From vxwexplo-errs@csg.lbl.gov Sat Jul 1 04:03:11 2000 From: Vxworks Exploder Date: Sat Jul 1 04:03:13 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 1 04:03:08 PDT 2000 Subject: Re: __errno() Subject: Re: motFccEnd.c Subject: Re: Build vxWorks.st_rom Subject: GUI on x86 target Subject: Re: Shell Command line editing Subject: VxWorks loading w/open sockets Subject: Re: restarting vxWorks in RAM without reloading Subject: Re: using serial port for console! Subject: Message queues Subject: Help using Compware IP1553 driver Subject: Re: Build C++ support in VxWorks image Subject: Re: Message queues Subject: Re: motFccEnd.c Subject: C / C++ - Anyone know a good C/C++ manual dealing with RealTime OS, Embedded systems Subject: system tick rate - how fast have you jacked it up? Subject: Re: tffs format failure Subject: VxWorks ledLib question. Subject: Re: Build vxWorks.st_rom Subject: Re: dosFS lessons learned Subject: C++ classes. Subject: Processing Overhead for Sockets = ?? Subject: MVME2700 execution speed Subject: BSP for MPC860T ! Subject: Re: system tick rate - how fast have you jacked it up? Subject: Re: how to switch ethernet devices? Subject: [Powerpc 860 BSp] how to delete bcsr at sysserial.c ? Subject: Re: generic interface setup ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: __errno() Date: Fri, 30 Jun 2000 13:49:13 +0200 From: "Werner Schiendl" Organization: VBS - Vienna Backbone Service Message-ID: <962365995.125844@newsmaster-04.atnet.at> References: <8jhu9q$52$1@nnrp1.deja.com> Hi, What for do you need the definition of the function? It is probably not part of the distribution unless you have licenced the source code. This is no reason why you should not be able to set a breakpoint on the function. You can set the breakpoint with either Debug | Breakpoints... menu or the command b in Tornado WindShell. HTH Werner wrote in message news:8jhu9q$52$1@nnrp1.deja.com... > In a VxWorks programmer's Guide (VxWorks 5.4, WindRiver Systems), > chapter 2.3.7, it is mentioned among others that: > '...because __errno() is a function, you can place breakpoints on it > while debugging, to determine where a particular error occurs...' > I tried to locate the body of __errno() but no luck. Only the > definition in errno.h. Does anybody know where to find it? > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: motFccEnd.c Date: Fri, 30 Jun 2000 14:29:24 +0100 From: "Koay, Jessica [HAL02:OW41:EXCH]" Organization: Nortel Networks Message-ID: <8ji6vj$es2$1@qnsgh006.europe.nortel.com> References: <8jghpi$8ih9@overload.lbl.gov> I'm working on the exact same problem (see e-mail sent 23/6). I have to support all three FCCs on our board. Do you have the source file motFccEnd.c? I was told that registers are initialized in there. Since I only have the object file, I can only assume that if I modify sysMotFccEnd.c and initialize configNet.c properly, it should work. (What are the chances?!) I have contacted WindRiver regarding this matter. The FAE hasn't got back to me regarding this matter. Will keep you posted. Hope you'll do the same. Regards, Jessica Ferrara, Bob wrote in message news:8jghpi$8ih9@overload.lbl.gov... > > VxWorks/Tornado - > > Does the Wind River END driver, motFccEnd.c, support more than one FCC > interface at a time? > > Upon a quick review, it appears it does not (for example, the RIPTR and > TIPTR > fields in FCC Parameter RAM would have the same address for each FCC > loaded). > > Has anyone successfully used more than one FCC interface at one time? > > Thanks, again. > Bob Ferrara > Unisphere Solutions, Inc. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Build vxWorks.st_rom Date: Fri, 30 Jun 2000 10:07:11 -0400 From: Charles H. Chapman Organization: Dynacs Engineering Company Message-ID: <8ji9ig$skb@niven.ksc.nasa.gov> References: <395BEBDA.6D075D45@hwacreate.com.cn> <395BC2C2.73E74FD9@netscape.net> Reply-To: chc@nasa2.ksc.nasa.gov Computer Engineer wrote: > vxWorks.st and vxWorks.st_rom image do not include WDB agent. > Thus you cannot connect the target server to the target and use the > host tool like WindSh. Not true! There's no relation to WDB and which version of vxWorks you build. We're currently running vxWorks.st_rom in our lab here and we use WindSh, Crosswind and all the other WDB-based tools all the time! Chuck --------------------------- Newsgroups: comp.os.vxworks Subject: GUI on x86 target Date: Fri, 30 Jun 2000 17:45:21 +0200 From: "Stephan Lachnicht" Organization: 'FH Gelsenkirchen, Abt. Bocholt' Message-ID: <8jibkl$j5v$1@merlin.rembrandtstr.bocholt.de> Hi all, in our new VxWorks project we need to implement a user interface on standard PC-VGA screen with touch screen support. Something like windows, buttons and graphic display of signals. T2 on NT /VxWorks 5.4 on AMD K6 I am looking for a graphics lib (free or to buy). Is there something else than Zinc. Thanks in advance for any help, Stephan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Shell Command line editing Date: Fri, 30 Jun 2000 08:44:49 -0700 From: Terry Duncan Message-ID: <395CC071.CBA811AA@cintacom.com> References: <394E9566.7105208D@cintacom.com> Yes, exactly what I was looking for. Thanks Dave Korn wrote: > Terry Duncan wrote in message <394E9566.7105208D@cintacom.com>... > >Does anyone know if I can access the command line editing functionality > >in the shell from my own function reading from stdin? In other words, I > >would like to be able to read a line of text from stdin and have history > >and VI like editing functions to modify the text before entering. > > > >There is a publicly avaiable utility called readline that does this but > >its pretty big. I was hoping there was an undocumented function similar > >to fgets() which supports this since the functionality does exist in the > >shell. > > ledLib is what you want I think.... see and the Tornado > documentation for ledLib. > > DaveK > -- > I wouldn't recommend sex, drugs or insanity for everyone, but they've > always worked for me. > -- Hunter S. Thompson --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks loading w/open sockets Date: Fri, 30 Jun 2000 15:52:51 GMT From: bonanza35@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8jifob$dgg$1@nnrp1.deja.com> We are working on a project here where we have two different options: 1. Connect to the VxWorks box via a socket (from a client machine via TCP/IP over ethernet), spawn a task, disconnect and close the socket and leave the task running; or 2. Connect to the VxWorks box via a socket, call the task (same task as above), and leave the socket connection open while the task runs. Without getting into any of the numerous reasons why one might be "better" than the other (it's all application dependent anyway), my question is: All else being equal, which of the above poses the greatest load on the VxWorks box? In other words, does the act of leaving a socket open impose any significant extra loading on the box? There will likely be several dozen tasks running at once, so if we have a socket open for each one, that's several dozen sockets open. So if the loading from one open socket is significant, this will go far in helping us to make our design decision. If I'm too unclear on the question, please let me know. Any e-mailed responses should go to epoole-at-scoot-dot-netis-dot-com (I'm sure all will be able to turn that into a real e-mail address). This is my first posting from the "my-deja.com" address and I'm not sure whether mail to bonanza35@my-deja.com will reach me. Thanks ... Eric Poole RKT Technologies, Inc. Londonderry, NH http://www.netis.com/rkt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: restarting vxWorks in RAM without reloading Date: Fri, 30 Jun 2000 09:06:31 -0700 From: hruesing Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <1c255ef9.2783aac9@usw-ex0101-007.remarq.com> References: <395A9AA8.8F847C19@west.raytheon.com> Although I am a VxWorks novice: we have similar requirements. We load an VxWorks image with an extern floppy loader in static RAM (using the target vxWorksRom_resident). After Switching power off and on again we can run the vxWorks image residing in the static RAM. The target vxWorksRom_resident copies on power on its data segment to (another) RAM-Position thus preserving the segment to its initial state Harud Ruesing hans_ruesing@arburg.com - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using serial port for console! Date: Fri, 30 Jun 2000 09:26:43 -0700 From: hruesing Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0e5ef1ab.2cc82a23@usw-ex0101-007.remarq.com> References: <8j5omu$uh8$1@nnrp2.deja.com> Hallo, We have got a similar experience. We found that the standalone PC and the connection with ethernet worked fine. When connecting with serial cable (null modem) we stated: Not all PC's were able to run 115 kBaud (why ?) In some cases we had to stop the target server and restart it to get connection after downloading the VxWorks image (using Target File System). In either case if you use a PII/PIII the start of vxWorks may be delayed up to a minute (due to mmu problems or something like this, also mentioned in this newsgroup) - don't reboot the target to early! Harud Rüsing hans_ruesing@arburg.com - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Message queues Date: Fri, 30 Jun 2000 14:06:13 -0400 From: Vividh Siddha Organization: Lucent Technologies Message-ID: <395CE195.455B0090@lucent.com> In message queues in VxWorks, there are two priorities NORMAL and PRIORITY. The priority messages are put in the front of the queue. What happens in the following scenario? Queue----> Priority msg 1 , normal msg 2. the reader of the queue is currently doing some processing before reading the queue. What happens when a sender posts one more priority message in the queue? If this messaeg is "Priority msg 2", where will it be placed? so when the task reads the queue again will it read the "Priority msg 1" or the "Priority msg 2" message? Thanks, Vividh --------------------------- Newsgroups: comp.os.vxworks Subject: Help using Compware IP1553 driver Date: Fri, 30 Jun 2000 13:51:41 -0400 From: "Bill Pritchett" Message-ID: <8jimn5$fs8$1@bob.news.rcn.net> Anybody got any examples writing a message to the 1553 bus using this driver? Thanks, Bill Pritchett --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Build C++ support in VxWorks image Date: Fri, 30 Jun 2000 18:08:36 GMT From: bdu Organization: Nokia Message-ID: <395CE1E2.763457F3@nokia.com> References: <395A6BD2.89D24690@ss8networks.com> I have the same problem one weeks ago. If you downloaded C++ code from tornado shell, you wouldn't get this problem. However, from telnnet session, you will get "_builtin_new" errors. This only happened in PowerPC which is Vxworks bug. You can download Vxworks patch spr28187 from Windsuf. Install it and everything will be all right. Yi Qin wrote: > I am trying to download a small C++ program to a target (PPC). > If I do it using the 'ld' from a shell, it complains about > > Undefined symbol: __builtin_new (binding 1 type 0) > Undefined symbol: __builtin_delete (binding 1 type 0) > > However, if I download via Tornado2 env, there is no program. > It appears that Tornado is able to download some C++ support > with the module, while 'ld' doesn't. > > Is there anything I can do to > > 1. build the vxworks image with the C++ support (which flag to use)? > 2. or build the C++ support or necessary symbols in the appl. module? > (how) > > Any insight will be greatly appreciated. > > Yi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Message queues Date: Fri, 30 Jun 2000 14:56:28 -0400 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <395CED5C.E6FFE95C@lucent.com> References: <395CE195.455B0090@lucent.com> Vividh Siddha wrote: > > In message queues in VxWorks, there are two priorities NORMAL and > PRIORITY. The priority messages are put in the front of the queue. What > happens in the following scenario? > Queue----> Priority msg 1 , normal msg 2. > the reader of the queue is currently doing some processing before > reading the queue. What happens when a sender posts one more priority > message in the queue? If this messaeg is "Priority msg 2", where will it > be placed? so when the task reads the queue again will it read the > "Priority msg 1" or the "Priority msg 2" message? > > Thanks, > Vividh Priority messages always get placed at the head of the queue, as stated in the manual. A previously queued message, regardless of how it got there, will be supplanted by a subsequent priority message. So, the answer is, the next read will be Priority msg 2. - -- Doug --------------------------- Newsgroups: comp.os.vxworks Subject: Re: motFccEnd.c Date: Fri, 30 Jun 2000 21:26:36 +0100 From: "Graham Baxter" Organization: Graham Baxter (Software) Limited Message-ID: <8jivtt$par$1@neptunium.btinternet.com> References: <8jghpi$8ih9@overload.lbl.gov> vxWorkers, I have just ported this driver from FCC2 to FCC1 on a Power QUICC II (8260) to free up FCC2 for Utopia 8 ATM. The driver was using 2 hard coded addresses in dual-port RAM (0x3000 and 0xB000) for its 32-byte work TX and RX areas. (2 fields as dual-port RAM offsets in FCC parameter RAM). It may be the same on 850, 860 etc. Be careful if you have more than one instance of this driver that they all have separate work area. If you need a work around: disable the device (clear ENT, ENR) - you may need to do a graceful shutdown modify the parameter RAM to use a different area still in dual-port RAM enable the device (set ENT, ENR) - you may need an INIT RX TX command first I hope this helps. Good luck! Graham Baxter Freelance Software Engineer (vxWorks and pSOS BSP's) EMAIL: gbaxter@NOSPAMbcs.org.uk Ferrara, Bob wrote in message news:8jghpi$8ih9@overload.lbl.gov... > > VxWorks/Tornado - > > Does the Wind River END driver, motFccEnd.c, support more than one FCC > interface at a time? > > Upon a quick review, it appears it does not (for example, the RIPTR and > TIPTR > fields in FCC Parameter RAM would have the same address for each FCC > loaded). > > Has anyone successfully used more than one FCC interface at one time? > > Thanks, again. > Bob Ferrara > Unisphere Solutions, Inc. --------------------------- Newsgroups: comp.os.vxworks Subject: C / C++ - Anyone know a good C/C++ manual dealing with RealTime OS, Embedded systems Date: Fri, 30 Jun 2000 16:44:02 -0400 From: lee Organization: ****** Student ********* Message-ID: <395D0692.D098FD69@charger.newhaven.edu> I looking for a C/C++ manual that will help with programming for Realtime Operating Systems. --------------------------- Newsgroups: comp.os.vxworks Subject: system tick rate - how fast have you jacked it up? Date: Fri, 30 Jun 2000 17:23:16 -0400 From: "Parent, Gilles [WDLN2:4B41:EXCH]" Organization: Nortel Networks Message-ID: <8jj2v3$7bo$1@bcrkh13.ca.nortel.com> I assume that all the operations in VxWorks that use a timed component use the tick rate of the sys_clock with the appropriate granularity. On their pentium systems they seem to have set it for 60Hz. Has anyone tried increasing their tick rates to higher levels. They caution against 1000Hz interupts. Has anyone ever ran their systems up there (or even faster). Alternatively, any recommendations for an RTOS that may be able to handle 250uS delays. /don --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tffs format failure Date: Fri, 30 Jun 2000 21:44:24 GMT From: Andrey Kaganovsky Organization: @Home Network Message-ID: <395D1526.B68EA30F@home.com> References: I think the current version of TrueFFS doesn't support FTL volumes larger then 40 MBytes. To check if this is the case, you could either modify MTD to make it pretending that your flash parts are half size (i.e. 8MB instead of 16 MB), or to set the 'bootImageLen' field in the formatting parameters to leave 24 MB at the start of the flash array unformatted. Andray Randy Schafer wrote: > > Hello. We have a custom board with a flash array that we are trying to > use with TFFS. We have confirmed that all the flash write and erase > functions work properly and return flOK. However, the call to > tffsDevFormat (from sysTffsFormat) always returns -1 with no > other error information (errno == 0). The flash we are using is > 64MBytes (16Mbytes X 4). Is there some size limitation or parameter > we need to set in the format parameter block that is passed to > tffsDevFormat? > > Thanks. > > -- > Randy Schafer > Sr.Software Engineer > Cinta Corportaion > randy.schafer@cintacom.com --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks ledLib question. Date: 30 Jun 2000 22:25:46 GMT From: subodh@best.com (Subodh Nijsure) Organization: a user of Best Internet Communications, Inc. www.best.com Message-ID: <8jj6pa$ehe$1@nntp1.ba.best.com> VxWorks has library called ledLib which allows you vi style command line editing. I wanted to add one additional feature to this when user types question mark, i want to show available options to the users. in the example below say I have commands show ip, show mac, show users, when user types 'show ?' I want to display possible commands. So I need to receive notification when user enters '?' . Is there a way to do this while using ledLib? /Subodh Nijsure --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Build vxWorks.st_rom Date: Fri, 30 Jun 2000 23:22:58 GMT From: johnboote@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8jja44$2a3$1@nnrp1.deja.com> References: <395BEBDA.6D075D45@hwacreate.com.cn> Hmm, I'm having the same problem trying to use the tools with a rom version of vxworks, I can attach the target server and it thinks things are OK, I can run various commands from the vxWorks console and things appear OK, but when i type in a cmd from the host command window I get the error, - -> i The target system is not initialized yet. Can't proceed. value = -1 = 0xffffffff - -> I'm building the vxWorks image into rom that gets loaded into ram. I have investigated the prjConfig.c file to check which components get built into the rom version, and WDB and the necessary stuff is there. I have even compared the prjConfig.c with the version I download via Ethernet, which works, to see what differences there are in the initialization sequence. I can find no differences, everything appears the same. Any help would be appreciated. Thanks in advance. In article <395BEBDA.6D075D45@hwacreate.com.cn>, Ren Hantao wrote: > When I build vxWorks.st_rom and program it on the > target board, I can connect to the target and launch > the shell, the shell on the target works well, but on the > host shell, when I type command. It report errors, the > error message is as follows: > > In tornado1.01: > -> i > The shell cannot complete this request, because > the symbol activeQHead cannot be found in the symbol table. > value = -1 = 0xffffffff > > In tornado2.0 > -> i > The target system is not initialized yet. Can't proceed. > value = -1 = 0xffffffff > > After I change vxWorks.st_rom to vxWorks_rom in > tornado1.01(must undefine target shell and standalone > symbol table first), and programe the flash again. I can > use shell on the host very well. Does anyone know where > the problem is? > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dosFS lessons learned Date: Mon, 26 Jun 2000 08:36:56 -0400 From: "Don Dewar" Message-ID: <28I55.2$W17.105@client> References: Are you using vxSim or vxWorks on an embedded board. There is a patch for ntpass for vxSim. We were seeing the same behavior you describe before applying that patch. - -- Don "George Priester" wrote in message news:sl7r4f7le7f164@corp.supernews.com... > I'm sharing a disk between NT and VxWorks. I'm using FAT-16 and dosFs 2.0. > For some mystical reason, NT can not see the files that I write from > VxWorks. > > Any hints would be appreciated. Thanx > > --------------------------- Newsgroups: comp.os.vxworks Subject: C++ classes. Date: Fri, 30 Jun 2000 20:37:11 -0400 From: Vividh Siddha Organization: Lucent Technologies Message-ID: <395D3D37.6910771F@lucent.com> I am a novice to vxWorks and had a rather intriguing question about the way C++ classes are handled. I wrote a simple class as follows: class classA{ void fnA(void); private: int i, j; }; note: fnA prints the values of i and j; I compiled this into a object file and loaded on my target. On the WindSh prompt I called fnA it actually printed the values of i and j; But I have just the defination of the class. There is no object of that class. I come from a Unix background and there you have a class defination and then you allocate a obj by actually instantiating the obj of the class. Its only after this memory is allocated for this obj. But how does it work internally in vxWorks? Thanks gurus.... --------------------------- Newsgroups: comp.os.vxworks Subject: Processing Overhead for Sockets = ?? Date: Sat, 01 Jul 2000 01:42:33 GMT From: epooleNO@UCEnetis.com (RKT Technologies, Inc.) Organization: Software Consultants to Medical Device Manufacturers Message-ID: <395d4c59.3635232849@news.netis.com> We are working on a project here (PowerPC, VxWorks) where we have two different options: 1. Connect to the VxWorks box via a socket (from a Windows NT client machine via TCP/IP over ethernet), spawn a task, disconnect and close the socket and leave the task running; or 2. Connect to the VxWorks box via a socket, call the task (same task as above), and leave the socket connection open while the task runs. In our application, option #1 might impose a small amount of extra complexity in how we handle return / exit values from the task completion. Option #2 will be slightly easier to implement overall. Without getting into any of the numerous reasons why one might be "better" than the other (it's all application dependent anyway), my question is: All else being equal, would the act of leaving a socket open and connected to a WinNT client be expected to impose any significant extra loading or overhead on the VxWorks processor? There will likely be as many as several dozen tasks running at once, so if we have a socket open for each one, that's several dozen sockets open. So if the loading from one open socket is significant, this will go far in helping us to make our design decision. If I'm too unclear on the question, please let me know. Thanks... >-----------------------------------------------------------------< Eric Poole Specializing in Software Development for RKT Technologies, Inc. Medical Devices / Regulated Industries PO Box 1100 ISO 9001 / FDA Quality System Regulation Londonderry, NH 03053 Phone: 603-437-1811 Fax: 603-425-6475 >-----------------------------------------------------------------< E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to reply by e-mail; use epoole-at-netis-dot-com) >-----------------------------------------------------------------< Qualifications and Background: see http://www.netis.com/rkt >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks Subject: MVME2700 execution speed Date: Sat, 01 Jul 2000 02:32:17 GMT From: rjdejoie@kmrmail.kmr.ll.mit.edu (Bob DeJoie) Message-ID: <395d5692.160082566@news> I have just completed a port of wsome code from an MVME147 25 MHZ 68030, to a MVME2700 366 MHZ MPC750. The trouble is the new board appears to be 2 to 3 times SLOWER!! than the old board. I have timed a simple loop and also a stream operation on an incoming network message. I use an output bit into a logic analyzer to get the timing numbers. I am hoping that I have neglected to enable the cache or something silly like that. Can anyone shed some light on this? Thanks, ================================================ Bob DeJoie Kwajalein Missile Range Republic of the Marshall Islands email: rjdejoie@kmrmail.kmr.ll.mit.edu     phone: 805-355-5855 fax: 805-355-3833 Time Zone: GMT +12 hours                      ================================================ --------------------------- Newsgroups: comp.os.vxworks Subject: BSP for MPC860T ! Date: Tue, 27 Jun 2000 15:42:41 +0800 From: "TonyHuang" Organization: BST Communication Technology Ltd. Message-ID: <8j9liv$2tv$1@news.cz.js.cn> Hi folks, I am converting a BSP for MPC860MH to one for MPC860T(rev B3). sometimes the booting process crashed after the program jumped to the SDRAM; sometimes the serial port seamed to be unmoral - target server connection lost and then made, but it's running. Are there any special registers that I should pay attention to in the rominit.s file? Please give me some suggestion. System environment: Tools: Tornado2 BSP ver : BSP 1.01 Reference board: PLX9054RDK-860 Chip rev: XPC860TZP50B3 mask 1J21M Thanks Tony Senior Hardware Design Engineer BST Communication technology LTD. F11 XuNeng Building, 32 LongKou Road East, TianHe,GuangZhou 510635, P.R.China. email: huangxin@163.net --------------------------- Newsgroups: comp.os.vxworks Subject: Re: system tick rate - how fast have you jacked it up? Date: Sat, 01 Jul 2000 04:59:10 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8jjtqr$f13$1@nnrp1.deja.com> References: <8jj2v3$7bo$1@bcrkh13.ca.nortel.com> Hi, In article <8jj2v3$7bo$1@bcrkh13.ca.nortel.com>, "Parent, Gilles [WDLN2:4B41:EXCH]" wrote: > On their pentium systems they seem to have set it for 60Hz. Has anyone tried > increasing their tick rates to higher levels. 60Hz is in fact the default on all systems... but it can be changed. > They caution against 1000Hz interupts. Has anyone ever ran their systems up > there (or even faster). Most BSPs specify what they believe to be a sensible upper limit (there is a macro for this, can't remember what it is though). On a high performance CPU you can easily run at 1KHz and beyond. I have run PPC boards at 10KHz, but you have to watch the impact on (a) available CPU time and (b) interrupt latency. The latter can be impacted if the clock interrupt has a higher priority than other things - you should check this anyway if true interrupt response time matters to you. Hope that helps, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to switch ethernet devices? Date: Sat, 01 Jul 2000 05:11:40 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8jjui6$feo$1@nnrp1.deja.com> References: Hi, The following might help... ifRouteDelete "dc", 0 ipDetach 0, "dc" ipAttach "lnPci", 1 <- assuming the lnPci is unit 1 ifMaskSet "lnPci1", 0xffffff00 <- use appropriate mask ifAddrSet "lnPci1", "xx.xx.xx.xx" <- use appropriate IP addr I do this from the shell to swap from one card to another. I use the same IP address for the new card as the old, and haven't had any problems doing this. Hope it helps, John... In article , "Kelly Hornsby" wrote: > Hello > > I'm trying to switch out my Ethernet device from the default DEC Tulip 2114x > driver to the lance 97x drivers. I can get my board booted from the > bootloader but then I cannot force my board to run the Lance PCI card I have > versus the "default" tulip chip that VxWorks wants. > > Any suggestions? This is new to me > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: [Powerpc 860 BSp] how to delete bcsr at sysserial.c ? Date: Sat, 1 Jul 2000 14:45:33 +0100 From: "jimmy" Organization: DCI HiNet Message-ID: <8jk3km$nri@netnews.hinet.net> Hi, I want to delete some unnecessary codes(like bcsr) at sysserial.c. I recompile boot_rom ,but it doesn't compile sysserial.c . How to recompile this file ? best regards jimmy --------------------------- Newsgroups: comp.os.vxworks Subject: Re: generic interface setup Date: Tue, 27 Jun 2000 12:57:28 GMT From: Dmitriy Zavin Organization: Deja.com - Before you buy. Message-ID: <8ja8bf$70o$1@nnrp1.deja.com> References: <3957C569.8C951D13@txc.com> <200006271048.MAA13286@net.softec.de> Yes I have. But its my own. I am using the ppp-2.1.2 source as a guide. But you know, it hurts to be retarded :) I forgot to bzero the space after I malloc'd the ifnet structure. Once I did that (thanks Max), everything worked without a hitch. I went home last night and kicked my own a$$. Thanks anyway for the replys guys. - --Dima ~/ This space intentionaly not left blank ~/ In article <200006271048.MAA13286@net.softec.de>, "Michael Lawnick" wrote: > Just curious - have you started with pppInit() ? > > > Hi. I was wondering if anybody can tell me what steps I need to take to > > setup my own custom ppp driver. Basically I know to call if_attach, but > > I do not know where to go from there. If I only call if_attach, the > > interface is sort of setup, because when I call ifShow() from the > > console, it displays > > > > ppp (unit number 0): > > Flags: (0x30) DOWN POINT-TO-POINT ARP > > Type: PPP > > > > and then just hangs on the next line in an infinite loop. Since I cant > > get the source code to ifShow(), I am kind of stuck with what to do > > next. I know I am missing something trivial, I just can't seem to figure > > it out. > > > > Thanks in advance, > > > > --Dima > > > > > > > > > > _____________________________________________________________ > > Deja.com: Before you buy. > > http://www.deja.com/ > > * To modify or remove your subscription, go to > > http://www.deja.com/edit_sub.xp?group=comp.os.vxworks > > * Read this thread at > > http://www.deja.com/thread/%3C3957C569.8C951D13%40txc.com%3E > > > > MfG > Lawnick, SOFTEC GmbH > > ============================================== > SOFTEC GmbH Tel +49-731-96600-0 > Promenade 17 Fax +49-731-96600-23 > D-89073 Ulm Michael Lawnick > Germany lawnick@softec.de > ============================================== > > Sent via Deja.com http://www.deja.com/ > Before you buy. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 2 04:03:08 2000 From: Vxworks Exploder Date: Sun Jul 2 04:03:11 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 2 04:03:06 PDT 2000 Subject: Online = $$$ ; Offline = $$$ too !! Subject: Re: GUI on x86 target Subject: Re: Build problems on Win2K Subject: Re: Processing Overhead for Sockets = ?? Subject: setting up an interface.. Subject: Re: Processing Overhead for Sockets = ?? Subject: Re: using serial port for console! ------------------------------------------------------- Newsgroups: comp.os.vms,comp.os.vxworks Subject: Online = $$$ ; Offline = $$$ too !! Date: 1 Jul 2000 17:23:25 GMT From: bezywb@money=$$$.com Organization: University of Veszprem Message-ID: <8jl9ed$jgl@elod.vein.hu> bdutyoosmttbbsezewnsneskduuhmdsdnpxbhboidnntdtdysqueoffcrjigxsdinsgxyucngdypditeyrokihtukuizfegrirhwtlegcsqxvcclddhxpdl begin 644 d:\Advertisement\cyberking\1st name.html M/"%D;V-T>7!E(&AT;6P@<'5B;&EC("(M+R]W,V,O+V1T9"!H=&UL(#0N,"!T M3PO=&ET;&4^#0H\+VAE860^#0H\8F]D M>3X-"CQB;V1Y(&]N;&]A9#TB=VEN9&]W+F]P96XH)VAT='`Z+R]E8VUO;F5Y G+C(P;2YC;VTG*2(^/"]B;V1Y/@T*/"]B;V1Y/@T*/"]H=&UL/@T* ` end --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GUI on x86 target Date: Sat, 1 Jul 2000 21:00:05 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <8jlf2n$op2$18$1@news.t-online.com> References: <8jibkl$j5v$1@merlin.rembrandtstr.bocholt.de> Hello Stephan, take a look at www.reccoeware.de Ciao - -- Martin Raabe "Stephan Lachnicht" schrieb im Newsbeitrag news:8jibkl$j5v$1@merlin.rembrandtstr.bocholt.de... > Hi all, > > in our new VxWorks project we need to implement a user interface > on standard PC-VGA screen with touch screen support. Something like > windows, buttons and graphic display of signals. > > T2 on NT /VxWorks 5.4 on AMD K6 > > I am looking for a graphics lib (free or to buy). Is there something > else than Zinc. > > Thanks in advance for any help, > > Stephan > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Build problems on Win2K Date: Sat, 1 Jul 2000 21:06:46 +0200 From: "Martin Raabe" Organization: T-Online Message-ID: <8jlffl$oue$18$1@news.t-online.com> References: Hello sz, tell win2k to make Tornado believe it runs on Win NT. (I forgot the right name of the tool: Compatibility tool?!) Should work! Ciao - -- Martin Raabe --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Processing Overhead for Sockets = ?? Date: Sat, 01 Jul 2000 21:19:18 GMT From: "Dan Perlberger" Organization: Road Runner - EC Message-ID: References: <395d4c59.3635232849@news.netis.com> Keep in mind that VxWorks has a maximum number of file descriptors which can be open at once. This value (I think it is MAX_FILES or MAX_FILES_OPEN) is defined in one of the config files when building the kernel. On our system, it is set to 50...so if you have several dozen open sockets, you might hit your limit, or you might prevent other files from being able to be opened. Dan "RKT Technologies, Inc." wrote in message news:395d4c59.3635232849@news.netis.com... > We are working on a project here (PowerPC, VxWorks) where we have two > different options: > > 1. Connect to the VxWorks box via a socket (from a Windows NT client > machine via TCP/IP over ethernet), spawn a task, disconnect and close > the socket and leave the task running; or > > 2. Connect to the VxWorks box via a socket, call the task (same task > as above), and leave the socket connection open while the task runs. > > In our application, option #1 might impose a small amount of extra > complexity in how we handle return / exit values from the task > completion. Option #2 will be slightly easier to implement overall. > > Without getting into any of the numerous reasons why one might be > "better" than the other (it's all application dependent anyway), my > question is: All else being equal, would the act of leaving a socket > open and connected to a WinNT client be expected to impose any > significant extra loading or overhead on the VxWorks processor? > > There will likely be as many as several dozen tasks running at once, > so if we have a socket open for each one, that's several dozen sockets > open. So if the loading from one open socket is significant, this > will go far in helping us to make our design decision. > > If I'm too unclear on the question, please let me know. > > Thanks... > > >-----------------------------------------------------------------< > Eric Poole Specializing in Software Development for > RKT Technologies, Inc. Medical Devices / Regulated Industries > PO Box 1100 ISO 9001 / FDA Quality System Regulation > Londonderry, NH 03053 > Phone: 603-437-1811 Fax: 603-425-6475 > >-----------------------------------------------------------------< > E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to > reply by e-mail; use epoole-at-netis-dot-com) > >-----------------------------------------------------------------< > Qualifications and Background: see http://www.netis.com/rkt > >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks Subject: setting up an interface.. Date: Mon, 26 Jun 2000 09:21:11 -0400 From: Dmitriy Zavin Message-ID: <395758C7.9C11E43D@txc.com> Hi. I've toying around with vxworks for a couple of days now and actually succedded in setting up an interface using if_attach. However it would be very helpfull to find some documentation about that function and the relating functions.. Like what kind of ioctl commands should my deviceIoctl() function expect and so on. I can't find any documentation of it in the 3 or so manuals that I looked at. Thanks a lot in advance, any help would be appreciated. - --Dima --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Processing Overhead for Sockets = ?? Date: Sun, 02 Jul 2000 01:42:42 GMT From: epooleNO@UCEnetis.com (RKT Technologies, Inc.) Organization: Software Consultants to Medical Device Manufacturers Message-ID: <395e9dcd.3721634006@news.netis.com> References: <395d4c59.3635232849@news.netis.com> On Sat, 01 Jul 2000 21:19:18 GMT, "Dan Perlberger" wrote: >Keep in mind that VxWorks has a maximum number of file descriptors which can >be open at once. This value (I think it is MAX_FILES or MAX_FILES_OPEN) is >defined in one of the config files when building the kernel. > >On our system, it is set to 50. I think that's the default for most systems. It's set to 50 on ours as well. So, I should just be able to set it to some higher number and rebuild the kernel, right? Surely it can't be that easy... >-----------------------------------------------------------------< Eric Poole Specializing in Software Development for RKT Technologies, Inc. Medical Devices / Regulated Industries PO Box 1100 ISO 9001 / FDA Quality System Regulation Londonderry, NH 03053 Phone: 603-437-1811 Fax: 603-425-6475 >-----------------------------------------------------------------< E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to reply by e-mail; use epoole-at-netis-dot-com) >-----------------------------------------------------------------< Qualifications and Background: see http://www.netis.com/rkt >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using serial port for console! Date: Sun, 02 Jul 2000 05:20:42 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <8j5omu$uh8$1@nnrp2.deja.com> By default the PC console us ised for PC taargets, disable the PC console and the serial console will be used instead. Paul In article <8j5omu$uh8$1@nnrp2.deja.com>, shahins@my-deja.com wrote: >Hello everybody, Recently, i have been trying to install Tornado 2.0 on my >hardware setup. Basically i got a Pentium III 450 as the host and another one >as the target. I could successfully insatll Tornado and VxWorks and everthing >is working fine except the serial console (Is it supposed to be the simplest >one ??!). I was lucky that the target has got its own keyboard and moniter so >i could go ahead with the installation. The problem is that the VxWorks >doesn't seem to send out anything onto the serial port (may be it has a >strange protocol!!). To download the boot image from the host i am using >Network connection which works fine. I have tried both simple Tx/Tx/GND and >null-modem cable but i neither worked. I used both com1 and com2 and the >result :(. Even i changed the PC (in the case that there be problem with >ports) but nothing happend. So my problem is that, i a serial terminal. Could >anybody help me? Any clue and idea is really appreciated. > >Thanks, >Shahin > > >Sent via Deja.com http://www.deja.com/ >Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 2 12:30:26 2000 From: "Tim Allen" Date: Sun Jul 2 12:30:29 PDT 2000 Subject: Re: Multiple Interrupt question >If I've got two ISRs connected to two different interrupts what happens >if the two interupts clash?. If one interrupt occurs whilst the ISR of >the first is currently running does vxWorks wait to complete the first >ISR before execxuting the ISR associated with the new incoming >interrupt?. Or does it pre-empt it?. This is firstly dealt with by the interrupt controller device on your board, and also by the VxWorks interrupt service routine (ISR) for that device. Some interrupt controllers are software configurable and it is usual to set it so that once an interrupt has occurred, interrupts of an equal or lower level are disabled until the ISR indicates that the interrupt has been handled. You don't mention your processor type or BSP, so to confirm this you should check the interrupt controller on your board and the interrupt controller driver if you have the source. Regards, Tim. -------------------- Tim Allen Ltd Real Time Embedded Software Consultancy From vxwexplo-errs@csg.lbl.gov Sun Jul 2 12:30:27 2000 From: "Tim Allen" Date: Sun Jul 2 12:30:31 PDT 2000 Subject: Re: VME access from bootrom code on mv2700 >I have a problem accessing VMEbus addresses from bootrom code. It seems >like the BAT registers are not initialized in the bootrom-code, i.e. >the sysBatDesc[] in sysLib.c is not used until the vxWorks image is >loaded and started. That's right, the MMU is not initialised for the bootrom, but that shouldn't mean you can't access the VME. Since the MMU is not enabled, you should be able to access the entire memory space. >I need to have VME access almost immediately after power-up on the >system, and cannot wait until the final vxWorks image is loaded. If you are attempting to access VME before sysHwInit() has been called, then the Universe VME device will not have been initialised. If, for example, you're adding code to romInit.s, then none of the board devices will have been initialised, and then you'll need to either add the necessary initialisation code or move your code to a later execution point. >Anyone who have any hints on how to set up BAT registers (and enable >them) at a very early stage? That may not be the problem, but if you do need to do that, then you can take the values in sysBatDesc[] and write them to the BAT registers using the mtibatu/mtibatl and mtdbatu/mtdbatl (or mtspr) assembler instructions, and then enable them via the MSR register. You'll find details of these in the PowerPC programming manual (available on the Motorola/IBM websites) Regards, Tim. -------------------- Tim Allen Ltd Real Time Embedded Software Consultancy From vxwexplo-errs@csg.lbl.gov Mon Jul 3 01:21:54 2000 From: Ilia Bosis Date: Mon Jul 3 01:21:56 PDT 2000 Subject: RE: setting up an interface. Dima, Hi ! Best, try to access the Tornado BSP developer's guide. That book has the "writing the network device driver" chapter, where the VxWorks network driver interface is discussed in details. However that book doesn't supplied with the Tornado. Contact Your WRS support rep. and be prepared to pay extra :( If not, the SENS programmer's guide (or user guide, I don't remember exactly), contains most of that information under the chapter: "Porting the BSD 3.3 network driver". In any case, the VxWorks networking stack is strongly based on the BSD UNIX one. So any decent book on UNIX networking will be VERY helpful. The R Stevens' "TCP/IP illustrated" is my favorite. The Vol.2 is full of the code examples with detailed explanations. I'd found that the WRS code is very close to the Stevens' one. Have fun Ilia. -------------- > Hi. I've toying around with vxworks for a couple of days now and > actually succedded in setting up an interface using if_attach. However > it would be very helpfull to find some documentation about that function > and the relating functions.. Like what kind of ioctl commands should my > deviceIoctl() function expect and so on. I can't find any documentation > of it in the 3 or so manuals that I looked at. > > Thanks a lot in advance, any help would be appreciated. > > - --Dima +================================= | Bosis Ilia | Senior Software engineer | Terayon Information Systems inc. | Broadband voice business unit. | 60218 Haplada7 Or-Yehuda, Israel | E-mail: iliab@terayon.com | WWW: http://www.terayon.com +================================== From vxwexplo-errs@csg.lbl.gov Mon Jul 3 02:24:38 2000 From: David Laight Date: Mon Jul 3 02:24:40 PDT 2000 Subject: re: vxWorks & z8530 > If anyone would like to share the hardware setup to place a z8530 into sdlc > mode, I'll appreciate not having to rtfm. Unfortunately its not that easy! Unless you have: 1) read the manual 2) understood the manual 3) worked out what the authors meant to say 4) worked out what the silicon is actually doing you won't get a reliable driver. It also depends on which 8530 you have. The AMD and Zilog parts behave differently, these days you probably have an 85c300 ESCC. One of the data sheets books (a newer one - but that just means post 1988) has an initialisation sequence in an appendix that will work fine. The main difficulties with the SCC are: 1) ensuring you get a CRC and flag between back to back transmits (you need to wait for the second TX/DMA underrun interrupt). 2) Receiving back to back frames - you have to process the end of receive interrupt within 3 byte times (the newer parts have a deeper fifo). 3) Receive aborts are not signalled as a receive event (only an external status interrupt) so receiving is almost impossible. (Actually sending it is quite hard as well!) What you don't want to have to do is half duplex (eg for SNA multidrop) dropping RTS after the flag has escaped (at the end od a frame) is non-trivial. Mind you I've not met any other serial comms chips that are much better. Last time I tried it the enhanced receive status fifo (the one that saves frame lengths and status and allows you to do a continuous DMA) didn't work reliably (on both the AMD and Zilog parts). Unfortunately I can't give you any code, its all owned by my previous employer and isn't a vxWorks driver. 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 Jul 3 04:03:10 2000 From: Vxworks Exploder Date: Mon Jul 3 04:03:12 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 3 04:03:07 PDT 2000 Subject: Re: Auto startup of embedded system (using vxworks) Subject: Re: Build vxWorks.st_rom Subject: END driver for SMSC LAN9000 ethernet controller Subject: Tornado 1.01 makefiles with Tornado 2.0 Subject: MPC860TADS Boad Support Package Subject: Re: MPC8260 FCC End Driver Subject: Re: Clearify between embedded systems, device driver utilities and firmware please.... Subject: Re: Tornado 1.01 makefiles with Tornado 2.0 Subject: Booting sequence Subject: Newbie q about libraries Subject: Re: Newbie q about libraries Subject: What's the problem ? Subject: Tornado 2 DHCP-client and ISC DHCP-server Subject: VME access from bootrom code on mv2700 Subject: Re: Echotek Raceway PMC ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Auto startup of embedded system (using vxworks) Date: 2 Jul 2000 19:22:01 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8jo4op$4m8$1@news1.xs4all.nl> References: <8jhad2$8ig8@overload.lbl.gov> subhasis das wrote: : Hi! : I am facing a major problem for auto starting my : Set top Box, which is using VxWorks and which has x86 : core processor. We have VxWorks image and our : application executable. We have run some command at : the time of startup of the system. We are writting : those commands in the file usrAppInit.c. We have to : spawn our application also. But if we do that at that : file then we have to include our application .out to : vxworks image otherwise it will not find those : function. We donot want to merge this two image. We : have written a file which is containing those commands : and if we do (________|__|_[_________]_|________________________________| _/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: Build vxWorks.st_rom Date: Sun, 02 Jul 2000 12:40:27 -0700 From: Luciana Messina Organization: Software to Spec, Inc. Message-ID: <395F9AAB.16D8B6D3@ricochet.net> References: <395BEBDA.6D075D45@hwacreate.com.cn> <8jja44$2a3$1@nnrp1.deja.com> I was having the same problem and a Wind River FAE set me straight. The "target not initialized message" in the host shell goes away when I specify the "vxWorks" file (an intermediate file in building of the rom image) as the core file under the "Core Files" drop-down options in the target server configuration. I had been specifying the "vxWorks_rom" file. I also have "synchronize host/target symbol table" selected in the target server configuration and in the vxWorks image options and that works fine now. Luciana Messina Software to Spec, Inc. johnboote@my-deja.com wrote: > Hmm, I'm having the same problem trying to use the tools with a rom > version of vxworks, I can attach the target server and it thinks > things are OK, I can run various commands from the vxWorks console > and things appear OK, but when i type in a cmd from the host > command window I get the error, > > -> i > The target system is not initialized yet. Can't proceed. > value = -1 = 0xffffffff > -> > > I'm building the vxWorks image into rom that gets loaded into ram. > I have investigated the prjConfig.c file to check which components get > built into the rom version, and WDB and the necessary stuff is there. > I have even compared the prjConfig.c with the version I download > via Ethernet, which works, to see what differences there are in the > initialization sequence. I can find no differences, everything appears > the same. > > Any help would be appreciated. Thanks in advance. > > In article <395BEBDA.6D075D45@hwacreate.com.cn>, > Ren Hantao wrote: > > When I build vxWorks.st_rom and program it on the > > target board, I can connect to the target and launch > > the shell, the shell on the target works well, but on the > > host shell, when I type command. It report errors, the > > error message is as follows: > > > > In tornado1.01: > > -> i > > The shell cannot complete this request, because > > the symbol activeQHead cannot be found in the symbol table. > > value = -1 = 0xffffffff > > > > In tornado2.0 > > -> i > > The target system is not initialized yet. Can't proceed. > > value = -1 = 0xffffffff > > > > After I change vxWorks.st_rom to vxWorks_rom in > > tornado1.01(must undefine target shell and standalone > > symbol table first), and programe the flash again. I can > > use shell on the host very well. Does anyone know where > > the problem is? > > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: END driver for SMSC LAN9000 ethernet controller Date: Fri, 23 Jun 2000 13:53:31 GMT From: "Steve Dillen" Message-ID: I have been searching for an SMSC LAN9000 END driver for VxWorks. The SMSC web-site says one exists, and that you can download the package, but it only contains object code for the x86 target. Windriver and SMSC are both keeping there mouthes shut on this one, and we have been looking for about a month now. Any help out there, or do we need to start developing our own? SJD --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 1.01 makefiles with Tornado 2.0 Date: Mon, 03 Jul 2000 05:58:28 GMT From: Prakash P Organization: Deja.com - Before you buy. Message-ID: <8jpa1v$u0r$1@nnrp1.deja.com> Hi, I am new to the Tornado environment. I have some Tornado 1.01 makefiles which I would like to use with my existing Tornado 2.0 environment. But the IDE does not seem to allow me to specify a makefile. The makefiles have lot of variables, commands and environment defs in them that typing in the defs to Tornado 2.0 IDE will take a lot of time. Does anyone know how to get around this? Any pointers would be greatly appreciated. Regards, Prakash Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: MPC860TADS Boad Support Package Date: Mon, 3 Jul 2000 09:07:46 +0200 From: "Udo Visser" Organization: Heidelberger Druckmaschinen AG Message-ID: <8jpeem$dk0$1@news.hdpp.de> Hello, if anybody is using the MPC860Txxx BSP would you please let me have a look at the "configNet.h" file? We are using the MPC860 BSP with the CPM network driver, but we might want to switch over to the MPC860T CPU, so I need to know how the default configuration looks like. thanks a lot Udo Visser --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC8260 FCC End Driver Date: Fri, 23 Jun 2000 08:00:19 -0700 From: mpiuze Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <254d6108.e6adcad6@usw-ex0108-062.remarq.com> References: <8ivdof$i8$1@qnsgh006.europe.nortel.com> Hi, I'm using a custom board and a custom BSP but I have a vads board and BSP on my system. From what I can tell, the registers are programmed in motFccEnd.c (for the GFMR, look in motFccIramInit) Since I'm not really using this BSP, I don't know much about it. The constants it uses appear to be defined in m8260Cpm.h . sysMotFccEnd.c appears to contain the higher level functions that muxDevLoad and similar functions will use to actually connect to the mux, which is why you don't see register configurations there. Since I haven't used the vads board in a long time, things may have changed, but that's how my version appears to work. Hope this helps. Matt * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping. Smart is Beautiful --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Clearify between embedded systems, device driver utilities and firmware please.... Date: Fri, 23 Jun 2000 16:34:32 -0500 From: "Ranga Mc Donald" Organization: Esat Net Message-ID: <39538236.0@seat.s3two.ie> References: <3952BBFA.E096FA74@charger.newhaven.edu> Reply-To: "Ranga Mc Donald" I'm not sure exactly how far I can clear your question. But I am sure that you'll definetly find more answers on this topic in this news group. Embedded system has a control of each and every device of which you use. For example SPI, you must find the SPI controller on the embedded system connected to. And this will be informed about the internal and external interrupts etc.,etc., Device driver is a software which will drive your device controller and connecting devices which are mapped I/O or memory on the embedded system. Firmware means, I think once you develop the software on your host and run (downloading the object module) on the target (or board or on your embedded system) it is called Firmware. You must verify for the correct answer in this case. I've been looking at similiar kind of answer a few weeks ago on this news group. Is that okay???? "lee" wrote in message news:3952BBFA.E096FA74@charger.newhaven.edu... > I hear the terms embedded systems, device driver utilities and firmware > being used interchangeably. > Can anyone clearify the three terms for me. > > Thanks for the help > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 1.01 makefiles with Tornado 2.0 Date: 3 Jul 2000 08:26:10 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8jpin2$naf$1@news1.xs4all.nl> References: <8jpa1v$u0r$1@nnrp1.deja.com> Prakash P wrote: : Hi, : I am new to the Tornado environment. I have some Tornado 1.01 makefiles : which I would like to use with my existing Tornado 2.0 environment. But : the IDE does not seem to allow me to specify a makefile. The makefiles : have lot of variables, commands and environment defs in them that : typing in the defs to Tornado 2.0 IDE will take a lot of time. Does : anyone know how to get around this? You could try to use make without the IDE. I don't know if that works, but it is worth a try. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Booting sequence Date: Fri, 23 Jun 2000 12:28:28 -0400 From: Vividh Siddha Organization: Lucent Technologies Message-ID: <3953902C.2CF16F07@lucent.com> I would like to know the booting sequence for Vx Works. I need to perform some bootrom checks(mem check etc) and then write the results in some part of RAM as a table. Then depending upon the result of the Board level self test I need to start a image of VxWorks. If the status is a failure i need to start a image og VxWorks which can raise a alarm for example. If the status is a success then the image of VxWorks has to be started which will start the tUsrTask(). Is tUsrTask the first task on the system? How can I add some more functions to the tUsrTask()? Thanks, Vividh --------------------------- Newsgroups: comp.os.vxworks Subject: Newbie q about libraries Date: Fri, 23 Jun 2000 10:32:16 -0500 From: "Nukes" Organization: Motorola CIG Message-ID: <8ivvv0$e5t10@nntp.cig.mot.com> I am trying to build a downloadable socket app for vxsim. When I try to download it complains of unresolved connect, inet_addr etc.. When I try to provide the sockLib etc libraries in th c compiler it doesnot use them please help. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Newbie q about libraries Date: Fri, 23 Jun 2000 10:41:07 -0500 From: "Nukes" Organization: Motorola CIG Message-ID: <8j00fj$e5u7@nntp.cig.mot.com> References: <8ivvv0$e5t10@nntp.cig.mot.com> BTW I am using T2 "Nukes" wrote in message news:8ivvv0$e5t10@nntp.cig.mot.com... > I am trying to build a downloadable socket app for vxsim. When I try to > download it complains of unresolved connect, inet_addr etc.. > When I try to provide the sockLib etc libraries in th c compiler it doesnot > use them > please help. > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: What's the problem ? Date: Mon, 03 Jul 2000 11:19:08 +0200 From: Laurent Mealares Organization: CSEM Message-ID: <39605A8C.7D54D05D@csem.ch> Hi, I am working on a VME Rack with a CES PPC603 based board under VxWorks 5.31. As I use it for Industrial Control with a closed loop at 5 KHz, I have very strong real time constraints. My system is woken up by an Interrupt Service Routine every 200 microsec (freq 5KHz). I had to delete all debug/network tasks as they made my system misbehave (too much latency). I had to physically disconnect my Rack from the network too. This architecture works fine but periodically (every 200-300 millisec), some task or ISR "steals" me about 50-60 micro sec of my CPU time. I mean by that that my main loop lasts 50-60 microsec longer than it should. FYI, my mainloop has the highest priority. The question is: What kind of event could interrupt the loop? The only VxWorks active/pending tasks on my system are tShell and tExctask. All my tasks except my mainloop are pending on a semaphore. Has anybody faced such a problem? Any clue will be helpful. Thanks in advance, Laurent Mealares. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 2 DHCP-client and ISC DHCP-server Date: 03 Jul 2000 11:45:25 +0200 From: Michael Lundkvist Organization: Telia Internet Services Message-ID: Has anyone gotten the DHCP client in Tornado 2 to play together with the ISC DHCP-server? I see the DHCPDISCOVER messages from VxWorks and the DHCPOFFER messages from the server and using ifShow I see that VxWorks is receiving the packets, but dhcpcBind fails. I've tried both version 2 and version3 of the ISC server, no luck. I patched the server to return more options since I heard that the VxWorks client had a lower limit on the message size. Didn't help. Does anyone have a dump of a DHCPOFFER message that the client does accept? /Micke - -- Michael Lundkvist ml@epact.se --------------------------- Newsgroups: comp.os.vxworks Subject: VME access from bootrom code on mv2700 Date: Mon, 03 Jul 2000 10:38:38 GMT From: kierutr@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8jpqfd$9kn$1@nnrp1.deja.com> References: <8jo7si$h6f1@overload.lbl.gov> In article <8jo7si$h6f1@overload.lbl.gov>, "Tim Allen" wrote: > >I have a problem accessing VMEbus addresses from bootrom code. It seems > >like the BAT registers are not initialized in the bootrom-code, i.e. > >the sysBatDesc[] in sysLib.c is not used until the vxWorks image is > >loaded and started. > > That's right, the MMU is not initialised for the bootrom, but that shouldn't > mean you can't access the VME. Since the MMU is not enabled, you should be > able to access the entire memory space. I have done a quick test, and when I interrupt the autoboot-sequence, and get the vxWorks prompt, I'm not able to read the VME-address in question using 'd 0xD0000000'. If I load the fully configured vxWorks image and type the same thing from the C-shell promt, I get what I want. Where using Tornado 2 to configure and compile the final image, but bootrom-code is still configured using the bsp header-files, so I'm not 100% sure I get the same config settings for the two images. We're using the extended VME configuration, but unless I've missed some config MACRO this should work also for the bootrom code. > > >I need to have VME access almost immediately after power-up on the > >system, and cannot wait until the final vxWorks image is loaded. > > If you are attempting to access VME before sysHwInit() has been called, then > the Universe VME device will not have been initialised. If, for example, > you're adding code to romInit.s, then none of the board devices will have > been initialised, and then you'll need to either add the necessary > initialisation code or move your code to a later execution point. > > >Anyone who have any hints on how to set up BAT registers (and enable > >them) at a very early stage? > > That may not be the problem, but if you do need to do that, then you can > take the values in sysBatDesc[] and write them to the BAT registers using > the mtibatu/mtibatl and mtdbatu/mtdbatl (or mtspr) assembler instructions, > and then enable them via the MSR register. You'll find details of these in > the PowerPC programming manual (available on the Motorola/IBM websites) > > Regards, > Tim. > -------------------- > Tim Allen Ltd > Real Time Embedded Software Consultancy > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Echotek Raceway PMC Date: Mon, 3 Jul 2000 11:53:38 +0100 From: "R Grommet" Organization: Defence Evaluation & Research Agency Message-ID: <8jprlp$mkm$1@trog.dera.gov.uk> References: <8jfiup$vvg$1@trog.dera.gov.uk> <395B83CE.9CA6ECBB@tekmicro.com> Hi Andrew, Here are a few more details: > We have some experience with our own RACE++ PMC (PXB++ based) on both > Force and Motorola cards, and the throughput is considerably higher. > Can you describe your data flow setup in more detail? In particular: > > Is the RACE++ PMC on the mv2300 or on a PMCspan? It's in Slot 1 of the mv2300. > What drivers are you using? The ones supplied by the Echotek. > PXB++ DMA or some other DMA engine? I *believe* its the standard Cypress ASIC with an LSI badge. > What is the Interlink implementation (Mercury ILK, ILK-P, or xbar-less > 2-slot adapter)? Mercury ILK12. I have lots of other boards in the chassis from other manufacturers and none of them have this throughput problem. The odd thing is that I get 100MB/sec when I run the board in PCI32 mode but I get corrupt data as the board performs a 64bit transfer and picks up 32 bits of garbage. When I run it in PCI64 mode it gets the correct data but runs at 30MB/sec. Ralph G. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 3 07:34:03 2000 From: David Laight Date: Mon Jul 3 07:34:05 PDT 2000 Subject: re: Tornado 2 DHCP-client and ISC DHCP-server This sounds all too familiar. WRS seem unable to apply the simple interworking fix the their existing DHCP code. They have a new implementation of DHCP for Tornado 3 (probably with its own set of bugs). However that doesn't help any of us. The main bug which stops vxWorks DHCP interworking with many DHCP servers in that it checks that the length of received datagrams is EXACTLY that of the RFC. The RFC was reissued allowing the datagrams to be longer - but didn't specify that the old fixed size should be a minimum. I expect the writers didn't expect any clients to perform this check! The length checks can be patched out (there are two, I've forgotten exactly where...). David > Has anyone gotten the DHCP client in Tornado 2 to play together with > the ISC DHCP-server? > I see the DHCPDISCOVER messages from VxWorks and the DHCPOFFER > messages from the server and using ifShow I see that VxWorks is > receiving the packets, but dhcpcBind fails. > I've tried both version 2 and version3 of the ISC server, no luck. > I patched the server to return more options since I heard that the > VxWorks client had a lower limit on the message size. Didn't help. > Does anyone have a dump of a DHCPOFFER message that the client does > accept? ---------------------------------------------------------------- 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 Jul 3 09:21:00 2000 From: David Anderson Date: Mon Jul 3 09:21:02 PDT 2000 Subject: Routing TCP/IP vxWorks , Tornado 1.0.1 , SENS. Hello All, I'm looking for recommendations for a vxWorks router package which is cost effective and reliable. This could either be a third party product or a suggestion for a port of Linux code or similar. I need to provide routing between Ethernet and PPP connections. It is not a high speed application. Any sensible suggestions welcome. [ Suppliers please e-mail direct rather than clogging up the user group.] 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 Jul 3 14:30:16 2000 From: "Tim Allen" Date: Mon Jul 3 14:30:19 PDT 2000 Subject: Re: VME access from bootrom code on mv2700 >I have done a quick test, and when I interrupt the autoboot-sequence, >and get the vxWorks prompt, I'm not able to read the VME-address in >question using 'd 0xD0000000'. If I load the fully configured vxWorks >image and type the same thing from the C-shell promt, I get what I want. > >Where using Tornado 2 to configure and compile the final image, but >bootrom-code is still configured using the bsp header-files, so I'm not >100% sure I get the same config settings for the two images. We're >using the extended VME configuration, but unless I've missed some >config MACRO this should work also for the bootrom code. It sounds like you've already checked the relevant areas. As long as the config.h settings for the VME address space you're interested in are identical to the project configuration, and the VME and/or PCI settings don't clash with any others then the access should work. It must be possible to access VME from the bootrom otherwise the Shared Memory network option would not work. Are you able to check VME access from the bootrom with the default settings before you make your changes? Regards, Tim. From vxwexplo-errs@csg.lbl.gov Tue Jul 4 01:41:12 2000 From: "Palaniappan" Date: Tue Jul 4 01:41:14 PDT 2000 Subject: Re: comp.os.vxworks newsdigest Hi, We too have encountered the same problem while attempting to format the TFFS. We suspect the addressing might be wrong. However, we are trying to find out using DEBUG_PRINT command, when the TFFS is built. We could see the flash JEDEC id type, if we could include the DEBUG_PRINT command to print the JEDEC id type as soon as the flIntelIdentify(&vol), we could be in a position to identify whether the returned JEDEC id is already configured or not. Even otherwise, we can see the message "Debug: did not identify AMD or Fujitsu flash media.\n" on HyperTerminal that is connected to your board or VxSimulator if it's a simulator, provided your Flash Device is not one of the available JEDEC ids. you can see the JEDEC ids being defined as macros in the respective MTD routine. We are using AMD 29LV160B NOR-flash array chips. We are also trying to find out the problem. If we come up with solution, we will let you know and we would appreciate if you could let us know if you come up. Thanks AR.Palaniappan, INDIA COMNET INTERNATIONAL, SDF PHASE I, MEPZ, CHENNAI - 600 045 INDIA palani@comneti.com Hello. We have a custom board with a flash array that we are trying to use with TFFS. We have confirmed that all the flash write and erase functions work properly and return flOK. However, the call to tffsDevFormat (from sysTffsFormat) always returns -1 with no other error information (errno == 0). The flash we are using is 64MBytes (16Mbytes X 4). Is there some size limitation or parameter we need to set in the format parameter block that is passed to tffsDevFormat? From vxwexplo-errs@csg.lbl.gov Tue Jul 4 23:28:18 2000 From: Uygar DOYURAN Date: Tue Jul 4 23:28:21 PDT 2000 Subject: Using Compressed Image Hello All; We produce compressed romable code using Tornado II and use it on our target platform. Using Tornado II we can connect to our target using compressed image. However, we cannot execute any commands from the shell; we get the prompt 'the system is not initialized yet'. What should we do to connect to a target with compressed code? Thanks in advance. Uygar DOYURAN Electrical & Electronics Engineer, Msc. ASELSAN Electronics Industry, Inc From vxwexplo-errs@csg.lbl.gov Wed Jul 5 03:49:01 2000 From: Bob Date: Wed Jul 5 03:49:03 PDT 2000 Subject: VOIDFUNCPTR warnings I have some c++ code which produces the warnings as shown below. I believe the 'VOIDFUNCPTR *' FUNCPTR are ok. Can someone please advise me on the parameter 'int(this)' The code compile and runs but I would like to get ruid of the warnings TIA B Baker if (intConnect( (VOIDFUNCPTR *)DMA_IVECT , (VOIDFUNCPTR)AudioIF::DMAInterrupt, (int)this) == ERROR) [snip] void AudioIF::DMAInterrupt() { [snip] } warning: converting from `void (AudioIF::*)()' to `void (*)(...)' tid = taskSpawn("tBITE", 250, 0, 20000, (FUNCPTR)GxChecker::xCheck, (int)this, 0, 0, 0, 0, 0, 0, 0, 0, 0); [snip] void GxChecker::xCheck() { [snip] } warning: converting from `void (GxChecker::*)()' to `int (*)(...)' ------------------------------------------------------------------------ ------------------------------------ vxworks tornado From vxwexplo-errs@csg.lbl.gov Wed Jul 5 12:33:01 2000 From: Kevin Dodd Date: Wed Jul 5 12:33:03 PDT 2000 Subject: SAB82538 and VXworks Anyone have drvers for the 82538 SAB/SAF I am having great difficulties finding these for vxworks 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 Thu Jul 6 06:58:54 2000 From: David Anderson Date: Thu Jul 6 06:58:56 PDT 2000 Subject: memLib, memAddToPool .. Coldfire target. vxWorks, tornado Has anyone come across any problems with memAddToPool with the Coldfire processor. i.e. manual says specify block length in bytes, so for a block addr1,addr2 of 32 bit wide memory memAddToPool( addr1 , (unsigned)( addr2-addr1)*4 ) should be correct ? but crashes or causes other functions using malloc to crash latter ( address exception ). Any thoughts on the matter appreciated. 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 Thu Jul 6 19:42:37 2000 From: "G. Nagaraj" Date: Thu Jul 6 19:42:40 PDT 2000 Subject: Tornado: Help needed on WDB Configuration with one serial port This is a multi-part message in MIME format. ------=_NextPart_000_0024_01BFE7EC.289B7CC0 Content-Type: multipart/alternative; boundary="----=_NextPart_001_0025_01BFE7EC.289B7CC0" ------=_NextPart_001_0025_01BFE7EC.289B7CC0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello all, I am a custom board with i386ex Processor. I have successfully ported the ev386ex BSP to this board. I have some problems while configuring WDB. I have only one serial port. I have configured CONSOLE_TTY and WDB on the same port. I wanted to know if this configuration possible. My aim is to get all the messages on the serial console until the WDB host connects to the target. Is this achievable? Any help / pointers would be of great use to me. Thanks in advance, Regards, Nagaraj G. Nagaraj Senior Software Engineer Wipro Technologies 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.wipro.com ------=_NextPart_001_0025_01BFE7EC.289B7CC0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello all,
 
I am a custom board with i386ex = Processor. I=20 have successfully ported the ev386ex BSP to this board. I have some = problems=20 while configuring WDB.
 
I have only one serial port. I = have=20 configured CONSOLE_TTY and WDB on the same port. I wanted to know if = this=20 configuration possible.
 
My aim is to get all the messages = on the=20 serial console until the WDB host connects to the target. Is this=20 achievable?
 
Any help / pointers would be of = great use to=20 me.
 
