From vxwexplo-errs@csg.lbl.gov Wed Mar 1 01:54:18 2000 From: teit.uldal@philips.com Date: Wed Mar 1 01:54:20 PST 2000 Subject: Re: using vxVMI with a pc486 BSP Dear VxWorkers, We are trying to use the vxVMI option for an x86 target - but don't seem to be able to get it working under Tornado 1.0.1. Both when INCLUDE_MMU_BASIC and INCLUDE_MMU_FULL are defined, the downloaded image hangs when calling 'usrMmuInit'. We are using a 'pc486' BSP as a starting point and we do not have a BIOS on our target but our own boot code. Thanks in Advance, Teit Tornado vxWorks Here is the sysPhysMemDesc entry from 'syslib.c': PHYS_MEM_DESC sysPhysMemDesc [] = { /* adrs and length parameters must be page-aligned (multiples of 0x1000) */ /* lower memory */ { (void *) LOCAL_MEM_LOCAL_ADRS, (void *) LOCAL_MEM_LOCAL_ADRS, 0xa0000, /* 640 KB */ VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE }, /* video ram, etc */ { (void *) 0xa0000, (void *) 0xa0000, 0x60000, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, /* upper memory */ { (void *) 0x100000, (void *) 0x100000, LOCAL_MEM_SIZE - 0x100000, /* it is changed in sysMemTop() */ VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE }, /* TODO -The default ROM is CACHEABLE, change if necessary. { (void *) ROM_BASE_ADRS, (void *) ROM_BASE_ADRS, ROM_SIZE, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE_NOT | VM_STATE_CACHEABLE }, */ #ifdef INCLUDE_PCI { (void *) FEI0_MEMBASE0, (void *) FEI0_MEMBASE0, FEI0_MEMSIZE0, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, { (void *) FEI1_MEMBASE0, (void *) FEI1_MEMBASE0, FEI1_MEMSIZE0, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, { (void *) FEI2_MEMBASE0, (void *) FEI2_MEMBASE0, FEI2_MEMSIZE0, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, { (void *) FEI3_MEMBASE0, (void *) FEI3_MEMBASE0, FEI3_MEMSIZE0, VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, #endif /* INCLUDE_PCI */ /* TODO - local I/O devices & NVRAM */ /* TODO- Add additional entries for VME space, special I/O devices, etc */ }; #ifdef INCLUDE_FLASH #endif #ifdef INCLUDE_VME /* TODO - one-time setup of control registers for master window mapping */ #endif ---------------------------------------------------------------------------------------------------------- Teit Uldal Software Engineer Digital Networks Product Services Philips Industrial Activities nv Kempische Steenweg 293 B-3500 Hasselt, Belgium tlf / fax: +32 (0)11 305816 / 958 email: teit.uldal@philips.com From vxwexplo-errs@csg.lbl.gov Wed Mar 1 04:03:10 2000 From: Vxworks Exploder Date: Wed Mar 1 04:03:12 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Mar 1 04:03:06 PST 2000 Subject: Slow download over network Subject: VxWorks Patches Subject: Re: VxWorks Patches Subject: Re: Stethoscope 5.0 Subject: Re: memory autosizing using sysPhysMemTop Subject: Re: memory autosizing using sysPhysMemTop Subject: Re: GUIs for Embedded Systems using VxWorks Subject: Use VisionIce+VxWorks Subject: Pass Through Filesystem for VxSim Subject: RIP Problem Subject: Re: tempnam ? Subject: non-blocking pipes Subject: Re: Windows Managers for VxWorks Subject: Re: Stethoscope 5.0 Subject: Re: memory leak deleting a pipe Subject: Drivers & signals Subject: Re: STD_IN failing on target shell Subject: Re: Embedded int 3 handing with vxWorks. Subject: Re: builtin_new/builtin_delete Subject: Re: STD_IN failing on target shell ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Slow download over network Date: 29 Feb 2000 14:25:13 +0100 From: "Jakob Pousard" Message-ID: <01bf82b8$3f5399a0$0525e38a@sesub-w2375> When I load an image to the target is very slow. The network traffic is wired. First the frames contains >1300 bytes data. But the frame seems to contain only <100 bytes of data and the rest is padding (zeros). After a while ther frame size is reduced to <100 and the padding seems to be gone. This is what happened with T2 on NT4 and using ethernet network as target connection. If I load the same image from T1 it loads very quickly. The BSP of the target is not changed. What I have done wrong Jakob Pousard jakob.pousard@se.abb.com --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Patches Date: Tue, 29 Feb 2000 10:13:28 -0500 From: John Sestak Organization: Raytheon, Falls Church Message-ID: <38BBE218.F11805CA@fallschurch.esys.com> Hi all, I was told that VxWorks has a number of patches that are available for it. When I look around the Wind River site, and do a search on "patch" or "update", etc. I get nothing related to software patches. Does anyone have a link to the hidden software patches page? Thanks, John Sestak --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks Patches Date: 29 Feb 2000 08:27:44 -0700 From: taashlo@sandia.gov Organization: Sandia National Laboratories, Albuquerque, NM USA Message-ID: References: <38BBE218.F11805CA@fallschurch.esys.com> John Sestak writes: > Hi all, > I was told that VxWorks has a number of patches that are available > for it. When I look around the Wind River site, and do a search on > "patch" or "update", etc. I get nothing related to software patches. > Does anyone have a link to the hidden software patches page? From the home page, click on "Support" and then "WindSurf", log in, click on "Download" and then click on "Patches". Or just go here: http://www.windriver.com/csdocs/kplocator/patchList.shtml Tad --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Stethoscope 5.0 Date: Tue, 29 Feb 2000 16:23:14 GMT From: Ted Dennison Organization: Deja.com - Before you buy. Message-ID: <89grpc$vc7$1@nnrp1.deja.com> References: <38BB3124.2A45485F@sympatico.ca> In article <38BB3124.2A45485F@sympatico.ca>, mike.mcnaughton@sympatico.ca wrote: > I have purchased Steth 5.0 and wish to install it tornado 1.01, > vxworks 5.3.1. > > So far it has done nothing but frustrate the Hell out of any attempts > to install it.. > > Is there someone among you who wouln't mind taking a few minutes to > outline a detailed procedure of what you had to do to get it going. > and helped me out with this in a manner which led to my getting it up > and running, I would buy you a 24-pack of beer. Hmm. You aren't in Oklahoma perchance? :-) We've had little problem with developers installing Stethescope themselves and getting it to work. The one that's a bear is Tornado. Are you sure you are getting it installed properly? Can you build BSP's, for instance? - -- T.E.D. http://www.telepath.com/~dennison/Ted/TED.html Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: memory autosizing using sysPhysMemTop Date: Tue, 29 Feb 2000 10:39:29 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38BC1261.62E7@despam.pangea.ca> References: <20000229003531.15846.00002347@ng-cd1.aol.com> TMiller652 wrote: >[...] > Does anyone know a reason why determining memory using sysPhysMemTop WOULDN"T > override the previous definition of LOCAL_MEM_SIZE? Any assistance would be > greatly appreciated. > Hi Tom, I can't give you a definite answer, but I suspect that the macro(s) have some compile time meaning. There is a bunch of memory macros some of which refer to each other. You might want to check this. The result of the generic sysPhysMemTop() calculation is stuffed into sysPhysMemDesc in sysLib.c and used in MMU initialization. - -het - --- "Assembly of Japanese bicycle require great peace of mind." - Instruction manual mentioned by R. Pirsig Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: Re: memory autosizing using sysPhysMemTop Date: Tue, 29 Feb 2000 13:15:53 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <38BC1AE9.F9FABABD@aps.anl.gov> References: <20000229003531.15846.00002347@ng-cd1.aol.com> TMiller652 wrote: > > Hello, > > I'm trying to have my vxworks system detect the amount of ram available to the > system at startup and configure accordingly. > > I have a method that I can use to test for the size of physical memory, and > i've included it in the function sysPhysMemTop. I thought this would override > the "#defined" memory size as described in "LOCAL_MEM_SIZE" - but it doesn't > seem to work. > > Does anyone know a reason why determining memory using sysPhysMemTop WOULDN"T > override the previous definition of LOCAL_MEM_SIZE? Any assistance would be > greatly appreciated. Does your sysHwInit routine properly set the size of the sysPhysMemDesc[] element that refers to the local memory to (sysPhysMemTop()-LOCAL_MEM_LOCAL_ADRS)? If you don't it would look like the memory doesn't exist. You don't say what BSP this is or what your "it doesn't work" symptoms are though. The standard way of switching from fixed to auto-sizing is by defining the macro LOCAL_MEM_AUTOSIZE and a #ifdef inside sysPhysMemTop() to override the fixed LOCAL_MEM_SIZE, but I don't think this switch is used anywhere else. If you've got something like a VMEbus I/O memory area immediately following the local memory then you need to make sure that the VME window register calculations make use of your sysPhysMemTop() value too, and of course that might need another modification in sysPhysMemDesc[]. - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GUIs for Embedded Systems using VxWorks Date: Tue, 29 Feb 2000 16:25:54 +0800 From: "tim" Organization: amlogic Message-ID: <89fvp7$2h6u$1@news.cz.js.cn> References: <896dc8$q2o3@overload.lbl.gov> Reply-To: "tim" Maybe ZeN for VxWorks? Fitch, Chris wrote in message news:896dc8$q2o3@overload.lbl.gov... > Fellow VxWorkers, > > We are trying to find a replacement for our current method of building GUIs: > RTX Windows and an X-Windows based GUI Builder. What other products are > available? For additional background: our system employs a Power PC Single > Board Computer with an attached Video Card. > > Thank you in advance, > Chris Fitch > LM NE&SS --------------------------- Newsgroups: comp.os.vxworks Subject: Use VisionIce+VxWorks Date: Mon, 28 Feb 2000 09:03:22 -0800 From: Antonio Fontan Organization: Alcatel USA Message-ID: <38BAAA59.AF758229@usa.alcatel.com> Hello, When can I find information on how to configure the host-target connection to work not using the target server but using instead the VisionIce emulator?. I would like to do something as what is described in the "tornado users guide" that you can do with the NETROM emulator. Thank you very much in advance, Antonio --------------------------- Newsgroups: comp.os.vxworks Subject: Pass Through Filesystem for VxSim Date: Tue, 29 Feb 2000 15:39:28 -0500 From: "Clare Sun" Organization: Frontier GlobalCenter Inc. Message-ID: <89han2$c69$1@nnrp02.primenet.com> Hi, There, I am doing some development work with Tornado 2.0 facility with VxSim. The host is running Windows 95. I am trying to access a disk file on the host from a downloadable application that is running on VxSim. According to the information from VxWorks Programmer's Guide I can accomplish it through pass through filesystem. I followed every clue in the reference material but could not get it work. Has anyone done it before or does anyone know how to do it? Thanks! --------------------------- Newsgroups: comp.os.vxworks Subject: RIP Problem Date: Tue, 29 Feb 2000 15:32:27 -0500 From: John Sestak Organization: Raytheon, Falls Church Message-ID: <38BC2CDA.96E4E7DD@fallschurch.esys.com> - --------------FA15A376FD9DEF61A3A6AE77 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello again, I'm currently having a problem with getting RIP (Routing Information Protocol) to work. I'm running a Radstone PPC4-750 with their version 2.1 BSP. When I include RIP using the default BSD network driver there doesn't seem to be any problem, but when I switch to the END (Enhanced Network Driver - for it's multicast support) driver and try and include RIP, I get the following error when I boot and press "i": - -> i NAME ENTRY TID PRI STATUS PC SP ERRNO DELAY - ---------- ------------ -------- --- ---------- -------- -------- - ------ ----- ... ... tRipTask ripTask 1dc04c0 100 PEND 1859d0 1dc0328 46 0 value = 0 = 0x0 - -> From looking at "errno.h" it says: #define EPFNOSUPPORT 46 /* Protocol family not supported */ meaning that RIP isn't supported (at least I think so?). Now, I talked to the guys at Radstone and they claim that RIP should indeed be supported with the END driver, so does anyone understand why I'm getting this error? Thank you, John Sestak - --------------FA15A376FD9DEF61A3A6AE77 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hello again,
    I'm currently having a problem with getting RIP (Routing Information Protocol) to work.  I'm running a Radstone PPC4-750 with their version 2.1 BSP.  When I include RIP using the default BSD network driver there doesn't seem to be any problem, but when I switch to the END (Enhanced Network Driver - for it's multicast support) driver and try and include RIP, I get the following error when I boot and press "i":

-> i
  NAME        ENTRY       TID    PRI   STATUS      PC       SP     ERRNO  DELAY
---------- ------------ -------- --- ---------- -------- -------- - ------  -----
...
...
tRipTask   ripTask       1dc04c0 100 PEND         1859d0  1dc0328      46     0
value = 0 = 0x0
->

From looking at "errno.h" it says:

#define EPFNOSUPPORT 46  /* Protocol family not supported */

meaning that RIP isn't supported (at least I think so?).  Now, I talked to the guys at Radstone and they claim that RIP should indeed be supported with the END driver, so does anyone understand why I'm getting this error?

Thank you,
John Sestak - --------------FA15A376FD9DEF61A3A6AE77-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: tempnam ? Date: Tue, 29 Feb 2000 14:31:06 -0800 From: "Mitch Stermer" Message-ID: <38bc48ab$0$211@nntp1.ba.best.com> References: <899hmg$6da$1@newsmaster01.nextra.at> "Mario Semo" wrote in message news:899hmg$6da$1@newsmaster01.nextra.at... > hello, > > i need a way to create a tempory filename / temporary file. i always use > tempnam() for this, but tempnam is just in one include i dont knoe in > vxworks, and it is not documented > > K:\LC32\L\RT\OOS\c\LCoosPersistenceFileSystem.cpp(304): warning: implicit > declaration of function `int tempnam(...)' > F:\Tornado>find32 tempnam *.h > > "F:\Tornado\target\h\private\stdioP.h" > "F:\Tornado\host\x86-win32\i386-pc-mingw32\sys-include\stdio.h" > > F:\Tornado>find32 stdiop.h *.h > > "F:\Tornado\target\h\private\stdioP.h" > > what is this include ditectory. > > any other idea how to create a temporary filename? > > -- > regards, Mario Semo. http://www.kirchnersoft.com > > /* tempnam emulation for vXworks. */ #include #include char *tempnam(const char *dir, const char *pfx) { int size; char *szbuf; char * tmp = tmpnam( NULL ); size = strlen( dir ) + strlen( pfx ) + strlen( tmp ) + 5; szbuf = malloc( size ); if ( NULL == szbuf ) { return NULL; } strcpy(szbuf, dir); strcat(szbuf, pfx); strcat(szbuf, tmp); return szbuf; } --------------------------- Newsgroups: comp.os.vxworks Subject: non-blocking pipes Date: Tue, 29 Feb 2000 14:42:37 -0800 From: "Mitch Stermer" Message-ID: <38bc4b5d$0$206@nntp1.ba.best.com> This is a multi-part message in MIME format. - ------=_NextPart_000_0056_01BF82C3.38B73E70 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I am porting a very large Unix application to VxWorks. I cannot change = the structure of the code so I am forced into using a select call.=20 The code used non-blocking fifo's on Unix and I am attempting to use = pipeDevCreate as the fd's for the select. What I would like is to something like a=20 fcntl(fd, F_SETFL, _FNONBLOCK );=20 on the fd of the created pipe. Is there such a way to set this type of = option. Thanks. - ------=_NextPart_000_0056_01BF82C3.38B73E70 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

I am porting a very large Unix application to VxWorks. I cannot = change the=20 structure of

the code so I am forced into using a select call.

The code used non-blocking fifo's on Unix and I am attempting to use=20 pipeDevCreate

as the fd's for the select. What I would like is to something like a =

fcntl(fd, F_SETFL, _FNONBLOCK );

on the fd of the created pipe. Is there such a way to set this type = of=20 option.

Thanks.

