PH : 65-7764798 ext: 105 FAX: 7760157 --------------------------- Newsgroups: comp.os.vxworks Subject: SNiFF+ in Tornado II Date: Tue, 18 Jan 2000 21:00:31 +0100 From: "Martin Raabe" Organization: T-Online Message-ID: <862gpf$i2l$1@news07.btx.dtag.de> Hello folks, I look for people who did sucessfully integrate the Source Code navigator and manager SNiFF+ in Tornado II. Do you have any hint what the best kind of integration would be? Ciao __ Martin Raabe Project Manager Integration of SNiFF+ in the embedded area TakeFive Software Im Leuschnerpark 4 D-64347 Griesheim T: (+49) 6155 / 878912 F: (+49) 6155 / 605-257 M: (+49) 170 / 1838405 E: MRaabe@TakeFive.de T A K E F I V E S O F T W A R E "the SOURCE CODE engineering company" --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CPU util measurement? Date: Mon, 31 Jan 2000 04:11:46 GMT From: bill.beckwith@realtime-corba.com (Bill Beckwith) Organization: Objective Interface Message-ID: <8ECBDE36Abeckwbnewscom@sigma.ois.com> References: <3874C795.732E9B0C@email.mot.com> <38782C28.E11BD702@bitband.com> <3878ac4d.367860@news.swbell.net> <3879775B.D33052ED@bitband.com> <85ibpt$8ao$1@nnrp1.deja.com> Ted Dennison wrote in <85ibpt$8ao$1@nnrp1.deja.com>: > >Often times what you really need to do is show that CPU utilization >never drops below a certian percentage, without having to go through the >rigor of a formal proof. What we do is write a cpu-eater task that gets >called at the highest rate and does not relinquish the CPU until after a >set percentage (say half) of the frequency has passed. If we can process >normally with no reported frame-overruns from the other tasks, then we >have shown that we have at least %50 CPU time free. "called at the highest rate" would have to be the highest priority hardware interrupt since a task at the highest software priority couldn't guarantee that it was actually getting half of the CPU time (since all hardware interrupts would take precidence). - -- Bill --------------------------- Newsgroups: comp.os.vxworks Subject: Re: CPU util measurement? Date: Mon, 31 Jan 2000 04:18:23 GMT From: bill.beckwith@realtime-corba.com (Bill Beckwith) Organization: Objective Interface Message-ID: <8ECBD213Cbeckwbnewscom@sigma.ois.com> References: <3874C795.732E9B0C@email.mot.com> <38782C28.E11BD702@bitband.com> <3878ac4d.367860@news.swbell.net> <3879775B.D33052ED@bitband.com> <85ibpt$8ao$1@nnrp1.deja.com> <8ECBDE36Abeckwbnewscom@sigma.ois.com> Bill Beckwith wrote in <8ECBDE36Abeckwbnewscom@sigma.ois.com>: >Ted Dennison wrote in ><85ibpt$8ao$1@nnrp1.deja.com>: >> >>Often times what you really need to do is show that CPU utilization >>never drops below a certian percentage, without having to go through the >>rigor of a formal proof. What we do is write a cpu-eater task that gets >>called at the highest rate and does not relinquish the CPU until after a >>set percentage (say half) of the frequency has passed. If we can process >>normally with no reported frame-overruns from the other tasks, then we >>have shown that we have at least %50 CPU time free. > >"called at the highest rate" would have to be the highest >priority hardware interrupt since a task at the highest >software priority couldn't guarantee that it was actually >getting half of the CPU time (since all hardware interrupts >would take precidence). I spoke too soon. The task at the highest software priority could turn off interrupts while it is eating CPU. :-) - -- Bill --------------------------- Newsgroups: comp.os.vxworks Subject: Booting from MVME147? Date: 31 Jan 2000 00:11:44 -0800 From: Andreas Kabel Message-ID: Reply-To: Andreas.Kabel@slac.stanford.edu Hi, I recently got 3 Motorola MVME147 boards cheaply. They have "VxWorks" boot EPROMS on them. Obvioulsy, they can boot an image off a remote FTP server. Can anyone point me in the right direction for information about format, parameters, ... of that boot image? I'd like to be able to boot netBSD on them. Thanks in advance for any help, Andreas - -- Andreas Kabel | akabel@slac.stanford.edu Stanford Linear Accelerator Center | +1(650)926-5069 (office) 2575 Sand Hill Road, MS 26 | +1(650)926-5368 (fax) Menlo Park, CA 94025 | +1(650)917-8559 (home) --------------------------- Newsgroups: comp.os.vxworks Subject: SNMP : Entries in Interfaces MIB Date: Tue, 18 Jan 2000 21:23:09 GMT From: "Dennis Wang" Organization: Bell Solutions Message-ID: <1R4h4.529$Vz6.601@198.235.216.4> Greetings, Is there a way to add some entries in Interfaces MIB? Right now , I got two entries , nt0 and lo0. Is there a way to delete that two entries? Thanks Dongsheng Wang --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Losing PowerPC in PCI write Date: Tue, 18 Jan 2000 16:06:29 -0500 From: "Mario Marquis" Organization: Motorola Message-ID: <862kfi$ruf$1@schbbs.mot.com> References: <38834338.F3B9B80E@trw.com> Check this link http://adwww.fnal.gov/controls/micro_p/nnfe/nnfe5.html Mike Johnston wrote in message <38834338.F3B9B80E@trw.com>... >I'm trying to interface with a new PCI device. I can read from the >device with no problem but when I post a PCI write, though the data >makes it to the device, the powerPC (mvme2307 with vxWorks 5.4) hangs in >a heavy CPU state (at least by the LED). > >when probed, we see that the #IRDY never is deasserted, though #TRDY is >asserted and deasserted. >#IRDY seems to behave properly in a read operation. (asserted first and >then deasserted concurrent with #TRDY being deasserted) > >Has anyone seen this? or if this is not apropos for a vxWorks group, >could you direct me in the right direction. > --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado II<->SourceSafe Date: Mon, 31 Jan 2000 10:45:15 +0100 From: BLP Organization: Technische Universitaet Graz, Austria Message-ID: <389559AB.1D3F2168@hotmail.com> I'm neither a SourceSafe nor a Tornado II expert, so maybe you can give me a hint, how to connect Tornado II with sourcesafe. (Windows NT4.0, Tornado II, SourceSafe 6.0) Under "Tools|Options|Version Control|CheckOut:" I start a batchfile with the following line: l:\temp\vsstst\ss.bat Checkout $filename -C$comment -I-N The content of the batchfile SS.BAT: @echo off rem echo para:%1 %2 %3 %4 %5 %6 %7 %8 %9 ' path to ss.exe path=%path%;d:\msdev\common\vss\win32 ' path to the sourcesafe database set SSDIR=\\pmed101\VSS\VSSTest ' select project ??? (depends on file!) ss.exe Cp $/VSSTestMain ' execute command (checkout/checkin) ss.exe %1 %2 %3 %4 %5 %6 %7 %8 %9 - --- SS assumes, that the current project is the container for the wanted file ($filename). Since our SW-project consists of several subprojects located in different subfolders (worst case: sometimes different folder and project names) I have the problem to select the "current project" ss.exe Cp $/VSSTestMain ss.exe Cp $/Subprj1, ... How to select the SourceSave project? Can I get the path of the selected file in Tornado? ($filepath doesn't work) Many Thanks for any hints. Bernhard --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cannot boot VxWorks on MCPN750 Date: Mon, 31 Jan 2000 14:47:57 +0300 From: "Arthur V. Grigorian" Organization: Auriga, Inc. Message-ID: <3895766D.5DD5D87B@auriga.ru> References: <86shh2$54q3@nntp.cig.mot.com> <3891E9F0.5C2101C4@auriga.ru> <86t2na$54q5@nntp.cig.mot.com> Reply-To: archy@auriga.ru Hello, > > - Please make sure that you build your vxWorks images with optimization > > level = 2 ( see properties for your project build(s) ) > > This problem can also be a result of your changing some header files > > in the BSP. > > Changed it to level 2... the image still won't start. Okay, keep building MCPN750 kernels with -O2 as it documented; let's check some other things. > > - As for MCPN750 bootrom.bin in T2 -- > > have you tried to flash the bootrom.bin file which came with the BSP > > and what were the results? > > As far as I know, there was no pre-built bootrom.bin file in the MCPN750 BSP > we got. We built a bootrom_uncmp, which we then ELFToBin'd to a bootrom.bin > file. After flashing and rebooting the board, the result was nothing -- > blank screen (on the serial-port terminal) and no response. We tried > building the compressed bootrom also...same result. > > > - If you use transition module TMCPN750 with IDE HD then such