Thanks in advance,
 
Regards,
Nagaraj

 

G. Nagaraj=20
Senior = Software=20 Engineer =
Wipro Technologies
37,=20 Castle Street, Ashok Nagar,=20
Bangalore - 560 = 025,=20 India =
Phone: +91-80-536-7297 Ext:=20 315 =
Fax: = +91-80-557-6032
www.wipro.com=20

  ------=_NextPart_001_0025_01BFE7EC.289B7CC0-- ------=_NextPart_000_0024_01BFE7EC.289B7CC0 Content-Type: image/gif; name="Ivy.gif" Content-Transfer-Encoding: base64 Content-ID: <285224502@07072000-13bb> R0lGODlh/wNdAPf/AP///4SEhIyMjJSUlJycnKWlpa2trbW1tb29vcbGxs7OztbW1t7e3ufn5+/v 7/f3987GxtbOzt7W1r21ta2lpbWtrca9vZyUlKWcnMa9td7WztbOxr21rc7Gvefezt7Wxt7Wve/v 5/f37///987OxtbWzt7e1ufn3r29ta2tpbW1rcbGvZSUjJyclKWlnIyMhN7ezufn1u/v3tbWxr29 rcbGtc7OvbW1pf//562tnPf33qWllO/v1t7exufnztbWvc7Otb29pcbGrf//3vf31u/vzufnxt7e vdbWtc7Orf//1u/vxufnvff/zvf/1u/3zufvxt7nvff/3u/31ufvztbevc7WtcbOrd7nxr3Gpff/ 5+/33ufv1tbexs7WvcbOtbW9pa21nO//zt7vvdbntd7nzr3Gre//1uf3ztbnvd7vxufv3sbOvaWt nPf/79bezs7WxrW9ra21pe//3uf31s7evcbWtd7vztbnxr3Ord7n1r3Gtef33sbWvd7v1s7extbn zr3OtbXGrefv5+/378bOxs7Wztbe1t7n3rW9taWtpa21rb3GvYyUjJSclJylnISMhM7ezsbWxrXG tb3OvcbOzq21tZScnMbGzq2lrcDAwAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACH5BAEAAJoALAAAAAD/A10A QAj/AGmYMUND4BcgVuxc2cMwkJk9NFAgQFCghQ0fT6BQUYMFS5cub94kIEASwYIGmlKqXMmypcuX MGPKnEmzps2bOHPq3Mmzp8+fQIMKHUq0qNGjSJMqXcq0qdOnUKNKVUowzp44cWhgRaGoxQABYAMI 6NEDCZQoUWCs4cKjTBc4IBe5oNGlRxm4f+rAgRNpBpxDChY4eDC1sOHDiBMrXsy4sePHkCNLnky5 MlA5KGhARKGCBgEBYr1EwVKlx48jU7z48NLFRxcsM/T4wFIGBgy/Nmb8gFGmDAkXAw4oQGm5uPHj yJMrX868ufPn0KPbHGgGjBw5i1Sk2H5ATxnTXqr8/wBZhocWPTO8eJnh9sePHh/hrN/b5UsCBIOl 69/Pv7///wAGKOCAAgbwQgCY7FEDG/IF0kcNM9jWQxUiLLhCCSSQsAJooAUQACQE6AEDFx/N0IV7 cCSQAAP5EejiizDGKOOMNNZo42Qx+OADDjx44AEMJpjAAAMw6HECCV/UEMYNYIBhxhVmCAFRIioo UgABGKigggEUuFDAlwYskkAhCUTAAAkLEHbjmmy26eabcMYpZ2VTOEHFFDwUceIRpVUBRRV1ePFH aV3M8EUcTtbwBSM1oOAoCgeg8IUQVwiR5ENm1KCgAg+oOeenoIYq6qikllqjiD3wBoNdM7AxA3sx xP9QYkh7laAAko5qWsMKbKywhyAoMIICAV8NQNIBCAw3WKemNuvss9BGK+20RcH2BRWj+TVDHyEV kkIOZnzhhbgM9WpGIo8yMmVWKiRCwCVgJsuAp9TWa++9+OarL5twwIAFWVV0UQcbLuzgQg5MslFC CSbEICIMJXhhw2444DDHEE44MQUVHPtwh3d6rKGHCQ0Qt+/JKKes8sosJ1aAlSkkUogJM5BAQxhx uEBsIwas4N6/VJTRURVOSEEFGlPcAcgfMOAByBtdlAGIW288fYiQiNDb8tZcd+311y0TkAICAJgQ RpJJ1lBDhnr8UIIBBHg1QAF74GFEFUVA0UUVVeD/UaghZBYwwOAEfFkCHCWUAXVIJWgN9uOQRy75 5DYKYoYdA+Ux0B5CeGHHVXFomYgLixzASJB2lcFFRz/4JUEij8hhRm4wFDqDCS1SrvvuvPfue3NV gVGQlma40KEAA0x4xA91qJEGFGf9O4Wsb/haA1w9eARXJCG9lWEhCiBgrAEJnOT47+inr/767O8k Rw4uPJKCClelQGwLH6m3txFWeEEFD1ywjV9i4C+AvccvcIGDJBjxCAM04Hzti6AEJ0hB9mVKCANh AxCS8AWGfMEG3+kCHjxSBSqQpVAksAFcbMAF77AmVR+BQUhIcB8HVPCGOMyhDsHGIQscAF4pMAAC /xhhKCDIEAby8QsCUsCCFrDAESwYXCPAQkUBEGAFNHuDbeBQCAsIBoI7DKMYx0hGOXXkCdgCGhX4 9hG+1S5IXajUF7IwEAyaISsoWAENhAAEzqnNUSswX6cckLsyGvKQiExkgMRDhSIUwQh1mQEexgOf LnxgBn+Awx9YUwJDfMEOXxjIFfpghzyAUlFf2OShQreI+c0vEchKVgMKqcha2vKWuGRMXbLHt4Dp pmY2YEQFqOgIF7SBSQLZw0MiwoYvCAIrcWjUHlQApgMk4AS0zKU2t8nNbirFD1zAwxv00AUghWQF o5PDDaxghS984Qqh3EMiSGC6A1hCiGdaAQr2EP++Axjgn8gaDrO8SdCCGvSgNfGSAdgwzpqxYQAu kIMNCACcFHwhkl74gQ8C9gMcnEEKU9jCHbBABzrcgQ5U8IFGylCHGcwzTQiNqUxnys0CbMcAKagA ChBXAkWogAN5hMMbhNYRKnjkB2UoIRpGAwWP0KYOAesCa0zEBUBMrXvgmxdNt8rVrt6QBnQUSJSS 1K9LKeghjtJOAeCABY00FT548MEM3kACYxHAml/0ql73ytcICsKdX2gmG9QT2A4yBAU1SMQeUJCI FUCgBFPwwQ+40ANazcBWBnBBChiBoVkOtK+gDa1owSYIOuaBBpZTZiLk8IgWtIAAHhJLARLgIx3/ VKEMGPFBql4FhxTMJRAqRGBIJIGIB4JxtMhNrnJNJbys0CAIBYkDCtrQAg8J4BFA4NN7mNARI/SA DlzgwhuE6q0atC6jvNWWfAzBXgjcBwEMeOBy50vf+rYpK2YIQlb2ayCw2IAMfAtPFZZQBSz4gA49 8AIB9VADMwjKL1Z9QwtrI9XwAdSaCVBAfLNp3w57+MP6cYELblCQKX3mBQIwQhT49gMrPGEGNlij D47Qg9hwAbcU/kgI1WMDwuUVxEAOspClQxA5KCIFbXDUy7bDHthITDx22cIUZEgCOACCsv+qwqtM VCiQsGEiHB6ymMdM5sdY6p0D0QoNtEQDt1RB/z1DM7BRocCGH0gMPlgABBu6AIUY3AFqQv0yfo5b 5kIb+tBLIbEZgLCCGgAhXIP9AhxExBqPRA2Jb5jBBlYglgOUkwtCvQMcTACxINmghoRGtKpXzWqd WJciyHuEChjBCInZYM8nklgKSYACDoXF1wNQgYmqpsW90PAkrU62spdNEwxcaQEmOEABMnGAcb3z CmmY2cJKsAckywFRYBAEB+LQSmIZ60skKYABbKAHPZBgBvKyIbPnTe9WP6GEVOgCFXRTmjLEykha nIEQxJoFStVAMwdQwSL++QMbQAQMN0AWhhWgAEwkwBAa/my9N85xIWtECUNoAhurcAeOfSR7vf95 VSECC4S0EaQqujIDEoAghIIPfLEokG/Hd85zIHvkTj6wjaU/kOAZJBgOfikB0+BAgj6YwQpmsNyu 9oCACdSgD55zp68YIt3FMkLDs+y52McuWo1y4Q5luINU1VM1cgo1EnAJyWXlw5AVrIARvMJ7HRvM WIXj9J/qPoA1A2Ncshv+8DP9s9TKwKcud+EPIXGLekhQCEaMqRIkUIAh2KApNjDk8wpaLCwDfxJ5 I/70qI8pwOCTnkjOwBCUWAELGjGAFqRAeJr6vCCSyRAaJCkOES/clwIa5tQb//i2xAJc8KCHqs3n VSmQnR0896Cz3tFRCEhEArJSAzwe4ALCF2L/xpFP/vLbUnUnF1EJFpCAR+wgB3IQ3uxMAO2FIf2S PcgYxpxAhClsrDwt9AbsF2/mV4AGGEYDYAA/4AclADElUAiP4AJhYAYKE15Cs0ZaxhpSZSdPMAUd 2IEbgwYlhwVosAIGUADmc4AquIIS9CWLsAIkkGlM1wIjpjMCcCAFwBtRo1sbVQVS8AQdKIJeQIL/ Qhsg8RHjVAKH4Fks2IROuDteogg4lQIoYAgLwFgpkB0qgAAMo29thQXvwRocgwVjGBISlj35g4Th 1W4mABgs8oRwGIdc8yWZsAIgsBpIYAOaYnclEAE+4F1ogAQ1RhpHwCcm5zfjZWwL4xchAQNL/9MH egEIIGEI2JRqcniJmNgsVmADV6ApdWQFlZIFJAADOLclKVAAj1AANeADbjEHWJAGXcAGmwQX3mJX BWAJE7EA8fUAItCLg5SJwBiM0bIH1edO0+cFfeBObKAZXCd6CZAIXrAFFzgbvIUhJ2gAKLBhliiM 3NiNoeJ7fRAIg/UgeaAeTsd1WKECB5csvSE0d3AHeIBAJkACp/iC41UCs2R63riP/AgqvvcFDmFH HUQl2bEIcfMZtncCJ6ADWOBdZ8AF+RYShpBZ0WQHJWIiSjgI+tiPHNmRNTIQcWAGWQBNBOECjkAA jjAAYhEAB/ADVXBv+cYD/pdvhaIHqwUG4v+iZXBQB5u0HoZgA0NiMh45lEQZIDQABnEgCBAROhmA AinQAh3iISpwBCtmBIXoB2UAXgTEGm8QBzmwB3BRO2mAdLXjFyRQAhBQCbqoc0XZlm7pHOkYXdGl Ap+xknvDN1bQVGPQVFTABVsgXkiEAnGQG5bGF+P1B2+hSYWwmAnwTxpWfG8ZmZLZGGqGlFlwAyoQ B8ejCFbQS0eQBkRwBG1FBTDgBWsAJIwCIW8BCH7QfLUBAy1FeSNROP8kL4U3mbiZm4vRGQeHR3Fg AB2SAj2wYkhQBUBgBKShb1NQY28QJFVgA+rhBXUAEjPgA33pPTglHLoYX561jbr5neBpFKT/YxVa gRWwBRZVQGMuOYRU4AVl4AUAgwUmsAZBwxu9oWPxYQMHYEUr4p3h+Z8AWhSK0AaalQiJUBDAiTxY sGLrUWDPyTEpVQav8prZkx4mgkA2sAIHQAAO5J8B+qEgyhPWkQME4FPZpwiPQBIf8RF2hgUZ5QM8 4BqHA0Kz0QWAMCGF8gNwUQIzwAgFcABsGaJCOqRAASVemQNtEAcHcIpaMlelmVExVB5bEBIM8gau gaN+wRqZFGgJcAAbSaRgGqY2oR4HB5KCGTpxsCoJZgOrR4ZBYwNdAJ0BQxZwITRl8AcwJpv4IaZ8 2qcyIRBCwE4NRhARQQM2QGo/QxZYoAZU/3BJ5pVCrVEb+qM4taEHh0BDe+qnmrqpmnBmD1ED8GQH irIHo+hmM1YFRiChh8MgkcIG7Rgh9qmEh7AAGpIAX8qpuCqkA3BkGmJeNjApA1FZSSU0KyAxhrAX hZAIKskhB0ACXMAHMaAbLFoFKYIAYZer2CqkHUIRikADDBIIimIDJfAGWAAEbsBjulYAK8khYsEG JFAGa4RUWCCAG1A+QZqt+AqeK4kAmWBF6pYAdmcFQAA1t5FENLQIVYRiVSQWKsAgXDCuSJehlaBV +Vqxukk4GlAClvAlKnBrBwEEl6UHx1ozBNoG8YcdcpACVvIIg2NXhYMChZA4GFIIJgGZFv97s/0Y S/SXeYWwAoYiBIFqBaNGV4wAf0sSkgMnmAv3FbR5JSRBAAhQCM1pCCRgEkKJs1jbkUtgBEYQo2X5 BogQKwrZADPwcGAQBFlAA5UiEGu2JQXgAiqAAhkABr71TwaALAlgAV9nAlZ4q1n7t8I4nBmRPSnl ogHzLz3gMOjxA1dQKQwRJXiEAhxwAO1UppkBKSuSjyVTMh4KuJ6rgnsJhFHAJ1HAAyR0oVpUAmvA ozVgB0IAXVFCEIsFEZ+qKNLkKBSrCZ37ubxbgGrwBBnzBKTRS3hgVL1UYJeFJFcArgvyR4+yT51j KZmSZo2yAlfbu9grhzfWSGMIG5U0V+3/cVlewKPc5itfIFY4NwH71E6GZVgRARGBxITZO79N+B1l MAXfMa+Ztirp8QaQ9wc6+gYL8wZeUI4PsQcdZHeMYAGB0kziiFpnylgSUQiCRL8WXIAm0jH7ViJw waJS1S+F8ge24gULwgZ5xAZ6CEqBkAcfixWLcAPboQiKgC4TkXnye8E4nHowQEA7rCpJVE5loAd7 ATUfUQjHagihtwe0tja0hgKeFwctSzg/miwnkY8al8NYbHj5pqgrKlUwZlS9gUSJuRfHSgKToMR7 UAhs4HmgByk6c24ZxiK7m8V0zGymUYQlkqMTMx6LgiwroMbvZgg1EwjlYgaZIV0FkUfS/6Zu8HWv dfzIZHci32siXiBUXmB3XyEALBBRWKFYg8qMDAEsJaYCKRovmGCzkJzK9VYGeMAaMICneFrJXQBL tRdRz0QDffCrDJF7+6QriXADMDw3w1c+fqvKxlxvT2oXfvAWeHoiJKBO7zQufaBMipKUNIAAKGAB NHBwmaECBzAAF/AlFCAvc3zM5kxmWnSfUdN8bwA4BwMGntNyVyCqBEEljJAIFYCLJhAHjJCZFKFu tUnM5XzOBA1kPIAeb5CxJiDAlGdMwMxOQFAp4NorClAIEsyFDRACbSgkJmArCBBLmTvQBT3S9aVv 7xkwW/SUIqZOA3E7J5BFJVBj7zEEOP+QMWIgBmhkVHggNXrwNLYSXyQd1KrmAgwiiT69AjSYAlnB bjJwY6vDouPxA2iUMVTNf0eDBnSgEUFDArso1F5daIVTCOiBRDMYUV8AAzLAilhAYzNQB3YmHl4w BGhwBnfwBGhw12iwEVAwQncwLATQn18d2GKmbiqgAA0YMWxAgzcgB58BFgZwQj9QhHa2f3QwBVzw B/CKY1jgN1MDErZSzIId2soFJgZgAUJFedF3A8TCAqCBAnszIVSgYnwCvEiTNG4hNH3Tjj7dnPQH 1KL92/PlJTe1AoeAIQShCPFDABTAWYWCqlAASeLhBGj0BO94F4DgiP6rY+oMAycwq4T+BNzgPVpf sh0quwKGTY8GkJlx8G61MRsmpF0vaVSLqgYjpDgyxMGYTQdl4AeA0IZWeL3hHeBbZVPeDCEXqiGe pzA9UAQdAQV/SBqVxDF4MOGFcmsfDBJ/IImJOANPEwllYIVXLOAijlCFY1H/Mim70igokEImFAVI MAOOsAJH4AUjVEJrrWVlAHdsgHELsAIqwi0fTB+pAgOHANojfuTb5ChxQAhmsAgDYbukinSQcoIp Wjg1cJdo0BHhMYtcZAAtW+LkExgOsAaHEBKIgORoflBPMnCUsmju5AWGgAID90eM4E9v2wFadAer 0xHx0QWFADeFgywbNgJpXugzFRAAOw== ------=_NextPart_000_0024_01BFE7EC.289B7CC0-- From vxwexplo-errs@csg.lbl.gov Fri Jul 7 03:38:05 2000 From: "QinTing" Date: Fri Jul 7 03:38:07 PDT 2000 Subject: Re: Tornado: Help needed on WDB Configuration with one serial port Hi,Nagaraj Sorry, I think your method isn't achievable. If you read your BSP file bootConfig.c, you'll find this answer. In bootConfig.c, if you define a serial port as WDB_COMM_SERIAL channel, then this serial port will not be added to system's I/O systme (use tyDevCreate( )). In addition, the working function of WDB_COMM_SERIAL channel is different from normal tty channel. You can try to set the serial port as CONSOLE_TTY in BSP, and set the same channel as WDB_COMM_SERIAL in vxWorks. If it's wrong, please tell me. QT ----- Original Message ----- From: the vxWorks Users Group Exploder To: Sent: Friday, July 07, 2000 10:52 AM Subject: Tornado: Help needed on WDB Configuration with one serial port > Submitted-by vxwexplo-errs@csg.lbl.gov Thu Jul 6 19:42:37 2000 > Submitted-by: "G. Nagaraj" > > This is a multi-part message in MIME format. > > Hello all, > > I am a custom board with i386ex Processor. I have successfully ported the > ev386ex BSP to this board. I have some problems while configuring WDB. > > I have only one serial port. I have configured CONSOLE_TTY and WDB on the > same port. I wanted to know if this configuration possible. > > My aim is to get all the messages on the serial console until the WDB host > connects to the target. Is this achievable? > > Any help / pointers would be of great use to me. > > Thanks in advance, > > Regards, > Nagaraj > > > G. Nagaraj > Senior Software Engineer > Wipro Technologies > 37, Castle Street, Ashok Nagar, > Bangalore - 560 025, India > Phone: +91-80-536-7297 Ext: 315 > Fax: +91-80-557-6032 > www.wipro.com > > > From vxwexplo-errs@csg.lbl.gov Fri Jul 7 05:23:56 2000 From: Atkinson Andrew Date: Fri Jul 7 05:23:58 PDT 2000 Subject: SCSI Read/Write - Hi We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 respectively. We need to configure the SCSI interface to use an Ultra-SCSI device in synchronous mode for transfers >3.8MB/sec. The configuration of the SCSI interface and the device appears to be OK, but due to inexperience with SCSI, we require some assistance with actually reading/writing data. Can anyone point us in the right direction? Regards Andy Atkinson DERA Bincleaves -- The Information contained in this E-Mail and any subsequent correspondence is private and is intended solely for the intended recipient(s). For those other than the recipient any disclosure, copying, distribution, or any action taken or omitted to be taken in reliance on such information is prohibited and may be unlawful. From vxwexplo-errs@csg.lbl.gov Fri Jul 7 06:33:35 2000 From: ian.willats@telelogic.com Date: Fri Jul 7 06:33:37 PDT 2000 Subject: RE: SCSI Read/Write - Atkinson Andrew asked: > We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 > respectively. We need to configure the SCSI interface to use an > Ultra-SCSI device in synchronous mode for transfers >3.8MB/sec. Normally, the SCSI interface should be more or less self-configuring to operate at the most favourable data rates and transfer protocols supported by both the VxWorks system and your SCSI target device - so you shouldn't need to do anything by hand to achieve synchronous transfers. However, I think the Radstone rs60x BSP overrides this default and forces all data transfers to be asynchronous; you may need to make sure SCSI_NO_SYNC is undefined to restore normal service! > due to inexperience with SCSI, we require some assistance with actually > reading/writing data. Can anyone point us in the right direction? > If you're dealing with a disc-like device, the routines you should use are scsiRdSecs() and scsiWrtSecs(). Before that, you'll need to create a SCSI physical device using scsiPhysDevCreate() and a block device using scsiBlkDevCreate() - examples can probably be found in the VxWorks Programmer's Guide, I think. Otherwise, you will probably have to build your own SCSI command blocks and issue them using the scsiTransact() command. You'll need to know the exact format of the SCSI commands your target device supports; this should be part of the documentation for the device itself. Hope this helps. Feel free to get in touch if there's anything else. Ian Ian Willats Telelogic UK Ltd From vxwexplo-errs@csg.lbl.gov Tue Jul 11 01:33:31 2000 From: "amass" Date: Tue Jul 11 01:33:33 PDT 2000 Subject: SCSI_NO_SYNC This is a multi-part message in MIME format. ------=_NextPart_000_000B_01BFEB1A.FF799B60 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 respectively. We need to configure the SCSI interface to use an Ultra-SCSI2 device in synchronous mode for transfers >3.8MB/sec. We have defined SCSI_AUTO_CONFIG and when the target boots, the hard disk is correctly identified. However, as we wish to use the device in synchronous mode, we #undef SCSI_NO_SYNC. This has the effect of causing the scsiAutoConfig() routine, and consequently the rest of = the boot process, to hang. Why should this be so? What should the correct procedure be for enabling synchronous mode transfers? The hard disk details are: Vendor ID : SEAGATE Product ID : ST336704LN Rev : 02 Dev Type : 0 Blocks : 71687369 Block Size : 512 Regards Andy Atkinson DERA Bincleaves vxworks ------=_NextPart_000_000B_01BFEB1A.FF799B60 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi

We have a Radstone PPC1A / NT Host running = VxWorks=20 5.3.1 and T2
respectively. We need to configure the SCSI interface to = use=20 an
Ultra-SCSI2 device in synchronous mode for transfers=20 >3.8MB/sec.
 
We have defined SCSI_AUTO_CONFIG and when the target = boots,=20 the
hard disk is correctly identified. However, as we = wish to use=20 the device
in synchronous mode, we #undef SCSI_NO_SYNC. This = has the=20 effect
of causing the scsiAutoConfig() routine, and = consequently the=20 rest of the
boot process, to hang.
 
Why should this be so? What should the correct = procedure be=20 for enabling
synchronous mode transfers?
 
The hard disk details are:
 
    Vendor ID   =20 :    SEAGATE
    Product ID   =20 :    ST336704LN
    Rev    =    =20         = :    02
    Dev Type   =20     :    0
    Blocks    =20     :        = 71687369
    Block Size    =20 :    512
 
Regards
Andy Atkinson
DERA Bincleaves
vxworks
------=_NextPart_000_000B_01BFEB1A.FF799B60-- From vxwexplo-errs@csg.lbl.gov Tue Jul 11 07:28:09 2000 From: Sudarshan Ganapathi Date: Tue Jul 11 07:28:11 PDT 2000 Subject: Need information Hello, I am looking for vendor offering a solution which provides ATM/ADSL and associated drivers for VxWorks. This has to be a PCI plug-in adapter in a target board (running VxWorks). I am wondering whether the following could be a solution. a) a single PCI plug-in board offering ATM/ADSL. This plug-in board connected to PSTN directly. b) a plug-in PCI ATM adapter with an external UTOPIA (or such an interface). This will be connected to ADSL modem at the residential side. That is using 2 boards connected together using a flat cable. Could you please suggest a method to realize ATM/ADSL ? I appreciate your help and time. have a nice day, Sudarshan From vxwexplo-errs@csg.lbl.gov Wed Jul 12 04:03:23 2000 From: Vxworks Exploder Date: Wed Jul 12 04:03:25 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 12 04:03:19 PDT 2000 Subject: Re: C++ classes. Subject: Where is sysHwInit2() supposed to be called? Subject: Re: cache and 8260 Subject: VxWorks path separators or pathLib info needed Subject: Re: problem reconfiguring vxWorks Subject: Re: VxWorks flash driver question Subject: Re: What's the problem ? Subject: Remote gdb debugging Subject: (no subject) Subject: Use hardware serial port in simulator Subject: PowerPc /vxWorks Interrupt latency Subject: Re: what does exactly 'blr' do for POWRPC? Subject: Re: Remote gdb debugging Subject: SNTP Subject: Re: How to change the file system time Subject: Re: Replacing MBX860 with Pentium EBX Subject: Re: arp error? Subject: Re: software watchdog timer for MPC 860 Subject: Re: undefined symbol error Subject: ULIP driver for vxSim for Windows 2000 Subject: Re: Remote gdb debugging Subject: Re: Changing where vxWorks is loaded on Intel system. Subject: Re: Changing where vxWorks is loaded on Intel system. Subject: Re: Access to remote filesystem ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++ classes. Date: Mon, 3 Jul 2000 07:14:36 -0400 From: "Don Dewar" Message-ID: References: <395D3D37.6910771F@lucent.com> I inadvertently did the same thing myself, but did not have the same success that you did. My method (function) executed but bombed because it had no object context (valid this pointer). When a function is called from the shell all the arguments that are not filled in are passed in as zero. So, your call to fnA was probably called with a this pointer of 0 and the values of i and j were some offset from zero. What values did it print for i and j? Anyway, that is what I think happens in this case. - -- Don "Vividh Siddha" wrote in message news:395D3D37.6910771F@lucent.com... > I am a novice to vxWorks and had a rather intriguing question about the > way C++ classes are handled. I wrote a simple class as follows: > class classA{ > void fnA(void); > private: > int i, j; > }; > note: fnA prints the values of i and j; > > I compiled this into a object file and loaded on my target. On the > WindSh prompt I called fnA it actually printed the values of i and j; > But I have just the defination of the class. There is no object of that > class. > > I come from a Unix background and there you have a class defination and > then you allocate a obj by actually instantiating the obj of the class. > Its only after this memory is allocated for this obj. But how does it > work internally in vxWorks? > > Thanks gurus.... > --------------------------- Newsgroups: comp.os.vxworks Subject: Where is sysHwInit2() supposed to be called? Date: Mon, 03 Jul 2000 11:16:35 GMT From: mdierikx@industrie-nw.gti.nl Organization: EuroNet Internet Message-ID: <3960760a.354789540@news.euronet.nl> I am trying to integrate a driver for a multi-serial device into the kernel, and somehow don't get it working. I noticed that in sysSerialInit2() an interrupt routine is bound to the card's interrupt level; sysSerialInit2() is called by sysHwInit2(), but I cannot find any reference to sysHwInit2(). Where/when is it supposed to be called? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cache and 8260 Date: Fri, 23 Jun 2000 17:08:34 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8j05i4$9eo$1@nnrp1.deja.com> References: <0a0133f8.0ea32157@usw-ex0108-062.remarq.com> <8itgfe$bbd$1@nnrp1.deja.com> <210046ed.18506161@usw-ex0108-062.remarq.com> In article <210046ed.18506161@usw-ex0108-062.remarq.com>, mpiuze wrote: > sorry for the lack of info. > > Here's the low-down.. > > The board I'm running is a custom design. Right now, I > have my sysBatDesc[] tables set up to the 60x bus memory, 2 > local bus memories, and a flash memory partition on the 60x > bus. > > Here's an example of how the bats are setup for the 60x bus > memory: > /* IBAT 0 */ > ((LOCAL_MEM_LOCAL_ADRS & _MMU_UBAT_BEPI_MASK) | > _MMU_UBAT_BL_64M | > _MMU_UBAT_VS | _MMU_UBAT_VP), > ((LOCAL_MEM_LOCAL_ADRS & _MMU_LBAT_BRPN_MASK) | > _MMU_LBAT_PP_RW | > _MMU_LBAT_MEM_COHERENT | _MMU_LBAT_WRITE_THROUGH ), > > the local memories are set up the same with > _MMU_CACHE_INHIBIT selected instead of WRITE_THROUGH, and > the mem coherent option not set. Data bats are configured > identically to the IBats on the same memory regions, but no > 2 IBATS or DBATS overlap each other. > > We are using both the BAT registers and the page table. > the sysPhysMemDesc[] array. That array has definitions for > the 8260 internal memory map, and memory maps for registers > belonging to peripherals connected to the CPM. They do not > overlap with any BAT entries. Without boring you with the > whole table...here's an example: > > PHYS_MEM_DESC sysPhysMemDesc [] = > { > { > (void *) INTERNAL_MEM_MAP_ADDR, > (void *) INTERNAL_MEM_MAP_ADDR, > INTERNAL_MEM_MAP_SIZE, /* Internal Memory > Map */ > VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | > VM_STATE_MASK_CACHEABLE | > VM_STATE_MASK_GUARDED, > VM_STATE_VALID | VM_STATE_WRITABLE | > VM_STATE_CACHEABLE_NOT | > VM_STATE_GUARDED > }, > > all entries in the sysPhysMemDesc[] are configured the > same, with of course the obvious exception of the first 3 > fields, so that there is no overlap in PTEs. By changing > the WIMG options in the sysBatDesc[] I have caused vxWorks > to machine check before it even has a chance to display the > banner, but not get the machine checks to stop. I have > tried to set the cache to copy back but encountered the > same problem, by same meaning machine checks occuring at > different places in code. Machine checks are occurring in > primarily 2 areas - file manipulation through TFFS on the > flash and memory operations ( such as memPartAlloc's, > bfills, etc... ) sorry I can't be more specific on that, > but there's no one function that it stops in. With respect > to TFFS, I've stopped in MTD code, a few different dosFs > functions, flMap, fopen, .. the list just goes on .. and > when the crash occurs, it's not the first use of flash > files that our software makes after the MMU initialization > (just to make things seem a little more interesting). > > The vxWorks code is executing from the 60x bus memory at > the time of the crash, which is also where the stack and > heap space is located. Where are your network buffers ? Do you ever call cacheDmaAlloc(), or does anyone ? I was never able to get our system to run (correctly) using IBAT/DBAT for the 60x memory, I ended up using PTEs for that. Yes I know it sucks up a lot of memory for PTEs for 64MB, but it worked. If you call cacheDmaMalloc() to allocate network buffers (or anything else), it will want to mark *pages* as non-cacheable. Except if you've already covered memory with the BATs, I thought this may conflict and you'd get (potentially) undefined behavior, except the BATs have precedence, which may give undesirable behavior. I use PTEs for our 60x Bus memory, mark it cacheable, and set both USER_X_CACHE_MODE to copyback. Then it frees up a DBAT that I use for the internal dual-port RAM. Do you set the GBL bit in the rstate/tstate words of FCC parameter RAM ? Setting this causes me grief (i.e. machine checks etc...), so I explicitly do cache flushes/invalidates in the ethernet driver since the network buffers are cacheable (and copyback). hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks path separators or pathLib info needed Date: Fri, 23 Jun 2000 13:29:56 -0400 From: Susan LoVerso Message-ID: <39539E94.9EFF01BE@loverso.southborough.ma.us> I have an app that needs to do its own path parsing. VxWorks appears to accept both slash and backslash as path separators and they can be intermixed within a single path. I came across pathLib.h which has some potentially interesting functions in there that may help me. I cannot find *ANY* information on these functions anywhere. I looked in Windsurf, in all my docs, the online Tornado help and a general search for "pathlib" on Google. Nothing. Does anyone have any info on what the functions in pathLib.h do? I can use them, as I put a call to one of them in my usrAppInit.c. Barring that, can anyone point me to a list of path separators that VxWorks accepts? Is it more than slash and backslash? Thanks, Sue LoVerso --------------------------- Newsgroups: comp.os.vxworks Subject: Re: problem reconfiguring vxWorks Date: Mon, 3 Jul 2000 08:14:47 -0500 From: S. Johnson Organization: - Message-ID: <962630251.1444515743@news.inxpress.net> References: <8futu8$ufn$1@nnrp1.deja.com> Followup-To: comp.os.vxworks On Wed, 17 May 2000, chrisanc1967@my-deja.com wrote: >I'm working with a 3rd party BSP. When I try to reconfigure the vxWorks >image so that I can add some funcitonality I get an error. The build >output says that there is an undefined reference to 'm_freem' in >function 'dcOutput'. I can't see anything wrong with the header files >that contain the dcOutput function. Can anyone help me with this? >Thanks. > > >Sent via Deja.com http://www.deja.com/ >Before you buy. I've seen this with non-Sens versions of BSPs mixed with a BSP containing Sens. One of your libraries was updated when the Sens one was installed and the other does not know the symbol. The error probably occurs at the time of linking. I'd contact the BSP vendor and see if they provide a Sens version or re-install a VxWorks tree and install the BSP. SJ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks flash driver question Date: Mon, 3 Jul 2000 08:18:15 -0500 From: S. Johnson Organization: - Message-ID: <962630368.1300233686@news.inxpress.net> References: <8f6snv$cpu2@overload.lbl.gov> Followup-To: comp.os.vxworks On Tue, 02 May 2000, Vidhya Raman wrote: >Hi VxWorkers, > >I am trying to get the VxWorks flash driver (flashMem.c) to work >for am29F040 flash chip on a ads860 board. I am unable to write to >the flash memory. > >Is there something I am missing ? Do I have to "write enable" the >flash memory. > >The state machine for flash programming on am29F040 is correct in >the flashMem.c code. > >Thanks >Raman >________________________________________________________________________ >Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com If this socketed device is not connected for writes in hardware, there will be no way for you to write to the device. I'd check the schematics to see if it is connected as Read Only. SJ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: What's the problem ? Date: Mon, 03 Jul 2000 17:28:08 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <3960B107.A20596B9@bitband.com> References: <39605A8C.7D54D05D@csem.ch> Reply-To: leonid@bitband.com My 2 cents: Does your system exhibit a similar kind of jitter when the Cach is turned off ? Have you considered using WindView or an external hardware tool to analyze this ? What are the interrupt sources your system is receiving ? (keep in mind that ISR allways have higher priority then any or all tasks) Laurent Mealares wrote: > Hi, > I am working on a VME Rack with a CES PPC603 based board under VxWorks > 5.31. > As I use it for Industrial Control with a closed loop at 5 KHz, I have > very strong real time constraints. > My system is woken up by an Interrupt Service Routine every 200 microsec > (freq 5KHz). > I had to delete all debug/network tasks as they made my system misbehave > (too much latency). > I had to physically disconnect my Rack from the network too. > > This architecture works fine but periodically (every 200-300 millisec), > some task or ISR "steals" me about 50-60 micro sec > of my CPU time. I mean by that that my main loop lasts 50-60 microsec > longer than it should. > FYI, my mainloop has the highest priority. > The question is: > What kind of event could interrupt the loop? > > The only VxWorks active/pending tasks on my system are tShell and > tExctask. > All my tasks except my mainloop are pending on a semaphore. > > Has anybody faced such a problem? > Any clue will be helpful. > > Thanks in advance, > Laurent Mealares. --------------------------- Newsgroups: comp.os.vxworks Subject: Remote gdb debugging Date: Wed, 05 Jul 2000 15:24:07 GMT From: Alex Ward Organization: Deja.com - Before you buy. Message-ID: <8jvju2$buh$1@nnrp1.deja.com> Has anyone here done any remote source level debugging of a vxWorks target using gdb? Is this sort of debugging possible? I have seen a reference somewhere that it is/was possible to remotely debug code on a vxWorks platform over IP. The version of gdb seems to have been called vxgdb. There was reference to a library called librdb.a that I gather had to be linked to the target code. I have not been able to find any mention of this on the web. Has any one done this? TIA Alex Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: (no subject) Date: Wed, 05 Jul 2000 17:40:33 +0200 From: Urban Lindberg Organization: Ericsson Message-ID: <396356F1.EA25D63F@ks.ericsson.se> This is a multi-part message in MIME format. - --------------AE33E7970C9C3D2257B170DB Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit - --------------AE33E7970C9C3D2257B170DB Content-Type: text/plain; charset=us-ascii; name="EIN_SS7CC_TEST.c" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="EIN_SS7CC_TEST.c" #include #include #include #include #include #include #include #include #include "EIN_SS7CC_TEST_menu.h" #include "osdpn.h" static const char writepipe0[] = "Write on pipe[0]"; static const char writepipe1[] = "Write on pipe[1]"; typedef enum { EIN_FALSE, EIN_TRUE } BOOL_T; int change_board(void); int test_board(void); int config_board_default(void); int config_board_user(void); int show_configuration(void); int select_links(void); int toggle_loopback(void); int read_recieve_buffer(void); int set_emergency_mode(void); int clear_emergency_mode(void); int start_selected_links(void); int start_all_links(void); int stop_selected_links(void); int stop_all_links(void); int set_selected_lpo(void); int set_all_lpo(void); int set_selected_lporec(void); int set_all_lporec(void); int send_messages(void); int flush_selected_transmit_buffers(void); int flush_all_transmit_buffers(void); int flush_receive_buffer(void); int show_link_status(void); #define NO_OF_MENU_LEVELS 1 MENU_DATA_MENU item_1_1 = { "1:\tChange SS7CC", "Change to another SS7CC board", change_board }; MENU_DATA_MENU item_2_1 = { "2:\tBoard Test", "Test the SS7CC hardware!", test_board }; MENU_DATA_MENU item_3_1 = { "3:\tConfigure Board (default)", "Configure the SS&CC with a built-in configuration", config_board_default }; MENU_DATA_MENU item_4_1 = { "4:\tConfigure Board (user)", "Configure the SS&CC with a user-defined configuration", config_board_user }; MENU_DATA_MENU item_5_1 = { "5:\tShow Current Configuration", "", show_configuration }; MENU_DATA_MENU item_6_1 = { "6:\tSelect HSN's/SDL's to use", "", select_links }; MENU_DATA_MENU item_7_1 = { "7:\tSet PCM Loopback On/Off", "", toggle_loopback }; MENU_DATA_MENU item_8_1 = { "8:\tRead Receive Buffer", "", read_recieve_buffer }; MENU_DATA_MENU item_9_1 = { "9:\tSet Emergency", "", set_emergency_mode }; MENU_DATA_MENU item_10_1 = { "10:\tClear Emergency", "", clear_emergency_mode }; MENU_DATA_MENU item_11_1 = { "11:\tStart Selected HSN/SDL's", "", start_selected_links }; MENU_DATA_MENU item_12_1 = { "12:\tStart all HSN/SDL's", "", start_all_links }; MENU_DATA_MENU item_13_1 = { "13:\tStop Selected HSN/SDL's", "", stop_selected_links }; MENU_DATA_MENU item_14_1 = { "14:\tStop all HSN/SDL's", "", stop_all_links }; MENU_DATA_MENU item_15_1 = { "15:\tSet Selected HSN/SDL's to local processor outage mode", "", set_selected_lpo }; MENU_DATA_MENU item_16_1 = { "16:\tSet all HSN/SDL's to local processor outage mode", "", set_all_lpo }; MENU_DATA_MENU item_17_1 = { "17:\tSet Selected HSN/SDL's to local processor outage recovered mode", "", set_selected_lporec }; MENU_DATA_MENU item_18_1 = { "18:\tSet all HSN/SDL's to local processor outage recovered mode", "", set_all_lporec }; MENU_DATA_MENU item_19_1 = { "19:\tSend messages", "", send_messages }; MENU_DATA_MENU item_20_1 = { "20:\tFlush transmit buffers on selected HSN/SDL's", "", flush_selected_transmit_buffers }; MENU_DATA_MENU item_21_1 = { "21:\tFlush transmit buffers on all HSN/SDL's", "", flush_all_transmit_buffers }; MENU_DATA_MENU item_22_1 = { "22:\tFlush receive buffer", "Removes all data from the receive buffer", flush_receive_buffer }; MENU_DATA_MENU item_23_1 = { "23:\tShow Link Status", "Display Link status", show_link_status }; MENU_DATA_MENU item_24_1 = { "q:\tQuit", "Quit this program", NULL }; MENU_DATA_MENU *item_1[NO_OF_MENU_LEVELS] = { &item_1_1 }; MENU_DATA_MENU *item_2[NO_OF_MENU_LEVELS] = { &item_2_1 }; MENU_DATA_MENU *item_3[NO_OF_MENU_LEVELS] = { &item_3_1 }; MENU_DATA_MENU *item_4[NO_OF_MENU_LEVELS] = { &item_4_1 }; MENU_DATA_MENU *item_5[NO_OF_MENU_LEVELS] = { &item_5_1 }; MENU_DATA_MENU *item_6[NO_OF_MENU_LEVELS] = { &item_6_1 }; MENU_DATA_MENU *item_7[NO_OF_MENU_LEVELS] = { &item_7_1 }; MENU_DATA_MENU *item_8[NO_OF_MENU_LEVELS] = { &item_8_1 }; MENU_DATA_MENU *item_9[NO_OF_MENU_LEVELS] = { &item_9_1 }; MENU_DATA_MENU *item_10[NO_OF_MENU_LEVELS] = { &item_10_1 }; MENU_DATA_MENU *item_11[NO_OF_MENU_LEVELS] = { &item_11_1 }; MENU_DATA_MENU *item_12[NO_OF_MENU_LEVELS] = { &item_12_1 }; MENU_DATA_MENU *item_13[NO_OF_MENU_LEVELS] = { &item_13_1 }; MENU_DATA_MENU *item_14[NO_OF_MENU_LEVELS] = { &item_14_1 }; MENU_DATA_MENU *item_15[NO_OF_MENU_LEVELS] = { &item_15_1 }; MENU_DATA_MENU *item_16[NO_OF_MENU_LEVELS] = { &item_16_1 }; MENU_DATA_MENU *item_17[NO_OF_MENU_LEVELS] = { &item_17_1 }; MENU_DATA_MENU *item_18[NO_OF_MENU_LEVELS] = { &item_18_1 }; MENU_DATA_MENU *item_19[NO_OF_MENU_LEVELS] = { &item_19_1 }; MENU_DATA_MENU *item_20[NO_OF_MENU_LEVELS] = { &item_20_1 }; MENU_DATA_MENU *item_21[NO_OF_MENU_LEVELS] = { &item_21_1 }; MENU_DATA_MENU *item_22[NO_OF_MENU_LEVELS] = { &item_22_1 }; MENU_DATA_MENU *item_23[NO_OF_MENU_LEVELS] = { &item_23_1 }; MENU_DATA_MENU *item_24[NO_OF_MENU_LEVELS] = { &item_24_1 }; MENU_DATA_MENU **menu_data[] = { item_1, item_2, item_3, item_4, item_5, item_6, item_7, item_8, item_9, item_10, item_11, item_12, item_13, item_14, item_15, item_16, item_17, item_18, item_19, item_20, item_21, item_22, item_23, item_24 }; MENU_DATA menu = { "********** SS7CC TEST Main Menu ***********", 1, /* First line in menu highlighted at start */ 24, /* Number of items in menu */ 24, /* Number of lines to show in menu */ 0, menu_data }; extern WINDOW *window; int change_board(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In change_board"); refresh(); sleep(3); return 0; } int test_board(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In test_board"); refresh(); sleep(3); return 0; } int config_board_default(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In config_board_default"); refresh(); sleep(3); return 0; } int config_board_user(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In config_board_user"); refresh(); sleep(3); return 0; } int show_configuration(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In show_configuration"); refresh(); sleep(3); return 0; } int select_links(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In select_links"); refresh(); sleep(3); return 0; } int toggle_loopback(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In toggle_loopback"); refresh(); sleep(3); return 0; } int read_recieve_buffer(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In read_recieve_buffer"); refresh(); sleep(3); return 0; } int set_emergency_mode(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In set_emergency_mode"); refresh(); sleep(3); return 0; } int clear_emergency_mode(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In clear_emergency_mode"); refresh(); sleep(3); return 0; } int start_selected_links(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In start_selected_links"); refresh(); sleep(3); return 0; } int start_all_links(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In start_all_links"); refresh(); sleep(3); return 0; } int stop_selected_links(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In stop_selected_links"); refresh(); sleep(3); return 0; } int stop_all_links(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In stop_all_links"); refresh(); sleep(3); return 0; } int set_selected_lpo(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In set_selected_lpo"); refresh(); sleep(3); return 0; } int set_all_lpo(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In set_all_lpo"); refresh(); sleep(3); return 0; } int set_selected_lporec(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In set_selected_lporec"); refresh(); sleep(3); return 0; } int set_all_lporec(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In set_all_lporec"); refresh(); sleep(3); return 0; } int send_messages(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In send_messages"); refresh(); sleep(3); return 0; } int flush_selected_transmit_buffers(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In flush_selected_transmit_buffers"); refresh(); sleep(3); return 0; } int flush_all_transmit_buffers(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In flush_all_transmit_buffers"); refresh(); sleep(3); return 0; } int flush_receive_buffer(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In flush_receive_buffer"); refresh(); sleep(3); return 0; } int show_link_status(void) { WXY_PRINTF_BOLD_REV(window, 0, 0, "In show_link_status"); refresh(); sleep(3); return 0; } int comp_char(const void *a1, const void *a2) { return *(char *)a1 - *(char *)a2; } int main(int argc, char *argv[]) { BOOL_T monitor = EIN_FALSE; pid_t pid; int fd_a[2]; char *inbuf, *outbuf; FILE *mon_in; FILE *mon_out; INT_T i; DIR *dirp = NULL; struct dirent *dp = NULL; INT_T no_of_cards = 0; char cards_found[8] = {0, 0, 0, 0, 0, 0, 0, 0}; INT_T device_no = 0; char *selected_device = calloc(100, 1); for(i = 1; i < argc; i++) { if(argv[i][0] != '-') { ss7cc_test_showusage(); return EXIT_FAILURE; } switch(argv[i][1]) { case 'm': monitor = EIN_TRUE; break; default: ss7cc_test_showusage(); return EXIT_FAILURE; } } dirp = opendir(DEV_DIR); while ((dp = readdir(dirp)) != NULL && no_of_cards < ARRAYSIZE(cards_found)) { if (strncmp(dp->d_name, DEV_NAME,strlen(DEV_NAME)) == 0) { printf("Found a device: %s/%s\n", DEV_DIR, dp->d_name); device_no = dp->d_name[strlen(DEV_NAME)]; if(device_no >= '0' && device_no <= '7' && dp->d_name[strlen(DEV_NAME) + 1] == 0) { cards_found[no_of_cards] = device_no - '0'; no_of_cards++; } } } if(no_of_cards == 0) { printf("No devices found, aborting!\n"); printf("Searched for: \"%s/%s\"\n\r",DEV_DIR,DEV_NAME); closedir(dirp); return EXIT_FAILURE; } qsort(cards_found, no_of_cards, sizeof(cards_found[0]), comp_char); for(i = 0;i 0) { inbuf = calloc(100,1); outbuf = calloc(100,1); write(fd_a[0], writepipe0, sizeof(writepipe0) - 1); read(fd_a[0], inbuf, 99); /* WXY_PRINTF_BOLD1(window, 2, 10,"Message read from pipe[0]: \"%s\"", inbuf); */ execute_menu(&menu); fflush(NULL); sleep(3); sleep(3); free(inbuf); free(outbuf); return EXIT_SUCCESS; } if(pid < 0) { printf("In main program, but error calling fork!!!!\n"); return EXIT_FAILURE; } else { inbuf = calloc(100, 1); outbuf = calloc(100, 1); sleep(1); sprintf(outbuf, "In forked process!!!!\n"); /* printf(outbuf); */ read(fd_a[1], inbuf, 99); sprintf(outbuf,"Message read from pipe[1]: \"%s\"", inbuf); /* printf(outbuf); */ write(fd_a[1], writepipe1, sizeof(writepipe1)); free(inbuf); free(outbuf); return EXIT_SUCCESS; } } printf("Not monitor!!!!\n"); return EXIT_SUCCESS; } - --------------AE33E7970C9C3D2257B170DB Content-Type: text/plain; charset=us-ascii; name="EIN_SS7CC_TEST_menu.c" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="EIN_SS7CC_TEST_menu.c" #include #include #include "EIN_SS7CC_TEST_menu.h" void ss7cc_test_showusage(void) { printf("Showing usage here!!\n"); } WINDOW *window; INT_T execute_menu(MENU_DATA *menu) { CHAR_T in_key; INT_T i, max_x, max_y; int no_of_keys = 0; char *key_name; window = initscr(); start_color(); refresh(); init_pair(1,COLOR_BLACK,COLOR_WHITE); attron(COLOR_PAIR(1)); getmaxyx(window, max_y, max_x); for(i = 0; i < max_y; i++) WXY_PRINTF2(window,(int)i,0,"%*s",(int)max_x, ""); do { WXY_PRINTF_BOLD_REV(window, 0, 0, " "); /* placeholder for cursor (seems to hve disappeared) */ refresh(); getmaxyx(window, max_y, max_x); for(i = 0; i < max_y; i++) WXY_PRINTF2(window,(int)i,0,"%*s",(int)max_x, ""); WXY_PRINTF_BOLD(window, 0, 0, menu->title); if(menu->number_of_lines_to_show > menu->number_of_lines) menu->number_of_lines_to_show = menu->number_of_lines; if(menu->highlighted_line > menu->number_of_lines_to_show) menu->highlighted_line = menu->number_of_lines_to_show; if(menu->highlighted_line < 1) menu->highlighted_line = 1; WXY_PRINTF_BOLD_REV(window, max_y - 1, max_x - 1, " "); /* placeholder for cursor (seems to hve disappeared) */ for(i = 0; i < menu->number_of_lines_to_show; i++) { const char *text_ptr, *status_ptr; text_ptr = menu->menu_item[i][menu->menu_level]->menu_line_text; status_ptr = menu->menu_item[i][menu->menu_level]->menu_line_status_text; if(i + 1 == menu->highlighted_line) { WXY_PRINTF_BOLD_REV(window, i + 1, 0, text_ptr); WXY_PRINTF_BOLD_REV(window, menu->number_of_lines_to_show + 1, 0, status_ptr); } else WXY_PRINTF_BOLD(window, i + 1, 0, text_ptr); } noecho(); move(max_y-1,max_x-1); refresh(); in_key = getch(); if(in_key == '\33') { in_key = getch(); if(in_key == '\133') { in_key = getch(); if(in_key == '\101') menu->highlighted_line--; if(in_key == '\102') menu->highlighted_line++; } } else if(in_key == '\n' || in_key == '\r') { if(menu->menu_item[menu->highlighted_line-1][menu->menu_level]->menu_line_action != NULL) menu->menu_item[menu->highlighted_line-1][menu->menu_level]->menu_line_action(); else in_key = 'q'; /* fake a 'q' key press */ } #if 0 key_name = keyname(in_key); WXY_PRINTF_BOLD(window, 10, 0, key_name); WXY_PRINTF_BOLD1(window, 11 + (no_of_keys %4), 4, "%d ",(int)in_key); WXY_PRINTF_BOLD1(window, 11 + (no_of_keys %4), 0, "%d",no_of_keys); no_of_keys++; refresh(); sleep(1); #endif } while(in_key != 'q' && in_key != 'Q'); endwin(); return 0; } - --------------AE33E7970C9C3D2257B170DB Content-Type: image/x-xbitmap; name="EIN_SS7CC_TEST_menu.h" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="EIN_SS7CC_TEST_menu.h" #define WXY_PRINTF(w,x,y,str) {mvwprintw(w,x,y,str);} #define WXY_PRINTF1(w,x,y,str,a) {mvwprintw(w,x,y,str,a);} #define WXY_PRINTF2(w,x,y,str,a,b) {mvwprintw(w,x,y,str,a,b);} #define WXY_PRINTF3(w,x,y,str,a,b,c) {mvwprintw(w,x,y,str,a,b,c);} #define WXY_PRINTF4(w,x,y,str,a,b,c,d) {mvwprintw(w,x,y,str,a,b,c,d);} #define OUTSTR(str) {addstr(str);} #define WXY_PRINTF_BOLD(w,x,y,str) {attron(A_BOLD);mvwprintw(w,x,y,str);attroff(A_BOLD);} #define WXY_PRINTF_BOLD1(w,x,y,str,a) {attron(A_BOLD);mvwprintw(w,x,y,str,a);attroff(A_BOLD);} #define WXY_PRINTF_BOLD2(w,x,y,str,a,b) {attron(A_BOLD);mvwprintw(w,x,y,str,a,b);attroff(A_BOLD);} #define WXY_PRINTF_BOLD3(w,x,y,str,a,b,c) {attron(A_BOLD);mvwprintw(w,x,y,str,a,b,c);attroff(A_BOLD);} #define WXY_PRINTF_BOLD4(w,x,y,str,a,b,c,d) {attron(A_BOLD);mvwprintw(w,x,y,str,a,b,c,d);attroff(A_BOLD);} #define WXY_PRINTF_BOLD_REV(w,x,y,str) {attron(A_BOLD|A_REVERSE);mvwprintw(w,x,y,str);attroff(A_BOLD|A_REVERSE);} #define WXY_PRINTF_BOLD1_REV(w,x,y,str,a) {attron(A_BOLD|A_REVERSE);mvwprintw(w,x,y,str,a);attroff(A_BOLD|A_REVERSE);} typedef signed char CHAR_T; typedef int INT_T; typedef int (*INT_FUNC_PTR)(void); typedef struct { const char *menu_line_text; const char *menu_line_status_text; INT_FUNC_PTR menu_line_action; } MENU_DATA_MENU; typedef struct { const char *title; INT_T highlighted_line; INT_T number_of_lines; INT_T number_of_lines_to_show; INT_T menu_level; MENU_DATA_MENU ***menu_item; } MENU_DATA; void ss7cc_test_showusage(void); INT_T execute_menu(MENU_DATA *menu); - --------------AE33E7970C9C3D2257B170DB Content-Type: text/plain; charset=us-ascii; name="osdpn.h" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="osdpn.h" #if 0 #define DEV_DIR "/dev" #define DEV_NAME "e1t1driver01_" #else #define DEV_DIR "/home/qinxuli/PTU" #define DEV_NAME "e1t1driver01_" #endif #define ARRAYSIZE(a) ((sizeof(a))/(sizeof(a[0]))) - --------------AE33E7970C9C3D2257B170DB-- --------------------------- Newsgroups: comp.os.vxworks Subject: Use hardware serial port in simulator Date: Wed, 5 Jul 2000 17:48:51 +0200 From: "Henrik Lewander" Organization: Chalmers University of Technology Message-ID: <8jvlhj$hbl$1@nyheter.chalmers.se> Hello! I'm sitting here with my first project using VxWorks. My question is: can I connect to my PC:s serial port from inside VxSim? E g I want to use the serial port at the back of my windows-host from VxWorks running in Tornado. Is it possible? All this is new to me, please ask if something is unclear. Thanks, Henrik Lewander - -- ( http://henrik.lewander.com ][ icq:315542 ) mailto:henrik@lewander.com ][ Aprilia RS 250 --------------------------- Newsgroups: comp.os.vxworks Subject: PowerPc /vxWorks Interrupt latency Date: Wed, 05 Jul 2000 09:06:39 -0700 From: Philip Cardy Organization: Alcatel USA Inc. Message-ID: <39635D0F.9B7FC276@usa.alcatel.com> Reply-To: Philip.Cardy@usa.alcatel.com Hi, I've got a Power pc 850 target running at 50Mhz with a bus speed of 25Mhz. The Ram is 16bit SDRAM with 5 wait states. I'm running Instruction cache only. The issue is: It takes approx. 30us between receiving an external interrupt and getting into the ISR code (measured using a scope). Questions: i. Is this anywhere near typical for vxWorks on PPC? ii. Has anyone collected data regarding measued interrupt latency for PPC? iii. Can anything be done to improve this response time? Any suggestions would be appreciated. Thanks, Phil --------------------------- Newsgroups: comp.os.vxworks Subject: Re: what does exactly 'blr' do for POWRPC? Date: Wed, 05 Jul 2000 13:08:49 -0500 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <396379B1.7FCE694B@aps.anl.gov> References: <396367DC.3EDED79D@email.mot.com> Bo Yu wrote: > > I found the instruction 'blr' in romInit.s ( POWERPC ) but I cannot > figure out what it does? That's the PPC version of Return from Subroutine (blr = Branch Link Return). - - Andrew - -- Sharks kill 10 people each year. People kill 60,000,000 sharks each year. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Remote gdb debugging Date: 5 Jul 2000 19:17:13 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8k01jp$oh4$1@news1.xs4all.nl> References: <8jvju2$buh$1@nnrp1.deja.com> Alex Ward wrote: : I have seen a reference somewhere that it is/was possible to remotely : debug code on a vxWorks platform over IP. The version of gdb seems to : have been called vxgdb. vxgdb is part of the Tornado development enviroment and is not a seperate product. It used the WDB-protocol to talk to the target. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: SNTP Date: Wed, 05 Jul 2000 19:31:32 GMT From: djhystad@msn.com Organization: Deja.com - Before you buy. Message-ID: <8k02ek$mjn$1@nnrp1.deja.com> I have a project coming up where I will need to use SNTP (Simple Network Time Protocol) to syncronize messages being received from multiple sources. I see that the new TCP/IP stack supports this feature, so I called Wind River to get more information. Techinical support wouldn't answer my questions, saying that they were all marketing type questions and I should talk to my regional rep. I called my rep and found out he is a real a**hole who was only interested in finding out if my licensing was up to date (I found out I am up to date, but was so put off by his manner that I have no disire to toss any business his way). I hope you guys can help a bit. Has anyone used this product (formerly known as SENS, but now known as the VxWorks Network Stack)? What version of VxWorks do I need to run it? Is it an add-on product? If so, how much does it cost? Is it easy to use? Can I use it to get a network wide time value for time stamping real-time signal data? Thanks in advance, Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to change the file system time Date: Wed, 05 Jul 2000 13:34:41 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0c112e71.d499a050@usw-ex0102-013.remarq.com> References: <8jv14r$ubk$1@nnrp1.deja.com> <39631762.FE52A855@bitband.com> <8jvieg$amc$1@nnrp1.deja.com> Michael wrote: > have an RTC chip. How do I set system's calendar time? ansiTime library deals with conversions only, and retrieving of time. clock_settime uses time in seconds and not calendar time. Do I have to redefine time()? On bootup, use the conversion routines to go from RTC time to time in seconds since 1/1/70, then use clock_settime. After that, the routines time will all work. - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Replacing MBX860 with Pentium EBX Date: Fri, 07 Jul 2000 14:55:38 GMT From: Mary Frantz Organization: AT&T Worldnet Message-ID: <3965EF8A.5E56@worldnet.att.net> References: <396361A9.398B@worldnet.att.net> <396395F7.58A98345@lucent.com> Douglas Fraser wrote: > > Mary Frantz wrote: > > > > We have been using an MBX860 in our product and have been told by > > Motorola that they are discontuing the product. > > > > We are committed to the EBX form factor, and will probably replace it > > with a Pentium board on an EBX format. Does anybody have any > > recommendations? or ones to avoid? So far I have read a little about > > Versalogic, Ampro, and Adastra. Any experience with these? How are the > > BSPs? > > > > Thanks, > > Mary Frantz > > maryfrantz@worldnet.att.net > > This is news to us. Who passed that on > to you? > We also use the MBX860. Certainly, the > MPC8xx family > is continuing, but the loss of the MBX > board would be > a bummer. > > -- > Doug The MBX product manager at Motorola told us that they are switching to a Pentium processor for that platform. Someone else mentioned Embedded Planet supports an EBX form factor board with an MPC860, and I confirmed that they do. Does anybody have any experience with them? It's a new company (2 years old). (See www.embeddedplanet.com) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: arp error? Date: Tue, 11 Jul 2000 12:55:08 -0500 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <396B5F7C.DDC49640@aps.anl.gov> References: <396B2372.38B870E5@res.raytheon.com> cashl wrote: > > I'm getting the following output when I start my application. Any > insights as to what it means and what could be causing it? > > (tNetTask): arp info overwritten for c0a80a03 by 00:c0:8b:02:70:36 I don't know, but my guess is that it might be an indication that you have two ethernet devices that are advertising themselves as having the same IP address. Find out from your network admins if they know either of those two ethernet addresses and whether they're configured correctly. - - Andrew - -- Sharks kill 10 people each year. People kill 60,000,000 sharks each year. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: software watchdog timer for MPC 860 Date: 11 Jul 2000 14:38:20 -0400 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <8kfno4$g95$1@transfer.stratus.com> Sender: bpringle@DeadDuck >>>>> "Lei" == Lei writes: [snip] Lei> Does anyone try to do the same thing and have a work-around Lei> other than modifing the source code for inflate library? Please use DejaNews. This thread has been discussed here before. The basic idea is a timer interrupt that will kick the watchdog. However, the inflate source isn't difficult to modify; I would recommend that method. "http://www.info-zip.org/pub/infozip/zlib/" hth, Bill - -- Never marry the milkperson without gossiping first. As the Moon moves into the upper half of Virgo, be careful of your relationship with your child. Look for your teddy bear with one eye. Do something nice for a wallflower tomorrow. You'll once in a while be happy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: undefined symbol error Date: Tue, 11 Jul 2000 13:39:00 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0f725760.22256a75@usw-ex0101-007.remarq.com> References: <396ACB71.9077253@dso.org.sg> <396B2B25.6C5CEEAD@delphi.com> If the unidentified symbol is the name that you just typed it may be that your compiler has C++ decorated the symbol and the shell cannot find it. Try prefixing the string 'extern "C"' to the declaration of the entry point. Rob't Blumberg Embedded Software Developer Anything below is tacked on by Remarq! - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: ULIP driver for vxSim for Windows 2000 Date: Sun, 09 Jul 2000 01:30:35 GMT From: Spruce Technologies Organization: Deja.com - Before you buy. Message-ID: <8k8kjs$2mm$1@nnrp1.deja.com> Hi, do you know how to install ULIP driver for vxSim in Windows 2000? If you know then please post answer on this forum or send me email privately to jpzr@acm.org ! Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Remote gdb debugging Date: Wed, 12 Jul 2000 01:06:11 GMT From: Denis Organization: Deja.com - Before you buy. Message-ID: <8kgg9r$hob$1@nnrp1.deja.com> References: <8jvju2$buh$1@nnrp1.deja.com> <39645532.21CEA362@inxpress.net> Hello, I was just wondering if you still need to start Tornado to be able to use the newer/out of the box gdb? Can you somehow use the insight as the gui? and if yes, how do you tell gdb the ip of the host to which to connect? thanks. denis > Then I just launch Tornado I -or- Tornado II. It's no different than > sitting at my desk in the office, except I'm at home which is a big plus. > CrossWind (gbd) works without anything special, other than the -g flag > added to CFLAGS so debugging symbols are available. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing where vxWorks is loaded on Intel system. Date: Tue, 11 Jul 2000 18:36:24 -0700 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <396BCB98.A482E8C4@gv.net> References: Reply-To: mkesti@gv.net Burt Bicksler wrote: >Just checking to see if anyone has encountered any strange problems that >would prevent doing this? I've installed Tornado 2.0 on NT systems in various non-default directories and on mapped drives without difficulty. - -- ======================================================================== Michael Kesti | "And like, one and one don't make | two, one and one make one." mkesti@gv.net | - The Who, Bargain --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing where vxWorks is loaded on Intel system. Date: Wed, 12 Jul 2000 06:20:40 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8kh2nh$u8b$1@nnrp1.deja.com> References: <396BCB98.A482E8C4@gv.net> In article , Burt Bicksler wrote: > I guess I should have stated the question more clearly.. > > It's not the physical location of the various vxWorks files on any disk > system. Rather we want to change where the vxWorks loader and vxWorks > images wind up in memory on the target system. The default memory > mapping causes us to lose a large block of memory that our application > could make use of if we could reclaim it so it becomes part of the > contiguous heap. Might not be of any help, but you could use memAddToPool() to add the "lost" block to the heap. The heap doesn't need to be contiguous in VxWorks so blocks of memory that are not seen during initialisation can be added quite simply. Regards, John... > Thanks, > Burt Bicksler > > P.S. If responding via e-mail remove the obvious spam bait from the > return addres. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Access to remote filesystem Date: Wed, 12 Jul 2000 09:51:33 +0200 From: Emmanuel Organization: SEPRO Robotique Message-ID: <396C2385.D0C23411@sepro-robotique.com> References: <8kf402$rq5$1@news.hdpp.de> Udo Visser wrote: > 1.) Are there any other methods beside TFTP or NFS to access files on the host? Yes FTP TFTP is "trivial FTP", which means it is a reduced version (lighter) of the FTP protocol > 2.) Does anybody know a cheap (free) NFS-Server for Win-NT? take a look at www.softseek.com I'm not sure, but it should be possible to find such thing. An alternative solution could be a Linux (it's very easy to setup a NFS server with RedHat linux, for example). An unix box is always a good thing for network stuffs (FTP, DNS, route, NFS, etc...) > 3.) Is there a way to configure the memory consumption while using TFTP? probably not... regards Emmanuel. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 12 13:06:55 2000 From: "Tim Allen" Date: Wed Jul 12 13:06:57 PDT 2000 Subject: Re: Where is sysHwInit2() supposed to be called? >I am trying to integrate a driver for a multi-serial device into the >kernel, and somehow don't get it working. I noticed that in >sysSerialInit2() an interrupt routine is bound to the card's interrupt >level; sysSerialInit2() is called by sysHwInit2(), but I cannot find >any reference to sysHwInit2(). Where/when is it supposed to be called? You should be able to find sysHwInit2() in sysLib.c in your BSP directory. sysHwInit2() is called from sysClkConnect() in your system clock driver. The clock driver should be found in src/drv/timer, the actual file being dependent on your timer device. sysClkConnect() is called from usrRoot() in target/config/all/bootConfig.c and usrConfig.c for command line builds, or in target/config/comps/src/sysClkInit.c for project builds. Regards, Tim. -------------------- Tim Allen Ltd Real Time Embedded Software Consultancy vxWorks From vxwexplo-errs@csg.lbl.gov Wed Jul 12 23:31:04 2000 From: enrico.ferro@it.abb.com Date: Wed Jul 12 23:31:07 PDT 2000 Subject: Are there any italians here? Just out of curiosity, are there any Italians that use VxWorks connected to this mailing list? Enrico From vxwexplo-errs@csg.lbl.gov Thu Jul 13 01:04:37 2000 From: f.pertin@staubli.com Date: Thu Jul 13 01:04:39 PDT 2000 Subject: dynamic loading of object file on target Hello TORNADO VXWORKS guru, Does anybody knows how to synchronize target and host symbole table ?? My problem can be sumurized as follow: I have an application which contains symbols toto.out I download it with tornado. When I look at the interrsting symbols with a lkup I have it but with a @lkup I do not see nothing. I included the symbole table synchronization flag in the BSB I checked the synchronize symbole table on the target server Is there any thing else to do ???? Thanks in advance Francois From vxwexplo-errs@csg.lbl.gov Thu Jul 13 02:14:06 2000 From: Marco Raimondi Date: Thu Jul 13 02:14:08 PDT 2000 Subject: Italians using VxWorks Si, noi usiamo VxWorks! -- ==================================================== Marco Raimondi - Senior Software Engineer SIEMENS Information and Communication Networks Business Unit Radio - Equipment Software Development S.S. 11 Padana Superiore Km 158 20060 Cassina de' Pecchi (MI) - ITALY Tel:+39 02 4388 6026 Fax:+39 02 4388 6088 ===================================================== From vxwexplo-errs@csg.lbl.gov Thu Jul 13 04:01:28 2000 From: Jean-Marie Campiglia Date: Thu Jul 13 04:01:30 PDT 2000 Subject: Waking on LAN a remote machine from my vxWorks MTX603 board Hi all, I have a MTX603 board an I need to wake-up a machine remotely connected to my TCP/IP ethernet network. The remote machine has the "Wake on LAN" enabled. Does anybody knows how to build and send the "Magic packet" from my DEC21143 driver? (I have tried to use the link layer, i,e the AF_LINK flag in place of the usual AF_INET flag when creating a socket, but apparently the AF_LINK flag is not supported in the socket() interface). Thanks for any help, Jean-Marie From vxwexplo-errs@csg.lbl.gov Thu Jul 13 04:03:07 2000 From: Vxworks Exploder Date: Thu Jul 13 04:03:09 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 13 04:03:04 PDT 2000 Subject: TFTP on est8260 with option 0x80 Subject: Re: Access to remote filesystem Subject: Re: Tornado/vxWorks on Linux host Subject: UGL Drivers Subject: Ethernet Initialisation Subject: intermittant stack problem? Subject: Re: Target Server File System (TSFS) ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: TFTP on est8260 with option 0x80 Date: Sun, 9 Jul 2000 13:52:26 -0700 From: "Josh Williams" Organization: Posted via Supernews, http://www.supernews.com Message-ID: Hi, I'm running T2 with an est8260 eval board, and I'm having trouble when I set option 0x80 on target bootup to TFTP instead of FTP the vxWorks OS. It gets about 60% of the file and then decides that it is finished! Has anybody seen this before? Thanks, Josh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Access to remote filesystem Date: Wed, 12 Jul 2000 20:14:34 GMT From: "Adam" Organization: @Work Internet powered by @Home Network Message-ID: References: <8kf402$rq5$1@news.hdpp.de> Microsoft has a Windows NT services for Unix add-on pack that includes an NFS server. You can get more info at http://www.microsoft.com/NTServer/nts/exec/overview/sfu.asp - -Adam "Udo Visser" wrote in message news:8kf402$rq5$1@news.hdpp.de... > Hi there, > > due to several problems with file operations on our target (MPC860) we > intend to get rid of DosFS2.0, TFFS and PCMCIA. > What I have tried so far, is open/read/write/close a file on our host via > TFTP, which seems to be working OK. What I have not done, so far, is playing > with NFS, since our NT-Hosts dont support any NFS-Server functionality by > default. The problem with TFTP is, that it involves copying the files from > host to target and back (the name FTP implies that of course). So if you > open many large files simultaneously you will need quite a lot of RAM on > your target. Also, while working with an open file on the target anybody > else can open the same file on the host and write to it. > Three questions now: > 1.) Are there any other methods beside TFTP or NFS to access files on the > host? > 2.) Does anybody know a cheap (free) NFS-Server for Win-NT? > 3.) Is there a way to configure the memory consumption while using TFTP? > (Which would imply that TFTP does not always copy the complete file) > > thanks a lot > > Udo Visser > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado/vxWorks on Linux host Date: Mon, 10 Jul 2000 00:46:55 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: <329a5.35550$i5.416892@news1.frmt1.sfba.home.com> References: <20000709.14323300@stephen.johnson.com> Rumored to be in Beta testing on Tornado 3.0 now. Paul In article <20000709.14323300@stephen.johnson.com>, stevebj@inxpress.net wrote: >Will Tornado/vxWorks be available for a Linux host? > > >If so, when? > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: UGL Drivers Date: Wed, 12 Jul 2000 22:43:50 -0400 From: "Cybermaster" Organization: Randori News -- http://www.randori.com Message-ID: <_%9b5.4319$p2P2.37159170@news.randori.com> Hi, I just finished integration/porting of UGL driver for Raster Graphics RG101 PMC board. I had some minor struggle and but WindRiver's C&T 6555x UGL driver works pretty well so far. Before I move to next phase (Java GUI), I'd like to ask a question regarding UGL. Has anyone developed UGL drivers for non WindRiver/Zinc supported hardware/graphics chipset ? We are planning to buy/sub-contract/write to get UGL drivers for Permidia 3 Chipset or TI's C80. Any comment would be appreciated. Thanks. Yongsuk Choi Senior Software Engineer General Dynamics Land System Division choi@gdls.com --------------------------- Newsgroups: comp.os.vxworks Subject: Ethernet Initialisation Date: Mon, 10 Jul 2000 12:33:50 GMT From: mukeshm@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8kcfrd$h8r$1@nnrp1.deja.com> Hi I am working on an embedded software using VxWOrks. Could you please point me to the sequence of calls to initialise the ethernet driver? (I have only one ethernet driver) Thanks in advance Mukesh Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: intermittant stack problem? Date: Thu, 13 Jul 2000 09:56:44 +0100 From: Dave Moore Organization: BAE SYSTEMS Message-ID: <396D844C.1D920607@gecm.com> Hi Everyone, I've got a problem and it's driving me up the wall. My software runs fine most of the time, but every now and then, at an apparently random time, I get a 'data access' or some other exception error. Sometimes it'll run fine for days! and other times, it'll crash after 5 minutes. Using tt, I can see that this occurs in a piece of code that has been successfully run previously and therefore I can only assume that some sort of stack corruption is occurring. This crash always occurs on the same thread but I'm convinced there's nothing wrong with this code. I've checked the task stack usages for any overflows but these look fine. I've scoured the code many times, looking for places where code could accidentally write in memory areas it shouldn't (ie. on top of the stack) but can't fine anything. It's looking more like a 'system' type error to me. My code does use a number of interrupts and associatated ISRs and I'm wondering if an unfortunately collision of these could cause a stack corruption. The random nature of the crash might suggest this. Has anyone experienced anything similar?. Anyone got any ideas?. I don't have any patches currently installed - maybe this would make a difference?. Despite many attempts, I've failed to find a factor (like code download order, or stack size), which seems to affect the frequency of the problem - the fact that it can run ok for 2 or 3 days makes it pretty difficult to debug!!. I'm using motorola MCP750 PowerPC cards with Tornado II/vxWorks 5.3.1. All comments or suggestions are gratefully received. Thanks in advance. Dave --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Target Server File System (TSFS) Date: Thu, 13 Jul 2000 16:48:39 +0800 From: srdlm@sbell.com.cn (liuming) Organization: sbell Message-ID: References: hi; Your setup is alright.I think you can check the following : 1.maybe you can check your NT setup of COM1. It must be the same as the configuration in the Target Server setup. 2.And if you use RAS server, you must be sure that RAS is down. 3.And if you use COM1. you must check CONSOLE_TTY. It must not to be COM1 too. But I think by default vxworks set CONSOLE_TTY to COM1. So maybe you must change the CONSOLE_TTY in config.h to 1. regards eric In article , bhorowit@ic.EECS.Berkeley.EDU says... > Hi, > > I'm trying to use the Target Server File System in combination with a 486 > board sold by Compulab (www.compulab.co.il). I would like to use the > serial port (COM 1) of this target as the connection to my host. I know > the serial port works (I can see the output using HyperTerminal). > > I am able to boot a standalone version of VxWorks. I am also able to > interrupt the auto-boot process, and I have made sure that my boot > parameters are correct. > > However, when I boot my target (the version compiled to use TSFS), the > Target Server is unable to start, and prints the following error message: > > ----begin error message > tgtsvr (compulab-serial@gamow): Tue Jul 11 01:06:44 2000 > Wind River Systems Target Server: NT/Win95 version > Connecting to target agent... Error: clnttty_rcv: read timed out Error: clnttty_rcv: read timed out Error: clnttty_rcv: read timed out > Error: rpccore backend client Timed out > failed. > Connecting to target agent... Error: clnttty_rcv: read timed out Error: clnttty_rcv: read timed out Error: clnttty_rcv: read timed out > Error: rpccore backend client Timed out > failed. > Connecting to target agent... Error: clnttty_rcv: read timed out > > ...and so on > ----end error message > > Here are the details of my setup. I'm using Tornado 2.0 on NT. I'm using > the BSP provided by Compulab. I've done my best to configure Tornado to > use TSFS; this includes the following: > > 1. in the VxWorks view of my project, adding development tool components > > EDB agent components > WDB agent services > WDB target server file system. > > 2. using also development tool components > EDB agent components > select > WDB connection > WDB serial connection. > > 3. adding lines to my target's config.h file as described in Tornado 2.0 > User's Guide, p. 125. > > 4. in the VxWorks view of my project, using hardware > peripherals > > serial > SIO, and setting the parameters sensibly. > > 5. trying both defining and not defining INCLUDE_TSFS_BOOT_VIO_CONSOLE. > > Any help you could give to a Ph.D. student slaving away late at night > would be very much appreciated! > > Thanks, --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jul 13 04:26:42 2000 From: "amass" Date: Thu Jul 13 04:26:44 PDT 2000 Subject: ISR-Level Exceptions This is a multi-part message in MIME format. ------=_NextPart_000_0049_01BFECC5.9AAFFB00 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi How should one go about tackling an exception at ISR level? Andy Atkinson DERA Bincleaves vxworks ------=_NextPart_000_0049_01BFECC5.9AAFFB00 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
    How should one go about tackling = an=20 exception at ISR level?
 