- ------=_NextPart_000_0056_01BF82C3.38B73E70-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Windows Managers for VxWorks Date: Tue, 29 Feb 2000 23:21:57 -0000 From: "Matthew Barratt" Message-ID: <951866527.4204.0.nnrp-08.9e987a34@news.demon.co.uk> References: <38bb67c0.135161586@news.easynet.co.uk> The motif package for vxwindows (rtx windows now) came with the motif window manager. Windriver would probably want money for it though. I suspect that given some time and effort one of the window managers for Linux could possibly be compiled up. Matthew Barratt "Philippe" wrote in message news:38bb67c0.135161586@news.easynet.co.uk... > > Is anyone aware of any Windows Managers (X-Windows) supported by the > VxWorks Platform ? > > Thanks > Phillipe --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Stethoscope 5.0 Date: Wed, 01 Mar 2000 03:16:04 GMT From: MJM Organization: mjm et al Message-ID: <38BC89CF.54B62979@sympatico.ca> References: <38BB3124.2A45485F@sympatico.ca> <89grpc$vc7$1@nnrp1.deja.com> Reply-To: mike.mcnaughton@sympatico.ca Ted: Thanks for the response: I have tornado running nicely and stethoscope even runs. It asserts a message on a status line at the bottm of the screen that says somehtin like "can't connect to server" The problem also seems to be in getting the symbol library linked or it could be in the server agent. I'm using the host based target server. The manual goes along great but flames out when it gets to the part where it talks about the target server-server. It also talks about sym_table_synch. When I configure SYM_TABLE_SYNCH into my build the result at run time is something imaginative and informative like "Unable to link symbol table". It might be looking for the symbol table and not being able to find it. A clue might be in the shell behavior. When I try something like > printf("test message") , I get an error message like AGENT_NO_AGENT_PROC. My main run time is built right ionto the vxWorks object. Evidence is tha the shell can't see even th globals in my vxWorks module. I could use some assistance. thanks MJM Ted Dennison wrote: > In article <38BB3124.2A45485F@sympatico.ca>, > mike.mcnaughton@sympatico.ca wrote: > > I have purchased Steth 5.0 and wish to install it tornado 1.01, > > vxworks 5.3.1. > > > > So far it has done nothing but frustrate the Hell out of any attempts > > to install it.. > > > > Is there someone among you who wouln't mind taking a few minutes to > > outline a detailed procedure of what you had to do to get it going. > > and helped me out with this in a manner which led to my getting it up > > and running, I would buy you a 24-pack of beer. > Hmm. You aren't in Oklahoma perchance? :-) > > We've had little problem with developers installing Stethescope > themselves and getting it to work. The one that's a bear is Tornado. Are > you sure you are getting it installed properly? Can you build BSP's, for > instance? > > -- > T.E.D. > > http://www.telepath.com/~dennison/Ted/TED.html > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: memory leak deleting a pipe Date: Wed, 01 Mar 2000 04:39:53 GMT From: "Will Darby" Organization: bCandid - Powering the world's discussions - http://bCandid.com Message-ID: References: We searched far and wide for an answer to this problem, to no avail. Because we frequently create/destroy pipes, we could not use the simple vxWorks supplied pipes. However, we didn't really need named pipes, and so we ended up writing our own simple driver that supported delete. Will Darby Xiox Corp. wdarby@xiox.com "Markus MItterer" wrote in message news:AQ4t4.2$VL1.280@nreader3.kpnqwest.net... > Hi VxWorkers, > In my code I create a pipe-device with pipeDevCreate(). When I don`t need > the pipe any more I want to delete it. For this purpose I work with the > functions iosDevFind() and iosDevDelete(). After calling this functions the > device was deleted from system, but now I have a memory leak, which was > created by pipeDevCreate(). > How can I delete the pipe completly or what is the workaround? > I can`t find a sence why there is no delete function for a pipe. > > Thank you in advance, > Markus Mitterer, STK > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Drivers & signals Date: Wed, 01 Mar 2000 05:28:43 GMT From: "Will Darby" Organization: bCandid - Powering the world's discussions - http://bCandid.com Message-ID: Question on correct interaction between signals & drivers. I have an application that uses signals & select, with a simple "trigger" driver, which basically acts like a selectable semaphore. The way it is implemented, I can lock up the task entering a select call when it receives a signal during the ioctl handling of the trigger fd as it enters the select. In other words, if a task entering select gets a signal that causes an operation on the same fd that is currently being manipulated by the select, then a semaphore protecting the fd's private data prevents the signal from completing, but the task cannot continue because it is handling a signal. Here's some pseudo code. int triggerIoctl( internals, function ) { semTake( internals->lock, WAIT_FOREVER ); switch( function ) { case FIOSELECT: /* Handle select code */ ... } semGive( internals->lock ); } int triggerWrite( internals, buffer, bufSize ) { semTake( internals->lock, WAIT_FOREVER ); /* Handle write */ ... /* Determine if select needs to be woken up */ if( lstCount( &internals->buffer ) == 1 ) /* Wake up pending tasks on select & give the empty lock */ selWakeupAll( &internals->selWakeupList, SELREAD ); semGive( internals->lock ); } So a task enters this ioctl during a select call on the specified descriptor. A signal handler within the task causes the write to occur while the task is within the locked region in triggerIoctl. The write then blocks at the initial semTake, preventing the signal handler from running, and the task cannot resume until the signal handler is complete. While some of these details are construed to illustrate the problem, (1) is there a general mechanism for appropriately dealing with signals from within a driver? (2) I could avoid the use of the write operation all together if the signal would force the select statement to return EINTR, is there a way to force this standard behavior? (3) Are there standard approaches to returning EINTR from system calls? (4) Are there more precise guard mechanisms I can use from within a driver? Thanks for any advice. Will Darby Xiox Corp. wdarby@xiox.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STD_IN failing on target shell Date: Tue, 29 Feb 2000 08:13:09 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89iees$i6f$1@newsmaster01.nextra.at> References: can youi give a small sample? i always use getchar() (target: VxSim / VxWorks-X86) . i start my app never via Tornado. - -- regards, Mario Semo. http://www.kirchnersoft.com Øyvind Holmeide wrote in message ... >I have an application that is based on input/output through the target shell >(INCLUDE_SHELL is defined). Output to the shell (STD_OUT) is OK, but input >(STD_IN) is failing. The target shell gives the response "undefined symbol" >for all inputs. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Embedded int 3 handing with vxWorks. Date: Tue, 29 Feb 2000 08:26:17 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89ief0$i6f$3@newsmaster01.nextra.at> References: can you post how you got iot running? - -- regards, Mario Semo. http://www.kirchnersoft.com Burt Bicksler wrote in message ... >Hi, > >I've been tasked with finding a way to invoke / break into the debugger >conditionally from a running vxWorks application (vxWorks target). > >I have a kludge that works, but wondered if anyone had ever gotten an >embedded int 3 to break into the debugger with vxWorks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: builtin_new/builtin_delete Date: Tue, 29 Feb 2000 08:25:06 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89ieev$i6f$2@newsmaster01.nextra.at> References: <38BAA71C.3C96@fzi.de> this is a well known Vxworks bug. as we run into it, we searched Deja and we found the following 3 msg which even contains the source code. - -- regards, Mario Semo. http://www.kirchnersoft.com Torsten Rupp wrote in message <38BAA71C.3C96@fzi.de>... >Hi, > >currently I try to load some program code including C++ new()/delete() >calls on a MV2400/PPC 750 board. I have a problem with the functions >__builtin_new and __builtin_delete, which seems not to be included begin 666 1.txt M1G)O;3H@=&AO;6%S+F)A:65R0&MI6]U(&9R;VT@1&5J82YC;VT@8GD@=&AO;6%S+F)A:65R0&MI6]U M(&)U>2P@=FES:70@#0IH='1P.B\O=W=W+F1E:F$N8V]M+R4U8E-47W)N)3-D M<',E-60O#0HM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM M+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2TM+2T-"@T* M*&)E9VEN;FEN9R!O9B!O'=O2!W6]U('=A;G0@=&\@879O M:60@=&AA="!T:&5S92!T:&EN9W,@9G)O;2!L:6)G8V,@9V5T(&QI;FME9"!I M;G1O#0IY;W5R(&%P<&QI8V%T:6]N+"!R97!L86-E('1H96T@8GD@>6]UPT*(" @(" @=F]I9" J<#L-"B @(" @#0H@(" @(" O M*B!M86QL;V,@*# I(&ES('5N<')E9&EC=&%B;&4[(&%V;VED(&ET+B @*B\- M"B @(" @(&EF("AS>B ]/2 P*0T*"7-Z(#T@,3L-"B @(" @(' @/2 H=F]I M9" J*2!M86QL;V,@*'-Z*3L-"B @(" @(&EF("AP(#T](# I#0H@(" @("![ M#0H);&]G37-G*$U%4U-!1T4L(# L(# L(# L(# L(# L(# I.PT*"5]E>&ET M*"TQ*3L-"B @(" @('T-"B @(" @(')E='5R;B!P.PT*(" @('T-"B @(" @ M#0H@(" @=F]I9" J#0H@(" @7U]B=6EL=&EN7W9E8U]N97<@*&EN="!S>BD- M"B @("![#0H@(" @(" @2YX<#]S96%R8V@] M=&AR96%D)G-V8V-L87-S/61N6]U(&1O;B=T(&1O M(&$@:6YC6]U('=I;&P@ M#0IH879E('1O(")M=6YC:"(@=&AI;F=S+"!T:&%T(&=L=64@8V]D92!F;W(@ M9VQO8F%L(&-O;G-T2!G96YE7EY>7EY>7EY>7EY>7EY>7DL#0H^#0H^5V4@8V%N;F]T(&=E="!T M:&4@9F]L;&]W:6YG('-I;7!L92!C;V1E("AT='0N8V,I('1O(&QO860@;VX@ M82 V-$U"(#(W,# -"CYB;V%R9"X@(%=E#0H^9V5T('1H92!I;F9A;6]U7!E(# I#0H^#0H^*$D@=&AO=6=H="!T:&ES('=A M6]N92!K;F]W(&$@%=OWT[#0H^?3L-"CX-"CYV;VED('1T="AI;G0@;BD-"CY[#0H^("!-05)+ M7T-,05-3("IM.PT*/B @;2 ](&YE=R!-05)+7T-,05-3*"D[#0H^#0H^?0T* M/CT]/3T]/3T]/3T]/3T](&5N9"!O9B!T='0N8V,-"CX-"CXJ*BHJ*BHJ*BH@ M0V]M<&EL871I;VX@;65T:&]D.@T*/B]U6=N=7,M,BXW+C(M.38P M,3(V(&5G8W,M.36=N=7,M,BXW+C(M.38P,3(V#0H^#0H^#0H^#0H^#0H^#0H- M"@T*+2T@#0H-"B @("!!2!T:&]M87,N8F%I97) :VER8VAN97)S;V9T M+F-O;2X-"D1E:F$N8V]M(&]F9F5R2(@/&-H87)L97,N8W5R;&5Y0&]R M:6]N+FMS8RYN87-A+F=O=CX-"D1A=&4Z(#$Y.3DO,#,O,3 -"DYE=W-G6=N=7,M,BXW+C(M.38P M,3(V+"!0;W=E2!G M:79E#0IE>'!L86YA=&EO;B!A"!T:&4@<')O8FQE;2X-"@T* M3W5R('-Y2!T:&%T(&=I=F5S('1H92!E2P@4V5N:6]R(%-O9G1W87)E($5N9VEN965R+"!$>6YA8W,@16YG M:6YE97)I;F<@0V\N+"!);F,N#0I%;F=I;F5E6]U('-O;VX@870@1&5J82YC M;VTN#0I"969O6]U(&)U>2X-"FAT=' Z+R]W=W Organization: Telia Internet Public Access Message-ID: <6R6v4.1505$F3.181570048@news.telia.no> References: <89iees$i6f$1@newsmaster01.nextra.at> I have now added the symbol table also for the target (sync host and target symbo tables + symbol table initialization components). The target shell has now all the support that is found in the corresponding Tornado shell. If my program that is based on STDIN from the console is started from the target shell, then STDIN is received through e.g. fioRdString(), and everything is fine. However, if the application is spawned then STDIN is not available regardless of any calls to ioGlobalStdGet()/ioGlobalStdSet() or ioTaskStdGet()/ioTaskStdSet(). Mario Semo wrote in message news:89iees$i6f$1@newsmaster01.nextra.at... > can youi give a small sample? i always use getchar() (target: VxSim / > VxWorks-X86) . i start my app never via Tornado. > > -- > regards, Mario Semo. http://www.kirchnersoft.com > Øyvind Holmeide wrote in message ... > >I have an application that is based on input/output through the target > shell > >(INCLUDE_SHELL is defined). Output to the shell (STD_OUT) is OK, but input > >(STD_IN) is failing. The target shell gives the response "undefined > symbol" > >for all inputs. > > > --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Mar 1 05:56:44 2000 From: jvlao@lucent.com Date: Wed Mar 1 05:56:46 PST 2000 Subject: vxwork 5.3.1 windriver:2.5 write function corrupt memory Hi all, I have a problem. When vxWorks's write function is called and the socket file descriptor no longer exist, it corrupted memory. The socket file descriptor was derived using the socket function. After it have successfully open, during a telnet session while using the sun workstation to open up telenet session, the code was able to write and read to the session. However, when the sun workstation's window is closed, the moment vxworks write to the socket file descriptor, it randomly corrupted the memory. According to the vxWork write manual, it should return ERROR when file descriptor no longer exit. But instead, it never return. We are using vxWorks 5.3.1 with the kernel windriver: 2.5. Any idea you could point me, I would great appreciated. Thanks, James From vxwexplo-errs@csg.lbl.gov Wed Mar 1 07:59:37 2000 From: John_W_Cosgrove@res.raytheon.com Date: Wed Mar 1 07:59:40 PST 2000 Subject: booting from a SCSI harddrive Hello, Folks, I am trying to boot my mvme2700 board from a Seagate ST32430N (2 gig SCSI) hard drive. I am using vxWorks 5.3.1, BSP 1.1/4.2 (The Motorola patch to the mv2604 BSP to make a 2700 BSP). I am encountering boot failures often with occasional success at the booting. When it fails it looks like: Press any key to stop auto-boot... 0 auto-booting... boot device : scsi processor number : 0 host name : emperor file name : /sd0/vxworks inet on ethernet (e) : 192.168.56.113 host inet (h) : 138.125.194.121 gateway inet (g) : 192.168.56.1 user (u) : jwc flags (f) : 0x8 target name (tn) : f100-hp-ray other (o) : dc ID LUN VendorID ProductID Rev. Type Blocks BlkSize pScsiPhysDev -- --- -------- ---------------- ---- ---- -------- ------- ------------ 0 0 SEAGATE ST32430N 0640 0 4197405 512 0x01ffdcb0 Can't load boot file!! and if it does "attaching to SCSI", everything is fine. The boot device of scsi or scsi=0,0 both yeild this result! Using the boot device of scsi sometimes the scsiConfig output is filled in, sometimes it is not. Has anyone had a similar problem, or does anyone know of a fix? John Cosgrove (401)842-4167 (desk) (401)842-5740 (lab) From vxwexplo-errs@csg.lbl.gov Wed Mar 1 11:24:28 2000 From: John_W_Cosgrove@res.raytheon.com Date: Wed Mar 1 11:24:31 PST 2000 Subject: Re: booting from a SCSI harddrive I hate followups to my own messages, but I have found out more. In scsiLoad(), I found that I reach the first scsiTestUnitRdy() call on powerup. On the first warm start (^X or reset button), the system now boots fine. On the second warm start (no matter whether ^X or reset button) the system reaches the scsiPhysDevCreate() call (for scsiTestUnitRdy() call) and stops... Does any of this sound familiar to anyone? John Cosgrove (401)842-4167 (desk) (401)842-5740 (lab) John Cosgrove writes: Hello, Folks, I am trying to boot my mvme2700 board from a Seagate ST32430N (2 gig SCSI) hard drive. I am using vxWorks 5.3.1, BSP 1.1/4.2 (The Motorola patch to the mv2604 BSP to make a 2700 BSP). I am encountering boot failures often with occasional success at the booting. When it fails it looks like: Press any key to stop auto-boot... 0 auto-booting... boot device : scsi processor number : 0 host name : emperor file name : /sd0/vxworks inet on ethernet (e) : 192.168.56.113 host inet (h) : 138.125.194.121 gateway inet (g) : 192.168.56.1 user (u) : jwc flags (f) : 0x8 target name (tn) : f100-hp-ray other (o) : dc ID LUN VendorID ProductID Rev. Type Blocks BlkSize pScsiPhysDev -- --- -------- ---------------- ---- ---- -------- ------- ------------ 0 0 SEAGATE ST32430N 0640 0 4197405 512 0x01ffdcb0 Can't load boot file!! and if it does "attaching to SCSI", everything is fine. The boot device of scsi or scsi=0,0 both yeild this result! Using the boot device of scsi sometimes the scsiConfig output is filled in, sometimes it is not. Has anyone had a similar problem, or does anyone know of a fix? John Cosgrove (401)842-4167 (desk) (401)842-5740 (lab) From vxwexplo-errs@csg.lbl.gov Wed Mar 1 15:09:32 2000 From: Navin Kumar Agrawal Date: Wed Mar 1 15:09:34 PST 2000 Subject: Raw Socket on Vxworks Tornado Hi Folks, I would really appreciate if somebody can help me in solving a problem of mine which is related with the usage of Raw Socket on VxWorks. I have put my effort in the following way. -------------------------------------------------------- struct sockaddr_in serveraddr; bzero ((char *)&serveraddr, sizeof(struct sockaddr_in)); serveraddr.sin_family = AF_INET; serveraddr.sin_len = sizeof(struct sockaddr_in); serveraddr.sin_port = htons (6007); serveraddr.sin_addr.s_addr = htonl (inet_addr ("192.168.100.140")); if ((fd = socket (AF_INET, SOCK_RAW, IPPROTO_ICMP)) == ERROR) { perror ("socket"); return -1; } if (bind (fd, (struct sockaddr *)&serveraddr, sizeof(struct sockaddr_in)) == ERROR) { perror ("bind failed"); return -1; } -------------------------------------------------------- My target system is based on PPC604 (PowerPC). On executing the above program, it is failing at bind () and the error code returned is EADDRNOTAVAIL ("Can't assign requested address"). Can anybody help me in sovling this problem? Regards, Navin From vxwexplo-errs@csg.lbl.gov Thu Mar 2 01:08:39 2000 From: Rainer Stelzer Date: Thu Mar 2 01:08:42 PST 2000 Subject: How to speed up netDev in FTP-mode ? Hi vxworkers, when our application is started, we begin with parsing about 40 files to get runtime parameters for our control system. Some of our control systems do not have a local filesystem, so we use a netDev to access the parameter files over network (FTP-mode). Reading the files this way, is very much slower than accessing a local filesystem. (800ms vs.45s !!!, We have up to seven files opened. We don't use the seek() command ) The netDev library seems to be implemented in a way, that it consumes only a small memory area. (and this is a good solution for some of our systems which are low in memory) causing the disadvantage that it has to use the the FTP server very often. Any hint to speed up the transfer (without changing our parser)? thanks in advance regards rainer vxworks, Tornado II ===================================================== Gebrüder Stoye GmbH department: traffic control systems Name: Rainer Stelzer Adr: Longericher Strasse 177 city: Cologne, Germany Zip: 50739 phone: ++49/221/2616-551 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== From vxwexplo-errs@csg.lbl.gov Thu Mar 2 01:08:39 2000 From: Rainer Stelzer Date: Thu Mar 2 01:08:43 PST 2000 Subject: SmartMedia driver for vxworks ? Hi vxworkers, can somebody give me hint where to find (free or commercial) SmartMedia driver software for vxworks ? Is my assumption correct, that I only have to write a low level driver to fit the vxworks BlockDev driver (to handle the SmartCard over vxworks MSDOS Filesystem ) ? Are the control commands in the low level driver ATA commands ? The problem must already be solved, because there are digital cameras running vxworks equiped with SmartMedia data storage. Any hint is welcome regards rainer vxworks, Tornado II ===================================================== Gebrüder Stoye GmbH department: traffic control systems Name: Rainer Stelzer Adr: Longericher Strasse 177 city: Cologne, Germany Zip: 50739 phone: ++49/221/2616-551 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== From vxwexplo-errs@csg.lbl.gov Thu Mar 2 04:03:10 2000 From: Vxworks Exploder Date: Thu Mar 2 04:03:13 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Mar 