problem > > can be caused by your chassis (I saw it on CPX8216) Contact me if you > > are facing the same problem. > > We're using a TMCPN710... The chassis is a CPX1204. > > I've looked through all the T2 release notes, BSP how-to guides, etc., and > made sure that I followed all the directions correctly. The question was about IDE HD connected to TMCPN710. Try to extract TMCPN710 before MCPN750 boot. > Someone here suggested that it may be a problem with the Raven chipset; the > newest MCPN750's have new chipsets that may not work correctly with the boot > loader (some addressing issue). Yes, there was such bug in this BSP about 6 months ago; It has been fixed and we received fresh MCPN750 BSP from WRS. Try to ask WRS support about the BSP updates for MCPx750 if you aren't sure that your BSP is fresh. Regards, __ R2 --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jan 31 06:18:38 2000 From: David Anderson Date: Mon Jan 31 06:18:41 PST 2000 Subject: memPartAlloc and vxWorks I/O, sockets Fellow vxWorkers, I have also seen the memPartAlloc problem with sockets. In my case what was happening was NOT due to the I/O system.. I actually traced it to a stack/pointer corruption. void mysocketfunction(int socket) { char *mytemppt; char hostname[20]; int anothersocket; int something; ..... send( anothersocket , mytemppt , strlen(mytemppt) , 0 ); close( anothersocket); ..... } I had the task calling mysocketfunction( ... ) unexpectedly suspend on the close(...). Shell would indicate that the task was SUSPENDED. On investigation I accidently discovered ( and not an unusual error in new code ) that hostname[20] was too small for one of the strings in use. This had the effect of overwriting 'anothersocket'. Once the socket number was destroyed the i/o hung up as described. I did not get an error message ( task errno not set still at zero ). I've also seen similar problems when the task stack has been too small for a ( re-entrant ) function call, although the effect was similarily subtle. It would appear that the application software is expected to be coded correctly !. I hope that helps resolve someones problem. If WRS admit to a problem in the BSD stack then let us know !!. David Anderson. mailto:danderson@dtrack.demon.co.uk 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 Mon Jan 31 07:55:00 2000 From: "Kevin Dodd [Healdsburg]" Date: Mon Jan 31 07:55:02 PST 2000 Subject: Due to a series of requirements, I need to run tornado on Win95/98 and not on NT. Does anyone have experience with Tornado II on Win95/98. Any help would be appreciated. Kevin E. Dodd SW Engineer Wescam-Healdsburg Fax: 707-433-7110 Email: kdodd@versatron.wescam.com Direct Phone:707-473-1114 Those who say a thing cannot be done..... Should not hamper those who are doing it From vxwexplo-errs@csg.lbl.gov Mon Jan 31 09:27:28 2000 From: David Laight Date: Mon Jan 31 09:27:31 PST 2000 Subject: re: fwrite and mempartAlloc error Christian: A suggestion which should bypass the error and runs a lot, lot faster. If you are sequentially writing a file, use rcmd directly, avoiding netLib. fd = rcmd( sysBootParams.hostName, 514, sysBootParams.usr, sysBootParams.usr, "cat >file", 0 ); ... write( fd, ... ); ... close( fd ); This has the additional advantage that the file is visible on the server as it is written. It should be possible to do the same for ftp, but the vxworks ftp interface is more complicated. Maybe WRS should change open to support O_APPEND (with and without O_TRUNC) in netLib and use code similar the the above - instead of locally buffering the entire file. But that might be useful:-) David ---------------------------------------------------------------- David Laight email: dsl@tadpole.co.uk Tadpole Technology plc phone: +44 1223 428 232 Cambridge, UK fax: +44 1223 428 201 From vxwexplo-errs@csg.lbl.gov Mon Jan 31 12:24:12 2000 From: "Blake, Don" Date: Mon Jan 31 12:24:15 PST 2000 Subject: Maximum Number of Open File Descriptors What is the maximum number of open file descriptors in VxWorks? Is it configurable? Don Donald R. Blake Sr. Programmer Lockheed Martin Federal Systems Owego, NY