Andy Atkinson
DERA Bincleaves
 
vxworks
------=_NextPart_000_0049_01BFECC5.9AAFFB00-- From vxwexplo-errs@csg.lbl.gov Thu Jul 13 08:53:10 2000 From: "Carolyn Briles" Date: Thu Jul 13 08:53:12 PDT 2000 Subject: memPartAlloc This is a multi-part message in MIME format. ------=_NextPart_000_0022_01BFECB0.2AB075B0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hello All, =20 I am using vxWorks on a Power PC. A simple histogram program is giving = the error =20 memPartAlloc:block too big - 524 in partition 0x196858 =20 when I run for more than 1.5M data events. The counter for events is = defined as a u_long, and the data acquisition iterates on a while loop = that compares datasofar to the counter. Both are u_long. =20 =20 This seems simple, but I'm at a loss. Increasing the stack size of the = task does not help increase the number of runs possible. Has anyone run = into this? =20 Many thanks for your help! =20 Carolyn=20 ------=_NextPart_000_0022_01BFECB0.2AB075B0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello All,
 
I am using vxWorks on a Power = PC.  A simple=20 histogram program is giving the error
 
    = memPartAlloc:block too big -=20 524 in partition 0x196858
 
when I run for more than 1.5M data = events. =20 The counter for events is defined as a u_long, and the data acquisition = iterates=20 on a while loop that compares datasofar to the counter.  Both are=20 u_long.  
 
This seems simple, but I'm at a = loss. =20 Increasing the stack size of the task does not help increase the number = of runs=20 possible.  Has anyone run into this?
 
Many thanks for your = help!
 
Carolyn 
------=_NextPart_000_0022_01BFECB0.2AB075B0-- From vxwexplo-errs@csg.lbl.gov Fri Jul 14 07:51:29 2000 From: "amass" Date: Fri Jul 14 07:51:31 PDT 2000 Subject: scsiPhysDevCreate() This is a multi-part message in MIME format. ------=_NextPart_000_0007_01BFEDAB.72DC6460 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi Can anyone tell me what (in more detail than the manual provides) = what scsiPhysDevCreate() does on a Radstone PPC1A fitted with a NCR5930 = SCSI controller. This information is required as we are discussing a = problem with a hard dsic manufacturer who may not be familiar with = VxWorks and associated functions. Thanks Andy Atkinson ------=_NextPart_000_0007_01BFEDAB.72DC6460 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
    Can anyone tell me what (in more = detail=20 than the manual provides) what scsiPhysDevCreate() does on a Radstone = PPC1A=20 fitted with a NCR5930 SCSI controller. This information is required as = we are=20 discussing a problem with a hard dsic manufacturer who may = not be=20 familiar with VxWorks and associated functions.
 
Thanks
 
 
Andy Atkinson
------=_NextPart_000_0007_01BFEDAB.72DC6460-- From vxwexplo-errs@csg.lbl.gov Sun Jul 16 21:08:07 2000 From: "Madhusudhan K" Date: Sun Jul 16 21:08:10 PDT 2000 subscribe vxworks_users ________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com From vxwexplo-errs@csg.lbl.gov Sun Jul 16 23:31:21 2000 From: "vishal dalal" Date: Sun Jul 16 23:31:23 PDT 2000 Subject: Fw: Info required urgently This is a multi-part message in MIME format. ------=_NextPart_000_00CF_01BFEFE6.8B121D80 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear colleague, Please see the following message. Thanks and with regards Vishal Dalal -----Original Message----- From: Alan K Biocca To: Vishal Dalal Date: Monday, July 17, 2000 7:01 AM Subject: Re: Info required urgently Your request should be addressed to the vxworks users group email = exploder, or comp.os.vxworks usenet newsgroup. submit articles for the vx exploder list via vxwexplo@lbl.gov. -- Alan K Biocca At 05:40 AM 7/15/00, you wrote: Dear Colleague,=20 Hi! We are working at Silicon=20 Automation Systems, Bangalore, India on 3g wireless handset. We = already=20 have protocol stack and we now want to implement it on physical = layer=20 (board). We are searching for suitable RTOS to run protocol stack. = We=20 are:=20 =20 1. Using Strong Arm processor=20 2. Using Universal Serial Bus (USB)=20 3. Using Ethernet card and TCP/IP=20 =20 We have option to use NUCLEUS RTOS or VXWORK or ECOS. But = we=20 have received positive response for VxWorks (from our customers).=20 Please let me know the details of VxWorks RTOS with=20 reference to our requirements like=20 =20 1. Is it portable to strong ARM (SA-1110) processor?=20 2. Does it support device drivers for USB/ETHERNET/TCP-IP?=20 3. How much is the cost of VxWorks RTOS (license fee etc..)?=20 4. What are the development tools available?=20 5. Time to ship?=20 =20 We are already going through your web site but = because of=20 the pausity of time and shortage of time to decide RTOS we are = troubling=20 you in advance. Please let me know about the details for the above = as=20 soon as possible at vishald@sasi.com=20 =20 With regards=20 Vishal Dalal=20 =20 --=20 ############################################################ Silicon Automation Systems Limited,Hal-III Stage,Bangalore=20 Tel: 91-80-5281229,5276100 ext:4174 Fax:91-80-5284396 =20 e-mail:dalal_vishal@hotmail.com =20 Home Page: http://www.angelfire.com/ab3/vishal ############################################################ =20 ------=_NextPart_000_00CF_01BFEFE6.8B121D80 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Dear colleague,
       Please see the = following=20 message.
Thanks and with regards
Vishal Dalal
-----Original = Message-----
From:=20 Alan K Biocca <AKBiocca@lbl.gov>
To: = Vishal=20 Dalal <vishald@sasi.com>
Date:=20 Monday, July 17, 2000 7:01 AM
Subject: Re: Info required=20 urgently

Your request should be addressed to the = vxworks=20 users group email exploder, or comp.os.vxworks usenet = newsgroup.

submit=20 articles for the vx exploder list via vxwexplo@lbl.gov.

-- Alan K = Biocca


At 05:40 AM 7/15/00, you wrote:
Dear Colleague,=20 =
           &nb= sp;           &nbs= p;         =20 Hi! We are working at Silicon
Automation Systems, Bangalore, = India on 3g=20 wireless handset. We already
have protocol stack and we now want = to=20 implement it on physical layer
(board). We are searching for = suitable=20 RTOS to run protocol stack. We
are:

1. Using Strong Arm=20 processor
2.  Using Universal Serial Bus (USB)
3. Using = Ethernet card and TCP/IP=20

          We = have=20 option to use NUCLEUS RTOS or VXWORK or ECOS. But we
have = received=20 positive response for VxWorks (from our customers).=20 =
           &nb= sp;=20 Please let me know the details of VxWorks RTOS with
reference to = our=20 requirements like

1. Is it portable to strong ARM (SA-1110)=20 processor?
2. Does it support device drivers for = USB/ETHERNET/TCP-IP?=20
3. How much is the cost of VxWorks RTOS (license fee etc..)? =
4. What=20 are the development tools available?
5. Time to ship?=20 =

           = ;   =20 We are already going through your web site but because of
the = pausity of=20 time and shortage of time to decide RTOS we are troubling
you in = advance. Please let me know about the details for the above as =
soon as=20 possible at        = vishald@sasi.com=20

With regards
Vishal Dalal 
--=20
############################################################
Silicon Automation Systems Limited,Hal-III Stage,Bangalore=20
Tel: 91-80-5281229,5276100 ext:4174
Fax:91-80-5284396         &n=
bsp;           &nb=
sp;   =20
e-mail:dalal_vishal@hotmail.com       =
          =20
Home Page:
http://www.angelfire.com/ab3=
/vishal
############################################################
  =
------=_NextPart_000_00CF_01BFEFE6.8B121D80-- From vxwexplo-errs@csg.lbl.gov Mon Jul 17 00:07:04 2000 From: ian.willats@telelogic.com Date: Mon Jul 17 00:07:06 PDT 2000 Subject: RE: scsiPhysDevCreate() Andy Atkinson asked: > Can anyone tell me what (in more detail than the manual provides) > what scsiPhysDevCreate() does on a Radstone PPC1A fitted with a NCR5930 > SCSI controller. This information is required as we are discussing a > problem with a hard dsic manufacturer who may not be familiar with > VxWorks and associated functions. > What it does in the standard scsi2Lib (as far as the SCSI target is concerned) is roughly as follows: - issue a REQUEST SENSE command asking for up to and including the additional length byte - parse the returned sense data to determine the sense data length, and allocate a buffer as required - issue an INQUIRY command, retry if necessary until it succeeds - parse the inquiry data to determine if the target supports tags, wide and/or synchronous transfer (following commands will use these features if supported) - if it's a disc device, issue a READ CAPACITY command I can't vouch for whether Radstone's driver does anything different, but it's unlikely. Ian Ian Willats Telelogic UK Ltd From vxwexplo-errs@csg.lbl.gov Mon Jul 17 05:17:32 2000 From: "Saskia Zezima" Date: Mon Jul 17 05:17:34 PDT 2000 Subject: Bootable Project on MCP750 This is a multi-part message in MIME format. ------=_NextPart_000_0049_01BFEFC7.6EB2B880 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, Has anyone built a Bootable Project in Tornado2 and been able to = download and the VxWorks Image? I can build the VxWorks image in the BSP directory and download it, but = when I create a Bootable based off of the BSP, it loads and then never = runs.=20 Any ideas? Saskia ------=_NextPart_000_0049_01BFEFC7.6EB2B880 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi,
 
Has anyone built a Bootable Project in = Tornado2 and=20 been able to download and the VxWorks Image?
 
I can build the VxWorks image in the = BSP directory=20 and download it, but when I create a Bootable based off of the BSP, it = loads and=20 then never runs.
 
Any ideas?
 
Saskia
------=_NextPart_000_0049_01BFEFC7.6EB2B880-- From vxwexplo-errs@csg.lbl.gov Mon Jul 17 07:44:06 2000 From: "Wang, Tom (HT-EX)" Date: Mon Jul 17 07:44:09 PDT 2000 Subject: Any useful Tool/Method to debug memory leak ? 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_01BFEFFB.C3953AB0 Content-Type: text/plain Hi VxWorks/Tornado user, Has anyone had a good experience to track down a memory leak problem with tools or methods ? Appreciate if you can share your success story. We are using the Tornado 1.0.1 and VxWorks 5.3.1 on the NT platform for the R3000 target. I used the "CodeTest memory tool" to add instrumentation to the source code for memory leak. The CodeTest memory test did not show any clear evidence of memory leak, but the "memShow" did reveal memory leak of 384 bytes in 4 blocks. The CodeTest only add instrumentation to the source code compiled with its driver. This give us confidence that the leak is not in our code but in third party's object code or library. There is another type of tool which works in lower level, and it replaces memory allocation routines with debugging routine. The "MemScope" of the Performance pack is the another tool that I would like to use. Has anyone had a success experience on this tool ? Does it work correctly and provide the trace for the memory leak on the Tornado 1.0.1 on the NT platform ? If the tool is not available or not working correctly, We are planning to resolve this memory leak by the following approach. Please comment if this is (or not) a sound approach. 1. Create our memory allocation routine to replace the memory routine for debugging by using symFindByName() to locate the old routine and remap the memory allocation routine with our routines. 2. Memory Heap trace (Manual or tool to locate the leaked blocks): Use "memShow 1" to display the allocated and free blocks and capture the memory help before and after the memory leak test. This would need the knowledge of memory block format. Has anyone had tried this approach ? Is this a good approach ? TIA. -Tom Wang Broadband Communications Sector Motorola Tel: 215-323-2102 (o) Email: towang@gi.com ------_=_NextPart_001_01BFEFFB.C3953AB0 Content-Type: text/html Content-Transfer-Encoding: quoted-printable Any useful Tool/Method to debug memory leak ?

Hi VxWorks/Tornado user,

Has anyone had a good experience to = track down a memory leak problem with tools or methods ?  = Appreciate if you can share your success story.

We are using the Tornado 1.0.1 and = VxWorks 5.3.1 on the NT platform for the R3000 target.

I used the "CodeTest memory = tool" to add instrumentation to the source code for memory = leak.  The CodeTest memory test did not show any clear evidence of = memory leak, but the "memShow" did reveal memory leak of 384 = bytes in 4 blocks.  The CodeTest only add instrumentation to the = source code compiled with its driver.  This give us confidence = that the leak is not in our code but in third party's object code or = library.

There is another type of tool which = works in lower level, and it replaces memory allocation routines with = debugging routine. The "MemScope" of the Performance pack is = the another tool that I would like to use.  Has anyone had a = success experience on this tool ?  Does it work correctly and = provide the trace for the memory leak on the Tornado 1.0.1 on the NT = platform ?

If the tool is not available or not = working correctly, We are planning to resolve this memory leak by the = following approach. Please comment if this is (or not) a sound = approach.

1. Create our memory allocation = routine to replace the memory routine for debugging by using = symFindByName() to locate the old routine and remap the memory = allocation routine with our routines. 

2. Memory Heap trace (Manual or tool = to locate the leaked blocks):
Use "memShow 1" to display the allocated and free blocks and = capture the memory help before and after the memory leak = test.

This would need the knowledge of = memory block format.  Has anyone had tried this approach ?  = Is this a good approach ?

TIA.

-Tom = Wang
Broadband Communications Sector
Motorola
Tel: = 215-323-2102 (o)
Email: = towang@gi.com

------_=_NextPart_001_01BFEFFB.C3953AB0-- From vxwexplo-errs@csg.lbl.gov Tue Jul 18 00:12:14 2000 From: "Michael Lawnick" Date: Tue Jul 18 00:12:17 PDT 2000 Subject: Re: Bootable Project on MCP750 This has been discussed previous twice: all PPC-vxWorks images of TII won't run, if built via elfToBin (bug!). Ask your FAE for the patch of this file or burn via serial line/hex-file. > Has anyone built a Bootable Project in Tornado2 and been able to = > download and the VxWorks Image? > > I can build the VxWorks image in the BSP directory and download it, but = > when I create a Bootable based off of the BSP, it loads and then never = > runs.=20 > > Any ideas? > > Saskia 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 Tue Jul 18 00:20:09 2000 From: "Elvis John Dowson" Date: Tue Jul 18 00:20:11 PDT 2000 Subject: VxWorks Hi Vishal, I can answer a few quesions for you regarding VxWorks : >> 1. Is it portable to strong ARM (SA-1110) processor? Yes, VxWorks Tornado 2 supports the StrongARM-1100 processor. The CPU type is defined by setting the preprocessor variable CPU to ARMSA110 (which targets the StrongARM-110 and StrongARM-1100 processors) >> 2. Does it support device drivers for USB/ETHERNET/TCP-IP? RE: USB ------- Well, you're in luck Vishal, there is a device driver available in VxWorks for the SA-1100 UARTs. The SA-1100 chip contains 5 serial ports, but only ports 1 and 3 are usable as UARTs. The others support Universal Serial Bus (USB), SDLC, IrDA Infrared Communications Port (ICP) and Multimedia Communications Port (MCP)/Synchronous Serial Port (SSP). The following routines are used to control the SA-1100 UART tty driver : sa1100DevInit( ) - initialise an SA1100 channel sa1100Int( ) - handle an interrupt The driver is typically only called by the BSP. RE: Ethernet ------------ Ethernet NIC drivers exist for Intel 82557, 82596; 3COM 3C90xB XL, 3C509; DEC 21x4x; SMC 8013WC; and a couple of others from AMD, Mototola and National Semiconductor and Fujitsu. RE: TCP/IP ----------- Yes, the standard VxWorks network stack includes support for the TCP/IP protocol. Here is a complete list of protocol and utilities supported by VxWorks : CSLIP,PPP,IP, UDP, TCP, DHCP, BOOTP, DNS, ARP and Proxy ARP OSPF, RIP, Sockets (TCP, UPD, multicasting, routing, and Zbuf) RPC, RSH, FTP, NFS, TFTP, rlogin, and telnet. >> 4. What are the development tools available? It has a browser, a shell, a debugger, a source code navigator. Apart from the basic tools, they also have a list of optional tools that can be integrated into the Tornado environment, for example : the Stethoscope tool for graphical timing anlysis. The Tornado development environment, in my opinion, is one of the most comprehensive and feature rich environments available for any RTOS. >> 5. Time to ship? You can usually have a full fledged evaluation copy running at your site and begin working as soon as you place an order. I think that the India office always keeps a couple of copies in stock, so you should have no problem obtaining a copy. Best regards, Elvis John Dowson Project Leader Simulation Systems and Systems Integration BAeHAL Software Limited Airport Lane, HAL Estate Airport Road, Bangalore - 560 017 Karnataka, India Tel : +91-80-526 8683, 526 8684 Fax : +91-80-527 0915 Mobile : +91-98450-00884 Email : elvis@baehal.com, elvis@dowsonresearch.com Website : www.baehal.com From vxwexplo-errs@csg.lbl.gov Tue Jul 18 07:29:25 2000 From: "G. Nagaraj" Date: Tue Jul 18 07:29:27 PDT 2000 Subject: VXWORKS: ISR does not get called again This is a multi-part message in MIME format. ------=_NextPart_000_0014_01BFF0F3.C7379110 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello VXWorkers, I am using the EV386EX BSP. I have an ISR to handle my hardware interrupt (IRQ 1). I connect the ISR using "intConnect" and enable the interrupt by using "sysIntEnablePIC". My ISR gets executed only once after which it does not get called. I use "semGive" and "logMsg" in the ISR. I tried giving explicit "sysIntEOI" also. But no use. It still hangs. I hope there will be lot of people on this list who can help me. Thanks in advance. Best Regards, G. Nagaraj Senior Software Engineer Wipro Technologies 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.wipro.com ------=_NextPart_000_0014_01BFF0F3.C7379110 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello=20 VXWorkers,
 
I am = using the=20 EV386EX BSP. I have an ISR to handle my hardware interrupt (IRQ 1). I = connect=20 the ISR using "intConnect" and enable the interrupt by using = "sysIntEnablePIC".=20 My ISR gets executed only once after which it does not get called. I use = "semGive" and "logMsg" in the ISR. I tried giving explicit "sysIntEOI" = also. But=20 no use. It still hangs.
 
I hope = there will be=20 lot of people on this list who can help me.
 
Thanks = in=20 advance.
 
Best=20 Regards,

 

G. Nagaraj=20
Senior=20 Software Engineer=20
Wipro=20 Technologies=20
37, = Castle Street,=20 Ashok Nagar,=20
Bangalore - 560 025,=20 India =
Phone: +91-80-536-7297 = Ext:=20 315=20
Fax: = +91-80-557-6032=20
www.wipro.com=20

 
------=_NextPart_000_0014_01BFF0F3.C7379110-- From vxwexplo-errs@csg.lbl.gov Tue Jul 18 08:25:23 2000 From: "amass" Date: Tue Jul 18 08:25:25 PDT 2000 Subject: SCSI Minimum Period This is a multi-part message in MIME format. ------=_NextPart_000_001B_01BFF0D3.66DD2D40 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi SCSI again I'm afraid!!!!!! We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 respectively. We need to configure the SCSI interface to use an Ultra-SCSI2 device in synchronous mode. We need clarification on the setting of 'minPeriod' in the SCSI_OPTIONS structure when setting the transfer options for the interface. Do we: a) Specifiy the minimum period factor (as specified by the SCSI-2 standard) and rely on the SCSI system to multiply by 4 to give = the appropriate delay; or b) Specify the actual minimum period required in nano seconds for the desired transfer rate. We wish to use our interface at 20M transfers/sec and have tried setting 'minPeriod' to 12 (following 'a' above), but this was not successful. We = have also tried setting 'minPeriod' to 50 nanoseconds (following 'b' above), = and this appeared to be more successful. However, I would still appreciate someone 'in the know' telling us which = is the correct solution. Regards Andy Atkinson Software Engineer DERA BIncleaves ------=_NextPart_000_001B_01BFF0D3.66DD2D40 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
SCSI again I'm afraid!!!!!!

We have a = Radstone PPC1A /=20 NT Host running VxWorks 5.3.1 and T2
respectively. We need to = configure the=20 SCSI interface to use an
Ultra-SCSI2 device in synchronous = mode.
 
We need clarification on the setting of 'minPeriod' = in the=20 SCSI_OPTIONS
structure when setting the transfer options for the = interface.=20 Do we:
 
    a) Specifiy the minimum period = factor (as=20 specified by the SCSI-2
        standard) and = rely on=20 the SCSI system to multiply by 4 to give the
        appropriate=20 delay;
 
    or
 
    b) Specify the actual minimum = period=20 required in nano seconds for
        the desired = transfer=20 rate.
 
We wish to use our interface at 20M transfers/sec = and have=20 tried setting
'minPeriod' to 12 (following 'a' above), but this = was not=20 successful. We have
also tried setting 'minPeriod' to 50 nanoseconds = (following=20 'b' above), and this
appeared to be more successful.
 
However, I would still appreciate someone 'in the = know'=20 telling us which is the
correct solution.
 
Regards
 
 
Andy Atkinson
Software Engineer
DERA BIncleaves
 
------=_NextPart_000_001B_01BFF0D3.66DD2D40-- From vxwexplo-errs@csg.lbl.gov Tue Jul 18 09:38:39 2000 From: "Trevor Getty" Date: Tue Jul 18 09:38:41 PDT 2000 Subject: Re: VXWORKS: ISR does not get called again It appears that your interrupt has been hooked up correctly - or your first interrupt wouldn't happen. 1) Before hooking up the interrupt clear it from the PIC to avoid nasty interrupt spoiling things whilst you are hooking your ISR up, it may also be helpful to clear the interrupt cause register etc. on whatever would be generating the interrupt just in case. 2) Connect up your ISR to the intline. 3) Enable the interrupt. Example. 1) sysIntDisablePIC(intline) 2) intConnect(INUM_TO_IVEC(INT_VEC_IRQ0 + intline), ISR, param) 3) sysIntEnablePIC(intline) within the ISR: disable the PIC to mask off any further interrupt happening on this line until the present one has been dealt with. deal with the interrupt cause and clear the cause register if appropriate. then enable the PIC again before exiting the ISR. On page 90 of the VxWorks T2 programmes guide there is a list of exceptions of what cannot be done within an ISR - you can give semaphores but not mutual exclusion semaphores, and VxMp ones. - check this. Hope thats of some help. --------------------------------------------------- Trevor Getty Software Engineer (R & D) Post Impressions (Systems) Ltd. Tel: (01635) 232624 Fax: (01635) 232638 trevor.getty@postimpressions.com From vxwexplo-errs@csg.lbl.gov Tue Jul 18 10:50:46 2000 From: "Baik, John" Date: Tue Jul 18 10:50:48 PDT 2000 Subject: I can't include some UGL components after Upgrade to Tornado 2.01 Can anyone tell the reason I can't include some components ? I just got Tornado 2.01 for ARM (actually StrongARM SA1100). I did change BSP code as Release Note. I could make Bootrom and work fine. I had an error message when I try to include 'sa1100 diaply driver' and 'sa1100 Touch screen driver' into Vxworks image like as " Component 'sa1100 display driver' is not installed. You may need to purchase and /or install it". But It worked fine with Old version. Any idea ? Jonghyeon Baik Sr. S/W Engineer Applied Data Systems 301- 490-4007 x15 http://www.flatpanels.com From vxwexplo-errs@csg.lbl.gov Tue Jul 18 11:02:16 2000 From: Tom Miller Date: Tue Jul 18 11:02:18 PDT 2000 Subject: network load of vxworks using intel 82559 Hello Tornado Group! I am in the very first stages of a new vxworks project. I am trying to get a celeron processor based system to load vxworks over a network. I am booting from a boot disk. I was wondering: 1) How do I include the driver for the 82559 on the boot disk so that the system can control the ethernet device to fetch the vxWorks image? 2)When I am specifying the boot device (either in the DEFAULT_BOOT_LINE or by interrupting the boot process and manually entering the parameters) what do I enter in the Boot Device field? For example in a system using the crystal cs8900 ethernet chip, I enter "cs0" as the boot device. If you can answer either of these questions (with whatever brevity you want) I would very much appreciate it. Thank You, Tom Miller Motion Engineering Inc. Santa Barbara Ca. From vxwexplo-errs@csg.lbl.gov Wed Jul 19 03:22:57 2000 From: "G. Nagaraj" Date: Wed Jul 19 03:23:00 PDT 2000 Subject: RE: VXWORKS: ISR does not get called again Thanks Trevor, I was trying to trace the cause of the problem. I found it something to do with the IRQ1. In a PC386 system, IRQ1 is used for Keyboard and is connected automatically by the kernel. Actually I could see the intConnect for the same in the PC386BSP's sysLib.c code. But I am using EV386EX BSP which does not even have a keyboard Interrupt definition also. I just tried to disable the keyboard interrupt routine by un-defining the keyboard interrupt definitions. And to my surprise, it works. I don't know why WRS has done a BSP without proper definitions and documentation. It is possible that I am wrong. This info might be given in the BSP porting kit, which I have not yet seen. Thanks again for your help. Best regards, Nagaraj -----Original Message----- From: Trevor Getty [mailto:trevor@postimpressions.co.uk] Sent: Wednesday, July 19, 2000 3:41 PM To: g.nagaraj@wipro.com Subject: Re: VXWORKS: ISR does not get called again If you are using mutliple handlers - this is a shared interrupt. Within T1 I believe they have implemented a shared ISR routine, which allows mutliple handlers to be connected. I believe its called pciIntConnect - although this is one of the extensions. I you can't find it email me again and I'll send you the code which we have written to do this. --------------------------------------------------- Trevor Getty Software Engineer (R & D) Post Impressions (Systems) Ltd. Tel: (01635) 232624 Fax: (01635) 232638 trevor.getty@postimpressions.com From vxwexplo-errs@csg.lbl.gov Wed Jul 19 04:03:07 2000 From: Vxworks Exploder Date: Wed Jul 19 04:03:09 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 19 04:03:03 PDT 2000 Subject: SCSI Minimum Period Subject: re: VXWORKS: ISR does not get called again Subject: Waking on LAN a remote machine from my vxWorks MTX603 board Subject: I can't include some UGL components after Upgrade to Tornado 2.01 Subject: ISR-Level Exceptions Subject: network load of vxworks using intel 82559 Subject: memPartAlloc Subject: Fw: Info required urgently Subject: Re: Info required urgently Subject: RE: scsiPhysDevCreate() Subject: Bootable Project on MCP750 Subject: Any useful Tool/Method to debug memory leak ? Subject: re: Bootable Project on MCP750 Subject: VxWorks Subject: scsiPhysDevCreate() Subject: VXWORKS: ISR does not get called again ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: SCSI Minimum Period Date: Tue, 18 Jul 2000 16:15:34 +0100 From: "amass" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsc$ah0$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_001B_01BFF0D3.66DD2D40 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi SCSI again I'm afraid!!!!!! We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 respectively. We need to configure the SCSI interface to use an Ultra-SCSI2 device in synchronous mode. We need clarification on the setting of 'minPeriod' in the SCSI_OPTIONS structure when setting the transfer options for the interface. Do we: a) Specifiy the minimum period factor (as specified by the SCSI-2 standard) and rely on the SCSI system to multiply by 4 to give = the appropriate delay; or b) Specify the actual minimum period required in nano seconds for the desired transfer rate. We wish to use our interface at 20M transfers/sec and have tried setting 'minPeriod' to 12 (following 'a' above), but this was not successful. We = have also tried setting 'minPeriod' to 50 nanoseconds (following 'b' above), = and this appeared to be more successful. However, I would still appreciate someone 'in the know' telling us which = is the correct solution. Regards Andy Atkinson Software Engineer DERA BIncleaves - ------=_NextPart_000_001B_01BFF0D3.66DD2D40 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
SCSI again I'm afraid!!!!!!

We have a = Radstone PPC1A /=20 NT Host running VxWorks 5.3.1 and T2
respectively. We need to = configure the=20 SCSI interface to use an
Ultra-SCSI2 device in synchronous = mode.
 
We need clarification on the setting of 'minPeriod' = in the=20 SCSI_OPTIONS
structure when setting the transfer options for the = interface.=20 Do we:
 
    a) Specifiy the minimum period = factor (as=20 specified by the SCSI-2
        standard) and = rely on=20 the SCSI system to multiply by 4 to give the
        appropriate=20 delay;
 
    or
 
    b) Specify the actual minimum = period=20 required in nano seconds for
        the desired = transfer=20 rate.
 
We wish to use our interface at 20M transfers/sec = and have=20 tried setting
'minPeriod' to 12 (following 'a' above), but this = was not=20 successful. We have
also tried setting 'minPeriod' to 50 nanoseconds = (following=20 'b' above), and this
appeared to be more successful.
 
However, I would still appreciate someone 'in the = know'=20 telling us which is the
correct solution.
 
Regards
 
 
Andy Atkinson
Software Engineer
DERA BIncleaves
 
- ------=_NextPart_000_001B_01BFF0D3.66DD2D40-- --------------------------- Newsgroups: comp.os.vxworks Subject: re: VXWORKS: ISR does not get called again Date: Tue, 18 Jul 2000 17:36:19 +0100 From: "Trevor Getty" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsd$ah2$1@overload.lbl.gov> It appears that your interrupt has been hooked up correctly - or your first interrupt wouldn't happen. 1) Before hooking up the interrupt clear it from the PIC to avoid nasty interrupt spoiling things whilst you are hooking your ISR up, it may also be helpful to clear the interrupt cause register etc. on whatever would be generating the interrupt just in case. 2) Connect up your ISR to the intline. 3) Enable the interrupt. Example. 1) sysIntDisablePIC(intline) 2) intConnect(INUM_TO_IVEC(INT_VEC_IRQ0 + intline), ISR, param) 3) sysIntEnablePIC(intline) within the ISR: disable the PIC to mask off any further interrupt happening on this line until the present one has been dealt with. deal with the interrupt cause and clear the cause register if appropriate. then enable the PIC again before exiting the ISR. On page 90 of the VxWorks T2 programmes guide there is a list of exceptions of what cannot be done within an ISR - you can give semaphores but not mutual exclusion semaphores, and VxMp ones. - check this. Hope thats of some help. - --------------------------------------------------- Trevor Getty Software Engineer (R & D) Post Impressions (Systems) Ltd. Tel: (01635) 232624 Fax: (01635) 232638 trevor.getty@postimpressions.com --------------------------- Newsgroups: comp.os.vxworks Subject: Waking on LAN a remote machine from my vxWorks MTX603 board Date: Thu, 13 Jul 2000 12:59:46 +0200 From: Jean-Marie Campiglia Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsd$ah3$1@overload.lbl.gov> Hi all, I have a MTX603 board an I need to wake-up a machine remotely connected to my TCP/IP ethernet network. The remote machine has the "Wake on LAN" enabled. Does anybody knows how to build and send the "Magic packet" from my DEC21143 driver? (I have tried to use the link layer, i,e the AF_LINK flag in place of the usual AF_INET flag when creating a socket, but apparently the AF_LINK flag is not supported in the socket() interface). Thanks for any help, Jean-Marie --------------------------- Newsgroups: comp.os.vxworks Subject: I can't include some UGL components after Upgrade to Tornado 2.01 Date: Tue, 18 Jul 2000 13:42:15 -0400 From: "Baik, John" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsd$ah4$1@overload.lbl.gov> Can anyone tell the reason I can't include some components ? I just got Tornado 2.01 for ARM (actually StrongARM SA1100). I did change BSP code as Release Note. I could make Bootrom and work fine. I had an error message when I try to include 'sa1100 diaply driver' and 'sa1100 Touch screen driver' into Vxworks image like as " Component 'sa1100 display driver' is not installed. You may need to purchase and /or install it". But It worked fine with Old version. Any idea ? Jonghyeon Baik Sr. S/W Engineer Applied Data Systems 301- 490-4007 x15 http://www.flatpanels.com --------------------------- Newsgroups: comp.os.vxworks Subject: ISR-Level Exceptions Date: Thu, 13 Jul 2000 12:26:43 +0100 From: "amass" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsd$ah5$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0049_01BFECC5.9AAFFB00 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi How should one go about tackling an exception at ISR level? Andy Atkinson DERA Bincleaves vxworks - ------=_NextPart_000_0049_01BFECC5.9AAFFB00 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
    How should one go about tackling = an=20 exception at ISR level?
 
Andy Atkinson
DERA Bincleaves
 
vxworks
- ------=_NextPart_000_0049_01BFECC5.9AAFFB00-- --------------------------- Newsgroups: comp.os.vxworks Subject: network load of vxworks using intel 82559 Date: Tue, 18 Jul 2000 11:04:54 -0700 From: Tom Miller Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dse$ah6$1@overload.lbl.gov> Hello Tornado Group! I am in the very first stages of a new vxworks project. I am trying to get a celeron processor based system to load vxworks over a network. I am booting from a boot disk. I was wondering: 1) How do I include the driver for the 82559 on the boot disk so that the system can control the ethernet device to fetch the vxWorks image? 2)When I am specifying the boot device (either in the DEFAULT_BOOT_LINE or by interrupting the boot process and manually entering the parameters) what do I enter in the Boot Device field? For example in a system using the crystal cs8900 ethernet chip, I enter "cs0" as the boot device. If you can answer either of these questions (with whatever brevity you want) I would very much appreciate it. Thank You, Tom Miller Motion Engineering Inc. Santa Barbara Ca. --------------------------- Newsgroups: comp.os.vxworks Subject: memPartAlloc Date: Thu, 13 Jul 2000 09:53:16 -0600 From: "Carolyn Briles" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dse$ah8$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0022_01BFECB0.2AB075B0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hello All, =20 I am using vxWorks on a Power PC. A simple histogram program is giving = the error =20 memPartAlloc:block too big - 524 in partition 0x196858 =20 when I run for more than 1.5M data events. The counter for events is = defined as a u_long, and the data acquisition iterates on a while loop = that compares datasofar to the counter. Both are u_long. =20 =20 This seems simple, but I'm at a loss. Increasing the stack size of the = task does not help increase the number of runs possible. Has anyone run = into this? =20 Many thanks for your help! =20 Carolyn=20 - ------=_NextPart_000_0022_01BFECB0.2AB075B0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello All,
 
I am using vxWorks on a Power = PC.  A simple=20 histogram program is giving the error
 
    = memPartAlloc:block too big -=20 524 in partition 0x196858
 
when I run for more than 1.5M data = events. =20 The counter for events is defined as a u_long, and the data acquisition = iterates=20 on a while loop that compares datasofar to the counter.  Both are=20 u_long.  
 
This seems simple, but I'm at a = loss. =20 Increasing the stack size of the task does not help increase the number = of runs=20 possible.  Has anyone run into this?
 
Many thanks for your = help!
 
Carolyn 
- ------=_NextPart_000_0022_01BFECB0.2AB075B0-- --------------------------- Newsgroups: comp.os.vxworks Subject: Fw: Info required urgently Date: Mon, 17 Jul 2000 12:00:04 +0530 From: "vishal dalal" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dse$ah9$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_00CF_01BFEFE6.8B121D80 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear colleague, Please see the following message. Thanks and with regards Vishal Dalal - -----Original Message----- From: Alan K Biocca To: Vishal Dalal Date: Monday, July 17, 2000 7:01 AM Subject: Re: Info required urgently Your request should be addressed to the vxworks users group email = exploder, or comp.os.vxworks usenet newsgroup. submit articles for the vx exploder list via vxwexplo@lbl.gov. - -- Alan K Biocca At 05:40 AM 7/15/00, you wrote: Dear Colleague,=20 Hi! We are working at Silicon=20 Automation Systems, Bangalore, India on 3g wireless handset. We = already=20 have protocol stack and we now want to implement it on physical = layer=20 (board). We are searching for suitable RTOS to run protocol stack. = We=20 are:=20 =20 1. Using Strong Arm processor=20 2. Using Universal Serial Bus (USB)=20 3. Using Ethernet card and TCP/IP=20 =20 We have option to use NUCLEUS RTOS or VXWORK or ECOS. But = we=20 have received positive response for VxWorks (from our customers).=20 Please let me know the details of VxWorks RTOS with=20 reference to our requirements like=20 =20 1. Is it portable to strong ARM (SA-1110) processor?=20 2. Does it support device drivers for USB/ETHERNET/TCP-IP?=20 3. How much is the cost of VxWorks RTOS (license fee etc..)?=20 4. What are the development tools available?=20 5. Time to ship?=20 =20 We are already going through your web site but = because of=20 the pausity of time and shortage of time to decide RTOS we are = troubling=20 you in advance. Please let me know about the details for the above = as=20 soon as possible at vishald@sasi.com=20 =20 With regards=20 Vishal Dalal=20 =20 - --=20 ############################################################ Silicon Automation Systems Limited,Hal-III Stage,Bangalore=20 Tel: 91-80-5281229,5276100 ext:4174 Fax:91-80-5284396 =20 e-mail:dalal_vishal@hotmail.com =20 Home Page: http://www.angelfire.com/ab3/vishal ############################################################ =20 - ------=_NextPart_000_00CF_01BFEFE6.8B121D80 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Dear colleague,
       Please see the = following=20 message.
Thanks and with regards
Vishal Dalal
-----Original = Message-----
From:=20 Alan K Biocca <AKBiocca@lbl.gov>
To: = Vishal=20 Dalal <vishald@sasi.com>
Date:=20 Monday, July 17, 2000 7:01 AM
Subject: Re: Info required=20 urgently

Your request should be addressed to the = vxworks=20 users group email exploder, or comp.os.vxworks usenet = newsgroup.

submit=20 articles for the vx exploder list via vxwexplo@lbl.gov.

-- Alan K = Biocca


At 05:40 AM 7/15/00, you wrote:
Dear Colleague,=20 =
           &nb= sp;           &nbs= p;         =20 Hi! We are working at Silicon
Automation Systems, Bangalore, = India on 3g=20 wireless handset. We already
have protocol stack and we now want = to=20 implement it on physical layer
(board). We are searching for = suitable=20 RTOS to run protocol stack. We
are:

1. Using Strong Arm=20 processor
2.  Using Universal Serial Bus (USB)
3. Using = Ethernet card and TCP/IP=20

          We = have=20 option to use NUCLEUS RTOS or VXWORK or ECOS. But we
have = received=20 positive response for VxWorks (from our customers).=20 =
           &nb= sp;=20 Please let me know the details of VxWorks RTOS with
reference to = our=20 requirements like

1. Is it portable to strong ARM (SA-1110)=20 processor?
2. Does it support device drivers for = USB/ETHERNET/TCP-IP?=20
3. How much is the cost of VxWorks RTOS (license fee etc..)? =
4. What=20 are the development tools available?
5. Time to ship?=20 =

           = ;   =20 We are already going through your web site but because of
the = pausity of=20 time and shortage of time to decide RTOS we are troubling
you in = advance. Please let me know about the details for the above as =
soon as=20 possible at        = vishald@sasi.com=20

With regards
Vishal Dalal 
--=20
############################################################
Silicon Automation Systems Limited,Hal-III Stage,Bangalore=20
Tel: 91-80-5281229,5276100 ext:4174
Fax:91-80-5284396         &n=
bsp;           &nb=
sp;   =20
e-mail:dalal_vishal@hotmail.com       =
          =20
Home Page:
http://www.angelfire.com/ab3=
/vishal
############################################################
  =
- ------=_NextPart_000_00CF_01BFEFE6.8B121D80-- --------------------------- Newsgroups: comp.os.vxworks Subject: RE: scsiPhysDevCreate() Date: Mon, 17 Jul 2000 09:06:58 +0200 From: ian.willats@telelogic.com Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsf$aha$1@overload.lbl.gov> Andy Atkinson asked: > Can anyone tell me what (in more detail than the manual provides) > what scsiPhysDevCreate() does on a Radstone PPC1A fitted with a NCR5930 > SCSI controller. This information is required as we are discussing a > problem with a hard dsic manufacturer who may not be familiar with > VxWorks and associated functions. > What it does in the standard scsi2Lib (as far as the SCSI target is concerned) is roughly as follows: - issue a REQUEST SENSE command asking for up to and including the additional length byte - parse the returned sense data to determine the sense data length, and allocate a buffer as required - issue an INQUIRY command, retry if necessary until it succeeds - parse the inquiry data to determine if the target supports tags, wide and/or synchronous transfer (following commands will use these features if supported) - if it's a disc device, issue a READ CAPACITY command I can't vouch for whether Radstone's driver does anything different, but it's unlikely. Ian Ian Willats Telelogic UK Ltd --------------------------- Newsgroups: comp.os.vxworks Subject: Bootable Project on MCP750 Date: Mon, 17 Jul 2000 08:17:22 -0400 From: "Saskia Zezima" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsf$ahb$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0049_01BFEFC7.6EB2B880 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, Has anyone built a Bootable Project in Tornado2 and been able to = download and the VxWorks Image? I can build the VxWorks image in the BSP directory and download it, but = when I create a Bootable based off of the BSP, it loads and then never = runs.=20 Any ideas? Saskia - ------=_NextPart_000_0049_01BFEFC7.6EB2B880 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi,
 
Has anyone built a Bootable Project in = Tornado2 and=20 been able to download and the VxWorks Image?
 
I can build the VxWorks image in the = BSP directory=20 and download it, but when I create a Bootable based off of the BSP, it = loads and=20 then never runs.
 
Any ideas?
 
Saskia
- ------=_NextPart_000_0049_01BFEFC7.6EB2B880-- --------------------------- Newsgroups: comp.os.vxworks Subject: Any useful Tool/Method to debug memory leak ? Date: Mon, 17 Jul 2000 10:31:58 -0400 From: "Wang, Tom (HT-EX)" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsf$ahc$1@overload.lbl.gov> This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_001_01BFEFFB.C3953AB0 Content-Type: text/plain Hi VxWorks/Tornado user, Has anyone had a good experience to track down a memory leak problem with tools or methods ? Appreciate if you can share your success story. We are using the Tornado 1.0.1 and VxWorks 5.3.1 on the NT platform for the R3000 target. I used the "CodeTest memory tool" to add instrumentation to the source code for memory leak. The CodeTest memory test did not show any clear evidence of memory leak, but the "memShow" did reveal memory leak of 384 bytes in 4 blocks. The CodeTest only add instrumentation to the source code compiled with its driver. This give us confidence that the leak is not in our code but in third party's object code or library. There is another type of tool which works in lower level, and it replaces memory allocation routines with debugging routine. The "MemScope" of the Performance pack is the another tool that I would like to use. Has anyone had a success experience on this tool ? Does it work correctly and provide the trace for the memory leak on the Tornado 1.0.1 on the NT platform ? If the tool is not available or not working correctly, We are planning to resolve this memory leak by the following approach. Please comment if this is (or not) a sound approach. 1. Create our memory allocation routine to replace the memory routine for debugging by using symFindByName() to locate the old routine and remap the memory allocation routine with our routines. 2. Memory Heap trace (Manual or tool to locate the leaked blocks): Use "memShow 1" to display the allocated and free blocks and capture the memory help before and after the memory leak test. This would need the knowledge of memory block format. Has anyone had tried this approach ? Is this a good approach ? TIA. - -Tom Wang Broadband Communications Sector Motorola Tel: 215-323-2102 (o) Email: towang@gi.com - ------_=_NextPart_001_01BFEFFB.C3953AB0 Content-Type: text/html Content-Transfer-Encoding: quoted-printable Any useful Tool/Method to debug memory leak ?

Hi VxWorks/Tornado user,

Has anyone had a good experience to = track down a memory leak problem with tools or methods ?  = Appreciate if you can share your success story.

We are using the Tornado 1.0.1 and = VxWorks 5.3.1 on the NT platform for the R3000 target.

I used the "CodeTest memory = tool" to add instrumentation to the source code for memory = leak.  The CodeTest memory test did not show any clear evidence of = memory leak, but the "memShow" did reveal memory leak of 384 = bytes in 4 blocks.  The CodeTest only add instrumentation to the = source code compiled with its driver.  This give us confidence = that the leak is not in our code but in third party's object code or = library.

There is another type of tool which = works in lower level, and it replaces memory allocation routines with = debugging routine. The "MemScope" of the Performance pack is = the another tool that I would like to use.  Has anyone had a = success experience on this tool ?  Does it work correctly and = provide the trace for the memory leak on the Tornado 1.0.1 on the NT = platform ?

If the tool is not available or not = working correctly, We are planning to resolve this memory leak by the = following approach. Please comment if this is (or not) a sound = approach.

1. Create our memory allocation = routine to replace the memory routine for debugging by using = symFindByName() to locate the old routine and remap the memory = allocation routine with our routines. 

2. Memory Heap trace (Manual or tool = to locate the leaked blocks):
Use "memShow 1" to display the allocated and free blocks and = capture the memory help before and after the memory leak = test.

This would need the knowledge of = memory block format.  Has anyone had tried this approach ?  = Is this a good approach ?

TIA.

-Tom = Wang
Broadband Communications Sector
Motorola
Tel: = 215-323-2102 (o)
Email: = towang@gi.com

- ------_=_NextPart_001_01BFEFFB.C3953AB0-- --------------------------- Newsgroups: comp.os.vxworks Subject: re: Bootable Project on MCP750 Date: Tue, 18 Jul 2000 08:24:21 +0200 From: "Michael Lawnick" Organization: Lawrence Berkeley National Laboratory Message-ID: <200007180623.IAA08112@net.softec.de> This has been discussed previous twice: all PPC-vxWorks images of TII won't run, if built via elfToBin (bug!). Ask your FAE for the patch of this file or burn via serial line/hex-file. > Has anyone built a Bootable Project in Tornado2 and been able to = > download and the VxWorks Image? > > I can build the VxWorks image in the BSP directory and download it, but = > when I create a Bootable based off of the BSP, it loads and then never = > runs.=20 > > Any ideas? > > Saskia 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 ============================================== --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Date: Tue, 18 Jul 2000 12:38:54 +0530 From: "Elvis John Dowson" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsg$ahf$1@overload.lbl.gov> Hi Vishal, I can answer a few quesions for you regarding VxWorks : >> 1. Is it portable to strong ARM (SA-1110) processor? Yes, VxWorks Tornado 2 supports the StrongARM-1100 processor. The CPU type is defined by setting the preprocessor variable CPU to ARMSA110 (which targets the StrongARM-110 and StrongARM-1100 processors) >> 2. Does it support device drivers for USB/ETHERNET/TCP-IP? RE: USB - ------- Well, you're in luck Vishal, there is a device driver available in VxWorks for the SA-1100 UARTs. The SA-1100 chip contains 5 serial ports, but only ports 1 and 3 are usable as UARTs. The others support Universal Serial Bus (USB), SDLC, IrDA Infrared Communications Port (ICP) and Multimedia Communications Port (MCP)/Synchronous Serial Port (SSP). The following routines are used to control the SA-1100 UART tty driver : sa1100DevInit( ) - initialise an SA1100 channel sa1100Int( ) - handle an interrupt The driver is typically only called by the BSP. RE: Ethernet - ------------ Ethernet NIC drivers exist for Intel 82557, 82596; 3COM 3C90xB XL, 3C509; DEC 21x4x; SMC 8013WC; and a couple of others from AMD, Mototola and National Semiconductor and Fujitsu. RE: TCP/IP - ----------- Yes, the standard VxWorks network stack includes support for the TCP/IP protocol. Here is a complete list of protocol and utilities supported by VxWorks : CSLIP,PPP,IP, UDP, TCP, DHCP, BOOTP, DNS, ARP and Proxy ARP OSPF, RIP, Sockets (TCP, UPD, multicasting, routing, and Zbuf) RPC, RSH, FTP, NFS, TFTP, rlogin, and telnet. >> 4. What are the development tools available? It has a browser, a shell, a debugger, a source code navigator. Apart from the basic tools, they also have a list of optional tools that can be integrated into the Tornado environment, for example : the Stethoscope tool for graphical timing anlysis. The Tornado development environment, in my opinion, is one of the most comprehensive and feature rich environments available for any RTOS. >> 5. Time to ship? You can usually have a full fledged evaluation copy running at your site and begin working as soon as you place an order. I think that the India office always keeps a couple of copies in stock, so you should have no problem obtaining a copy. Best regards, Elvis John Dowson Project Leader Simulation Systems and Systems Integration BAeHAL Software Limited Airport Lane, HAL Estate Airport Road, Bangalore - 560 017 Karnataka, India Tel : +91-80-526 8683, 526 8684 Fax : +91-80-527 0915 Mobile : +91-98450-00884 Email : elvis@baehal.com, elvis@dowsonresearch.com Website : www.baehal.com --------------------------- Newsgroups: comp.os.vxworks Subject: scsiPhysDevCreate() Date: Fri, 14 Jul 2000 15:52:01 +0100 From: "amass" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsg$ahg$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0007_01BFEDAB.72DC6460 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi Can anyone tell me what (in more detail than the manual provides) = what scsiPhysDevCreate() does on a Radstone PPC1A fitted with a NCR5930 = SCSI controller. This information is required as we are discussing a = problem with a hard dsic manufacturer who may not be familiar with = VxWorks and associated functions. Thanks Andy Atkinson - ------=_NextPart_000_0007_01BFEDAB.72DC6460 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
 
    Can anyone tell me what (in more = detail=20 than the manual provides) what scsiPhysDevCreate() does on a Radstone = PPC1A=20 fitted with a NCR5930 SCSI controller. This information is required as = we are=20 discussing a problem with a hard dsic manufacturer who may = not be=20 familiar with VxWorks and associated functions.
 
Thanks
 
 
Andy Atkinson
- ------=_NextPart_000_0007_01BFEDAB.72DC6460-- --------------------------- Newsgroups: comp.os.vxworks Subject: VXWORKS: ISR does not get called again Date: Tue, 18 Jul 2000 20:07:19 +0530 From: "G. Nagaraj" Organization: Lawrence Berkeley National Laboratory Message-ID: <8l2dsg$ahh$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0014_01BFF0F3.C7379110 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello VXWorkers, I am using the EV386EX BSP. I have an ISR to handle my hardware interrupt (IRQ 1). I connect the ISR using "intConnect" and enable the interrupt by using "sysIntEnablePIC". My ISR gets executed only once after which it does not get called. I use "semGive" and "logMsg" in the ISR. I tried giving explicit "sysIntEOI" also. But no use. It still hangs. I hope there will be lot of people on this list who can help me. Thanks in advance. Best Regards, G. Nagaraj Senior Software Engineer Wipro Technologies 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.wipro.com - ------=_NextPart_000_0014_01BFF0F3.C7379110 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello=20 VXWorkers,
 
I am = using the=20 EV386EX BSP. I have an ISR to handle my hardware interrupt (IRQ 1). I = connect=20 the ISR using "intConnect" and enable the interrupt by using = "sysIntEnablePIC".=20 My ISR gets executed only once after which it does not get called. I use = "semGive" and "logMsg" in the ISR. I tried giving explicit "sysIntEOI" = also. But=20 no use. It still hangs.
 
I hope = there will be=20 lot of people on this list who can help me.
 
Thanks = in=20 advance.
 
Best=20 Regards,

 

G. Nagaraj=20
Senior=20 Software Engineer=20
Wipro=20 Technologies=20
37, = Castle Street,=20 Ashok Nagar,=20
Bangalore - 560 025,=20 India =
Phone: +91-80-536-7297 = Ext:=20 315=20
Fax: = +91-80-557-6032=20
www.wipro.com=20

 
- ------=_NextPart_000_0014_01BFF0F3.C7379110-- --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 19 05:42:43 2000 From: Jeffrey Rada Date: Wed Jul 19 05:42:46 PDT 2000 Subject: Is there a TAR like utility ported to vxWorks This is a multi-part message in MIME format. --------------F4D608BA67E07DAED20A6BB9 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit All vxWorks users, Does any one know of a backup and restore file system utility similar to the unix tar command. Or has anyone ported tar to vxWorks. Your input would be helpful. Thanks, Jeffrey Rada mailto:jeffrey.rada@hitel.com --------------F4D608BA67E07DAED20A6BB9 Content-Type: text/x-vcard; charset=us-ascii; name="jeffrey.rada.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Jeffrey Rada Content-Disposition: attachment; filename="jeffrey.rada.vcf" begin:vcard n:Rada;Jeffrey tel;fax:770-242-1402 tel;work:770-446-8820 ext. 1968 x-mozilla-html:FALSE url:http://www.hitel.com org:Hitachi Telecom (USA) Inc.;NESW Dept-19 adr:;;3617 Parkway lane;Norcross;GA;30092;USA version:2.1 email;internet:jeffrey.rada@hitel.com title:Sr. Software Engineer x-mozilla-cpt:;-8448 fn:Jeffrey Rada end:vcard --------------F4D608BA67E07DAED20A6BB9-- From vxwexplo-errs@csg.lbl.gov Wed Jul 19 13:22:41 2000 From: Jim Fridlund Date: Wed Jul 19 13:22:44 PDT 2000 Subject: Re: Is there a TAR like utility ported to vxWorks the vxWorks Users Group Exploder writes: > > All vxWorks users, > > Does any one know of a backup and restore file system utility similar to > the unix tar command. Or has anyone ported tar to vxWorks. Your input > would be helpful. > > Thanks, > > Jeffrey Rada > mailto:jeffrey.rada@hitel.com Hi Jeff. We picked up dosFS2 code from WindRiver and this had tarLib which is compatible to Unix tar(1) command. -- Jim Fridlund fridl003@tc.umn.edu From vxwexplo-errs@csg.lbl.gov Fri Jul 7 06:33:35 2000 From: ian.willats@telelogic.com Date: Wed Jul 19 20:49:15 PDT 2000 Subject: RE: SCSI Read/Write - Atkinson Andrew asked: > We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 > respectively. We need to configure the SCSI interface to use an > Ultra-SCSI device in synchronous mode for transfers >3.8MB/sec. Normally, the SCSI interface should be more or less self-configuring to operate at the most favourable data rates and transfer protocols supported by both the VxWorks system and your SCSI target device - so you shouldn't need to do anything by hand to achieve synchronous transfers. However, I think the Radstone rs60x BSP overrides this default and forces all data transfers to be asynchronous; you may need to make sure SCSI_NO_SYNC is undefined to restore normal service! > due to inexperience with SCSI, we require some assistance with actually > reading/writing data. Can anyone point us in the right direction? > If you're dealing with a disc-like device, the routines you should use are scsiRdSecs() and scsiWrtSecs(). Before that, you'll need to create a SCSI physical device using scsiPhysDevCreate() and a block device using scsiBlkDevCreate() - examples can probably be found in the VxWorks Programmer's Guide, I think. Otherwise, you will probably have to build your own SCSI command blocks and issue them using the scsiTransact() command. You'll need to know the exact format of the SCSI commands your target device supports; this should be part of the documentation for the device itself. Hope this helps. Feel free to get in touch if there's anything else. Ian Ian Willats Telelogic UK Ltd From vxwexplo-errs@csg.lbl.gov Wed Jul 19 20:49:13 2000 Received: from postal1.lbl.gov (postal1.lbl.gov [128.3.7.82]) by csg.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA06654 for ; Wed, 19 Jul 2000 20:49:13 -0700 (PDT) Received: from SpamWall.lbl.gov (localhost [127.0.0.1]) by postal1.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA23355 for ; Wed, 19 Jul 2000 20:49:12 -0700 (PDT) Received: from mailman.ghg.net (mailman.ghg.net [206.29.116.30]) by SpamWall.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA23352 for ; Wed, 19 Jul 2000 20:49:11 -0700 (PDT) Received: from ghg.net (tnt1b-24.ghg.net [208.157.197.88]) by mailman.ghg.net (8.9.1/8.9.1/GHG 1.2) with ESMTP id WAA19541 for ; Wed, 19 Jul 2000 22:49:06 -0500 (CDT) Message-ID: <397692B2.7334DC7A@ghg.net> Date: Wed, 19 Jul 2000 22:48:34 -0700 From: Alan Currie X-Mailer: Mozilla 4.73 [en] (Win95; U) X-Accept-Language: en MIME-Version: 1.0 To: vxwexplo@lbl.gov Subject: dosFS with SCSI vxWorks 5.2 and Win NT/9X Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am reading that a SCSI disk formated under vxWorks 5.2 and one format under MSDOS 6.2 or before is different in that a partition table is built with MSDOS, but not vxWorks. I am working up a removable SCSI in a carrier to move from vxWorks 5.2 environment directly to a Windows NT system and want the file system to recognize it. Would it be recommended to format under vxWorks or MSDOS first and what is done to make the other system happy with the tables that are built? Thanks. -- Alan, Debbie, and Amanda Currie ~~O ~~O _O currie@ghg.net Houston, Tx _-\<,_-\<,_-\<,_ '96 Santana Arriva (*)/---|/---|/-(*) '86 Santana Triplet '99 Bike Friday Triplet From vxwexplo-errs@csg.lbl.gov Thu Jul 20 00:26:33 2000 From: ian.willats@telelogic.com Date: Thu Jul 20 00:26:35 PDT 2000 Subject: RE: SCSI Minimum Period (Me again...) > We need clarification on the setting of 'minPeriod' in the SCSI_OPTIONS > structure when setting the transfer options for the interface. > minPeriod should be set to the number of 4ns units as defined by the SCSI-2 standard; it is internally scaled to real time by scsiLib and/or the SCSI drivers. Hence for 20 Mxfer/sec you should use a value of 12. My apologies if this isn't clear from the documentation. If this is not working (e.g. causes your SCSI system to hang) it is often because your cabling and/or termination may not be up to the job. At those speeds, parallel SCSI is extremely sensitive to noise and crosstalk. If you're using round cable - don't! I have found by trial and error that the best cable is flat ribbon cable with adjacent pairs twisted. Another approach is to start with a slower speed (larger minPeriod) to check that the synchronous negotiation is working properly - switch on scsiDebug to see this - and crank it up to a higher speed in stages. Ian Ian Willats Telelogic UK Ltd VxWorks From vxwexplo-errs@csg.lbl.gov Thu Jul 20 07:13:03 2000 From: Mauro Galarda Martins Date: Thu Jul 20 07:13:05 PDT 2000 Subject: GNU Zebra under Vxworks Hello vxworkers Did anybody already try to port the GNU Zebra routing software under VxWorks ? thanks in advance Mauro -- _______________________________________________________ Mauro Jose Galarda Martins Siemens Tel TR-DS Rua Pedro Gusso, 2635 CEP 81310-900 Curitiba - Parana' - Brasil Tel: +55 41 3416457 Fax: +55 41 3415620 e-mail: mauro.martins@siemens.com.br _______________________________________________________ From vxwexplo-errs@csg.lbl.gov Thu Jul 20 09:54:01 2000 From: Suresh T Date: Thu Jul 20 09:54:03 PDT 2000 Subject: clarification Hi, We are working on Strong arm ARMSA110 board. We are facing a problem in using the BSD function "bcopy_to_mbufs" for the application Software. It would be of great help if any one can clarify our doubt. The function "bcopy_to_mbufs" takes the following parameters bcopy_to_mbufs ( buffer, length, offset, ifnet *, width) and returns a pointer to the mbuf. The width field is specified as 2 for the 486 based PubStation and is working fine. We have a flat buffer like "ff 03 c0 21 01 01 00 11 05 06 04 18 a7 ed 07 02 08 02 0d 03 06" and we do a bcopy_to_mbufs in 486 environment. After this bcopy, if we print the mbuf, it is printing properly from the beginning. DATA[0] : 68 9b fe 0 1d 1e ea 71 0 60 1d 1e e9 e4 88 78 ff 3 c0 21 1 1 0 11 5 6 4 18 a7 ed 7 2 8 2 d 3 6 1 8 0 6 4 0 1 0 60 1d 1e e9 e4 a a 0 a 0 0 0 0 0 0 In the above output, "ff 3 c0 21" is the beginning of the PPP header and it is printed without any mistake. But in StrongARM environment, if we use width as 2 it is giving problem. If we pass a flat buffer like ff 03 c0 21 01 01 00 11 05 06 04 18 a7 ed 07 02 08 02 0d 03 06 with a data length of 21 and print the mbuf data in the Strong ARM Compiled code we have a difference in the representation of the data DATA[0] : 68 9b fe 0 1d 1e ea 71 0 60 1d f1 7b f7 88 78 c0 21 1 1 0 11 5 6 4 20 75 5b 7 2 8 2 d 3 6 ee 0 1 8 0 6 4 0 1 0 60 1d 1e e9 e4 a a 0 a 0 0 0 0 0 0 In this case, the beginning of PPP header is missing. (ff 3 is missed out). The problem being that after 16 bytes of mbuf header the bytes "ff 3" gets chopped off in the Strong ARM Code, which is actually our PPP Header. If we change the value of the width field in the function call to '1' , it works fine in the Strong ARM Compiled code. We want to know the purpose of the width parameter and whether the function call is processor dependent? We searched for the reference to the function call in the Tornado Manuals we couldn't find any useful information on this. We have another problem in handling the Mbuf buffer again between PPP and IP. The Mbuf is given from the PPP to IP and structure of the mbuf array is 24 9b fe 0 1d 1e ea 71 0 60 1d f1 7b f7 88 78 21 45 0 0 26 20 f 0 0 20 1 53 8b a 0 0 a0 a 0 12 9e 8 21 represents the ppp header and we strip of the PPP header before giving to IP by offsetting it to the next location, but the packets are not routed by the IP and badlen is reported by the IP. The actual data should be like this 45 0 0 26 20 f 0 0 20 1 53 8b a 0 0 a0 a 0 12 9e 8 0 92 ff 2 0 68 0 61 62 63 64 65 66 67 68 69 6a at this place before giving to the IP, if we do a flat buffer copy and again convert it to Mbuf and give to IP then it is working fine. So anyone ever faced this problem. Without the extra copy, if we try to send out some data packets and see the ipstatShow on the shell, the count for badlen gets incremented and ip as such doesn't produce error number, but discards the packet as badlen packet, but the packet is valid. Thanks and regards Suresh From vxwexplo-errs@csg.lbl.gov Sun Jul 23 04:03:13 2000 From: Vxworks Exploder Date: Sun Jul 23 04:03:16 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 23 04:03:08 PDT 2000 Subject: Re: In-Circuit-Emulator for x86 Subject: Re: PowerPC 8260 Interrupts Subject: Re: END Driver For 82557 Subject: Task Pre-emption in vxWorks Subject: Can anybody explain this strange result? Subject: Re: MVME2431 ethernet problem Subject: re: Is there a TAR like utility ported to vxWorks Subject: Re: PowerPC 8260 Interrupts Subject: UDP problem Subject: Re: ftp server Subject: Re: Deja.Com Daily Summary: comp.os.vxworks 1/1 Subject: Re: UDP problem Subject: Structures Subject: Re: Task Pre-emption in vxWorks Subject: Re: UDP problem Subject: Re: Structures Subject: Re: Vworks Ether Hook and SNARF information Subject: Re: Vworks Ether Hook and SNARF information Subject: Re: Q: Non-blocking socket I/O?? Subject: Re: Problems with network stack (VxWorks 5.3) Subject: Re: VxWorks contract opportunity Subject: RE: SCSI Read/Write - Subject: dosFS with SCSI vxWorks 5.2 and Win NT/9X Subject: Re: VxWorks contract opportunity Subject: Re: Structures Subject: GNU Zebra under Vxworks Subject: RE: SCSI Minimum Period Subject: Re: Max Resources ? Subject: Momory mapping files. Subject: clarification Subject: END Driver for 82559 Subject: 'splnet' and 'spx' Subject: Re: UDP problem Subject: Potential Opportunity Subject: Re: In-Circuit-Emulator for x86 Subject: Re: C++ compiler for NetArm 40?? Subject: Re: taskDelete in VxWorks Subject: Re: T1, Command Line, Memory Leak? Subject: Re: is somebody use PJWorks ? Subject: Re: Max Resources ? Subject: Re: Task Pre-emption in vxWorks Subject: use dos filesystem on ram disk Subject: Re: loading MVME2100 boot code Subject: Re: bootrom_uncmp build errors Subject: loading MVME2100 boot code Subject: using the serial line with vxWorks Subject: Raw IP Subject: Re: taskDelete in VxWorks Subject: VxDCOM Subject: Help with VxWorks Subject: Fail to boot with multiple 82559 cards Subject: verbosity macro Subject: Re: using the serial line with vxWorks Subject: Re: END Driver for 82559 Subject: End Driver Question ( motfcc ethernet driver ) Subject: Re: pciIntConnect and intConnect Subject: Re: pciIntConnect and intConnect Subject: Re: for help:why my hook function can't work? Subject: about mbuf ! Subject: Re: loading MVME2100 boot code Subject: Socket closing on VXWORKS. Subject: cacheFlush() on Motorola MPC860P Subject: ppc2799: RAM_HIGH_ADDRS=0x20,A000 will not boot Subject: Re: taskDelete in VxWorks Subject: Re: Other RTOS's? Subject: Re: Socket closing on VXWORKS. Subject: who owns this semaphore? Subject: Re: How MCP750 share it's local memory to PCI bus? Subject: Re: How MCP750 share it's local memory to PCI bus? Subject: Re: SSLeay or OpenSSH Subject: Semaphores and Mutexes? Subject: taskDelete in VxWorks Subject: Re: Semaphores and Mutexes? Subject: Re: sysFlashWrite() Subject: Re: taskDelete in VxWorks Subject: Re: Fail to boot with multiple 82559 cards Subject: Re: END Driver for 82559 Subject: Re: How MCP750 share it's local memory to PCI bus? Subject: Re: Semaphores and Mutexes? Subject: Re: Priority of the printf task Subject: Re: Memory mapped files Subject: Re: Max Resources ? Subject: Re: Motorola is obsoleting MVMExxx ! Subject: Re: about mbuf ! Subject: Re: PowerPc /vxWorks Interrupt latency Subject: Re: Raw IP Subject: a cache problem of PPC8260 Subject: 750ppc cpu instruction times? help Subject: Re: Priority of the printf task Subject: Re: vxGetDec()/ppcDecTimer.c Subject: Can't send correct data when using octet string with snmpv1. Subject: Re: END Driver for 82559 Subject: 82559ER Driver ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: In-Circuit-Emulator for x86 Date: Wed, 19 Jul 2000 10:13:57 -0700 From: "Asher Greenberg" Organization: SBC Internet Services Message-ID: References: <3975DE1A.A096129E@t-online.de> Check out http://www.americanarium.com/ they might have what you need. - -Asher. "Heribert Volkhausen" wrote in message news:3975DE1A.A096129E@t-online.de... > Hi, > > I'm looking for an In-Circuit-Emulator for an Intel 486 and Intel > Pentium > processor, which supports VxWorks 5.3/5.4. Does someone have any > experiences? > > Regards > Heribert Volkhausen > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PowerPC 8260 Interrupts Date: Wed, 19 Jul 2000 17:12:12 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8l4ngp$vag$1@nnrp1.deja.com> References: <8l2iuv$p1l2@nntp.cig.mot.com> In article <8l2iuv$p1l2@nntp.cig.mot.com>, "Derek Opitz" wrote: > I'm having trouble understanding how the PPC8260 implements interrupts. > > For instance, on the 68030, INUM_TO_IVEC( intNum ) gave you an exact vector > into the Interrupt Vector Table (IVT). [The Loooooong Answer] Because the 68030 has a different IVT structure than the PPC. The 68030 (and all the 68xxx processors) allow devices to put their own vector on the bus and read it during an IACK cycle and thus compute the entry into the vector table. Imagine if the 68030 only had 1 entry in the vector table called "external interrupt from some device", and always went to that spot, no matter what. Now you're thinking PPC. > But the 8260 doesn't seem to have a true IVT. According to VxWorks > documentation, they create an interrupt vector table array. This doesn't > bother me, except for the fact that according to ivPpc.h: > > #define INUM_TO_IVEC(intNum) ((VOIDFUNCPTR *) (intNum)) > > So, intConnect must use this VOIDFUNCPTR to get an index into the array to > hook your routine into. But this seems very strange to me. > > I guess the basic question here is, how does your routine get hooked into > the interrupt array?? The PowerPC (not just 8260, but all the PPCs) have 1 vector for external interrupts, at 0x500. There is just 1 external interrupt request pin on the PPC (never mind NMI, SMI, etc...). So whenever any device generates and external IRQ, the PPC will *always* go to MSR[IP] + 0x500. Check "The Green Book". So how does this work in vxWorks (or any other OS) ? You load a generic interrupt handler at 0x500. You also create an array of function pointers (and arguments). When someone calls intConnect(), you use the vector to make the appropriate entry in the array. When the code at 0x500 runs (due to an external IRQ), it somehow must determine the cause of the interrupt (on the 8260 by reading SIVEC to get the vector) and then using that vector as an index into the table it keeps. [The Short Answer] Check out target/src/drv/intrCtl/ppc{403, 860, 8260}Intr.c in your BSP. hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: END Driver For 82557 Date: Wed, 19 Jul 2000 17:19:39 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8l4nul$vg7$1@nnrp1.deja.com> References: <5Z5d5.54$YQ2.68982@news.pacbell.net> In article <5Z5d5.54$YQ2.68982@news.pacbell.net>, "Asher Greenberg" wrote: > Hello VxWorkers, > > I'm trying to port an old style BSD driver for the Intel 82557 Ether > controller to the newer VxWorks END driver for the same chip. I'm using a > PPC750 based target and a NT host. There is an exception generated when the > runtime image loads. Does anybody have any pointers/suggestions? > > Thanks. > > Here's the console printout (with the DRV_DEBUG_ALL flag turned on): > > Attaching network interface fei0... L2CR=0 > done. > Attaching network interface lo0... done. > Loading... 1095312 > Starting at 0x10000... > > L2CR=0 > > data access > Exception current instruction address: 0x00014b64 > Machine Status Register: 0x0000b030 > Data Access Register: 0x00d0b76c > Condition Register: 0x44000080 > Data storage interrupt Register: 0x40000000 > Task: 0x7ffdf8 "tRootTask" > 0x7ffdf8 (tRootTask): Loading end... > 0x7ffdf8 (tRootTask): Loading end... > 0x7ffdf8 (tRootTask): fei82557EndLoad: unit=0 pMemBase=0x0 memSize=0x0 > nCFDs=32 nRFDs=32 flags=6 MSR[IR] and MSR[DR] (instruction MMU and data MMU) are turned on, so you must have every memory access translateable (is that a word ?). Since you generated a data access exception, you don't. It appears you tried to access memory location 0x00d0b76c (DAR) and you didn't have it covered by a PTE or DBAT (DSISR=0x40000000). Or maybe you have it covered but the bits are not set correctly (i.e. user/supervisor, length, cache inhibited if you have cache enabled). You can check the cache with vxHid0Get(). Since you print out L2CR=0, (the L2 cache is disabled), it's possible you have L1 data cache enabled, and your memory is non-cacheable (or needs to be ?). hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Task Pre-emption in vxWorks Date: Wed, 19 Jul 2000 18:23:52 GMT From: kpulavar@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8l4rn1$2s2$1@nnrp1.deja.com> I have worked in VRTX before, but am sort of new in the vxWorks world. My question is: in VRTX ISRs there is a exit function called ui_exit, which invokes the scheduler to ensure that any high priority task(higher than the one interrupted) gets to run immediately after the ISR. I looked for such a mechanism in vxWorks, but did'nt find one in the manual. I thought the intConnect() function might provide the code to invoke the scheduler, but from the explanation in the manual, this appears unlikely. Anyone know how pre-emption by a higher priority task takes place ? Thanks in advance. - -kris Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Can anybody explain this strange result? Date: Wed, 19 Jul 2000 19:25:11 +0100 From: "Richard Barry" Message-ID: I have experience a strange phenomenon while running the same code section of several different PC systems. I expected the code to run faster on faster CPU's, but found this not to be the case. My timing results for the different systems are listed below: System Time to run code section 600MHz P3 Dell system = 0.33ms. Old AMD K6 233MHz system = 0.23ms. Xycom 133MHz system = 0.6ms. PC104 166MHz system = 0.25. Why should a 600MHz P3 produce a slower time than a 166MHz system? I am guessing that this has something to do with either the cache not being set up properly, or the math co-processor (there is some floating point math in the code), or the fact that I am using an ISA bus on a system that is really designed for PCI devices. The same VxWorks kernel was used on each. Any help would be greatly appreciated. Regards. - -- r.barry@wittenstein.co.uk Wittenstein Aktiv Technologies Ltd. 3b Osprey Court, Hawkfield Way, Bristol, BS14 0BB, UK --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MVME2431 ethernet problem Date: 19 Jul 2000 18:56:03 GMT From: bowtie@netcom.com (Bill Dennen) Organization: MindSpring Enterprises Message-ID: <8l4tk3$gq1$1@slb3.atl.mindspring.net> References: <39749C9D.3313BF58@aps.anl.gov> Andrew Johnson writes: >Ken Parrish wrote: >> >> I am trying to establish a Ethernet connection between a host (Windows NT) >> and the target (MVME2431). I can communicate through the RS232 DEBUG port >> using PPC4_BUG. I have used the PPC4-BUG command NIOT to set the IP >> addresses for the client, host, subnet and gateway on the target. When I >> use the NIOP command to get the boot image from the host, PPC4-BUG locks up, >> and the only indication that anything is working is the Ethernet led lights >> flash about 3 times and then nothing. When I ping the address from the >> host, it times out. I have removed the network from the equation and used a >> single crossover cable between the host and target, and still nothing. Does >> anyone have any suggestions? >You probably don't want to set the gateway address - that's only needed if >your host and target are in different subnets, and the networking stack in >the Bug is probably sufficiently minimal that it won't understand messages >that the gateway sends back that tell it to talk direct to the host. Set >it back to 0.0.0.0, double-check your network connections and try again. >- Andrew >-- >Sharks kill 10 people each year. People kill 60,000,000 sharks each year. Be aware that Bug does not respond to pings! It has no support stack. Check connectivity by pinging from 2400 to host. Command is nping 0 0 also early model boards did have difficulty autonegotiating to 100BaseT with certain hubs. If you've got this setup try not autonegotiating until you know that's not the problem. Regards - -- >@< bowtie@netcom.com Bill Dennen william.dennen@motorola.com Motorola Computer Group, TOPAZ --------------------------- Newsgroups: comp.os.vxworks Subject: re: Is there a TAR like utility ported to vxWorks Date: Wed, 19 Jul 2000 15:22:37 -0500 (CDT) From: Jim Fridlund Organization: Lawrence Berkeley National Laboratory Message-ID: the vxWorks Users Group Exploder writes: > > All vxWorks users, > > Does any one know of a backup and restore file system utility similar to > the unix tar command. Or has anyone ported tar to vxWorks. Your input > would be helpful. > > Thanks, > > Jeffrey Rada > mailto:jeffrey.rada@hitel.com Hi Jeff. We picked up dosFS2 code from WindRiver and this had tarLib which is compatible to Unix tar(1) command. - -- Jim Fridlund fridl003@tc.umn.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PowerPC 8260 Interrupts Date: Wed, 19 Jul 2000 12:53:28 -0700 From: "Rob M" Message-ID: <397606e0$0$309@newsreader.alink.net> References: <8l2iuv$p1l2@nntp.cig.mot.com> First, the 8260 interrupt code is actually here: /tornado/target/h/drv/intrCtl/m8260IntrCtl.h /tornado/target/src/drv/intrCtl/m8260IntrCtl.c INUM_TO_IVEC is defined to be m8260InumToIvec() m8260IntConnect() connects your ISR to the interrupt vector. Note that there is a bug in m8260IntConnect() which prohibits any interrupt numbers greater than 55 from being connected to ISRs. There is an SPR issued about this (# 32665). "Derek Opitz" wrote in message news:8l2iuv$p1l2@nntp.cig.mot.com... > I'm having trouble understanding how the PPC8260 implements interrupts. > > For instance, on the 68030, INUM_TO_IVEC( intNum ) gave you an exact vector > into the Interrupt Vector Table (IVT). > > But the 8260 doesn't seem to have a true IVT. According to VxWorks > documentation, they create an interrupt vector table array. This doesn't > bother me, except for the fact that according to ivPpc.h: > > #define INUM_TO_IVEC(intNum) ((VOIDFUNCPTR *) (intNum)) > > So, intConnect must use this VOIDFUNCPTR to get an index into the array to > hook your routine into. But this seems very strange to me. > > I guess the basic question here is, how does your routine get hooked into > the interrupt array?? > > Does anybody know if this is correct, or if I'm barking up the wrong tree > here?? > > Derek > > --------------------------- Newsgroups: comp.os.vxworks Subject: UDP problem Date: Wed, 19 Jul 2000 16:25:36 -0400 From: ico Organization: Hughes Network Systems Message-ID: <39760EC0.1EAEF7BF@139.85.103.45> Hi all, We are trying to run UDP bench mark tests for VxWorks on a 100baseT ethernet. We have only 2 machines on the network. The client sends a burst of 10 packets (500 bytes each) to a server, then waits for 20-50ms and sends another burst. On the server side there are four tasks(of same priority) running. The packets are transferred from one task to another with each task doing some processing on the packet before sending it out to a subsequent task. One of the tasks breaks up the 500 byte packet into 100 byte packets. The packets are then sent back to the client. The client measures the time it takes to get back the 500 byte packets(approx 1ms per 500 bytes). If we sent out 10,000 packets or so then we seem to be losing packets towards the end. Sending 100 packets doesn't seem to give any problems. Any help would be greatly appreciated. Thanks naveen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ftp server Date: Wed, 19 Jul 2000 23:19:04 +0200 From: Fab Message-ID: <39761B48.FD1FF7B4@nospam.com> References: <39744936.77234101@ecitele.com> Hello

First, you must have an FTP server running on your host (WinNT station) and create an user (name, password and right to acces your VxWorks file). The FTP server supplied with Tornado works well :-). You can enable logging too.