2 04:03:07 PST 2000 Subject: Re: STD_IN failing on target shell Subject: Re: gcc 2.7.2 and C++ Template Classes - NFG! Subject: How do I reduce data sizes in flash? Subject: Re: C++ RTTI Overhead Subject: Tornado 2 and STL Subject: Tornado host<->target via terminal concentrator Subject: Re: Tornado host<->target via terminal concentrator Subject: Re: Stethoscope 5.0 Subject: Get Socket Send Buffer Size?l Subject: ¿ESTÁS BUSCANDO TRABAJO? Subject: Re: Tornado 2 and STL Subject: Re: Is munchless C++ linking possible Subject: VxFusion Subject: local filesystem/ change drive / ... Subject: Changing usrNetwork.c Subject: Re: How do I reduce data sizes in flash? Subject: Re: Stethoscope 5.0 Subject: Re: Changing usrNetwork.c Subject: Re: Changing usrNetwork.c Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Tornado 2 and STL ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: STD_IN failing on target shell Date: Wed, 01 Mar 2000 12:23:23 GMT From: Thomas Rahn Organization: Deja.com Message-ID: <89j23p$gn7$1@nnrp1.deja.com> References: <89iees$i6f$1@newsmaster01.nextra.at> <6R6v4.1505$F3.181570048@news.telia.no> In article <6R6v4.1505$F3.181570048@news.telia.no>, "Øyvind Holmeide" wrote: > I have now added the symbol table also for the target (sync host and target > symbo tables + symbol table initialization components). The target shell > has now all the support that is found in the corresponding Tornado shell. > > If my program that is based on STDIN from the console is started from the > target shell, then STDIN is received through e.g. fioRdString(), and > everything is fine. However, if the application is spawned then STDIN is > not available regardless of any calls to ioGlobalStdGet()/ioGlobalStdSet() > or ioTaskStdGet()/ioTaskStdSet(). The reason might be that the shell is still running (at a very high priority). Your application will be spawned at prio 100 by default, so the shell gets all the input.... Thomas Rahn Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: gcc 2.7.2 and C++ Template Classes - NFG! Date: 1 Mar 2000 13:56:58 +0100 From: hartmutp@bidro145.bbn.hp.com (Hartmut Pohl) Message-ID: <38bd139a_1@isoit370.bbn.hp.com> References: <8676lu$a53$1@sulawesi.lerc.nasa.gov> <38b41715_1@isoit370.bbn.hp.com> In article <38b41715_1@isoit370.bbn.hp.com>, hartmutp@bidro145.bbn.hp.com (Hartmut Pohl) writes: > [Posted and mailed] > > > Dave, > > we've had the same problem with our old g++ compiler. It seems > like you have the same version of '96 that comes with Tornado 1.0 > > We claimed the problem at Wind River and got a patch. With the > updgrade to Tornado 1.01 we got a newer g++ (gcc version 2.7.2.3) > also without the template problem. If you can afford it, switch > to Tornado 1.01. Sorry, I've provded you with the wrong information. Actually, we have the same compiler version as you (gcc version cygnus-2.7.2-960126) but with a patch from WRS. I looked back to the TSR, and the problem has been something with a weak symbol generation of the GNU toolchain. I tried your piece of code with the patched i86 compiler and it worked fine. Trying the same with the not patched 68k compiler produced exactly the same problem you reported. Thus it would be a good idea to contact WRS to get the same patch. Our TSR numer for this issue has been TSR 59034 (back in 1997) Hope this helps, Hartmut - -- Hartmut_Pohl@agilent.com - -- Die Ehe ist beliebt, weil sie ein Hoechstmass an Versuchung mit einem Hoechstmass an Gelegenheit verbindet. George Bernard Shaw --------------------------- Newsgroups: comp.os.vxworks Subject: How do I reduce data sizes in flash? Date: Wed, 01 Mar 2000 13:09:42 GMT From: steve.folly@rdel.co.uk (Steve Folly) Organization: RDEL Message-ID: <38bd1465.418562180@news.rrds.co.uk> Hi, I hope someone can provide some help on the following problem: We have a program running on a DY4 PowerPC 603e with VxWorks 5.3. The program is in Ada95 and compiled with Rational ApexWorks 3.0.0b There are currently some very large static data structures (up to about 3Mb) which, when the program is linked and programmed into flash take up about 3Mb of flash! This seems rather wasteful as the initialization could be done programmatically and the data structures dynamically allocated at run time. Well, I had a go at doing this but the resulting executable and S records file came out at the same size. Curious thing is, according to the link map the structures in the program_bss section had disappeared. Should I expect the size of the executable or S records file to have reduced? Or am I missing something here? Are there other ways to reduce flash usage? BTW, We are also looking at compressing the flash image. Thanks for any help. - -- Regards, Steve Folly, Software Engineer, Racal Defence Electronics Ltd. mailto:steve.folly@rdel.co.uk http://www.rdel.co.uk/ +44 (0)1293 528787 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++ RTTI Overhead Date: Wed, 01 Mar 2000 15:26:30 +0100 From: Artem Alimarine Organization: Philips Medical Systems Message-ID: <38BD2896.CB28FFC9@best.ms.philips.com> References: <893uhb$4k8$1@nnrp1.deja.com> This is a multi-part message in MIME format. - --------------510E59F51F12816D88C06595 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi normpgh@home.com wrote: > Hi, > > I was wondering if anyone out there has been using Runtime Type > Information (RTTI) in a VxWorks C++ application. My question relates > to the amount of overhead that is incurred by turning this feature > on. > > I beleive that just enabling RTTI causes additional overhead within > the construction of each object in the system. And that there is > additional overhead if/when this information is actually used. But > I have no way to quantify any performance degradation. I do not think there is an overhead on creation of objects because the RTTI information is normally stored in the virtual method table, which is created at compile time. When an object of class A is created a vmt pointer in the object is assigned to vmt of class. This is the run-time effect that does not depend on RTTI. Sorry, I do not have any performance figures. > > > If anyone has experience using RTTI under VxWorks, I would be very > interested in hearing your opinion. Thanks in advance for any info > you might provide. > > - NormPgh > > Sent via Deja.com http://www.deja.com/ > Before you buy. Artem - --------------510E59F51F12816D88C06595 Content-Type: text/x-vcard; charset=us-ascii; name="aalimari.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Artem Alimarine Content-Disposition: attachment; filename="aalimari.vcf" begin:vcard n:Alimarine;Artem tel;home:(+31) 40-2440633 tel;work:(+31) 40-2763078 x-mozilla-html:TRUE adr:;;;;;; version:2.1 email;internet:aalimari@best.ms.philips.com fn:Artem Alimarine end:vcard - --------------510E59F51F12816D88C06595-- --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 2 and STL Date: Wed, 01 Mar 2000 15:39:22 +0100 From: Artem Alimarine Organization: Philips Medical Systems Message-ID: <38BD2B9A.76F08025@best.ms.philips.com> This is a multi-part message in MIME format. - --------------DE3F3C1D650779C8BD7B3016 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello I am trying to compile some software written in C++ for on Tornado2. cc386 --version reports: cygnus-2.7.2-960126 egcs-971225 tornado 2.0 The problem is that it is two old C++ compiler. It reports errors on nromal usage of templates and STL stuff. Is there a newer Tornado 2 compiler (for pentium and i960)? Any help would be appreciated, thanks in advance Artem - --------------DE3F3C1D650779C8BD7B3016 Content-Type: text/x-vcard; charset=us-ascii; name="aalimari.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Artem Alimarine Content-Disposition: attachment; filename="aalimari.vcf" begin:vcard n:Alimarine;Artem tel;home:(+31) 40-2440633 tel;work:(+31) 40-2763078 x-mozilla-html:TRUE adr:;;;;;; version:2.1 email;internet:aalimari@best.ms.philips.com fn:Artem Alimarine end:vcard - --------------DE3F3C1D650779C8BD7B3016-- --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado host<->target via terminal concentrator Date: Wed, 01 Mar 2000 16:44:14 +0100 From: Steve McKinty - SunSoft ICNC Grenoble Organization: SunSoft Networking Products Message-ID: <38BD3ACE.F68A933E@france.sun.com> Reply-To: smckinty@france I'm using Tornado 2 and VxWorks 5.4, on a Solaris 2 system. The target is a Force PMC860. This works fine when I use the wdbserial target server and a direct serial link between the host and target. This isn't a convenient solution, and I would prefer to link to the target via a Bay Networks Annex 3 terminal server. The terminal server allws the use of Reverse Telnet to link a pseudo-terminal on the Solaris system to a serial port on the Concetrator. When I try this I can see that data is being exchanged, the tgtsvr process sends data to the target and the target responds. Unfortunately the tgtsvr process simply ignores the response, and continues polling. Can anyone suggest any way that I can use the terminal server, or any special parameter settingsthat I need? failing that, how can I debug the wdbserial tgtsvr process, to find out why it is ignoring the data? Pointers to code, FAQs, etc. gratefully received. - -- Steve McKinty | Sun Microsystems ICNC | 38240 Meylan, France | Unsolicited e-mail advertising email: smckinty (AT) france.sun.com | is not welcome. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado host<->target via terminal concentrator Date: Wed, 01 Mar 2000 11:41:52 -0500 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <38BD4850.724A52A2@lucent.com> References: <38BD3ACE.F68A933E@france.sun.com> Steve, We used these same severs (Annex 3) to concentrate our target shell connections, but not for wdbserial. I would check that you aren't suffering some character mapping in the server, NL -> CRNL, stuff like that. The port would have to operate in raw mode. I do know that I once had a problem on an earlier Annex device that would not allow raw settings. Even though the ioctl() in SunOS returned with no errors, the actual settings at the Annex port did not reflect the request. You may be encountering a similar problem. I would look there first. Sorry about editing the original post, but my NNTP service only allows replies that are bigger than the quoted text. Good look. Doug Steve McKinty - SunSoft ICNC Grenoble wrote: <..> > This works fine when I use the wdbserial target server and a direct > serial link between the host and target. This isn't a convenient > solution, and I would prefer to link to the target via a Bay Networks > Annex 3 terminal server. > > The terminal server allws the use of Reverse Telnet to link a > pseudo-terminal on the Solaris system to a serial port on the > Concetrator. When I try this I can see that data is being exchanged, > the tgtsvr process sends data to the target and the target responds. > > Unfortunately the tgtsvr process simply ignores the response, and > continues polling. > <..> --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Stethoscope 5.0 Date: Wed, 01 Mar 2000 16:56:50 GMT From: Ted Dennison Organization: Deja.com - Before you buy. Message-ID: <89ji4e$s9v$1@nnrp1.deja.com> References: <38BB3124.2A45485F@sympatico.ca> <89grpc$vc7$1@nnrp1.deja.com> <38BC89CF.54B62979@sympatico.ca> In article <38BC89CF.54B62979@sympatico.ca>, mike.mcnaughton@sympatico.ca wrote: > I have tornado running nicely and stethoscope even runs. It asserts a > message on a status line at the bottm of the screen that says somehtin like > "can't connect to server" > A clue might be in the shell behavior. When I try something like > > printf("test message") , > I get an error message like AGENT_NO_AGENT_PROC. When I connect my target server in Tornado there's a line that says: Connecting to target agent... succeeded. The terminology is suspiciously similar. You are running a target server, right? What does that line say on yours? - -- T.E.D. http://www.telepath.com/~dennison/Ted/TED.html Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Get Socket Send Buffer Size?l Date: Wed, 01 Mar 2000 18:23:24 GMT From: "Michael Murphy" Organization: bCandid - Powering the world's discussions - http://bCandid.com Message-ID: Hi All, I'm send large buffers via TCP using send(). After a few MB or so, send will return an error 0x20. I'm assuming this is a Send Buffer Full Error. Does anyone know how to peek into send buffer to get its current size? All calls to send() afterwards return an error 0x23. Anyone know where I can get info on these error codes? Thanx, Michael --------------------------- Newsgroups: comp.os.vxworks Subject: ¿ESTÁS BUSCANDO TRABAJO? Date: Wed, 01 Mar 2000 19:00:56 GMT From: "Jose Angel Marin Losantos" Organization: Clientes_Teleline Message-ID: Si estás buscando trabajo te ofrezco la posibilidad de conseguir información sobre más de 370 empresas o personas que ofrecen trabajo en casa. Si tienes titulación, información sobre más de 240 grandes empresas a nivel nacional que están en periodo de selección. Si estas pensando en establecerte por tu cuenta como trabajador autónomo, información sobre más de 350 empresas franquiciadoras. Para ampliar información escribe una carta a: José Angel Marín Losantos C/Concepción Nº 8, 4º Ida 50.008 Zaragoza También me puedes enviar tu dirección postal por correo electrónico a la siguiente dirección MARINN@teleline.es NOTA: debido al volumen de la información, la misma no se envía por correo electrónico. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 2 and STL Date: Wed, 1 Mar 2000 11:36:13 -0800 From: "Dave Craig" Organization: Verio Message-ID: References: <38BD2B9A.76F08025@best.ms.philips.com> Hi Artem, There are a lot of problems to overcome with templates, STL and vxWorks. However, almost all of these come at link time and not at compile time (there is an -frepo patch which fixes some template instantiation problems - look on Windsurf for this). What are the compilation errors you are seeing? Dave Craig Palo Alto, CA. "Artem Alimarine" wrote in message news:38BD2B9A.76F08025@best.ms.philips.com... > Hello > > I am trying to compile some software written in C++ for on Tornado2. > cc386 --version reports: > cygnus-2.7.2-960126 egcs-971225 tornado 2.0 > > The problem is that it is two old C++ compiler. It reports errors on > nromal usage of templates and STL stuff. > > Is there a newer Tornado 2 compiler (for pentium and i960)? > > Any help would be appreciated, thanks in advance > Artem > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is munchless C++ linking possible Date: Wed, 01 Mar 2000 12:41:12 -0700 From: Doug Gibbons Organization: Lucent Technologies Message-ID: <38BD7258.22685AD@lucent.com> References: <891a6b$hiq$1@bcrkh13.ca.nortel.com> Jim Nance wrote: > > Hello All, > I am working on a VxWorks based project that is written in C++. Part of > the build process involves partially linking the executable and running a > tcl script called munch that takes the symbols table for this executable > and generating global constructor calls. This is slow, and I was wondering > if there was another way to link things that did not require running munch? > > Thanks, > > Jim Back in "the old days" when I was loosely connected with the C++ efforts at AT&T, the Cfront C++ translator evolved two postprocessing strategies to handle this. They were known as "munch" and "patch." In both cases, post-processing/editing on the a.out arranged for the invocation of functions (known then as "stc" and "std") functions which would construct/destroy the static objects defined in a particular translation unit (read "dot-oh"). When UNIX System V Release 4 (SVR4) came along with ELF a.outs, another method became possible which eliminated post-processing. The UNIX SVR4 scheme was to coax the compiler to drop calls to the "stc" and "std" functions into the .init and .fini sections of the ELF a.out. Since a UNIX SVR4 process executes the contents of .init before anything else, constructors for static objects were called in advance of main(). The .fini section was handled similarly on process exit, and calls to destructors would occur after main() returned. This technique is still in use today, and is supported in embedded environments by at least Diab if not others. I've used it successfully in VxWorks/Diab projects. It may be that GNU did not go in this direction (did they?) because it is dot-oh dependent. - -- Lee Douglas Gibbons L u c e n t T e c h n o l o g i e s ldgibbons@lucent.com Bell Labs Innovations --------------------------- Newsgroups: comp.os.vxworks Subject: VxFusion Date: Wed, 1 Mar 2000 20:24:10 GMT From: David & Bella Aldrich Message-ID: <2000030120241083839@zetnet.co.uk> Hi I wonder if anyone has experience of using VxFusion? It is expensive (6k pounds) and I would like some opinions before our company goes ahead with it. Does anyone know of the existance of a VxFusion Adapter layer for PCI? Thank you David Aldrich --------------------------- Newsgroups: comp.os.vxworks Subject: local filesystem/ change drive / ... Date: Wed, 1 Mar 2000 21:30:52 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89juks$gfs$2@newsmaster01.nextra.at> Hello, current environment: HOST: NT 4.05 Workstation Target : VxWorks from Tornado 2.0, X86 Communication: Ethernet. I boot the trarget via LOCAL floppy, and load vxworks itself via FTP from HOST. as filesystem i have a remote file system via netDrv (FTP FileSystem). at a command prompt (at the target) i start in my homedir in HOSTNAME:n:\vxworks\home i can do cd ".." or cd "something" to change directory on drive N: 1st problem: how can i change to a different drive on the host? when i do cd "m:\" i see HOSTNAME:n:\vxworks\home\m: which is not what i want. when i change to \ and then to m: i see HOSTNAME:/m: which does not work too. :-( ?? 2nd problem: how can i switch to my local A: from where i booted? can i access this drive from the host too? i want to copy some files from host to target and have them local on the target. 3rd problem: the target machine has an SCSI Harddisk local. can i - and if, how - access this local harddrive from vxWorks? there is a e.g. DOS formatted C: drive on the SCSI-Harddisk. i use a NCR 810 SCSI adapter. how can i switch to this harddisk? 