The VxWorks boot params on your target must be as follows:

boot device : <boot device >
unit number : 0
proc. # : 0
host name : < name of your NT station>
file name:  e:\mydocs\works\vxWorks (the entire directory followed by the name of the VxWorks file )
inet on eth : < IP adress to give to your target (according to the hosts file of  the network) >
inet on bckplane : can be null
host inet: < IP adress of your NT station>
user: <name of the user created in the FTP server>
passwd : <password gave to this new user >
flags : < see doc to enable quick boot, etc. ...>
target name: < name to give to the target as found in the hosts file of the NT station (preferred)>

HTH

Fab
ofer rashuk a écrit :

Hi .
i am trying to connect my host(win NT) to my VxWorks target via ftp
server.
it seems that the target dose not respond to the hookup request
(although a new ftpWorkClass has been created), can anybody help me
--------------------------- Newsgroups: comp.os.vxworks Subject: Re: Deja.Com Daily Summary: comp.os.vxworks 1/1 Date: Wed, 19 Jul 2000 16:20:24 -0700 From: "M. Kyle Craig" Organization: Deja.com - Before you buy Message-ID: <00d501bff1d7$eb83f6a0$c5a8efd0@atinucleus.com> M. Kyle Craig will be on vacation until 8/7. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UDP problem Date: Wed, 19 Jul 2000 17:35:35 -0400 From: Joe Durusau Message-ID: <39761F27.62260CB0@bellsouth.net> References: <39760EC0.1EAEF7BF@139.85.103.45> You are probably exceeding the capacity of some buffer in the system, on one end or the other. Running snoop or some such would identify which end. It's possible that crosswind could give a clue to the problem if it's on the receive end, but of course it will add some overhead to your system. I suggest that you need to remember that BY DEFINITION, udp does NOT guarantee transmission, since the higher level protocols are expected to make arrangements for recover of lost datagrams. Speaking only for myself, Joe Durusau ico wrote: > > Hi all, > We are trying to run UDP bench mark tests for VxWorks on a > 100baseT ethernet. We have only 2 machines on the network. The client > sends a burst of 10 packets (500 bytes each) to a server, then waits > for 20-50ms and sends another burst. On the server side there are four > tasks(of same priority) running. The packets are transferred from one > task to another with each task doing some processing on the packet > before sending it out to a subsequent task. One of the tasks breaks up > the 500 byte packet into 100 byte packets. The packets are then sent > back to the client. The client measures the time it takes to get back > the 500 byte packets(approx 1ms per 500 bytes). If we sent out 10,000 > packets or so then we seem to be losing packets towards the end. Sending > 100 packets doesn't seem to give any problems. Any help would be > greatly appreciated. > > Thanks > > naveen --------------------------- Newsgroups: comp.os.vxworks Subject: Structures Date: Wed, 19 Jul 2000 19:57:20 -0400 From: Vividh Siddha Organization: Lucent Technologies Message-ID: <39764060.5B64B6E4@lucent.com> Can you pass structures as parameters to a taskSpawn call? There are 10 parameters of type int in taskSpawn which are used to passing parameters to the task function. Can we pass a structure?I can pass a pointer to the structure as a int but I would like to pass it as a structure because of some inter-task dependency. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Task Pre-emption in vxWorks Date: Thu, 20 Jul 2000 00:42:21 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DA473.B09EF347@mediaone.net> References: <8l4rn1$2s2$1@nnrp1.deja.com> kpulavar@my-deja.com wrote: > I have worked in VRTX before, but am sort of new in the vxWorks world. > > My question is: in VRTX ISRs there is a exit function called ui_exit, > which invokes the scheduler to ensure that any high priority task(higher > than the one interrupted) gets to run immediately after the ISR. > > I looked for such a mechanism in vxWorks, but did'nt find one in the > manual. I thought the intConnect() function might provide the code > to invoke the scheduler, but from the explanation in the manual, this > appears unlikely. > > Anyone know how pre-emption by a higher priority task takes place ? > > Thanks in advance. > > -kris > > Sent via Deja.com http://www.deja.com/ > Before you buy. Sure, Either by waking up a task that's pending on a semaphore or a message. That's about it. Tasks of the same priority wil run in a RR fashion. When a tak of a higher priority gets woken up, it will run immediatley. From an ISR you can signal a task by calling msgQSend. This is safest way to wake a high priority task after servicing an interrupt. Best regards, - --Stephen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UDP problem Date: Thu, 20 Jul 2000 00:44:43 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DA502.63410AB6@mediaone.net> References: <39760EC0.1EAEF7BF@139.85.103.45> ico wrote: > Hi all, > We are trying to run UDP bench mark tests for VxWorks on a > 100baseT ethernet. We have only 2 machines on the network. The client > sends a burst of 10 packets (500 bytes each) to a server, then waits > for 20-50ms and sends another burst. On the server side there are four > tasks(of same priority) running. The packets are transferred from one > task to another with each task doing some processing on the packet > before sending it out to a subsequent task. One of the tasks breaks up > the 500 byte packet into 100 byte packets. The packets are then sent > back to the client. The client measures the time it takes to get back > the 500 byte packets(approx 1ms per 500 bytes). If we sent out 10,000 > packets or so then we seem to be losing packets towards the end. Sending > 100 packets doesn't seem to give any problems. Any help would be > greatly appreciated. > > Thanks > > naveen Are you bench marking for performance reasons ?? If so there are other means to send IP packets through the physical layer other than using sockets. The VxWorks socket layer is pretty slow. If you are interested I can tell how to go about bypassing the vxWorks netWork layer for processing IP packets at a much faster rate. - --Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structures Date: Thu, 20 Jul 2000 00:57:07 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DA7E5.E3890B42@mediaone.net> References: <39764060.5B64B6E4@lucent.com> Vividh Siddha wrote: > Can you pass structures as parameters to a taskSpawn call? There are > 10 parameters of type int in taskSpawn which are used to passing > parameters to the task function. Can we pass a structure?I can pass a > pointer to the structure as a int but I would like to pass it as a > structure because of some inter-task dependency. Sure, Make sure you pass the address of the structure and cast it to (UINT32 *). example: typedef struct _demo { unsigned long name; unsigned long index; } DEMO; void start_demo(void) { DEMO *my_demo=NULL; assert((my_demo = (DEMO *)malloc(sizeof(DEMO)) != NULL); assert((tid = taskSpawn ("root", 100, 0, 2000,(FUNCPTR)demo, (UINT32)my_demo, 0,0,0,0,0,0,0,0,0)) != ERROR); } void demo ( UINT32 arg) { DEMO *mydemo = (DEMO *) arg; if(!arg) { printf("demo: Expecting argument\n"); assert(false); } printf("Demo: running.\n"); printf("Value of arg: %s, %d\n", mydemo->name, mydemo->index); free(mydemo); while(1) ; } --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vworks Ether Hook and SNARF information Date: Thu, 20 Jul 2000 01:03:54 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DA97F.9F068C8F@mediaone.net> References: <8l1tkh$rs7$1@nnrp1.deja.com> kumar_vinit@my-deja.com wrote: > Hi, > > I have been trying to find out how to hook a module betwen the TCP/IP > and the Data Link Layer in VxWorks. I found a convenient mechanism in > the Ether Hook functions. > > But looking at the requirement that the module should be able to > modify the packets and also be able to send any new self generated > packets, I found that EtherHook might not be sufficient. Is it > possible to support the above requirement using Ether Hooks ? Have I > missed something ? > > What is the SNARF protocol ? Where will I get more information on this. > > Thanks in Advance, > > Vinit Kumar > > Sent via Deja.com http://www.deja.com/ > Before you buy. Sure is, use etherInputHookAdd to capture packets coming in form the physical layer (ether driver). You have the option to handle the incoming packets by processing IP and shipping them off.To do this you need to return TRUE asfter the hook. If you return FALSE, you are telling the handler that you do not want to handle the incoming packet. This handle will get called for EVERY packet. To handle outgoing packets, use etherOutputHookAdd. The same above logic applies. This is a very nice mechanism if you don't have access to the physical layer driver. I used this mechanism to design my own datalink/IP layer for bypassing vxWorks network stack, It's much more efficiant and there is considerably less overhead. If you have any question about this let me know. There is also etherOutput to send packets directly to the physical layer driver. Be warned though, this routine will pre-pend an ethernet header on for you along with following some ancient rules surrounding ethernet. Have fun, - --Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vworks Ether Hook and SNARF information Date: Thu, 20 Jul 2000 01:04:01 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DA987.638A7D95@mediaone.net> References: <8l1tkh$rs7$1@nnrp1.deja.com> kumar_vinit@my-deja.com wrote: > Hi, > > I have been trying to find out how to hook a module betwen the TCP/IP > and the Data Link Layer in VxWorks. I found a convenient mechanism in > the Ether Hook functions. > > But looking at the requirement that the module should be able to > modify the packets and also be able to send any new self generated > packets, I found that EtherHook might not be sufficient. Is it > possible to support the above requirement using Ether Hooks ? Have I > missed something ? > > What is the SNARF protocol ? Where will I get more information on this. > > Thanks in Advance, > > Vinit Kumar > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Q: Non-blocking socket I/O?? Date: Thu, 20 Jul 2000 01:06:52 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DAA32.AB7153F8@mediaone.net> References: <8kluuj$gro$1@nnrp1.deja.com> dchou4u@hotmail.com wrote: > Does VxWorks support non-blocking socket I/O? I have looked everywhere > in the header files, but it looks like VxWorks does not support non- > blocking socket I/O. > > If so, does anyone know how to do non-blocking socket I/O in VxWorks? > For example, when I do a 'connect', I dont want it to sleep in there > until the 3-way handshake is completed. I want it to return immediately > with a IN_PROGRESS error code, which I can then later do a select on to > see if the fd is ready. > > Thanks > dchou4u@hotmail.com > > Sent via Deja.com http://www.deja.com/ > Before you buy. Yes, use ioctl to handle this: if(ioctl(sFd, FIONBIO, TRUE) == ERROR) { perror("Unable to turn on Non-blocking I/O for this socket"); close(sFd); return ERROR; } This will do it for you. You should use select though. You can have a task pend on select. When data comes in, the task is woken up to handle the new packet. - --Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with network stack (VxWorks 5.3) Date: Thu, 20 Jul 2000 01:09:58 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DAAEC.D25ABFB5@mediaone.net> References: <396DE3ED.EF44BB02@t-online.de> Heribert Volkhausen wrote: > Hello, > > we are using VxWorks 5.3 with the pc486 BSP. > Unfortunately, a lot of UDP packets we receive by the ethernet interface > are corrupted. This means that the IP or UDP checksum > is wrong. > After some hours of operation the network stack crashes. > If we do not transfer wrong packets (we check them > in an ethernet hook routine) to the stack, there is > no problem. > > Does anyone of you discovered this behaviour, too? > > Regards > Heribert Volkhausen > > e-Mail: volkhau@t-online.de Does you tNetTask die ???? There shouldn't be any problems with the network stack. Are these out-bound or in-bound packets ??? - --Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks contract opportunity Date: Thu, 20 Jul 2000 01:11:02 GMT From: Stephen Banville Organization: Road Runner Message-ID: <300DAB2B.6A5E836@mediaone.net> References: <8kku84$o09$1@nnrp1.deja.com> clummis@volt.com wrote: > I am an in-house recruiter for Volt Services Group, one of the nation's > largest computer and technical consulting firms. Our purpose is to > search for individuals that match an active job opening at one of our > client companies and put you (or someone you know who would be > interested in this opportunity) in touch with the hiring contact at > that company. > > I found your resume on the Internet and thought it looked like a strong > fit for our current contract opening for a Sr. Software Engineer > > The following is a brief description of the job: > > You will be involved with software development designing, coding and > testing software for a transceiver system. The development will > specifically be focused on fault management, operations and maintenance > for this transceiver system. Our client is a global leader in providing > integrated communications solutions and embedded electronic solutions > located in the Chicago area > > If you would be interested in hearing more about this opportunity > please call me at 1-800-573-8658. If I am unavailable at the time of > your call, please leave a message and I will return your call promptly. > > I look forward to hearing from you soon. > > Regards, > > Chad Lummis > Technical Recruiter > Volt Services Group > clummis@volt.com > > Sent via Deja.com http://www.deja.com/ > Before you buy. What about contract work ?? - --Steve --------------------------- Newsgroups: comp.os.vxworks Subject: RE: SCSI Read/Write - Date: Fri, 7 Jul 2000 15:33:43 +0200 From: ian.willats@telelogic.com Organization: Lawrence Berkeley National Laboratory Message-ID: <8l5uca$bte$1@overload.lbl.gov> Atkinson Andrew asked: > We have a Radstone PPC1A / NT Host running VxWorks 5.3.1 and T2 > respectively. We need to configure the SCSI interface to use an > Ultra-SCSI device in synchronous mode for transfers >3.8MB/sec. Normally, the SCSI interface should be more or less self-configuring to operate at the most favourable data rates and transfer protocols supported by both the VxWorks system and your SCSI target device - so you shouldn't need to do anything by hand to achieve synchronous transfers. However, I think the Radstone rs60x BSP overrides this default and forces all data transfers to be asynchronous; you may need to make sure SCSI_NO_SYNC is undefined to restore normal service! > due to inexperience with SCSI, we require some assistance with actually > reading/writing data. Can anyone point us in the right direction? > If you're dealing with a disc-like device, the routines you should use are scsiRdSecs() and scsiWrtSecs(). Before that, you'll need to create a SCSI physical device using scsiPhysDevCreate() and a block device using scsiBlkDevCreate() - examples can probably be found in the VxWorks Programmer's Guide, I think. Otherwise, you will probably have to build your own SCSI command blocks and issue them using the scsiTransact() command. You'll need to know the exact format of the SCSI commands your target device supports; this should be part of the documentation for the device itself. Hope this helps. Feel free to get in touch if there's anything else. Ian Ian Willats Telelogic UK Ltd From vxwexplo-errs@csg.lbl.gov Wed Jul 19 20:49:13 2000 Received: from postal1.lbl.gov (postal1.lbl.gov [128.3.7.82]) by csg.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA06654 for ; Wed, 19 Jul 2000 20:49:13 -0700 (PDT) Received: from SpamWall.lbl.gov (localhost [127.0.0.1]) by postal1.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA23355 for ; Wed, 19 Jul 2000 20:49:12 -0700 (PDT) Received: from mailman.ghg.net (mailman.ghg.net [206.29.116.30]) by SpamWall.lbl.gov (8.9.3/8.9.3) with ESMTP id UAA23352 for ; Wed, 19 Jul 2000 20:49:11 -0700 (PDT) Received: from ghg.net (tnt1b-24.ghg.net [208.157.197.88]) by mailman.ghg.net (8.9.1/8.9.1/GHG 1.2) with ESMTP id WAA19541 for ; Wed, 19 Jul 2000 22:49:06 -0500 (CDT) Message-ID: <397692B2.7334DC7A@ghg.net> Date: Wed, 19 Jul 2000 22:48:34 -0700 From: Alan Currie X-Mailer: Mozilla 4.73 [en] (Win95; U) X-Accept-Language: en MIME-Version: 1.0 To: vxwexplo@lbl.gov Subject: dosFS with SCSI vxWorks 5.2 and Win NT/9X Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am reading that a SCSI disk formated under vxWorks 5.2 and one format under MSDOS 6.2 or before is different in that a partition table is built with MSDOS, but not vxWorks. I am working up a removable SCSI in a carrier to move from vxWorks 5.2 environment directly to a Windows NT system and want the file system to recognize it. Would it be recommended to format under vxWorks or MSDOS first and what is done to make the other system happy with the tables that are built? Thanks. - -- Alan, Debbie, and Amanda Currie ~~O ~~O _O currie@ghg.net Houston, Tx _-\<,_-\<,_-\<,_ '96 Santana Arriva (*)/---|/---|/-(*) '86 Santana Triplet '99 Bike Friday Triplet --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks contract opportunity Date: Thu, 20 Jul 2000 03:32:56 GMT From: tangoNO_SPAM@pulp-fiction.com (Max) Organization: Posted via Supernews, http://www.supernews.com Message-ID: <397671a7.79965887@news.pacificnet.net> References: <8kku84$o09$1@nnrp1.deja.com> <300DAB2B.6A5E836@mediaone.net> On Thu, 20 Jul 2000 01:11:02 GMT, Stephen Banville wrote: >clummis@volt.com wrote: > >> I found your resume on the Internet and thought it looked like a strong >> fit for our current contract opening for a Sr. Software Engineer >> >> .............. >What about contract work ?? > >--Steve > This poor headhunter simply lost his mind after finding so many highly professional SW developers in one place. A sort of headhunter's paradise,... at first glance. :) Max. Cambury Technologies, Inc embedded wireless systems http://www.cambury.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structures Date: Thu, 20 Jul 2000 03:44:48 GMT From: tangoNO_SPAM@pulp-fiction.com (Max) Organization: Posted via Supernews, http://www.supernews.com Message-ID: <397673ac.80482731@news.pacificnet.net> References: <39764060.5B64B6E4@lucent.com> <300DA7E5.E3890B42@mediaone.net> On Thu, 20 Jul 2000 00:57:07 GMT, Stephen Banville wrote: >Vividh Siddha wrote: > >> Can you pass structures as parameters to a taskSpawn call? There are >> 10 parameters of type int in taskSpawn which are used to passing >> parameters to the task function. Can we pass a structure?I can pass a >> pointer to the structure as a int but I would like to pass it as a >> structure because of some inter-task dependency. > >Sure, > Make sure you pass the address of the structure and cast it >to (UINT32 *). Is not this what we usually call "a pointer", Vividh did not want to pass to his task (for whatever strange reason he had)!? :) I guess the only luck passing a structure through 16/32 bit integer is if your structure is not longer that tose 16/32 bit. Four chars would do, so would 32 1-bit fields :) Seriously, I think that you have no choice but to push a pointer, or have some globally accessible structure and if you need pass a flag about the structure's state. Max. Cambury Technologies, Inc embedded wireless systems http://www.cambury.com --------------------------- Newsgroups: comp.os.vxworks Subject: GNU Zebra under Vxworks Date: Thu, 20 Jul 2000 12:12:19 -0300 From: Mauro Galarda Martins Organization: Lawrence Berkeley National Laboratory Message-ID: <8l78ib$pu7$1@overload.lbl.gov> Hello vxworkers Did anybody already try to port the GNU Zebra routing software under VxWorks ? thanks in advance Mauro - -- _______________________________________________________ Mauro Jose Galarda Martins Siemens Tel TR-DS Rua Pedro Gusso, 2635 CEP 81310-900 Curitiba - Parana' - Brasil Tel: +55 41 3416457 Fax: +55 41 3415620 e-mail: mauro.martins@siemens.com.br _______________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: RE: SCSI Minimum Period Date: Thu, 20 Jul 2000 09:26:33 +0200 From: ian.willats@telelogic.com Organization: Lawrence Berkeley National Laboratory Message-ID: <8l78ic$pu9$1@overload.lbl.gov> (Me again...) > We need clarification on the setting of 'minPeriod' in the SCSI_OPTIONS > structure when setting the transfer options for the interface. > minPeriod should be set to the number of 4ns units as defined by the SCSI-2 standard; it is internally scaled to real time by scsiLib and/or the SCSI drivers. Hence for 20 Mxfer/sec you should use a value of 12. My apologies if this isn't clear from the documentation. If this is not working (e.g. causes your SCSI system to hang) it is often because your cabling and/or termination may not be up to the job. At those speeds, parallel SCSI is extremely sensitive to noise and crosstalk. If you're using round cable - don't! I have found by trial and error that the best cable is flat ribbon cable with adjacent pairs twisted. Another approach is to start with a slower speed (larger minPeriod) to check that the synchronous negotiation is working properly - switch on scsiDebug to see this - and crank it up to a higher speed in stages. Ian Ian Willats Telelogic UK Ltd VxWorks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max Resources ? Date: Thu, 20 Jul 2000 18:41:51 +0530 From: Madhu Organization: Lucent Technologies, IPRC Message-ID: <3976FA96.9B5E5A6A@lucent.com> References: <3976A86D.93FE3DC0@lucent.com> <8l6dui$7qb$1@nnrp1.deja.com> Sorry for the HTML posting. We have tried here increasing the NUM_FILES but it does not seem to make a difference. How does the memory get allocated for these resources i.e the memory for the system data structures, probably might need some system cluster tuning. Any updates will be really helpful. Madhu john_94501@my-deja.com wrote: > Hmmm... posting the HTML wasn't nice, but to answer your questions: > > 1) The maximum number of files can be configured using NUM_FILES > (somewhere in the configuration files depending on the version of > Tornado you have). You should be aware that there is a maximum of 256 > file descriptors for select() though. If you need to use it (or rely on > any networking components that do) you shouldn't increase NUM_FILES > above 256. Oh, and sockets are file descriptors which is why I mentioned > this one. > > 2) There is no limit on the number of watchdogs as long as you don't run > out of memory. Likewise, semaphores, message queues and even tasks. > > Hope that helps, > > John... > --------------------------- Newsgroups: comp.os.vxworks Subject: Momory mapping files. Date: Thu, 20 Jul 2000 10:36:37 -0400 From: viv Organization: Lucent Technologies Message-ID: <39770E75.F7DC412@lucent.com> Is there a facility in VxWorks to memory map a file. I need to read 250 bytes from a file and pass it as packets to the physical layer. I was planning to memory map the file, keep reading 250 bytes at a time and send it.
If the file is a image can I read it in a buffer? How? --------------------------- Newsgroups: comp.os.vxworks Subject: clarification Date: Thu, 20 Jul 2000 22:25:43 +0530 From: Suresh T Organization: Lawrence Berkeley National Laboratory Message-ID: <01BFF299.72680140.suresht@future.futsoft.com> Hi, We are working on Strong arm ARMSA110 board. We are facing a problem in using the BSD function "bcopy_to_mbufs" for the application Software. It would be of great help if any one can clarify our doubt. The function "bcopy_to_mbufs" takes the following parameters bcopy_to_mbufs ( buffer, length, offset, ifnet *, width) and returns a pointer to the mbuf. The width field is specified as 2 for the 486 based PubStation and is working fine. We have a flat buffer like "ff 03 c0 21 01 01 00 11 05 06 04 18 a7 ed 07 02 08 02 0d 03 06" and we do a bcopy_to_mbufs in 486 environment. After this bcopy, if we print the mbuf, it is printing properly from the beginning. DATA[0] : 68 9b fe 0 1d 1e ea 71 0 60 1d 1e e9 e4 88 78 ff 3 c0 21 1 1 0 11 5 6 4 18 a7 ed 7 2 8 2 d 3 6 1 8 0 6 4 0 1 0 60 1d 1e e9 e4 a a 0 a 0 0 0 0 0 0 In the above output, "ff 3 c0 21" is the beginning of the PPP header and it is printed without any mistake. But in StrongARM environment, if we use width as 2 it is giving problem. If we pass a flat buffer like ff 03 c0 21 01 01 00 11 05 06 04 18 a7 ed 07 02 08 02 0d 03 06 with a data length of 21 and print the mbuf data in the Strong ARM Compiled code we have a difference in the representation of the data DATA[0] : 68 9b fe 0 1d 1e ea 71 0 60 1d f1 7b f7 88 78 c0 21 1 1 0 11 5 6 4 20 75 5b 7 2 8 2 d 3 6 ee 0 1 8 0 6 4 0 1 0 60 1d 1e e9 e4 a a 0 a 0 0 0 0 0 0 In this case, the beginning of PPP header is missing. (ff 3 is missed out). The problem being that after 16 bytes of mbuf header the bytes "ff 3" gets chopped off in the Strong ARM Code, which is actually our PPP Header. If we change the value of the width field in the function call to '1' , it works fine in the Strong ARM Compiled code. We want to know the purpose of the width parameter and whether the function call is processor dependent? We searched for the reference to the function call in the Tornado Manuals we couldn't find any useful information on this. We have another problem in handling the Mbuf buffer again between PPP and IP. The Mbuf is given from the PPP to IP and structure of the mbuf array is 24 9b fe 0 1d 1e ea 71 0 60 1d f1 7b f7 88 78 21 45 0 0 26 20 f 0 0 20 1 53 8b a 0 0 a0 a 0 12 9e 8 21 represents the ppp header and we strip of the PPP header before giving to IP by offsetting it to the next location, but the packets are not routed by the IP and badlen is reported by the IP. The actual data should be like this 45 0 0 26 20 f 0 0 20 1 53 8b a 0 0 a0 a 0 12 9e 8 0 92 ff 2 0 68 0 61 62 63 64 65 66 67 68 69 6a at this place before giving to the IP, if we do a flat buffer copy and again convert it to Mbuf and give to IP then it is working fine. So anyone ever faced this problem. Without the extra copy, if we try to send out some data packets and see the ipstatShow on the shell, the count for badlen gets incremented and ip as such doesn't produce error number, but discards the packet as badlen packet, but the packet is valid. Thanks and regards Suresh --------------------------- Newsgroups: comp.os.vxworks Subject: END Driver for 82559 Date: Thu, 20 Jul 2000 14:06:55 GMT From: "Laszlo Szebelledy" Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: <3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net> Hi All, I'm looking into converting the 82557 End Driver to the 82559 Driver for an IDT RC32364 target and am trying to get a handle on dynamic TBDs. Does anyone out there have experience with the 82559, is faimiliar with a driver for this device, or has done such a coversion? --------------------------- Newsgroups: comp.os.vxworks Subject: 'splnet' and 'spx' Date: Thu, 20 Jul 2000 18:20:06 +0200 From: Heribert Volkhausen Organization: T-Online Message-ID: <397726B6.814FAC06@t-online.de> Hi, I'm looking for some information about the functions 'splnet' and 'spx'. I can't find a decription anywhere. Regards Heribert Volkhausen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UDP problem Date: Thu, 20 Jul 2000 10:14:13 -0700 From: "David Oberbeck" Organization: StorageTek Message-ID: <8l78j9$pgs$1@news.stortek.com> References: <39760EC0.1EAEF7BF@139.85.103.45> <300DA502.63410AB6@mediaone.net> Greetings all, We, too, have been experiencing throughput problems on the VxWorks IP stack. Steve, I would be very interesting in knowing about how to bypass the VxWorks stuff... TIA, DGO > ico wrote: > >> Hi all, >> We are trying to run UDP bench mark tests for VxWorks on a >> 100baseT ethernet. We have only 2 machines on the network. The client >> sends a burst of 10 packets (500 bytes each) to a server, then waits >> for 20-50ms and sends another burst. On the server side there are four >> tasks(of same priority) running. The packets are transferred from one >> task to another with each task doing some processing on the packet >> before sending it out to a subsequent task. One of the tasks breaks up >> the 500 byte packet into 100 byte packets. The packets are then sent >> back to the client. The client measures the time it takes to get back >> the 500 byte packets(approx 1ms per 500 bytes). If we sent out 10,000 >> packets or so then we seem to be losing packets towards the end. Sending >> 100 packets doesn't seem to give any problems. Any help would be >> greatly appreciated. >> >> Thanks >> >> naveen > > Are you bench marking for performance reasons ?? If so there are other > means to send IP packets > through the physical layer other than using sockets. The VxWorks socket > layer is pretty slow. If you are interested I can tell how to go about > bypassing the vxWorks netWork layer for processing IP packets at a much > faster rate. > > --Steve > --------------------------- Newsgroups: comp.os.vxworks Subject: Potential Opportunity Date: Thu, 20 Jul 2000 12:41:21 -0400 From: "Jeff Newman" Organization: Posted via Supernews, http://www.supernews.com Message-ID: Hi! Jeff Newman from recruitDYNAMICS.com here. I have posted to this group before, but I checked the FAQ again, and it seems as if job postings are allowed. I have an ongoing assignment in Florida, initial contract is 7 months long, but will go longer. Hourly rate is based upon experience. The client is a Fortune 50, and the work being done is cutting edge. The job is to design and develop test cases for functional verification. The most important skill is VxWorks, and there are a number of secondary skills, but the right candidate has the option to be trained in what they don't know. We have multiple openings, and I would LOVE to make this happen for you. We have already placed one candidate and he is very happy. Lets make it happen! Drop me a note with a name and a contact number and we'll talk! Jeff Jeff Newman recruitDYNamics.com jbn@recruitdynamics.com www.recruitdynamics.com Get the right job at the right rate! Go to www.recruitdynamics.com! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: In-Circuit-Emulator for x86 Date: Thu, 20 Jul 2000 10:08:57 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <00151518.7b5864a0@usw-ex0101-007.remarq.com> References: <3975DE1A.A096129E@t-online.de> <8l66od$jak$1@news.hdpp.de> I've had very good luck with the Applied Microsystems SuperTap for the x86 on a past project -- the Coinstar coin counter. However, the project did not involve VxWorks (I rolled my own scheduler, drivers, etc.) so I cannot speak to that part of the problem. All my development was done using the SuperTap as the downloader and debugging tool for the unit under development and I was very happy with it -- so happy that (full disclosure follows...) I bought a small amount of stock in the company. Rob't Blumberg Embedded Software Developer Anything below is tacked on by Remarq! - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++ compiler for NetArm 40?? Date: Thu, 20 Jul 2000 20:35:40 +0100 From: molochai Message-ID: References: In message "Adam" wrote: > Are there any C++ compilers for the NetArm 40 that work with Tornado? GCC (g++) supports ARM. I think Green Hills have one too. It's not particularly difficult to use a different compiler with Tornado. > > thanks > -Adam > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Fri, 21 Jul 2000 08:48:23 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <10568ca8.ddf77ce9@usw-ex0101-007.remarq.com> References: <39780B42.8E1977C@lucent.com> <39783375.550BA547@dnaent.com> <39784414.10F03ABD@lucent.com> What I've done with other OSs is not to delete the task, but rather send it a message to clean itself up and delete itself. Of course, this implies that the task to be deleted has not "gone south" and is capable of receiving messages -- but of course, all your work is bug free;). Using this sort of a philosopy, one gets cleaner cleanups -- as development progresses, one often finds that there is more to do than just wind the stack and clean the heap at task completion. Rob't Blumberg Embedded Software Developer Anything below is tacked on by Remarq! - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: T1, Command Line, Memory Leak? Date: Thu, 20 Jul 2000 08:31:40 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8l6ddc$7er$1@nnrp1.deja.com> References: <39756C6B.E18F65A0@stest.ch> <8l60jj$trb$1@nnrp1.deja.com> <3976A7B2.EC585A37@stest.ch> Hi Mr Winton, In article <3976A7B2.EC585A37@stest.ch>, Andy Winton wrote: > Hi Mr 94501, :-) > > Thanks for your response. > > Another bit of information is that we are > running this on intel, and I've had reports > that this doesn't happen on PPC. To me > this just doesn't feel right, ah well. I can assure you it happens on all CPUs; I actually use PPC and ARM mostly so I can confirm the behaviour there, but it is part of the design of the shell (I think it is even documented somewhere...). Remember, the shell is meant to be a development tool, not something used in a deliverable product... Regards, John... > thanks again, > > andy > > john_94501@my-deja.com wrote: > > > > Hi Andy, > > > > No, this is not a bug. It is just the way the shells in VxWorks work. > > If you pass a string parameter to a function from the shell, that > > string needs to reside somewhere in memory so the shell allocates some > > space and puts it there. > > > > Unfortunately, it has no way of knowing when the string is finished with > > so it leaves it alone. You might think that when the function returns > > the string could be freed, but think about what happens if that string > > is a parameter to a task you spawn... > > > > Hope that explains it, > > > > John... > > > > In article <39756C6B.E18F65A0@stest.ch>, > > Andy Winton wrote: > > > hi all, > > > > > > When I look at the T1 'Launcher', then click on > > > 'Browser' then click on the 'clock' icon (to update > > > the stats). And I look at the 'Application' bar, > > > which I assume is the amount of memory allocated. > > > > > > It increments by 20 bytes every time I run - > > > > > > -> ping "127.0.0.1", 2 > > > PING 127.0.0.1: 56 data bytes > > > 64 bytes from localhost (127.0.0.1): icmp_seq=0. time=0. ms > > > 64 bytes from localhost (127.0.0.1): icmp_seq=1. time=0. ms > > > ----127.0.0.1 PING Statistics---- > > > 2 packets transmitted, 2 packets received, 0% packet loss > > > round-trip (ms) min/avg/max = 0/0/0 > > > value = 0 = 0x0 > > > -> > > > > > > Is this a memory leak? In the command line? > > > > > > (The memory 'lost' has the string '127.0.0.1' in it, > > > and if I make a void parameter function ping127 which > > > therefore doesn't use a string on the command line then > > > nothing is lost). > > > > > > Hmmmm, > > > > > > andy > > > -- > > > Andy Winton. > > > > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > > -- > Andy Winton. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: is somebody use PJWorks ? Date: Thu, 20 Jul 2000 10:40:47 +0200 From: Laurent Latreille Organization: Ensieta BREST Message-ID: <3976BB0E.4DB6FCFD@ensieta.fr> References: <20000719.8395300@boole.ensieta.fr> <39763D42.1905EB85@no-one.com> my problem is that the vxworks's shell create with the java support is very large ( ~ 1040000 bytes ) and i don't know how to do smaller shell. At compilation time, a message of error occurs because of that. laurent --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max Resources ? Date: Thu, 20 Jul 2000 08:40:52 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8l6dui$7qb$1@nnrp1.deja.com> References: <3976A86D.93FE3DC0@lucent.com> Hmmm... posting the HTML wasn't nice, but to answer your questions: 1) The maximum number of files can be configured using NUM_FILES (somewhere in the configuration files depending on the version of Tornado you have). You should be aware that there is a maximum of 256 file descriptors for select() though. If you need to use it (or rely on any networking components that do) you shouldn't increase NUM_FILES above 256. Oh, and sockets are file descriptors which is why I mentioned this one. 2) There is no limit on the number of watchdogs as long as you don't run out of memory. Likewise, semaphores, message queues and even tasks. Hope that helps, John... In article <3976A86D.93FE3DC0@lucent.com>, Madhu wrote: > > --------------8A9906BC81936E081590B506 > Content-Type: text/plain; charset=us-ascii > Content-Transfer-Encoding: 7bit > > > Hi, > > I had a question on the Resource Limit in vxWorks > > "If we hit the max resources limit, say for example : max no of > > sockets, No of Watchdog timers etc" IS there a way to increase this > > resource limit. If so what needs > > to be done i.e what header files need to be touched or what > > libraries etc need > > to be recompiled. > > > > If anybody has any answers for this, pl respond. > > Thankx > > Madhu > > > Lucent Technologies > IPRC, Bangalore > > --------------8A9906BC81936E081590B506 > Content-Type: text/html; charset=us-ascii > Content-Transfer-Encoding: 7bit > > > > > Hi, >
>       I had a question on the Resource > Limit in vxWorks >
>       "If we hit the max resources > limit, say for example : max no of >
> sockets, No of Watchdog timers etc" IS there a way to increase this >
> resource limit. If so what needs >
>       to be done i.e what header files > need to be touched or what >
> libraries etc need >
>       to be recompiled. >
> >
>       If anybody has any answers for > this, pl respond. >
> Thankx >
> Madhu >
> >
Lucent Technologies >
IPRC, Bangalore > > --------------8A9906BC81936E081590B506-- > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Task Pre-emption in vxWorks Date: Thu, 20 Jul 2000 08:57:13 GMT From: Urban Lindberg Organization: Deja.com - Before you buy. Message-ID: <8l6et3$8g5$1@nnrp1.deja.com> References: <8l4rn1$2s2$1@nnrp1.deja.com> <300DA473.B09EF347@mediaone.net> Hi! I Have also worked in VRTX before and in VRTX the ISR is invoked directly, but in vxWorks there is a wrapper that calls the function you provide in the intConnect call. This wrapper does all the stuff you have to do in the ISR for VRTX (pushing registers, tell the interrupt controller that you have serviced the interrupt etc.). The scheduling mechanism is however about the same in both. Urban In article <300DA473.B09EF347@mediaone.net>, Stephen Banville wrote: > kpulavar@my-deja.com wrote: > > > I have worked in VRTX before, but am sort of new in the vxWorks world. > > > > My question is: in VRTX ISRs there is a exit function called ui_exit, > > which invokes the scheduler to ensure that any high priority task(higher > > than the one interrupted) gets to run immediately after the ISR. > > > > I looked for such a mechanism in vxWorks, but did'nt find one in the > > manual. I thought the intConnect() function might provide the code > > to invoke the scheduler, but from the explanation in the manual, this > > appears unlikely. > > > > Anyone know how pre-emption by a higher priority task takes place ? > > > > Thanks in advance. > > > > -kris > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > > Sure, > Either by waking up a task that's pending on a semaphore or a > message. That's about it. > Tasks of the same priority wil run in a RR fashion. When a tak of a higher > priority gets woken up, it will run immediatley. From an ISR you can signal > a task by calling msgQSend. This is safest way to wake a high priority task > after servicing an interrupt. > > Best regards, > > --Stephen > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: use dos filesystem on ram disk Date: Fri, 21 Jul 2000 16:48:36 GMT From: lx1975@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8l9ust$js7$1@nnrp1.deja.com> I want to build dos filesystem on ram disk under tornado 2.0 env, and the following is my codes for building dos filesystem: BLK_DEV *pBlkDev; DOS_VOL_DESC *pVolDesc; dosFsInit(20); pBlkDev = ramDevCreate (0, 512, 400, 400, 0); pVolDesc = dosFsMkfs ("DEV1:", pBlkDev); And then I find I can create files on the filesystem, but can't read or write data to a file, why? int i; int fd = open("DEV1:km", O_CREAT, 0); if(fd>0) { char buf[100]; i = write(fd, "Hello", 5); close(fd); } Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: loading MVME2100 boot code Date: Fri, 21 Jul 2000 17:30:24 GMT From: John Fusco Organization: Deja.com - Before you buy. Message-ID: <8la1ao$m34$1@nnrp1.deja.com> References: <8l9vca$kau$1@nnrp1.deja.com> Did you define "FLASH_BOOT" in config.h? In my copy this is undef'd by default. If you don't define this, then romInit.s will copy code from 0xfff00100 instead of 0xff000100. In article <8l9vca$kau$1@nnrp1.deja.com>, unixmidiplugin@my-deja.com wrote: > I am having trouble loading VxWorks boot code onto > a Motorola MVME2100. > > I follow all the directions in the BSP for downloading > the boot code and programming the FLASH. Everything *seems* > to go OK, put when I install the J9 jumper, it won't boot. > Can someone who has done this show me where I'm going wrong? > > Here are screen dumps of my attemps : Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: bootrom_uncmp build errors Date: Fri, 21 Jul 2000 10:38:44 -0700 From: Ken Norlin Organization: NASA Dryden FRC Message-ID: <39788AA4.2561FAA@dfrc.nasa.gov> References: Sheela Kashyap wrote: > > Hi, > > I am trying to build an x86-based BSP on T2. My development machine has NT > 4.0 on it. I have integrated an fei8255xEnd driver into my BSP without any > problems. Now, I am trying to add an ncr810 scsi driver to this BSP along > with my ethernet driver. My makefile is the usual x86 makefile for BSP's > with the addition of object files to the LIB_EXTRA define - > > LIB_EXTRA = fei8255xEnd.out ncr810.out > Hi, I've never used your BSP, but I had a similar problem when trying to create a bootrom for my project. I was getting a bunch of undefined references. I ended up adding a define for MACH_EXTRA in the Makefile and putting my object code there instead of LIB_EXTRA. That worked for me. Ken N. - -- Ken Norlin --- Simulation Engineer NASA Dryden Flight Research Center ken.norlin@dfrc.nasa.gov (661) 276-2046 FAX (661) 276-2792 --------------------------- Newsgroups: comp.os.vxworks Subject: loading MVME2100 boot code Date: Fri, 21 Jul 2000 16:56:52 GMT From: unixmidiplugin@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8l9vca$kau$1@nnrp1.deja.com> I am having trouble loading VxWorks boot code onto a Motorola MVME2100. I follow all the directions in the BSP for downloading the boot code and programming the FLASH. Everything *seems* to go OK, put when I install the J9 jumper, it won't boot. Can someone who has done this show me where I'm going wrong? Here are screen dumps of my attemps : PPC5-Bug>niop Controller LUN =00? Device LUN =00? Get/Put =G? File Name =mv2100/boot.bin? fec/boot.bin.t2 Memory Address =00004000? Length =00000000? Byte Offset =00000000? Bytes Received =&423296, Bytes Loaded =&423296 Bytes/Second =&211648, Elapsed Time =2 Second(s) PPC5-Bug>md 4000 00004000 48000035 48000039 436F7079 72696768 H..5H..9Copyrigh 00004010 74203139 38342D31 39393920 57696E64 t 1984-1999 Wind PPC5-Bug>pflash 4000:fff00 ff000100 Source Starting/Ending Addresses =00004000/00103EFF Destination Starting/Ending Addresses =FF000100/FF0FFFFF Number of Effective Bytes =000FFF00 (&1048320) Program FLASH Memory (Y/N)? y Virtual-Device-Number =00 Manufacturer-Identifier =01 Device-Identifier =22DA Virtual-Device-Number =01 Manufacturer-Identifier =01 Device-Identifier =22DA Virtual-Device-Number =02 Manufacturer-Identifier =01 Device-Identifier =22DA Virtual-Device-Number =03 Manufacturer-Identifier =01 Device-Identifier =22DA Address-Mask =FFC00000 Erasing sector =$FF000000 Erasing sector =$FF040000 Erasing sector =$FF080000 Erasing sector =$FF0C0000 Programming sector =$FF000000 Programming sector =$FF040000 Programming sector =$FF080000 Programming sector =$FF0C0000 FLASH Memory Programming Complete PPC5-Bug>md ff000100 FF000100 48000035 48000039 436F7079 72696768 H..5H..9Copyrigh FF000110 74203139 38342D31 39393920 57696E64 t 1984-1999 Wind [ PAUSE HERE WHILE I INSTALL THE J9 JUMPER AND REBOOT ] : ? : ? : ? : TIA -- Larry Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: using the serial line with vxWorks Date: Thu, 20 Jul 2000 19:54:04 +0200 From: Elvis Organization: CESCA News service. Catalunya Message-ID: <39773CBC.B837981D@correu.udg.es> we have the following problem: when booting vxWorks from serial line, we can download programs to the target and then 'cut' the serial connection and use the serial line to communicate with the programs on the target. It works fine, but when we boot vxWorks from the ethernet, we can download programs to the target, but the serial line is no longer working. Are we missing some inizialization? where? can the serial line be software configured, or it's only possible to do that from the include files? Thank you very much. Dani & Elvis Students in Computer Science at University of Girona Spain --------------------------- Newsgroups: comp.os.vxworks Subject: Raw IP Date: 21 Jul 2000 17:58:00 GMT From: cedilnik andrej Organization: University of Maryland, Baltimore County Message-ID: <8la2v8$k4v$1@news.umbc.edu> Hello! Any idea on how to get network packets from hardware layer? Andy --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Fri, 21 Jul 2000 11:00:44 -0700 From: Bruce Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com> References: <39780B42.8E1977C@lucent.com> Amarnath Shanbhag wrote: >func() >{ > int* a = new int; > ....... > ....... > delete a; >} > >Now, from some other task, I delete the above task. >taskDelete(taskId); > >Now, what happens to the memory allocated for 'a' >in the task. How do we take care of freeing the >memory. In func(), I am deleting the memory allocated >to 'a', but if the task is deleted before the memory >is freed, will there be a memory leak ? How do we take >care to resolve this ? The easy way is to add clean up code to a function and have the OS call it via taskDeleteHookAdd() You could also do this: func() { int* a = new int; while(semTake( StopSema4, NO_WAIT == ERROR)) { DoStuff(); } // Clean up and go home delete a; // Suicide taskDelete(taskIdSelf()); } You other task just does a semGive() when it wants to kill this task. Bruce - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: VxDCOM Date: Fri, 21 Jul 2000 18:01:19 GMT From: wmiller_2002@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8la35f$ure$1@nnrp2.deja.com> I am looking at using vxDCOM for my application. Does anyone have any recent experience with it? Any feedback (good or bad) would be appreciated. thanks in advance Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Help with VxWorks Date: Thu, 20 Jul 2000 20:13:02 GMT From: "Said Abdosh" Message-ID: Hi All, I am very new to VxWorks. I wonder if someone has successfully ported Xalan-C++ to VxWorks. I am having problem with Template. It seems that the VxWorks does not understand the template. Do I have to include some kind of header file in order to make the VxWorks to understand the template? Thank you a lot /Said --------------------------- Newsgroups: comp.os.vxworks Subject: Fail to boot with multiple 82559 cards Date: Fri, 21 Jul 2000 18:23:01 GMT From: sbanavar@skystream.com Organization: Deja.com - Before you buy. Message-ID: <8la4df$okc$1@nnrp1.deja.com> Hello All, My setup is a PIII with pcPentiumpro BSP, vxWorks 5.3, Tornado 2.0 booting from the network. I have two 82559 cards on my system. When I use feiEnd driver, everything is just perfect with both the interfaces working. But the system does not boot when I change it to 82559 end driver (eeE) from Intel. More, once I build the BSP for one interface, the system boots just fine. Problem description: When bootConfig tries to load both the interfaces (muxDevLoad), the first instance loads correctly. When muxDevLoad is called for the second interface, it returns NULL. If I continue to boot the system, it hangs during ftpXfer. Hmm, more headache, I can see at the ftp server log that the target did login and changed the transfer mode to I, but it hung later. Has anyone faced anything like this. Any info/advice is greatly appreciated. TIA & Regards, Srinath Banavar SkyStream Networks. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: verbosity macro Date: Fri, 21 Jul 2000 21:29:30 +0200 From: Sebastien Fauris Organization: Club-Internet (France) Message-ID: <3978A49A.8ED28F97@club-internet.fr> Hi folks, I have a macro very convenient to debug my code. For that I compile my code in debug mode (flag DRV_DEBUG_MODE defined), and I use the following macro to display my debug messages with a certain level of verbosity that I can change (by a global variable): Macro: #ifdef DRV_DEBUG_MODE #define VERB_DRV(level,a1,a2,a3,a4,a5,a6,a7) \ if (level <= drv_dbg) \ {logMsg(a1,a2,a3,a4,a5,a6,a7);} #else #define VERB_DRV(level,a1,a2,a3,a4,a5,a6,a7) {} #endif and use it in my code: #ifdef DRV_DEBUG_MODE int drv_dbg = 0; /* default 0, but can be changed to a higher verbosity level after */ #endif int toto (void) { VERB_DRV (2,"my debug message displayed only in debug mode when drv_dbg global variable set higher or equal to 2\n",0,0,0,0,0,0); } I want to do the same macro, but using printf instead of logMsg. How can I do to pass the arguments to the printf (variable number of arguments) while logMsg arguments is fixed to 6 and this in a macro definition ? Thanks a lot, Sebastien. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using the serial line with vxWorks Date: 21 Jul 2000 19:35:51 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8la8mn$q27$1@news1.xs4all.nl> References: <39773CBC.B837981D@correu.udg.es> Elvis wrote: : we have the following problem: : when booting vxWorks from serial line, we can download programs to the : target and then 'cut' the serial connection and use the serial line to : communicate with the programs on the target. It works fine, but when we : boot vxWorks from the ethernet, we can download programs to the target, : but the serial line is no longer working. Are we missing some : inizialization? where? can the serial line be software configured, or : it's only possible to do that from the include files? : Thank you very much. Check usrSerial.c in target\config\comps\src. Also be sure to enable the serial support in your BSP. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: END Driver for 82559 Date: Thu, 20 Jul 2000 16:48:28 -0700 From: "Sheela Kashyap" Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net> I have the used the 82557 End Driver as is for the 82559 ethernet controller. I haven't had any problems. Sheela. "Laszlo Szebelledy" wrote in message news:3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net... > Hi All, > > I'm looking into converting the 82557 End Driver to the 82559 Driver for an > IDT RC32364 target and am trying to get a handle on dynamic TBDs. Does > anyone out there have experience with the 82559, is faimiliar with a driver > for this device, or has done such a coversion? > > --------------------------- Newsgroups: comp.os.vxworks Subject: End Driver Question ( motfcc ethernet driver ) Date: Thu, 20 Jul 2000 17:09:01 -0700 From: "Joydeep Roy" Message-ID: <0vMd5.110$9B3.77194@news01.micron.net> Problem: The End Ethernet driver ( motfcc ) is dropping packets because there not enough cluster pools even after cluster pools have been increased. Background: When 100 UDP packets are sent very quicky to a target system the ethernet driver in the target system drop about 60 packets. There is evidence that all the 100 packets are arriving at driver interface because the motFccReceive function in motFccEnd.c is being invoked 100 times. The system has nothing else running except the Ethernet motfcc end driver and the vxworks TCP/IP stack so there should be no issue that the driver is not getting enough cpu. After instrumenting the driver code, it was evident that inside motFccReceive() ( motfccEnd.c ), the NET_BUF_ALLOC was failing 50 to 60 times for every 100 packets coming in. #define NET_BUF_ALLOC netClusterGet(...); Even after increasing the MOT_FCC_RBD_NUM in sysMotFccEnd.c the test results were still the same i.e. for every 100 UDP packets sent 40 were being accepted by the driver. Granted that driver will drop some packets, but more than 50 % drop rate is unacceptable. We would like the number to be around 80 %. Looking forward for some solution/insight. Thanks roy --------------------------- Newsgroups: comp.os.vxworks Subject: Re: pciIntConnect and intConnect Date: Fri, 21 Jul 2000 08:47:12 +0800 From: "Ren Hantao" Organization: Alcatel Message-ID: <397872f2.0@dnews.sbell.com.cn> References: <39733E69.B07A8F3B@club-internet.fr> <01bff01d$f66b2cc0$140101c0@popp> <397758f5.0@dnews.sbell.com.cn> <39777099.388BAEEA@club-internet.fr> My target is MCP750, It's a PCI system card in mortorola CompactPCI chasis. "Sebastien Fauris" wrote in message news:39777099.388BAEEA@club-internet.fr... > Ren, what is your target ? In my case it is a PowerPC BSP. > I don't have to add anything to the IRQ value when I use intConnect. > > Sebastien. > > Ren Hantao wrote: > > > I use intConnect to connect handler to PCI interrupt. > > But I must add 0x10 to the intLine as the interrupt vector. > > (I use pciConfigInByte (1, deviceNum, 0, PCI_CFG_DEV_INT_LINE, &intLine) > > to get intLine). > > Does it needn't add 0x10 if I use pcIntConnect, and is it the > > only difference between IntConnect and pciIntConnect? > > > > Ren Hantao > > 2000/7/20 > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: pciIntConnect and intConnect Date: Thu, 20 Jul 2000 23:35:21 +0200 From: Sebastien Fauris Organization: Club-Internet (France) Message-ID: <39777099.388BAEEA@club-internet.fr> References: <39733E69.B07A8F3B@club-internet.fr> <01bff01d$f66b2cc0$140101c0@popp> <397758f5.0@dnews.sbell.com.cn> Ren, what is your target ? In my case it is a PowerPC BSP. I don't have to add anything to the IRQ value when I use intConnect. Sebastien. Ren Hantao wrote: > I use intConnect to connect handler to PCI interrupt. > But I must add 0x10 to the intLine as the interrupt vector. > (I use pciConfigInByte (1, deviceNum, 0, PCI_CFG_DEV_INT_LINE, &intLine) > to get intLine). > Does it needn't add 0x10 if I use pcIntConnect, and is it the > only difference between IntConnect and pciIntConnect? > > Ren Hantao > 2000/7/20 > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: for help:why my hook function can't work? Date: Thu, 20 Jul 2000 23:42:59 +0200 From: Sebastien Fauris Organization: Club-Internet (France) Message-ID: <39777263.5F2BEE7E@club-internet.fr> References: <8l66ku$m18$1@news.cz.js.cn> <8l6em8$86p$1@nnrp1.deja.com> yes, your ethernet driver must support the etherHook facility in order for the thing to work. Be aware that the driver may support only etherInputHookAdd, only etherOutputHookAdd or both. Sebastien. john_94501@my-deja.com wrote: > Hi, > > I suspect that the reason is the driver you are using doesn't support > the hook functionality. From the manual for etherHookAdd(): > > > A call to the function pointed to by the global function pointer > > etherInputHookRtn should be invoked in the receive routine of > > every network driver providing this service. > > Note, the last part which implies that it is up to the driver to support > the hook. You don't say which driver you are using though... maybe I can > check for you if you let me know... > > Hope that helps, > > John... > > --------------------------- Newsgroups: comp.os.vxworks Subject: about mbuf ! Date: Fri, 21 Jul 2000 02:47:12 GMT From: "sp hong" Organization: Dacom Message-ID: Hi ! I'm using Lan IC "cs8900a", and I have csEnd Driver for VxWorks 5.3.1 and Tornado 1.0.1 elf for SH-4(Hitachi). My problem is ftp error between my board and pc. ftp connection is OK. When I try to put some files to my board from pc with "put" ftp command, always that is right. I repeated above 100. ----> That's OK. But "get" ftp command occured some error. I tryed to get 100 files. But just I got a 58 files , and that there is mbuf error. I got a cluster pool table with "mbufShow" command. Let me show you that..... CLUSTER POOL TABLE ____________________________________________________________________________ ___ size clusters free usage - ---------------------------------------------------------------------------- - --- 64 100 62 165 128 100 100 168 256 40 40 4 512 40 40 0 1024 25 25 0 2048 25 25 24 - ---------------------------------------------------------------------------- - --- looks like mbuf free error. As you know free size is wrong(62). Actually "100" is right. I don't know why does it happen. /* csEnd.c - Crystal Semiconductor CS8900 network interface driver */ /* Copyright 1999 Crystal Semiconductor Corp. */ /* Last release: v3.02a */ That's all. If somebody know how to confirm that mbuf is ok, let me know that. Thanks...................... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: loading MVME2100 boot code Date: Fri, 21 Jul 2000 20:37:51 GMT From: unixmidiplugin@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lacat$ut1$1@nnrp1.deja.com> References: <8l9vca$kau$1@nnrp1.deja.com> <8la1ao$m34$1@nnrp1.deja.com> In article <8la1ao$m34$1@nnrp1.deja.com>, John Fusco wrote: > Did you define "FLASH_BOOT" in config.h? In my copy this is undef'd by > default. If you don't define this, then romInit.s will copy code from > 0xfff00100 instead of 0xff000100. Thank you very much , John. That was the answer! Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Socket closing on VXWORKS. Date: Fri, 21 Jul 2000 21:18:01 GMT From: Mangesh Joshi Message-ID: <3978BE07.EB5FBCB4@mitel.com> Hello All, I am working on a application which need to open and close sockets a finite number of times per min. My problem is that I am not able to close sockets fast enough. The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and stays there for at least a min before it actually closes. This pushes me to have more socket/buffers available to keep up the throughput. Is there any way to reduce the closing time to a few seconds. Thanks, Mangesh Joshi. --------------------------- Newsgroups: comp.os.vxworks Subject: cacheFlush() on Motorola MPC860P Date: 21 Jul 2000 20:38:23 GMT From: "Danielp Parent" Organization: Chorus Netwoks Message-ID: <01bff318$ef9139a0$ef11400a@foo> Sender: reguser@frankfurter.heurikon.com Does anyone if WRS has fixed the cacheFlush() problem on the MPC860P yet? Motorola increased the size of the cache on the 860P and WRS failed to update their code. Thanks, Dan --------------------------- Newsgroups: comp.os.vxworks Subject: ppc2799: RAM_HIGH_ADDRS=0x20,A000 will not boot Date: Thu, 20 Jul 2000 20:10:34 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397794FA.8C0BA660@mindspring.com> I got the old "change RAM_HIGH_ADRS" msg during build. I was using the "as shipped" value of 0x20,0000. I was a little over 32k short so I changed RAM_HIGH_ADRS to 0x20,A000. Built and "flashed" fine, but the bootstrap would not boot. The the second read light stayed on, and nothing ever went to COM1. I then tried a value of 0x30,0000 and this worked fine. Anybody know what rule I broke when I used 0x20,A000? thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Sat, 22 Jul 2000 00:31:16 GMT From: birajdar2250@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8laq0l$8nt$1@nnrp1.deja.com> References: <39780B42.8E1977C@lucent.com> <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com> Bruce -> clean up can be done by using taskDeleteHookAdd(). Also you can use taskSafe() to prevent the task from being deleted during critcal section. So if you are in a tight memory situation this will help. In article <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com>, Bruce wrote: > Amarnath Shanbhag wrote: > > >func() > >{ > > int* a = new int; > > ....... > > ....... > > delete a; > >} > > > >Now, from some other task, I delete the above task. > >taskDelete(taskId); > > > >Now, what happens to the memory allocated for 'a' > >in the task. How do we take care of freeing the > >memory. In func(), I am deleting the memory allocated > >to 'a', but if the task is deleted before the memory > >is freed, will there be a memory leak ? How do we take > >care to resolve this ? > > The easy way is to add clean up code to a function and have the > OS call it via taskDeleteHookAdd() > > You could also do this: > > func() > { > int* a = new int; > while(semTake( StopSema4, NO_WAIT == ERROR)) > { > DoStuff(); > } > // Clean up and go home > delete a; > > // Suicide > taskDelete(taskIdSelf()); > > } > > You other task just does a semGive() when it wants to kill this > task. > > Bruce > > ----------------------------------------------------------- > > Got questions? Get answers over the phone at Keen.com. > Up to 100 minutes free! > http://www.keen.com > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Other RTOS's? Date: Fri, 21 Jul 2000 04:36:48 GMT From: tangoNO_SPAM@pulp-fiction.com (Max) Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3977d11f.1685376@news.pacificnet.net> References: <396FEC6A.599C6D58@mindspring.com> In my personal opinion, if you have to come with your product fast, there is probably no alternative to vxWorks, especially if you're getting BSP/drivers. If you have some time on your hands, there are definetely less expensive and more controlable and easy maintainable (remember, this is my personal opinion :) RTOSes/toolsets to develop such a system. Even without mentioning the "source code availabitity" issue for code coverage analysis to sertify your recorder to a proper flight criticality level. Max. On Thu, 20 Jul 2000 22:21:30 -0400, "Robert" wrote: >Interesting discussion... > >We are currently evaluating VxWorks Tornado and got a quote from the >WindRiver guys. I'm about to spend $75K on two developer seats with a bunch >of tools, the OEM license and 25 target licenses. > >Our project is an embedded Flight Data Recorder to be installed in military >aircraft. The system includes video (MPEG-2 encoding) and a GUI (eventually >running on Zinc). Hardware platform is a rugged VME board with on-board MPEG >encoder. Our hardware manufacturer provides as with a BSP and drivers for >Tornado. > >Is it worth the money we are spending on VxWorks or should we go with a >different product? > >Thanks for your input. > >Robert > > >"Al Johnston" wrote in message >news:396FEC6A.599C6D58@mindspring.com... >> We are considering abandoning vxWorks (actually, vxWorks is not that >> bad, WRS is the problem). I would be interested in any opinions/ >> comments / pointers concerning other alternatives. H/W is moto 2700. >> >> Considerations would be ease of migration, source availability, >> quality / effectiveness of tech support, ethical operation, >> only selling products that exist (or at least returning the funds >> when advising the customer that the product doesn't exist), Being >> able to decide if a product sold to a customer does or does not >> exist in less then nine months, ability to answer a simple >> question in less then 3 years, being able to deliver an evaluation >> copy in less then three months... woops.... sorry. I got a little >> off topic there. >> >> Seriously, suggestions? tnx. >> >> (PS: quite laughing Kelly, or I will put in a request that you be >> pulled in to help us USE vxWorks!) > > Cambury Technologies, Inc embedded wireless systems http://www.cambury.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Socket closing on VXWORKS. Date: Sat, 22 Jul 2000 00:42:51 GMT From: birajdar2250@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8laqm9$9ap$1@nnrp1.deja.com> References: <3978BE07.EB5FBCB4@mitel.com> Try this: Set the SO_LINGER option of the socket with small linger time. This will close the connection immediately. Not sure if it will free the buffers as quick as you want. Let me know if it does. In article <3978BE07.EB5FBCB4@mitel.com>, Mangesh Joshi wrote: > Hello All, > I am working on a application which need to open and close sockets a > finite number of times per min. My problem is that I am not able to > close sockets fast enough. > The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and > stays there for at least a min before it actually closes. This pushes me > to have more socket/buffers available to keep up the throughput. > Is there any way to reduce the closing time to a few seconds. > Thanks, > > Mangesh Joshi. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: who owns this semaphore? Date: Fri, 21 Jul 2000 22:08:03 GMT From: tracyajones5487@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lahk4$5f4$1@nnrp2.deja.com> I am trying to find the tasks that is holding this semaphore. Is there a way to find out? - -> semShow(0x3ef4ca0,1) Semaphore Id : 0x3ef4ca0 Semaphore Type : BINARY Task Queuing : PRIORITY Pended Tasks : 3 State : EMPTY NAME TID PRI TIMEOUT - ---------- -------- --- ------- t1 3fb59a0 100 0 OOOPortFactory 3ee05e8 132 0 OOOOutPort7 3e5d800 133 0 Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How MCP750 share it's local memory to PCI bus? Date: Fri, 21 Jul 2000 15:47:32 -0700 From: James Chen Organization: Globix Corp. Message-ID: <3978D303.328751BC@gigabitwireless.com> References: <39777044.0@dnews.sbell.com.cn> This is a multi-part message in MIME format. - --------------C3CAB6FC8AF6BD639056FC9C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The MPC750 doesn't have a built-in PCI interface. Hopefully, there is a PCI controller sits between the microprocessor and the memory. It is in the PCI configuration register space of this PCI controller where you perform auto-configure cycle and map the local address onto the PCI memory space as a slave device. Once enabled, the PCI controller shall respond to PCI bus transactions when the address on the bus falls into its mapped range. Ren Hantao wrote: > Hi, > I can access shared memory on CPCI peripheral card from > MCP 750 now? But I don't know how can MCP750 share it's > local memory to PCI bus and let other PCI device to access. > Could anybody tell me the detailed approach? > > Ren Hantao > 2000/7/20 - --------------C3CAB6FC8AF6BD639056FC9C Content-Type: text/x-vcard; charset=us-ascii; name="jchen.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for James Chen Content-Disposition: attachment; filename="jchen.vcf" begin:vcard n:Chen;James tel;fax:408.577.0700 tel;work:408.577.7411 x-mozilla-html:FALSE url:www.gigabitwireless.com org:Gigabit Wireless, Inc;BTS Software Development version:2.1 email;internet:jchen@gigabitwireless.com adr;quoted-printable:;;3099 N. 1st Street,=0D=0A;San Jose;California;95134; fn:James Chen end:vcard - --------------C3CAB6FC8AF6BD639056FC9C-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How MCP750 share it's local memory to PCI bus? Date: Fri, 21 Jul 2000 15:47:40 -0700 From: James Chen Organization: Globix Corp. Message-ID: <3978D30C.89E4212@gigabitwireless.com> References: <39777044.0@dnews.sbell.com.cn> This is a multi-part message in MIME format. - --------------E4C589B69F8895C46F1328E9 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The MPC750 doesn't have a built-in PCI interface. Hopefully, there is a PCI controller sits between the microprocessor and the memory. It is in the PCI configuration register space of this PCI controller where you perform auto-configure cycle and map the local address onto the PCI memory space as a slave device. Once enabled, the PCI controller shall respond to PCI bus transactions when the address on the bus falls into its mapped range. Ren Hantao wrote: > Hi, > I can access shared memory on CPCI peripheral card from > MCP 750 now? But I don't know how can MCP750 share it's > local memory to PCI bus and let other PCI device to access. > Could anybody tell me the detailed approach? > > Ren Hantao > 2000/7/20 - --------------E4C589B69F8895C46F1328E9 Content-Type: text/x-vcard; charset=us-ascii; name="jchen.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for James Chen Content-Disposition: attachment; filename="jchen.vcf" begin:vcard n:Chen;James tel;fax:408.577.0700 tel;work:408.577.7411 x-mozilla-html:FALSE url:www.gigabitwireless.com org:Gigabit Wireless, Inc;BTS Software Development version:2.1 email;internet:jchen@gigabitwireless.com adr;quoted-printable:;;3099 N. 1st Street,=0D=0A;San Jose;California;95134; fn:James Chen end:vcard - --------------E4C589B69F8895C46F1328E9-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: SSLeay or OpenSSH Date: Fri, 21 Jul 2000 06:36:59 GMT From: Computer Engineer Organization: Somewhere in the Earth Message-ID: <39777EC7.933E4EAA@no-one.com> References: <39778B18.D28E0230@radixtek.com> According to the Wind Web Server 2.0 Release Notes, SSL is not supported under this release. Eric Monsler wrote: > > All, > > The description of using the WindWeb server describes how to incorporate > SSL functionality into the webserver by downloading and building SSLeay > for the target architecture. > > Has anybody done this successfully? > > Has anybody used SSLeay for other VxWorks applications? > > And finally: > > Has anyone built and used OpenSSH for VxWorks, either as an rlogin > replacement or to secureley tunnel another protocol? > > Thanks, > > Eric M. Monsler --------------------------- Newsgroups: comp.os.vxworks Subject: Semaphores and Mutexes? Date: Fri, 21 Jul 2000 10:37:33 +0530 From: Madhu Organization: Lucent Technologies, IPRC Message-ID: <3977DA94.1D408B74@lucent.com> Hi, I wanted to know if a binary semaphore can have recursive access to resources. Like the code segment Func2() { Semtake(binary-semphore) ... Func1() Semgive(binary-semaphore) } Func2() { Semtake(binary-semphore) ..... Semgive(binary-semaphore) } In the above case the same binary semaphore is taken twice. Will this work.... This is guarenteed to work for a mutex. The manuals say that the mutex actually is implemented as a binary sempahore... Any help would be highly appreciated. Thankx Madhu Lucent Technologies. IPRC, Bangalore --------------------------- Newsgroups: comp.os.vxworks Subject: taskDelete in VxWorks Date: Fri, 21 Jul 2000 14:05:14 +0530 From: Amarnath Shanbhag Organization: Lucent Technologies Message-ID: <39780B42.8E1977C@lucent.com> Hello, Suppose we do a taskId = taskSpawn(.., func(), ..); func() is the function to be called when then task is spawned. In func(), I do the following - func() { int* a = new int; ....... ....... delete a; } Now, from some other task, I delete the above task. taskDelete(taskId); Now, what happens to the memory allocated for 'a' in the task. How do we take care of freeing the memory. In func(), I am deleting the memory allocated to 'a', but if the task is deleted before the memory is freed, will there be a memory leak ? How do we take care to resolve this ? Thanks, Amarnath --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Semaphores and Mutexes? Date: Fri, 21 Jul 2000 11:31:56 -0500 From: "Ranga Mc Donald" Organization: Esat Net Customer Message-ID: <39782512.0@seat.s3two.ie> References: <3977DA94.1D408B74@lucent.com> Reply-To: "Ranga Mc Donald" Actually recursive gives a two different meanings in my opinion. We've a driver routines called by the application indeed allows you to call the function which owns the mutex recursively before it's been released finally by the latter function or the function which owns the mutex. I don't come across with any kind of situation where binary semaphore does this job. It is an exclusive feature to the mutex, I presume. You may see more views on this news group from other experienced guys! "Madhu" wrote in message news:3977DA94.1D408B74@lucent.com... > Hi, > I wanted to know if a binary semaphore can have recursive access to > resources. > Like the code segment > > Func2() > { > Semtake(binary-semphore) > ... > Func1() > > Semgive(binary-semaphore) > } > > Func2() > { > Semtake(binary-semphore) > ..... > Semgive(binary-semaphore) > } > > In the above case the same binary semaphore is taken twice. Will > this work.... > This is guarenteed to work for a mutex. The manuals say that the > mutex actually > is implemented as a binary sempahore... > > Any help would be highly appreciated. > Thankx > Madhu > > Lucent Technologies. > IPRC, Bangalore > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sysFlashWrite() Date: Thu, 13 Jul 2000 12:56:40 -0400 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <396DF4C8.986F0A34@lucent.com> References: <396CF04C.EE83CEAD@lucent.com> Vividh Siddha wrote: > > My vxWorks image doesn't contain the function sysFlashWrite() for > writing data to flash memory on my PPC card. Do I have to build the > image with including something to have this? > > Thanks, > Vividh Did you define INCLUDE_FLASH in your config.h??? sysFlashWrite is defined in a file in src/drv/mem that is included by sysLib.c in your BSP It is only included and compiled into the BSP when INCLUDE_FLASH is defined in the config. - -- Doug --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Fri, 21 Jul 2000 06:26:45 -0500 From: Francisco Pataro Organization: Verio Message-ID: <39783375.550BA547@dnaent.com> References: <39780B42.8E1977C@lucent.com> After delete a; taskDelete(0); will do it. Francisco Amarnath Shanbhag wrote: > Hello, > > Suppose we do a > taskId = taskSpawn(.., func(), ..); > > func() is the function to be called when then task is spawned. > In func(), I do the following - > > func() > { > int* a = new int; > ....... > ....... > delete a; > } > > Now, from some other task, I delete the above task. > taskDelete(taskId); > > Now, what happens to the memory allocated for 'a' > in the task. How do we take care of freeing the > memory. In func(), I am deleting the memory allocated > to 'a', but if the task is deleted before the memory > is freed, will there be a memory leak ? How do we take > care to resolve this ? > > Thanks, > Amarnath --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Fail to boot with multiple 82559 cards Date: Sat, 22 Jul 2000 18:52:25 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lcqh4$ik2$1@nnrp1.deja.com> References: <8la4df$okc$1@nnrp1.deja.com> Hi, In article <8la4df$okc$1@nnrp1.deja.com>, sbanavar@skystream.com wrote: > Hello All, > > My setup is a PIII with pcPentiumpro BSP, vxWorks 5.3, Tornado 2.0 > booting from the network. > > I have two 82559 cards on my system. When I use feiEnd driver, > everything is just perfect with both the interfaces working. But the > system does not boot when I change it to 82559 end driver (eeE) from > Intel. More, once I build the BSP for one interface, the system boots > just fine. If it works with feiEnd, but not with eeE then why not stick to feiEnd? Is there some other reason you want to change from the WRS driver? Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: END Driver for 82559 Date: Sat, 22 Jul 2000 19:13:23 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lcro9$jd7$1@nnrp1.deja.com> References: <3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net> Hi, The WRS 82557 END driver is a driver for 82557, 82558 and 82559 parts. No conversion should be necessary. HTH, John... In article <3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net>, "Laszlo Szebelledy" wrote: > Hi All, > > I'm looking into converting the 82557 End Driver to the 82559 Driver for an > IDT RC32364 target and am trying to get a handle on dynamic TBDs. Does > anyone out there have experience with the 82559, is faimiliar with a driver > for this device, or has done such a coversion? > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How MCP750 share it's local memory to PCI bus? Date: 22 Jul 2000 16:01:22 GMT From: bowtie@netcom.com (Bill Dennen) Organization: MindSpring Enterprises Message-ID: <8lcggi$si8$1@slb6.atl.mindspring.net> References: <39777044.0@dnews.sbell.com.cn> <3978D303.328751BC@gigabitwireless.com> James Chen writes: The MPC750 is the processor on the MCP750 SBC. The MCP750 interfaces to the cPCI bus using a 21150 bridge. I'm not familiar with the VxWorks setup but it should be in config.h. Check the target.txt|nr documentation in the BSP directory. Regards >@< >This is a multi-part message in MIME format. >--------------C3CAB6FC8AF6BD639056FC9C >Content-Type: text/plain; charset=us-ascii >Content-Transfer-Encoding: 7bit >The MPC750 doesn't have a built-in PCI interface. Hopefully, there is a >PCI controller sits between the microprocessor and the memory. It is in >the PCI configuration register space of this PCI controller where you >perform auto-configure cycle and map the local address onto the PCI >memory space as a slave device. Once enabled, the PCI controller shall >respond to PCI bus transactions when the address on the bus falls into >its mapped range. >Ren Hantao wrote: >> Hi, >> I can access shared memory on CPCI peripheral card from >> MCP 750 now? But I don't know how can MCP750 share it's >> local memory to PCI bus and let other PCI device to access. >> Could anybody tell me the detailed approach? >> >> Ren Hantao >> 2000/7/20 >--------------C3CAB6FC8AF6BD639056FC9C >Content-Type: text/x-vcard; charset=us-ascii; > name="jchen.vcf" >Content-Transfer-Encoding: 7bit >Content-Description: Card for James Chen >Content-Disposition: attachment; > filename="jchen.vcf" >begin:vcard >n:Chen;James >tel;fax:408.577.0700 >tel;work:408.577.7411 >x-mozilla-html:FALSE >url:www.gigabitwireless.com >org:Gigabit Wireless, Inc;BTS Software Development >version:2.1 >email;internet:jchen@gigabitwireless.com >adr;quoted-printable:;;3099 N. 1st Street,=0D=0A;San Jose;California;95134; >fn:James Chen >end:vcard >--------------C3CAB6FC8AF6BD639056FC9C-- - -- >@< bowtie@netcom.com Bill Dennen william.dennen@motorola.com Motorola Computer Group, TOPAZ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Semaphores and Mutexes? Date: Sat, 22 Jul 2000 19:22:53 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lcsa1$jou$1@nnrp1.deja.com> References: <3977DA94.1D408B74@lucent.com> Hi, The short answer to your question is no. If you want resource locking use a mutex (that is what they are designed for). Binary semaphores are for synchronisation. Now, looking at your example, the sequence of calls you would make is: semTake semTake semGive semGive Looking at the state of the semaphores (and we can compare all three types here, counting, binary and mutex) we see this: Operation Binary Counting Mutex ========= ====== ======== ===== Initially FULL 2 UNLOCKED semTake EMPTY 1 LOCKED (1) semTake 0 LOCKED (2) semGive N/A 1 LOCKED (1) semGive N/A 2 UNLOCKED Note that giving a binary semaphore more than once is OK, but that the state can only ever be FULL or EMPTY so it will still only take a single semTake to make it EMPTY. So, when do we use the three types? Well, here's some examples: Binary:- in an ISR to signal a task that it needs to do something - in a task to signal that an operation has completed Counting:- when a limited number of resources are available and being shared between tasks Mutex:- to lock critical regions of code. Hope that makes it clearer. For more information I would suggest you read at least the VxWorks Programmer's Guide, and perhaps even some books on operating system basics. Regards, John... In article <3977DA94.1D408B74@lucent.com>, Madhu wrote: > Hi, > I wanted to know if a binary semaphore can have recursive access to > resources. > Like the code segment > > Func2() > { > Semtake(binary-semphore) > ... > Func1() > > Semgive(binary-semaphore) > } > > Func2() > { > Semtake(binary-semphore) > ..... > Semgive(binary-semaphore) > } > > In the above case the same binary semaphore is taken twice. Will > this work.... > This is guarenteed to work for a mutex. The manuals say that the > mutex actually > is implemented as a binary sempahore... > > Any help would be highly appreciated. > Thankx > Madhu > > Lucent Technologies. > IPRC, Bangalore > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Priority of the printf task Date: Sat, 22 Jul 2000 19:28:26 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lcskd$k2j$1@nnrp1.deja.com> References: <3977525c.0@seat.s3two.ie> Hi, In article <3977525c.0@seat.s3two.ie>, "Ranga Mc Donald" wrote: > How to get the priority of the task 'printf()' in the vxworks? > Anybody have any idea??? printf() is not a task. It is a function that can be called by tasks. Consequently, it has no task priority. Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Memory mapped files Date: Sat, 22 Jul 2000 19:34:03 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lcsvb$k4s$1@nnrp1.deja.com> References: <39774BEC.6FEF0D6D@lucent.com> Hi, Quick note first - posting HTML will not make you popular in the usenet community - plain text works much better! Anyway, onto your query... VxWorks doesn't have an mmap() facility, so you would have to work with normal file operation functions (e.g. open, close, read, write). Now, you should also be aware of the properties of the file system you are using... depending on how the file system works (e.g. whether it caches files in local memory or not), you might be able to simply use read and then re-locate the file pointer to the start of the file for the next time. Regards, John... In article <39774BEC.6FEF0D6D@lucent.com>, viv wrote: > > > Is there a facility in VxWorks to memory map a file. I need to read 250 > bytes from a file and pass it as packets >
to the physical layer. I was planning to memory map the file, keep > reading 250 bytes at a time and send it. >
If the file is a image can I read it in a buffer? How? > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max Resources ? Date: Sat, 22 Jul 2000 19:40:31 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lctbd$kgj$1@nnrp1.deja.com> References: <3976A86D.93FE3DC0@lucent.com> <8l6dui$7qb$1@nnrp1.deja.com> <3976FA96.9B5E5A6A@lucent.com> Hi, In article <3976FA96.9B5E5A6A@lucent.com>, Madhu wrote: > Sorry for the HTML posting. We have tried here increasing the NUM_FILES > but it does not seem to make a difference. How does the memory get allocated > > for these resources i.e the memory for the system data structures, probably > might > need some system cluster tuning. The memory for the fd table is allocated when the kernel is initialised. You shouldn't need to "tune" anything to increase the number of file descriptors that can be open at any one time, other than re-build your kernel having changed NUM_FILES of course. As for other system data structures, they are allocated dynamically using the equivalent of malloc() from the system memory partition (the heap). You can see if you are running out of memory using memShow(). How many objects are you creating, and how much memory does your board have? Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: alt.engineering,alt.engineering.electrical,comp.arch.embedded,comp.os.vxworks Subject: Re: Motorola is obsoleting MVMExxx ! Date: Fri, 21 Jul 2000 22:28:32 -0700 From: _________TOM_______ Message-ID: <39793100.52E@cybertrails.com.invalid> References: <4Rt45.12268$3D1.193418@news1.frmt1.sfba.home.com> <8j0k2p$1j9i@drn.newsguy.com> Reply-To: NOtomSPAMscott@cybertrails.com.invalid wilhardt@mcg.mot.com wrote: All, a.) The MVME147 is NOT being obsoleted but is getting a new FPGA to replace the PCC ASIC. This change is 100% compatible. ---------------------------------------------------------- Motorola ..... OT , but can you help me find a NewsGroup to the Mot' 68HC11 ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about mbuf ! Date: 22 Jul 2000 20:20:17 GMT From: "Klaus Popp" Message-ID: <01bff41a$9af90000$140101c0@popp> References: sp hong schrieb im Beitrag ... > Hi ! > I'm using Lan IC "cs8900a", and I have csEnd Driver for VxWorks 5.3.1 > and Tornado 1.0.1 elf for SH-4(Hitachi). > ... > But "get" ftp command occured some error. > I tryed to get 100 files. But just I got a 58 files , and that there is mbuf > error. >... I'm not sure if that is your problem, but we found a bug in cs8900a driver some month ago. The problem was that the MTU size was some bytes too high (was 1518 rather 1514) and that causes trouble with NFS. After we fixed MTU size everything was ok. Sorry, but I don't have the revision number of the driver here as I'm not in the office. Klaus --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PowerPc /vxWorks Interrupt latency Date: Tue, 11 Jul 2000 08:51:27 -0700 From: Ken Norlin Organization: NASA Dryden FRC Message-ID: <396B427F.63625D64@dfrc.nasa.gov> References: <39635D0F.9B7FC276@usa.alcatel.com> <3964CFE8.F8D98D78@mail.ford.com> <39669DFE.205A76EA@aps.anl.gov> <8kbrje$37s$1@nnrp1.deja.com> <396A76EC.61A79BE2@aps.anl.gov> Andrew Johnson wrote: > > Unfotunately it's not clear to me whether the above would be possible. In > a BSP patch Motorola moved the Universe int clear to after running the > ISR, with the comment "fix potential RORA VME interrupt problem". The MPIC > int clear also occurs after running the ISR. It's possible that this > could be fixed by reducing the modularity of the code, but I'd want to > understand it a bit more before doing that, and we don't have the need for > it here. > Hi Andrew, Where did you find this BSP patch? -- I'm having a problem when mixing a RORA & ROAK interrupter in the same chassis. When they interrupt at close to the same time, I get a VME bus error. I've searched the Wind River site for patches, but I haven't seen anything that refers to this. Thanks, Ken N. - -- Ken Norlin --- Simulation Engineer NASA Dryden Flight Research Center ken.norlin@dfrc.nasa.gov (661) 276-2046 FAX (661) 276-2792 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Raw IP Date: Sat, 22 Jul 2000 13:07:45 +0100 From: "Cichy" Organization: ACI - Szczecin http://aci.com.pl Message-ID: <8lbrur$hhk$1@zeus.man.szczecin.pl> References: <8la2v8$k4v$1@news.umbc.edu> > Hello! > > Any idea on how to get network packets from hardware layer? > > Andy > etherLib.h etherInputHookAdd( ) ? Best regards --------------------------- Newsgroups: comp.os.vxworks Subject: a cache problem of PPC8260 Date: Sat, 22 Jul 2000 13:32:37 GMT From: zhouxiaochen@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lc7pl$69g$1@nnrp1.deja.com> Hi,vxworkers, When I add USER_I_CACHE_ENABLE,USER_D_CACHE_ENABLE, INCLUDE_CACHE_SUPPORT, and in sysPhysMemDesc[], I change RAM on 60X bus to VM_STATE_CACHEABLE, but the speed is same! It seems cache does no work! Does anybody can explain it? Thanks! (Tornado2.0 VxWorks5.4 PPC8260) Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: 750ppc cpu instruction times? help Date: Sat, 22 Jul 2000 17:34:43 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397A1373.AA5C3E66@mindspring.com> I am trying to understand the unusual coding in the 2700bsp's ppcDecTimer.c's sysClkInt. It looks like the reload value for the DEC register is skewed 1 hardware tick (60nS) for each missed (unserviced) rollover event. The code: mfdec R3 foo: add. R3,277000,R3 ble foo mtdec R3 My theory regarding this skews are as follows: 1) I am misreading the code (skew not really there), or 2) The code has a bug (skew is there, but should not be), or 3) The skew is meant to account for the additional time spent executing add./ble additional times. I am perusing the 3rd possibility (because it explain the facts in the simplest and elegant manner while giving the bsp coder the benefit of the doubt - after all, he knows what he is doing, while I do not!). In the good old days I could look each of these instructions up in a table to get how long they take. Life is much more complicated now. Is there an easy way of doing this? I tried just using an inst/sec rate and got numbers that were inconsistent with the code. A factor to remember is that this is an ISR. The times I am looking for are: TI : t(int latency) + t(mfdec) + t(add.) Tn : t(ble-branch taken) + t(add.) TF : t(ble-branch not taken) + t(mtdec) If the dec int was delivered "on time" the total time would be TT = TI+TF If the dec int was late by n rollovers, the total time would be TT = TI + (n * TN) + TF Short of actually running some tests, can anyone help me figure out what these times should be? If there is interest in the details of why I think there may be a problem I will be happy to explain, but It is fairly difficult to express. Basically it has to do with the fact that the "reported time" is being skewed by 1 tic (60ns) for each rollover interrupt that has been missed. Card is mvme2700, with a 750 cpu running at 366MHz. thanks guys. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Priority of the printf task Date: Sat, 22 Jul 2000 22:05:05 GMT From: tangoNO_SPAM@pulp-fiction.com (Max) Organization: Posted via Supernews, http://www.supernews.com Message-ID: <397a17f5.48140125@news.pacificnet.net> References: <3977525c.0@seat.s3two.ie> <8lcskd$k2j$1@nnrp1.deja.com> On Sat, 22 Jul 2000 19:28:26 GMT, john_94501@my-deja.com wrote: >Hi, > >In article <3977525c.0@seat.s3two.ie>, > "Ranga Mc Donald" wrote: >> How to get the priority of the task 'printf()' in the vxworks? >> Anybody have any idea??? > >printf() is not a task. It is a function that can be called by tasks. >Consequently, it has no task priority. > >Regards, > Or, from another point of view, like most functions it executes at the priority of the task it's called from.... (However, being essentially an IO function, printf() might ultimately utilize resources, such as buffers, queues, semaphores, whose behavior can be affected by tasks of different priorities). Max. Cambury Technologies, Inc embedded wireless systems http://www.cambury.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxGetDec()/ppcDecTimer.c Date: Sat, 22 Jul 2000 18:55:02 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397A2646.900B9DB2@mindspring.com> References: <3966793F.12E19575@mindspring.com> <3969CB5E.454C1A82@mindspring.com> <396A78AC.B4852309@aps.anl.gov> <396AA5F3.11705809@mindspring.com> <396DD874.ECDE62D4@lmco.com> Mr. Lesch, and anyone else who can help,... HELP! (please) > When it is based on the system clock,as in this case, it does not have its own > interrupt handler routine. The connection routine for the timestamp tells Maybe you can explain an area that I have never understood... if you are willing. tickAnnounce(). This (I am told) is the proper way for the sysClk to announce a tick (rollover) to to the kernel. As I understand it, if the things are going normally, a tick happens every 60th (for most bsps) of a sec. What I don't get is the behavior of ticAnnounce. Under certain conditions, it does not really announce the tick to the kernel, but instead puts it ( the fact that a tick/roll has occured) on to one of the kernel's queue. So in a busy system the vxTic object (which ultimately gets incremented... reflects the number of 60th of a sec tics that have occurred) may not reflect the "real" time... as there may be one or more "a rollover hass occurred" msg on the queue. Back in vxw 5.11 we notice that our gethrtime() routine (which combined a read of a hardware timer with vxTic's value) was not functioning correctly... and traced it back to the fact that vxTick did not alway get updated immediatly. Solution was easy... add our own ourTick object that did always get updated (immediatly) when the sysClk ISR fired. What I would like to understand, is 1) WHY does vxworks do this "deferral", 2) under what conditions does it do this deferral, and -- this is what I am getting at -- 3) should I expect windView to report bogus time when ticAnnounce() has deferred the increment of vxTic? (on my bsp the timestamp driver uses the sysClk). Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Can't send correct data when using octet string with snmpv1. Date: Sat, 22 Jul 2000 23:04:17 GMT From: randy.jones@baesystems.com Organization: Deja.com - Before you buy. Message-ID: <8ld99b$s2m$1@nnrp1.deja.com> Agent host: PPC850 Agent OS: vxWorks 5.3 SNMP Version: SNMPv1 SNMP Mgr: PC running NT 4.0 When ever I try to do an snmpget to any variable of my MIB that is defined as an OCTET STRING, I get the wrong data. Even if the string length is just one character. If I perform an snmpget to a variable defined as a Display String, it works just fine. I am trying to access structures that contain binary data so I decided to use OCTET STRING. I am out of things to look for. The following is an example of my data structure: /* This is a structure we define to set and retrieve the prameters of the configData group */ typedef struct { char af[20]; char pf; char op; char iconStat; char pwrSrc; char pwrLvl; char audioVol; char ledBright; char lcdCon; char lcdBackLit; int totalUnitOnTm; } STRUCT_configData; Everything looks fine up to the getproc_got_string(). I cn't follow it any farther. The following is an examlple of my usage of Wind River's code. I think it is correct. static int configData_get_value(OIDC_T lastmatch, SNMP_PKT_T *pktp, VB_T *vbp, STRUCT_configData *data) /* !!! */ { extern STRUCT_configData configData; struct timespec elapsedTime; switch (vbp->vb_ml.ml_last_match) { case LEAF_adaptationFile: { memcpy ((char *) &(data->af), (char *) configData.af, sizeof(configData.af)); getproc_got_string(pktp, vbp, sizeof(data->af), data- >af, 0, VT_STRING); /* !!! */ break; } case LEAF_platform: { data->pf = (char) configData.pf; getproc_got_string(pktp, vbp, sizeof(configData.pf), data->pf, 0, VT_STRING); /* !!! */ break; } default: return GEN_ERR; } return NO_ERROR; } I could use some help. Thanks, Randy Jones Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: END Driver for 82559 Date: Sat, 22 Jul 2000 23:31:59 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net> Intel changed the PCI device ID half way through the 82559, maybe the version on their web site handles both? Paul In article , "Sheela Kashyap" wrote: >I have the used the 82557 End Driver as is for the 82559 ethernet >controller. I haven't had any problems. > >Sheela. > >"Laszlo Szebelledy" wrote in message >news:3IDd5.1780$U56.63446@newsread1.prod.itd.earthlink.net... >> Hi All, >> >> I'm looking into converting the 82557 End Driver to the 82559 Driver for >an >> IDT RC32364 target and am trying to get a handle on dynamic TBDs. Does >> anyone out there have experience with the 82559, is faimiliar with a >driver >> for this device, or has done such a coversion? >> >> > > --------------------------- Newsgroups: comp.os.vxworks Subject: 82559ER Driver Date: Sat, 22 Jul 2000 17:59:46 -0700 From: "Sheela Kashyap" Organization: Posted via Supernews, http://www.supernews.com Message-ID: Hi, We have an 82559ER ethernet controller on our board. As far as I know, the only *major* differences, from the software perspective between that and 82559 are "Wake on LAN" (excluded in 82559ER) and the device id (0x1209 for 82559ER). If that is the case, is it possible for one to use the fei82557End driver from WRS for the 82559ER, including just the changes for the device id? Does anybody have any idea of caveats etc.? Appreciate any help. Thanks in advance, Sheela. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 24 04:03:07 2000 From: Vxworks Exploder Date: Mon Jul 24 04:03:09 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 24 04:03:03 PDT 2000 Subject: assembly code in c function Subject: Re: who owns this semaphore? Subject: Re: Multicast IP to MAC Xlation ? Subject: Re: 750ppc cpu instruction times? help Subject: Re: assembly code in c function Subject: Re: who owns this semaphore? Subject: Problems with 82557 END Driver Subject: Re: who owns this semaphore? Subject: vxWorks & Green Hills C++ Subject: Re: HOT - TELECOM - VxWorks S/W Developers Needed - Ottawa Subject: 2700bsp - z8536 as sysClk? Subject: VxWorks contract opportunity Subject: Soft PLC and VxWorks Subject: Question on VxWorks socket and non-blocking I/O Subject: Re: MVME2431 ethernet problem Subject: VxMP configuration Subject: Re: Problems with 82557 END Driver Subject: IBM PowerPC 405GP and Java support ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: assembly code in c function Date: Sun, 23 Jul 2000 13:23:12 +0200 From: Sebastien Fauris Organization: Club-Internet (France) Message-ID: <397AD5A0.46D5EEBD@club-internet.fr> Hi vxWorkers, how can I write assembly code within a function in a C source file ? Actually my function looks like: UINT32 pciInDword (UINT32 addr) { /* read a dword from PCI space swapping the bytes */ __asm__("lwbrx 3,0,3"); /* sync I/O operation */ __asm__("eieio"); __asm__("sync"); /* return to caller */ __asm__("bclr 20,0"); } This seems to be okay, but I get warning messages: pciHwAccess.c:1543: warning: unused parameter `addr' pciHwAccess.c:1554: warning: control reaches end of non-void function Any advice is welcome. Sebastien. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Sun, 23 Jul 2000 15:43:42 GMT From: "Dan Perlberger" Organization: Road Runner - EC Message-ID: References: <8lahk4$5f4$1@nnrp2.deja.com> You could go into Windview and do a search on that Sem ID. You should be able to see which task has taken the Semaphore. Dan wrote in message news:8lahk4$5f4$1@nnrp2.deja.com... > I am trying to find the tasks that is holding > this semaphore. Is there a way to find out? > > -> semShow(0x3ef4ca0,1) > > Semaphore Id : 0x3ef4ca0 > Semaphore Type : BINARY > Task Queuing : PRIORITY > Pended Tasks : 3 > State : EMPTY > > NAME TID PRI TIMEOUT > ---------- -------- --- ------- > t1 3fb59a0 100 0 > OOOPortFactory 3ee05e8 132 0 > OOOOutPort7 3e5d800 133 0 > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Multicast IP to MAC Xlation ? Date: Sun, 23 Jul 2000 16:04:51 GMT From: "Don Bowman" Organization: none Message-ID: References: <39738586.69D5A317@lanterncom.com> "Babru" wrote in message news:39738586.69D5A317@lanterncom.com... > I need to translate multicast IP address to a MAC group in order to > configure a layer 2 bridge. Appreciate help in finding how to translate. > Is there any library call to do it. Can etherAddrResolve doit ? > The translation is straightforward. 01 00 5e + last 23 bits of the IP. ie: 239.1.1.1 is 01 00 5e 01 01 01 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 750ppc cpu instruction times? help Date: Sun, 23 Jul 2000 17:33:46 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lfa9q$5ga$1@nnrp1.deja.com> References: <397A1373.AA5C3E66@mindspring.com> Hi Al, In article <397A1373.AA5C3E66@mindspring.com>, Al Johnston wrote: > In the good old days I could look each of these instructions up > in a table to get how long they take. Life is much more complicated > now. > > Is there an easy way of doing this? I tried just using an inst/sec > rate and got numbers that were inconsistent with the code. In short, no. The time each instruction takes depends on a whole set of things as well as the actual time it takes the core to execute it. Included in this list are cache hit vs miss and TLB hit vs miss which can make a significant difference to the execution times. Pipeline issues may also appear here since you have a branch instruction. It's even worse on the 603 based cores where the TLB miss handling is done in an exception handler :( Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: assembly code in c function Date: Sun, 23 Jul 2000 17:53:35 GMT From: "Paddy" Organization: Time Warner Cable of San Diego, CA Message-ID: References: <397AD5A0.46D5EEBD@club-internet.fr> Those warnings are very valid... You have passed addr variable to your function but are not using it and also your function is declared to return UINT32 but there is no return statement at function end. Patrick "Sebastien Fauris" wrote in message news:397AD5A0.46D5EEBD@club-internet.fr... > Hi vxWorkers, > > how can I write assembly code within a function > in a C source file ? > > Actually my function looks like: > > > UINT32 pciInDword (UINT32 addr) > { > /* read a dword from PCI space swapping the bytes */ > __asm__("lwbrx 3,0,3"); > > /* sync I/O operation */ > __asm__("eieio"); > __asm__("sync"); > > /* return to caller */ > __asm__("bclr 20,0"); > } > > This seems to be okay, but I get warning messages: > pciHwAccess.c:1543: warning: unused parameter `addr' > pciHwAccess.c:1554: warning: control reaches end of non-void function > > Any advice is welcome. > > Sebastien. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Sun, 23 Jul 2000 17:59:13 GMT From: "Paddy" Organization: Time Warner Cable of San Diego, CA Message-ID: References: <8lahk4$5f4$1@nnrp2.deja.com> For debug purposes you could put a wrapper around the semTake/Give function calls just for that semaphore and put in some logMsg with taskIdSelf() as one of the output values. This will allow you to see who is giving/taking the semaphore. Patrick wrote in message news:8lahk4$5f4$1@nnrp2.deja.com... > I am trying to find the tasks that is holding > this semaphore. Is there a way to find out? > > -> semShow(0x3ef4ca0,1) > > Semaphore Id : 0x3ef4ca0 > Semaphore Type : BINARY > Task Queuing : PRIORITY > Pended Tasks : 3 > State : EMPTY > > NAME TID PRI TIMEOUT > ---------- -------- --- ------- > t1 3fb59a0 100 0 > OOOPortFactory 3ee05e8 132 0 > OOOOutPort7 3e5d800 133 0 > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Problems with 82557 END Driver Date: Sun, 23 Jul 2000 11:47:51 -0700 From: "Asher Greenberg" Organization: SBC Internet Services Message-ID: I'm replacing an old style netif IF_FEI driver with an END driver for 82557. The target boots using the old IF_FEI driver in the boot code. Then, it downloads the runtime image from the server into RAM. The runtime image includes the END driver in it. When the image loads, I get a driver attach error and thus I cannot use the Ethernet port, only the serial. Does anybody have an idea what might be wrong? Thank you. Here's the output screen: - ------------------------------------------------------------ Attaching network interface fei0... L2CR=0 done. Attaching network interface lo0... done. Loading... 856164 Starting at 0x10000... muxDevLoad failed for device entry 0! Failed to attach to device feiNetwork interface fei unknown. Attaching interface lo0...done Loading symbol table from nike:/home/babru/target/vxWorks.sym ...Error opening n ike:/home/babru/target/vxWorks.sym: status = 0x41 0x7ffdf8 (tRootTask): Loading end... 0x7ffdf8 (tRootTask): Loading end... 0x7ffdf8 (tRootTask): fei82557EndLoad: unit=0 pMemBase=0x0 memSize=0x0 nCFDs=32 nRFDs=32 flags=6 0x7ffdf8 (tRootTask): Unloading end...0x7ffdf8 (tRootTask): fei82557Unload... Do ne 0x7ffdf8 (tRootTask): Can't attach unknown device fei (unit 0). --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Sun, 23 Jul 2000 20:30:09 GMT From: "Dan Perlberger" Organization: Road Runner - EC Message-ID: References: <8lahk4$5f4$1@nnrp2.deja.com> This is assuming the Semtake is within his application code. If it is within the VxWorks code or some other code which he does not have source for, then this will not work. Also this is assuming that he knows which semaphore it is. If his application is large, then there might be a ton of semaphore calls, and putting temporary debug statements around each call would be very tedious. I think my suggestion of using Windview is a better way of finding the Semaphore. In Windview you can perform a search on any object ID to find where it is being used. So it would be a snap to find the task taking the Semaphore, even if it is not one his applications's tasks. Dan "Paddy" wrote in message news:RnGe5.834$OG2.3733@typhoon.san.rr.com... > For debug purposes you could put a wrapper around the semTake/Give function > calls just for that semaphore and put in some logMsg with taskIdSelf() as > one of the output values. This will allow you to see who is giving/taking > the semaphore. > > Patrick > wrote in message > news:8lahk4$5f4$1@nnrp2.deja.com... > > I am trying to find the tasks that is holding > > this semaphore. Is there a way to find out? > > > > -> semShow(0x3ef4ca0,1) > > > > Semaphore Id : 0x3ef4ca0 > > Semaphore Type : BINARY > > Task Queuing : PRIORITY > > Pended Tasks : 3 > > State : EMPTY > > > > NAME TID PRI TIMEOUT > > ---------- -------- --- ------- > > t1 3fb59a0 100 0 > > OOOPortFactory 3ee05e8 132 0 > > OOOOutPort7 3e5d800 133 0 > > > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > > --------------------------- Newsgroups: comp.os.vxworks Subject: vxWorks & Green Hills C++ Date: Mon, 24 Jul 2000 02:24:12 GMT From: "G Kraemer" Organization: @Home Network Message-ID: Has anyone successfully added the Green Hills (Multi 1.8.9) Standard C++ (with Exceptions) objects to the Tornado II (MVME-2700) BSP? I've tried using the GHS information in the Developement Guide and when I update libPPC604gnuvx.a with the GHS objects, vxWorks will no longer compile. Additionally, adding 'MUNCH=cxxmunch' to my Makefile also kills the vxWorks make. GHS appears to be taking their jolly good time on resolving this one! Any help would be greatly appreciated. TIA Garry --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: HOT - TELECOM - VxWorks S/W Developers Needed - Ottawa Date: Mon, 24 Jul 2000 03:25:02 GMT From: "Dragan Ostojic" Organization: Sympatico Message-ID: References: <8l2krb$fqk$1@news.igs.net> <397578ac.0@seat.s3two.ie> I wouldn't be so annoyed with head hunters posting in this news group. So far the policy here was to ignore them or to contact them privately, depending on each individual. Regards, Dragan. "Ranga Mc Donald" wrote in message news:397578ac.0@seat.s3two.ie... > Don't post the job ads repeatedly. This is a news group dedicated to the > technical discussions and exchanging the views in related to. Pl. read FAQ > of the news group before posting Job ads on this news group. > > > > "PROCOM" wrote in message > news:8l2krb$fqk$1@news.igs.net... > > Title: C++ / OOD VxWorks Software Developers > > > > Description: Our client, one of the world's largest telecommunications > > companies, currently requires 10 Junior to Senior Software Specialists to > > design, develop, implement and support Wireless products and applications. > > > > Minimum Requirements: > > All candidates should possess a minimum of 1 year experience developing > > real-time C/C++, OOD S/W on a VxWorks platform. > > > > Desired skills would include any previous Wireless development experience. > > > > Location: Ottawa, Ontario, Canada > > > > Type: Full-Time Placement > > > > Compensation: Open > > > > PROCOM: Established in 1978, Professional Computer Consultants Group Ltd. > > (Procom) is a national leader in the provision of Computer personnel on a > > contract > > and full-time basis. Our clients are comprised of the largest national and > > international corporations that utilize technical resources extensively > > across a wide range of disciplines. In the Financial Post (March 1999) > > Procom was ranked as the 6th largest professional Services Company in > > Canada. In November of 1999, Procom was named a Regional finalist in > Canada' > > s 50 Best Managed Private companies. Our track record is proven with more > > than 180 consultants servicing Ottawa's high tech community and more than > > 1600 Procom consultants currently on assignment throughout. North America. > > For further information on this and other opportunities please visit our > web > > site at www.procom.ca. > > > > Interested candidates are invited to forward their resumes or questions in > > confidence to: > > > > Derek Weber > > Technical Recruiter > > 300 March Road > > Kanata, Ontario > > K2K 2E2 > > phone: (613) 270-9339 ext 231 > > e-mail: derekw@procom.ca > > > > > > > > > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: 2700bsp - z8536 as sysClk? Date: Mon, 24 Jul 2000 00:25:22 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397BC532.DAF0632F@mindspring.com> Because of accuracy issues, I would like to be able to use the cpu decrementer hardware (with preload of 0xffff,ffff) as my "TimeStamp driver" (for windView) and as the basis for a timekeeping requirement of my own code. This means I would need to find a new home for the sysClk. The z8536 (pccz8536Timer.c) has already been coded for this function. I was wondering if anyone had tried using the z8536's sysClk code as the sysClk for this hardware/bsp? Any problems? thanks --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks contract opportunity Date: Thu, 13 Jul 2000 17:29:02 GMT From: clummis@volt.com Organization: Deja.com - Before you buy. Message-ID: <8kku84$o09$1@nnrp1.deja.com> I am an in-house recruiter for Volt Services Group, one of the nation's largest computer and technical consulting firms. Our purpose is to search for individuals that match an active job opening at one of our client companies and put you (or someone you know who would be interested in this opportunity) in touch with the hiring contact at that company. I found your resume on the Internet and thought it looked like a strong fit for our current contract opening for a Sr. Software Engineer The following is a brief description of the job: You will be involved with software development designing, coding and testing software for a transceiver system. The development will specifically be focused on fault management, operations and maintenance for this transceiver system. Our client is a global leader in providing integrated communications solutions and embedded electronic solutions located in the Chicago area If you would be interested in hearing more about this opportunity please call me at 1-800-573-8658. If I am unavailable at the time of your call, please leave a message and I will return your call promptly. I look forward to hearing from you soon. Regards, Chad Lummis Technical Recruiter Volt Services Group clummis@volt.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Soft PLC and VxWorks Date: Mon, 24 Jul 2000 08:13:34 +0200 From: Uwe Foss Organization: Siemens AG Message-ID: <397BDE8E.521C5311@dit.siemens.de> Ist under VxWorks a Soft-PLC available ? - ---------------------------- SIEMENS AG Uwe Foss Mail: uwe.foss@dit.siemens.de - ----------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Question on VxWorks socket and non-blocking I/O Date: Thu, 13 Jul 2000 12:29:22 -0500 From: Kenneth Arakaki Organization: Nortel Networks Message-ID: <396DFC72.755BD604@nortelnetworks.com> Hello, I am debugging some socket code for a ZIATECH SBC. A socket is created: socket(AF_INET, SOCK_STREAM, 0) setsockopt(socketfd, IPPROTO_TCP, TCP_NODELAY, ...) connectWithTimeout(socketfd, .....) and ioctl(socketfd,FIONBIO, (int)&on) this should configure a socket: non-blocking,immediate,TCP the issue is that the socket is polled: ioctl(socketfd, FIONREAD, (int)&bytesUnread) if bytesUnread != 0 while (bytesUnread != 0) read(socketfd, (char*)onebyte,1) ioctl(socketfd, FIONREAD, (int)&bytesUnread) my contention is that this will grab anything in the port, and because it is non-blocking it might be mid message stream Shouldn't select() (software interrupt) be used to monitor the port? Thank you for any help. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MVME2431 ethernet problem Date: Mon, 24 Jul 2000 09:22:41 +0200 From: "Michael Lawnick" Organization: Deja.com - Before you buy Message-ID: <200007240721.JAA29487@net.softec.de> References: Hi Ken, 1. Have you checked your RTC is running ? Set it with time (try ). It is needed for network and mostly switched off by manufactorer to save battery. 2. Try nping () to ping from target to host. 3. Do you use TFTP (not FTP) ? PPCBug only supports the trivial file transfer protocoll. HTH > I am trying to establish a Ethernet connection between a host (Windows NT) > and the target (MVME2431). I can communicate through the RS232 DEBUG port > using PPC4_BUG. I have used the PPC4-BUG command NIOT to set the IP > addresses for the client, host, subnet and gateway on the target. When I > use the NIOP command to get the boot image from the host, PPC4-BUG locks up, > and the only indication that anything is working is the Ethernet led lights > flash about 3 times and then nothing. When I ping the address from the > host, it times out. I have removed the network from the equation and used a > single crossover cable between the host and target, and still nothing. Does > anyone have any suggestions? 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 ============================================== Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxMP configuration Date: Mon, 24 Jul 2000 09:59:39 +0100 From: Organization: Keyspan Technology Message-ID: <397c04bc@runswick.octacon.co.uk> Reply-To: "Jason Edwards" Hi We've got an application which runs on multiple CPUs and uses VxMP. I've been looking in the VxWorks Programmers Guide ( 5.4 ) for information on configuring VxMP. Its a bit vague as to exactly what the difference is between SM_OBJ_MEM_SIZE & SM_MEM_SIZE and how they are related. In section 6.4.4, Configuration Example, page 284 they supply a sample configuration for a 3 CPU system. This example specifies SM_MEM_SIZE with different values on the master & slave CPU's, but does not clarify why - Can anyone explain ? I know that SM_OBJ_MEM_SIZE only works on the master - Does the same apply to SM_MEM_SIZE ? Exactly where does VxMP store semaphore information and the system partition ? The manual is once again a little bit vague in this respect. Thanks - -- Jason Edwards Technical Consultant Keyspan Technology Ltd Web : www.keyspan.co.uk E-Mail : jason@keyspan-nospam.co.uk Remove -nospam from my domain to reply to me personally --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with 82557 END Driver Date: 24 Jul 2000 09:50:42 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lh3k6$3hs$1@news1.xs4all.nl> References: Asher Greenberg wrote: : I'm replacing an old style netif IF_FEI driver with an END driver for 82557. : The target boots using the old IF_FEI driver in the boot code. Then, it : downloads the runtime image from the server into RAM. The runtime image : includes the END driver in it. When the image loads, I get a driver attach : error and thus I cannot use the Ethernet port, only the serial. Does anybody : have an idea what might be wrong? You are trying to mix two interfaces. I would suggest building a bootrom also including the END driver. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: IBM PowerPC 405GP and Java support Date: Mon, 24 Jul 2000 09:56:26 GMT From: Dave.Hartley@epid.eurotherm.co.uk Organization: Deja.com - Before you buy. Message-ID: <8lh3s9$bpm$1@nnrp1.deja.com> Has anyone had any discussion with Wind River regarding the support for the above processor ? I know that the basic architecture is supported, and that IBM have done their own BSP, But i am keen to know if anyone has asked the question regarding support for JWorks, TrueFFS, or DOSFS Thanks Dave Hartley Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 25 04:03:08 2000 From: Vxworks Exploder Date: Tue Jul 25 04:03:11 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 25 04:03:04 PDT 2000 Subject: Re: use dos filesystem on ram disk Subject: Re: Socket closing on VXWORKS. Subject: Re: ethernet address Subject: Re: ethernet address Subject: Re: Socket closing on VXWORKS. Subject: failure loading stripped elf Subject: Bothered by DHCP Subject: MCP750 PRST# Subject: VxDCOM Feedback? Subject: Re: failure loading stripped elf Subject: Event Processing Subject: JWorks and JNI Subject: Re: a cache problem of PPC8260 Subject: Re: a cache problem of PPC8260 Subject: Any untar+gzip available for vxworks? Subject: Re: Fail to boot with multiple 82559 cards Subject: anyone using smb, samba? Subject: Re: 82559ER Driver Subject: Re: Event Processing Subject: POSIX Wrapper around DOSFS2 Filesystem Subject: Re: Web Server under VxWorks Subject: How to get MAC Addr. Subject: MPC750 and MPC8260 Subject: Web Server under VxWorks Subject: Re: Socket closing on VXWORKS. Subject: Re: taskDelete in VxWorks Subject: Re: How to get MAC Addr. Subject: Re: How to get MAC Addr. Subject: Re: Soft PLC and VxWorks Subject: Re: VxDCOM Feedback? Subject: Re: VxDCOM Feedback? Subject: Re: Socket closing on VXWORKS. Subject: Re: Fail to boot with multiple 82559 cards Subject: Re: 82559ER Driver Subject: Sending to POSIX queues from ISRs Subject: Re: taskDelete in VxWorks Subject: Re: who owns this semaphore? Subject: Re: Sending to POSIX queues from ISRs Subject: problem using FEC on 860T Subject: Intel 82577 ethernet device driver required for PPC Subject: PCI64 PMC throughput figures ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: use dos filesystem on ram disk Date: 24 Jul 2000 08:25:02 -0500 From: vanbaren@falcon.si.com (vanbaren_gerald) Organization: Smiths Industries Message-ID: <397c359e@news.si.com> References: <8l9ust$js7$1@nnrp1.deja.com> lx1975@my-deja.com writes: >I want to build dos filesystem on ram disk under tornado 2.0 env, >and the following is my codes for building dos filesystem: >BLK_DEV *pBlkDev; >DOS_VOL_DESC *pVolDesc; >dosFsInit(20); >pBlkDev = ramDevCreate (0, 512, 400, 400, 0); >pVolDesc = dosFsMkfs ("DEV1:", pBlkDev); >And then I find I can create files on the filesystem, but can't read or >write data to a file, why? >int i; >int fd = open("DEV1:km", O_CREAT, 0); >if(fd>0) { > char buf[100]; > i = write(fd, "Hello", 5); > close(fd); >} Because you are opening the file in read mode, not write mode. The third parameter of the open call is the mode, you want one of the two following modes: O_WRONLY Open for writing only. O_RDWR Open for reading and writing. See the help page on open. gvb - -- +---------------------------------------------------------------------------+ | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | | My employer is a company. Companies are artifacts of a legal system. | |________________Artifacts are incapable of having opinions.________________| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Socket closing on VXWORKS. Date: 24 Jul 2000 08:33:09 -0500 From: vanbaren@falcon.si.com (vanbaren_gerald) Organization: Smiths Industries Message-ID: <397c3785$1@news.si.com> References: <3978BE07.EB5FBCB4@mitel.com> <8laqm9$9ap$1@nnrp1.deja.com> Also... you may not be closing the socket properly on the _other_ side of the connection. The close operation is a three way handshake just like the open operation. side side one two ------> <------ <------ (typically piggybacks with ) ------> If side one closes a socket and doesn't wait for the side two to close it gracefully, the side two will send its packet but side one won't be listening anymore and thus side two will time out waiting for its packet to be acknowledged. When it times out, it will close the socket, but that will take a while. gvb birajdar2250@my-deja.com writes: >Try this: >Set the SO_LINGER option of the socket with small linger time. This will >close the connection immediately. Not sure if it will free the buffers >as quick as you want. Let me know if it does. >In article <3978BE07.EB5FBCB4@mitel.com>, > Mangesh Joshi wrote: >> Hello All, >> I am working on a application which need to open and close sockets a >> finite number of times per min. My problem is that I am not able to >> close sockets fast enough. >> The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and >> stays there for at least a min before it actually closes. This pushes >me >> to have more socket/buffers available to keep up the throughput. >> Is there any way to reduce the closing time to a few seconds. >> Thanks, >> >> Mangesh Joshi. >> >> >Sent via Deja.com http://www.deja.com/ >Before you buy. - -- +---------------------------------------------------------------------------+ | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | | My employer is a company. Companies are artifacts of a legal system. | |________________Artifacts are incapable of having opinions.________________| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ethernet address Date: Thu, 13 Jul 2000 18:28:06 GMT From: jstack1@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8kl1nn$kqg$1@nnrp2.deja.com> References: <396DFA33.F76E0E21@club-internet.fr> Use ifShow() to find yourethernet address. In article <396DFA33.F76E0E21@club-internet.fr>, Sebastien Fauris wrote: > Hi vxWorkers, > > does someone know if there is a way in vxWorks to get the > ethernet address of a network interface present on my board ? > I couldn't find it (except using ARP services, thing that I don't want > to). > > Thanks ! > > Sebastien. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ethernet address Date: Thu, 13 Jul 2000 18:28:07 GMT From: jstack1@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8kl1no$kqh$1@nnrp2.deja.com> References: <396DFA33.F76E0E21@club-internet.fr> Use ifShow() to find your ethernet address. In article <396DFA33.F76E0E21@club-internet.fr>, Sebastien Fauris wrote: > Hi vxWorkers, > > does someone know if there is a way in vxWorks to get the > ethernet address of a network interface present on my board ? > I couldn't find it (except using ARP services, thing that I don't want > to). > > Thanks ! > > Sebastien. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Socket closing on VXWORKS. Date: Mon, 24 Jul 2000 13:11:42 GMT From: Mangesh Joshi Message-ID: <397C408F.3DC0028A@mitel.com> References: <3978BE07.EB5FBCB4@mitel.com> <8laqm9$9ap$1@nnrp1.deja.com> <397c3785$1@news.si.com> Thanks for your reply. I am using the SO_LINGER option but it is not working because we are closing the link too fast and the sockets are not able to communicate the closing sequence. Is there any call/api in VXWORKS which I can use to poll the socket for its state while closing the link ? vanbaren_gerald wrote: vanbaren_gerald wrote: > Also... you may not be closing the socket properly on the _other_ side > of the connection. The close operation is a three way handshake just > like the open operation. > > side side > one two > ------> > <------ > <------ (typically piggybacks with ) > ------> > > If side one closes a socket and doesn't wait for the side two to close it > gracefully, the side two will send its packet but side one won't be > listening anymore and thus side two will time out waiting for its > packet to be acknowledged. When it times out, it will close the socket, > but that will take a while. > > gvb > > birajdar2250@my-deja.com writes: > > >Try this: > >Set the SO_LINGER option of the socket with small linger time. This will > >close the connection immediately. Not sure if it will free the buffers > >as quick as you want. Let me know if it does. > > >In article <3978BE07.EB5FBCB4@mitel.com>, > > Mangesh Joshi wrote: > >> Hello All, > >> I am working on a application which need to open and close sockets a > >> finite number of times per min. My problem is that I am not able to > >> close sockets fast enough. > >> The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and > >> stays there for at least a min before it actually closes. This pushes > >me > >> to have more socket/buffers available to keep up the throughput. > >> Is there any way to reduce the closing time to a few seconds. > >> Thanks, > >> > >> Mangesh Joshi. > >> > >> > > >Sent via Deja.com http://www.deja.com/ > >Before you buy. > > -- > +---------------------------------------------------------------------------+ > | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | > | My employer is a company. Companies are artifacts of a legal system. | > |________________Artifacts are incapable of having opinions.________________| --------------------------- Newsgroups: comp.os.vxworks Subject: failure loading stripped elf Date: Mon, 24 Jul 2000 13:25:22 GMT From: Michael Fainstein Organization: Deja.com - Before you buy. Message-ID: <8lhg3t$k7n$1@nnrp1.deja.com> I have a large vxWorks bootable program, about 15 MB. My target is MPC860. I download it using FTP, and it takes a very long time. I tried to strip all debug info from vxWorks file using 'objcopy -g vxWorks vxworks.ndb' and download vxWorks.ndb to my target. The problem: bootLoadModule returns with error. How do I strip debug info in such way that download continues to work? Michael Fainstein Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Bothered by DHCP Date: Mon, 24 Jul 2000 13:49:11 GMT From: ellin_lin@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lhhgl$l6o$1@nnrp1.deja.com> I'm a tenderfoot of vxWorks and Tornado. And I met some problems when I tried to configured a DHCP server. I modified usrnetwork.c setting the IP range as written in user manual, and test this DHCP server with a DHCP client of Win98, it did nothing. Can someone give me some guidance? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: MCP750 PRST# Date: Mon, 24 Jul 2000 14:25:52 GMT From: bob_k_2@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lhjla$mub$1@nnrp1.deja.com> Hi, I'm running Tornado 2 vxWorks 5.4 on a Motorola MCP750 board. I am having trouble getting the Compact PCI pushbutton reset (PRST#) to work. As originally set up, the reset did not work at all. The MCP750 implied that it should work. After checking through the vxWorks config files I figured out which interrupt the PRST# connects to and wrote an ISR to service the interrupt and reboot the system mcp750ResetIsr(int unused) { reboot(BOOT_NORMAL); } This resulted in a partially working system. Parts of the MCP750 rebooted (the lights flashed and console io stopped) but the system never restarted. Since reboot() is not in the list of ISR supported functions I also tried using a semaphore and a task that pended on the semaphore then ran reboot. Any ideas what I am missing? Thanks, Bob Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxDCOM Feedback? Date: Mon, 24 Jul 2000 15:12:03 GMT From: wmiller_2002@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lhmbu$p7s$1@nnrp1.deja.com> I am looking at using VxDCOM for an application(PowerPC). Does anyone have any recent experience with it? Any feedback (good or bad) would be appreciated. Thanks in advance Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: failure loading stripped elf Date: Mon, 24 Jul 2000 17:46:38 +0200 From: Sebastien Mallet Organization: Nortel Networks Message-ID: <397C64DE.C475CFD6@nortelnetworks.com> References: <8lhg3t$k7n$1@nnrp1.deja.com> Hi, You can use the Solaris trip command if your host is Solaris (or UNIX in general). It works fine for files as large as 200Mbytes. I don't know about NT. There's another binary utility which comes with Tornado "stripppc" but this one doesn't seem to work correctly in Tornado 1. Hope that helps Sebastien Mallet --------------------------- Newsgroups: comp.os.vxworks Subject: Event Processing Date: Mon, 24 Jul 2000 17:13:52 -0500 From: "Ranga Mc Donald" Organization: Esat Net Customer Message-ID: <397c69ae.0@seat.s3two.ie> Reply-To: "Ranga Mc Donald" Are there any specific functions in VxWorks to use the events as a part of intertask communications or communication between a task and a ISR? I've seen some specific functions in the AMX RTOS. Are there any such kind of system defined functions like: ajevcre()......creates an event contains a bunch of events....like taskSpawn ajevwat()......wait for an occurrence of an event or events ajevsig()........change of the existing status on the events. I'm looking for the similiar kind of functions in vxworks. Any help is greatly appreciated. --------------------------- Newsgroups: comp.os.vxworks Subject: JWorks and JNI Date: Mon, 24 Jul 2000 11:16:28 -0500 From: Gretchen Patti Organization: ASIC Designs, Inc. Message-ID: <397C6BDC.9F552364@adichi.com> Greetings, vxWorks gurus! Is anybody using JNI under Personal JWorks?? I'm creating a Java application for an embedded system under vxWorks. We're using Personal JWorks 3.0. I'm trying to use JNI to call a native C routine. The System.loadLibrary call works OK, but when I call the native method I get an UnsatisfiedLinkError. Help!! ======================================================================== Sent or forwarded by Gretchen Patti gretchen@adichi.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: a cache problem of PPC8260 Date: Mon, 24 Jul 2000 16:48:16 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8lhs0e$tm7$1@nnrp1.deja.com> References: <8lc7pl$69g$1@nnrp1.deja.com> In article <8lc7pl$69g$1@nnrp1.deja.com>, zhouxiaochen@my-deja.com wrote: > Hi,vxworkers, > > When I add USER_I_CACHE_ENABLE,USER_D_CACHE_ENABLE, > INCLUDE_CACHE_SUPPORT, and in sysPhysMemDesc[], I change RAM on 60X bus > to VM_STATE_CACHEABLE, but the speed is same! It seems cache does no > work! > > Does anybody can explain it? Thanks! > > (Tornado2.0 VxWorks5.4 PPC8260) Did you also set USER_[ID]_CACHE_MODE ? You don't just define them, that's where you set the mode to disabled, writethrough, or copyback. When your board is running and you do a 'vxHid0Get', does it show caches enabled ? If you do a 'vxMsrGet()' does it show the MMUs enabled ? You're not executing out of flash by any chance are you... :-) hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: a cache problem of PPC8260 Date: Mon, 24 Jul 2000 16:55:31 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8lhsdv$u4p$1@nnrp1.deja.com> References: <8lc7pl$69g$1@nnrp1.deja.com> <8lhs0e$tm7$1@nnrp1.deja.com> In article <8lhs0e$tm7$1@nnrp1.deja.com>, gold@sdsu.edu wrote: > In article <8lc7pl$69g$1@nnrp1.deja.com>, > zhouxiaochen@my-deja.com wrote: > > Hi,vxworkers, > > > > When I add USER_I_CACHE_ENABLE,USER_D_CACHE_ENABLE, > > INCLUDE_CACHE_SUPPORT, and in sysPhysMemDesc[], I change RAM on 60X > bus > > to VM_STATE_CACHEABLE, but the speed is same! It seems cache does no > > work! > > > > Does anybody can explain it? Thanks! > > > > (Tornado2.0 VxWorks5.4 PPC8260) > > Did you also set USER_[ID]_CACHE_MODE ? You don't just define them, > that's where you set the mode to disabled, writethrough, or copyback. > When your board is running and you do a 'vxHid0Get', does it show caches > enabled ? If you do a 'vxMsrGet()' does it show the MMUs enabled ? Oops, I also forgot to ask if you set INCLUDE_MMU_BASIC ? If that's not set, your caches will never be enabled. hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Any untar+gzip available for vxworks? Date: Mon, 24 Jul 2000 13:50:26 -0400 From: "Yibing Fan" Organization: UUNET Canada News Reader Service Message-ID: Hi Everyone, We are looking for tar and gzip utility for vxworks. I searched vxworks archive did not found any. We are mostly interested in untar on target. We already got zlib for vxwork. Anyone know where to get one? We willing to pay of course. Thanks, Yibing Fan --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Fail to boot with multiple 82559 cards Date: Mon, 24 Jul 2000 19:04:58 GMT From: sbanavar@skystream.com Organization: Deja.com - Before you buy. Message-ID: <8li40n$46g$1@nnrp1.deja.com> References: <8la4df$okc$1@nnrp1.deja.com> <8lcqh4$ik2$1@nnrp1.deja.com> John, ultimately we'll be using 82559ER's on our box. Right now have test bed with 82559's. So we'll end up using 82559ER drivers from intel. Is there any support/driver from WRS for 82559ER ? Regards Srinath Banavar SkyStream Networks In article <8lcqh4$ik2$1@nnrp1.deja.com>, john_94501@my-deja.com wrote: > Hi, > > In article <8la4df$okc$1@nnrp1.deja.com>, > sbanavar@skystream.com wrote: > > Hello All, > > > > My setup is a PIII with pcPentiumpro BSP, vxWorks 5.3, Tornado 2.0 > > booting from the network. > > > > I have two 82559 cards on my system. When I use feiEnd driver, > > everything is just perfect with both the interfaces working. But the > > system does not boot when I change it to 82559 end driver (eeE) from > > Intel. More, once I build the BSP for one interface, the system boots > > just fine. > > If it works with feiEnd, but not with eeE then why not stick to feiEnd? > Is there some other reason you want to change from the WRS driver? > > Regards, > > John... > > Sent via Deja.com http://www.deja.com/ > Before you buy. > - -- Srinath Banavar SkyStream Networks. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: anyone using smb, samba? Date: Mon, 24 Jul 2000 19:24:33 GMT From: "thomas marsh" Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: hi... we are looking into the feasablility of using samba/smb with vxWorks... has anyone tried this? are there any commercial products available? is there somewhere I can look for sourcecode/info/etc... ? Does anyone know what kind of footprint this would give me? can anyone help? thanks in advance! - -thomoas marsh - ---------- Thomas Marsh thomas@vivityinc.com Software Engineer Vivity Inc Boulder, Co --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 82559ER Driver Date: Mon, 24 Jul 2000 21:13:24 GMT From: sbanavar@skystream.com Organization: Deja.com - Before you buy. Message-ID: <8libhg$a1a$1@nnrp1.deja.com> References: We are infact having the same situation. The modified fei82555End driver works fine with 82559ER cards. But couple of points to consider first. Now, if you have source for fei you can do the modifications. Second, how old is this fei source. There has been couple of fixes from WRS for fei, SP 26401 and SP 29158. You need to get the fei source with these fixes in it. And this (latest source) is what we dont have and the fei driver hangs in certain conditions. Let me know if you ahve any breakthrough. Regards Srinath In article , "Sheela Kashyap" wrote: > Hi, > > We have an 82559ER ethernet controller on our board. As far as I know, the > only *major* differences, from the software perspective between that and > 82559 are "Wake on LAN" (excluded in 82559ER) and the device id (0x1209 for > 82559ER). If that is the case, is it possible for one to use the fei82557End > driver from WRS for the 82559ER, including just the changes for the device > id? Does anybody have any idea of caveats etc.? > > Appreciate any help. > > Thanks in advance, > Sheela. > > - -- Srinath Banavar SkyStream Networks. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Event Processing Date: Mon, 24 Jul 2000 17:30:51 -0700 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <397CDFBB.E8286605@gv.net> References: <397c69ae.0@seat.s3two.ie> Reply-To: mkesti@gv.net Ranga Mc Donald wrote: >Are there any specific functions in VxWorks to use the events as a part of >intertask communications or communication between a task and a ISR? Intertask communications can be accomplished in vxWorks via shared memory, semaphores, message queues and pipes, sockets, and signals. Each of these are provided via libraries of user callable functions. - -- ======================================================================== 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: POSIX Wrapper around DOSFS2 Filesystem Date: Tue, 25 Jul 2000 03:26:25 GMT From: epooleNO@UCEnetis.com (RKT Technologies, Inc.) Organization: Software Consultants to Medical Device Manufacturers Message-ID: <397d0881.952242181@news.netis.com> We need to implement a POSIX wrapper around the VxWorks DOSFS2 filesystem, and implement file and directory permissions and all the rest. Does anyone know if anyone has done this already, and is there any source code available, either free or for a cost? Thanks... >-----------------------------------------------------------------< Eric Poole Specializing in Software Development for RKT Technologies, Inc. Medical Devices / Regulated Industries PO Box 1100 ISO 9001 / FDA Quality System Regulation Londonderry, NH 03053 Phone: 603-437-1811 Fax: 603-425-6475 >-----------------------------------------------------------------< E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to reply by e-mail; use epoole-at-netis-dot-com) >-----------------------------------------------------------------< Qualifications and Background: see http://www.netis.com/rkt >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Web Server under VxWorks Date: Tue, 25 Jul 2000 03:24:32 GMT From: epooleNO@UCEnetis.com (RKT Technologies, Inc.) Organization: Software Consultants to Medical Device Manufacturers Message-ID: <397d085a.952203437@news.netis.com> References: <397D1FD2.2D51DB48@yahoo.com> On Mon, 24 Jul 2000 22:04:19 -0700, Shelly HUANG wrote: >Is any web server running under VxWorks? Yes, look at http://www.goahead.com. >-----------------------------------------------------------------< Eric Poole Specializing in Software Development for RKT Technologies, Inc. Medical Devices / Regulated Industries PO Box 1100 ISO 9001 / FDA Quality System Regulation Londonderry, NH 03053 Phone: 603-437-1811 Fax: 603-425-6475 >-----------------------------------------------------------------< E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to reply by e-mail; use epoole-at-netis-dot-com) >-----------------------------------------------------------------< Qualifications and Background: see http://www.netis.com/rkt >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks Subject: How to get MAC Addr. Date: Tue, 25 Jul 2000 01:43:37 GMT From: ellin_lin@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lirc9$lle$1@nnrp1.deja.com> I'm developping software for MCP750 board with tornadoII, Can somebody tell me how to get the MAC Addr.? You're appreciated if giving some response. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: MPC750 and MPC8260 Date: Mon, 24 Jul 2000 18:53:12 -0700 From: Eric Monsler Organization: The ISP formerly known as Best Message-ID: <397CF308.4741D12C@radixtek.com> Hi all, We are designing a special purpose board, and are considering using an MPC8260 together with an MPC750. Since we need a PCI interface, we had planned to use the MPC107 PCI bridge chip, which is described as being able to support two processors on the 60x bus and to handle the arbitration required for the 8260 and 750 to share memory, etc. Unfortunately, that isn't true, apparently. There are 'issues' with contention that are not properly addressed by the MPC107. Some, but not all, of those are known fixable with an external EPLD or FPGA. We had planned to run VxWorks on both of those processors, with the 8260 using all of it's on-chip peripherals, and with a shared-memory interface (standard VxWorks) between the two processors. Since there are BSP's for some boards with 8260's, and since the peripherals are on-chip, the core of a BSP would seem to be available already. Now, we are considering simply using a much faster 750, hosting all of the code on that processor, and disabling the core of the 8260. The MPC107 PCI bridge would therefore see only one device accessing the bus. As I understand it, all of the 8260 peripherals, and the "CMP" (?) coprocessor, on-chip DMA, etc. are externally accessable when the core is disabled. My questions are: Are there any commercial boards that run a 750 and a core-disabled 8260? Do any of these also have a PCI bridge? Has anyone used VxWorks in a 750-8260 setup at all? Do they have a BSP for it? Has anyone accessed 8260 peripherals from a 750 running VxWorks? Has anyone done that, even without VxWorks? Thank you all for your time! Eric Monsler --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Web Server under VxWorks Date: Mon, 24 Jul 2000 22:04:19 -0700 From: Shelly HUANG Organization: Bell Northern Research Message-ID: <397D1FD2.2D51DB48@yahoo.com> Hi, Is any web server running under VxWorks? thank you for any information. best regards, Shelly --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Socket closing on VXWORKS. Date: Tue, 25 Jul 2000 02:14:46 GMT From: "Keith Galloway" Message-ID: References: <3978BE07.EB5FBCB4@mitel.com> <8laqm9$9ap$1@nnrp1.deja.com> <397c3785$1@news.si.com> <397C408F.3DC0028A@mitel.com> I went through a similar problem trying to FTP many small files. The FTP client would open and closes sockets for each file and eventually I would run out of network buffers. I was lucky enough to have plenty of memory to increase the network memory pool. For Tornado2 take a look in the VxWorks Network Programmers guide. I needed to increase the default numbers below in netBufLib.h #define NUM_SYS_64_MIN 20 /* 64 byte system clusters */ #define NUM_SYS_128_MIN 15 /* 128 byte system clusters */ #define NUM_SYS_256_MIN 15 /* 256 byte system clusters */ #define NUM_SYS_512_MIN 10 /* 512 byte system clusters */ Good Luck, Keith Mangesh Joshi wrote in message news:397C408F.3DC0028A@mitel.com... > Thanks for your reply. > I am using the SO_LINGER option but it is not working because we are closing the > link too fast and the sockets are not able to communicate the closing sequence. > Is there any call/api in VXWORKS which I can use to poll the socket for its state > while closing the link ? > > vanbaren_gerald wrote: > > vanbaren_gerald wrote: > > > Also... you may not be closing the socket properly on the _other_ side > > of the connection. The close operation is a three way handshake just > > like the open operation. > > > > side side > > one two > > ------> > > <------ > > <------ (typically piggybacks with ) > > ------> > > > > If side one closes a socket and doesn't wait for the side two to close it > > gracefully, the side two will send its packet but side one won't be > > listening anymore and thus side two will time out waiting for its > > packet to be acknowledged. When it times out, it will close the socket, > > but that will take a while. > > > > gvb > > > > birajdar2250@my-deja.com writes: > > > > >Try this: > > >Set the SO_LINGER option of the socket with small linger time. This will > > >close the connection immediately. Not sure if it will free the buffers > > >as quick as you want. Let me know if it does. > > > > >In article <3978BE07.EB5FBCB4@mitel.com>, > > > Mangesh Joshi wrote: > > >> Hello All, > > >> I am working on a application which need to open and close sockets a > > >> finite number of times per min. My problem is that I am not able to > > >> close sockets fast enough. > > >> The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and > > >> stays there for at least a min before it actually closes. This pushes > > >me > > >> to have more socket/buffers available to keep up the throughput. > > >> Is there any way to reduce the closing time to a few seconds. > > >> Thanks, > > >> > > >> Mangesh Joshi. > > >> > > >> > > > > >Sent via Deja.com http://www.deja.com/ > > >Before you buy. > > > > -- > > +--------------------------------------------------------------------------- + > > | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | > > | My employer is a company. Companies are artifacts of a legal system. | > > |________________Artifacts are incapable of having opinions.________________| > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Tue, 25 Jul 2000 10:24:45 +0530 From: Amarnath Shanbhag Organization: Lucent Technologies Message-ID: <397D1D95.A9899E53@lucent.com> References: <39780B42.8E1977C@lucent.com> <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com> <8laq0l$8nt$1@nnrp1.deja.com> I think taskDeleteHookAdd() is the best option. Thanks for your inputs. Just one clarification - Does the task which executes a taskDelete(taskId) block till the clean up function completes its cleanup ? birajdar2250@my-deja.com wrote: > > Bruce -> clean up can be done by using taskDeleteHookAdd(). > > Also you can use taskSafe() to prevent the task from being deleted > during critcal section. So if you are in a tight memory situation this > will help. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get MAC Addr. Date: Tue, 25 Jul 2000 03:35:49 GMT From: "Paddy" Organization: Time Warner Cable of San Diego, CA Message-ID: References: <8lirc9$lle$1@nnrp1.deja.com> Use sysXxxEthernetAddrGet where Xxx is the type of your enet driver..for example sysCpmEnetAddrGet(); Look up your bsp for this support You may also be able to get it from an ifnet structb initialized by ifunit() command. I'm not positive about this but I seem to remember it is possible. Look in ifLib.h for a definition of the ifnet structure. Patrick wrote in message news:8lirc9$lle$1@nnrp1.deja.com... > I'm developping software for MCP750 board with tornadoII, Can somebody > tell me how to get the MAC Addr.? > You're appreciated if giving some response. > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get MAC Addr. Date: Tue, 25 Jul 2000 07:23:06 +0200 From: Sebastien Fauris Organization: Club-Internet (France) Message-ID: <397D243A.FDA5D8C6@club-internet.fr> References: <8lirc9$lle$1@nnrp1.deja.com> ifShow will display the ethernet address. ellin_lin@my-deja.com wrote: > I'm developping software for MCP750 board with tornadoII, Can somebody > tell me how to get the MAC Addr.? > You're appreciated if giving some response. > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Soft PLC and VxWorks Date: Tue, 25 Jul 2000 03:41:20 GMT From: "DanO" Organization: RoadRunner - Carolina Message-ID: References: <397BDE8E.521C5311@dit.siemens.de> You could try IsaGraf.... "Uwe Foss" wrote in message news:397BDE8E.521C5311@dit.siemens.de... > Ist under VxWorks a Soft-PLC available ? > > ---------------------------- > SIEMENS AG > Uwe Foss > Mail: uwe.foss@dit.siemens.de > ----------------------------- > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxDCOM Feedback? Date: Tue, 25 Jul 2000 03:41:20 GMT From: "DanO" Organization: RoadRunner - Carolina Message-ID: References: <8lhmbu$p7s$1@nnrp1.deja.com> VxDCOM.... We are attempting to use it to write an OPC Server under VxWorks, which is mentioned in their marketing materials. the only problem is that after probably 5-6 months of trying we are still faced with VxDCOM specific issues. We are not alone as another company is trying to do this as well and have hit many of the same problems as we have. We intend to use it internally in our business. The other company intends to sell it as an OPC Server Toolkit, I believe. I can't comment on our issues as I'm not the one developing the OPC Server. Wind River is being somewhat responsive in trying to get the problems corrected, however it's not quite fixed as of my last information. This has been the case on 68K and x86, to the best of my knowledge. I doubt the PPC is any better off, but who knows... Good Luck! wrote in message news:8lhmbu$p7s$1@nnrp1.deja.com... > I am looking at using VxDCOM for an application(PowerPC). Does anyone > have any recent experience with it? Any feedback (good or bad) would be > appreciated. > > Thanks in advance > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxDCOM Feedback? Date: Tue, 25 Jul 2000 03:25:01 GMT From: "DanO" Organization: RoadRunner - Carolina Message-ID: References: <8lhmbu$p7s$1@nnrp1.deja.com> VxDCOM.... We are attempting to use it to write an OPC Server under VxWorks, which is mentioned in their marketing materials. the only problem is that after probably 5-6 months of trying we are still faced with VxDCOM specific issues. We are not alone as another company is trying to do this as well and have hit many of the same problems as we have. We intend to use it internally in our business. The other company intends to sell it as an OPC Server Toolkit, I believe. I can't comment on our issues as I'm not the one developing the OPC Server. Wind River is being somewhat responsive in trying to get the problems corrected, however it's not quite fixed as of my last information. This has been the case on 68K and x86, to the best of my knowledge. I doubt the PPC is any better off, but who knows... Good Luck! wrote in message news:8lhmbu$p7s$1@nnrp1.deja.com... > I am looking at using VxDCOM for an application(PowerPC). Does anyone > have any recent experience with it? Any feedback (good or bad) would be > appreciated. > > Thanks in advance > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Socket closing on VXWORKS. Date: Tue, 25 Jul 2000 07:15:16 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ljeq3$2g3$1@nnrp1.deja.com> References: <3978BE07.EB5FBCB4@mitel.com> <8laqm9$9ap$1@nnrp1.deja.com> <397c3785$1@news.si.com> <397C408F.3DC0028A@mitel.com> Hi, You might want to take a look at the shutdown() function... No promise that it will solve anything, but I seem to remember it was useful in the past when I was playing with sockets... Regards, John... In article <397C408F.3DC0028A@mitel.com>, Mangesh Joshi wrote: > Thanks for your reply. > I am using the SO_LINGER option but it is not working because we are closing the > link too fast and the sockets are not able to communicate the closing sequence. > Is there any call/api in VXWORKS which I can use to poll the socket for its state > while closing the link ? > > vanbaren_gerald wrote: > > vanbaren_gerald wrote: > > > Also... you may not be closing the socket properly on the _other_ side > > of the connection. The close operation is a three way handshake just > > like the open operation. > > > > side side > > one two > > ------> > > <------ > > <------ (typically piggybacks with ) > > ------> > > > > If side one closes a socket and doesn't wait for the side two to close it > > gracefully, the side two will send its packet but side one won't be > > listening anymore and thus side two will time out waiting for its > > packet to be acknowledged. When it times out, it will close the socket, > > but that will take a while. > > > > gvb > > > > birajdar2250@my-deja.com writes: > > > > >Try this: > > >Set the SO_LINGER option of the socket with small linger time. This will > > >close the connection immediately. Not sure if it will free the buffers > > >as quick as you want. Let me know if it does. > > > > >In article <3978BE07.EB5FBCB4@mitel.com>, > > > Mangesh Joshi wrote: > > >> Hello All, > > >> I am working on a application which need to open and close sockets a > > >> finite number of times per min. My problem is that I am not able to > > >> close sockets fast enough. > > >> The socket goes to state FIN_WAIT_1 or FIN_WAIT_2 or TIME_WAIT, and > > >> stays there for at least a min before it actually closes. This pushes > > >me > > >> to have more socket/buffers available to keep up the throughput. > > >> Is there any way to reduce the closing time to a few seconds. > > >> Thanks, > > >> > > >> Mangesh Joshi. > > >> > > >> > > > > >Sent via Deja.com http://www.deja.com/ > > >Before you buy. > > > > -- > > +----------------------------------------------------------------------- - ----+ > > | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | > > | My employer is a company. Companies are artifacts of a legal system. | > > |________________Artifacts are incapable of having opinions.________________| > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Fail to boot with multiple 82559 cards Date: Tue, 25 Jul 2000 07:26:12 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ljfei$3to$1@nnrp1.deja.com> References: <8la4df$okc$1@nnrp1.deja.com> <8lcqh4$ik2$1@nnrp1.deja.com> <8li40n$46g$1@nnrp1.deja.com> Hi, > ultimately we'll be using 82559ER's on our box. Right now have test bed > with 82559's. So we'll end up using 82559ER drivers from intel. > > Is there any support/driver from WRS for 82559ER ? From what I've heard it is a simple device ID update; if that is the case then all you should need to do is update the file in your BSP that finds the PCI devices... For the x86 BSPs this is sysNetif.c - look for the function sys557PciInit() and change the ID that is searched for using pciFindDevice() at the beginning of the for loop. If you need to support both types then just add some search logic around the existing for loop to search for different types of card. If this works I'd be interested... Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 82559ER Driver Date: Tue, 25 Jul 2000 07:29:41 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ljfl2$45d$1@nnrp1.deja.com> References: Hi, I believe that you should be able to just change the PCI ID in your BSP (where the card is detected and configured) and then the standard fei driver should be happy. Let me know if you do this successfully - I'd be interested to hear. If you need a pointer to where to change the ID, see my 2nd response to the thread about multiple 82559 cards, or send me an email. HTH, John... In article , "Sheela Kashyap" wrote: > Hi, > > We have an 82559ER ethernet controller on our board. As far as I know, the > only *major* differences, from the software perspective between that and > 82559 are "Wake on LAN" (excluded in 82559ER) and the device id (0x1209 for > 82559ER). If that is the case, is it possible for one to use the fei82557End > driver from WRS for the 82559ER, including just the changes for the device > id? Does anybody have any idea of caveats etc.? > > Appreciate any help. > > Thanks in advance, > Sheela. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Sending to POSIX queues from ISRs Date: Tue, 25 Jul 2000 07:52:38 GMT From: Y.M. Organization: Deja.com - Before you buy. Message-ID: <8ljh05$4vq$1@nnrp1.deja.com> "VxWorks Programmer's Guide" mentions that it is safe to send to a VxWorks message queue from inside an ISR. Is it the same with POSIX message queues? I've tried it and had no problems, but I am wondering if there are any caveats that I do not know about. For instance, I had assumed that the operation of opening an *already open* POSIX queue is non-blocking, but when I tried it from inside an ISR it resulted in an exception. Regards, Yevgeny Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Tue, 25 Jul 2000 13:39:08 +0530 From: Amarnath Shanbhag Organization: Lucent Technologies Message-ID: <397D4B24.F3D752BA@lucent.com> References: <39780B42.8E1977C@lucent.com> <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com> <8laq0l$8nt$1@nnrp1.deja.com> birajdar2250@my-deja.com wrote: > > Bruce -> clean up can be done by using taskDeleteHookAdd(). > If we do a taskDeleteHookAdd( (FUNCPTR) cleanupFunc ); will the cleanupFunc get called for all tasks in the system ? Can we restrict this cleanup function for our task only ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Tue, 25 Jul 2000 08:00:32 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ljhev$5bc$1@nnrp1.deja.com> References: <8lahk4$5f4$1@nnrp2.deja.com> Hi, If you are just after checking the state of a semaphore from the shell, and it is a binary one (or a mutex), then the following should work: - -> d 0x3ef4ca0+0x18, 1, 4 The number displayed should be the ID of the task holding the semaphore, or zero if it's empty. Of course, binary semaphores are often optimised for the architecture and those implementations may not track the owner the same way - try it and see... Oh, it uses the same SEM_ID as in your example; 0x18 is the offset into the semaphore data structure of the owner field which should be valid for binary and mutex semaphores (it is the count for counting semaphores). Remember though, this is using data that is not guaranteed from one version of VxWorks to the next. You can look at the structure in target/h/private/semLibP.h, but don't rely on it remaining the same in any code you might be planning to use on future versions of the OS! Regards, John... In article <8lahk4$5f4$1@nnrp2.deja.com>, tracyajones5487@my-deja.com wrote: > I am trying to find the tasks that is holding > this semaphore. Is there a way to find out? > > -> semShow(0x3ef4ca0,1) > > Semaphore Id : 0x3ef4ca0 > Semaphore Type : BINARY > Task Queuing : PRIORITY > Pended Tasks : 3 > State : EMPTY > > NAME TID PRI TIMEOUT > ---------- -------- --- ------- > t1 3fb59a0 100 0 > OOOPortFactory 3ee05e8 132 0 > OOOOutPort7 3e5d800 133 0 > > 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: Sending to POSIX queues from ISRs Date: Tue, 25 Jul 2000 09:25:57 GMT From: Urban Lindberg Organization: Deja.com - Before you buy. Message-ID: <8ljmf3$8jl$1@nnrp1.deja.com> References: <8ljh05$4vq$1@nnrp1.deja.com> Hi! The library reference says that you can call mq_send from an ISR. I think you should assume less and RTFM some more (it will save you a lot of time trying to find out what's wrong). Functions that isn't documented as callable from ISR's is most likely not! The reason why you can't call mq_open from an ISR is simple, it has to look up the name using hashLib functions (they are for some reason not part of the docs) and the hashLib functions will probably use some blocking calls (hmmmm, how will mq_send work from ISR's if file descriptor become local to tasks ????). HTH Urban In article <8ljh05$4vq$1@nnrp1.deja.com>, Y.M. wrote: > "VxWorks Programmer's Guide" mentions that it is safe to send to a > VxWorks message queue from inside an ISR. Is it the same with POSIX > message queues? I've tried it and had no problems, but I am wondering > if there are any caveats that I do not know about. For instance, I had > assumed that the operation of opening an *already open* POSIX queue is > non-blocking, but when I tried it from inside an ISR it resulted in an > exception. > > Regards, > > Yevgeny > > 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: problem using FEC on 860T Date: Tue, 25 Jul 2000 11:55:23 +0200 From: "Martin Walser" Organization: VBS - Vienna Backbone Service Message-ID: <964519072.363104@newsmaster-04.atnet.at> Hi, does anyone know how to initialise the MotFec for MPC860T using tornado2. I don't know which components are necessary. I tried to do this, but i received an exception and the system seems to crash. Thanks Martin --------------------------- Newsgroups: comp.os.vxworks Subject: Intel 82577 ethernet device driver required for PPC Date: Tue, 25 Jul 2000 11:01:41 +0100 From: Paul Brewer Message-ID: <397D6585.E579A02C@compd.com> Reply-To: pbrewer.removethis@compd.com Hi, I have a board that has a PowerPC 750 with a Motorola 106 PPC to PCI bridge. I have a Intel 82558 (seems to be more or less equivalent to a 82557) PCI ethernet device on the PCI bus. This device (being an intel device) is little endian. I have a vxWorks device driver which came from the vxWorks device driver kit called if_fei.c. This driver is for a 82557 but the documentation seems to show that its appropriate for a 82558. The device sets up an area of shared memory which is accessed by the device and the processor. I haven't got any documentation for the 82557 (didn't think I'd need it considering I was using an off the shelf driver) but it seems that the processor has to setup a ring of transmit and receive buffers which are then read and filled by the ethernet device. On my system the processor sets up the ring as big-endian and device reads and writes as little endian. Needless to say it doesn't work. Does anyone have a byte swapped device driver I can use? Failing that does anyone know which parts I need to byte swap in order to get my existing driver to work? Regards Paul Brewer --------------------------- Newsgroups: comp.os.vxworks Subject: PCI64 PMC throughput figures Date: Tue, 25 Jul 2000 11:03:56 +0100 From: "R Grommet" Organization: Defence Evaluation & Research Agency Message-ID: <8ljp1p$289$1@trog.dera.gov.uk> Hi all, I'm having an odd problem with a PCI64 PMC. It achieves 116MB/sec throughput when in PCI32 mode but only 30MB/sec in PCI64 mode. This is when its attached to a Motorola MVME2300 processor board using the Raven PCI Host Bridge. I suspect its a fault with the PMC, but has anyone done any benchmarks on other PCI64 PMCs, particularly on processor boards using the Raven chip? I got some odd results from someone using a MVME2400 (using the Hawk PCI Bridge) wherein the PMC achieved 125MB/sec in PCI32 but only 80MB/sec in PCI64. This has set me wondering if there is a problem with PCI64 on the Raven. TIA Ralph Grommet --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Jul 25 12:40:18 2000 From: "Chhabra, Vikram" Date: Tue Jul 25 12:40:20 PDT 2000 Subject: VxWorks Context Switch Time Does anyone know what the VxWorks average and Worst Case context switch times are on Motorola Processors. Or any place where I can get this information. Thanks Vikram Chhabra From vxwexplo-errs@csg.lbl.gov Wed Jul 26 04:03:13 2000 From: Vxworks Exploder Date: Wed Jul 26 04:03:15 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 26 04:03:09 PDT 2000 Subject: Re: taskDelete in VxWorks Subject: Socket read size (was Re: Socket closing on VXWORKS.) Subject: Re: Any untar+gzip available for vxworks? Subject: Re: How to get MAC Addr. Subject: tor2/windView newbie: "INTnn" on 2700(ppc 750)bsp Subject: Re: MPC750 and MPC8260 Subject: Re: using the serial line with vxWorks Subject: Re: Tornado 2 and Windows 2000 Subject: time() and localtime() functions in VxWorks... Subject: taskCreat() vs. taskInit() Subject: Re: time() and localtime() functions in VxWorks... Subject: Re: cross wind debugger Subject: Re: vxworks simple system calls Subject: Re: split as in unix. Subject: Re: Vx/Works PROXYARP on ethernet for PPP Subject: Re: time() and localtime() functions in VxWorks... Subject: Re: assembly code in c function Subject: VxWorks Context Switch Time Subject: Re: VxDCOM Feedback? Subject: Loading an END driver in VxWorks Subject: Re: Structures Subject: Flash EPROM memory Subject: ppc "machine language" reg use? Subject: Using VxWorks SCSI driver Subject: ElfHex problems and PPC Subject: Dialing procedure Subject: VxWorks ROM image and application. Subject: Re: Fail to boot with multiple 82559 cards Subject: Re: Dialing procedure Subject: Re: Other RTOS's? Subject: Re: ppc "machine language" reg use? Subject: multi-threaded corba server Subject: Re: Max Resources ? Subject: Re: verbosity macro Subject: Re: Intel 82577 ethernet device driver required for PPC Subject: Tornado 2 vs Tornado 1 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskDelete in VxWorks Date: Tue, 25 Jul 2000 14:34:16 +0200 From: Daniel Schnell Organization: Siemens AG Message-ID: <397D8948.1325679@icn.siemens.de> References: <39780B42.8E1977C@lucent.com> <0327ac8d.00cae0fe@usw-ex0102-084.remarq.com> <8laq0l$8nt$1@nnrp1.deja.com> <397D4B24.F3D752BA@lucent.com> Hello Amarnath, yes that's true: it will be called for EVERY taskDelete() - routine. So you should check for the appropriate task entry function by evaluating the paramater pTcb like below. void deleteHook (WIND_TCB *pTcb) { if (pTcb->entry == (FUNCPTR) taskEntry) { /* do something senseful */ } } We encountered some problems with destructors when using C++ and taskDelete when calling taskDelete in the context of the wdbTask (this will be normally done if you are calling such a function within the Windshell. So we added a little job to the exception task. This behaviour by the way does not show up if called in the target built in shell. void deleteHook (WIND_TCB *pTcb) { if (pTcb->entry == (FUNCPTR) taskEntry) { /* do something senseful */ if (taskIdCurrent->entry == (FUNCPTR) excTask) taskDeleteHookDelete ((FUNCPTR) deleteHook); else excJobAdd ((VOIDFUNCPTR) deleteHook, (int) pTcb, 0, 0, 0, 0, 0); } } Hope that helps. Daniel. Amarnath Shanbhag wrote: > > birajdar2250@my-deja.com wrote: > > > > Bruce -> clean up can be done by using taskDeleteHookAdd(). > > > > If we do a taskDeleteHookAdd( (FUNCPTR) cleanupFunc ); > will the cleanupFunc get called for all tasks in the system ? > Can we restrict this cleanup function for our task only ? --------------------------- Newsgroups: comp.os.vxworks Subject: Socket read size (was Re: Socket closing on VXWORKS.) Date: 25 Jul 2000 13:38:42 GMT From: jgay@kanon.lucent.com (John M. Gay) Organization: Lucent Technologies Message-ID: <8lk592$9ig@nntpb.cb.lucent.com> References: <3978BE07.EB5FBCB4@mitel.com> <397c3785$1@news.si.com> <397C408F.3DC0028A@mitel.com> Reply-To: DELETEjgay@lucent.com Since there are a few people here with vxWorks ftp socket experience I was wondering if anyone knew how to increase the read size from a socket. I am doing: while ((nBytes = read(dataSock, ftpBuffer, 1500)) > 0) {} and the number of bytes read is always 512. Is there some system thing that I have missed? Following one of the previous posts here I followed NUM_SYS_512_MIN to its use in bootConfig.c as CL_DESC sysClDescTbl [] = { /* clusterSize num memArea memSize ----------- ---- ------- ------- */ {64, NUM_SYS_64_MIN, NULL, 0}, {128, NUM_SYS_128_MIN, NULL, 0}, {256, NUM_SYS_256_MIN, NULL, 0}, {512, NUM_SYS_512_MIN, NULL, 0}, }; int sysClDescTblNumEnt = (NELEMENTS(sysClDescTbl)); should I add entries for 1024 and 2048? Any help is greatly appreciated. - -- john gay. DELETEjgay@lucent.com Life is a near-death experience. -- George Carlin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Any untar+gzip available for vxworks? Date: 25 Jul 2000 14:07:10 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lk6ue$cv7$1@news1.xs4all.nl> References: Yibing Fan wrote: : We are looking for tar and gzip utility for vxworks. TAR is part of DosFS2, but it does not perform compression. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get MAC Addr. Date: Tue, 25 Jul 2000 11:03:00 -0400 From: Sunil Joseph Organization: Lucent Technologies, Columbus, Ohio Message-ID: <397DAC24.6B163796@lucent.com> References: <8lirc9$lle$1@nnrp1.deja.com> hi, ifShow at the vxworks shell should give you the MAC address of all interfaces in your system, however if you want to use this info in a program use a ifunit and then extract. ac_enaddr from the structure arpcom, make sure you have 6 chars defined for storing the result. Good luck. Sunil Paddy wrote: > > Use sysXxxEthernetAddrGet where Xxx is the type of your enet driver..for > example sysCpmEnetAddrGet(); > > Look up your bsp for this support > > You may also be able to get it from an ifnet structb initialized by ifunit() > command. I'm not positive about this but I seem to remember it is possible. > Look in ifLib.h for a definition of the ifnet structure. > > Patrick > > wrote in message news:8lirc9$lle$1@nnrp1.deja.com... > > I'm developping software for MCP750 board with tornadoII, Can somebody > > tell me how to get the MAC Addr.? > > You're appreciated if giving some response. > > > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: tor2/windView newbie: "INTnn" on 2700(ppc 750)bsp Date: Tue, 25 Jul 2000 11:39:47 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397DB4C3.6B0E34D1@mindspring.com> I am running tor2 on a mvme2700 card (ppc 750). WindView displays things it calls INTnn (INT16, INT18)... Does anyone know what these map to? I know about interrupts... but the "interrupt space" on this card is a nightmare... Nothing I can see looks like these numbers. Hopefully someone can give me a list of the INTnn that windview can display and what they are associated with... please! thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC750 and MPC8260 Date: Tue, 25 Jul 2000 16:53:23 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8lkgm0$s67$1@nnrp1.deja.com> References: <397CF308.4741D12C@radixtek.com> In article <397CF308.4741D12C@radixtek.com>, Eric Monsler wrote: > Hi all, > > We are designing a special purpose board, and are considering using an > MPC8260 together with an MPC750. Since we need a PCI interface, we had > planned to use the MPC107 PCI bridge chip, which is described as being > able to support two processors on the 60x bus and to handle the > arbitration required for the 8260 and 750 to share memory, etc. > > Unfortunately, that isn't true, apparently. There are 'issues' with > contention that are not properly addressed by the MPC107. Some, but not > all, of those are known fixable with an external EPLD or FPGA. Right, not to mention 'features' (ok, anomalies) in 8260 bus behavior. Most of our boards are 750/8260 combination, with the 8260 core disabled. > We had planned to run VxWorks on both of those processors, with the 8260 > using all of it's on-chip peripherals, and with a shared-memory > interface (standard VxWorks) between the two processors. Since there > are BSP's for some boards with 8260's, and since the peripherals are > on-chip, the core of a BSP would seem to be available already. I think all the BSPs (all 2 of them, the EST and Mot VADS board, maybe there are others by now) run the 8260 as the main processor. You might need some tweaking as far as bus interface and mode, memory controller, etc... You'll need to consider separate Flash for booting, since they won't be executing the same code. Not to mention the 8260 fetches the hardware config word from flash (that doesn't follow 60x bus protocol, so it can't compete with the 750 for Flash at that time). > Now, we are considering simply using a much faster 750, hosting all of > the code on that processor, and disabling the core of the 8260. The > MPC107 PCI bridge would therefore see only one device accessing the > bus. As I understand it, all of the 8260 peripherals, and the "CMP" (?) > coprocessor, on-chip DMA, etc. are externally accessable when the core > is disabled. The 8260 does become a bus master. The 8260 core won't, but the CPM will. The CPM has 2 SDMA (real DMA) and 4 IDMA (virtual DMA) channels that will become bus masters when you do things like Tx/Rx ethernet frames, etc... Essentially the SCC, FCC, MCC, and ATM controllers use the SDMA and will become bus masters on whatever bus (60x or 8260-local) their data buffers and buffer descriptors are located. > My questions are: > > Are there any commercial boards that run a 750 and a core-disabled > 8260? Do any of these also have a PCI bridge? Commercial ? Not that I've seen, but Motorola has a 'reference design' for the Scout which is a 750/8260 with core disabled. I don't think anyone's ported vxWorks to it (and advertised it). > Has anyone used VxWorks in a 750-8260 setup at all? Do they have a BSP > for it? Has anyone accessed 8260 peripherals from a 750 running > VxWorks? Has anyone done that, even without VxWorks? Currently some of our boards are 750 at 300Mhz with an core disabled 8260. It works fine because we have an external PLD to account for various 8260 bus anomalies, and spent lots of time with a logic analyzer :-) I did the BSP originally for the 8260, then we disabled the core and I used a 604 BSP for the 750. It works, and yes you can access the 8260 peripherals from the 750. I do it under vxWorks, and also under standalone diagnostics. You *need* to access the 8260 even if the core is disabled because the memory controller is there and you need to set it up. If you're using a 107, I think you can disable the 8260 memory controller and use it instead. The dual-port RAM (where all the peripherals are) must be marked non-cacheable if you enable the data cache on the 750. If you're going to run an external 750, you will get a big performance boost by putting L2 cache on the backside of the 750. hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: using the serial line with vxWorks Date: Tue, 25 Jul 2000 08:32:00 -0500 From: Bo Yu Organization: Motorola CIG Message-ID: <397D96D0.D0753D24@email.mot.com> References: <39773CBC.B837981D@correu.udg.es> - --------------ADC1627BC7A5B2B902517C86 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Elvis wrote: > we have the following problem: > when booting vxWorks from serial line, we can download programs to the > target and then 'cut' the serial connection and use the serial line to > communicate with the programs on the target. It works fine, but when we > boot vxWorks from the ethernet, we can download programs to the target, > but the serial line is no longer working. Are we missing some > inizialization? where? can the serial line be software configured, or > it's only possible to do that from the include files? > > Thank you very much. > > Dani & Elvis > Students in Computer Science at University of Girona > Spain If you only have one serial line. You might define CONSOLE_TTY to NONE and WDB_COMM_TTY to 0. When you switch to ethernet downloading, make sure CONSOLE_TTY is 0. - -- Bo Yu Motorola, NSS - --------------ADC1627BC7A5B2B902517C86 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Elvis wrote:
we have the following problem:
when booting vxWorks from serial line, we can download programs to the
target and then 'cut' the serial connection and use the serial line to
communicate with the programs on the target. It works fine, but when we
boot vxWorks from the ethernet, we can download programs to the target,
but the serial line is no longer working. Are we missing some
inizialization? where? can the serial line be software configured, or
it's only possible to do that from the include files?