4) do i need NFS? what do i need to do to use NFS? - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: Changing usrNetwork.c Date: Wed, 01 Mar 2000 21:01:47 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> Hi all, Under Tornado II I need to modify \src\config\usrNetwork.c. How do I implement the changed source file in my Tornado II project? Thanx, Michael --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How do I reduce data sizes in flash? Date: Wed, 1 Mar 2000 19:50:29 -0500 From: "Jeff Creem" Organization: MV Communications, Inc. Message-ID: <89kdsg$p1a$1@pyrite.mv.net> References: <38bd1465.418562180@news.rrds.co.uk> Obviously this is an area where Rational support ought to be able to help give you a pointer. You mention the bss -- what do the other segments look like? I've worked with the VADS tools which output a map file after link that shows you where your space is being allocated down to the package/procedure level. I assume ApexWorks has something similar. Try a simple hello world and see what that works out to. At least in the early versions of ApexWorks the entire run time got linked into your executable -- perhaps this is where some of your space is being eaten up. If you get no where with Rational support I'd recommend a switch to GNAT. Steve Folly wrote in message news:38bd1465.418562180@news.rrds.co.uk... > Hi, > > I hope someone can provide some help on the following problem: > We have a program running on a DY4 PowerPC 603e with VxWorks 5.3. > > The program is in Ada95 and compiled with Rational ApexWorks 3.0.0b > > There are currently some very large static data structures (up to about 3Mb) > which, when the program is linked and programmed into flash take up about 3Mb > of flash! > > This seems rather wasteful as the initialization could be done programmatically > and the data structures dynamically allocated at run time. > > Well, I had a go at doing this but the resulting executable and S records file > came out at the same size. Curious thing is, according to the link map the > structures in the program_bss section had disappeared. > > Should I expect the size of the executable or S records file to have reduced? > Or am I missing something here? Are there other ways to reduce flash usage? > > BTW, We are also looking at compressing the flash image. > > Thanks for any help. > > -- > Regards, > Steve Folly, Software Engineer, Racal Defence Electronics Ltd. > mailto:steve.folly@rdel.co.uk > http://www.rdel.co.uk/ > +44 (0)1293 528787 > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Stethoscope 5.0 Date: Thu, 02 Mar 2000 01:47:40 GMT From: MJM Organization: mjm et al Message-ID: <38BDC697.3CF2BAF6@sympatico.ca> References: <38BB3124.2A45485F@sympatico.ca> <89grpc$vc7$1@nnrp1.deja.com> <38BC89CF.54B62979@sympatico.ca> <89ji4e$s9v$1@nnrp1.deja.com> Reply-To: mike.mcnaughton@sympatico.ca Hello Ted: I do have a target server going. I read some pretty good material in the Release notes that indicate the need to INCLUDE_NET_SYM_TABLE and INCLUDE_ANSI_ALL. I had missed a couple of the components in the ansi_all and have not yet tried full compliance with the relaes notes. I have high hopes that the fix is in this area. My FS Engineer is away on a conference and hasn't been able to get into this much. I was using the Steth 5.0 manual as an installation guide and just discovered the release notes. Thanks for your thoughts, MJM Ted Dennison wrote: > In article <38BC89CF.54B62979@sympatico.ca>, > mike.mcnaughton@sympatico.ca wrote: > > I have tornado running nicely and stethoscope even runs. It asserts a > > message on a status line at the bottm of the screen that says somehtin > like > > "can't connect to server" > > > A clue might be in the shell behavior. When I try something like > > > printf("test message") , > > I get an error message like AGENT_NO_AGENT_PROC. > > When I connect my target server in Tornado there's a line that says: > Connecting to target agent... succeeded. > > The terminology is suspiciously similar. You are running a target > server, right? What does that line say on yours? > > -- > T.E.D. > > http://www.telepath.com/~dennison/Ted/TED.html > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 02 Mar 2000 02:37:50 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: <2ukv4.937$l7.10115@news1.frmt1.sfba.home.com> References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> So modify it and rebuild vxWorks. Paul In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, "Michael Murphy" wrote: >Hi all, > >Under Tornado II I need to modify \src\config\usrNetwork.c. How do I >implement the changed source file in my Tornado II project? > >Thanx, > >Michael > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 2 Mar 2000 07:52:17 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <89l308$3ej$1@crusher.de.colt.net> References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> <2ukv4.937$l7.10115@news1.frmt1.sfba.home.com> copy the file to your BSP directory and add it to the workspace (project), the changed file will be used when you rebuild. You don't have to touch the original in your Tornado path. Achim Paul Whicker schrieb in im Newsbeitrag: 2ukv4.937$l7.10115@news1.frmt1.sfba.home.com... > So modify it and rebuild vxWorks. > > Paul > > > In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, "Michael Murphy" > wrote: > >Hi all, > > > >Under Tornado II I need to modify \src\config\usrNetwork.c. How do I > >implement the changed source file in my Tornado II project? > > > >Thanx, > > > >Michael > > > > > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Thu, 2 Mar 2000 08:03:17 +0100 From: "Markus MItterer" Organization: KPNQwest customer news service Message-ID: References: Hi, Have a look at the socket setsockopt()-function. There is a specific param for the size of send buffer (SO_SNDBUF). I don`t know the meaning of the error codes. Greetings Markus Mitterer, STK --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado 2 and STL Date: Thu, 02 Mar 2000 10:58:59 +0100 From: Artem Alimarine Organization: Philips Medical Systems Message-ID: <38BE3B63.5A390931@best.ms.philips.com> References: <38BD2B9A.76F08025@best.ms.philips.com> This is a multi-part message in MIME format. - --------------3E8313D8F05119027CB5481B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Dave Dave Craig wrote: > Hi Artem, > > There are a lot of problems to overcome with templates, STL and vxWorks. > However, almost all of these come at link time and not at compile time I have not linked it yet. So, I'll get more surprises then. > > (there > is an -frepo patch which fixes some template instantiation problems - look > on Windsurf for this). What are the compilation errors you are seeing? > The code I'm trying to compile has been written for VMS CXX and NT VC6.0 compilers that are, as far as i understand, much more advanced with respect to streams and STL support. The problems: - - basic_string does not exist, i have to use string instead - - basic_*stream do not exist - - string streams are different I can ovecome all of it with defines and rewriting parts of code, but it is still annoying. The code does not use STL containers and algirithms yet. When we are going to develop new portable code for VMS/NT/VxWorks we can use only intersection of the features supported by the compilers. This is annoying if VxWorks compiler makes me to refrain from using of many C++ and, streams and STL things. > > Dave Craig > Palo Alto, CA. > Thanks for the information Artem > > "Artem Alimarine" wrote in message > news:38BD2B9A.76F08025@best.ms.philips.com... > > Hello > > > > I am trying to compile some software written in C++ for on Tornado2. > > cc386 --version reports: > > cygnus-2.7.2-960126 egcs-971225 tornado 2.0 > > > > The problem is that it is two old C++ compiler. It reports errors on > > nromal usage of templates and STL stuff. > > > > Is there a newer Tornado 2 compiler (for pentium and i960)? > > > > Any help would be appreciated, thanks in advance > > Artem > > > > > > - --------------3E8313D8F05119027CB5481B Content-Type: text/x-vcard; charset=us-ascii; name="aalimari.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Artem Alimarine Content-Disposition: attachment; filename="aalimari.vcf" begin:vcard n:Alimarine;Artem tel;home:(+31) 40-2440633 tel;work:(+31) 40-2763078 x-mozilla-html:TRUE adr:;;;;;; version:2.1 email;internet:aalimari@best.ms.philips.com fn:Artem Alimarine end:vcard - --------------3E8313D8F05119027CB5481B-- --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Mar 2 04:18:20 2000 From: Anil Vijay K Date: Thu Mar 2 04:18:22 PST 2000 Subject: snmp agent in vxworks hello all, I am looking for availability of master SNMP agent in VxWorks which allows registration of subagents (e.g. using RFC 1592 or similar mechanism). These subagents are specific for manageability for new protocols. Could you give me references ? I am looking at a scenario where in addition of new protocol manageability is supported without making any change (not even building afresh) to master SNMP agent. I thank you for your time. anil From vxwexplo-errs@csg.lbl.gov Thu Mar 2 07:21:39 2000 From: Miri Epstein Date: Thu Mar 2 07:21:41 PST 2000 Subject: mRouteShow() vxWorks Tornado ================== Hi vxWorkers, We are using SENS 2.0 for our proprietary router (ipForwarding = 1). We call several times the following routeLib utilities: mRouteEntryAdd ()- add a protocol-specific route to the routing table mRouteEntryDelete () - delete a protocol-specific route from the routing table where : proto=0 /* static route */ After successfully adding about eight routing entries to the routing table and being able to ping each destination of the routes added, when typing mRouteShow() from the shell prompt - the target is stuck. By the way, when typing routeShow() (instead of mRouteShow), we see all the appropriate routing entries, but without network mask (as expected). Did somebody encounter this problem ? Does someone know if there is a mRouteEntryShow utility ? or is something wrong with the mRouteShow ? Is mRouteShow working only if mRouteAdd/mRouteDelete were used ? Any help is appreciated. Thx, Miri From vxwexplo-errs@csg.lbl.gov Fri Mar 3 04:03:11 2000 From: Vxworks Exploder Date: Fri Mar 3 04:03:14 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Mar 3 04:03:08 PST 2000 Subject: Re: STD_IN failing on target shell Subject: Re: Is munchless C++ linking possible Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Changing usrNetwork.c Subject: Re: Changing usrNetwork.c Subject: Re: How to speed up netDev in FTP-mode ? Subject: Real-time Developer Subject: Trouble detecting DVD ROM device with no media? Subject: Re: Changing usrNetwork.c Subject: Re: Changing usrNetwork.c Subject: nfsExportShow : how to include it? Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Raw Socket on Vxworks Subject: Re: non-blocking pipes Subject: Re: non-blocking pipes Subject: Re: Raw Socket on Vxworks Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Real-time Developer Subject: Re: nfsExportShow : how to include it? Subject: Re: Get Socket Send Buffer Size?l Subject: muxReceive Subject: Telephony API support on VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: STD_IN failing on target shell Date: Thu, 02 Mar 2000 13:41:38 GMT From: "Øyvind Holmeide" Organization: Telia Internet Public Access Message-ID: References: <89iees$i6f$1@newsmaster01.nextra.at> <6R6v4.1505$F3.181570048@news.telia.no> <89j23p$gn7$1@nnrp1.deja.com> Thanks, higher priority made the difference. Øyvind Thomas Rahn wrote in message news:89j23p$gn7$1@nnrp1.deja.com... > > > In article <6R6v4.1505$F3.181570048@news.telia.no>, > "Øyvind Holmeide" wrote: > > I have now added the symbol table also for the target (sync host and > target > > symbo tables + symbol table initialization components). The target > shell > > has now all the support that is found in the corresponding Tornado > shell. > > > > If my program that is based on STDIN from the console is started from > the > > target shell, then STDIN is received through e.g. fioRdString(), and > > everything is fine. However, if the application is spawned then STDIN > is > > not available regardless of any calls to > ioGlobalStdGet()/ioGlobalStdSet() > > or ioTaskStdGet()/ioTaskStdSet(). > > > The reason might be that the shell is still running (at a very > high priority). Your application will be spawned at prio 100 > by default, so the shell gets all the input.... > > Thomas Rahn > > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Is munchless C++ linking possible Date: 2 Mar 2000 14:01:25 GMT From: jnance@wrtph0qb.nortelnetworks.com (Jim Nance) Organization: Bell Northern Research Message-ID: <89ls7l$jhj$1@bcrkh13.ca.nortel.com> References: <891a6b$hiq$1@bcrkh13.ca.nortel.com> <38BD7258.22685AD@lucent.com> On Wed, 01 Mar 2000 12:41:12 -0700, Doug Gibbons wrote: >This technique is still in use today, and is supported in embedded >environments by at least Diab if not others. I've used it successfully >in VxWorks/Diab projects. It may be that GNU did not go in this direction >(did they?) because it is dot-oh dependent. You dont have to munch if you use g++ under Unix which is what prompted me to ask if it was really necessary under VxWorks. Of course under VxWorks, you want the constructors to run after the OS itself is up and running and its part of the same image. This may account for the difference. Jim --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Thu, 02 Mar 2000 15:50:02 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: Yes, I'm aware of setting the size. But I need to know how much is currently being used. Michael "Markus MItterer" wrote in message news:wmov4.2$U54.139@nreader2.kpnqwest.net... > Hi, > Have a look at the socket setsockopt()-function. There is a specific param > for the size of send buffer (SO_SNDBUF). I don`t know the meaning of the > error codes. > > Greetings > Markus Mitterer, STK > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 02 Mar 2000 16:00:41 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> <2ukv4.937$l7.10115@news1.frmt1.sfba.home.com> <89l308$3ej$1@crusher.de.colt.net> If I copy usrNetwork.c to my project directory and then include in my project, I get alot of undefines during build. Here is a sample cc386 -g -mpentium -ansi -nostdinc -DRW_MULTI_THREAD -D_REENTRANT -fvolatile -nostdlib -fn o-builtin -fno-defer-pop -IE:\Tornado\target\proj\rtperiph1 -IE:\Tornado\tar get\config\RTP ERIPH1 -IE:\Tornado\target\h -IE:\Tornado\target\config\comps\src -IE:\Torna do\target\src\ drv -DCPU=PENTIUM -DPRJ_BUILD -DCPU_VARIANT=PENTIUM -c E:\Tornado\target\proj\rtperiph1\us rNetwork.c E:\Tornado\target\proj\rtperiph1\usrNetwork.c:296: syntax error before `ipCfgParams' E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: `IP_FLAGS_DFLT' undeclared here (not in a function) E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: initializer element for `ipCfgParams' i s not constant E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: `IP_TTL_DFLT' undeclared here (not in a function) E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: warning: excess elements in scalar init ializer after `ipCfgParams' E:\Tornado\target\proj\rtperiph1\usrNetwork.c:300: `IP_QLEN_DFLT' undeclared here (not in "Achim Zimmer" wrote in message news:89l308$3ej$1@crusher.de.colt.net... > copy the file to your BSP directory and add it to the workspace (project), > the changed file will be used when you rebuild. > You don't have to touch the original in your Tornado path. > > Achim > > Paul Whicker schrieb in im Newsbeitrag: > 2ukv4.937$l7.10115@news1.frmt1.sfba.home.com... > > So modify it and rebuild vxWorks. > > > > Paul > > > > > > In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, "Michael > Murphy" > > wrote: > > >Hi all, > > > > > >Under Tornado II I need to modify \src\config\usrNetwork.c. How do I > > >implement the changed source file in my Tornado II project? > > > > > >Thanx, > > > > > >Michael > > > > > > > > > > > > > > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 2 Mar 2000 13:51:47 -0700 From: "Mike Lechner" Organization: Dimensional Communications Message-ID: <7uAv4.182$J4.5700@wormhole.dimensional.com> References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> <2ukv4.937$l7.10115@news1.frmt1.sfba.home.com> <89l308$3ej$1@crusher.de.colt.net> Try not including it into your project. Just put it into the BSP directory and rebuild. The prjConfig.c file probably does an #include that'll pick it up from the BSP directory first. - --Mike Michael Murphy wrote in message news:Jewv4.83844$ox5.22532568@tw11.nn.bcandid.com... > If I copy usrNetwork.c to my project directory and then include in my > project, I get alot of undefines during build. Here is a sample > > cc386 -g -mpentium -ansi -nostdinc -DRW_MULTI_THREAD -D_REENTRANT -fvolatile > -nostdlib -fn > o-builtin -fno-defer-pop -IE:\Tornado\target\proj\rtperiph1 -IE:\Tornado\tar > get\config\RTP > ERIPH1 -IE:\Tornado\target\h -IE:\Tornado\target\config\comps\src -IE:\Torna > do\target\src\ > drv -DCPU=PENTIUM -DPRJ_BUILD -DCPU_VARIANT=PENTIUM -c > E:\Tornado\target\proj\rtperiph1\us > rNetwork.c > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:296: syntax error before > `ipCfgParams' > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: `IP_FLAGS_DFLT' > undeclared here (not in > a function) > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: initializer element for > `ipCfgParams' i > s not constant > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: `IP_TTL_DFLT' undeclared > here (not in a > function) > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: warning: excess elements > in scalar init > ializer after `ipCfgParams' > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:300: `IP_QLEN_DFLT' undeclared > here (not in > > "Achim Zimmer" wrote in message > news:89l308$3ej$1@crusher.de.colt.net... > > copy the file to your BSP directory and add it to the workspace (project), > > the changed file will be used when you rebuild. > > You don't have to touch the original in your Tornado path. > > > > Achim > > > > Paul Whicker schrieb in im Newsbeitrag: > > 2ukv4.937$l7.10115@news1.frmt1.sfba.home.com... > > > So modify it and rebuild vxWorks. > > > > > > Paul > > > > > > > > > In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, "Michael > > Murphy" > > > wrote: > > > >Hi all, > > > > > > > >Under Tornado II I need to modify \src\config\usrNetwork.c. How do I > > > >implement the changed source file in my Tornado II project? > > > > > > > >Thanx, > > > > > > > >Michael > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to speed up netDev in FTP-mode ? Date: Thu, 02 Mar 2000 14:38:42 -0600 From: Andrew Johnson Organization: APS, Argonne National Laboratory Message-ID: <38BED152.6B8B85E9@aps.anl.gov> References: <89lbgg$bo32@overload.lbl.gov> Rainer Stelzer wrote: > > Some of our control systems do not have a local filesystem, so we use a > netDev > to access the parameter files over network (FTP-mode). > > Reading the files this way, is very much slower than accessing a local > filesystem. > (800ms vs.45s !!!, We have up to seven files opened. We don't use the > seek() command ) Every time you open a file, netDev has to connect and login to the host using FTP, download the whole file into memory and close the FTP connection. The call blocks until the whole file is in memory, thus it's usually very slow as you've found. > The netDev library seems to be implemented in a way, that it consumes > only a small memory > area. (and this is a good solution for some of our systems which are low > in memory) > causing the disadvantage that it has to use the the FTP server very > often. This implies that you're only using small files - it uses as much memory as it needs to hold the complete file contents while the file is open. If your files get bigger, the memory usage will grow to match. > Any hint to speed up the transfer (without changing our parser)? The simplest speedup is to switch to the NFS filesystem, which is *much* faster than netDrv, and for large files it doesn't need as much memory as it doesn't load the whole file at once. If you can't do that for some reason then reducing the number of files you open will help, but would probably need you to change the code. - - Andrew - -- Complexity comes for free, Simplicity you have to work for. --------------------------- Newsgroups: comp.os.vxworks Subject: Real-time Developer Date: Thu, 02 Mar 2000 21:09:29 GMT From: "TalentLAB Inc." Organization: Magma Communications Ltd. Message-ID: Our Client is looking for a software developer who can work in a real-time UNIX environment. Working with the existing database (sybase) running on a UNIX platform, the software developer will work to optimize the performance of the database using C++. The successful candidate will have the experience with embedded SQL and understand large systems. The product is a satellite tracking device used in the trucking industry. The database is used in the communication of information. This is an opportunity to work in a well-funded satellite telecommunications company, which has spent 4 successful years of R&D designing and building commercial mobile terminal products. The company has recently begun commercializing the product. Contact Information To apply this and other career opportunities please submit your resume to: stuartm@talentlab.com Be sure to sign up for the Technology Job Report and check out our website at http://www.talentlab.com --------------------------- Newsgroups: comp.os.vxworks Subject: Trouble detecting DVD ROM device with no media? Date: Thu, 2 Mar 2000 21:26:41 -0000 From: "aparry" Organization: Tesco ISP Message-ID: <89mmp4$201$1@barcode.tesco.net> Can anyone help me? My programme can't see a DVD device when there is no media in the tray. I have two work-arounds; fix the block size during the device creation or boot up my system, place a disc into the DVD device then re-boot the system again!!! Since I am green to this area and my support is poor, please can anyone tell me how I can detect the device like any other device. Thanks in advance --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 02 Mar 2000 22:10:22 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> <2ukv4.937$l7.10115@news1.frmt1.sfba.home.com> <89l308$3ej$1@crusher.de.colt.net> <7uAv4.182$J4.5700@wormhole.dimensional.com> usrNetwork.c is not included in prjConfig.c. So placing in the BSP dir does not seem to do anything. I tried. Michael "Mike Lechner" wrote in message news:7uAv4.182$J4.5700@wormhole.dimensional.com... > Try not including it into your project. Just put it into the BSP directory > and rebuild. The prjConfig.c file probably does an #include that'll pick it > up from the BSP directory first. > > --Mike > > Michael Murphy wrote in message > news:Jewv4.83844$ox5.22532568@tw11.nn.bcandid.com... > > If I copy usrNetwork.c to my project directory and then include in my > > project, I get alot of undefines during build. Here is a sample > > > > > cc386 -g -mpentium -ansi -nostdinc -DRW_MULTI_THREAD -D_REENTRANT -fvolatile > > -nostdlib -fn > > > o-builtin -fno-defer-pop -IE:\Tornado\target\proj\rtperiph1 -IE:\Tornado\tar > > get\config\RTP > > > ERIPH1 -IE:\Tornado\target\h -IE:\Tornado\target\config\comps\src -IE:\Torna > > do\target\src\ > > drv -DCPU=PENTIUM -DPRJ_BUILD -DCPU_VARIANT=PENTIUM -c > > E:\Tornado\target\proj\rtperiph1\us > > rNetwork.c > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:296: syntax error before > > `ipCfgParams' > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: `IP_FLAGS_DFLT' > > undeclared here (not in > > a function) > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:298: initializer element for > > `ipCfgParams' i > > s not constant > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: `IP_TTL_DFLT' > undeclared > > here (not in a > > function) > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:299: warning: excess > elements > > in scalar init > > ializer after `ipCfgParams' > > E:\Tornado\target\proj\rtperiph1\usrNetwork.c:300: `IP_QLEN_DFLT' > undeclared > > here (not in > > > > "Achim Zimmer" wrote in message > > news:89l308$3ej$1@crusher.de.colt.net... > > > copy the file to your BSP directory and add it to the workspace > (project), > > > the changed file will be used when you rebuild. > > > You don't have to touch the original in your Tornado path. > > > > > > Achim > > > > > > Paul Whicker schrieb in im Newsbeitrag: > > > 2ukv4.937$l7.10115@news1.frmt1.sfba.home.com... > > > > So modify it and rebuild vxWorks. > > > > > > > > Paul > > > > > > > > > > > > In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, "Michael > > > Murphy" > > > > wrote: > > > > >Hi all, > > > > > > > > > >Under Tornado II I need to modify \src\config\usrNetwork.c. How do I > > > > >implement the changed source file in my Tornado II project? > > > > > > > > > >Thanx, > > > > > > > > > >Michael > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Changing usrNetwork.c Date: Thu, 02 Mar 2000 22:27:29 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com> <89l75t$2dt$1@nnrp1.deja.com> O.K., This got me going... Thanx, Michael wrote in message news:89l75t$2dt$1@nnrp1.deja.com... > For project builds the file is slightly different, and is located in a > new place... look in target/config/comps/src/net and you should find a > collection of network related "configlettes" (the new name for these > small C fragments that get compiled at project build time). > > You'll still need to modify the one in src/config if you plan to build > from the BSP directory (for example, to create a boot ROM image). > > Hope that helps, > > John... > > In article <%yfv4.82173$ox5.22069561@tw11.nn.bcandid.com>, > "Michael Murphy" wrote: > > Hi all, > > > > Under Tornado II I need to modify \src\config\usrNetwork.c. How do I > > implement the changed source file in my Tornado II project? > > > > Thanx, > > > > Michael > > > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: nfsExportShow : how to include it? Date: Fri, 3 Mar 2000 01:02:03 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89mvd5$fqr$1@newsmaster01.nextra.at> Hello. i have a general Tornado (2) development question and a specialized question. i want to build a new VxWorks image (486). i want a given function to be included (e.g. nfsExportShow) what do i have to include when i build vxWorks? where can i find a relation nbetween functions and INCLUDE_xxx sections? - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 00:40:34 GMT From: snaphook@southwesternbell.net (Bruce) Organization: SBC Internet Services Message-ID: <38bf09ce.6753726@news.swbell.net> References: In comp.os.vxworks "Michael Murphy" wrote: >Yes, I'm aware of setting the size. But I need to know how much is currently >being used. There is an analog to setsockopt() called getsockopt(). Or do you want to know how much is IN the buffer. You should be able to do that with ioctl() --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 01:53:39 GMT From: Daniel Wong Organization: Deja.com - Before you buy. Message-ID: <89n5v2$hir$1@nnrp1.deja.com> References: <38bf09ce.6753726@news.swbell.net> Which option would you use with ioctl() to get the number of bytes that are currently in the send buffer and what's the syntax for that option? Daniel In article <38bf09ce.6753726@news.swbell.net>, snaphook@southwesternbell.net (Bruce) wrote: > In comp.os.vxworks > "Michael Murphy" wrote: > > >Yes, I'm aware of setting the size. But I need to know how much is currently > >being used. > > There is an analog to setsockopt() called getsockopt(). Or do you want to > know how much is IN the buffer. You should be able to do that with ioctl() > > - -- Daniel Wong daniel_k_wong@my-dejan.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Raw Socket on Vxworks Date: Thu, 02 Mar 2000 13:17:29 -0500 From: Peter Colby Organization: Raytheon Systems Company C3I Message-ID: <38BEB039.1AF58EC4@res.raytheon.com> References: <89k8bg$bnr1@overload.lbl.gov> try serveraddr.sin_addr.s_addr = inet_addr ("192.168.100.140"); (in other words, I don't believe that htonl is necessary) or you might want to try mucking around with hostent if vxWorks has it. Navin Kumar Agrawal wrote: > > Tornado > > Hi Folks, > > I would really appreciate if somebody can help me in solving a problem > of > mine which is related with the usage of Raw Socket on VxWorks. I have > put my effort in the following way. > > -------------------------------------------------------- > struct sockaddr_in serveraddr; > bzero ((char *)&serveraddr, sizeof(struct sockaddr_in)); > serveraddr.sin_family = AF_INET; > serveraddr.sin_len = sizeof(struct sockaddr_in); > serveraddr.sin_port = htons (6007); > serveraddr.sin_addr.s_addr = htonl (inet_addr ("192.168.100.140")); > > if ((fd = socket (AF_INET, SOCK_RAW, IPPROTO_ICMP)) == ERROR) > { > perror ("socket"); > return -1; > } > > if (bind (fd, (struct sockaddr *)&serveraddr, sizeof(struct > sockaddr_in)) == ERROR) > { > perror ("bind failed"); > return -1; > } > -------------------------------------------------------- > > My target system is based on PPC604 (PowerPC). On executing the above > program, > it is failing at bind () and the error code returned is > EADDRNOTAVAIL ("Can't > assign requested address"). > > Can anybody help me in sovling this problem? > > Regards, > Navin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: non-blocking pipes Date: Thu, 02 Mar 2000 13:02:32 -0500 From: Peter Colby Organization: Raytheon Systems Company C3I Message-ID: <38BEACB8.F9E29757@res.raytheon.com> References: <38bc4b5d$0$206@nntp1.ba.best.com> try int on = TRUE; ioctl(fd, FIONBIO, (int)&on); > Mitch Stermer wrote: > > I am porting a very large Unix application to VxWorks. I cannot change > the structure of > > the code so I am forced into using a select call. > > The code used non-blocking fifo's on Unix and I am attempting to use > pipeDevCreate > > as the fd's for the select. What I would like is to something like a > > fcntl(fd, F_SETFL, _FNONBLOCK ); > > on the fd of the created pipe. Is there such a way to set this type of > option. > > Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: non-blocking pipes Date: Thu, 2 Mar 2000 20:43:37 -0700 From: Rich Ireland Organization: XMission http://www.xmission.com/ Message-ID: References: <38bc4b5d$0$206@nntp1.ba.best.com> In article <38bc4b5d$0$206@nntp1.ba.best.com>, Mitch Stermer wrote: > I am porting a very large Unix application to VxWorks. I cannot change the structure of > > the code so I am forced into using a select call. > > The code used non-blocking fifo's on Unix and I am attempting to use pipeDevCreate > > as the fd's for the select. What I would like is to something like a > > fcntl(fd, F_SETFL, _FNONBLOCK ); > > on the fd of the created pipe. Is there such a way to set this type of option. > > Thanks. Do you *have* to use pipes? If the data you are passing can be placed into fixed blocks, I would suggest that message queues are more efficient and have a more predicable behavior. By using message queues you can do a non-blocking read without using select. - -- Rich Ireland mailto:r.ireland@computer.org Embedded & Real Time Engineering --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Raw Socket on Vxworks Date: Thu, 02 Mar 2000 19:58:14 -0500 From: Thomas Davis Organization: Interpath Communications, Inc. Message-ID: <38BF0E26.27ADE3B1@galois.net> References: <89k8bg$bnr1@overload.lbl.gov> Navin Kumar Agrawal wrote: > > Tornado > > Hi Folks, > > I would really appreciate if somebody can help me in solving a problem > of > mine which is related with the usage of Raw Socket on VxWorks. I have > put my effort in the following way. > > -------------------------------------------------------- > struct sockaddr_in serveraddr; > bzero ((char *)&serveraddr, sizeof(struct sockaddr_in)); > serveraddr.sin_family = AF_INET; > serveraddr.sin_len = sizeof(struct sockaddr_in); > serveraddr.sin_port = htons (6007); > serveraddr.sin_addr.s_addr = htonl (inet_addr ("192.168.100.140")); > > if ((fd = socket (AF_INET, SOCK_RAW, IPPROTO_ICMP)) == ERROR) > { > perror ("socket"); > return -1; > } > > if (bind (fd, (struct sockaddr *)&serveraddr, sizeof(struct > sockaddr_in)) == ERROR) > { > perror ("bind failed"); > return -1; > } > -------------------------------------------------------- > > My target system is based on PPC604 (PowerPC). On executing the above > program, > it is failing at bind () and the error code returned is > EADDRNOTAVAIL ("Can't > assign requested address"). > > Can anybody help me in sovling this problem? > > Regards, > Navin Peter is right, htonl is not needed because ip address are not numbers they are 4 bytes that are just easy to carry around that way. Thomas - -- Galois, Inc. http://www.galois.net We color the world! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 05:42:29 GMT From: snaphook@southwesternbell.net (Bruce) Organization: SBC Internet Services Message-ID: <38bf505d.24819374@news.swbell.net> References: <38bf09ce.6753726@news.swbell.net> <89n5v2$hir$1@nnrp1.deja.com> In comp.os.vxworks Daniel Wong wrote: >Which option would you use with ioctl() to get the number of bytes that >are currently in the send buffer and what's the syntax for that option? I don't have my development system with me, but I was looking at serial code like this tonight. I believe it was: ioctl( socknum, FIONREAD, &nBytes); And the number to be read was stored in nBytes. You might also have a look at select() if this doesn't work. Bruce --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Real-time Developer Date: Fri, 03 Mar 2000 06:15:27 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: <3MIv4.1880$WI1.23108@news1.frmt1.sfba.home.com> References: Goodness! Real-time sybase, isn't that an oxymoron? Whatever is the world coming to? Paul In article , "TalentLAB Inc." wrote: > >Our Client is looking for a software developer who can work in a real-time >UNIX environment. Working with the existing database (sybase) running on a >UNIX platform, the software developer will work to optimize the performance >of the database using C++. The successful candidate will have the experience >with embedded SQL and understand large systems. The product is a satellite >tracking device used in the trucking industry. The database is used in the >communication of information. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: nfsExportShow : how to include it? Date: Fri, 03 Mar 2000 06:52:30 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89nnfe$tb8$1@nnrp1.deja.com> References: <89mvd5$fqr$1@newsmaster01.nextra.at> Hi Mario, > i want a given function to be included (e.g. nfsExportShow) > > what do i have to include when i build vxWorks? nfsExportShow is part of nfsLib.o (at least it is on my ARM version... should be the same for yours). Since this is an NFS client facility, the component you would need should be INCLUDE_NFS. > where can i find a relation > nbetween functions and INCLUDE_xxx sections? This is a little bit more involved... For the example you gave, you just needed to look in the manual for nfsLib (the library level manual entry for the function) and it says that you need the INCLUDE_NFS configuration macro - i.e. the NFS client software. To find the relationship between the component INCLUDE_xxx names and the more readable names shown in the GUI look in target/config/comps/vxWorks for the component's description. Now, for others you may need to find the name of the .o file (library) and search in target/config/comps/vxWorks for the component that matches. If none matches, then you are looking for something that the auto-dependency analysis will find for you, and you'll need to guess what the parent component is I think... Hope that's at least a little help, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 3 Mar 2000 09:39:40 +0100 From: "Markus MItterer" Organization: KPNQwest customer news service Message-ID: <_SKv4.11$wq4.364@nreader2.kpnqwest.net> References: Okay, this seems to be a bigger problem. Sorry, I don`t know, but please tell me if you can solve it. --------------------------- Newsgroups: comp.os.vxworks Subject: muxReceive Date: Fri, 3 Mar 2000 10:44:38 +0100 From: "Yves NORTIER" Organization: [Posted via] Easynet France Message-ID: <1SLv4.179$b3.3692@tengri.easynet.fr> Hi, Why the function muxReceive isn't described in muxLib (Tornado II Ref Manual)? We can find it in the old SENS TI as one of the method for reception in the END-MUX interface , and in the muxLib.h file. Does anybody use it whithout any problem ? What is the good way for using (error codes, ...)? Thanks for help. Yves --------------------------- Newsgroups: comp.os.vxworks Subject: Telephony API support on VxWorks Date: Fri, 03 Mar 2000 09:56:24 GMT From: parameshbabu@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89o287$53j$1@nnrp1.deja.com> Hi, We are right now evaluating Windows CE and VxWorks for the use of Telephony based embedded applications. I do know that, Windows CE supports Telephony with a rich set of APIs. Do VxWorks support Telephony APIs? Or Are there Third party vendors available for the Telephony Plugins for VxWorks? Could anyone throw some light on this direction... Awaiting earliest response, Regards, Paramesh Babu Software Engineer Philips Software Centre. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Mar 4 00:07:15 2000 From: Navin Kumar Agrawal Date: Sat Mar 4 00:07:17 PST 2000 Subject: fcntl () call in VxWorks Tornado Hi Folks, How do we use fcntl () system call in a vxworks based application program? Do I have to include some specific include file to call this function? I used fcntl () call in one of my application program. While compiling there was no problem. But while downloading the .out file on the target, it gave me 'fcntl' symbol unresolved error. Can anybody help me in solving this? Regards, Navin From vxwexplo-errs@csg.lbl.gov Sat Mar 4 04:03:09 2000 From: Vxworks Exploder Date: Sat Mar 4 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Mar 4 04:03:06 PST 2000 Subject: VxWorks on a PC/104 with a CAN bus interface? Subject: [OT] Re: Real-time Developer Subject: Difference Subject: Test Subject: Re: Get Socket Send Buffer Size?l Subject: Re: Difference Subject: Re: Get Socket Send Buffer Size?l Subject: ucmp_error in T2 Subject: PSOS // VxWorks Subject: Snmp NOTIFICATION-TYPE Subject: Re: Get Socket Send Buffer Size?l Subject: ECP/EPP support. Any hints? Subject: How to set IP address on a running system Subject: Re: How to get current time & date for timestamps for DOS-FS files Subject: Re: How to get current time & date for timestamps for DOS-FS files Subject: changing the BSP Makefile Subject: Re: [OT] Re: Real-time Developer ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks on a PC/104 with a CAN bus interface? Date: Fri, 03 Mar 2000 14:12:17 +0100 From: "Mattias Sjögren" Organization: Uppsala University Message-ID: <38BFBA31.E1077447@student.docs.uu.se> Hi all. I'm in a project to build soccer playing robots for the RoboCup challenge and we have now come to the point where we have to choose an OS and peripherals for the robots. Has anyone any experience with using a PC/104 and a CAN module together with VxWorks? What CAN modules are supported by VxWorks? Which one is the best? Thanks for your time, Mattias Sjögren http://www.docs.uu.se/robocup/ --------------------------- Newsgroups: comp.os.vxworks Subject: [OT] Re: Real-time Developer Date: 03 Mar 2000 11:11:51 -0500 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <3MIv4.1880$WI1.23108@news1.frmt1.sfba.home.com> Sender: bpringle@DeadDuck It is real-time; a really long time. Real-time generally means the ability to meet a dead line (could be hard or soft). It is a much maligned phrase. You can use with other words of its ilk, like object oriented, RAD, wireless Internet and Java DOM/SAX XML polymorphic data device. Collect many phrases like this, put them together with a random key and you have created your average high tech marketing genius with an MBA. They say things like 'Calculus, what's that?' and 'Who cares about physics'. Bill >>>>> "Paul" == Paul Whicker writes: Paul> Goodness! Real-time sybase, isn't that an oxymoron? Whatever Paul> is the world coming to? >>>> "TalentLAB Inc." wrote: >> Our Client is looking for a software developer who can work in >> a real-time UNIX environment. Working with the existing >> database (sybase) running on a UNIX platform, the software >> developer will work to optimize the performance of the database >> using C++. The successful candidate will have the experience >> with embedded SQL and understand large systems. The product is >> a satellite tracking device used in the trucking industry. The >> database is used in the communication of information. - -- crc16("oust")=0xdddd --------------------------- Newsgroups: comp.os.vxworks Subject: Difference Date: Fri, 3 Mar 2000 16:53:53 -0600 From: "Rao" Organization: Esat Net Message-ID: <38bfedf4.0@seat.s3two.ie> Could any one give me a hint about the difference between Character-Oriented Devices and Block-Oriented Devices please Rao --------------------------- Newsgroups: comp.os.vxworks Subject: Test Date: Fri, 3 Mar 2000 16:55:18 -0600 From: "Rao" Organization: Esat Net Message-ID: <38bfee49.0@seat.s3two.ie> Test only...sorry to disturb you. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 17:28:20 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: <38bf09ce.6753726@news.swbell.net> <89n5v2$hir$1@nnrp1.deja.com> <38bf505d.24819374@news.swbell.net> Yes, I'am trying to get the free bytes left in the buffer. I have tried ioctrl( socknum, FIONREAD, &nBytes). It always returns zero in nBytes. Michael "Bruce" wrote in message news:38bf505d.24819374@news.swbell.net... > In comp.os.vxworks > Daniel Wong wrote: > > >Which option would you use with ioctl() to get the number of bytes that > >are currently in the send buffer and what's the syntax for that option? > > I don't have my development system with me, but I was looking at serial > code like this tonight. I believe it was: > > ioctl( socknum, FIONREAD, &nBytes); > > And the number to be read was stored in nBytes. You might also have a look > at select() if this doesn't work. > > Bruce > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Difference Date: Fri, 03 Mar 2000 11:30:29 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38C012D5.3A0E@despam.pangea.ca> References: <38bfedf4.0@seat.s3two.ie> Rao wrote: > Could any one give me a hint about the difference between Character-Oriented > Devices and Block-Oriented Devices please > I would suggest you check some of the classical references; Bach or one of the BSD devil books for background. Consider what is the smallest unit of data writable to a given piece of hardware. In a character device it will typically be a byte; while in a block device it will be a block [512/1024/2048 bytes]. There may be a filesystem in the picture as well. - -het - --- "Assembly of Japanese bicycle require great peace of mind." - Instruction manual mentioned by R. Pirsig Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 17:40:04 GMT From: "Michael Murphy" Organization: Imaging Business Machines, LLC Message-ID: References: <38bf09ce.6753726@news.swbell.net> <89n5v2$hir$1@nnrp1.deja.com> <38bf505d.24819374@news.swbell.net> Does anyone have a sample for checking free bytes in a socket using select? Thanx, Michael "Michael Murphy" wrote in message news:UCSv4.86629$ox5.23162579@tw11.nn.bcandid.com... > Yes, I'am trying to get the free bytes left in the buffer. I have tried > ioctrl( socknum, FIONREAD, &nBytes). It always returns zero in nBytes. > > Michael > > "Bruce" wrote in message > news:38bf505d.24819374@news.swbell.net... > > In comp.os.vxworks > > Daniel Wong wrote: > > > > >Which option would you use with ioctl() to get the number of bytes that > > >are currently in the send buffer and what's the syntax for that option? > > > > I don't have my development system with me, but I was looking at serial > > code like this tonight. I believe it was: > > > > ioctl( socknum, FIONREAD, &nBytes); > > > > And the number to be read was stored in nBytes. You might also have a > look > > at select() if this doesn't work. > > > > Bruce > > > > --------------------------- Newsgroups: comp.os.vxworks Subject: ucmp_error in T2 Date: Fri, 03 Mar 2000 17:38:37 GMT From: fatihozcan@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89otat$oji$1@nnrp1.deja.com> When working with T1 icmp_error is a defined in lib but merging to T2 we have no prototype for it. And the source code of it is available in the internet but the prototypes that it calls differs from the T2 has defined in xx_icmp.h files. Can you help me? Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: PSOS // VxWorks Date: Fri, 03 Mar 2000 13:22:16 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38C02D08.6E6A@despam.pangea.ca> Some here may find this story relevant... http://www.eetimes.com/story/OEG20000302S0038 - -het - --- "Assembly of Japanese bicycle require great peace of mind." - Instruction manual mentioned by R. Pirsig Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: Snmp NOTIFICATION-TYPE Date: Fri, 03 Mar 2000 18:08:48 GMT From: rliu@inter-tel.com Organization: Deja.com - Before you buy. Message-ID: <89ov3b$q0m$1@nnrp1.deja.com> I am using WindNet SNMPv1/v2c Agent and its mibcomp. Here is my mib for notifications: sdTraps OBJECT IDENTIFIER ::= { myproduct 4 } sysAccessDenied NOTIFICATION-TYPE STATUS current DESCRIPTION "." ::= { sdTraps 1 } When it is compiled, I got this warning ./snmp/superdsp.mib:854: Warning: The next-to-last sub-identifier in the name of any newly-defined notification must have the value zero. Can anyone tell me what this means? Thanks Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size?l Date: Fri, 03 Mar 2000 20:27:26 GMT From: bwedding@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89p778$hs$1@nnrp1.deja.com> References: <38bf09ce.6753726@news.swbell.net> <89n5v2$hir$1@nnrp1.deja.com> <38bf505d.24819374@news.swbell.net> In article , "Michael Murphy" wrote: > Yes, I'am trying to get the free bytes left in the buffer. I have > tried ioctrl( socknum, FIONREAD, &nBytes). It always returns zero in >nBytes. That is ioctl() correct? That should work according to the manual. It says a socket is an ordinary FD like any other. It looks like select() won't work, but this should: char tmpBuf[REC_BUF_SZ]; nBytes = recv( socknum, tmpBuf, REC_BUF_SZ, MSG_PEEK); The problem is the data is copied into the tmpBuf, but it isn't removed from the socket receive buffer. Bruce Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: ECP/EPP support. Any hints? Date: Fri, 03 Mar 2000 20:38:59 +0100 From: "Henk W.A.M. van Rooij" Organization: Eindhoven University of Technology Message-ID: <38C014D3.881C8FD@tue.nl> Hi, We want to use the parallel port of an ordinary PC (468 or P1) in ECP or EPP mode. Officially EPP/ECP is not supported by VWorks. Do any of you have any experience (good or bad) in using EPP/ECP under VxWorks? Can it be done easily, or is it a difficult task to write a driver for ECP/EPP use? Thanks for your help. Henk. --------------------------- Newsgroups: comp.os.vxworks Subject: How to set IP address on a running system Date: Fri, 3 Mar 2000 22:16:32 +0100 From: "Øyvind Holmeide" Organization: Telenor Online Public Access Message-ID: Two questions: 1.) How can the IP address be changed on a running system. 2.) How can the MAC address be extracted? regards Øyvind Holmeide --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get current time & date for timestamps for DOS-FS files Date: Fri, 3 Mar 2000 22:36:39 +0100 From: "Zakelijk" Organization: UUNET-NL (http://www.nl.uu.net) Message-ID: <89pb9i$t77$1@porthos.nl.uu.net> References: <38bb9cf1.9562099@news.ronet.de> <38BC10F1.4F07@despam.pangea.ca> Harvey. In the message below u mention that u should use clock_settime() to set the system time. That's all nice etc. but where can i get for example the time from the bios. What i like to do: - - having a standalone (PCPEntium boarded) VXWorks machine which could autoboot without interference from the outside. (or rather if i tried to do a reboot i would want to move to the location of the machine to reset the time and date). Do u (or anyone else) knows how to set the time on such a standalone system. thx.. Leendert Biemans Harvey Taylor wrote in message news:38BC10F1.4F07@despam.pangea.ca... > Manfred Fischer wrote: > > I'm working with a MV2700 PowerPC board with a DOS FS on a harddisk. > > Is there a possibilty to get current time and date as timestamp for > > new created or changed files. > > > > Is there a special mechanism to apply a timestamp automatically on > > files. > > > > Yes look at the dosFsLib docs. The function is called > dosFsDateTimeInstall(). > > It requires a handler function which typically looks like: > > STATUS DateTimeFunc(DOS_DATE_TIME * pDosDateTime) > { > time_t tNow = time(NULL); > struct tm* t = localtime(&tNow); > > pDosDateTime->dosdt_year = t->tm_year ; > pDosDateTime->dosdt_year += ( (t->tm_year > 70) ? 1900 : 2000); > pDosDateTime->dosdt_month = t->tm_mon + 1 ; /* Map 0..11 to 1..12 */ > pDosDateTime->dosdt_day = t->tm_mday ; > pDosDateTime->dosdt_hour = t->tm_hour ; > pDosDateTime->dosdt_minute = t->tm_min ; > pDosDateTime->dosdt_second = t->tm_sec ; > > return( OK ) ; > } > > You should probably add error handling to taste. > Oh and of cousre you must have done a clock_settime() to > make this work. > > > -het > > > > > --- > "Assembly of Japanese bicycle require great peace of mind." > - Instruction manual mentioned by R. Pirsig > > Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get current time & date for timestamps for DOS-FS files Date: Fri, 03 Mar 2000 16:21:14 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38C056FA.1071@despam.pangea.ca> References: <38bb9cf1.9562099@news.ronet.de> <38BC10F1.4F07@despam.pangea.ca> <89pb9i$t77$1@porthos.nl.uu.net> Zakelijk wrote: > In the message below u mention that u should use clock_settime() to set the > system time. > That's all nice etc. but where can i get for example the time from the bios. > Hi Leendert, I am not familiar with the MV2700 PowerPC, but it depends upon your particular RTC chip. If it uses something different, you will have to write some RTC reading code yourself (or bug the MV2700 supplier). If it looks like a generic PC, you can use Linux source (or something like snippets.org). I'm sorry but I don't happen to have a sample to hand. Anybody? - -het > What i like to do: > - having a standalone (PCPEntium boarded) VXWorks machine which could > autoboot without interference from the outside. > > (or rather if i tried to do a reboot i would want to move to the location of > the machine to reset the time and date). > > Do u (or anyone else) knows how to set the time on such a standalone > system. > > thx.. Leendert Biemans > > Harvey Taylor wrote in message > news:38BC10F1.4F07@despam.pangea.ca.. > > Manfred Fischer wrote: > > > I'm working with a MV2700 PowerPC board with a DOS FS on a harddisk. > > > Is there a possibilty to get current time and date as timestamp for > > > new created or changed files. > > > > > > Is there a special mechanism to apply a timestamp automatically on > > > files. > > > > > > > Yes look at the dosFsLib docs. The function is called > > dosFsDateTimeInstall(). > > > > It requires a handler function which typically looks like: > > > > STATUS DateTimeFunc(DOS_DATE_TIME * pDosDateTime) > > { > > time_t tNow = time(NULL); > > struct tm* t = localtime(&tNow); > > > > pDosDateTime->dosdt_year = t->tm_year ; > > pDosDateTime->dosdt_year += ( (t->tm_year > 70) ? 1900 : 2000); > > pDosDateTime->dosdt_month = t->tm_mon + 1 ; /* Map 0..11 to 1..12 */ > > pDosDateTime->dosdt_day = t->tm_mday ; > > pDosDateTime->dosdt_hour = t->tm_hour ; > > pDosDateTime->dosdt_minute = t->tm_min ; > > pDosDateTime->dosdt_second = t->tm_sec ; > > > > return( OK ) ; > > } > > > > You should probably add error handling to taste. > > Oh and of cousre you must have done a clock_settime() to > > make this work. > > > > > > -het > > > > > > > > > > --- > > "Assembly of Japanese bicycle require great peace of mind." > > - Instruction manual mentioned by R. Pirsig > > > > Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het - -- - --- "Assembly of Japanese bicycle require great peace of mind." - Instruction manual mentioned by R. Pirsig Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: changing the BSP Makefile Date: Sat, 04 Mar 2000 05:27:15 GMT From: "Dragan Ostojic" Organization: Sympatico Message-ID: I moved portions of BSP Makefile into another file and then I included it in the same Makefile. The idea is to keep board specific definitions separate from the core BSP. When I tried to build the project based on such a BSP, final project Makefile would ignore macros that were included instead of inlined. Does anyone know what is going on? Dragan. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: [OT] Re: Real-time Developer Date: Sat, 04 Mar 2000 09:19:13 GMT From: pwhicker@home.com (Paul Whicker) Organization: @Home Network Message-ID: References: <3MIv4.1880$WI1.23108@news1.frmt1.sfba.home.com> Realtime, to me, also means deterministic. I want to know that when I expect something to happen then it will take a particular time. A database just isn't realtime because it will take however long it takes, the performance is not reliable. paul In article , Bill Pringlemeir wrote: > >It is real-time; a really long time. Real-time generally means >the ability to meet a dead line (could be hard or soft). It is >a much maligned phrase. You can use with other words of its ilk, >like object oriented, RAD, wireless Internet and Java DOM/SAX XML >polymorphic data device. > >Collect many phrases like this, put them together with a random >key and you have created your average high tech marketing genius >with an MBA. They say things like 'Calculus, what's that?' and >'Who cares about physics'. > >Bill > >>>>>> "Paul" == Paul Whicker writes: > > Paul> Goodness! Real-time sybase, isn't that an oxymoron? Whatever > Paul> is the world coming to? > >>>>> "TalentLAB Inc." wrote: > >> Our Client is looking for a software developer who can work in > >> a real-time UNIX environment. Working with the existing > >> database (sybase) running on a UNIX platform, the software > >> developer will work to optimize the performance of the database > >> using C++. The successful candidate will have the experience > >> with embedded SQL and understand large systems. The product is > >> a satellite tracking device used in the trucking industry. The > >> database is used in the communication of information. > --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Mar 4 22:11:36 2000 From: Ilia Maller Date: Sat Mar 4 22:11:38 PST 2000 Subject: RE: non-blocking pipes Hi, You have to implement your own pipe driver which will respond to FIONBIO ioctl. I suggest to use message queue as an underlying mechanism. It's pretty easy thing to do. VxWorks manuals describe how to write and install your own driver. See VxWorks' Programmer Guide, 3.9. Internal structure, p.140. You may also look at memDrv.c (tornado/target/src/usr) as example. I had implemented it a few years ago, and hope I still have sources. You may contact me at iliam@radware.com Ilia Maller. > > > > Mitch Stermer wrote: > > > > ..... > > > > The code used non-blocking fifo's on Unix and I am attempting to use > > pipeDevCreate > > > > as the fd's for the select. What I would like is to something like a > > > > fcntl(fd, F_SETFL, _FNONBLOCK ); > > > > on the fd of the created pipe. Is there such a way to set > this type of > > option. > > > > Thanks. > > ------------------------------------------------------------------------ Ilia Maller                                     RadWare Ltd.   Tel: +972-3-7657907 Fax: +972-3-6485641 Mailto:IliaM@radware.com ICQ#:17064929 ------------------------------------------------------------------------ From vxwexplo-errs@csg.lbl.gov Sun Mar 5 04:03:07 2000 From: Vxworks Exploder Date: Sun Mar 5 04:03:09 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Mar 5 04:03:04 PST 2000 Subject: Re: changing the BSP Makefile Subject: Re: nfsExportShow : how to include it? Subject: Re: a problem when using "ClearCase" and "Tornado2.0" together Subject: Re: PPC740 and MMU problem Subject: Very simple question on open() and fopen() failure Subject: Re: changing the BSP Makefile Subject: ***FREE MOBILE CALLS*** This works! 6206 Subject: ***FREE MOBILE CALLS*** This works! 7205 Subject: Re: nfsExportShow : how to include it? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: changing the BSP Makefile Date: Sat, 04 Mar 2000 11:12:03 -0700 From: James Scott Organization: Newsfeeds.com http://www.newsfeeds.com 60,000+ UNCENSORED Newsgroups. Message-ID: <38C151F0.F251CC7A@amug.org> References: Are you using Tornado 1 or 2? Builds in Tornado 1 seem to follow "normal" make rules and such. Builds in Tornado 2 happen through the configuration manager, and the rules are setup when you define a project based upon a BSP. When you modify a BSP you need to re-create the project for the BSP to build properly. James Dragan Ostojic wrote: > > I moved portions of BSP Makefile into another file and then I included it in > the same Makefile. The idea is to keep board specific definitions separate > from the core BSP. When I tried to build the project based on such a BSP, > final project Makefile would ignore macros that were included instead of > inlined. Does anyone know what is going on? > > Dragan. - -----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- http://www.newsfeeds.com - The #1 Newsgroup Service in the World! - -----== Over 80,000 Newsgroups - 16 Different Servers! =----- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: nfsExportShow : how to include it? Date: Fri, 3 Mar 2000 22:40:05 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <89rmp8$hub$1@newsmaster01.nextra.at> References: <89mvd5$fqr$1@newsmaster01.nextra.at> <89nnfe$tb8$1@nnrp1.deja.com> Hello, thx for the reply. looks as things are different on HOST=NT and your host. a) after haveing activated ALL the nfs* parts, the function nfsExportShow is still not avail on the target. i had to add "NetWorkShowRoutines." b) i have not a single idea, where do you see this INCLUDE_NFS in the manuals. i opened the help, manual index, selected the function, and here is the start of the .html file which contains the function: - ---------------------------------------------------------------------------- - ---- VxWorks Reference Manual : Libraries nfsLib NAME nfsLib - Network File System (NFS) library ROUTINES nfsHelp( ) - display the NFS help menu nfsExportShow( ) - display the exported file systems of a remote host nfsAuthUnixPrompt( ) - modify the NFS UNIX authentication parameters nfsAuthUnixShow( ) - display the NFS UNIX authentication parameters nfsAuthUnixSet( ) - set the NFS UNIX authentication parameters nfsAuthUnixGet( ) - get the NFS UNIX authentication parameters nfsIdSet( ) - set the ID number of the NFS UNIX authentication parameters DESCRIPTION ... i cannot see the include_xxx anywhere. more, i search lowlevel all the files in the vxworks directory, and i didnt find a single documentation which shows a relation between the function and a include_xxx. when i knwo the includexxx value, then it is no problem anymore. - -- regards, Mario Semo. http://www.kirchnersoft.com john_94501@my-deja.com wrote in message <89nnfe$tb8$1@nnrp1.deja.com>... >Hi Mario, > >> i want a given function to be included (e.g. nfsExportShow) >> >> what do i have to include when i build vxWorks? > >nfsExportShow is part of nfsLib.o (at least it is on my ARM version... >should be the same for yours). Since this is an NFS client facility, the >component you would need should be INCLUDE_NFS. > >> where can i find a relation >> nbetween functions and INCLUDE_xxx sections? > >This is a little bit more involved... For the example you gave, you just >needed to look in the manual for nfsLib (the library level manual >entry for the function) and it says that you need the INCLUDE_NFS >configuration macro - i.e. the NFS client software. To find the >relationship between the component INCLUDE_xxx names and the more >readable names shown in the GUI look in target/config/comps/vxWorks for >the component's description. > >Now, for others you may need to find the name of the .o file (library) >and search in target/config/comps/vxWorks for the component that >matches. If none matches, then you are looking for something that the >auto-dependency analysis will find for you, and you'll need to guess >what the parent component is I think... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: a problem when using "ClearCase" and "Tornado2.0" together Date: 4 Mar 2000 20:39:58 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <89rsau$5rd$1@news1.xs4all.nl> References: <38BB266B.98845913@hns.com> rwang wrote: : Hello, : We met a problem when we used "ClearCase" as version control tool in : "Tornado" environment. What version of Tornado are you using? : The problem is we can not share the "Makefile" without checkout. : For example, two users have their own views, j:\user1 and j:\user2, : which are mapped from a unix server to a NT client. In T2 the Makefile is an intermediate file, so you should not put this file in an version control tool. : When user1 setup a : new project in the sharing workspace, the Makefile is created : automatically based on user1's mapping directory. The first two lines in : the core information part of the makefile looks like: : ## core information : PRJ_DIR = J:/user1/projectname : PRJ_FILE = projectname.wpj Try to make these rules relative, so for example: PRJ_DIR = . : I tried to set some environment variables, like "$(WIND_BASE)", in the : PRJ_DIR in the Makefile. Such as: : PRJ_DIR = $(MAPPED_USER_PRJDIR) When using T2, you can define this variable as an extra macro. 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: PPC740 and MMU problem Date: Sat, 04 Mar 2000 21:10:58 GMT From: Karen Musgrove & Stephane Couture Organization: Sympatico Message-ID: <38C17EB9.50130724@sympatico.ca> References: <38BA4D40.B15BBDF8@nortelnetworks.com> What about disabling the unused BAT registers? Setting VP and VS bit to zero instead of using a 128KB size would probably help. Also, I would make sure that your PTE are not overlapping your BAT description. Ofer Goren wrote: > Hi. > I have the following configuration: > A PPC740 based board with up to 384MB of memory. > I can reach this amount of memory using three memory slots of 128MB > each, but each slot can have also 64MB simm, and here comes my problem. > > In the BAT table, I defined 2 IBATs and 2 DBATs. I initialize each for > 256MB in sysLib.c. > Then in sysSetDramBat(), I'm fixing the BATs, according to the real > amount of memory I have. > Please pay attention, that I can have memeory simms of 64MB also. > Well, all went well for 128MB and 384MB, but for 256MB the board refused > to come up. So I have 2 problems here: > 1. If I have only 128MB, I deed to fix my First IBAT/DBAT to 128MB. What > do I do with my second IBAT/DBAT? For now, I put there 128K, but this is > probably a bug which will result unpredictable behaviour. So what do I > do? > 2. For 256MB, I put 128MB in my first IBAT/DBAT and 128MB in my second > IBAT/DBAT. With that scenarion, my board did not come up. So I fixed it > so my first IBAT/DBAT will be 256MB, and my second DBAT/IBAT will be > 128K. again, bug, but at least the board comes up. > > Here is my initialized BATS: > /* I BAT 0 - code resides within CPU SDRAM */ > ((LOCAL_MEM_LOCAL_ADRS & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | > _MMU_UBAT_VS | _MMU_UBAT_VP), > ((LOCAL_MEM_LOCAL_ADRS & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW), > > /* I BAT 1 - additional CPU bank */ > ((LOCAL_MEM_LOCAL_ADRS2 & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | > > _MMU_UBAT_VS | _MMU_UBAT_VP), > ((LOCAL_MEM_LOCAL_ADRS2 & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW), > .... > ... > /* D BAT 0 - data resides within CPU SDRAM */ > ((LOCAL_MEM_LOCAL_ADRS & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | > _MMU_UBAT_VS | _MMU_UBAT_VP), > ((LOCAL_MEM_LOCAL_ADRS & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW), > > /* D BAT 1 - additional CPU bank */ > ((LOCAL_MEM_LOCAL_ADRS2 & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | > > _MMU_UBAT_VS | _MMU_UBAT_VP), > ((LOCAL_MEM_LOCAL_ADRS2 & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW), > > ... > } > > And here is the fix done later in sysSetDramBat(): > > ULONG sysSetDramBat() > { > ULONG bs1, bs2; > ULONG offset; > > switch (sysDramSize) > { > case (512*1024*1024): /* 512 MB */ > /* I guess third bank is empty, and first two banks are 256 > each */ > bs1 = _MMU_UBAT_BL_256M; > bs2 = _MMU_UBAT_BL_256M; > offset = (256 * 1024 * 1024); > break; > > case (384*1024*1024): /* 384 MB */ > /* 128 in each bank */ > bs1 = _MMU_UBAT_BL_256M; /* bank 1 and 2 */ > bs2 = _MMU_UBAT_BL_128M; /* bank 3 */ > offset = (256 * 1024 * 1024); > break; > > case (256*1024*1024): /* 256 MB */ > bs1 = _MMU_UBAT_BL_256M; /* bank 1 and 2 */ > bs2 = _MMU_UBAT_BL_128K; /* bank 3 */ > offset = (256 * 1024 * 1024); > /* > here, I know I have a bug. bs2 should be 0 or the same as > bs1. But 0 is not possible, and when I tried to put the > same > value of bs1 in bs2, the board wont come up. So I put > there > the minimal memory avaliable, hoping later I'll have the > time > to research the bug, and figure it out. > */ > break; > > case (128*1024*1024): /* 128 MB */ > bs1 = _MMU_UBAT_BL_128M; /* bank 1 and 2 */ > bs2 = _MMU_UBAT_BL_128K; /* bank 3 */ > offset = (128 * 1024 * 1024); > /* > here, I know I have a bug. bs2 should be 0 or the same as > bs1. But 0 is not possible, and when I tried to put the > same > value of bs1 in bs2, the board wond come up. So I put > there > the minimal memory avaliable, hoping later I'll have the > time > to research the bug, and figure it out. > */ > break; > > default: /* 128 MB */ > bs1 = _MMU_UBAT_BL_128M; /* bank 1 and 2 */ > bs2 = _MMU_UBAT_BL_128K; /* bank 3 */ > offset = (128 * 1024 * 1024); > break; > } > /* I BAT 0 */ > sysBatDesc[0] = ( (LOCAL_MEM_LOCAL_ADRS & _MMU_UBAT_BEPI_MASK) | > bs1 | _MMU_UBAT_VS | _MMU_UBAT_VP); > /* I BAT 1 */ > sysBatDesc[2] = ( ((LOCAL_MEM_LOCAL_ADRS + offset) & > _MMU_UBAT_BEPI_MASK) | > bs2 | _MMU_UBAT_VS | _MMU_UBAT_VP); > /* D BAT 0 */ > sysBatDesc[8] = ( (LOCAL_MEM_LOCAL_ADRS & _MMU_UBAT_BEPI_MASK) | > bs1 | _MMU_UBAT_VS | _MMU_UBAT_VP); > /* D BAT 1 */ > sysBatDesc[10] = ( ((LOCAL_MEM_LOCAL_ADRS + offset) & > _MMU_UBAT_BEPI_MASK) | > bs2 | _MMU_UBAT_VS | _MMU_UBAT_VP); > > return(bs1); > } > > Can any one help? > > -- > Embedded SW Team Leader > Email : ogoren@nortelnetworks.com > Phone number: +9723-6456023 > Fax number : +9723-6479579 > ESN : 826-6023 > http://www.nortelnetworks.com > ---------------------------------------------------- > If you lend someone $20, and never see that person again, > it was probably worth the investment. --------------------------- Newsgroups: comp.os.vxworks Subject: Very simple question on open() and fopen() failure Date: Sat, 04 Mar 2000 14:43:29 -0800 From: Marco Zandonadi Organization: SBC Internet Services Message-ID: <38C19191.5B7859E5@pacbell.net> Reply-To: marcoz@teja.com Hello, I have VXWorks 5.4 and my target is VXSim for Windows. My problem is that open() and fopen() always fail. Here's what I did. - - I created a new project - - I double clicked "Create a bootable VXWorks image (custom configured)" - - I pressed Next - - I selected "An existing project" and pressed Next - - I pressed Finish - - I double clicked on userAppInit.c and modified it as follows: #include ... void usrAppInit (void) { ... FILE * f = fopen ("myfile", "r"); if (!f) printf ("File not found!\n"); else printf ("File found!\n"); } - - I selected Build (compilation works ok) - - I opened Object Modules, right clicked on vxWorks.exe and selected Start 'vxWorks.exe' The simulator starts properly and prints out File not found! even though myfile is in the same directory as vxWorks.exe. I also tried opening other files like c:/autoexec.bat but the result is the same. Note that I didn't touch the default VxWorks component configuration. Does anybody know what I'm missing? Thanks very much in advance. Marco Zandonadi --------------------------- Newsgroups: comp.os.vxworks Subject: Re: changing the BSP Makefile Date: Sun, 05 Mar 2000 01:48:05 GMT From: "Dragan Ostojic" Organization: Sympatico Message-ID: References: <38C151F0.F251CC7A@amug.org> I'm using Tornado 2. I'm sorry I was a bit vague in my description. What I did is I modified BSP and then used poject manager to create project Makefile. Project manager left those parameters that were included undifined in the generated project Makefile. I would like to import through the BSP automaticaly certain number of parameters before I start to change project settings. - -- Dragan Ostojic ostojic@sympatico.ca "James Scott" wrote in message news:38C151F0.F251CC7A@amug.org... > Are you using Tornado 1 or 2? Builds in Tornado 1 seem to > follow "normal" make rules and such. Builds in Tornado 2 > happen through the configuration manager, and the rules > are setup when you define a project based upon a BSP. When > you modify a BSP you need to re-create the project for the > BSP to build properly. > > James > > Dragan Ostojic wrote: > > > > I moved portions of BSP Makefile into another file and then I included it in > > the same Makefile. The idea is to keep board specific definitions separate > > from the core BSP. When I tried to build the project based on such a BSP, > > final project Makefile would ignore macros that were included instead of > > inlined. Does anyone know what is going on? > > > > Dragan. > > > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- > http://www.newsfeeds.com - The #1 Newsgroup Service in the World! > -----== Over 80,000 Newsgroups - 16 Different Servers! =----- --------------------------- Newsgroups: comp.os.v,comp.os.vms,comp.os.vxworks,comp.os.windows95,comp.os.xinu Subject: ***FREE MOBILE CALLS*** This works! 6206 Date: Sun, 05 Mar 2000 02:36:00 GMT From: mregpg@force9.net Message-ID: Get it while you can.....dont know how long site will be up:) http://www.geocities.com/nutech2/payandgo.zip ksjmzimorvoqsdqzgrrxsyyuuuockhtw --------------------------- Newsgroups: comp.os.v,comp.os.vms,comp.os.vxworks,comp.os.windows95,comp.os.xinu Subject: ***FREE MOBILE CALLS*** This works! 7205 Date: Sun, 05 Mar 2000 03:21:39 GMT From: lfywdn@btinternet.com Message-ID: <7pkw4.23634$O5.221847@stones> Run this file, and after 20 seconds of looking at optical visuals you will WANT to ring all your friends...damn amazing!!! www.fortunecity.com/westwood/makeover/759/optical.exe bcqhjxyctljfwdoctchqitbweyuyhtgmnbuxwbsngkgglejzfendxtvkdqerlzzobzcntfozbshesfkjlfgpwxheocntvxc --------------------------- Newsgroups: comp.os.vxworks Subject: Re: nfsExportShow : how to include it? Date: Sun, 05 Mar 2000 09:27:50 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89t9ak$mk2$1@nnrp1.deja.com> References: <89mvd5$fqr$1@newsmaster01.nextra.at> <89nnfe$tb8$1@nnrp1.deja.com> <89rmp8$hub$1@newsmaster01.nextra.at> Hi, > a) after haveing activated ALL the nfs* parts, the function nfsExportShow is > still not avail on the target. i had to add "NetWorkShowRoutines." Ah, could be... I tend to have all the show routines in by default so I wouldn't have seen that one as a problem. > b) i have not a single idea, where do you see this INCLUDE_NFS in the > manuals. i opened the help, manual index, selected the function, and here is > the start of the .html file which contains the function: > when i knwo the includexxx value, then it is no problem anymore. I found it on the web based version (I was at home when I replied). The second paragraph of the description says: "VxWorks is delivered with NFS disabled. The configuration macro for NFS is INCLUDE_NFS." TTFN, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sun Mar 5 19:08:43 2000 From: "Administrator" Date: Sun Mar 5 19:08:46 PST 2000 Subject: Message not deliverable --=_sb2000/03/06-03-04-01-538=_sb Content-Type: text/plain; charset=us-ascii ----------------------------------- Returned ----------------------------------- From: vxwexplo@lbl.gov (the vxWorks Users Group Exploder) at internet-gateway Date: 3/4/00 4:12AM -0800 *To: vxworks_users@csg.lbl.gov at internet-gateway Subject: comp.os.vxworks newsdigest ------------------------------------------------------------------------------- --=_sb2000/03/06-03-04-01-538=_sb Content-Type: application/octet-stream; name="UNKNOWN.BIN" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="UNKNOWN.BIN" U3VibWl0dGVkLWJ5IHZ4d2V4cGxvLWVycnNAY3NnLmxibC5nb3YgIFNhdCBNYXIgIDQgMDQ6MDM6 MDkgMjAwMA0KU3VibWl0dGVkLWJ5OiBWeHdvcmtzIEV4cGxvZGVyIDx2eF9leHBsb0Bjc2cubGJs Lmdvdj4NCg0KQ29tcC5Pcy5WeHdvcmtzIERhaWx5IERpZ2VzdCAgICBTYXQgTWFyICA0IDA0OjAz OjA2IFBTVCAyMDAwDQoNCiAgICAgICAgU3ViamVjdDogVnhXb3JrcyBvbiBhIFBDLzEwNCB3aXRo IGEgQ0FOIGJ1cyBpbnRlcmZhY2U/DQogICAgICAgIFN1YmplY3Q6IFtPVF0gUmU6IFJlYWwtdGlt ZSBEZXZlbG9wZXINCiAgICAgICAgU3ViamVjdDogRGlmZmVyZW5jZQ0KICAgICAgICBTdWJqZWN0 OiBUZXN0DQogICAgICAgIFN1YmplY3Q6IFJlOiBHZXQgU29ja2V0IFNlbmQgQnVmZmVyIFNpemU/ bA0KICAgICAgICBTdWJqZWN0OiBSZTogRGlmZmVyZW5jZQ0KICAgICAgICBTdWJqZWN0OiBSZTog R2V0IFNvY2tldCBTZW5kIEJ1ZmZlciBTaXplP2wNCiAgICAgICAgU3ViamVjdDogdWNtcF9lcnJv ciBpbiBUMg0KICAgICAgICBTdWJqZWN0OiBQU09TIC8vIFZ4V29ya3MNCiAgICAgICAgU3ViamVj dDogU25tcCBOT1RJRklDQVRJT04tVFlQRQ0KICAgICAgICBTdWJqZWN0OiBSZTogR2V0IFNvY2tl dCBTZW5kIEJ1ZmZlciBTaXplP2wNCiAgICAgICAgU3ViamVjdDogRUNQL0VQUCBzdXBwb3J0LiBB bnkgaGludHM/DQogICAgICAgIFN1YmplY3Q6IEhvdyB0byBzZXQgSVAgYWRkcmVzcyBvbiBhIHJ1 bm5pbmcgc3lzdGVtDQogICAgICAgIFN1YmplY3Q6IFJlOiBIb3cgdG8gZ2V0IGN1cnJlbnQgdGlt ZSAmIGRhdGUgZm9yIHRpbWVzdGFtcHMgZm9yIERPUy1GUyBmaQ0KbGVzDQogICAgICAgIFN1Ympl Y3Q6IFJlOiBIb3cgdG8gZ2V0IGN1cnJlbnQgdGltZSAmIGRhdGUgZm9yIHRpbWVzdGFtcHMgZm9y IERPUy1GUyBmaQ0KbGVzDQogICAgICAgIFN1YmplY3Q6IGNoYW5naW5nIHRoZSBCU1AgTWFrZWZp bGUNCiAgICAgICAgU3ViamVjdDogUmU6IFtPVF0gUmU6IFJlYWwtdGltZSBEZXZlbG9wZXINCg0K LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLQ0K DQpOZXdzZ3JvdXBzOiBjb21wLm9zLnZ4d29ya3MNClN1YmplY3Q6IFZ4V29ya3Mgb24gYSBQQy8x MDQgd2l0aCBhIENBTiBidXMgaW50ZXJmYWNlPw0KRGF0ZTogRnJpLCAwMyBNYXIgMjAwMCAxNDox MjoxNyArMDEwMA0KRnJvbTogIk1hdHRpYXMgU2r2Z3JlbiIgPGl0OTZtYXNAc3R1ZGVudC5kb2Nz LnV1LnNlPg0KT3JnYW5pemF0aW9uOiBVcHBzYWxhIFVuaXZlcnNpdHkNCk1lc3NhZ2UtSUQ6IDwz OEJGQkEzMS5FMTA3NzQ0N0BzdHVkZW50LmRvY3MudXUuc2U+DQoNCkhpIGFsbC4NCg0KSSdtIGlu IGEgcHJvamVjdCB0byBidWlsZCBzb2NjZXIgcGxheWluZyByb2JvdHMgZm9yIHRoZSBSb2JvQ3Vw DQpjaGFsbGVuZ2UgYW5kIHdlIGhhdmUgbm93IGNvbWUgdG8gdGhlIHBvaW50IHdoZXJlIHdlIGhh dmUgdG8gY2hvb3NlIGFuDQpPUyBhbmQgcGVyaXBoZXJhbHMgZm9yIHRoZSByb2JvdHMuDQoNCkhh cyBhbnlvbmUgYW55IGV4cGVyaWVuY2Ugd2l0aCB1c2luZyBhIFBDLzEwNCBhbmQgYSBDQU4gbW9k dWxlIHRvZ2V0aGVyDQp3aXRoIFZ4V29ya3M/DQoNCldoYXQgQ0FOIG1vZHVsZXMgYXJlIHN1cHBv cnRlZCBieSBWeFdvcmtzPyBXaGljaCBvbmUgaXMgdGhlIGJlc3Q/DQoNClRoYW5rcyBmb3IgeW91 ciB0aW1lLA0KDQpNYXR0aWFzIFNq9mdyZW4NCmh0dHA6Ly93d3cuZG9jcy51dS5zZS9yb2JvY3Vw Lw0KDQotLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52 eHdvcmtzDQpTdWJqZWN0OiBbT1RdIFJlOiBSZWFsLXRpbWUgRGV2ZWxvcGVyDQpEYXRlOiAwMyBN YXIgMjAwMCAxMToxMTo1MSAtMDUwMA0KRnJvbTogQmlsbCBQcmluZ2xlbWVpciA8YnByaW5nbGVt ZWlyQHlhaG9vLmNvbT4NCk9yZ2FuaXphdGlvbjogRmFjdG9yeSBvZiB0aGUgbWluZA0KTWVzc2Fn ZS1JRDogPHV1MmlvbmlkNC5mc2ZfLV9AeWFob28uY29tPg0KUmVmZXJlbmNlczogPGRNQXY0LjYw MTUkTDMuMTY4MjI4NDVAbmV3cy5tYWdtYS5jYT4gPDNNSXY0LjE4ODAkV0kxLjIzMTA4QG5ld3Mx LmYNCnJtdDEuc2ZiYS5ob21lLmNvbT4NClNlbmRlcjogYnByaW5nbGVARGVhZER1Y2sNCg0KDQpJ dCBpcyByZWFsLXRpbWU7IGEgcmVhbGx5IGxvbmcgdGltZS4gIFJlYWwtdGltZSBnZW5lcmFsbHkg bWVhbnMNCnRoZSBhYmlsaXR5IHRvIG1lZXQgYSBkZWFkIGxpbmUgKGNvdWxkIGJlIGhhcmQgb3Ig c29mdCkuICBJdCBpcw0KYSBtdWNoIG1hbGlnbmVkIHBocmFzZS4gIFlvdSBjYW4gdXNlIHdpdGgg b3RoZXIgd29yZHMgb2YgaXRzIGlsaywNCmxpa2Ugb2JqZWN0IG9yaWVudGVkLCBSQUQsIHdpcmVs ZXNzIEludGVybmV0IGFuZCBKYXZhIERPTS9TQVggWE1MDQpwb2x5bW9ycGhpYyBkYXRhIGRldmlj ZS4NCg0KQ29sbGVjdCBtYW55IHBocmFzZXMgbGlrZSB0aGlzLCBwdXQgdGhlbSB0b2dldGhlciB3 aXRoIGEgcmFuZG9tDQprZXkgYW5kIHlvdSBoYXZlIGNyZWF0ZWQgeW91ciBhdmVyYWdlIGhpZ2gg dGVjaCBtYXJrZXRpbmcgZ2VuaXVzDQp3aXRoIGFuIE1CQS4gIFRoZXkgc2F5IHRoaW5ncyBsaWtl ICdDYWxjdWx1cywgd2hhdCdzIHRoYXQ/JyBhbmQNCidXaG8gY2FyZXMgYWJvdXQgcGh5c2ljcycu DQoNCkJpbGwNCg0KPj4+Pj4gIlBhdWwiID09IFBhdWwgV2hpY2tlciA8cHdoaWNrZXJAaG9tZS5j b20+IHdyaXRlczoNCg0KICAgIFBhdWw+IEdvb2RuZXNzISBSZWFsLXRpbWUgc3liYXNlLCBpc24n dCB0aGF0IGFuIG94eW1vcm9uPyBXaGF0ZXZlcg0KICAgIFBhdWw+IGlzIHRoZSB3b3JsZCBjb21p bmcgdG8/DQoNCj4+Pj4gIlRhbGVudExBQiBJbmMuIiAgPGluZm9AdGFsZW50bGFiLmNvbT4gd3Jv dGU6DQogICAgPj4gIE91ciBDbGllbnQgaXMgbG9va2luZyBmb3IgYSBzb2Z0d2FyZSBkZXZlbG9w ZXIgd2hvIGNhbiB3b3JrIGluDQogICAgPj4gYSByZWFsLXRpbWUgVU5JWCBlbnZpcm9ubWVudC4g V29ya2luZyB3aXRoIHRoZSBleGlzdGluZw0KICAgID4+IGRhdGFiYXNlIChzeWJhc2UpIHJ1bm5p bmcgb24gYSBVTklYIHBsYXRmb3JtLCB0aGUgc29mdHdhcmUNCiAgICA+PiBkZXZlbG9wZXIgd2ls bCB3b3JrIHRvIG9wdGltaXplIHRoZSBwZXJmb3JtYW5jZSBvZiB0aGUgZGF0YWJhc2UNCiAgICA+ PiB1c2luZyBDKysuIFRoZSBzdWNjZXNzZnVsIGNhbmRpZGF0ZSB3aWxsIGhhdmUgdGhlIGV4cGVy aWVuY2UNCiAgICA+PiB3aXRoIGVtYmVkZGVkIFNRTCBhbmQgdW5kZXJzdGFuZCBsYXJnZSBzeXN0 ZW1zLiBUaGUgcHJvZHVjdCBpcw0KICAgID4+IGEgc2F0ZWxsaXRlIHRyYWNraW5nIGRldmljZSB1 c2VkIGluIHRoZSB0cnVja2luZyBpbmR1c3RyeS4gVGhlDQogICAgPj4gZGF0YWJhc2UgaXMgdXNl ZCBpbiB0aGUgY29tbXVuaWNhdGlvbiBvZiBpbmZvcm1hdGlvbi4NCg0KLSAtLQ0KY3JjMTYoIm91 c3QiKT0weGRkZGQNCg0KLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tDQoNCk5ld3Nncm91cHM6 IGNvbXAub3Mudnh3b3Jrcw0KU3ViamVjdDogRGlmZmVyZW5jZQ0KRGF0ZTogRnJpLCAzIE1hciAy MDAwIDE2OjUzOjUzIC0wNjAwDQpGcm9tOiAiUmFvIiA8cmFvQHMzZ3JvdXAuY29tPg0KT3JnYW5p emF0aW9uOiBFc2F0IE5ldA0KTWVzc2FnZS1JRDogPDM4YmZlZGY0LjBAc2VhdC5zM3R3by5pZT4N Cg0KQ291bGQgYW55IG9uZSBnaXZlIG1lIGEgaGludCBhYm91dCB0aGUgZGlmZmVyZW5jZSBiZXR3 ZWVuIENoYXJhY3Rlci1PcmllbnRlZA0KRGV2aWNlcyBhbmQgQmxvY2stT3JpZW50ZWQgRGV2aWNl cyBwbGVhc2UNCg0KUmFvDQoNCg0KDQotLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3 c2dyb3VwczogY29tcC5vcy52eHdvcmtzDQpTdWJqZWN0OiBUZXN0DQpEYXRlOiBGcmksIDMgTWFy IDIwMDAgMTY6NTU6MTggLTA2MDANCkZyb206ICJSYW8iIDxyYW9AczNncm91cC5jb20+DQpPcmdh bml6YXRpb246IEVzYXQgTmV0DQpNZXNzYWdlLUlEOiA8MzhiZmVlNDkuMEBzZWF0LnMzdHdvLmll