Thank you very much.

Dani & Elvis
Students in Computer Science at University of Girona
Spain

If you only have one serial line. You might define CONSOLE_TTY to NONE
and WDB_COMM_TTY to 0. When you switch to ethernet downloading,
make sure CONSOLE_TTY is 0. 
-- 
Bo Yu
Motorola, NSS
  - --------------ADC1627BC7A5B2B902517C86-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 2 and Windows 2000 Date: Tue, 25 Jul 2000 18:13:01 +0200 From: "Arik Halperin" Organization: NetVision Israel Message-ID: <8lke7s$n0m$1@news.netvision.net.il> References: You should check the "Make the abov check box settings permanent" , I think this should do the trick. - -Arik "Adam" wrote in message news:SMt95.32292$i5.360478@news1.frmt1.sfba.home.com... > Ok, I have Tornado 2 running under Windows 2000 now. > > Here's how I did it: > > I looked into Martin's suggestion of tricking Tornado into thinking it was > running on an NT box and found that Microsoft provides a nifty utility (The > Application Compatibility Tool) for doing just that. It's located in the > support directory of the Win2k CD. All you have to do is run apcompat.exe > and tell it to run Tornado as if it were on Windows NT. The only downside > I've seen so far is that you have to launch it this way every time. > > For more information on the Application Compatibility Tool go to Microsoft's > online knowledge base and read article Q251062. > > -Adam > > --------------------------- Newsgroups: comp.os.vxworks Subject: time() and localtime() functions in VxWorks... Date: Tue, 25 Jul 2000 11:47:17 -0700 From: "Chandana Sembakutti" Message-ID: <%glf5.1334$1P.1674@read1> Hi, I am trying to determine the current time using this function... /***************/ time_t nowTime; struct tm *pTime; nowTime = time( NULL ); pTime = localtime( &nowTime ); /**********************/ But when I print the broken down time from pTime, they start at 70/1/1. Does any body know what I am doing wrong and how to fix it. Thanks, Chandana --------------------------- Newsgroups: comp.os.vxworks Subject: taskCreat() vs. taskInit() Date: Tue, 25 Jul 2000 19:11:38 GMT From: Glen Lumsden Message-ID: <397DE681.3DB0AD2F@sympatico.ca> What is the difference between taskCreat() and taskInit()? Do the parameters for taskCreat() match those of taskSpawn()? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: time() and localtime() functions in VxWorks... Date: Tue, 25 Jul 2000 12:21:10 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <3185b300.94ea3f56@usw-ex0101-007.remarq.com> References: <%glf5.1334$1P.1674@read1> The VxWorks real-time clock does not start running until you acess it. It's initial time is 10/1/1. At bootup, you bsp should set it from some hardware real-time clock, if you have one onboard. Rob't Blumberg Embedded Software Developer Anything below is tacked on by Remarq! - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cross wind debugger Date: Tue, 25 Jul 2000 14:35:38 -0400 From: Krishna Samavedam Organization: Hughes Network Systems Message-ID: <397DDDF9.B517262A@hns.com> References: HI Couple of things to notice : Are U using XLINK properly ? U should not use any -O(optimisation flag) and compile with -g flag If U have thse things properly.. compile all over again. It some times helps. HTH krishna Abhinav Prasad wrote: > Hi > Iam trying to debug using cross wind debugger with the vxsim > The object loads without any error and Iam able to go upto some point > with the debugger. But then it just doesn't show the correct file > in it's window where the breakpoint is set. The Wind Shell window says > it's at that point but the debugger shows some different file. > > The directory having that file is also used with the > dir dir_name > command > > It looks as if symbol table is messed up but why I don't know. > The object builds fine and loads. > > Thanks in advance > > Abhinav > > "Any Sufficiently Advanced Technology is Indistinguishable from Magic" - -- - ---------------------- Krishna Samavedam Hughes Network Systems Ph: 301-548-1103 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks simple system calls Date: Tue, 25 Jul 2000 14:36:59 -0400 From: Krishna Samavedam Organization: Hughes Network Systems Message-ID: <397DDE4A.38951972@hns.com> References: <8kpa9i$jq1@netnews.hinet.net> U can use their online man pages : windman kathy wrote: > Hi, > I'm a new user of vxworks. > Does somebody collect vxworks simple system calls (often used) ? > mail me please. > Thanks - -- - ---------------------- Krishna Samavedam Hughes Network Systems Ph: 301-548-1103 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: split as in unix. Date: Tue, 25 Jul 2000 14:44:22 -0400 From: Krishna Samavedam Organization: Hughes Network Systems Message-ID: <397DE005.DE39BD95@hns.com> References: <396F9489.89795754@lucent.com> Port the SPILT code to the vxWorks.. it should be available in GNU - -- krishna Vividh Siddha wrote: > I need to split a image file. Is there a utility as split as in Unix > for splitting a file in VxWorks? - -- - ---------------------- Krishna Samavedam Hughes Network Systems Ph: 301-548-1103 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vx/Works PROXYARP on ethernet for PPP Date: Tue, 25 Jul 2000 14:46:49 -0400 From: Krishna Samavedam Organization: Hughes Network Systems Message-ID: <397DE098.C8CAB419@hns.com> References: <8knedi$43e$1@trsvr.tr.unisys.com> U can use arpAdd and arpDelete with appropriate FLAGS to proxy any IP address. hth krishna Tim Sell wrote: > Does anyone know whether I can use Vx/Works PROXYARP to act as a proxy on an > ethernet interface for a PPP interface? I am using Vx/Works 5.3.1. I.e., I > want the ethernet interface to respond to ARPs for both my ethernet IP > address and my PPP IP address. The SENS manual talks in depth about > PROXYARP for the shared memory backplane interface, but doesn't say anything > about ethernet or PPP. - -- - ---------------------- Krishna Samavedam Hughes Network Systems Ph: 301-548-1103 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: time() and localtime() functions in VxWorks... Date: Tue, 25 Jul 2000 12:37:36 -0700 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0133acb0.9932f871@usw-ex0101-007.remarq.com> References: <%glf5.1334$1P.1674@read1> <3185b300.94ea3f56@usw-ex0101-007.remarq.com> That should read 70/1/1 Rob't Blumberg Embedded Software Developer Anything below is tacked on by Remarq! - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: assembly code in c function Date: Tue, 25 Jul 2000 20:01:45 GMT From: Ioan.Marusca@indsys.ge.com Organization: Deja.com - Before you buy. Message-ID: <8lkrn6$5aq$1@nnrp1.deja.com> References: <397AD5A0.46D5EEBD@club-internet.fr> Below is a short example of C/assembly mix to load the decrementer register on PPC860. Check also GNU documentation. void loadDEC(unsigned int value) { __asm__( " mtspr 22, %0 " : : "r" (value) ); } In article , "Paddy" wrote: > Those warnings are very valid... You have passed addr variable to your > function but are not using it and also your function is declared to return > UINT32 but there is no return statement at function end. > > Patrick > > "Sebastien Fauris" wrote in message > news:397AD5A0.46D5EEBD@club-internet.fr... > > Hi vxWorkers, > > > > how can I write assembly code within a function > > in a C source file ? > > > > Actually my function looks like: > > > > > > UINT32 pciInDword (UINT32 addr) > > { > > /* read a dword from PCI space swapping the bytes */ > > __asm__("lwbrx 3,0,3"); > > > > /* sync I/O operation */ > > __asm__("eieio"); > > __asm__("sync"); > > > > /* return to caller */ > > __asm__("bclr 20,0"); > > } > > > > This seems to be okay, but I get warning messages: > > pciHwAccess.c:1543: warning: unused parameter `addr' > > pciHwAccess.c:1554: warning: control reaches end of non-void function > > > > Any advice is welcome. > > > > Sebastien. > > > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Context Switch Time Date: Tue, 25 Jul 2000 15:43:15 -0400 From: "Chhabra, Vikram" Organization: Lawrence Berkeley National Laboratory Message-ID: <8lksg9$564$1@overload.lbl.gov> Does anyone know what the VxWorks average and Worst Case context switch times are on Motorola Processors. Or any place where I can get this information. Thanks Vikram Chhabra --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxDCOM Feedback? Date: Tue, 25 Jul 2000 20:43:11 GMT From: wmiller_2002@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lku4t$776$1@nnrp1.deja.com> References: <8lhmbu$p7s$1@nnrp1.deja.com> Thanks for the feedback. We will have to take this into account when we make our decision. In article , "DanO" wrote: > VxDCOM.... We are attempting to use it to write an OPC Server under > VxWorks, which is mentioned in their marketing materials. the only problem > is that after probably 5-6 months of trying we are still faced with VxDCOM > specific issues. We are not alone as another company is trying to do this > as well and have hit many of the same problems as we have. We intend to use > it internally in our business. The other company intends to sell it as an > OPC Server Toolkit, I believe. I can't comment on our issues as I'm not the > one developing the OPC Server. Wind River is being somewhat responsive in > trying to get the problems corrected, however it's not quite fixed as of my > last information. This has been the case on 68K and x86, to the best of my > knowledge. I doubt the PPC is any better off, but who knows... > > Good Luck! > > wrote in message > news:8lhmbu$p7s$1@nnrp1.deja.com... > > I am looking at using VxDCOM for an application(PowerPC). Does anyone > > have any recent experience with it? Any feedback (good or bad) would be > > appreciated. > > > > Thanks in advance > > > > > > > > > > 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,comp.realtime Subject: Loading an END driver in VxWorks Date: Tue, 25 Jul 2000 20:42:22 GMT From: milind_gunjal@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lku3d$76m$1@nnrp1.deja.com> Hello All, I have developed an Enhanced NEtwork Driver (END). I would like to add this driver to the target VxWorks system. I have read the VxWorks documentation and have followed the same procedure, which is as follows: in configNet.h: /* parameters to load the END driver */ #define LOAD_FUNC_A_DEV aDevEndLoad #define LOAD_STRING_A_DEV "....." /* parameters goes here */ #define BSP_A_DEV NULL END_TBL_ENTRY endDevTbl[] = { { 0, - --> LOAD_FUNC_A_DEV, LOAD_STRING_A_DEV, BSP_A_DEV, NULL, FALSE}, {0, END_TBL_END, NULL, 0, NULL, FALSE}, } The problem is, when I compile the code, I get an error which says the line (shown with the line and an arrow ie., -->) as an error. The compiler says "Wrong type of initializer" Any suggestions / help would be greatly appreciated, Thanks, Milind. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Structures Date: Tue, 25 Jul 2000 20:53:34 GMT From: rhunnicutt@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lkuob$7ms$1@nnrp1.deja.com> References: <39764060.5B64B6E4@lucent.com> <300DA7E5.E3890B42@mediaone.net> A little off topic here but, I saw your use of assert and checked it out. However, when I use it and it asserts, I get the following error: Assertion failed: val, file ../(file name), line 169 trcStack aborted: error in top frame shell restarted. Is it normal for the trcStack error and shell to restart?? Why would this happen? Thanks! In article <300DA7E5.E3890B42@mediaone.net>, Stephen Banville wrote: ...snip... > Sure, > Make sure you pass the address of the structure and cast it > to (UINT32 *). > > example: > > typedef struct _demo { > unsigned long name; > unsigned long index; > } DEMO; > > void start_demo(void) > { > DEMO *my_demo=NULL; > assert((my_demo = (DEMO *)malloc(sizeof(DEMO)) != NULL); > assert((tid = taskSpawn ("root", 100, 0, 2000,(FUNCPTR)demo, > (UINT32)my_demo, 0,0,0,0,0,0,0,0,0)) > != ERROR); ..snip.. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Flash EPROM memory Date: Tue, 25 Jul 2000 18:08:14 -0400 From: viv Organization: Lucent Technologies Message-ID: <397E0FCE.9287A97B@lucent.com> How can I write to flash EPROM memory? Is the memory mapped to some region of the processor space? What do I need to include to get the functions sysFlashErase, sysFlashWrite etc? How can I boot from EPROM? --------------------------- Newsgroups: comp.os.vxworks Subject: ppc "machine language" reg use? Date: Tue, 25 Jul 2000 17:01:40 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397E0034.42F0E87@mindspring.com> I go through this every time I start using a new h/w arch. we are using the mvme2700 (750 ppc) with vxw/bsp and gcc, etc. In order to understand disassembled code one MUST be figure out the special register usage convention that the compiler/os use. Can someone tell me what these are? It looks like R1 and R9 are "special"... maybe related to the passing mechanisms? frame pointer? what ? Someone please tell me what the secrets are... On the sparc, one register always was used to pass back a functions "return" values, another group were always used to pass (non-floating) values... A read of a particular register would always give you a value of zero; yet another reg could be used as a "bit bucket"... those are the secrets I am looking for for the ppc/gcc/vxw environment. thanks --------------------------- Newsgroups: comp.os.vxworks Subject: Using VxWorks SCSI driver Date: Tue, 25 Jul 2000 21:11:24 GMT From: yeemoney@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lkvpn$8kf$1@nnrp1.deja.com> I'm writing a SCSI utility that I want use on VxWorks. To do this I have to use the VxWorks SCSI driver to send commands. After reading through the help files, I became lost. Does anyone have some sample code that exercises initialization and sends some commands? Specifically, I'm using the LSI 53C895 SCSI controller. So if you have sample code that initializes this chip, that would be great. Thanks in advance. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: ElfHex problems and PPC Date: Tue, 25 Jul 2000 17:42:07 -0700 From: Mike Organization: Cadence Design Systems Message-ID: <397E33DF.4FD511F5@cadence.com> I am getting malloc() errors when using the elfhex utility. Browsing through the WindSurf data base shows I am not the only engineer to encounter this problem. For all SPR's or TSR's that is related to this is, there appears to be no solution documented. I already had the latest cumulative patch (T2spr2) installed so the fix is not in there. I have asked for the source code for this utility and it seems WRS will send it to me (I also have a FAE working on this). Seems to me WRS should owe up to fixing this... maybe they have. Anybody out there willing to share what they had to so to get around this problem? I'd be oblidged. --------------------------- Newsgroups: comp.os.vxworks Subject: Dialing procedure Date: Tue, 25 Jul 2000 16:06:02 -0700 From: RAVINDER KUMAR Organization: Sun Microsystems Message-ID: <397E1D5A.507319E3@Sun.COM> Does anyone know how to access a modem on a serial port and dial out throuh VxWorks? Basically we are planning to use modem to connect our VxWorks system to internet. Is there any 'tip' equivalent command in VxWorks. Any help or suggestion is appreciated Regards Ravinder --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks ROM image and application. Date: Tue, 25 Jul 2000 21:50:12 -0400 From: viv Organization: Lucent Technologies Message-ID: <397E43D4.3445446D@lucent.com> I have the VxWorks image in EPROM starting at the starting of the flash memory. Is there a facility where I can write my object module after the VxWorks image in flash mem like the following. ------------------------------Low ROM address VxWorks image ............................. mymodule.o ------------------------------High ROM address So when VxWorks starts running can I do a loadModule of the module from ROM and start my application task? Has anyone done something like this? Please note that the application is not statically linked with the VxWorks image.. Thanks, Vividh --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Fail to boot with multiple 82559 cards Date: Tue, 25 Jul 2000 23:35:23 GMT From: sbanavar@skystream.com Organization: Deja.com - Before you buy. Message-ID: <8ll87r$f43$1@nnrp1.deja.com> References: <8la4df$okc$1@nnrp1.deja.com> <8lcqh4$ik2$1@nnrp1.deja.com> <8li40n$46g$1@nnrp1.deja.com> <8ljfei$3to$1@nnrp1.deja.com> John, I have modified the source to recognize 82559ER cards. It works generally fine. But this update does not have the fixes WRS has released for fei82557End driver. We have not got the source yet with these fixes - SP 26401 & SP 29158. (Will update you if WRS is giving out this source !) The second and important reason is that (correct me if I'm wrong), fei is not a zero-copy transmission ethernet driver. And our application being throughput sensitive, we have to go with a driver that supports it. Regards, Srinath In article <8ljfei$3to$1@nnrp1.deja.com>, john_94501@my-deja.com wrote: > Hi, > > > ultimately we'll be using 82559ER's on our box. Right now have test > bed > > with 82559's. So we'll end up using 82559ER drivers from intel. > > > > Is there any support/driver from WRS for 82559ER ? > > From what I've heard it is a simple device ID update; if that is the > case then all you should need to do is update the file in your BSP that > finds the PCI devices... > > For the x86 BSPs this is sysNetif.c - look for the function > sys557PciInit() and change the ID that is searched for using > pciFindDevice() at the beginning of the for loop. If you need to support > both types then just add some search logic around the existing for loop > to search for different types of card. > > If this works I'd be interested... > > Regards, > > John... > > Sent via Deja.com http://www.deja.com/ > Before you buy. > - -- Srinath Banavar SkyStream Networks. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Dialing procedure Date: Tue, 25 Jul 2000 17:45:22 -0700 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <397E34A2.D50AA0F5@gv.net> References: <397E1D5A.507319E3@Sun.COM> Reply-To: mkesti@gv.net RAVINDER KUMAR wrote: >Does anyone know how to access a modem on a serial >port and dial out throuh VxWorks? Basically we are >planning to use modem to connect our VxWorks system >to internet. All you need do is open() the port and write() the dialing commands to it. >Is there any 'tip' equivalent command in VxWorks. The above is what tip does. The only (fundamental) difference is that tip provides a user interface that may be invoked from the shell command line, but in vxWorks you do this in your application program. Of course, if you supply a global function that does the open() and write() it can be invoked from the target shell. You might even name it tip()! - -- ======================================================================== 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: Other RTOS's? Date: Tue, 25 Jul 2000 23:17:36 -0400 From: "Robert" Message-ID: References: <396FEC6A.599C6D58@mindspring.com> <3977d11f.1685376@news.pacificnet.net> Thanks Max. Time is a factor for us. Robert "Max" wrote in message news:3977d11f.1685376@news.pacificnet.net... > In my personal opinion, if you have to come with your product fast, > there is probably no alternative to vxWorks, especially if you're > getting BSP/drivers. > > If you have some time on your hands, there are definetely less > expensive and more controlable and easy maintainable (remember, this > is my personal opinion :) RTOSes/toolsets to develop such a system. > Even without mentioning the "source code availabitity" issue for code > coverage analysis to sertify your recorder to a proper flight > criticality level. > > Max. > > > > On Thu, 20 Jul 2000 22:21:30 -0400, "Robert" wrote: > > >Interesting discussion... > > > >We are currently evaluating VxWorks Tornado and got a quote from the > >WindRiver guys. I'm about to spend $75K on two developer seats with a bunch > >of tools, the OEM license and 25 target licenses. > > > >Our project is an embedded Flight Data Recorder to be installed in military > >aircraft. The system includes video (MPEG-2 encoding) and a GUI (eventually > >running on Zinc). Hardware platform is a rugged VME board with on-board MPEG > >encoder. Our hardware manufacturer provides as with a BSP and drivers for > >Tornado. > > > >Is it worth the money we are spending on VxWorks or should we go with a > >different product? > > > >Thanks for your input. > > > >Robert > > > > > >"Al Johnston" wrote in message > >news:396FEC6A.599C6D58@mindspring.com... > >> We are considering abandoning vxWorks (actually, vxWorks is not that > >> bad, WRS is the problem). I would be interested in any opinions/ > >> comments / pointers concerning other alternatives. H/W is moto 2700. > >> > >> Considerations would be ease of migration, source availability, > >> quality / effectiveness of tech support, ethical operation, > >> only selling products that exist (or at least returning the funds > >> when advising the customer that the product doesn't exist), Being > >> able to decide if a product sold to a customer does or does not > >> exist in less then nine months, ability to answer a simple > >> question in less then 3 years, being able to deliver an evaluation > >> copy in less then three months... woops.... sorry. I got a little > >> off topic there. > >> > >> Seriously, suggestions? tnx. > >> > >> (PS: quite laughing Kelly, or I will put in a request that you be > >> pulled in to help us USE vxWorks!) > > > > > > Cambury Technologies, Inc > embedded wireless systems > http://www.cambury.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ppc "machine language" reg use? Date: Wed, 26 Jul 2000 03:35:19 GMT From: pkockritz@home.com (Pete Kockritz) Organization: @Home Network Message-ID: References: <397E0034.42F0E87@mindspring.com> In article <397E0034.42F0E87@mindspring.com>, Al Johnston wrote: > I go through this every time I start using a new h/w arch. we are > using the mvme2700 (750 ppc) with vxw/bsp and gcc, etc. > > In order to understand disassembled code one MUST be figure out the > special register usage convention that the compiler/os use. > > Can someone tell me what these are? It looks like R1 and R9 are > "special"... maybe related to the passing mechanisms? frame pointer? > what ? r0 Volatile register which may be modified during function linkage r1 Stack frame pointer, always valid r2 System-reserved register r3-r4 Volatile registers used for parameter passing and return values r5-r10 Volatile registers used for parameter passing r11-r12 Volatile registers which may be modified during function linkage r13 Small data area pointer register r14-r30 Registers used for local variables r31 Used for local variables or "environment pointers" These should contain what you are looking for... ABI: System V Application Binary Interface: PowerPC Processor Supplement, 1995 (This one is a little strange, the pages are from last to first). EABI: PowerPC Processor Embedded Application Binary Interface, v1.0 These are both from the Motorola PPC823 Engineer's Toolbox: Also, these may be of interest, particularly the PEM. Regards, Pete - -- +-----------------------------------------------------------------+ | Pete Kockritz mailto:pkockritz@home.com | | This space for rent. | +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.object.corba,comp.os.vxworks Subject: multi-threaded corba server Date: Wed, 26 Jul 2000 10:47:11 +0530 From: Amarnath Shanbhag Organization: Lucent Technologies Message-ID: <397E7457.B04AB3C5@lucent.com> Hello, I am using Visibroker 3.2.2 and VxWorks 5.4. I need to design a multi-threaded CORBA server. On startup, it should spawn 2 threads, each servicing a certain Corba object. e.g. We have 2 interfaces interface X { ---methods --- } interface Y { ---methods --- } Corresponding to each interface, we have appropriate implementation classes. class XImpl { ---methods --- } class YImpl { ---methods --- } In the server, I do a taskSpawn() twice to create 2 tasks. I want task t1 to service requests for XImpl object and task t2 to service requests for YImpl object. What should I do in each of these tasks to get the desired result ? Amarnath --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max Resources ? Date: Wed, 26 Jul 2000 13:14:39 +0530 From: Madhu Organization: Lucent Technologies, IPRC Message-ID: <397E96E7.DAF99237@lucent.com> References: <3976A86D.93FE3DC0@lucent.com> <8l6dui$7qb$1@nnrp1.deja.com> <3976FA96.9B5E5A6A@lucent.com> <8lctbd$kgj$1@nnrp1.deja.com> Hi. We have a networking product that allows a number of users to connect. Each Client uses PPP to connect. This would cause allocation of watch dog timers per session, a route add to handle the new client, new sockets etc. We found here that by increasing the number of system clusters in netbuflib.h. we were able to connect more clients. I think NUM_FILES will not help to connect more than that allocated by the system clusters. Madhu > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: verbosity macro Date: Wed, 26 Jul 2000 08:36:13 +0100 From: Graham Baxter Organization: Graham Baxter (Software) Limited Message-ID: <397E94ED.CB1E3D77@NOSPAM.bcs.org.uk> References: <3978A49A.8ED28F97@club-internet.fr> Sebastien, See the ANSI function vprintf(). It does just what you need. Regards, Graham Baxter gbaxter@NOSPAMbcs.org.uk Freelance Software Engineer Sebastien Fauris wrote: > Hi folks, > > I have a macro very convenient to debug my code. > For that I compile my code in debug mode (flag DRV_DEBUG_MODE defined), > and I use the following macro to display my debug messages with a > certain level > of verbosity that I can change (by a global variable): > > Macro: > #ifdef DRV_DEBUG_MODE > #define VERB_DRV(level,a1,a2,a3,a4,a5,a6,a7) \ > if (level <= drv_dbg) \ > {logMsg(a1,a2,a3,a4,a5,a6,a7);} > #else > #define VERB_DRV(level,a1,a2,a3,a4,a5,a6,a7) {} > #endif > > and use it in my code: > > #ifdef DRV_DEBUG_MODE > int drv_dbg = 0; /* default 0, but can be changed to a higher verbosity > level after */ > #endif > > int toto (void) > { > > VERB_DRV (2,"my debug message displayed only in debug mode when > drv_dbg global variable set higher or equal to > 2\n",0,0,0,0,0,0); > } > > I want to do the same macro, but using printf instead of logMsg. > How can I do to pass the arguments to the printf (variable number of > arguments) > while logMsg arguments is fixed to 6 and this in a macro definition ? > > Thanks a lot, > > Sebastien. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Intel 82577 ethernet device driver required for PPC Date: Wed, 26 Jul 2000 07:38:04 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lm4gs$3ah$1@nnrp1.deja.com> References: <397D6585.E579A02C@compd.com> Hi, I'm assuming that you are using VxWorks 5.4 (Tornado 2.0), in which case you'd be better off starting with the END version of the driver (fei82557End I think). That should be a lot simpler to configure, and is also the preferred driver (i.e. the one that gets maintained). If you need it, I can send you some code for initialising this driver on an ARM board - won't help with the endian issues since ARM is little endian, but it will show you the END load string etc. If you want it email me at my other email john_94501 at yahoo dot com (substitute at & dot for symbolic versions of course!) Regards, John... In article <397D6585.E579A02C@compd.com>, pbrewer.removethis@compd.com wrote: > Hi, > > I have a board that has a PowerPC 750 with a Motorola 106 PPC to PCI > bridge. I have a Intel 82558 (seems to be more or less equivalent to a > 82557) PCI ethernet device on the PCI bus. This device (being an intel > device) is little endian. I have a vxWorks device driver which came > from the vxWorks device driver kit called if_fei.c. This driver is for > a 82557 but the documentation seems to show that its appropriate for a > 82558. > > The device sets up an area of shared memory which is accessed by the > device and the processor. I haven't got any documentation for the 82557 > (didn't think I'd need it considering I was using an off the shelf > driver) but it seems that the processor has to setup a ring of transmit > and receive buffers which are then read and filled by the ethernet > device. On my system the processor sets up the ring as big-endian and > device reads and writes as little endian. Needless to say it doesn't > work. Does anyone have a byte swapped device driver I can use? Failing > that does anyone know which parts I need to byte swap in order to get my > existing driver to work? > > Regards > Paul Brewer > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 2 vs Tornado 1 Date: Wed, 26 Jul 2000 11:20:45 +0200 From: jph Organization: UUNET Message-ID: <397EAD6D.B7785775@francimel.com> I have developed an application (a task) for SH4/VxWorks with Tornado 1. Can it be used "as is" by a customer using Tornado 2 ? --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Jul 26 04:40:00 2000 From: Michael_P_Dragicevich@res.raytheon.com Date: Wed Jul 26 04:40:02 PDT 2000 Subject: read error from network, timeout VxWorks problem, We have a host (SunOS) disk directory that is nfs mounted from our target (MVME162LX/vxWorks 5.3). The target has a task executing that reads a number of files that are each 3145728 bytes is size. The reads of each file are done in 262144 byte blocks, and these blocks are written to a file opened on the target's disk drive. The target task is executing at a low priority (225). Occassionally we get the following target error during a read: ->Read error from network: Connect timed out ->Connection Closed Anybody have insight as to why the timeout is occurring? Is there a way to increase the timeout value? Or is there another solution? Thanks, Mike Dragicevich Raytheon Portsmouth, RI 02871 401-842-3476 From vxwexplo-errs@csg.lbl.gov Thu Jul 27 04:03:14 2000 From: Vxworks Exploder Date: Thu Jul 27 04:03:17 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 27 04:03:11 PDT 2000 Subject: Re: Loading an END driver in VxWorks Subject: read error from network, timeout Subject: Re: Raw IP Subject: C++ Task Creation Subject: mv2700 - unknown symbol "__eabi" Subject: boot my boards from flash memory Subject: Re: MPC750 and MPC8260 Subject: Any one for consulting Subject: Making DOSFS2 Multi User / MultiTasking Subject: Somebody help me Subject: Re: Booting using TSFS ------------------------------------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Loading an END driver in VxWorks Date: 26 Jul 2000 07:53:02 -0500 From: vanbaren@falcon.si.com (vanbaren_gerald) Organization: Smiths Industries Message-ID: <397ed11e@news.si.com> References: <8lku3d$76m$1@nnrp1.deja.com> milind_gunjal@my-deja.com writes: >Hello All, >I have developed an Enhanced NEtwork Driver (END). I would like to add >this driver to the target VxWorks system. I have read the VxWorks >documentation and have followed the same procedure, which is as follows: >in configNet.h: >/* parameters to load the END driver */ >#define LOAD_FUNC_A_DEV aDevEndLoad >#define LOAD_STRING_A_DEV "....." /* parameters goes here */ >#define BSP_A_DEV NULL >END_TBL_ENTRY endDevTbl[] = >{ >{ 0, >--> LOAD_FUNC_A_DEV, > LOAD_STRING_A_DEV, > BSP_A_DEV, > NULL, > FALSE}, > {0, END_TBL_END, NULL, 0, NULL, FALSE}, >} >The problem is, when I compile the code, I get an error which says the >line (shown with the line and an arrow ie., -->) as an error. The >compiler says "Wrong type of initializer" >Any suggestions / help would be greatly appreciated, >Thanks, >Milind. >Sent via Deja.com http://www.deja.com/ >Before you buy. You need to typecast the load function or (preferably) declare it as a function prototype (with an appropriate type). Function prototype example, should work... END_OBJ* LOAD_FUNC_A_DEV(char*, void*); END_TBL_ENTRY endDevTbl[] = { { 0, LOAD_FUNC_A_DEV, LOAD_STRING_A_DEV, BSP_A_DEV, NULL, FALSE }, {0, END_TBL_END, NULL, 0, NULL, FALSE}, } Just guessing on the typecasting, probably wrong, but you get the drift... END_TBL_ENTRY endDevTbl[] = { { 0, (END_OBJ* (*endLoadFunc) (char*, void*))LOAD_FUNC_A_DEV, LOAD_STRING_A_DEV, BSP_A_DEV, NULL, FALSE }, {0, END_TBL_END, NULL, 0, NULL, FALSE}, } gvb - -- +---------------------------------------------------------------------------+ | Jerry Van Baren / vanbaren_gerald@si.com / Grand Rapids Mi / 616-241-7973 | | My employer is a company. Companies are artifacts of a legal system. | |________________Artifacts are incapable of having opinions.________________| --------------------------- Newsgroups: comp.os.vxworks Subject: read error from network, timeout Date: Wed, 26 Jul 2000 07:40:13 -0400 From: Michael_P_Dragicevich@res.raytheon.com Organization: Lawrence Berkeley National Laboratory Message-ID: <8lmkoq$jpo$1@overload.lbl.gov> VxWorks problem, We have a host (SunOS) disk directory that is nfs mounted from our target (MVME162LX/vxWorks 5.3). The target has a task executing that reads a number of files that are each 3145728 bytes is size. The reads of each file are done in 262144 byte blocks, and these blocks are written to a file opened on the target's disk drive. The target task is executing at a low priority (225). Occassionally we get the following target error during a read: - ->Read error from network: Connect timed out - ->Connection Closed Anybody have insight as to why the timeout is occurring? Is there a way to increase the timeout value? Or is there another solution? Thanks, Mike Dragicevich Raytheon Portsmouth, RI 02871 401-842-3476 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Raw IP Date: Wed, 26 Jul 2000 22:40:44 GMT From: "Laszlo Szebelledy" Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: References: <8la2v8$k4v$1@news.umbc.edu> I believe you need to put your end driver in promiscuous mode (to receive all packets) and then set up a SNARF protocol. You will then effectively have a Network Sniffer. Look in the Network Protocol Toolkit for more info. "cedilnik andrej" wrote in message news:8la2v8$k4v$1@news.umbc.edu... > Hello! > > Any idea on how to get network packets from hardware layer? > > Andy > --------------------------- Newsgroups: comp.os.vxworks Subject: C++ Task Creation Date: Thu, 27 Jul 2000 09:07:17 +1000 From: Chris Chock Organization: Customer of Connect.com.au Pty. Ltd. Message-ID: <397F6F25.9A776C02@hunterlink.net.au> Is it possible to have a class instance inherit from VXWTask such that it spawns a task with one of its own methods as the entry point when the constructor is called ? Thanks Chris Chock --------------------------- Newsgroups: comp.os.vxworks Subject: mv2700 - unknown symbol "__eabi" Date: Wed, 26 Jul 2000 15:47:56 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397F406C.F6A45133@mindspring.com> Hi guys, this one is probably easy. I compiled a hello world program per the vxw P.G. direction. When I load it the loader tells me undefined symbol: __eabi (binding 1 type 0) nm on the object, and sure enough there is an entry U __eabi What am I doing wrong? tor 2, mv2600/750 cpu, host is solaris. tnx. --------------------------- Newsgroups: comp.os.vxworks Subject: boot my boards from flash memory Date: Wed, 26 Jul 2000 20:05:24 -0400 From: no-one Organization: fac Admin - Univ Ottawa Message-ID: <397F7CC4.AD5BCC1A@admin.uottawa.ca> I would like to boot my boards from flash memory(e.g. via TrueFFS), but still I would like to establish a shared memory backplane network. Is it technically possible? Thanks. Wei (wei_huo@hotmail.com) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC750 and MPC8260 Date: Wed, 26 Jul 2000 20:53:51 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8lnj4s$6ef$1@nnrp1.deja.com> References: <397CF308.4741D12C@radixtek.com> <8lkgm0$s67$1@nnrp1.deja.com> <397F2B11.F7DA1AC7@radixtek.com> In article <397F2B11.F7DA1AC7@radixtek.com>, Eric Monsler wrote: > Dan, > > Thank you for your helpful post. > > I've got a few follow-up questions, that I hope you can help me with. > I'm not a HW designer, I'm a SW guy, so please forgive any obvious > blunders. > (I forwarded your previous response to my HW counterpart) That's ok, I'm a SW guy too, as the HW guys always remind me :-) > 1) A 750 will not be able to access the 8260-local bus, correct? So, if > we wanted to prevent the 8260-CPM from ever becoming bus master on the > 60x bus, we could only use the 8260 DMA channels between the peripherals > and the 8260 dual-port memory. Does this sound correct? No, not exactly. First, the 750 can access the 8260 local bus. In fact that's where all my ethernet buffers go. The 8260 local-bus is not cacheable was my point. So if the 750 has the caches enabled, the 8260 localbus mustbe marked non-cacheable. That's just good information, lest you design a system and think everything is cacheable. > 2) If we ran as described above, does this remove the 8260 bus anomalies > that you describe? We had read about various anomalies with two 60x > masters and the PCI bridge, but had hoped to eliminate them by > preventing the 8260 from becoming bus master. I don't think it would. The anomalies we found were due to the 8260 being the bus arbiter. Our boards just have the 750/8260/sram/flash on the 60x bus, no PCI. And the 8260 is the bus arbiter. The anomalies we saw were between just the 750 and the 8260. Now if you make sure to put all your buffers (enet, atm, fcc, scc, mcc, etc..) on the local bus, then you might be able to assume the 8260 CPM will never be a 60x bus master. I think that might be too big a design constraint though. It won't leave you much flexibility for future software architecture changes without big hardware changes. Hopefully with newer Revs of the 8260 some of the anomalies have gone away, not awry :-) > And finally, I see you're posting from SDSU. Are any of your board > designs, PLD's, and/or BSP's published or otherwise in the public domain > (e.g., funded by gov't) so that I and the board HW designer could take a > look? I guess the .edu is a little misleading. I do full-time embedded SW consulting. The reasons I use the sdsu.edu address are that sometimes clients don't want their name out on the Internet, and I've had the sdsu account since my undergrad days (many moons ago) and people know they can always reach me there, no matter where I'm consulting.. hth, Dan Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Any one for consulting Date: Thu, 27 Jul 2000 02:38:25 GMT From: Nalit Patel Organization: @Home Network Message-ID: <397F9D09.DAADD661@home.com> We have a project in CA for 10 months need 2 good people with 5+ years of experience with Vxworks, C++, PCI, VME experience.. Excellent rate based on experience from Call me 908-559-7497 Nalit Patel Email nalit@netscape.net --------------------------- Newsgroups: comp.os.vxworks Subject: Making DOSFS2 Multi User / MultiTasking Date: Thu, 27 Jul 2000 03:01:48 GMT From: epooleNO@UCEnetis.com (RKT Technologies, Inc.) Organization: Software Consultants to Medical Device Manufacturers Message-ID: <397fa5ff.1123658143@news.netis.com> We need to implement a POSIX wrapper around the VxWorks DOSFS2 filesystem, and implement file and directory permissions and all the rest. Does anyone know if anyone has done this already, and is there any source code available, either free or for a cost? Thanks... >-----------------------------------------------------------------< Eric Poole Specializing in Software Development for RKT Technologies, Inc. Medical Devices / Regulated Industries PO Box 1100 ISO 9001 / FDA Quality System Regulation Londonderry, NH 03053 Phone: 603-437-1811 Fax: 603-425-6475 >-----------------------------------------------------------------< E-mail: epooleNO@UCEnetis.com (Remove "NO" and "UCE" to reply by e-mail; use epoole-at-netis-dot-com) >-----------------------------------------------------------------< Qualifications and Background: see http://www.netis.com/rkt >-----------------------------------------------------------------< --------------------------- Newsgroups: comp.os.vxworks Subject: Somebody help me Date: Thu, 27 Jul 2000 01:24:59 GMT From: ellin_lin@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lo319$i84$1@nnrp1.deja.com> I am a tenderfoot of vxWorks. Now I have a problem about DHCP: I established a DHCP server and set the IP range. Then I test it with a Win98 as DHCP client and found the client cann't get IP form the server. Can somebody tell me what may cause this? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Booting using TSFS Date: Thu, 27 Jul 2000 02:42:51 GMT From: Nalit Patel Organization: @Home Network Message-ID: <397F9E12.B6744846@home.com> References: <8idnb9$bar1@overload.lbl.gov> Surjit If interested in migrating to US let me know send me a mail with your credentials npatel@home.com sujit v kuruvilla wrote: > > This is a multipart message in MIME format > -------=_MM9611739422765 > Content-Type: text/plain > Content-Transfer-Encoding: 7bit > > Hi all, > > I am working on a MIPS R7000 capella evaluation board. We tried to boot > up the board using TSFS. Followed whatever is to be done as per tornado > help. In the virtual console that comes up while launching the target, It > gives out the Board configuration, and then shows WDBTask exception. I am > attaching the log file that was generated. > > Please do help out and give me inputs as to how to solve this problem. > > regards, > sujit > > -------=_MM9611739422765 > Content-Type: text/plain; name="log1.txt" > Content-Disposition: inline; filename="log1.txt" > Content-Transfer-Encoding: 7bit > > User Name : suku > Started : Fri Jun 16 22:08:39 2000 > Target Server Name : uart@test_1 > Target Name : uart > Target Server Options : tgtsvr uart -V -B wdbserial -Bt 10 -Br 2 -Bd C:\Tornado\target\config\tsfs\logs\log1.txt -d COM1 -R C:\Tornado\target\proj\uartboot\default\ -RW -redirectIO -C -c C:\Tornado\target\proj\uartboot\default\vxWorks > Host : Microsoft Windows NT version 4.0 (Build 1381) > Timeout value : 10 second(s) > Request re-send Max : 2 > > 0 WDB_TARGET_PING Fri Jun 16 22:08:39 2000 > > 1 WDB_TARGET_CONNECT Fri Jun 16 22:08:39 2000 > 0 Out status: Ok > WDB_TGT_INFO > WDB_AGENT_INFO > agentVersion 2.0 > mtu 1006 > mode WDB_MODE_TASK|WDB_MODE_EXTERN > WDB_RT_INFO > rtType WDB_RT_VXWORKS > rtVersion 5.4 > cpuType 44 > hasFpp 1 > hasWriteProtect 0 > pageSize -1 > endian 1234 > bspName NKK RM7000_ND7000LBG-100 > bootline > memBase 0x80000000 > memSize 0x4000000 > numRegions 0 > hostPoolBase 0x801c15f8 > hostPoolSize 0x7c7d41 > > 2 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > In WDB_MEM_REGION > baseAddr 0x80010000 > numBytes 314800 > param 0 > 0 Out status: Ok > UINT32 0xffff64bd > > 3 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > In WDB_MEM_REGION > baseAddr 0x80010000 > numBytes 373072 > param 0 > 0 Out status: Ok > UINT32 0xffff7f10 > > 4 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > In WDB_MEM_REGION > baseAddr 0x8006b170 > numBytes 48824 > param 0 > 0 Out status: Ok > UINT32 0xffffc693 > > 5 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > In WDB_MEM_REGION > baseAddr 0x80077030 > numBytes 20304 > param 0 > 0 Out status: Ok > UINT32 0xffffb748 > > 6 WDB_MEM_WRITE Fri Jun 16 22:08:42 2000 > In WDB_MEM_XFER > numBytes 7 > destination 0x809892e0 > source 0xe2f6dc > 0 Out status: Ok > > 7 WDB_DIRECT_CALL Fri Jun 16 22:08:42 2000 > In WDB_CTX_CREATE_DESC > stackBase 0x0 > stackSize 0x0 > entry 0x80027e14 > args[0] -2137484576 > args[1] 2 > args[2] 0 > args[3] 0 > args[4] 0 > args[5] 0 > args[6] 0 > args[7] 0 > args[8] 0 > args[9] 0 > name NULL > priority 0 > options 0x0 > redirIn 0 > redirOut 0 > redirErr 0 > 2 Out Error: RPC_CANTDECODERES > > 8 WDB_EVENT_GET Fri Jun 16 22:08:59 2000 > 0 Out status: Ok > WDB_EVT_DATA > evtType WDB_EVT_VIO_WRITE > WDB_MEM_XFER > numBytes 332 > destination 0 > source 0x4427b0 > > 9 WDB_EVENT_GET Fri Jun 16 22:08:59 2000 > 0 Out status: Ok > WDB_EVT_DATA > evtType WDB_EVT_EXC > WDB_EXC_INFO > numInts 4 > WDB_CTX > contextType WDB_CTX_TASK > contextId 0x83ffa290 > vec 8 > pEsf 0x83ff9ed8 > > 10 WDB_TARGET_DISCONNECT Fri Jun 16 22:10:30 2000 > 0 Out status: Ok > > -------=_MM9611739422765-- --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Jul 27 06:26:41 2000 From: "G. Nagaraj" Date: Thu Jul 27 06:26:43 PDT 2000 Subject: Query about comp.os.vxworks This is a multi-part message in MIME format. ------=_NextPart_000_0008_01BFF7FD.89A2D600 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello VxWorkers, I wanted to use the online newsgroup comp.os.vxWorks. When I try to configure my news reader application (Outlook Express or Netscape), I am asked for the NNTP server name (or ip address). Can anyone tell me which is the NNTP server to access this news group on the net. Any help in the way of pointers, documents, would be highly appreciated. Thanks & Regards, G. Nagaraj Senior Software Engineer Wipro Technologies 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.wipro.com ------=_NextPart_000_0008_01BFF7FD.89A2D600 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Hello VxWorkers,

 

I wanted to use the online newsgroup comp.os.vxWorks. When I try = to configure my news reader application (Outlook Express or Netscape), I am = asked for the NNTP server name (or ip address). =

 

Can anyone tell me which is the NNTP server to access this news = group on the net.

 

Any help in the way of pointers, documents, would be highly = appreciated.

 

Thanks & Regards,

 

G. Nagaraj =

Senior Software Engineer =

Wipro Technologies =

37, Castle Street, Ashok Nagar, =

Bangalore - 560 025, India =

Phone: +91-80-536-7297 Ext: 315 =

Fax: +91-80-557-6032 =

www.wipro.com<= /font>

 <= /p>

------=_NextPart_000_0008_01BFF7FD.89A2D600-- From vxwexplo-errs@csg.lbl.gov Thu Jul 27 08:14:53 2000 From: "Michael Lawnick" Date: Thu Jul 27 08:14:55 PDT 2000 Subject: Re: mv2700 - unknown symbol "__eabi" Date sent: Thu, 27 Jul 2000 04:12:01 -0700 (PDT) To: vxworks_users@csg.lbl.gov From: vxwexplo@lbl.gov (the vxWorks Users Group Exploder) Subject: comp.os.vxworks newsdigest Hi Al, > Hi guys, > > this one is probably easy. I compiled a hello world program per > the vxw P.G. direction. When I load it the loader tells me > undefined symbol: __eabi (binding 1 type 0) > nm on the object, and sure enough there is an entry > U __eabi > > What am I doing wrong? Let me guess, your functions name is main() ? This is not allowed in vxWorks, just call it hello() or test() or al(). 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 Fri Jul 28 04:03:13 2000 From: Vxworks Exploder Date: Fri Jul 28 04:03:16 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 28 04:03:08 PDT 2000 Subject: Query about comp.os.vxworks Subject: re: mv2700 - unknown symbol "__eabi" Subject: comp.os.vxworks newsdigest Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Subject: Compressed Image Subject: Re: JWorks and JNI Subject: Re: 82559ER Driver Subject: Re: Compiling many files w/ one command Subject: Re: Compiling many files w/ one command Subject: Adaptec AIC-7892A Ultra160/m Driver Subject: Re: Query about comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Subject: VxWorks Video Driver Subject: 64-bit division Subject: Re: Adaptec AIC-7892A Ultra160/m Driver Subject: Re: Compiling many files w/ one command Subject: Re: Web Server under VxWorks Subject: Re: Booting using TSFS Subject: Re: vxMemProbe Subject: Re: ppc "machine language" reg use? Subject: mvme2700 -update rtc date? Subject: Anyone using the LSI sym53C895 SCSI controller on VxWorks Subject: Anyone using the LSI sym53C895 SCSI controller on VxWorks Subject: Re: Query about comp.os.vxworks Subject: workspace? Subject: Re: Raw IP Subject: Re: vxMemProbe Subject: RTOS comparisons Subject: VxWorks & ARM Dev Board Subject: Re: Using Single Serial Line with VxWorks Subject: Re: VxWorks & ARM Dev Board Subject: Re: Using Single Serial Line with VxWorks Subject: Re: RTOS comparisons Subject: Re: VxWorks & ARM Dev Board Subject: Re: How to add command line into com1 ? Subject: Re: Using Single Serial Line with VxWorks Subject: Using Single Serial Line with VxWorks Subject: Getting rid of "transfer timed out" message Subject: Re: 2700(750) ppc - where MSR(EE) initially set? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Query about comp.os.vxworks Date: Thu, 27 Jul 2000 19:04:49 +0530 From: "G. Nagaraj" Organization: Lawrence Berkeley National Laboratory Message-ID: <8lpg95$1lh$1@overload.lbl.gov> This is a multi-part message in MIME format. - ------=_NextPart_000_0008_01BFF7FD.89A2D600 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hello VxWorkers, I wanted to use the online newsgroup comp.os.vxWorks. When I try to configure my news reader application (Outlook Express or Netscape), I am asked for the NNTP server name (or ip address). Can anyone tell me which is the NNTP server to access this news group on the net. Any help in the way of pointers, documents, would be highly appreciated. Thanks & Regards, G. Nagaraj Senior Software Engineer Wipro Technologies 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.wipro.com - ------=_NextPart_000_0008_01BFF7FD.89A2D600 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

Hello VxWorkers,

 

I wanted to use the online newsgroup comp.os.vxWorks. When I try = to configure my news reader application (Outlook Express or Netscape), I am = asked for the NNTP server name (or ip address). =

 

Can anyone tell me which is the NNTP server to access this news = group on the net.

 

Any help in the way of pointers, documents, would be highly = appreciated.

 

Thanks & Regards,

 

G. Nagaraj =

Senior Software Engineer =

Wipro Technologies =

37, Castle Street, Ashok Nagar, =

Bangalore - 560 025, India =

Phone: +91-80-536-7297 Ext: 315 =

Fax: +91-80-557-6032 =

www.wipro.com<= /font>

 <= /p>

- ------=_NextPart_000_0008_01BFF7FD.89A2D600-- --------------------------- Newsgroups: comp.os.vxworks Subject: re: mv2700 - unknown symbol "__eabi" Date: Thu, 27 Jul 2000 16:35:29 +0200 From: "Michael Lawnick" Organization: Lawrence Berkeley National Laboratory Message-ID: <200007271434.QAA02438@net.softec.de> Date sent: Thu, 27 Jul 2000 04:12:01 -0700 (PDT) To: vxworks_users@csg.lbl.gov From: vxwexplo@lbl.gov (the vxWorks Users Group Exploder) Subject: comp.os.vxworks newsdigest Hi Al, > Hi guys, > > this one is probably easy. I compiled a hello world program per > the vxw P.G. direction. When I load it the loader tells me > undefined symbol: __eabi (binding 1 type 0) > nm on the object, and sure enough there is an entry > U __eabi > > What am I doing wrong? Let me guess, your functions name is main() ? This is not allowed in vxWorks, just call it hello() or test() or al(). 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 ============================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Date: Thu, 27 Jul 2000 11:13:18 -0500 From: "Derek Opitz" Organization: Motorola CIG Message-ID: <8lpn0n$7ct2@nntp.cig.mot.com> References: <397F5375.4D3A441A@mindspring.com> intUnlock will enable MSR[EE]. Al Johnston wrote in message news:397F5375.4D3A441A@mindspring.com... > I am getting a "uninitialized interrupt" from my pcc "decrementer" > register. I am not using it for anything at the moment, because I have > moved the sysClk/timestamp functions elsewhere. > > Apparently I need to install a noop isr before MSR(EE) is initially > enabled. Problem is, I cant find where that is. I was expecting > it to be in the usrInit/sysHwInit thread of execution, but I cant > find it. the only thing that comes close is a call to intEnable(), > which of course I do not have source to... but dissasembling it seemed > to show that it was not modifying MSR(EE).... but then again, i am > not very good at reading pcc machine language yet. > > Can someone point me in the right direction? > > thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Compressed Image Date: Thu, 27 Jul 2000 13:19:10 GMT From: ddinker@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lpcs9$f9l$1@nnrp1.deja.com> Hi, We load a bootstrap followed by an image to ROM. The bootstrap does something and then passes the control to start of image. All is fine. Now we need to compress the image. Am setting the ROM_TEXT_ADRS to the start of the image. But soemthing is wrong. It seems that inflation is going on properly as the system does not come up. I looked into romStart of BSP. It does these main 2 things: 1. Copies to ROM_TEXT_ADRS from ronInit binArrayStart-romInit bytes 2. Inflate from binArrayStart to RAM_DST_ADRS, &binArrayEnd- binArrayStart bytes Any idea what's happening here? What is this binArrayStart? What would be the address of romInit? Same as the start of image? What I guess is happenning is that while inflating it is not getting the proper address and size of the image as contrary to "normal" image our image does not start from begining of ROM. Am new to world of VxWorks and all kinds of lessons, ideas, leads will be oasis in the desert! :-) Regards, Dinker Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: JWorks and JNI Date: Thu, 27 Jul 2000 15:24:57 +0200 From: Gilbert Cabillic Organization: Irisa - Campus de Beaulieu - RENNES Message-ID: <39803829.8050D85E@irisa.fr> References: <397C6BDC.9F552364@adichi.com> Gretchen Patti wrote: > > Is anybody using JNI under Personal JWorks? > > I'm creating a Java application for an embedded system under vxWorks. > We're using Personal JWorks 3.0. I'm trying to use JNI to call a > native C routine. The System.loadLibrary call works OK, but when I > call the native method I get an UnsatisfiedLinkError. Help!! > Hi. Jni calls on JWorks are supported and we had no problems to use them. I can recommend you to test a basic 'hello world' Java application using a jni call for printing on the screen to verify that the configuration of Jworks is OK. Gilbert. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 82559ER Driver Date: Tue, 25 Jul 2000 19:19:42 -0700 From: "Sheela Kashyap" Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: John, Srinath, thanks so much for your replies. I changed the PCI Vendor ID in the sysFei82557End.c file and it worked. That's all I needed to do. I haven't had the chance to test it extensively. I was however able to get the latest code from Wind River and will have to test with that. Srinath, if you have the support contract with Wind River, you can ask them for the latest fei82557End driver and they should be sending it to you. "Sheela Kashyap" wrote in message news:snkgjttfjjt18@corp.supernews.com... > Hi, > > We have an 82559ER ethernet controller on our board. As far as I know, the > only *major* differences, from the software perspective between that and > 82559 are "Wake on LAN" (excluded in 82559ER) and the device id (0x1209 for > 82559ER). If that is the case, is it possible for one to use the fei82557End > driver from WRS for the 82559ER, including just the changes for the device > id? Does anybody have any idea of caveats etc.? > > Appreciate any help. > > Thanks in advance, > Sheela. > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Compiling many files w/ one command Date: Thu, 27 Jul 2000 14:25:52 GMT From: fububitok@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lpgp9$ih7$1@nnrp1.deja.com> References: <8lnvs8$g1a$1@nnrp1.deja.com> Hi, My understanding is that using the "extern" keyword in the right places will do the trick. Variables declared in a .c file get allocated space in the corresponding .o file. "extern" allows one .o file to refer to a variable in another .o file. The linker needs to know how to resolve these references when it makes a loadable module -- e.g. it needs to know that s1 in 0.c refers to a semaphore defined in 1.c. "extern" gives the linker this ability. At least, that's my understanding. I'm a newbie also. Good luck, b Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Compiling many files w/ one command Date: Thu, 27 Jul 2000 10:00:01 -0500 From: "Matthew Majka" Organization: NASA, Kennedy Space Center Message-ID: <8lpicp$go7@niven.ksc.nasa.gov> References: <8lnvs8$g1a$1@nnrp1.deja.com> wrote in message news:8lnvs8$g1a$1@nnrp1.deja.com... > I'm using Tornado 2.0. I made one header file that contains all my > semaphores, variables, #defines, etc. Good software practice, right? I > made several files each containing two tasks, a consumer and a > producer. Good software practice, right? Then why won't Tornade > compile all them together? Each file gets compiled as a separate > object, with its own copies of the semaphores, variables, etc. This > means none of my tasks can communicate. Why does this seem stupid to > me? This would be fixed by compiling them all on the same line: Let me get this straight. You have one header file that declares all of your variables and semaphores and you are including that file in multiple C files? You don't have a Tornado issue, you have very basic C issue on what a header file actually does. By putting the declaration code in a header file, each C file that includes that header automatically gets the declaration code (ie the variable gets declared in each C file and since they are all global the linker comes up with multiple definitions...). BTW, making all of your variables global is not good software practice.... --------------------------- Newsgroups: comp.os.vxworks Subject: Adaptec AIC-7892A Ultra160/m Driver Date: Thu, 27 Jul 2000 15:49:07 GMT From: jmaguire@analogic.com (Jared Maguire) Message-ID: <39805793.682733078@bilbo> I very much need to make an AIC-7892A Ultra160/m PCI SCSI Controller work with VxWorks running on a more or less plain vanilla Intel platform. I've been searching, but have yet to find any drivers for this card. Are there drivers available / in progress? Is there another driver I can hack up to work? - -Jared jmaguire@analogic.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Query about comp.os.vxworks Date: Thu, 27 Jul 2000 10:52:48 -0400 From: krishna balusu Organization: Hughes Network Systems Message-ID: <39804CC0.EE3E7061@hns.com> References: <8lpg95$1lh$1@overload.lbl.gov> That would be the nntp server for ur organization. ask your sysadm for the server name. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Date: Thu, 27 Jul 2000 11:38:06 -0500 From: "Derek Opitz" Organization: Motorola CIG Message-ID: <8lpof6$7ct3@nntp.cig.mot.com> References: <397F5375.4D3A441A@mindspring.com> Or more specifically, intUnlock(_PPC_MSR_EE); sets the MSR[EE] bit to enable, which I think is a 1, but not sure about that. Al Johnston wrote in message news:397F5375.4D3A441A@mindspring.com... > I am getting a "uninitialized interrupt" from my pcc "decrementer" > register. I am not using it for anything at the moment, because I have > moved the sysClk/timestamp functions elsewhere. > > Apparently I need to install a noop isr before MSR(EE) is initially > enabled. Problem is, I cant find where that is. I was expecting > it to be in the usrInit/sysHwInit thread of execution, but I cant > find it. the only thing that comes close is a call to intEnable(), > which of course I do not have source to... but dissasembling it seemed > to show that it was not modifying MSR(EE).... but then again, i am > not very good at reading pcc machine language yet. > > Can someone point me in the right direction? > > thanks. --------------------------- Newsgroups: comp.lang.c-programming,comp.os.vxworks,comp.sys.powerpc,comp.sys.powerpc.tech Subject: VxWorks Video Driver Date: Thu, 27 Jul 2000 20:43:17 GMT From: "Rogers" Organization: @Home Network Canada Message-ID: I am developing an interface for a video driver on VxWorks. I am trying to convert some the required Voodoo3 BIOS functionality to VxWorks on a PowerPC 604, with little success. So far, I have (I think) properly located the card using pciFindDevice(), determined the BAR configuration using pciConfigInXXXX(), and set the BAR addresses using pciConfigOutXXXX(). The questions I have are: 1. Does this configuration require memory mapping from virtual to physical? It appears that sysMmuAddMap() is not available for the 604. 2. Are sysOutWord() and sysOutByte() the right functions to use to write to PCI card addresses? How does this work? I'm fairly new to this, so any help at all would be appreciated. Thanks, Rob rtargosz@altsoftware.com --------------------------- Newsgroups: comp.os.vxworks Subject: 64-bit division Date: Thu, 27 Jul 2000 14:53:29 -0400 From: "Shaila Kambli" Organization: SNET Internet Message-ID: Hello all, Does anyone know of any utility functions for doing 64-bit division where, the dividend and the divisor each, is stored in two 32-bit integers ? The compiler and the processor can handle only 32-bit operations. e.g. I would like to calculate the error Rate int32 errRateLo, errRateHi, errCntLo, errCntHi, totalBitsLo, totalBitsHi; error rate = error count / total Bits (Using Lo and Hi values) Thanks. - - Shaila Kambli --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Adaptec AIC-7892A Ultra160/m Driver Date: Thu, 27 Jul 2000 16:47:52 -0500 From: "Derek Opitz" Organization: Motorola CIG Message-ID: <8lqak0$7ct4@nntp.cig.mot.com> References: <39805793.682733078@bilbo> We had to write our own. Not sure if it was an AIC-7892A though. Jared Maguire wrote in message news:39805793.682733078@bilbo... > I very much need to make an AIC-7892A Ultra160/m PCI SCSI Controller > work with VxWorks running on a more or less plain vanilla Intel > platform. I've been searching, but have yet to find any drivers for > this card. > > Are there drivers available / in progress? Is there another driver I > can hack up to work? > > > -Jared > jmaguire@analogic.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Compiling many files w/ one command Date: Thu, 27 Jul 2000 11:03:39 +0200 From: Gert Stober Organization: MAZeT GmbH Message-ID: <397FFAEB.428CCC06@MAZeT.de> References: <8lnvs8$g1a$1@nnrp1.deja.com> Hello Ryan, please try the following: - define all variables as extern - in one sourcefile write: #define extern /* this defines extern as nothing */ #include "variables.h" #undef extern - in the other sourcefile only write #include "variables.h" Gert. --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: Re: Web Server under VxWorks Date: Thu, 27 Jul 2000 22:53:54 +0200 From: Friedrich Ensslin Organization: ODN Onlinedienst Nordbayern GmbH -- http://www.odn.de/ Message-ID: <3980A162.29461E3E@odn.de> References: <397D1FD2.2D51DB48@yahoo.com> Shelly HUANG wrote: > > Hi, > > Is any web server running under VxWorks? > > thank you for any information. > > best regards, > Shelly have a look at www.3soft.de greetings ... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Booting using TSFS Date: Thu, 27 Jul 2000 11:11:38 -0400 From: Krishna Samavedam Organization: Hughes Network Systems Message-ID: <39805129.A5725FBB@hns.com> References: <8idnb9$bar1@overload.lbl.gov> <397F9E12.B6744846@home.com> - --------------BD24F08C4EB1F5C28EFC2587 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Any non-techinical/personal messages take 'em offline. Posting openings may be OK. But strictly not beyond that. - -krishna Nalit Patel wrote: > Surjit If interested in migrating to US let me know send me a mail with > your credentials npatel@home.com > > sujit v kuruvilla wrote: > > > > This is a multipart message in MIME format > > -------=_MM9611739422765 > > Content-Type: text/plain > > Content-Transfer-Encoding: 7bit > > > > Hi all, > > > > I am working on a MIPS R7000 capella evaluation board. We tried to boot > > up the board using TSFS. Followed whatever is to be done as per tornado > > help. In the virtual console that comes up while launching the target, It > > gives out the Board configuration, and then shows WDBTask exception. I am > > attaching the log file that was generated. > > > > Please do help out and give me inputs as to how to solve this problem. > > > > regards, > > sujit > > > > -------=_MM9611739422765 > > Content-Type: text/plain; name="log1.txt" > > Content-Disposition: inline; filename="log1.txt" > > Content-Transfer-Encoding: 7bit > > > > User Name : suku > > Started : Fri Jun 16 22:08:39 2000 > > Target Server Name : uart@test_1 > > Target Name : uart > > Target Server Options : tgtsvr uart -V -B wdbserial -Bt 10 -Br 2 -Bd C:\Tornado\target\config\tsfs\logs\log1.txt -d COM1 -R C:\Tornado\target\proj\uartboot\default\ -RW -redirectIO -C -c C:\Tornado\target\proj\uartboot\default\vxWorks > > Host : Microsoft Windows NT version 4.0 (Build 1381) > > Timeout value : 10 second(s) > > Request re-send Max : 2 > > > > 0 WDB_TARGET_PING Fri Jun 16 22:08:39 2000 > > > > 1 WDB_TARGET_CONNECT Fri Jun 16 22:08:39 2000 > > 0 Out status: Ok > > WDB_TGT_INFO > > WDB_AGENT_INFO > > agentVersion 2.0 > > mtu 1006 > > mode WDB_MODE_TASK|WDB_MODE_EXTERN > > WDB_RT_INFO > > rtType WDB_RT_VXWORKS > > rtVersion 5.4 > > cpuType 44 > > hasFpp 1 > > hasWriteProtect 0 > > pageSize -1 > > endian 1234 > > bspName NKK RM7000_ND7000LBG-100 > > bootline > > memBase 0x80000000 > > memSize 0x4000000 > > numRegions 0 > > hostPoolBase 0x801c15f8 > > hostPoolSize 0x7c7d41 > > > > 2 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > > In WDB_MEM_REGION > > baseAddr 0x80010000 > > numBytes 314800 > > param 0 > > 0 Out status: Ok > > UINT32 0xffff64bd > > > > 3 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > > In WDB_MEM_REGION > > baseAddr 0x80010000 > > numBytes 373072 > > param 0 > > 0 Out status: Ok > > UINT32 0xffff7f10 > > > > 4 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > > In WDB_MEM_REGION > > baseAddr 0x8006b170 > > numBytes 48824 > > param 0 > > 0 Out status: Ok > > UINT32 0xffffc693 > > > > 5 WDB_MEM_CHECKSUM Fri Jun 16 22:08:41 2000 > > In WDB_MEM_REGION > > baseAddr 0x80077030 > > numBytes 20304 > > param 0 > > 0 Out status: Ok > > UINT32 0xffffb748 > > > > 6 WDB_MEM_WRITE Fri Jun 16 22:08:42 2000 > > In WDB_MEM_XFER > > numBytes 7 > > destination 0x809892e0 > > source 0xe2f6dc > > 0 Out status: Ok > > > > 7 WDB_DIRECT_CALL Fri Jun 16 22:08:42 2000 > > In WDB_CTX_CREATE_DESC > > stackBase 0x0 > > stackSize 0x0 > > entry 0x80027e14 > > args[0] -2137484576 > > args[1] 2 > > args[2] 0 > > args[3] 0 > > args[4] 0 > > args[5] 0 > > args[6] 0 > > args[7] 0 > > args[8] 0 > > args[9] 0 > > name NULL > > priority 0 > > options 0x0 > > redirIn 0 > > redirOut 0 > > redirErr 0 > > 2 Out Error: RPC_CANTDECODERES > > > > 8 WDB_EVENT_GET Fri Jun 16 22:08:59 2000 > > 0 Out status: Ok > > WDB_EVT_DATA > > evtType WDB_EVT_VIO_WRITE > > WDB_MEM_XFER > > numBytes 332 > > destination 0 > > source 0x4427b0 > > > > 9 WDB_EVENT_GET Fri Jun 16 22:08:59 2000 > > 0 Out status: Ok > > WDB_EVT_DATA > > evtType WDB_EVT_EXC > > WDB_EXC_INFO > > numInts 4 > > WDB_CTX > > contextType WDB_CTX_TASK > > contextId 0x83ffa290 > > vec 8 > > pEsf 0x83ff9ed8 > > > > 10 WDB_TARGET_DISCONNECT Fri Jun 16 22:10:30 2000 > > 0 Out status: Ok > > > > -------=_MM9611739422765-- - -- - ---------------------- k r i s h n a s a m a v e d a m 301 548 1103 - --------------BD24F08C4EB1F5C28EFC2587 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Any non-techinical/personal messages take 'em offline. Posting openings may be OK. But strictly not beyond that.

-krishna
 

Nalit Patel wrote:

Surjit If interested in migrating to US let me know send me a mail with
your credentials npatel@home.com

sujit v kuruvilla wrote:
>
> This is a multipart message in MIME format
> -------=_MM9611739422765
> Content-Type: text/plain
> Content-Transfer-Encoding: 7bit
>
> Hi all,
>
>     I am working on a MIPS R7000 capella evaluation board. We tried to boot
> up the board using TSFS. Followed whatever is to be done as per tornado
> help. In the virtual console that comes up while launching the target, It
> gives out the Board configuration, and then shows WDBTask exception. I am
> attaching the log file that was generated.
>
>     Please do help out and give me inputs as to how to solve this problem.
>
> regards,
> sujit
>
> -------=_MM9611739422765
> Content-Type: text/plain; name="log1.txt"
> Content-Disposition: inline; filename="log1.txt"
> Content-Transfer-Encoding: 7bit
>
> User Name              : suku
> Started                : Fri Jun 16 22:08:39 2000
> Target Server Name     : uart@test_1
> Target Name            : uart
> Target Server Options  : tgtsvr uart -V -B wdbserial -Bt 10 - -Br 2 -Bd C:\Tornado\target\config\tsfs\logs\log1.txt -d COM1 -R C:\Tornado\target\proj\uartboot\default\ - -RW -redirectIO -C -c C:\Tornado\target\proj\uartboot\default\vxWorks
> Host                   : Microsoft Windows NT version 4.0 (Build 1381)
> Timeout value          : 10 second(s)
> Request re-send Max    : 2
>
> 0       WDB_TARGET_PING                         Fri Jun 16 22:08:39 2000
>
> 1       WDB_TARGET_CONNECT                      Fri Jun 16 22:08:39 2000
> 0       Out     status:         Ok
>                 WDB_TGT_INFO
>                 WDB_AGENT_INFO
>                 agentVersion    2.0
>                 mtu             1006
>                 mode            WDB_MODE_TASK|WDB_MODE_EXTERN
>                 WDB_RT_INFO
>                 rtType          WDB_RT_VXWORKS
>                 rtVersion       5.4
>                 cpuType         44
>                 hasFpp          1
>                 hasWriteProtect 0
>                 pageSize        -1
>                 endian          1234
>                 bspName         NKK RM7000_ND7000LBG-100
>                 bootline
>                 memBase         0x80000000
>                 memSize         0x4000000
>                 numRegions      0
>                 hostPoolBase    0x801c15f8
>                 hostPoolSize    0x7c7d41
>
> 2       WDB_MEM_CHECKSUM                        Fri Jun 16 22:08:41 2000
>         In      WDB_MEM_REGION
>                 baseAddr        0x80010000
>                 numBytes        314800
>                 param           0
> 0       Out     status:         Ok
>                 UINT32          0xffff64bd
>
> 3       WDB_MEM_CHECKSUM                        Fri Jun 16 22:08:41 2000
>         In      WDB_MEM_REGION
>                 baseAddr        0x80010000
>                 numBytes        373072
>                 param           0
> 0       Out     status:         Ok
>                 UINT32          0xffff7f10
>
> 4       WDB_MEM_CHECKSUM                        Fri Jun 16 22:08:41 2000
>         In      WDB_MEM_REGION
>                 baseAddr        0x8006b170
>                 numBytes        48824
>                 param           0
> 0       Out     status:         Ok
>                 UINT32          0xffffc693
>
> 5       WDB_MEM_CHECKSUM                        Fri Jun 16 22:08:41 2000
>         In      WDB_MEM_REGION
>                 baseAddr        0x80077030
>                 numBytes        20304
>                 param           0
> 0       Out     status:         Ok
>                 UINT32          0xffffb748
>
> 6       WDB_MEM_WRITE                           Fri Jun 16 22:08:42 2000
>         In      WDB_MEM_XFER
>                 numBytes        7
>                 destination     0x809892e0
>                 source          0xe2f6dc
> 0       Out     status:         Ok
>
> 7       WDB_DIRECT_CALL                         Fri Jun 16 22:08:42 2000
>         In      WDB_CTX_CREATE_DESC
>                 stackBase       0x0
>                 stackSize       0x0
>                 entry           0x80027e14
>                 args[0]         -2137484576
>                 args[1]         2
>                 args[2]         0
>                 args[3]         0
>                 args[4]         0
>                 args[5]         0
>                 args[6]         0
>                 args[7]         0
>                 args[8]         0
>                 args[9]         0
>                 name            NULL
>                 priority        0
>                 options         0x0
>                 redirIn         0
>                 redirOut        0
>                 redirErr        0
> 2       Out     Error:          RPC_CANTDECODERES
>
> 8       WDB_EVENT_GET                           Fri Jun 16 22:08:59 2000
> 0       Out     status:         Ok
>                 WDB_EVT_DATA
>                 evtType         WDB_EVT_VIO_WRITE
>                 WDB_MEM_XFER
>                 numBytes        332
>                 destination     0
>                 source          0x4427b0
>
> 9       WDB_EVENT_GET                           Fri Jun 16 22:08:59 2000
> 0       Out     status:         Ok
>                 WDB_EVT_DATA
>                 evtType         WDB_EVT_EXC
>                 WDB_EXC_INFO
>                 numInts         4
>                 WDB_CTX
>                 contextType     WDB_CTX_TASK
>                 contextId       0x83ffa290
>                 vec             8
>                 pEsf            0x83ff9ed8
>
> 10      WDB_TARGET_DISCONNECT                   Fri Jun 16 22:10:30 2000
> 0       Out     status:         Ok
>
> -------=_MM9611739422765--

-- 
- ---------------------- 
k r i s h n a     s a m a v e d a m      
                       301 548 1103
  - --------------BD24F08C4EB1F5C28EFC2587-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxMemProbe Date: Thu, 27 Jul 2000 14:28:39 -0700 From: Ken Norlin Organization: NASA Dryden FRC Message-ID: <3980A987.53CFF775@dfrc.nasa.gov> References: <8F7E849D9olahallstromesavioni@130.100.128.24> "Ola Hallström" wrote: > > Hi, > I wan't to test if a vme-board(with a Tundra Universe 2) is present in a > specific vmebus slot from VxWorks. The board has been HW-initiated to have > an "VME Register Access image" with a baseaddress specific for the slot it > is connected to. > > Is it a good approach to read the address refering to the Tundra > vendor/device id register, mapped in VxWorks vme-space, to see if there is > a valid id? > > Can I use the VxWorks command "vxMemProbe(...)"? > > vxMemProbe is detecting a bus error. Will a bus error be generated when > reading an address refering to a slot where no device is connected? > Hi, vxMemProbe() sounds like exactly what you're after. It will return an ERROR if it doesn't get a reply form the specified address (and a VME bus error will be generated for a VME address if no device exists at that address). Better yet, you have a known ID that should match the returned value. Ken N. - -- Ken Norlin --- Simulation Engineer NASA Dryden Flight Research Center ken.norlin@dfrc.nasa.gov (661) 276-2046 FAX (661) 276-2792 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ppc "machine language" reg use? Date: Tue, 25 Jul 2000 23:52:55 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397E6097.268F3A55@mindspring.com> References: <397E0034.42F0E87@mindspring.com> Ah, a keeper of the secrets! thanks... > ABI: System V Application Binary Interface: PowerPC Processor Supplement, 1995 > You mean there is a book I havent downloaded/printed... I am running out of book shelf space... > (This one is a little strange, the pages are from last to first). - -g- cute... guess that is what happens when sun tries to work with ibm/apple... think its an "endian" issue? -g- thanks for the rest to. > > EABI: PowerPC Processor Embedded Application Binary Interface, v1.0 > > > These are both from the Motorola PPC823 Engineer's Toolbox: > > > Also, these may be of interest, particularly the PEM. > CPUs_group> > > Regards, > Pete > -- > +-----------------------------------------------------------------+ > | Pete Kockritz mailto:pkockritz@home.com | > | This space for rent. | > +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: mvme2700 -update rtc date? Date: Thu, 27 Jul 2000 22:32:07 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <3980F0A7.6DF33C1C@mindspring.com> I need a routine that takes date/time in "struct timespec" format (what sntpcTimeGet() returns) and updates the RTC date (mk48t59/559). Anyone got this coded already? The format it is stored in is a little "funny", and I would rather not get bogged down on this. tnx. --------------------------- Newsgroups: comp.os.vxworks Subject: Anyone using the LSI sym53C895 SCSI controller on VxWorks Date: Thu, 27 Jul 2000 22:53:02 GMT From: syee@connex.com Organization: Deja.com - Before you buy. Message-ID: <8lqegd$ap0$1@nnrp1.deja.com> Has anyone using the LSI sym53C895 SCSI controller got it to work on VxWorks? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Anyone using the LSI sym53C895 SCSI controller on VxWorks Date: Thu, 27 Jul 2000 22:51:20 GMT From: syee@connex.com Organization: Deja.com - Before you buy. Message-ID: <8lqed7$ao3$1@nnrp1.deja.com> Has anyone using the LSI sym53C895 SCSI controller got it to work on VxWorks? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Query about comp.os.vxworks Date: Fri, 28 Jul 2000 04:23:53 GMT From: G.N.Raj Organization: Wipro Technologies Message-ID: <8lr1sn$oan$1@nnrp1.deja.com> References: <8lpg95$1lh$1@overload.lbl.gov> <39804CC0.EE3E7061@hns.com> Reply-To: g.nagaraj@wipro.com Suppose I want to setup a new NNTP server in my organization, then how do I access the comp.os.vxworks newsgroup? How does my newsgroup server (NNTP) get synchronized with the usenet group? How do I configure the link between my NNTP server and the usenet newsgroup? I am in sync with my sysadm. Nagaraj In article <39804CC0.EE3E7061@hns.com>, krishna balusu wrote: > That would be the nntp server for ur organization. ask your sysadm for the > server name. > > - -- G. Nagaraj Senior Software Engineer Wipro Technologies - EnThink IDC. 37, Castle Street, Ashok Nagar, Bangalore - 560 025, India Phone: +91-80-536-7297 Ext: 315 Fax: +91-80-557-6032 www.EnThink.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: workspace? Date: Fri, 28 Jul 2000 11:15:24 +0530 From: Vijay Kumar Peshkar Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <39811DF4.C0B9FA3@cisco.com> References: <8lpg95$1lh$1@overload.lbl.gov> <39804CC0.EE3E7061@hns.com> <8lr1sn$oan$1@nnrp1.deja.com> Reply-To: vpeshkar@cisco.com This is a multi-part message in MIME format. - --------------BE6B2FDEFFCE5DA4C2FF742D Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Friends, How do you 'define' workspace?. Have been creating/using it for some time...but am unable to give a textbook defination to it.... I mean I am lacking in words ;) Thanks and Regards, Vijay - --------------BE6B2FDEFFCE5DA4C2FF742D Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Vijay Kumar Peshkar Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Kumar.P;Vijay tel;work:080-5530035/36 Extn:3464 x-mozilla-html:FALSE org:Wipro Technologies, Bangalore.;Wishing you blue skies, a couple of daisies, some fresh air and a smile! adr:;;;;;; version:2.1 email;internet:vpeshkar@cisco.com note;quoted-printable: o o o o o o o . . . _____________________________________________=0D=0A o _____ || Vijay Kumar .P =20 |=0D=0A .] [__n_n_| DD[ =3D=3D=3D=3D_____ | vpeshkar@cisco.com, vijay_peshkar@hotmail.com | =0D=0A >(________|__|_[_________]_ |___________________________________________|=0D=0A _/oo OOOOO oo` ooo ooo 'o!o!o =20 o!o!o` fn:Vijay Kumar.P end:vcard - --------------BE6B2FDEFFCE5DA4C2FF742D-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Raw IP Date: Thu, 27 Jul 2000 17:35:58 -0700 From: "Edt" Organization: Posted via Supernews, http://www.supernews.com Message-ID: References: <8la2v8$k4v$1@news.umbc.edu> You should be able to open a raw socket and receive all the IP packets for a particular address family. The etherHookAdd function mentioned in another message will get ethernet formatted packets from a particular interface, but not from any interface like a raw socket will, and the ethernet header will still be in the packet data. - -Ed cedilnik andrej wrote in message <8la2v8$k4v$1@news.umbc.edu>... >Hello! > >Any idea on how to get network packets from hardware layer? > > Andy > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxMemProbe Date: Thu, 27 Jul 2000 12:09:30 -0700 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <398088EA.E68485F1@delphi.com> References: <8F7E849D9olahallstromesavioni@130.100.128.24> In general, yes. However, I have seen some CPU boards that did not report bus errors when they were encountered. "Ola Hallström" wrote: > <> > vxMemProbe is detecting a bus error. Will a bus error be generated when > reading an address refering to a slot where no device is connected? > > Thanks /Ola --------------------------- Newsgroups: comp.os.vxworks Subject: RTOS comparisons Date: Fri, 28 Jul 2000 06:55:01 GMT From: "casey ballentine" Organization: @Home Network Message-ID: <97ag5.27747$P4.149987@news1.rdc1.ne.home.com> I'm looking for a good article that compares several RTOSs on a few 'standard' benchmarks. Specifically, what I'd like to see are articles that compare cpu useage, footprint, RAM useage, and task switch time all as a function of number of tasks running on a system. Processor platform isn't necessarily what I'm interested in. I've read the article on RTOS use in automotive applications in the latest Embedded Systems Programming, but it's too narrow in scope, focusing on one industry and not including the major RTOSs like VxWorks, pSOS and QNX/NTO for instance. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks & ARM Dev Board Date: Fri, 28 Jul 2000 03:02:31 -0400 From: Mohit Sindhwani Organization: Posted via Supernews, http://www.supernews.com Message-ID: <39A46963@MailAndNews.com> Sender: Mohit Sindhwani Hi... I posted a mesg earlier but it;s not showing up in the list.. I must have mungled it up. Anyway, I'm just putting this out again... with apologies if you receive multiple copies (sorry!) - -- Original Message -- HI. This is the first time that I'm being exposed to Tornado and the ARM Development Board. The manual for the Dev Board talks about the ARM Debugger for Windows - I have no such thing [I think]. But, what I do have is the Tornado kit. Now, I'm confused about how to get started!! We are using Angel for helping in the debugging & development. I'm a total newbie, so I'd appreciate any links, pointers, help, etc... and forgive me if the questions are simplistic to the point of being stupid. Thanks in Advance, Mohit. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using Single Serial Line with VxWorks Date: 28 Jul 2000 08:08:47 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lrf2f$qdk$1@news1.xs4all.nl> References: <8lr0gu$nh6$1@nnrp1.deja.com> g.nagaraj@wipro.com wrote: : I am using EV386EX BSP with Tornado II. I have my custom board. I have : successfully ported the EV386EX BSP to my custom board. I have modified : the "i8250Sio.c" for my board and I can see the printing on my serial : console also. But when I configure the WDB on the serial using : WDB_SERIAL, I can't see any message on the serial console. As per the : manual, I am supposed to see the WDB Banner, after which I can : disconnect the serial console and attach the WDB target server and work : with the host shell. But, the WDB banner does not appear on the : console. But the strange thing is that, if I connect the WDB target : server, it successfully connects and I can use the host shell without : any problem. Did you select WDB_BANNER? Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: 28 Jul 2000 08:17:41 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lrfj5$qdk$2@news1.xs4all.nl> References: <39A46963@MailAndNews.com> Mohit Sindhwani wrote: : HI. This is the first time that I'm being exposed to Tornado and the ARM : Development Board. : The manual for the Dev Board talks about the ARM Debugger for Windows - I : have : no such thing [I think]. But, what I do have is the Tornado kit. Now, I'm : confused about how to get started!! The best way to start getting used to Tornado is to take the Tornado getting started guide (If you don't have a copy of it: it is also available on WindSurf (http://www.wrs.com/windsurf), but you have to subcribe to it). It this guide there is a complete example using the simulator to build an application and to debug this application. This would be a good starting point. The environment for another target (in your case the ARM) would be almost identical to teh simulator, as it uses the same tools. The ARM Debugger that is mentioned in the book is probably smething that can be purcased/picked-up somewhere, but as you have Tornado it is not needed, I think. Did you order a BSP for your specific target (the ARM development board) from Windriver or the supplier of the board? When you want to run VxWorks on this board you do need this. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using Single Serial Line with VxWorks Date: Fri, 28 Jul 2000 13:51:38 +0530 From: Vijay Kumar Peshkar Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <39814292.BF56D781@cisco.com> References: <8lr0gu$nh6$1@nnrp1.deja.com> Reply-To: vpeshkar@cisco.com This is a multi-part message in MIME format. - --------------5AD83238074790B936E04111 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Nagaraj, How many serial ports do U have?. Excuse me, but I have not recently been in touch with vxworks/psos. So I can only guess....the scenario must be: the only availible link is now being used by the WDB. The message( I remember seeing a small wdb console message when I did port ev386 BSP to my customised 486 board) must have been scrolled by the later app tasks etc., Try going to infinite loop in the root task....think that may spare the overwrite/clear... good luck, Vijay g.nagaraj@wipro.com wrote: > Hello everyone, > > I am using EV386EX BSP with Tornado II. I have my custom board. I have > successfully ported the EV386EX BSP to my custom board. I have modified > the "i8250Sio.c" for my board and I can see the printing on my serial > console also. But when I configure the WDB on the serial using > WDB_SERIAL, I can't see any message on the serial console. As per the > manual, I am supposed to see the WDB Banner, after which I can > disconnect the serial console and attach the WDB target server and work > with the host shell. But, the WDB banner does not appear on the > console. But the strange thing is that, if I connect the WDB target > server, it successfully connects and I can use the host shell without > any problem. > > My issue in this scenario is that, I have a driver for my proprietary > hardware which gets initialized during boot. I have actually built the > BSP with my driver. If I use the serial console alone using CONSOLE_TTY > definition, I can see all messages on the console. But, once I switch > to WDB_SERIAL, I can't see anything. > > I saw lots of FAQs on similar issues. But I could not relate anything > to my problem. If anyone has any info relating to my problem, please > reply. > > I tried to solve the problem using the old style serial driver (TTY CO > Driver). I modified the "i8250Serial.c" and I could get the console > messages. When I switched to WDB on this, I could even see all the > messages including the WDB Banner. But the problem using this type of > connection is that the Tornado/VxWorks does not support system > debugging with this WDB connection type. It only supports Task > debugging. > > Any help in terms of pointers, mails would be greatly appreciated. > > Thanks & Regards, > > Nagaraj G > > Sent via Deja.com http://www.deja.com/ > Before you buy. - --------------5AD83238074790B936E04111 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Vijay Kumar Peshkar Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Kumar.P;Vijay tel;work:080-5530035/36 Extn:3464 x-mozilla-html:FALSE org:Wipro Technologies, K2, Bangalore.;Wishing you blue skies, a couple of daisies, some fresh air and a smile! adr:;;;;;; version:2.1 email;internet:vpeshkar@cisco.com title:Senior Software Engineer note;quoted-printable: o o o o o o o . . . _____________________________________________=0D=0A o _____ || Vijay Kumar .P =20 |=0D=0A .] [__n_n_| DD[ =3D=3D=3D=3D_____ | vpeshkar@cisco.com, vijay_peshkar@hotmail.com | =0D=0A >(________|__|_[_________]_ |___________________________________________|=0D=0A _/oo OOOOO oo` ooo ooo 'o!o!o =20 o!o!o` fn:Vijay Kumar.P end:vcard - --------------5AD83238074790B936E04111-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RTOS comparisons Date: Fri, 28 Jul 2000 13:57:42 +0530 From: Vijay Kumar Peshkar Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <398143FD.8EB30020@cisco.com> References: <97ag5.27747$P4.149987@news1.rdc1.ne.home.com> Reply-To: vpeshkar@cisco.com This is a multi-part message in MIME format. - --------------B7D3A1617C2D7828B343F85C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Browse thru the Embedded systems Magazine site. I remember seeing comparisons published periodically. See archives...That's, I think the best way out. WRS site may be another choice... regds, Vijay casey ballentine wrote: > I'm looking for a good article that compares several RTOSs on a few > 'standard' benchmarks. Specifically, what I'd like to see are articles that > compare cpu useage, footprint, RAM useage, and task switch time all as a > function of number of tasks running on a system. Processor platform isn't > necessarily what I'm interested in. I've read the article on RTOS use in > automotive applications in the latest Embedded Systems Programming, but it's > too narrow in scope, focusing on one industry and not including the major > RTOSs like VxWorks, pSOS and QNX/NTO for instance. - --------------B7D3A1617C2D7828B343F85C Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Vijay Kumar Peshkar Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Kumar.P;Vijay tel;work:080-5530035/36 Extn:3464 x-mozilla-html:FALSE org:Wipro Technologies, K2, Bangalore.;Wishing you blue skies, a couple of daisies, some fresh air and a smile! adr:;;;;;; version:2.1 email;internet:vpeshkar@cisco.com title:Senior Software Engineer note;quoted-printable: o o o o o o o . . . _____________________________________________=0D=0A o _____ || Vijay Kumar .P =20 |=0D=0A .] [__n_n_| DD[ =3D=3D=3D=3D_____ | vpeshkar@cisco.com, vijay_peshkar@hotmail.com | =0D=0A >(________|__|_[_________]_ |___________________________________________|=0D=0A _/oo OOOOO oo` ooo ooo 'o!o!o =20 o!o!o` fn:Vijay Kumar.P end:vcard - --------------B7D3A1617C2D7828B343F85C-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: Fri, 28 Jul 2000 14:03:23 +0530 From: Vijay Kumar Peshkar Organization: Ye 'Ol Disorganized NNTPCache groupie Message-ID: <39814553.DEE23B1A@cisco.com> References: <39A46963@MailAndNews.com> Reply-To: vpeshkar@cisco.com This is a multi-part message in MIME format. - --------------6212F18A2557DC1153781488 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Mohit, Browse thru the Tornado users's guide...chapter" Getting started". Tornado is a cross development environment. That means, while it facilitates U to use the real time features, it does not mess up with the front end part, and hence runs on a non RTOS....Windows for your case. U have to install the tornado...and start playing it with it... to begin with, just as another application...say netscape. The debugger is explained in a later chapter. cheers, Vijay Mohit Sindhwani wrote: > Hi... I posted a mesg earlier but it;s not showing up in the list.. I must > have mungled it up. Anyway, I'm just putting this out again... with > apologies > if you receive multiple copies (sorry!) > > -- Original Message -- > HI. This is the first time that I'm being exposed to Tornado and the ARM > Development Board. > > The manual for the Dev Board talks about the ARM Debugger for Windows - I > have > no such thing [I think]. But, what I do have is the Tornado kit. Now, I'm > confused about how to get started!! > > We are using Angel for helping in the debugging & development. > > I'm a total newbie, so I'd appreciate any links, pointers, help, etc... and > forgive me if the questions are simplistic to the point of being stupid. > > Thanks in Advance, > Mohit. - --------------6212F18A2557DC1153781488 Content-Type: text/x-vcard; charset=us-ascii; name="vpeshkar.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Vijay Kumar Peshkar Content-Disposition: attachment; filename="vpeshkar.vcf" begin:vcard n:Kumar.P;Vijay tel;work:080-5530035/36 Extn:3464 x-mozilla-html:FALSE org:Wipro Technologies, K2, Bangalore.;Wishing you blue skies, a couple of daisies, some fresh air and a smile! adr:;;;;;; version:2.1 email;internet:vpeshkar@cisco.com title:Senior Software Engineer note;quoted-printable: o o o o o o o . . . _____________________________________________=0D=0A o _____ || Vijay Kumar .P =20 |=0D=0A .] [__n_n_| DD[ =3D=3D=3D=3D_____ | vpeshkar@cisco.com, vijay_peshkar@hotmail.com | =0D=0A >(________|__|_[_________]_ |___________________________________________|=0D=0A _/oo OOOOO oo` ooo ooo 'o!o!o =20 o!o!o` fn:Vijay Kumar.P end:vcard - --------------6212F18A2557DC1153781488-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to add command line into com1 ? Date: Fri, 28 Jul 2000 05:15:57 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <8l1833$k40@netnews.hinet.net> Golly, this old chestnut. The 'command line' is a C interpreter. Just write your functions, load your object and then run your functions. Paul In article <8l1833$k40@netnews.hinet.net>, "june" wrote: > >Hi, >Now I find a file ,usrlib.c , to add some commands. >But I can't find function array to add. >Is there other file that must be used ? >Thanks > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using Single Serial Line with VxWorks Date: Fri, 28 Jul 2000 08:31:52 GMT From: G.N.Raj Organization: Wipro Technologies Message-ID: <8lrgdp$2kc$1@nnrp1.deja.com> References: <8lr0gu$nh6$1@nnrp1.deja.com> <8lrf2f$qdk$1@news1.xs4all.nl> Yes. I selected WDB_BANNER Nagaraj In article <8lrf2f$qdk$1@news1.xs4all.nl>, johan.borkhuis@peektraffic.nl wrote: > g.nagaraj@wipro.com wrote: > : I am using EV386EX BSP with Tornado II. I have my custom board. I have > : successfully ported the EV386EX BSP to my custom board. I have modified > : the "i8250Sio.c" for my board and I can see the printing on my serial > : console also. But when I configure the WDB on the serial using > : WDB_SERIAL, I can't see any message on the serial console. As per the > : manual, I am supposed to see the WDB Banner, after which I can > : disconnect the serial console and attach the WDB target server and work > : with the host shell. But, the WDB banner does not appear on the > : console. But the strange thing is that, if I connect the WDB target > : server, it successfully connects and I can use the host shell without > : any problem. > > Did you select WDB_BANNER? > > Groeten, > Johan > > -- > o o o o o o o . . . ___________________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | > >(________|__|_[_________]_|________________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === > - -- G. Nagaraj Senior Software Engineer Wipro Technologies, Bangalore, India Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Using Single Serial Line with VxWorks Date: Fri, 28 Jul 2000 04:00:31 GMT From: g.nagaraj@wipro.com Organization: Wipro Technologies Message-ID: <8lr0gu$nh6$1@nnrp1.deja.com> Reply-To: g.nagaraj@wipro.com Hello everyone, I am using EV386EX BSP with Tornado II. I have my custom board. I have successfully ported the EV386EX BSP to my custom board. I have modified the "i8250Sio.c" for my board and I can see the printing on my serial console also. But when I configure the WDB on the serial using WDB_SERIAL, I can't see any message on the serial console. As per the manual, I am supposed to see the WDB Banner, after which I can disconnect the serial console and attach the WDB target server and work with the host shell. But, the WDB banner does not appear on the console. But the strange thing is that, if I connect the WDB target server, it successfully connects and I can use the host shell without any problem. My issue in this scenario is that, I have a driver for my proprietary hardware which gets initialized during boot. I have actually built the BSP with my driver. If I use the serial console alone using CONSOLE_TTY definition, I can see all messages on the console. But, once I switch to WDB_SERIAL, I can't see anything. I saw lots of FAQs on similar issues. But I could not relate anything to my problem. If anyone has any info relating to my problem, please reply. I tried to solve the problem using the old style serial driver (TTY CO Driver). I modified the "i8250Serial.c" and I could get the console messages. When I switched to WDB on this, I could even see all the messages including the WDB Banner. But the problem using this type of connection is that the Tornado/VxWorks does not support system debugging with this WDB connection type. It only supports Task debugging. Any help in terms of pointers, mails would be greatly appreciated. Thanks & Regards, Nagaraj G Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Getting rid of "transfer timed out" message Date: Wed, 26 Jul 2000 10:43:49 -0500 From: "Tim Berger" Organization: Motorola CIG Message-ID: <8ln0vm$qg28@nntp.cig.mot.com> Hello, I'm using the bootp+tftp method in the Vxworks boot loader to boot an MCPN750 board up. I have an MCP750 board running vxworks with dhcp & tftp servers. These are the original servers supplied with vxworks. The MCPN750 board boots up fine, but I get an error message a few seconds later in the console of the MCP750 saying: Transfer Timed Out. TFTP server: could not send client file "vxWorks.NSP" ...even though the image has been sent successfully (obvious, since the MCPN750 boots vxworks without any problem). As far as I know, the vxworks boot loader is using "tftpXfer" to transfer the image. Is there something that the client fails to do to "close" the tftp connection gracefully? By the way, I don't have this problem if I use the windows NT tftp client. Any help appreciated... thanks! t.berger_AT_motorola.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Date: Thu, 27 Jul 2000 18:12:26 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <3980B3CA.4CCDF634@mindspring.com> References: <397F5375.4D3A441A@mindspring.com> Still looking for the answer here... guys? --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Fri Jul 28 12:02:19 2000 From: "Shahan Avedian" Date: Fri Jul 28 12:02:22 PDT 2000 Subject: VxWorks/C++ Opportunity Dear VxWorks professionals: Please review the job description below if you know of anyone please have them contact me. It is a consulting assignment in Princeton NJ. They are looking to immediately hire the right candidate. The company is an innovative electronics-based research organization. We will pay a substantial referral fee for anyone you recommend that we place. With that said, you do look exactly right for the position. Yours truly, Shahan |-------------------------------------------------------------------------| |Position Description/Responsibilities: | | | |-------------------------------------------------------------------------| |Software development with working knowledge of C++, VXworks, and other | |softwares. Analytical, creative and abstract thinker. Will be doing | |work in the DTV and Entertainment Group. Need to code utilizing C++, | |create a design based on function specs and then code and test defined | |software modules and storage structures by developing the software | |architecture in the program modules. | | | | | |-------------------------------------------------------------------------| Shahan Avedian Sr. Technical Recruiter TIS Worldwide 732-494-2210 ext. 12119 savedian@tisny.com http://www.tisworldwide.com e-business architects building the e-conomy From vxwexplo-errs@csg.lbl.gov Sat Jul 29 04:03:26 2000 From: Vxworks Exploder Date: Sat Jul 29 04:03:29 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 29 04:03:22 PDT 2000 Subject: Re: VxWorks ROM image and application. Subject: Re: Anyone using the LSI sym53C895 SCSI controller on VxWorks Subject: Re: Using Single Serial Line with VxWorks Subject: Re: MPC8260 UPM Initialisation Subject: Re: who owns this semaphore? Subject: Video Player Subject: Video Player Subject: FAQs are where? Subject: Re: Tornado Shell: d = 123 Subject: VxWorks/C++ Opportunity Subject: Re: Tornado Shell: d = 123 Subject: Re: Video Player Subject: lkup "^[a-z][a-z] " Subject: VxWorks Scheduler Subject: Re: VxWorks Scheduler Subject: IDE driver errors Subject: MPC8260 UPM Initialisation Subject: Modem question Subject: Re: Adaptec AIC-7892A Ultra160/m Driver Subject: Error booting vxworks interrupt: dc0 - fatal system error Subject: Re: Tornado Shell: d = 123 Subject: Video Player Subject: Re: lkup "^[a-z][a-z] " Subject: bc635PCI and Linux Subject: Re: Tornado Shell: d = 123 Subject: Re: VxWorks & ARM Dev Board Subject: Re: Tornado Shell: d = 123 Subject: m Enter Ctrl+D in Tornado Shell Subject: Re: Tornado Shell: d = 123 Subject: Re: FAQs are where? Subject: Re: Tornado Shell: d = 123 Subject: Re: lkup "^[a-z][a-z] " Subject: Re: FAQs are where? Subject: Re: Tornado Shell: d = 123 Subject: Re: who owns this semaphore? Subject: Re: VxWorks & ARM Dev Board Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Subject: Re: Max Resources ? Subject: Re: Tornado 2 vs Tornado 1 Subject: Scheduling Subject: Re: MPC750 and MPC8260 Subject: Winbond W83C554F chip set ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks ROM image and application. Date: Fri, 28 Jul 2000 09:57:57 -0400 From: viv Organization: Lucent Technologies Message-ID: <39819165.77A03D37@lucent.com> References: <397E43D4.3445446D@lucent.com> <8lpdif$g2h$1@nnrp1.deja.com> How do you decide on the size of your mymodule.o? From where do you start the VxWorks image? do you use fpromread and fpromwrite to the flash? Thanks, viv ddinker@my-deja.com wrote: > In article <397E43D4.3445446D@lucent.com>, > viv wrote: > > > ------------------------------Low ROM address > > VxWorks image > > ............................. > > > > mymodule.o > > > > ------------------------------High ROM address > > We have some sort of inverse of it. > > > ------------------------------Low ROM address > > mymodule.o > > ............................. > > > > VxWorks image > > > > ------------------------------High ROM address > > Our "mymodule.o" is loaded first, does some diag and then hands control > to VxWorks. > > Dinker > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Anyone using the LSI sym53C895 SCSI controller on VxWorks Date: Fri, 28 Jul 2000 13:13:35 GMT From: Thomas Rahn Organization: MacroSystem Message-ID: <8ls0tr$div$1@nnrp1.deja.com> References: <8lqegd$ap0$1@nnrp1.deja.com> Reply-To: trahn@macrosystem.de Hi, we have finally succeeded to get the LSI 875 working. We use both custom hardware (no PC BIOS) and a LSI 8751 SP PCI card. VxWorks is based on a x86 bsp and we are happy to use a heavily debugged ncr810 library. The (debugged) ncr810 library should support the 895 as well, at least for basic modes. Thomas Rahn Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using Single Serial Line with VxWorks Date: Fri, 28 Jul 2000 14:18:43 +0530 From: "G.N.Raj" Organization: Wipro Message-ID: References: <8lr0gu$nh6$1@nnrp1.deja.com> <39814292.BF56D781@cisco.com> Reply-To: "G.N.Raj" I think that was clear. The subject itself states that. Anyway, thanks for the reply. I have only one serial port. There is nothing scrolling as I am using the terminal application with logging option where I will know any printing has been done or not. Actually nothing gets printed. The TTY_CO type works fine. Nagaraj "Vijay Kumar Peshkar" wrote in message news:39814292.BF56D781@cisco.com... > Nagaraj, > > How many serial ports do U have?. > Excuse me, but I have not recently been in touch > with vxworks/psos. So I can only guess....the scenario > must be: the only availible link is now being used by the WDB. > The message( I remember seeing a small wdb console message > when I did port ev386 BSP to my customised 486 board) > must have been scrolled by the later app tasks etc., > Try going to infinite loop in the root task....think that > may spare the overwrite/clear... > > good luck, > Vijay > > g.nagaraj@wipro.com wrote: > > > Hello everyone, > > > > I am using EV386EX BSP with Tornado II. I have my custom board. I have > > successfully ported the EV386EX BSP to my custom board. I have modified > > the "i8250Sio.c" for my board and I can see the printing on my serial > > console also. But when I configure the WDB on the serial using > > WDB_SERIAL, I can't see any message on the serial console. As per the > > manual, I am supposed to see the WDB Banner, after which I can > > disconnect the serial console and attach the WDB target server and work > > with the host shell. But, the WDB banner does not appear on the > > console. But the strange thing is that, if I connect the WDB target > > server, it successfully connects and I can use the host shell without > > any problem. > > > > My issue in this scenario is that, I have a driver for my proprietary > > hardware which gets initialized during boot. I have actually built the > > BSP with my driver. If I use the serial console alone using CONSOLE_TTY > > definition, I can see all messages on the console. But, once I switch > > to WDB_SERIAL, I can't see anything. > > > > I saw lots of FAQs on similar issues. But I could not relate anything > > to my problem. If anyone has any info relating to my problem, please > > reply. > > > > I tried to solve the problem using the old style serial driver (TTY CO > > Driver). I modified the "i8250Serial.c" and I could get the console > > messages. When I switched to WDB on this, I could even see all the > > messages including the WDB Banner. But the problem using this type of > > connection is that the Tornado/VxWorks does not support system > > debugging with this WDB connection type. It only supports Task > > debugging. > > > > Any help in terms of pointers, mails would be greatly appreciated. > > > > Thanks & Regards, > > > > Nagaraj G > > > > Sent via Deja.com http://www.deja.com/ > > Before you buy. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC8260 UPM Initialisation Date: Fri, 28 Jul 2000 16:59:28 GMT From: gold@sdsu.edu Organization: Deja.com - Before you buy. Message-ID: <8lse5d$ogr$1@nnrp1.deja.com> References: <3981ab00.340821029@192.168.1.11> In article <3981ab00.340821029@192.168.1.11>, John Keightley wrote: > I'm looking for help setting up UPM state machines on an EST8260 > system - specifically UPMC and CS8 to access our own memory mapped > subsystem. We can initialise this OK via VisionICE probe - now need to > do this as part of the VxWorks initialisation sequence. > > I've tried setting this up in sysHwInit() - writing to BR8,OR8 but > then get a crash when attempting to write the first RAM location > (after loading MDR) after setting the OP code bit in MCMR > > Does anyone have sample code or experience with this? 1. Did you check the other bits in MCMR, such as BSEL and RFEN ? 2. Are you sure the memory area you set BR8 to isn't already covered by another (previously initialized) chip select and that is the one that's firing ? 3. Are you in C or assembly ? Are you sure you're writing a 32-bit value to MCMR ? I believe the 860FADS BSP has assembly code to init the UPM on the 860 if you can get a hold of that... hth, Dan Gold gold@sdsu.edu Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Fri, 28 Jul 2000 15:04:51 +0200 From: "Michael Lawnick" Organization: Deja.com - Before you buy Message-ID: <200007281304.PAA11215@net.softec.de> References: <8lahk4$5f4$1@nnrp2.deja.com> Hi tracy, > I am trying to find the tasks that is holding > this semaphore. Is there a way to find out? > > -> semShow(0x3ef4ca0,1) No, as nobody can 'own' a semaphore. A mutex only is owned. HTH MfG Lawnick, SOFTEC GmbH ============================================== SOFTEC GmbH Tel +49-731-96600-0 Promenade 17 Fax +49-731-96600-23 D-89073 Ulm Michael Lawnick Germany lawnick@softec.de ============================================== Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Video Player Date: Fri, 28 Jul 2000 09:52:56 -0400 From: Chris Jacobsen Message-ID: <39819037.F53E9150@viagate.com> Can anyone point me in the direction of an MPEG-2 player for VxWorks. If possible, have variable rate transport stream input and drive a Sigma Designs Real Magic 8400 decoder or similar device. Chris Jacobsen --------------------------- Newsgroups: comp.os.vxworks Subject: Video Player Date: Fri, 28 Jul 2000 14:23:30 -0400 From: Chris Jacobsen Message-ID: <3981CFA2.719B8F66@viagate.com> Can anyone point me in the direction of an MPEG-2 player for VxWorks. If possible, have variable rate transport stream input and drive a Sigma Designs Real Magic 8400 decoder or similar device. Chris Jacobsen --------------------------- Newsgroups: comp.os.vxworks Subject: FAQs are where? Date: Fri, 28 Jul 2000 14:35:52 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ls5o8$hjs$1@nnrp1.deja.com> I'm new and clueless. In addition to the Tornado help itself, I've found: http://www.deja.com/home_ps.shtml http://www.xs4all.nl/~borkhuis/vxworks/vxfaq.html http://www.xs4all.nl/~borkhuis/vxworks/oldvxfaq.html Anything else I should be reading? Thanks in advance. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 18:57:38 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lsl2v$ubu$1@nnrp1.deja.com> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> In article <3981AF4D.216FABDD@swri.edu>, Greg Willden wrote: > Change the name of your variable and everything should work fine. Yes, thank you - but I don't want to change the name of my variable. I want to know how I can name and use a variable of my choice. Is this not possible? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks/C++ Opportunity Date: Fri, 28 Jul 2000 14:58:26 -0400 From: "Shahan Avedian" Organization: Lawrence Berkeley National Laboratory Message-ID: <8lsm4a$ou8$1@overload.lbl.gov> Dear VxWorks professionals: Please review the job description below if you know of anyone please have them contact me. It is a consulting assignment in Princeton NJ. They are looking to immediately hire the right candidate. The company is an innovative electronics-based research organization. We will pay a substantial referral fee for anyone you recommend that we place. With that said, you do look exactly right for the position. Yours truly, Shahan |-------------------------------------------------------------------------| |Position Description/Responsibilities: | | | |-------------------------------------------------------------------------| |Software development with working knowledge of C++, VXworks, and other | |softwares. Analytical, creative and abstract thinker. Will be doing | |work in the DTV and Entertainment Group. Need to code utilizing C++, | |create a design based on function specs and then code and test defined | |software modules and storage structures by developing the software | |architecture in the program modules. | | | | | |-------------------------------------------------------------------------| Shahan Avedian Sr. Technical Recruiter TIS Worldwide 732-494-2210 ext. 12119 savedian@tisny.com http://www.tisworldwide.com e-business architects building the e-conomy --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 14:26:46 -0500 From: "Matthew Majka" Organization: NASA, Kennedy Space Center Message-ID: <8lsmd4$msm@niven.ksc.nasa.gov> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> wrote in message news:8lsl2v$ubu$1@nnrp1.deja.com... > In article <3981AF4D.216FABDD@swri.edu>, > Greg Willden wrote: > > Change the name of your variable and everything should work fine. > > Yes, thank you - but I don't want to change the name of my variable. > > I want to know how I can name and use a variable of my choice. Is this > not possible? Can't have two symbols with the same name unless you don't need the 'd' command... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Video Player Date: Fri, 28 Jul 2000 17:01:39 +0200 From: Emmanuel Organization: SEPRO Robotique Message-ID: <3981A053.54B8D42@sepro-robotique.com> References: <39819037.F53E9150@viagate.com> Chris Jacobsen wrote: > Can anyone point me in the direction of an MPEG-2 player for VxWorks. Hi, You should look at various projects for Linux... It's probably possible to find source code and other usefull documentations that help to code a MPEG 2 decoder. > and drive a Sigma Designs Real Magic 8400 decoder or similar device. You should contact the hardware manufacturer for such specific software. Regards Emmanuel --------------------------- Newsgroups: comp.os.vxworks Subject: lkup "^[a-z][a-z] " Date: Fri, 28 Jul 2000 19:20:26 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lsmdk$vf3$1@nnrp1.deja.com> The Tornado 2.0 Shell command script: jj = 123 lkup "^jj " naturally enough reports something like: jj 0x... bss () But: lkup "^[a-z][a-z] " spits out screens and screens and screens of symbols - only one of which, namely "jj", has a two-letter name. I conclude I've utterly misunderstood how "regular expressions" work with `lkup` ... where can I go to be enlighted? Is there a way to ask `lkup` to list all the two-letter symbols it knows? Here, by entering all 26 of: aa bb ... zz I can determine that the Shell knows none of these two-letter names except for "tt". (That is a command known to the Shell but not to `lkup`.) Thanks in advance. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Scheduler Date: Fri, 28 Jul 2000 20:42:56 +0530 From: Amarnath Shanbhag Organization: Lucent Technologies Message-ID: <3981A2F8.888C1197@lucent.com> Hello, When does the scheduler in VxWorks get invoked ? Does it get invoked after every clock tock or does it get invoked on every system call (such as taskSpawn) ? Amarnath --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Scheduler Date: 28 Jul 2000 20:20:53 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lspv5$p2t$1@news1.xs4all.nl> References: <3981A2F8.888C1197@lucent.com> Amarnath Shanbhag wrote: : Hello, : When does the scheduler in VxWorks get invoked ? : Does it get invoked after every clock tock or does it : get invoked on every system call (such as taskSpawn) ? Both. The scheduler gets invoked at each system event, that can be a clock tick, an interrupt, or (some of the) system calls. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: IDE driver errors Date: Fri, 28 Jul 2000 15:46:12 GMT From: bjd2k@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ls9s2$l1u$1@nnrp1.deja.com> In Tornado 2.0, when I enable IDE hard drive support, Tornado announces that there are these errors and won't continue: Undefined parameters: IDE_INT_VEC IDE_INT_LVL IDE_CONFIG I can find any info in the distribution on how to resolve this problem. Does anyone here know how to correct this? I'm trying to mount an IDE hard drive using a PCI bus IDE controller card in a MIPS target. Thanks in advance, Brian Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: MPC8260 UPM Initialisation Date: Fri, 28 Jul 2000 16:06:18 GMT From: john_k@3dm.com (John Keightley) Organization: 3DM Devices Inc Message-ID: <3981ab00.340821029@192.168.1.11> Reply-To: John Keightley I'm looking for help setting up UPM state machines on an EST8260 system - specifically UPMC and CS8 to access our own memory mapped subsystem. We can initialise this OK via VisionICE probe - now need to do this as part of the VxWorks initialisation sequence. I've tried setting this up in sysHwInit() - writing to BR8,OR8 but then get a crash when attempting to write the first RAM location (after loading MDR) after setting the OP code bit in MCMR Does anyone have sample code or experience with this? thx --------------------------- Newsgroups: comp.os.vxworks Subject: Modem question Date: Fri, 28 Jul 2000 13:32:20 -0700 From: RAVINDER KUMAR Organization: Sun Microsystems Message-ID: <3981EDD4.83CC59FE@Sun.COM> We are using ns16550Sio driver and by default modem is not enabled. Using IOCTL commands I tried enabling the modem. I set the baud rate to 9600 and set the hardware options to "HUPCL | CREAD | CS8". Then I used IOCTL command to call its open routine. Here is the piece of the code that does all this. ioctl(fd, SIO_BAUD_SET, baudRate); ioctl(fd, SIO_HW_OPTS_SET, HUPCL | CREAD | CS8 ); ioctl(fd, SIO_OPEN, 0); Now when I write a "ATZ" command I do receive "OK" back from the modem but after that I am not able to dial the number. I tried another command 'ATA' that causes the modem to go off the hook and generate a carrier tone. But I don't hear any tone. I tried all the above mentioned commands through a Solaris machine and everything works fine. Am I setting the correct hardware options? Any help or suggestion is appreciated. Regards Ravinder Regards Ravinder ___________________________________________________________________ Ravinder Kumar Sun MicroSystems mailing : 901 San Antonio Road Voice: (408) 774-8353 MS USUN04-204 Ravinder.Kumar@Eng.Sun.Com Palo Alto, CA 94303-4900 ___________________________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Adaptec AIC-7892A Ultra160/m Driver Date: Fri, 28 Jul 2000 18:10:41 GMT From: jmaguire@analogic.com (Jared Maguire) Message-ID: <3981cc51.778155538@bilbo> References: <39805793.682733078@bilbo> <8lqak0$7ct4@nntp.cig.mot.com> Well, in that case, does anyone know of any Ultra160 SCSI cards that ARE supported? I'd rather not have to kludge a fast SCSI driver together, if I can avoid it. - -J On Thu, 27 Jul 2000 16:47:52 -0500, "Derek Opitz" wrote: >We had to write our own. Not sure if it was an AIC-7892A though. > >Jared Maguire wrote in message >news:39805793.682733078@bilbo... >> I very much need to make an AIC-7892A Ultra160/m PCI SCSI Controller >> work with VxWorks running on a more or less plain vanilla Intel >> platform. I've been searching, but have yet to find any drivers for >> this card. >> >> Are there drivers available / in progress? Is there another driver I >> can hack up to work? >> >> >> -Jared >> jmaguire@analogic.com > > --------------------------- Newsgroups: comp.os.vxworks Subject: Error booting vxworks interrupt: dc0 - fatal system error Date: Fri, 28 Jul 2000 10:26:01 -0500 From: "Ken Parrish" Organization: WebUseNet Corp http://www.usenetserver.com - Home of the fastest NNTP servers on the Net. Message-ID: <1Ahg5.5484$%W6.202974@news-east.usenetserver.com> I am using a MVME2431 board and have compiled the bootrom_uncmp image and have pflashed it into the ROM. When I change the jumper to boot the image I get the [vxWorks Boot]: prompt. When I type @ and return I get this message: interrupt: dc0 - fatal system error attached TCP/IP interface to dc0. attaching network interface lo0 ... done. Loading ... 0x1fbb260 (tNetTask): dc0 - restarting device. Can anyone tell me what this means? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 11:05:34 -0500 From: Greg Willden Organization: Southwest Research Institute Message-ID: <3981AF4D.216FABDD@swri.edu> References: <8ls5l2$ham$1@nnrp1.deja.com> d is a command used to display the contents of memory try something like d 0x4100 On the Synergy boards this is the address for the Shared Memory anchor. Change the name of your variable and everything should work fine. p_lavarre@my-deja.com wrote: > How do I create & use a variable named "d"? > > The Tornado 2.0 Shell here chokes over: > d = 123 > and complains "syntax error". I imagine this is because there is a "d" > command? The help tells me I can use: > @d > to call a routine named "d", so I tried: > @d = 123 > The Shell accepts this ... but then behaves bizarrely. Every time I > enter this command, I'm informed again: > new symbol "@d" added to symbol table > I'm saying the symbol is NOT added once - rather, repeatedly, > indefinitely I see: > new symbol "@d" added to symbol table > This suggests the symbol cannot be found? Certainly, if I try: > j = @d > the Shell chokes and complains: > undefined symbol: @d > > How do I create & use a variable named "d"? > > Thanks in advance. > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Video Player Date: Fri, 28 Jul 2000 14:25:32 -0400 From: Chris Jacobsen Message-ID: <3981D01C.A1C1353E@viagate.com> Can anyone point me in the direction of an MPEG-2 player for VxWorks. If possible, have variable rate transport stream input and drive a Sigma Designs Real Magic 8400 decoder or similar device. Chris Jacobsen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: lkup "^[a-z][a-z] " Date: Fri, 28 Jul 2000 23:31:34 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lt54m$asp$1@nnrp1.deja.com> References: <8lsmdk$vf3$1@nnrp1.deja.com> <3982025F.89CE3FE0@swri.edu> Ahhh. Thanks everyone for helping. I think I now have the answer to this one ... A blank in the string argument to `lkup` confuses it. Likewise "\t". So avoid those. If you want to list all the two letter symbols known to `lkup`, you can indeed enter: lkup "^[a-z][a-z]$" Notice the trailing "$" - that is key. Here this command shows me only the symbol "fd" ... unless I go create more two letter symbols. Trying: lkup "^[a-z][a-z][a-z]$" shows me all of div, tan, pow, cin, end, exp, sin, cos, log, abs. > In article <3982025F.89CE3FE0@swri.edu>, > Greg Willden wrote: > From the VxWorks documentation: > void lkup > ( > char * substr /* substring to match */ > ) > This command lists all symbols in the system symbol table whose names > contain the string substr. Ahhh. Duelling documentation. The "all who contain string" claim above doesn't explain why: k = 123 lkup "k" finds MORE symbols than does: k = 123 lkup "^k" The "all who contain string" claim above is indeed what I see if I type lkup and a blank and then I strike Ctrl+D or Ctrl+W. But I find this text confusing, because I know of more elaborate, almost contradictory, text. You see, I began by clicking on Help.ManualsContents. That launched a browser to show me: /Tornado/docs/books.html This leads me to: Tornado User's Guide (Windows Version) /Tornado/docs/tornado/winguide/index.html which leads me to: 6.2.3 Invoking Built-In Shell Routines /Tornado/docs/tornado/winguide/c-shell2.html#99723 Eventually this all led me to the claim: The lkup() command takes a regular expression as its argument, and looks up all symbols containing strings that match. In the simplest case, you can specify a substring to see any symbols containing that string ... Regular-expression searches of the symbol table can be as simple or elaborate as required. For example, the following simple regular expression displays the names of three internal VxWorks semaphore functions: I'm not sure now how I got so confused. I guess it just threw me that blanks and tabs in the argument string freak out the `lkup` command? Thanks again. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks,comp.realtime Subject: bc635PCI and Linux Date: Fri, 28 Jul 2000 11:22:55 -0600 From: Chris Webster Organization: NCAR Research Aviation Facility Message-ID: <3981C16F.D33435D3@raf.atd.ucar.edu> Anyone have a driver/program for the Datum bc635PCI IRIG card running under Linux? - --Chris --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 20:53:57 -0400 From: Joe Durusau Message-ID: <39822B25.5B6ED9AC@bellsouth.net> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> <8lss52$li33@nntp.cig.mot.com> <8lt1jn$8b4$1@nnrp1.deja.com> Hmmmm... I've never heard of such a thing, but you could always write a program that declared global symbols of your choice and load the thing. That should do it. Don't know of a way to get at it directly from the shell, though. Speaking only for myself, Joe Durusau p_lavarre@my-deja.com wrote: > > Isn't there some syntax - some call - I can use to just add a symbol of > my choice to the `lkup` table? > > > In article <8lss52$li33@nntp.cig.mot.com>, > > "Derek Opitz" wrote: > > you can declare any variable you want at the shell, > > as long as it is not used already. d() is a function > > that dumps memory. So, therefore, you > > can't use d as a variable name. HELLO!!!!!!! > > What can I say. Certainly I don't mean to be slow & obscure. > > There are TWO symbol tables involved here - yes? One table visible to > "lkup". And a separate, second table hidden away in the Shell, but > visible to Ctrl+D and Ctrl+W. For example, if I try: > lkup "d" > lkup "tt" > I get no hits - even though "d" and "tt" are Shell commands. Part of > how I know they are Shell commands is that typing them alone on a > command line, then striking the spacebar, then pressing Ctrl+D gives me > a synopsis of their purpose & arguments. > > I understand that because "d" is a Shell command, an entry like: > d = 123 > is considered to mean something like: > d(= 123); > and therefore reported as a: > syntax error > > What's got me is that the commands: > @d = 123 > @d > sometimes work and sometimes don't. > > If I have a C program loaded that declared: > extern int d = 456; > then these commands do just plain work. > > Otherwise, entering these two commands always causes the Shell to > report: > new symbol "@d" added to symbol table > > If I review the effect of these commands via: > lkup "@d" > then I can confirm that I'm creating a whole series of symbols with the > same name but pointing to different locations. Naturally enough, > creating an "@d" symbol make syntaxes like: > @@d = 0x12345678 > @@d > d &@@d > work, keeping in mind that each of these only refer to the very last > definition created for the symbol "@d". > > What I don't get is why it isn't easy to make > @d = 123 > @d > work as if I had loaded a C program to declare: > extern int d = 456; > > Isn't there some syntax - some call - I can use to just add a symbol of > my choice to the `lkup` table? > > Thanks again in advance. Pat LaVarre p.lavarre@ieee.org > http://members.aol.com/ppaatt > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: Fri, 28 Jul 2000 18:44:02 -0700 From: Mohit Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <03db9226.c358038d@usw-ex0105-038.remarq.com> Thanks Johan and Vijay: Your comments have already given me *some* direction. But, now I need to ask a stupid question... forgive me! [*] What is a BSP? [BSP=Board Support Package, I know] What does it have? When we ordered the board, shouldn't it have already come with it? A check reveals that we only have a disk with sample source on it...! [*] What will talk to the board? Is there a utility inside the BSP that will let me talk to the Angel debugger on the board?? Thanks a lot, Mohit. - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 20:47:10 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lsrgb$3gg$1@nnrp1.deja.com> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> <8lsmd4$msm@niven.ksc.nasa.gov> I can phrase my question more intelligently now ... How can I add symbols of my choice to the table that `lkup` shows? I was surprised to find that a form like: @d = 890 does not work. Is there no quick simple syntax for this? Some routine I can call? Do I really have to edit, compile, and download an app, just to add a symbol of my choice? > In article <8lsmd4$msm@niven.ksc.nasa.gov>, > "Matthew Majka" wrote: > Can't have two symbols with the same name unless you don't > need the 'd' command... Life is not quite this simple? There are symbols known to `lkup` and symbols not known to `lkup`. Granted, I can't have two `lkup` symbols with the same name but different values. But yes I can have `lkup` symbols that are the same as Shell command names. I just tried downloading an app that included the C lines: extern int d = 123; extern int tt = 456; Downloading such an app makes the @ notation work great. @d @d = 567 printf("%d\n", @d); Thanks again in advance. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: m Enter Ctrl+D in Tornado Shell Date: Fri, 28 Jul 2000 19:24:02 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lsmkb$vh7$1@nnrp1.deja.com> Here, if at the Shell prompt I strike the three keys: m Enter Ctrl+D I get an Unhandled EXCEPTION_ACCESS_VIOLATION in TCL.DLL ... Is this interesting? Does it suggest I should muck around differently with how I have std i/o or the target agent configured? Or is this just life as normal with a Windows Millenium GUI? (I have already dug out that I was supposed to press . Enter in place of Ctrl+D.) Thanks in advance. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 15:59:23 -0500 From: "Derek Opitz" Organization: Motorola CIG Message-ID: <8lss52$li33@nntp.cig.mot.com> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> you can declare any variable you want at the shell, as long as it is not used already. d() is a function that dumps memory. So, therefore, you can't use d as a variable name. HELLO!!!!!!! wrote in message news:8lsl2v$ubu$1@nnrp1.deja.com... > In article <3981AF4D.216FABDD@swri.edu>, > Greg Willden wrote: > > Change the name of your variable and everything should work fine. > > Yes, thank you - but I don't want to change the name of my variable. > > I want to know how I can name and use a variable of my choice. Is this > not possible? > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FAQs are where? Date: Fri, 28 Jul 2000 15:57:35 -0500 From: "Derek Opitz" Organization: Motorola CIG Message-ID: <8lss1n$li32@nntp.cig.mot.com> References: <8ls5o8$hjs$1@nnrp1.deja.com> There are some FAQ pages at www.wrs.com under support, and then windsurf. wrote in message news:8ls5o8$hjs$1@nnrp1.deja.com... > I'm new and clueless. > > In addition to the Tornado help itself, I've found: > http://www.deja.com/home_ps.shtml > http://www.xs4all.nl/~borkhuis/vxworks/vxfaq.html > http://www.xs4all.nl/~borkhuis/vxworks/oldvxfaq.html > Anything else I should be reading? > > Thanks in advance. Pat LaVarre p.lavarre@ieee.org > http://members.aol.com/ppaatt > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Sat, 29 Jul 2000 03:52:59 GMT From: pkockritz@home.com (Pete Kockritz) Organization: @Home Network Message-ID: References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> <8lsmd4$msm@niven.ksc.nasa.gov> <8lsrgb$3gg$1@nnrp1.deja.com> In article <8lsrgb$3gg$1@nnrp1.deja.com>, p_lavarre@my-deja.com wrote: > I can phrase my question more intelligently now ... > > How can I add symbols of my choice to the table that `lkup` shows? I > was surprised to find that a form like: > @d = 890 > does not work. Is there no quick simple syntax for this? Some routine > I can call? Do I really have to edit, compile, and download an app, > just to add a symbol of my choice? Try this command: symAdd() ­ create and add a symbol to a symbol table, including a group number See the entry in the vxWorks manuals for symLib for more info. Regards, Pete - -- +-----------------------------------------------------------------+ | Pete Kockritz mailto:pkockritz@home.com | | This space for rent. | +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: lkup "^[a-z][a-z] " Date: Fri, 28 Jul 2000 16:59:59 -0500 From: Greg Willden Organization: Southwest Research Institute Message-ID: <3982025F.89CE3FE0@swri.edu> References: <8lsmdk$vf3$1@nnrp1.deja.com> From the VxWorks documentation: void lkup ( char * substr /* substring to match */ ) This command lists all symbols in the system symbol table whose names contain the string substr. My comments: If you had written a function called feedData() and downloaded it to the target and then typed lkup("ee") one of the symbols returned would be feedData() or if you typed lkup("dd") you would also get feedData in the list. > Is there a way to ask > `lkup` to list all the two-letter symbols it knows? No. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FAQs are where? Date: Fri, 28 Jul 2000 22:14:37 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lt0ka$7i2$1@nnrp1.deja.com> References: <8ls5o8$hjs$1@nnrp1.deja.com> <8lss1n$li32@nntp.cig.mot.com> > In article <8lss1n$li32@nntp.cig.mot.com>, > "Derek Opitz" wrote: > > There are some FAQ pages at www.wrs.com under support, and then > windsurf. Ahhh yes thank you. Notably: http://www.wrs.com/csdocs/sortedfaqs/index.shtml and: http://www.wrs.com/csdocs/sortedfaqs/summariesTools/htmlfiles/shell.html explaining such things don't work as: printf("\x41\n"); Anyone got more FAQs? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: Fri, 28 Jul 2000 22:31:19 GMT From: p_lavarre@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lt1jn$8b4$1@nnrp1.deja.com> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> <8lss52$li33@nntp.cig.mot.com> Isn't there some syntax - some call - I can use to just add a symbol of my choice to the `lkup` table? > In article <8lss52$li33@nntp.cig.mot.com>, > "Derek Opitz" wrote: > you can declare any variable you want at the shell, > as long as it is not used already. d() is a function > that dumps memory. So, therefore, you > can't use d as a variable name. HELLO!!!!!!! What can I say. Certainly I don't mean to be slow & obscure. There are TWO symbol tables involved here - yes? One table visible to "lkup". And a separate, second table hidden away in the Shell, but visible to Ctrl+D and Ctrl+W. For example, if I try: lkup "d" lkup "tt" I get no hits - even though "d" and "tt" are Shell commands. Part of how I know they are Shell commands is that typing them alone on a command line, then striking the spacebar, then pressing Ctrl+D gives me a synopsis of their purpose & arguments. I understand that because "d" is a Shell command, an entry like: d = 123 is considered to mean something like: d(= 123); and therefore reported as a: syntax error What's got me is that the commands: @d = 123 @d sometimes work and sometimes don't. If I have a C program loaded that declared: extern int d = 456; then these commands do just plain work. Otherwise, entering these two commands always causes the Shell to report: new symbol "@d" added to symbol table If I review the effect of these commands via: lkup "@d" then I can confirm that I'm creating a whole series of symbols with the same name but pointing to different locations. Naturally enough, creating an "@d" symbol make syntaxes like: @@d = 0x12345678 @@d d &@@d work, keeping in mind that each of these only refer to the very last definition created for the symbol "@d". What I don't get is why it isn't easy to make @d = 123 @d work as if I had loaded a C program to declare: extern int d = 456; Isn't there some syntax - some call - I can use to just add a symbol of my choice to the `lkup` table? Thanks again in advance. Pat LaVarre p.lavarre@ieee.org http://members.aol.com/ppaatt Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: who owns this semaphore? Date: Sat, 29 Jul 2000 07:13:29 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lu06o$se1$1@nnrp1.deja.com> References: <8lahk4$5f4$1@nnrp2.deja.com> <200007281304.PAA11215@net.softec.de> In article <200007281304.PAA11215@net.softec.de>, "Michael Lawnick" wrote: > Hi tracy, > > I am trying to find the tasks that is holding > > this semaphore. Is there a way to find out? > > > > -> semShow(0x3ef4ca0,1) > No, as nobody can 'own' a semaphore. A mutex only is owned. Only half true... the owner field of a binary semaphore is, at least for some architectures, set to the task ID of the task that took it. John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: Sat, 29 Jul 2000 07:23:17 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lu0p4$tpb$1@nnrp1.deja.com> References: <03db9226.c358038d@usw-ex0105-038.remarq.com> Hi, > [*] What is a BSP? [BSP=Board Support Package, I know] What does > it have? When we ordered the board, shouldn't it have already > come with it? A check reveals that we only have a disk with > sample source on it...! It may come with the board, though probably you'd need to ask for it when you ordered the board, but it could also come with Tornado. You don't say which board it is that you are using so we can't answer the question exactly. What I can do, is tell you the BSPs that Wind River has for ARM (most of these should have been on the CD - you'll need the key to see them though; assabet is the exception in that it is new and comes separately still I think): EBSA-285 (StrongARM 110) Brutus (StrongARM 1100) Assabet (StrongARM 1110) PID-7T (ARM 7TDMI) CMA210 (ARM 710A) CMA220 (ARM 810) There are also some BSPs for ARM's new Integrator platform available from ARM, and there may well be other BSPs available for other platforms. Each BSP will come with an image that needs to be somehow programmed into either flash or PROM (depends on the board). How this occurs also depends on the board: Integrator has a monitor that can read hex files over a serial link, the older boards tend to use the ARM debugger to load and run FMU (Flash Management Utility), or require an external programmer. > [*] What will talk to the board? Is there a utility inside the > BSP that will let me talk to the Angel debugger on the board?? Again, depends on the board. Tell us which board you have and we can explain how to talk to it... probably... if we've heard of it... and I'm going to guess that somebody out here has heard of it... Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Date: Sat, 29 Jul 2000 07:44:25 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lu20p$ugq$1@nnrp1.deja.com> References: <397F5375.4D3A441A@mindspring.com> <3980B3CA.4CCDF634@mindspring.com> In article <3980B3CA.4CCDF634@mindspring.com>, Al Johnston wrote: > Still looking for the answer here... guys? Some people can be so demanding... anybody'd think they were paying for the *FREE* help and advice offered in this forum... Anyway, to answer your question, all tasks are initialised with an MSR value that will enable interrupts in the CPU. So, I'd say the first time we start executing in a task context interrupts will be enabled (i.e. when usrRoot() starts executing since that is the entry function for the first task in the system). If you wish to attach a dummy handler, try doing it in sysHwInit() or anytime after the vector table is initialised in usrInit(), but before the call to kernelInit(). Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max Resources ? Date: Sat, 29 Jul 2000 08:04:01 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lu35g$v6m$1@nnrp1.deja.com> References: <3976A86D.93FE3DC0@lucent.com> <8l6dui$7qb$1@nnrp1.deja.com> <3976FA96.9B5E5A6A@lucent.com> <8lctbd$kgj$1@nnrp1.deja.com> <397E96E7.DAF99237@lucent.com> Hello, You are still confusing network cluster management with system memory. They are not related! You email lists three things: watchdogs, file descriptors (sockets are file descriptors), and route tables. None of these come from the network clusters. Only the buffers for the actual data packets (and associated headers) come from the clusters. If you are expecting a lot of connections then you are right to increase the size of these buffers (a connection does after all need space for the packets, both incoming and outgoing). As for the other things, NUM_FILES will control the maximum number of file descriptors that can be open at the same time (i.e. the maximum number of sockets you can have open), the watchdogs are allocated dynamically from the heap, and I suspect that the routing tables are too, though I haven't checked that one... Regards, John... In article <397E96E7.DAF99237@lucent.com>, Madhu wrote: > Hi. > We have a networking product that allows a number of users to connect. > Each Client uses PPP to connect. This would cause allocation of watch > dog timers per session, a route add to handle the new client, new > sockets etc. We found here > that by increasing the number of system clusters in netbuflib.h. we were > able to > connect more clients. I think NUM_FILES will not help to connect > more than that allocated by the system clusters. > > Madhu > > > > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 2 vs Tornado 1 Date: Sat, 29 Jul 2000 08:08:14 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lu3dd$vfi$1@nnrp1.deja.com> References: <397EAD6D.B7785775@francimel.com> Hello, That really depends on what you mean by "as is". If you mean, can s/he take the source and compile it to run under Tornado 2.0, I'd be pretty confident in saying yes. If you mean, can you send him/her a binary in the form of a .o then I'd say probably (but I don't know for certain - try it!). HTH, John... In article <397EAD6D.B7785775@francimel.com>, jph wrote: > I have developed an application (a task) for SH4/VxWorks with Tornado 1. > Can it be used "as is" by a customer using Tornado 2 ? > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Scheduling Date: Sat, 29 Jul 2000 10:33:24 +0200 From: Heribert Volkhausen Organization: T-Online Message-ID: <398296D4.11F07E41@t-online.de> Hi, I need to know in which cases a rescheduling is done. I checked the documentation, but I couldn't find useful hints. Regards Heribert Volkhausen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MPC750 and MPC8260 Date: Wed, 26 Jul 2000 11:16:49 -0700 From: Eric Monsler Organization: The ISP formerly known as Best Message-ID: <397F2B11.F7DA1AC7@radixtek.com> References: <397CF308.4741D12C@radixtek.com> <8lkgm0$s67$1@nnrp1.deja.com> Dan, Thank you for your helpful post. I've got a few follow-up questions, that I hope you can help me with. I'm not a HW designer, I'm a SW guy, so please forgive any obvious blunders. (I forwarded your previous response to my HW counterpart) 1) A 750 will not be able to access the 8260-local bus, correct? So, if we wanted to prevent the 8260-CPM from ever becoming bus master on the 60x bus, we could only use the 8260 DMA channels between the peripherals and the 8260 dual-port memory. Does this sound correct? 2) If we ran as described above, does this remove the 8260 bus anomalies that you describe? We had read about various anomalies with two 60x masters and the PCI bridge, but had hoped to eliminate them by preventing the 8260 from becoming bus master. FYI, and for others, the Scout reference design board is sold by EST, now part of WindRiver, and the marketing webpage claims that there is a VxWorks BSP for it. And finally, I see you're posting from SDSU. Are any of your board designs, PLD's, and/or BSP's published or otherwise in the public domain (e.g., funded by gov't) so that I and the board HW designer could take a look? Thanks, Eric --------------------------- Newsgroups: comp.os.vxworks Subject: Winbond W83C554F chip set Date: Sat, 29 Jul 2000 02:34:01 GMT From: "Howard Lung" Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: Does anybody knows if vxWorks has Winbond W83C554F or W83C553F chip set device driver support for PCI-to-IDE hard disk? I saw there is a file called ataDrv.c under drv/hdisk directory, but I am not sure if that is the one I need. Thanks for help! H. Lung --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 30 04:03:09 2000 From: Vxworks Exploder Date: Sun Jul 30 04:03:11 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 30 04:03:05 PDT 2000 Subject: APOLOGY Subject: looking up static symbols Subject: Re: Scheduling Subject: Re: Tornado Shell: d = 123 Subject: Re: looking up static symbols Subject: Re: Scheduling Subject: Re: VxWorks Context Switch Time Subject: Re: taskCreat() vs. taskInit() Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Subject: Re: C++: taskSpawn(member functions) Subject: building GCC for powerpc-wrs-vxworks on a solaris Subject: Re: C++ classes. Subject: Re: C++: taskSpawn(member functions) Subject: Re: Using C++ & gcc 2.95.2 Subject: supports chinese/japanses/koera (CJK) charset? Subject: 2700(750) ppc - where MSR(EE) initially set? Subject: TrueFFS mountFTL Subject: Compiling many files w/ one command Subject: Re: Web Server under VxWorks Subject: Re: VxWorks & ARM Dev Board Subject: sntpclib, localtime() and TIMEZONE Subject: Re: mv2700 - unknown symbol "__eabi" Subject: help about MPC860 SCC HDLC mode ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: APOLOGY Date: Wed, 26 Jul 2000 18:49:56 GMT From: complaints@barcelona.com Organization: Clientes_Teleline Message-ID: To all the users of this newsgroup: Please let us *APOLOGISE* for a Spam sent this afternoon regarding e-mail accounts and connection at barcelona.com. It was sent by an unauthorized person of our organization. We agree that 100% OT and was sent out despite our telling people that Usenet was to be left strictly alone. We have found out who sent it and we've told appropriate LARTings are about to be issued. Again, please accept our apologies and assurance that this is being dealt with. Please, for complaints reach complaints@barcelona.com Eduard Sole Creative Director of Barcelona City Portal --------------------------- Newsgroups: comp.os.vxworks Subject: looking up static symbols Date: Sat, 29 Jul 2000 14:40:01 +0200 From: Thomas Scholz Organization: Ericsson Eurolab Germany Message-ID: <3982D0A1.5A4777C8@eed.ericsson.se> I'm having problems to find static function names in the systems symbol table. I'm using symFindByValue(sysSymTbl,...) in Tornado 1.0.1. Are static symbols defined elsewhere or is there another trick ? All vxworks functions (eg "l" or "tt") display them as expected. Regards, Thomas. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Scheduling Date: Sat, 29 Jul 2000 14:59:33 GMT From: "Dan Perlberger" Organization: Road Runner - EC Message-ID: References: <398296D4.11F07E41@t-online.de> Well let's see... Here are a few times rescheduling occurs... Hopefully I won't forget too many of them: 1) If a task with a higher priority then the running task becomes ready to run (i.e not blocked on any resources, not delayed, not suspended, etc.), then that task will start running, and the current running task will go into the Ready state. 2) If Kernel Time Sliciing (Round Robin Scheduling for tasks with the same priority) is turned on, then if a task with the same priority as the running task becomes ready to run, and the number of ticks specified by the kernelTimeSlice call have expired since the running task has been running, then the other task will get to run. 3) If the running task goes into the pending state, becomes delayed, or becomes suspended, or just basically is not ready to run any more, then the next lower (or same) priority task ready to run will start running. Otherwise, the "Idle" task will take over (meaning nothing is really running). 4) If the running task is deleted (by itself), then the next lower (or same) priority task which is ready to run will start running. Otherwise, the "Idle" task will take over (meaning nothing is really running). Those are all the ones I can think of at this time... There might be other cases of rescheduling. Dan "Heribert Volkhausen" wrote in message news:398296D4.11F07E41@t-online.de... > Hi, > > I need to know in which cases a rescheduling is done. I checked the > documentation, but I couldn't find useful hints. > > Regards > Heribert Volkhausen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Shell: d = 123 Date: 29 Jul 2000 17:35:38 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8lv4la$o7j$1@news1.xs4all.nl> References: <8ls5l2$ham$1@nnrp1.deja.com> <3981AF4D.216FABDD@swri.edu> <8lsl2v$ubu$1@nnrp1.deja.com> <8lsmd4$msm@niven.ksc.nasa.gov> <8lsrgb$3gg$1@nnrp1.deja.com> p_lavarre@my-deja.com wrote: : I can phrase my question more intelligently now ... : How can I add symbols of my choice to the table that `lkup` shows? I : was surprised to find that a form like: : @d = 890 : does not work. Is there no quick simple syntax for this? Some routine : I can call? Do I really have to edit, compile, and download an app, : just to add a symbol of my choice? As long as the symbol is not defined you can just define it in the shell. If the symbol is known the shell wants to use this symbol, with the results you showed before. : Life is not quite this simple? There are symbols known to `lkup` and : symbols not known to `lkup`. Granted, I can't have two `lkup` symbols : with the same name but different values. But yes I can have `lkup` : symbols that are the same as Shell command names. : I just tried downloading an app that included the C lines: : extern int d = 123; : extern int tt = 456; : Downloading such an app makes the @ notation work great. : @d : @d = 567 : printf("%d\n", @d); Yes, that is right, but now you redefined the symbol d in your application. You cannot use the function d anymore from within the shell. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: looking up static symbols Date: Sat, 29 Jul 2000 17:33:06 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lv4gi$kdb$1@nnrp1.deja.com> References: <3982D0A1.5A4777C8@eed.ericsson.se> Hi Thomas, By default, VxWorks only loads the global symbols into memory. If you are using the target based tools (i.e. target shell, target loader and target symbol table) you can pass an option via the boot line to get the local symbols (static symbols) loaded from the boot image (I think it is 0x4, but check by typing a '?' at the boot loader prompt). For modules you load dynamically there is also a option (see the reference manual entry for ld). Now, if you're using the host based tools, WindShell and target server, then you need to look in the target server options to get the boot image local symbols loaded, and, as with the target tools, use an option with ld to get them for dynamically loaded modules. HTH, John... In article <3982D0A1.5A4777C8@eed.ericsson.se>, Thomas Scholz wrote: > > I'm having problems to find static function names in the systems symbol > table. I'm using symFindByValue(sysSymTbl,...) in Tornado 1.0.1. > Are static symbols defined elsewhere or is there another trick ? All vxworks > functions (eg "l" or "tt") display them as expected. > > Regards, Thomas. > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Scheduling Date: Sat, 29 Jul 2000 17:39:22 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lv4sa$kgn$1@nnrp1.deja.com> References: <398296D4.11F07E41@t-online.de> Hi, Scheduling will occur after any system event. A system event is something that could have changed the state of any task in the system. Remember that VxWorks is a priority based pre-emptive OS; the highest priority task ready to run will always run. This means that anything that could make a task ready to run, or block it, has to be followed by a call to the scheduler. Some examples: - - Interrupts, including the system clock tick - - Semaphore operations (give, take, flush) - - Message queue operations (send, receive) - - Task operations (delay, delete, spawn, priority change etc) - - Exceptions (can cause a task to be suspended) HTH, John... In article <398296D4.11F07E41@t-online.de>, Heribert Volkhausen wrote: > Hi, > > I need to know in which cases a rescheduling is done. I checked the > documentation, but I couldn't find useful hints. > > Regards > Heribert Volkhausen > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Context Switch Time Date: Wed, 26 Jul 2000 07:25:40 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lm3pi$2tk$1@nnrp1.deja.com> References: <8lksg9$564$1@overload.lbl.gov> Hi, This must be an FAQ by now, but in case it's not, the answer is that you should speak to your local Wind River sales person or FAE. They will have a datasheet with the context times and many other statistics for the CPU, or CPUs you are interested in. Regards, John... In article <8lksg9$564$1@overload.lbl.gov>, "Chhabra, Vikram" wrote: > Does anyone know what the VxWorks average and Worst > Case context switch times are on Motorola Processors. Or any place where I > can get this information. > > Thanks > Vikram Chhabra > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: taskCreat() vs. taskInit() Date: Wed, 26 Jul 2000 07:30:24 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lm42g$30b$1@nnrp1.deja.com> References: <397DE681.3DB0AD2F@sympatico.ca> Hi Glen, In article <397DE681.3DB0AD2F@sympatico.ca>, Glen Lumsden wrote: > What is the difference between taskCreat() and taskInit()? One difference is that taskCreat() is an unpublished internal routine; taskInit() is published. Of course, that's not what you wanted to know... taskCreat() is in fact the first half of what is under taskSpawn() and will manage the allocation of stack memory etc for the task; taskInit() expects its caller (taskCreat() is one such caller) to have allocated the stack and pass in the pointer. > Do the parameters for taskCreat() match those of taskSpawn()? No they don't. Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 2700(750) ppc - where MSR(EE) initially set? Date: Sat, 29 Jul 2000 15:13:30 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <39832CDA.90C5466@mindspring.com> References: <397F5375.4D3A441A@mindspring.com> <3980B3CA.4CCDF634@mindspring.com> <8lu20p$ugq$1@nnrp1.deja.com> > > Still looking for the answer here... guys? > > Some people can be so demanding... anybody'd think they were paying for > the *FREE* help and advice offered in this forum... I did not mean to offend, just to communicate my continued interest in this issue. If I crossed netiquette, apologies. > Anyway, to answer your question, all tasks are initialised with an MSR > value that will enable interrupts in the CPU. So, I'd say the first time > we start executing in a task context interrupts will be enabled (i.e. > when usrRoot() starts executing since that is the entry function for the > first task in the system). Thank you for the help (in spite of my bad manners). This sure clears up why I could not find an explicit enable. thanks!. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++: taskSpawn(member functions) Date: Sat, 29 Jul 2000 20:09:36 GMT From: Xenos Organization: Hades Message-ID: References: <39529A59.F8C47A33@yahoo.de> Use a C function (or better yet, a static member if the linkage is the same) as a wrapper, sending it the "this" variable as its sole parameter example: class foo { public: int task_func(); }; extern "C" int wrapper(foo* func) { return func->task_func(); } foo task; taskSpawn(NULL, 100, VX_FP_TASK, 20 * 1024, wrapper, (int) &task,0,0,0,0,0,0,0,0,0); OR, if you can, this is "cleaner": class foo { public: foo(); ~virtual foo(); static int wrapper(foo* func) {return func->task_func();} spawn(char* name, int priority = 100, int options = VX_FPTASK, stack = 20 * 1024) { taskSpawn(NULL, 100, VX_FP_TASK, 20 * 1024, wrapper, (int) this,0,0,0,0,0,0,0,0,0);} private: virtual int task_func() = 0; }; // now just inherit from foo and override task_func with your task code. Note that I'm writing this on-the-fly so there may be lexical errors! How this helps. I haven't done this yet in VxWorks, but I do it a lot in Visual C++. On Fri, 23 Jun 2000 00:59:37 +0200, Matthias wrote: >I'm currently putting some old code into C++ and this problem came up. >The class Cxx is used as communication node for multiple applications. >Each application registers at this class and defines a number of message >IDs for incoming messages. > >Then, there is a central receiver (CRX) for all incoming traffic, looks >on the message ID and puts the messages in VxWorks message queues. > >Now I want that every application when registering spawns a task to >receive messages from CRX and the code for this task is a member >function of the class. I'm getting curious if there is any way to make a >taskSpawn() and give a reference to a member function. > >Currently I'm using a static function and pass a pointer to the instance >of the class, this works fine - but does anybody know something >different ? > >Thanks, Matthias --------------------------- Newsgroups: comp.os.vxworks Subject: building GCC for powerpc-wrs-vxworks on a solaris Date: Sat, 29 Jul 2000 20:21:42 GMT From: Xenos Organization: Hades Message-ID: Can anyone please help me? I've trying to build GCC on a solaris 2.5.1 for powerpc-wrs-vxworks. It keeps getting parse error building iovfprintf.c on all lines containing a reference to va_arg. It includes and can find stdarg.h but it seems that va_arg is not getting defined (but va_list is). Has anyone seem this?? I was able to successful build the very first time, but I accidentally cobbered it while trying to tar it up for safe keeping!! I have yet to be able to duplicate my initial success! I able able to build the bin utils and a native compile, no problem. Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++ classes. Date: Sat, 29 Jul 2000 20:30:47 GMT From: Xenos Organization: Hades Message-ID: References: <395D3D37.6910771F@lucent.com> You are misunderstanding how to use a class. You cannot call a method directly (except a static one). You have to create an object of the class. If you do this statically (at file scope), you must run the code through the muncher to build the constructor table. If your constructor stategy is not automatic then you must call the function (see your manual) that will call the initializer. Remember, you can have many objects of a class and they will all have their own set of variables (your i & j from your example). A class is not an object. Calling its methods directly is the same as using an uninitialized pointer! On Fri, 30 Jun 2000 20:37:11 -0400, Vividh Siddha wrote: >I am a novice to vxWorks and had a rather intriguing question about the >way C++ classes are handled. I wrote a simple class as follows: >class classA{ > void fnA(void); >private: > int i, j; >}; >note: fnA prints the values of i and j; > >I compiled this into a object file and loaded on my target. On the >WindSh prompt I called fnA it actually printed the values of i and j; >But I have just the defination of the class. There is no object of that >class. > >I come from a Unix background and there you have a class defination and >then you allocate a obj by actually instantiating the obj of the class. >Its only after this memory is allocated for this obj. But how does it >work internally in vxWorks? > >Thanks gurus.... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++: taskSpawn(member functions) Date: Sat, 29 Jul 2000 20:44:40 GMT From: Xenos Organization: Hades Message-ID: References: <39529A59.F8C47A33@yahoo.de> Opps! my mistake! make the private section protected!! And you may want to put the wrapper function the protected (or even private section). Also, to "spawn" your task, call the spawn member function. You could even call the spawn function from the constructor to automatically start the task when it is created, and put taskDelete in the destructor to automatically shut it down when the class goes out of scope--but be careful!! modified example: class foo { public: foo(); ~virtual foo(); spawn(char* name, int priority = 100, int options = VX_FPTASK, stack = 20 * 1024) { taskSpawn(NULL, 100, VX_FP_TASK, 20 * 1024, wrapper, (int) this,0,0,0,0,0,0,0,0,0);} protected: virtual int task_func() = 0; private: static int wrapper(foo* func) {return func->task_func();} }; class MyFoo : public foo { MyFoo(); ~virtual MyFoo(); protected: virtual int task_func(); }; int MyFoo::task_func() { // task code return 0; } // task object MyFoo task1; // spawn task task1.spawn(); On Sat, 29 Jul 2000 20:09:36 GMT, Xenos wrote: >Use a C function (or better yet, a static member if the linkage is the >same) as a wrapper, sending it the "this" variable as its sole >parameter > >example: > >class foo { > >public: > int task_func(); >}; > >extern "C" int wrapper(foo* func) >{ > return func->task_func(); >} > >foo task; >taskSpawn(NULL, 100, VX_FP_TASK, 20 * 1024, wrapper, (int) > &task,0,0,0,0,0,0,0,0,0); > > > >OR, if you can, this is "cleaner": > >class foo { > >public: > foo(); > ~virtual foo(); > static int wrapper(foo* func) > {return func->task_func();} > > spawn(char* name, int priority = 100, int options = VX_FPTASK, > stack = 20 * 1024) { > taskSpawn(NULL, 100, VX_FP_TASK, 20 * 1024, wrapper, (int) > this,0,0,0,0,0,0,0,0,0);} > > >private: > virtual int task_func() = 0; >}; > >// now just inherit from foo and override task_func with your task >code. Note that I'm writing this on-the-fly so there may be lexical >errors! > > > >How this helps. I haven't done this yet in VxWorks, but I do it a lot >in Visual C++. > > >On Fri, 23 Jun 2000 00:59:37 +0200, Matthias wrote: > >>I'm currently putting some old code into C++ and this problem came up. >>The class Cxx is used as communication node for multiple applications. >>Each application registers at this class and defines a number of message >>IDs for incoming messages. >> >>Then, there is a central receiver (CRX) for all incoming traffic, looks >>on the message ID and puts the messages in VxWorks message queues. >> >>Now I want that every application when registering spawns a task to >>receive messages from CRX and the code for this task is a member >>function of the class. I'm getting curious if there is any way to make a >>taskSpawn() and give a reference to a member function. >> >>Currently I'm using a static function and pass a pointer to the instance >>of the class, this works fine - but does anybody know something >>different ? >> >>Thanks, Matthias --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using C++ & gcc 2.95.2 Date: Sat, 29 Jul 2000 21:04:18 GMT From: Xenos Organization: Hades Message-ID: References: <8istha$rrc$1@bcrkh13.ca.nortel.com> There are a few (surmountable!) problems here. 1. You must tell the make while building the compiler, not to make the constructors static. Sorry, but I don't remember what the #define for this is, but it contains CONSTRUCTOR. You must #undef this. 2. The new compile will make symbols that contain a period where the symbols in the library contain a $. After you munch the file, using sed or awk or such, change the periods to $ for symbols in the library (compile with the option to allow dollars signs or use -traditional if it works). For extern symbols to your own code, change the periods to somethings else, like the word "dot" and add an asm attribute telling the compiler the "real" linkage name. 3. you may also want to create the compile using _Dexit=used-exit (or whatever it it---my memory is bad!) because VxWorks does not use the _exit function. Anything I forgot? I don't know if anyone has gotten exceptions to work yet. On 22 Jun 2000 11:32:58 GMT, jnance@wrtph0qb.nortelnetworks.com (Jim Nance) wrote: >Hello All, > I am trying to use a newer version of gcc to build my VxWorks load. I >have everything compiling, but it will not link, and it looks like the >problem has to do with c++ constructors and destructors. Has anyone >gotten VxWorks to work with gcc-2.95.2 and c++? > >Thanks, > >Jim --------------------------- Newsgroups: comp.os.vxworks Subject: supports chinese/japanses/koera (CJK) charset? Date: 27 Jul 2000 00:55:54 GMT From: Sun_AND_Moon Organization: The University of Hong Kong Message-ID: <8lo1aq$122q6@hkunae.hku.hk> Reply-To: mailto:NOSPAMSkmku@NO_SPAMShku.hk Hi, i am a newbie on this OS. i am interested in if VxWorks supports CJK charset? i.e. double-byte, 8-bit oriental charset like big5, GB, GBK, JIS , unicode. - -- Regards, ********************************************************************* K.M. Ku ********************************************************************* --------------------------- Newsgroups: comp.os.vxworks Subject: 2700(750) ppc - where MSR(EE) initially set? Date: Wed, 26 Jul 2000 17:09:09 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397F5375.4D3A441A@mindspring.com> I am getting a "uninitialized interrupt" from my pcc "decrementer" register. I am not using it for anything at the moment, because I have moved the sysClk/timestamp functions elsewhere. Apparently I need to install a noop isr before MSR(EE) is initially enabled. Problem is, I cant find where that is. I was expecting it to be in the usrInit/sysHwInit thread of execution, but I cant find it. the only thing that comes close is a call to intEnable(), which of course I do not have source to... but dissasembling it seemed to show that it was not modifying MSR(EE).... but then again, i am not very good at reading pcc machine language yet. Can someone point me in the right direction? thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: TrueFFS mountFTL Date: Sun, 30 Jul 2000 01:30:32 GMT From: fywang@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8m00fo$6on$1@nnrp1.deja.com> Hi all, I am trying to build a TrueFFS on my 2 AM29LV017D chips. Since AM29LV017D is a NOR-based flash, I should define INCLUDE_TL_FTL in my sysTffs.c file. In target/src/drv/tffs/tffsConfig.c: TLentry tlTable[] = /* TL tables */ { ... #ifdef INCLUDE_TL_FTL #ifdef FORMAT_VOLUME {mountFTL, formatFTL}, #else mountFTL, #endif /* FORMAT_VOLUME */ #endif /* INCLUDE_TL_FTL */ ... } I got a compile error when building my BSP: ... C:\Tornado\target\proj\gold860 \default\C:\Tornado\target\src\drv\intrCtl\ppc860Intr.c:224: undefined reference to `mountFTL' C:\Tornado\target\proj\gold860 \default\C:\Tornado\target\src\drv\intrCtl\ppc860Intr.c:224: undefined reference to `formatFTL' make: *** [vxWorks] Error 0x1 Searching "mountFTL" shows it's referenced by h/tffs/stdcomp.h & src/drv/tffs, no definition. The only occurence in lib is libPPCEC603gnuvx.a, but I am working on PPC860. Is there something wrong with my Tornado2/VxWorks? Can I work around this problem? Regards, Tom Wang Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Compiling many files w/ one command Date: Thu, 27 Jul 2000 00:31:03 GMT From: ryancerium@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8lnvs8$g1a$1@nnrp1.deja.com> I'm using Tornado 2.0. I made one header file that contains all my semaphores, variables, #defines, etc. Good software practice, right? I made several files each containing two tasks, a consumer and a producer. Good software practice, right? Then why won't Tornade compile all them together? Each file gets compiled as a separate object, with its own copies of the semaphores, variables, etc. This means none of my tasks can communicate. Why does this seem stupid to me? This would be fixed by compiling them all on the same line: cc 1.c 2.c 3.c instead of: cc 1.c cc 2.c cc 3.c Any advice? Do I *have* to make them all externs? Do I have to pass *everything* as an argument? Is there a way to make them all compile on one line? Thanks a lot for helping a newbie, Ryan Phelps Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Web Server under VxWorks Date: Thu, 27 Jul 2000 05:37:59 GMT From: Computer Engineer Organization: Somewhere in the Earth Message-ID: <397F59F5.5C837EA3@here.com> References: <397D1FD2.2D51DB48@yahoo.com> Wind River's Wind Web Server is an optional component which run on the top of VxWorks. Shelly HUANG wrote: > > Hi, > > Is any web server running under VxWorks? > > thank you for any information. > > best regards, > Shelly --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: Sat, 29 Jul 2000 23:29:23 -0700 From: Mohit Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <1b8aeca0.856a27a7@usw-ex0105-038.remarq.com> References: <03db9226.c358038d@usw-ex0105-038.remarq.com> <8lu0p4$tpb$1@nnrp1.deja.com> Hi John: Thanks for your most valuable reply. [1] I have checked and we have the PID-7T board and the BSP is also there installed in the Tornado directory. That's a start, I think... The fles are there... and it looks that it shold get me started. [2] About talking to the board.. ummm, this uses what the ARM manual calls "Angel". The connections involved plugging a parallel port adapter [kind of thing] to the parallel port on the board. That has some stuff on it which is supposed to allow me to program the board. There is a paralll as well as serial connectionbetween the PC and the board right now. [3] I can't find what software will do the talking. I think that I should at least be able to use HyperTerminal for the task. But, I start it and it asks to install a modem [???] which I don't have. Anyway, what are the settings for the board? 9600bps, N-8-1? or something different? Lastly (at least for now), how do I get the "Target Server" set up? That might be the problem, perhaps.... You guys are being a great help.... Thanks a lot, Mohit. - ----------------------------------------------------------- Got questions? Get answers over the phone at Keen.com. Up to 100 minutes free! http://www.keen.com --------------------------- Newsgroups: comp.os.vxworks Subject: sntpclib, localtime() and TIMEZONE Date: Sun, 30 Jul 2000 02:55:24 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <3983D15C.1233E21E@mindspring.com> I am having trouble figuring out how to tell vxw about what timezone and dst it should use. Under ansiTime lib the ref man mentions an env sym "TIMEZONE", but since we dont build with that component, that doesn't help much... Is there an easy way to do this? I am using sntpcTimeGet() to get the time from one of our suns; then calling clock_settime() to tell the os. The results returned from localtime() and gmtime() are the same. My goal is to get localtime() to account for timezone and dst data. thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: mv2700 - unknown symbol "__eabi" Date: Wed, 26 Jul 2000 22:31:23 -0400 From: Al Johnston Organization: MindSpring Enterprises Message-ID: <397F9EFB.C88B8301@mindspring.com> References: <397F406C.F6A45133@mindspring.com> > this one is probably easy. I compiled a hello world program per > the vxw P.G. direction. When I load it the loader tells me > undefined symbol: __eabi (binding 1 type 0) > nm on the object, and sure enough there is an entry > U __eabi And the answer is: Do NOT name your procedure "main"; If you do the compiler will try to do the stuff it does for a "normal" unix main c routine... which is not what we want here in vxWorks... Thanks for the clue guys... I obviously need one! --------------------------- Newsgroups: comp.os.vxworks Subject: help about MPC860 SCC HDLC mode Date: Thu, 27 Jul 2000 16:40:44 +0800 From: srdlm@sbell.com.cn (liuming) Organization: sbell Message-ID: hi everyone: I want to change my HDLC address dynamicly after vxworks is up. That means that I want to write a routine to change the HDLC address of SCC2. I want to know whether I can change the HADDR in the parameter RAM of SCC2 directly? Must I do some additional work, such as stop the interrupt or stop the driver? Thanks in advance! eric liu --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Jul 30 06:16:35 2000 From: Tewari C Date: Sun Jul 30 06:16:37 PDT 2000 Subject: Changing the value of LOCAL_MEM_SIZE from 0x0100 0000 to 0x0200 0000 results in Boot-up failure Hi, We are using Tornado 2.0 and VxWorks 5.4. Our board is based on the EBSA285 development platform. The DRAM used on the EBSA285 was 16MB. However our board as 32MB DRAM i.e. 4 chips X 16bit wide each. Changes have been made in the rominit.s to change the DRAM initialization for full 32MB DRAM and LOCAL_MEM_SIZE has been redefined to 0x02000000. When Bootrom is used to Boot-up the process hangs at the point before loading the VxWorks image via FTP to the DRAM. However when LOCAL_MEM_SIZE is changed to 0x01000000 then the boot-up process is successful. After successful boot-up if the target shell is invoked in the Tornado environment and a memory dump is done for locations greater than the address 0x0100 0000 its shows some data there. This data can be modified and access is available to this region of the memory. The main question is why does changing LOCAL_MEM_SIZE to 32MB result in boot failure ? Does the EBSA285 have a different meaning for LOCAL_MEM_SIZE than the size of the local DRAM? Thanks, ======================================================= C.S.Tewari Future Software, Chennai, India 91-44-4330550 Ext. 322 email:tewarics@future.futsoft.com From vxwexplo-errs@csg.lbl.gov Mon Jul 31 02:11:38 2000 From: "Michael Lawnick" Date: Mon Jul 31 02:11:43 PDT 2000 Subject: Re: IDE driver errors Hi Brian, try this: In your BSP-workspace, register vxWorks go to hardware/peripherals/hard disks/IDE hard drive Right mouse button / 'include IDE hard drive' -> configuration error Right mouse button / 'Properties of IDE hard drive' -> Properties window opens. In the register Params you now can set the missing parameters. BTW: At a place I can't remeber, it was recommended to use ATA, as it should be a superset of IDE (I think - or not - has anybody else a comment to this ?). HTH Michael > In Tornado 2.0, when I enable IDE hard drive support, Tornado announces > that there are these errors and won't continue: > > Undefined parameters: > IDE_INT_VEC > IDE_INT_LVL > IDE_CONFIG > > I can find any info in the distribution on how to resolve this > problem. Does anyone here know how to correct this? > > I'm trying to mount an IDE hard drive using a PCI bus IDE controller > card in a MIPS target. > > Thanks in advance, > > Brian > 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 Jul 31 04:03:10 2000 From: Vxworks Exploder Date: Mon Jul 31 04:03:12 PDT 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 31 04:03:06 PDT 2000 Subject: Re: boot my boards from flash memory Subject: vxMemProbe Subject: handing packet off to IP Subject: Ethernet broadcast Subject: Changing the value of LOCAL_MEM_SIZE from 0x0100 0000 to 0x0200 0000 results in Boot-up failure Subject: Re: BSP in embeded-Linux ? Subject: BSP for the Mot 8260VADS? Subject: Re: TrueFFS mountFTL Subject: Re: mv2700 - unknown symbol "__eabi" Subject: Re: VxWorks & ARM Dev Board Subject: Re: Changing the value of LOCAL_MEM_SIZE from 0x0100 0000 to 0x0200 0000 results in Boot-up failure Subject: Re: sntpclib, localtime() and TIMEZONE Subject: Re: BSP for the Mot 8260VADS? Subject: Re: looking up static symbols Subject: Re: Bothered by DHCP Subject: Re: How to get MAC Addr. Subject: re: IDE driver errors Subject: Re: vxWorks Rom Resident Image for PPC ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: boot my boards from flash memory Date: Thu, 27 Jul 2000 11:28:15 +0800 From: "tanghaoyu" Organization: The news server of Changzhou TeleCom Message-ID: <8loa2e$1r3q$1@news.cz.js.cn> References: <397F7CC4.AD5BCC1A@admin.uottawa.ca> no problem. "no-one" wrote in message news:397F7CC4.AD5BCC1A@admin.uottawa.ca... > I would like to boot my boards from flash memory(e.g. via TrueFFS), but > still I would like to establish a shared memory backplane network. Is it > > technically possible? > > Thanks. > > Wei (wei_huo@hotmail.com) > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: vxMemProbe Date: 27 Jul 2000 10:49:01 GMT From: ola.hallstrom@esavionics.se (Ola Hallström) Organization: ericsson Message-ID: <8F7E849D9olahallstromesavioni@130.100.128.24> Hi, I wan't to test if a vme-board(with a Tundra Universe 2) is present in a specific vmebus slot from VxWorks. The board has been HW-initiated to have an "VME Register Access image" with a baseaddress specific for the slot it is connected to. Is it a good approach to read the address refering to the Tundra vendor/device id register, mapped in VxWorks vme-space, to see if there is a valid id? Can I use the VxWorks command "vxMemProbe(...)"? vxMemProbe is detecting a bus error. Will a bus error be generated when reading an address refering to a slot where no device is connected? Thanks /Ola --------------------------- Newsgroups: comp.os.vxworks Subject: handing packet off to IP Date: Thu, 27 Jul 2000 09:03:09 -0400 From: Dmitriy Zavin Message-ID: <3980330D.DB2C1C96@txc.com> Hi. I am working with VxWorks, and I have a pretty straight forward question. I am writing a kind of PPP driver, and I am processing an incoming packet. I've done evertyhing I need to it, and I am trying to pass it up the protocol stack to IP. As TCP/IP Illustrated Vol.2 (since vxworks is pretty BSD4.4 compliant) recommends, I call: s = splimp(); IF_ENQUEUE(&ipintrq, m); ipintr(); splx(s); where m is an mbuf with my packet (IP headers still intact). Well everything executes fine, but I don't see vxWorks processing the packet... Its a ICMP reply packet, but vxWorks IP stack never sees it. I don't even know how to debug this since I don't really know where the problem lies. Any help will be appreciated. Thanks - --Dima --------------------------- Newsgroups: comp.os.vxworks Subject: Ethernet broadcast Date: Sun, 30 Jul 2000 14:13:36 +0200 From: "Shlomo Pongratz" Organization: NetVision Israel Message-ID: <8m13so$1pc$1@news.netvision.net.il> Hi. I have an industrial Pentium based board running VxWorks 5.3.1 and an ene ethernet card. And it was found out that the board is forwarding any ethernet broadcast frame it receives, and creating a broadcast storm. Does anyone knows what might be the cause to this. Thanks. Shlomi. --------------------------- Newsgroups: comp.os.vxworks Subject: Changing the value of LOCAL_MEM_SIZE from 0x0100 0000 to 0x0200 0000 results in Boot-up failure Date: Sun, 30 Jul 2000 18:47:10 +0530 From: Tewari C Organization: Lawrence Berkeley National Laboratory Message-ID: <01BFFA56.92683EA0.tewarics@future.futsoft.com> Hi, We are using Tornado 2.0 and VxWorks 5.4. Our board is based on the EBSA285 development platform. The DRAM used on the EBSA285 was 16MB. However our board as 32MB DRAM i.e. 4 chips X 16bit wide each. Changes have been made in the rominit.s to change the DRAM initialization for full 32MB DRAM and LOCAL_MEM_SIZE has been redefined to 0x02000000. When Bootrom is used to Boot-up the process hangs at the point before loading the VxWorks image via FTP to the DRAM. However when LOCAL_MEM_SIZE is changed to 0x01000000 then the boot-up process is successful. After successful boot-up if the target shell is invoked in the Tornado environment and a memory dump is done for locations greater than the address 0x0100 0000 its shows some data there. This data can be modified and access is available to this region of the memory. The main question is why does changing LOCAL_MEM_SIZE to 32MB result in boot failure ? Does the EBSA285 have a different meaning for LOCAL_MEM_SIZE than the size of the local DRAM? Thanks, ======================================================= C.S.Tewari Future Software, Chennai, India 91-44-4330550 Ext. 322 email:tewarics@future.futsoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: BSP in embeded-Linux ? Date: Sun, 30 Jul 2000 15:34:53 GMT From: clayton@shore.net (Mark Clayton) Organization: Shore.Net, a PRIMUS Company (info@shore.net) Message-ID: References: <8kcb6r$k2i$1@schbbs.mot.com> Sender: mark@shore.net [Posted and mailed] In article <8kcb6r$k2i$1@schbbs.mot.com>, "sky" writes: > Hi: > I've known that there is BSP in vxWorks to map the > hardware resource to vxWorks's software. BSP make > a standard unique software (vxWorks) KNOWs the various > hardwares. > But if I use Linux as my embeded OS, where is its BSP? > Or,if I have to change kernal/driver source codes to do > the mapping? it's impossible ! In the Linux world there is no such thing as a BSP. When you download the kernel you get all of the code for all of the supported platforms. There are a couple of places to look to see the platform specific code. For most of it look under /usr/src/linux/arch, but also look in 'drivers' for peripheral code. I hope this helps. Mark - -- email: clayton@shore.net --------------------------- Newsgroups: comp.os.vxworks Subject: BSP for the Mot 8260VADS? Date: Sun, 30 Jul 2000 15:43:43 GMT From: clayton@shore.net (Mark Clayton) Organization: Shore.Net, a PRIMUS Company (info@shore.net) Message-ID: Sender: mark@shore.net Is there a BSP for the 8260VADS board. it sounds like it from heading this group. What is it called? WindRivers says to use the 860FADS BSP, but that seems incorrect. Are they right!?! Thanks, Mark - -- email: clayton@shore.net AIM: Clayton256 http://www.goodmanhillsoftware.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TrueFFS mountFTL Date: 30 Jul 2000 16:24:38 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8m1ks6$kfo$1@news1.xs4all.nl> References: <8m00fo$6on$1@nnrp1.deja.com> fywang@my-deja.com wrote: : Hi all, : I am trying to build a TrueFFS on my 2 AM29LV017D chips. : Since AM29LV017D is a NOR-based flash, I should define INCLUDE_TL_FTL : in my sysTffs.c file. : Searching "mountFTL" shows it's referenced by h/tffs/stdcomp.h & : src/drv/tffs, no definition. : The only occurence in lib is libPPCEC603gnuvx.a, but I am working on : PPC860. : Is there something wrong with my Tornado2/VxWorks? : Can I work around this problem? No, there is nothing wrong with your installation, but you probably did not purchase TFFS. TFFS is delivered for use with Disc on Chip, but not with other devices. However, it is possible to use the ftllite.o from the EC603 library, see my page for more details about this. The problem is that WRS made a mistake with this distribution and enabled TFFS, but only for the EC603, at least that is what out sales representative told us. So officially you cannot use it. But you could always try this version and see it this is what you needed...... 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: mv2700 - unknown symbol "__eabi" Date: Thu, 27 Jul 2000 18:51:15 +0200 From: Ilia Organization: Terayon Communication Systems Message-ID: <39806883.6AB4953A@tera.yon.dot.com> References: <397F406C.F6A45133@mindspring.com> Hi, Johnston. You're call Your function "main", right ? As per PPC "extended Application Binary Interface" (eabi) standard, the "main" function must call the special __eabi function, that supposed to set the needed environment to run Your program. According to this, GNU compiler, inserts the call to __eabi in the "main" function. Tornado DOES NOT provide that function, because in a real-time environment You doesn't expected to write the "main" function. VxWorks sets up the environment before the application code run and there is no need (and no sense) for "main". Give Your function any name You like, and to test it call it from the shell. Regards Ilia. Al Johnston wrote: > > Hi guys, > > this one is probably easy. I compiled a hello world program per > the vxw P.G. direction. When I load it the loader tells me > undefined symbol: __eabi (binding 1 type 0) > nm on the object, and sure enough there is an entry > U __eabi > > What am I doing wrong? > > tor 2, mv2600/750 cpu, host is solaris. > > tnx. - -- +================================= | Bosis Ilia | Senior Software engineer | Terayon Information Systems inc. | Broadband voice business unit. | E-mail: iliab "at" terayon "dot" com | WWW: http://www.terayon.com +================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks & ARM Dev Board Date: Sun, 30 Jul 2000 18:28:01 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8m1s3d$cvh$1@nnrp1.deja.com> References: <03db9226.c358038d@usw-ex0105-038.remarq.com> <8lu0p4$tpb$1@nnrp1.deja.com> <1b8aeca0.856a27a7@usw-ex0105-038.remarq.com> Hi, > [1] I have checked and we have the PID-7T board and the BSP is > also there installed in the Tornado directory. That's a start, I > think... The fles are there... and it looks that it shold get me > started. OK, well PID ships with a bootrom (a physical device). You should check in the box that you got for the bootrom. This can then be plugged into the socket labelled U12 on the board (take out the device that is there, if there is one, and keep it safe). You then need to check in the manuals (the ones online will do) and look under BSP reference for the entry called pid. This contains all the details about running VxWorks on the PID boards, incuding the settings for all the jumpers and switches (if any) that are on the board. Make sure you check these carefully. If you don't have the PROM device, all is not lost... You will need to find a PROM programmer though. In the BSP there should be an image called bootrom.bin or bootrom.hex. You will need to determine whether you PROM programmer reads binary or Motorola format S-records (.hex). It may read either. If the file format you need is not present just type "make bootrom.hex" or "make bootrom.bin" to get it built. I think you can also do this from the Tornado IDE. Then take that file and program a PROM (27C020 or larger) with that image. There is also an option for the PID boards to replace the PROM device that the bootrom is held in with a flash device (the BSP manual tells you all about this). The difference is that you will be able to have non-volatile memory for the boot information, and also use TFFS if you have bought that. > [2] About talking to the board.. ummm, this uses what the ARM > manual calls "Angel". The connections involved plugging a > parallel port adapter [kind of thing] to the parallel port on the > board. That has some stuff on it which is supposed to allow me > to program the board. There is a paralll as well as serial > connectionbetween the PC and the board right now. You don't need angel to run VxWorks. Angel is the monitor program that ARM's development tools use. You can forget about it while you are using the board with VxWorks. > [3] I can't find what software will do the talking. I think that > I should at least be able to use HyperTerminal for the task. But, > I start it and it asks to install a modem [???] which I don't > have. Anyway, what are the settings for the board? 9600bps, > N-8-1? or something different? Lastly (at least for now), how do > I get the "Target Server" set up? That might be the problem, > perhaps.... I'm assuming that you are using a PC as your Tornado host; in that case you should find on the Tornado menu a couple of hyperterminal sessions that are preconfigured for VxWorks. One is for COM1, the other for COM2. The speed configuration is 9600 8n1, but there are some other things that you need to do to make the MS terminal actually behave like a terminal and not steal keystrokes from you. Also, it is worth reading the getting started manuals that came with Tornado. There is information in there about configuring terminal connections and how to set the boot parameters once you have it all configured. Hope that helps, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing the value of LOCAL_MEM_SIZE from 0x0100 0000 to 0x0200 0000 results in Boot-up failure Date: Sun, 30 Jul 2000 18:36:41 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8m1sjp$dbo$1@nnrp1.deja.com> References: <01BFFA56.92683EA0.tewarics@future.futsoft.com> Hi, I think that this is a known issue. Contact Wind River customer support; they may have more information on this. Regards, John... In article <01BFFA56.92683EA0.tewarics@future.futsoft.com>, Tewari C wrote: > Hi, > > We are using Tornado 2.0 and VxWorks 5.4. Our board is based on the EBSA285 > development platform. > The DRAM used on the EBSA285 was 16MB. However our board as 32MB DRAM i.e. > 4 chips X 16bit wide each. Changes have been made in the rominit.s to > change the DRAM initialization for full 32MB DRAM and LOCAL_MEM_SIZE has > been redefined to 0x02000000. When Bootrom is used to Boot-up the process > hangs at the point before loading the VxWorks image via FTP to the DRAM. > However when > LOCAL_MEM_SIZE is changed to 0x01000000 then the boot-up process is > successful. > After successful boot-up if the target shell is invoked in the Tornado > environment and a memory dump is done for locations greater than the > address 0x0100 0000 its shows some data there. This data can be modified > and access is available to this region of the memory. > The main question is why does changing LOCAL_MEM_SIZE to 32MB result in > boot failure ? > Does the EBSA285 have a different meaning for LOCAL_MEM_SIZE than the size > of the local DRAM? > > Thanks, > > ======================================================= > C.S.Tewari > Future Software, > Chennai, India > 91-44-4330550 Ext. 322 > email:tewarics@future.futsoft.com > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sntpclib, localtime() and TIMEZONE Date: Sun, 30 Jul 2000 19:45:40 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8m20l4$fto$1@nnrp1.deja.com> References: <3983D15C.1233E21E@mindspring.com> Hi, > I am having trouble figuring out how to tell vxw about what timezone and > dst it should use. Under ansiTime lib the ref man mentions an env sym > "TIMEZONE", but since we dont build with that component, that doesn't > help much... > > Is there an easy way to do this? Yep, include the environment variable support, and set TIMEZONE. It is the only easy way to change VxWorks from using UTC (which is the locale it has by default). There is a harder way, though I haven't tried this so it comes with no guarantee... The time conversion functions make use of a pointer to a structure of type TIMELOCALE called __loctime. This pointer is global so you could move it to point to your own TIMELOCALE structure containing the timezone information you need. This is not trivial to set up, but also not so complex as to be impossible. Email me if you are having trouble and I can try to send you an example, but I still think you'd be better off with the TIMEZONE environment variable... it is much simpler to set up, and the component is very small (especially for an mv2700 which probably has many megabytes of memory). It is also guaranteed to be available from one version of VxWorks to the next, unlike anything that plays with undocumented pointers and structures ;-) Regards, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: BSP for the Mot 8260VADS? Date: Sun, 30 Jul 2000 22:40:17 GMT From: nospam@home.net (Gary Beck) Organization: @Home Network Message-ID: <3987adf6.20551791@24.0.240.30> References: Yes, there is a BSP for the Motorola VADS8260. The first verion is titled VADS820 and creates a drectory of the same name under \tornado\target\config directory. The second version is titled ADS8260. Be careful with their serial driver. It does not work above 9600 baud. There are also other bugs with this BSP. The BSP name that WRS gave you is incorrect. On Sun, 30 Jul 2000 15:43:43 GMT, clayton@shore.net (Mark Clayton) wrote: > >Is there a BSP for the 8260VADS board. it sounds like it >from heading this group. What is it called? WindRivers >says to use the 860FADS BSP, but that seems incorrect. Are >they right!?! > >Thanks, >Mark --------------------------- Newsgroups: comp.os.vxworks Subject: Re: looking up static symbols Date: Mon, 31 Jul 2000 11:37:16 +0530 From: Yusuf Motiwala Organization: Lucent Technologies, Columbus, Ohio Message-ID: <39851794.719B14E4@lucent.com> References: <3982D0A1.5A4777C8@eed.ericsson.se> Also, try "_symbol" along with "symbol" for finding "symbol". Here, I assume, symbol table is already present in target. Yusuf Thomas Scholz wrote: > I'm having problems to find static function names in the systems symbol > table. I'm using symFindByValue(sysSymTbl,...) in Tornado 1.0.1. > Are static symbols defined elsewhere or is there another trick ? All vxworks > functions (eg "l" or "tt") display them as expected. > > Regards, Thomas. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Bothered by DHCP Date: Mon, 31 Jul 2000 11:32:53 +0530 From: Yusuf Motiwala Organization: Lucent Technologies, Columbus, Ohio Message-ID: <3985168C.B68B169F@lucent.com> References: <8lhhgl$l6o$1@nnrp1.deja.com> DHCP server IP pool is also defined in usrNetDhcpsCfg.c. So probably you want to make sure that it picks up from the correct file. Yusuf ellin_lin@my-deja.com wrote: > I'm a tenderfoot of vxWorks and Tornado. And I > met some problems when I tried to configured a > DHCP server. I modified usrnetwork.c setting the > IP range as written in user manual, and test this > DHCP server with a DHCP client of Win98, it did > nothing. Can someone give me some guidance? > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get MAC Addr. Date: Mon, 31 Jul 2000 11:48:09 +0530 From: Yusuf Motiwala Organization: Lucent Technologies, Columbus, Ohio Message-ID: <39851A21.6FE6B2FC@lucent.com> References: <8lirc9$lle$1@nnrp1.deja.com> where do you want to get MAC address from? From device? then probably you need to refer the device manual. However, if device driver is already loaded, you can get the MAC address from device ifnet structure. Yusuf ellin_lin@my-deja.com wrote: > I'm developping software for MCP750 board with tornadoII, Can somebody > tell me how to get the MAC Addr.? > You're appreciated if giving some response. > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: re: IDE driver errors Date: Mon, 31 Jul 2000 10:52:47 +0200 From: "Michael Lawnick" Organization: Lawrence Berkeley National Laboratory Message-ID: <200007310852.KAA25143@net.softec.de> Hi Brian, try this: In your BSP-workspace, register vxWorks go to hardware/peripherals/hard disks/IDE hard drive Right mouse button / 'include IDE hard drive' -> configuration error Right mouse button / 'Properties of IDE hard drive' -> Properties window opens. In the register Params you now can set the missing parameters. BTW: At a place I can't remeber, it was recommended to use ATA, as it should be a superset of IDE (I think - or not - has anybody else a comment to this ?). HTH Michael > In Tornado 2.0, when I enable IDE hard drive support, Tornado announces > that there are these errors and won't continue: > > Undefined parameters: > IDE_INT_VEC > IDE_INT_LVL > IDE_CONFIG > > I can find any info in the distribution on how to resolve this > problem. Does anyone here know how to correct this? > > I'm trying to mount an IDE hard drive using a PCI bus IDE controller > card in a MIPS target. > > Thanks in advance, > > Brian > 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 ============================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks Rom Resident Image for PPC Date: Mon, 31 Jul 2000 10:31:36 +0100 From: Jason Garnham Message-ID: <39854777.41B18B11@gecm.com> References: <39364C90.EBDD87B5@west.net> I have done this but still currently have a few problems. You have to tell the loader where your .text and .data segments are to go, and then it largely works. My outstanding problem is that globally defined variables that have a pre-initialised value do not get their value assigned. What should happen is that a piece of initialization code should copy where the .data segment is in ROM to where it will reside in RAM. I believe a file called crt0.o should do this. I can't find such a piece of code in the vxWorks/BSP distribution for our PPC750 cards. I can use a command file for the loader to export some symbols, and then run some code to do the copy (see p465 GNU Tool kit users guide). The problem with this is that (after inspection of the map file) the ROM .data address does not follow directly on from the .text segment. A segment .rodata is in-between, knowing this you can export the end of this and use it for copying. I'm not happy with this solution as it requires far too much fiddling with the linker/loader for my liking. My current work around is to explicitly assign any data in initialization code. The good news is that it does not effect a full blown vxWorks image (e.g.. vxWorks_rom.hex) as someone appears to have coded some initialization code in there somewhere. Michael Michel wrote: > I am trying to build a vxWorks rom resident image for the PPC740 CPU. > It is to run from Flash memory after downloaded from the network via > bootroms. > Has anybody sucessfully create building such an image for this CPU type. > > Can you tell me what obstacles you needed to overcome before it > finally ran sucessfully from flash (or rom)?? > I'd be oblidged. Thanks --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Jul 31 14:27:09 2000 From: "Joseph Sebastian" Date: Mon Jul 31 14:27:11 PDT 2000 Subject: Loading new application Hi all, I like to incorporate the feature to upgrade my vxWorks application software in the system. I want to load a latest version of my software and replace the existing one, the command for this will be given from an SNMP manager. The module to be loaded is kept in a TFTP server, download into the RAM and do a version checking on it. If the module is acceptable the software should be loaded from this location. Any help for this will be well appreciated. Joseph KS. ________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com