Pg0KDQpUZXN0IG9ubHkuLi5zb3JyeSB0byBkaXN0dXJiIHlvdS4NCg0KDQoNCi0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLQ0KDQpOZXdzZ3JvdXBzOiBjb21wLm9zLnZ4d29ya3MNClN1YmplY3Q6 IFJlOiBHZXQgU29ja2V0IFNlbmQgQnVmZmVyIFNpemU/bA0KRGF0ZTogRnJpLCAwMyBNYXIgMjAw MCAxNzoyODoyMCBHTVQNCkZyb206ICJNaWNoYWVsIE11cnBoeSIgPG1tdXJwaHlAaWJtbC5jb20+ DQpPcmdhbml6YXRpb246IEltYWdpbmcgQnVzaW5lc3MgTWFjaGluZXMsIExMQw0KTWVzc2FnZS1J RDogPFVDU3Y0Ljg2NjI5JG94NS4yMzE2MjU3OUB0dzExLm5uLmJjYW5kaWQuY29tPg0KUmVmZXJl bmNlczogPHdlZHY0LjgxOTE1JG94NS4yMjAwODMyNkB0dzExLm5uLmJjYW5kaWQuY29tPiA8d21v djQuMiRVNTQuMTM5QG5yZWENCmRlcjIua3BucXdlc3QubmV0PiA8SzR3djQuODM4MjUkb3g1LjIy NTI3MjE0QHR3MTEubm4uYmNhbmRpZC5jb20+IDwzOGJmMDljZS42NzUzDQo3MjZAbmV3cy5zd2Jl bGwubmV0PiA8ODluNXYyJGhpciQxQG5ucnAxLmRlamEuY29tPiA8MzhiZjUwNWQuMjQ4MTkzNzRA bmV3cy5zd2JlbA0KbC5uZXQ+DQoNClllcywgSSdhbSB0cnlpbmcgdG8gZ2V0IHRoZSBmcmVlIGJ5 dGVzIGxlZnQgaW4gdGhlIGJ1ZmZlci4gSSBoYXZlIHRyaWVkDQppb2N0cmwoIHNvY2tudW0sIEZJ T05SRUFELCAmbkJ5dGVzKS4gSXQgYWx3YXlzIHJldHVybnMgemVybyBpbiBuQnl0ZXMuDQoNCk1p Y2hhZWwNCg0KIkJydWNlIiA8c25hcGhvb2tAc291dGh3ZXN0ZXJuYmVsbC5uZXQ+IHdyb3RlIGlu IG1lc3NhZ2UNCm5ld3M6MzhiZjUwNWQuMjQ4MTkzNzRAbmV3cy5zd2JlbGwubmV0Li4uDQo+IElu IGNvbXAub3Mudnh3b3Jrcw0KPiBEYW5pZWwgV29uZyA8ZGFuaWVsX2tfd29uZ0BteS1kZWphLmNv bT4gd3JvdGU6DQo+DQo+ID5XaGljaCBvcHRpb24gd291bGQgeW91IHVzZSB3aXRoIGlvY3RsKCkg dG8gZ2V0IHRoZSBudW1iZXIgb2YgYnl0ZXMgdGhhdA0KPiA+YXJlIGN1cnJlbnRseSBpbiB0aGUg c2VuZCBidWZmZXIgYW5kIHdoYXQncyB0aGUgc3ludGF4IGZvciB0aGF0IG9wdGlvbj8NCj4NCj4g SSBkb24ndCBoYXZlIG15IGRldmVsb3BtZW50IHN5c3RlbSB3aXRoIG1lLCBidXQgSSB3YXMgbG9v a2luZyBhdCBzZXJpYWwNCj4gY29kZSBsaWtlIHRoaXMgdG9uaWdodC4gIEkgYmVsaWV2ZSBpdCB3 YXM6DQo+DQo+IGlvY3RsKCBzb2NrbnVtLCBGSU9OUkVBRCwgJm5CeXRlcyk7DQo+DQo+IEFuZCB0 aGUgbnVtYmVyIHRvIGJlIHJlYWQgd2FzIHN0b3JlZCBpbiBuQnl0ZXMuICBZb3UgbWlnaHQgYWxz byBoYXZlIGENCmxvb2sNCj4gYXQgc2VsZWN0KCkgaWYgdGhpcyBkb2Vzbid0IHdvcmsuDQo+DQo+ IEJydWNlDQo+DQoNCg0KDQotLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3Vw czogY29tcC5vcy52eHdvcmtzDQpTdWJqZWN0OiBSZTogRGlmZmVyZW5jZQ0KRGF0ZTogRnJpLCAw MyBNYXIgMjAwMCAxMTozMDoyOSAtMDgwMA0KRnJvbTogSGFydmV5IFRheWxvciA8aGV0QGRlc3Bh bS5wYW5nZWEuY2E+DQpPcmdhbml6YXRpb246IE9yZ2FuaXphdGlvbj8gV2hhdCBvcmdhbml6YXRp b24/IQ0KTWVzc2FnZS1JRDogPDM4QzAxMkQ1LjNBMEVAZGVzcGFtLnBhbmdlYS5jYT4NClJlZmVy ZW5jZXM6IDwzOGJmZWRmNC4wQHNlYXQuczN0d28uaWU+DQoNClJhbyB3cm90ZToNCj4gQ291bGQg YW55IG9uZSBnaXZlIG1lIGEgaGludCBhYm91dCB0aGUgZGlmZmVyZW5jZSBiZXR3ZWVuIENoYXJh Y3Rlci1PcmllbnRlZA0KPiBEZXZpY2VzIGFuZCBCbG9jay1PcmllbnRlZCBEZXZpY2VzIHBsZWFz ZQ0KPg0KDQogICAgIEkgd291bGQgc3VnZ2VzdCB5b3UgY2hlY2sgc29tZSBvZiB0aGUgY2xhc3Np Y2FsIHJlZmVyZW5jZXM7DQogICAgIEJhY2ggb3Igb25lIG9mIHRoZSBCU0QgZGV2aWwgYm9va3Mg Zm9yIGJhY2tncm91bmQuDQogICAgIENvbnNpZGVyIHdoYXQgaXMgdGhlIHNtYWxsZXN0IHVuaXQg b2YgZGF0YSB3cml0YWJsZSB0bw0KICAgICBhIGdpdmVuIHBpZWNlIG9mIGhhcmR3YXJlLiBJbiBh IGNoYXJhY3RlciBkZXZpY2UgaXQNCiAgICAgd2lsbCB0eXBpY2FsbHkgYmUgYSBieXRlOyB3aGls ZSBpbiBhIGJsb2NrIGRldmljZSBpdA0KICAgICB3aWxsIGJlIGEgYmxvY2sgWzUxMi8xMDI0LzIw NDggYnl0ZXNdLiAgVGhlcmUgbWF5IGJlDQogICAgIGEgZmlsZXN5c3RlbSBpbiB0aGUgcGljdHVy ZSBhcyB3ZWxsLg0KPGNpYW8+DQotIC1oZXQNCg0KDQoNCi0gLS0tDQoiQXNzZW1ibHkgb2YgSmFw YW5lc2UgYmljeWNsZSByZXF1aXJlIGdyZWF0IHBlYWNlIG9mIG1pbmQuIg0KICAgICAtIEluc3Ry dWN0aW9uIG1hbnVhbCBtZW50aW9uZWQgYnkgUi4gUGlyc2lnDQoNCkhhcnZleSBUYXlsb3IgIG1h aWx0bzpoZXRAZGVzcGFtLnBvcnRhbC5jYSAgaHR0cDovL3d3dy5wYW5nZWEuY2EvfmhldA0KDQot LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52eHdvcmtz DQpTdWJqZWN0OiBSZTogR2V0IFNvY2tldCBTZW5kIEJ1ZmZlciBTaXplP2wNCkRhdGU6IEZyaSwg MDMgTWFyIDIwMDAgMTc6NDA6MDQgR01UDQpGcm9tOiAiTWljaGFlbCBNdXJwaHkiIDxtbXVycGh5 QGlibWwuY29tPg0KT3JnYW5pemF0aW9uOiBJbWFnaW5nIEJ1c2luZXNzIE1hY2hpbmVzLCBMTEMN Ck1lc3NhZ2UtSUQ6IDxVTlN2NC44NjY0NyRveDUuMjMxNjcyNjBAdHcxMS5ubi5iY2FuZGlkLmNv bT4NClJlZmVyZW5jZXM6IDx3ZWR2NC44MTkxNSRveDUuMjIwMDgzMjZAdHcxMS5ubi5iY2FuZGlk LmNvbT4gPHdtb3Y0LjIkVTU0LjEzOUBucmVhDQpkZXIyLmtwbnF3ZXN0Lm5ldD4gPEs0d3Y0Ljgz ODI1JG94NS4yMjUyNzIxNEB0dzExLm5uLmJjYW5kaWQuY29tPiA8MzhiZjA5Y2UuNjc1Mw0KNzI2 QG5ld3Muc3diZWxsLm5ldD4gPDg5bjV2MiRoaXIkMUBubnJwMS5kZWphLmNvbT4gPDM4YmY1MDVk LjI0ODE5Mzc0QG5ld3Muc3diZWwNCmwubmV0PiA8VUNTdjQNCjg2NjI5JG94NS4yMzE2MjU3OUB0 dzExLm5uLmJjYW5kaWQuY29tPg0KDQpEb2VzIGFueW9uZSBoYXZlIGEgc2FtcGxlIGZvciBjaGVj a2luZyBmcmVlIGJ5dGVzIGluIGEgc29ja2V0IHVzaW5nIHNlbGVjdD8NCg0KVGhhbngsDQoNCk1p Y2hhZWwNCg0KIk1pY2hhZWwgTXVycGh5IiA8bW11cnBoeUBpYm1sLmNvbT4gd3JvdGUgaW4gbWVz c2FnZQ0KbmV3czpVQ1N2NC44NjYyOSRveDUuMjMxNjI1NzlAdHcxMS5ubi5iY2FuZGlkLmNvbS4u Lg0KPiBZZXMsIEknYW0gdHJ5aW5nIHRvIGdldCB0aGUgZnJlZSBieXRlcyBsZWZ0IGluIHRoZSBi dWZmZXIuIEkgaGF2ZSB0cmllZA0KPiBpb2N0cmwoIHNvY2tudW0sIEZJT05SRUFELCAmbkJ5dGVz KS4gSXQgYWx3YXlzIHJldHVybnMgemVybyBpbiBuQnl0ZXMuDQo+DQo+IE1pY2hhZWwNCj4NCj4g IkJydWNlIiA8c25hcGhvb2tAc291dGh3ZXN0ZXJuYmVsbC5uZXQ+IHdyb3RlIGluIG1lc3NhZ2UN Cj4gbmV3czozOGJmNTA1ZC4yNDgxOTM3NEBuZXdzLnN3YmVsbC5uZXQuLi4NCj4gPiBJbiBjb21w Lm9zLnZ4d29ya3MNCj4gPiBEYW5pZWwgV29uZyA8ZGFuaWVsX2tfd29uZ0BteS1kZWphLmNvbT4g d3JvdGU6DQo+ID4NCj4gPiA+V2hpY2ggb3B0aW9uIHdvdWxkIHlvdSB1c2Ugd2l0aCBpb2N0bCgp IHRvIGdldCB0aGUgbnVtYmVyIG9mIGJ5dGVzIHRoYXQNCj4gPiA+YXJlIGN1cnJlbnRseSBpbiB0 aGUgc2VuZCBidWZmZXIgYW5kIHdoYXQncyB0aGUgc3ludGF4IGZvciB0aGF0IG9wdGlvbj8NCj4g Pg0KPiA+IEkgZG9uJ3QgaGF2ZSBteSBkZXZlbG9wbWVudCBzeXN0ZW0gd2l0aCBtZSwgYnV0IEkg d2FzIGxvb2tpbmcgYXQgc2VyaWFsDQo+ID4gY29kZSBsaWtlIHRoaXMgdG9uaWdodC4gIEkgYmVs aWV2ZSBpdCB3YXM6DQo+ID4NCj4gPiBpb2N0bCggc29ja251bSwgRklPTlJFQUQsICZuQnl0ZXMp Ow0KPiA+DQo+ID4gQW5kIHRoZSBudW1iZXIgdG8gYmUgcmVhZCB3YXMgc3RvcmVkIGluIG5CeXRl cy4gIFlvdSBtaWdodCBhbHNvIGhhdmUgYQ0KPiBsb29rDQo+ID4gYXQgc2VsZWN0KCkgaWYgdGhp cyBkb2Vzbid0IHdvcmsuDQo+ID4NCj4gPiBCcnVjZQ0KPiA+DQo+DQo+DQoNCg0KDQotLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52eHdvcmtzDQpTdWJq ZWN0OiB1Y21wX2Vycm9yIGluIFQyDQpEYXRlOiBGcmksIDAzIE1hciAyMDAwIDE3OjM4OjM3IEdN VA0KRnJvbTogZmF0aWhvemNhbkBteS1kZWphLmNvbQ0KT3JnYW5pemF0aW9uOiBEZWphLmNvbSAt IEJlZm9yZSB5b3UgYnV5Lg0KTWVzc2FnZS1JRDogPDg5b3RhdCRvamkkMUBubnJwMS5kZWphLmNv bT4NCg0KV2hlbiB3b3JraW5nIHdpdGggVDEgaWNtcF9lcnJvciBpcyBhIGRlZmluZWQgaW4gbGli IGJ1dCBtZXJnaW5nIHRvIFQyDQp3ZSBoYXZlIG5vIHByb3RvdHlwZSBmb3IgaXQuIEFuZCB0aGUg c291cmNlIGNvZGUgb2YgaXQgaXMgYXZhaWxhYmxlIGluDQp0aGUgaW50ZXJuZXQgYnV0IHRoZSBw cm90b3R5cGVzIHRoYXQgaXQgY2FsbHMgZGlmZmVycyBmcm9tIHRoZSBUMiBoYXMNCmRlZmluZWQg aW4geHhfaWNtcC5oIGZpbGVzLiBDYW4geW91IGhlbHAgbWU/DQoNCg0KU2VudCB2aWEgRGVqYS5j b20gaHR0cDovL3d3dy5kZWphLmNvbS8NCkJlZm9yZSB5b3UgYnV5Lg0KDQotLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52eHdvcmtzDQpTdWJqZWN0OiBQ U09TIC8vIFZ4V29ya3MNCkRhdGU6IEZyaSwgMDMgTWFyIDIwMDAgMTM6MjI6MTYgLTA4MDANCkZy b206IEhhcnZleSBUYXlsb3IgPGhldEBkZXNwYW0ucGFuZ2VhLmNhPg0KT3JnYW5pemF0aW9uOiBP cmdhbml6YXRpb24/IFdoYXQgb3JnYW5pemF0aW9uPyENCk1lc3NhZ2UtSUQ6IDwzOEMwMkQwOC42 RTZBQGRlc3BhbS5wYW5nZWEuY2E+DQoNClNvbWUgaGVyZSBtYXkgZmluZCB0aGlzIHN0b3J5IHJl bGV2YW50Li4uDQogICAgIGh0dHA6Ly93d3cuZWV0aW1lcy5jb20vc3RvcnkvT0VHMjAwMDAzMDJT MDAzOA0KLSAtaGV0DQoNCg0KLSAtLS0NCiJBc3NlbWJseSBvZiBKYXBhbmVzZSBiaWN5Y2xlIHJl cXVpcmUgZ3JlYXQgcGVhY2Ugb2YgbWluZC4iDQogICAgIC0gSW5zdHJ1Y3Rpb24gbWFudWFsIG1l bnRpb25lZCBieSBSLiBQaXJzaWcNCg0KSGFydmV5IFRheWxvciAgbWFpbHRvOmhldEBkZXNwYW0u cG9ydGFsLmNhICBodHRwOi8vd3d3LnBhbmdlYS5jYS9+aGV0DQoNCi0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLQ0KDQpOZXdzZ3JvdXBzOiBjb21wLm9zLnZ4d29ya3MNClN1YmplY3Q6IFNubXAg Tk9USUZJQ0FUSU9OLVRZUEUNCkRhdGU6IEZyaSwgMDMgTWFyIDIwMDAgMTg6MDg6NDggR01UDQpG cm9tOiBybGl1QGludGVyLXRlbC5jb20NCk9yZ2FuaXphdGlvbjogRGVqYS5jb20gLSBCZWZvcmUg eW91IGJ1eS4NCk1lc3NhZ2UtSUQ6IDw4OW92M2IkcTBtJDFAbm5ycDEuZGVqYS5jb20+DQoNCkkg YW0gdXNpbmcgV2luZE5ldCBTTk1QdjEvdjJjIEFnZW50IGFuZCBpdHMgbWliY29tcC4gSGVyZSBp cyBteSBtaWIgZm9yDQpub3RpZmljYXRpb25zOg0KDQpzZFRyYXBzICAgICAgICAgT0JKRUNUIElE RU5USUZJRVIgOjo9IHsgbXlwcm9kdWN0IDQgfQ0Kc3lzQWNjZXNzRGVuaWVkIE5PVElGSUNBVElP Ti1UWVBFDQogICAgICAgICAgICAgICAgU1RBVFVTICBjdXJyZW50DQogICAgICAgICAgICAgICAg REVTQ1JJUFRJT04NCiAgICAgICAgICAgICAgICAgICAgICAgICAiLiINCiAgICAgICAgICAgICAg ICA6Oj0geyBzZFRyYXBzIDEgfQ0KDQpXaGVuIGl0IGlzIGNvbXBpbGVkLCBJIGdvdCB0aGlzIHdh cm5pbmcNCi4vc25tcC9zdXBlcmRzcC5taWI6ODU0OiAgV2FybmluZzogIFRoZSBuZXh0LXRvLWxh c3Qgc3ViLWlkZW50aWZpZXIgaW4NCnRoZSBuYW1lIG9mIGFueSBuZXdseS1kZWZpbmVkIG5vdGlm aWNhdGlvbiBtdXN0IGhhdmUgdGhlIHZhbHVlIHplcm8uDQoNCkNhbiBhbnlvbmUgdGVsbCBtZSB3 aGF0IHRoaXMgbWVhbnM/DQoNClRoYW5rcw0KDQoNCg0KDQoNClNlbnQgdmlhIERlamEuY29tIGh0 dHA6Ly93d3cuZGVqYS5jb20vDQpCZWZvcmUgeW91IGJ1eS4NCg0KLS0tLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tDQoNCk5ld3Nncm91cHM6IGNvbXAub3Mudnh3b3Jrcw0KU3ViamVjdDogUmU6IEdl dCBTb2NrZXQgU2VuZCBCdWZmZXIgU2l6ZT9sDQpEYXRlOiBGcmksIDAzIE1hciAyMDAwIDIwOjI3 OjI2IEdNVA0KRnJvbTogYndlZGRpbmdAbXktZGVqYS5jb20NCk9yZ2FuaXphdGlvbjogRGVqYS5j b20gLSBCZWZvcmUgeW91IGJ1eS4NCk1lc3NhZ2UtSUQ6IDw4OXA3NzgkaHMkMUBubnJwMS5kZWph LmNvbT4NClJlZmVyZW5jZXM6IDx3ZWR2NC44MTkxNSRveDUuMjIwMDgzMjZAdHcxMS5ubi5iY2Fu ZGlkLmNvbT4gPHdtb3Y0LjIkVTU0LjEzOUBucmVhDQpkZXIyLmtwbnF3ZXN0Lm5ldD4gPEs0d3Y0 LjgzODI1JG94NS4yMjUyNzIxNEB0dzExLm5uLmJjYW5kaWQuY29tPiA8MzhiZjA5Y2UuNjc1Mw0K NzI2QG5ld3Muc3diZWxsLm5ldD4gPDg5bjV2MiRoaXIkMUBubnJwMS5kZWphLmNvbT4gPDM4YmY1 MDVkLjI0ODE5Mzc0QG5ld3Muc3diZWwNCmwubmV0PiA8VUNTdjQNCjg2NjI5JG94NS4yMzE2MjU3 OUB0dzExLm5uLmJjYW5kaWQuY29tPg0KDQpJbiBhcnRpY2xlIDxVQ1N2NC44NjYyOSRveDUuMjMx NjI1NzlAdHcxMS5ubi5iY2FuZGlkLmNvbT4sDQogICJNaWNoYWVsIE11cnBoeSIgPG1tdXJwaHlA aWJtbC5jb20+IHdyb3RlOg0KDQo+IFllcywgSSdhbSB0cnlpbmcgdG8gZ2V0IHRoZSBmcmVlIGJ5 dGVzIGxlZnQgaW4gdGhlIGJ1ZmZlci4gSSBoYXZlDQo+IHRyaWVkIGlvY3RybCggc29ja251bSwg RklPTlJFQUQsICZuQnl0ZXMpLiBJdCBhbHdheXMgcmV0dXJucyB6ZXJvIGluDQo+bkJ5dGVzLg0K DQpUaGF0IGlzIGlvY3RsKCkgY29ycmVjdD8gIFRoYXQgc2hvdWxkIHdvcmsgYWNjb3JkaW5nIHRv IHRoZSBtYW51YWwuICBJdA0Kc2F5cyBhIHNvY2tldCBpcyBhbiBvcmRpbmFyeSBGRCBsaWtlIGFu eSBvdGhlci4NCg0KSXQgbG9va3MgbGlrZSBzZWxlY3QoKSB3b24ndCB3b3JrLCBidXQgdGhpcyBz aG91bGQ6DQoNCmNoYXIgdG1wQnVmW1JFQ19CVUZfU1pdOw0KDQpuQnl0ZXMgPSByZWN2KCBzb2Nr bnVtLCB0bXBCdWYsIFJFQ19CVUZfU1osIE1TR19QRUVLKTsNCg0KVGhlIHByb2JsZW0gaXMgdGhl IGRhdGEgaXMgY29waWVkIGludG8gdGhlIHRtcEJ1ZiwgYnV0IGl0IGlzbid0IHJlbW92ZWQNCmZy b20gdGhlIHNvY2tldCByZWNlaXZlIGJ1ZmZlci4NCg0KQnJ1Y2UNCg0KDQoNClNlbnQgdmlhIERl amEuY29tIGh0dHA6Ly93d3cuZGVqYS5jb20vDQpCZWZvcmUgeW91IGJ1eS4NCg0KLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tDQoNCk5ld3Nncm91cHM6IGNvbXAub3Mudnh3b3Jrcw0KU3ViamVj dDogRUNQL0VQUCBzdXBwb3J0LiBBbnkgaGludHM/DQpEYXRlOiBGcmksIDAzIE1hciAyMDAwIDIw OjM4OjU5ICswMTAwDQpGcm9tOiAiSGVuayBXLkEuTS4gdmFuIFJvb2lqIiA8SC5XLkEuTS52LlJv b2lqQHR1ZS5ubD4NCk9yZ2FuaXphdGlvbjogRWluZGhvdmVuIFVuaXZlcnNpdHkgb2YgVGVjaG5v bG9neQ0KTWVzc2FnZS1JRDogPDM4QzAxNEQzLjg4MUM4RkRAdHVlLm5sPg0KDQpIaSwNCg0KV2Ug d2FudCB0byB1c2UgdGhlIHBhcmFsbGVsIHBvcnQgb2YgYW4gb3JkaW5hcnkgUEMgKDQ2OCBvciBQ MSkgaW4gRUNQIG9yDQpFUFAgbW9kZS4gT2ZmaWNpYWxseSBFUFAvRUNQIGlzIG5vdCBzdXBwb3J0 ZWQgYnkgVldvcmtzLg0KDQpEbyBhbnkgb2YgeW91IGhhdmUgYW55IGV4cGVyaWVuY2UgKGdvb2Qg b3IgYmFkKSBpbiB1c2luZyBFUFAvRUNQIHVuZGVyDQpWeFdvcmtzPw0KDQpDYW4gaXQgYmUgZG9u ZSBlYXNpbHksIG9yIGlzIGl0IGEgZGlmZmljdWx0IHRhc2sgdG8gd3JpdGUgYSBkcml2ZXIgZm9y DQpFQ1AvRVBQIHVzZT8NCg0KVGhhbmtzIGZvciB5b3VyIGhlbHAuDQoNCkhlbmsuDQoNCi0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0tLQ0KDQpOZXdzZ3JvdXBzOiBjb21wLm9zLnZ4d29ya3MNClN1 YmplY3Q6IEhvdyB0byBzZXQgSVAgYWRkcmVzcyBvbiBhIHJ1bm5pbmcgc3lzdGVtDQpEYXRlOiBG cmksIDMgTWFyIDIwMDAgMjI6MTY6MzIgKzAxMDANCkZyb206ICLYeXZpbmQgSG9sbWVpZGUiIDxv bnRpbWVAZnJpc3VyZi5ubz4NCk9yZ2FuaXphdGlvbjogVGVsZW5vciBPbmxpbmUgUHVibGljIEFj Y2Vzcw0KTWVzc2FnZS1JRDogPFJQVnY0LjU3OTYkNmIxLjkzNDI0QG5ld3MxLm9ubGluZS5ubz4N Cg0KDQpUd28gcXVlc3Rpb25zOg0KDQoxLikgSG93IGNhbiB0aGUgSVAgYWRkcmVzcyBiZSBjaGFu Z2VkIG9uIGEgcnVubmluZyBzeXN0ZW0uDQoNCjIuKSBIb3cgY2FuIHRoZSBNQUMgYWRkcmVzcyBi ZSBleHRyYWN0ZWQ/DQoNCnJlZ2FyZHMNCth5dmluZCBIb2xtZWlkZQ0KDQoNCg0KDQotLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52eHdvcmtzDQpTdWJq ZWN0OiBSZTogSG93IHRvIGdldCBjdXJyZW50IHRpbWUgJiBkYXRlIGZvciB0aW1lc3RhbXBzIGZv ciBET1MtRlMgZmlsZXMNCkRhdGU6IEZyaSwgMyBNYXIgMjAwMCAyMjozNjozOSArMDEwMA0KRnJv bTogIlpha2VsaWprIiA8bGVlbmRlcnQuYmllbWFuc0BpY3Qubmw+DQpPcmdhbml6YXRpb246IFVV TkVULU5MIChodHRwOi8vd3d3Lm5sLnV1Lm5ldCkNCk1lc3NhZ2UtSUQ6IDw4OXBiOWkkdDc3JDFA cG9ydGhvcy5ubC51dS5uZXQ+DQpSZWZlcmVuY2VzOiA8MzhiYjljZjEuOTU2MjA5OUBuZXdzLnJv bmV0LmRlPiA8MzhCQzEwRjEuNEYwN0BkZXNwYW0ucGFuZ2VhLmNhPg0KDQpIYXJ2ZXkuDQoNCklu IHRoZSBtZXNzYWdlIGJlbG93IHUgbWVudGlvbiB0aGF0IHUgc2hvdWxkIHVzZSBjbG9ja19zZXR0 aW1lKCkgdG8gc2V0IHRoZQ0Kc3lzdGVtIHRpbWUuDQpUaGF0J3MgYWxsIG5pY2UgZXRjLiBidXQg d2hlcmUgY2FuIGkgZ2V0IGZvciBleGFtcGxlIHRoZSB0aW1lIGZyb20gdGhlIGJpb3MuDQoNCldo YXQgaSBsaWtlIHRvIGRvOg0KLSAtIGhhdmluZyBhIHN0YW5kYWxvbmUgKFBDUEVudGl1bSBib2Fy ZGVkKSBWWFdvcmtzIG1hY2hpbmUgd2hpY2ggY291bGQNCmF1dG9ib290IHdpdGhvdXQgaW50ZXJm ZXJlbmNlIGZyb20gdGhlIG91dHNpZGUuDQoNCihvciByYXRoZXIgaWYgaSB0cmllZCB0byBkbyBh IHJlYm9vdCBpIHdvdWxkIHdhbnQgdG8gbW92ZSB0byB0aGUgbG9jYXRpb24gb2YNCnRoZSBtYWNo aW5lIHRvIHJlc2V0IHRoZSB0aW1lIGFuZCBkYXRlKS4NCg0KRG8gdSAgIChvciBhbnlvbmUgZWxz ZSkga25vd3MgaG93IHRvIHNldCB0aGUgdGltZSBvbiBzdWNoIGEgc3RhbmRhbG9uZQ0Kc3lzdGVt Lg0KDQp0aHguLiBMZWVuZGVydCBCaWVtYW5zDQoNCkhhcnZleSBUYXlsb3IgPGhldEBkZXNwYW0u cGFuZ2VhLmNhPiB3cm90ZSBpbiBtZXNzYWdlDQpuZXdzOjM4QkMxMEYxLjRGMDdAZGVzcGFtLnBh bmdlYS5jYS4uLg0KPiBNYW5mcmVkIEZpc2NoZXIgd3JvdGU6DQo+ID4gSSdtIHdvcmtpbmcgd2l0 aCBhIE1WMjcwMCBQb3dlclBDIGJvYXJkIHdpdGggYSBET1MgRlMgb24gYSBoYXJkZGlzay4NCj4g PiBJcyB0aGVyZSBhIHBvc3NpYmlsdHkgdG8gZ2V0IGN1cnJlbnQgdGltZSBhbmQgZGF0ZSBhcyB0 aW1lc3RhbXAgZm9yDQo+ID4gbmV3IGNyZWF0ZWQgb3IgY2hhbmdlZCBmaWxlcy4NCj4gPg0KPiA+ IElzIHRoZXJlIGEgc3BlY2lhbCBtZWNoYW5pc20gdG8gYXBwbHkgYSB0aW1lc3RhbXAgYXV0b21h dGljYWxseSBvbg0KPiA+IGZpbGVzLg0KPiA+DQo+DQo+IFllcyBsb29rIGF0IHRoZSBkb3NGc0xp YiBkb2NzLiAgVGhlIGZ1bmN0aW9uIGlzIGNhbGxlZA0KPiBkb3NGc0RhdGVUaW1lSW5zdGFsbCgp Lg0KPg0KPiBJdCByZXF1aXJlcyBhIGhhbmRsZXIgZnVuY3Rpb24gd2hpY2ggdHlwaWNhbGx5IGxv b2tzIGxpa2U6DQo+DQo+IFNUQVRVUyBEYXRlVGltZUZ1bmMoRE9TX0RBVEVfVElNRSAqIHBEb3NE YXRlVGltZSkNCj4gew0KPiB0aW1lX3QgdE5vdyA9IHRpbWUoTlVMTCk7DQo+IHN0cnVjdCB0bSog dCA9IGxvY2FsdGltZSgmdE5vdyk7DQo+DQo+IHBEb3NEYXRlVGltZS0+ZG9zZHRfeWVhciAgID0g dC0+dG1feWVhciA7DQo+IHBEb3NEYXRlVGltZS0+ZG9zZHRfeWVhciArPSAoICh0LT50bV95ZWFy ID4gNzApID8gMTkwMCA6IDIwMDApOw0KPiBwRG9zRGF0ZVRpbWUtPmRvc2R0X21vbnRoICA9IHQt PnRtX21vbiArIDEgIDsgLyogTWFwIDAuLjExIHRvIDEuLjEyICovDQo+IHBEb3NEYXRlVGltZS0+ ZG9zZHRfZGF5ICAgID0gdC0+dG1fbWRheSA7DQo+IHBEb3NEYXRlVGltZS0+ZG9zZHRfaG91ciAg ID0gdC0+dG1faG91ciA7DQo+IHBEb3NEYXRlVGltZS0+ZG9zZHRfbWludXRlID0gdC0+dG1fbWlu ICA7DQo+IHBEb3NEYXRlVGltZS0+ZG9zZHRfc2Vjb25kID0gdC0+dG1fc2VjICA7DQo+DQo+IHJl dHVybiggT0sgKSA7DQo+IH0NCj4NCj4gWW91IHNob3VsZCBwcm9iYWJseSBhZGQgZXJyb3IgaGFu ZGxpbmcgdG8gdGFzdGUuDQo+IE9oIGFuZCBvZiBjb3VzcmUgeW91IG11c3QgaGF2ZSBkb25lIGEg Y2xvY2tfc2V0dGltZSgpIHRvDQo+IG1ha2UgdGhpcyB3b3JrLg0KPg0KPiA8bDhyPg0KPiAtaGV0 DQo+DQo+DQo+DQo+DQo+IC0tLQ0KPiAiQXNzZW1ibHkgb2YgSmFwYW5lc2UgYmljeWNsZSByZXF1 aXJlIGdyZWF0IHBlYWNlIG9mIG1pbmQuIg0KPiAtIEluc3RydWN0aW9uIG1hbnVhbCBtZW50aW9u ZWQgYnkgUi4gUGlyc2lnDQo+DQo+IEhhcnZleSBUYXlsb3IgIG1haWx0bzpoZXRAZGVzcGFtLnBv cnRhbC5jYSAgaHR0cDovL3d3dy5wYW5nZWEuY2EvfmhldA0KDQoNCg0KLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tDQoNCk5ld3Nncm91cHM6IGNvbXAub3Mudnh3b3Jrcw0KU3ViamVjdDogUmU6 IEhvdyB0byBnZXQgY3VycmVudCB0aW1lICYgZGF0ZSBmb3IgdGltZXN0YW1wcyBmb3IgRE9TLUZT IGZpbGVzDQpEYXRlOiBGcmksIDAzIE1hciAyMDAwIDE2OjIxOjE0IC0wODAwDQpGcm9tOiBIYXJ2 ZXkgVGF5bG9yIDxoZXRAZGVzcGFtLnBhbmdlYS5jYT4NCk9yZ2FuaXphdGlvbjogT3JnYW5pemF0 aW9uPyBXaGF0IG9yZ2FuaXphdGlvbj8hDQpNZXNzYWdlLUlEOiA8MzhDMDU2RkEuMTA3MUBkZXNw YW0ucGFuZ2VhLmNhPg0KUmVmZXJlbmNlczogPDM4YmI5Y2YxLjk1NjIwOTlAbmV3cy5yb25ldC5k ZT4gPDM4QkMxMEYxLjRGMDdAZGVzcGFtLnBhbmdlYS5jYT4gPDgNCjlwYjlpJHQ3NyQxQHBvcnRo b3MubmwudXUubmV0Pg0KDQpaYWtlbGlqayB3cm90ZToNCj4gSW4gdGhlIG1lc3NhZ2UgYmVsb3cg dSBtZW50aW9uIHRoYXQgdSBzaG91bGQgdXNlIGNsb2NrX3NldHRpbWUoKSB0byBzZXQgdGhlDQo+ IHN5c3RlbSB0aW1lLg0KPiBUaGF0J3MgYWxsIG5pY2UgZXRjLiBidXQgd2hlcmUgY2FuIGkgZ2V0 IGZvciBleGFtcGxlIHRoZSB0aW1lIGZyb20gdGhlIGJpb3MuDQo+DQoNCkhpICBMZWVuZGVydCwN CiAgICAgSSBhbSBub3QgZmFtaWxpYXIgd2l0aCB0aGUgTVYyNzAwIFBvd2VyUEMsIGJ1dCBpdCBk ZXBlbmRzIHVwb24NCiAgICAgeW91ciBwYXJ0aWN1bGFyIFJUQyBjaGlwLiAgSWYgaXQgdXNlcyBz b21ldGhpbmcgZGlmZmVyZW50LCB5b3UNCiAgICAgd2lsbCBoYXZlIHRvIHdyaXRlIHNvbWUgUlRD IHJlYWRpbmcgY29kZSB5b3Vyc2VsZiAob3IgYnVnIHRoZQ0KICAgICBNVjI3MDAgc3VwcGxpZXIp LiAgSWYgaXQgbG9va3MgbGlrZSBhIGdlbmVyaWMgUEMsIHlvdSBjYW4gdXNlDQogICAgIExpbnV4 IHNvdXJjZSAob3Igc29tZXRoaW5nIGxpa2Ugc25pcHBldHMub3JnKS4NCiAgICAgSSdtIHNvcnJ5 IGJ1dCBJIGRvbid0IGhhcHBlbiB0byBoYXZlIGEgc2FtcGxlIHRvIGhhbmQuDQoNCiAgICAgQW55 Ym9keT8NCjxjb3JkaWFsbHk+DQotIC1oZXQNCg0KDQoNCj4gV2hhdCBpIGxpa2UgdG8gZG86DQo+ IC0gaGF2aW5nIGEgc3RhbmRhbG9uZSAoUENQRW50aXVtIGJvYXJkZWQpIFZYV29ya3MgbWFjaGlu ZSB3aGljaCBjb3VsZA0KPiBhdXRvYm9vdCB3aXRob3V0IGludGVyZmVyZW5jZSBmcm9tIHRoZSBv dXRzaWRlLg0KPg0KPiAob3IgcmF0aGVyIGlmIGkgdHJpZWQgdG8gZG8gYSByZWJvb3QgaSB3b3Vs ZCB3YW50IHRvIG1vdmUgdG8gdGhlIGxvY2F0aW9uIG9mDQo+IHRoZSBtYWNoaW5lIHRvIHJlc2V0 IHRoZSB0aW1lIGFuZCBkYXRlKS4NCj4NCj4gRG8gdSAgIChvciBhbnlvbmUgZWxzZSkga25vd3Mg aG93IHRvIHNldCB0aGUgdGltZSBvbiBzdWNoIGEgc3RhbmRhbG9uZQ0KPiBzeXN0ZW0uDQo+DQo+ IHRoeC4uIExlZW5kZXJ0IEJpZW1hbnMNCj4NCj4gSGFydmV5IFRheWxvciA8aGV0QGRlc3BhbS5w YW5nZWEuY2E+IHdyb3RlIGluIG1lc3NhZ2UNCj4gbmV3czozOEJDMTBGMS40RjA3QGRlc3BhbS5w YW5nZWEuY2EuLg0KPiA+IE1hbmZyZWQgRmlzY2hlciB3cm90ZToNCj4gPiA+IEknbSB3b3JraW5n IHdpdGggYSBNVjI3MDAgUG93ZXJQQyBib2FyZCB3aXRoIGEgRE9TIEZTIG9uIGEgaGFyZGRpc2su DQo+ID4gPiBJcyB0aGVyZSBhIHBvc3NpYmlsdHkgdG8gZ2V0IGN1cnJlbnQgdGltZSBhbmQgZGF0 ZSBhcyB0aW1lc3RhbXAgZm9yDQo+ID4gPiBuZXcgY3JlYXRlZCBvciBjaGFuZ2VkIGZpbGVzLg0K PiA+ID4NCj4gPiA+IElzIHRoZXJlIGEgc3BlY2lhbCBtZWNoYW5pc20gdG8gYXBwbHkgYSB0aW1l c3RhbXAgYXV0b21hdGljYWxseSBvbg0KPiA+ID4gZmlsZXMuDQo+ID4gPg0KPiA+DQo+ID4gWWVz IGxvb2sgYXQgdGhlIGRvc0ZzTGliIGRvY3MuICBUaGUgZnVuY3Rpb24gaXMgY2FsbGVkDQo+ID4g ZG9zRnNEYXRlVGltZUluc3RhbGwoKS4NCj4gPg0KPiA+IEl0IHJlcXVpcmVzIGEgaGFuZGxlciBm dW5jdGlvbiB3aGljaCB0eXBpY2FsbHkgbG9va3MgbGlrZToNCj4gPg0KPiA+IFNUQVRVUyBEYXRl VGltZUZ1bmMoRE9TX0RBVEVfVElNRSAqIHBEb3NEYXRlVGltZSkNCj4gPiB7DQo+ID4gdGltZV90 IHROb3cgPSB0aW1lKE5VTEwpOw0KPiA+IHN0cnVjdCB0bSogdCA9IGxvY2FsdGltZSgmdE5vdyk7 DQo+ID4NCj4gPiBwRG9zRGF0ZVRpbWUtPmRvc2R0X3llYXIgICA9IHQtPnRtX3llYXIgOw0KPiA+ IHBEb3NEYXRlVGltZS0+ZG9zZHRfeWVhciArPSAoICh0LT50bV95ZWFyID4gNzApID8gMTkwMCA6 IDIwMDApOw0KPiA+IHBEb3NEYXRlVGltZS0+ZG9zZHRfbW9udGggID0gdC0+dG1fbW9uICsgMSAg OyAvKiBNYXAgMC4uMTEgdG8gMS4uMTIgKi8NCj4gPiBwRG9zRGF0ZVRpbWUtPmRvc2R0X2RheSAg ICA9IHQtPnRtX21kYXkgOw0KPiA+IHBEb3NEYXRlVGltZS0+ZG9zZHRfaG91ciAgID0gdC0+dG1f aG91ciA7DQo+ID4gcERvc0RhdGVUaW1lLT5kb3NkdF9taW51dGUgPSB0LT50bV9taW4gIDsNCj4g PiBwRG9zRGF0ZVRpbWUtPmRvc2R0X3NlY29uZCA9IHQtPnRtX3NlYyAgOw0KPiA+DQo+ID4gcmV0 dXJuKCBPSyApIDsNCj4gPiB9DQo+ID4NCj4gPiBZb3Ugc2hvdWxkIHByb2JhYmx5IGFkZCBlcnJv ciBoYW5kbGluZyB0byB0YXN0ZS4NCj4gPiBPaCBhbmQgb2YgY291c3JlIHlvdSBtdXN0IGhhdmUg ZG9uZSBhIGNsb2NrX3NldHRpbWUoKSB0bw0KPiA+IG1ha2UgdGhpcyB3b3JrLg0KPiA+DQo+ID4g PGw4cj4NCj4gPiAtaGV0DQo+ID4NCj4gPg0KPiA+DQo+ID4NCj4gPiAtLS0NCj4gPiAiQXNzZW1i bHkgb2YgSmFwYW5lc2UgYmljeWNsZSByZXF1aXJlIGdyZWF0IHBlYWNlIG9mIG1pbmQuIg0KPiA+ IC0gSW5zdHJ1Y3Rpb24gbWFudWFsIG1lbnRpb25lZCBieSBSLiBQaXJzaWcNCj4gPg0KPiA+IEhh cnZleSBUYXlsb3IgIG1haWx0bzpoZXRAZGVzcGFtLnBvcnRhbC5jYSAgaHR0cDovL3d3dy5wYW5n ZWEuY2EvfmhldA0KDQotIC0tDQoNCi0gLS0tDQoiQXNzZW1ibHkgb2YgSmFwYW5lc2UgYmljeWNs ZSByZXF1aXJlIGdyZWF0IHBlYWNlIG9mIG1pbmQuIg0KICAgICAtIEluc3RydWN0aW9uIG1hbnVh bCBtZW50aW9uZWQgYnkgUi4gUGlyc2lnDQoNCkhhcnZleSBUYXlsb3IgIG1haWx0bzpoZXRAZGVz cGFtLnBvcnRhbC5jYSAgaHR0cDovL3d3dy5wYW5nZWEuY2EvfmhldA0KDQotLS0tLS0tLS0tLS0t LS0tLS0tLS0tLS0tLS0NCg0KTmV3c2dyb3VwczogY29tcC5vcy52eHdvcmtzDQpTdWJqZWN0OiBj aGFuZ2luZyB0aGUgQlNQIE1ha2VmaWxlDQpEYXRlOiBTYXQsIDA0IE1hciAyMDAwIDA1OjI3OjE1 IEdNVA0KRnJvbTogIkRyYWdhbiBPc3RvamljIiA8b3N0b2ppY0BzeW1wYXRpY28uY2E+DQpPcmdh bml6YXRpb246IFN5bXBhdGljbw0KTWVzc2FnZS1JRDogPFQ4MXc0LjIxMTc2JHBmNC43NDEyNTZA bmV3czIwLmJlbGxnbG9iYWwuY29tPg0KDQpJIG1vdmVkIHBvcnRpb25zIG9mIEJTUCBNYWtlZmls ZSBpbnRvIGFub3RoZXIgZmlsZSBhbmQgdGhlbiBJIGluY2x1ZGVkIGl0IGluDQp0aGUgc2FtZSBN YWtlZmlsZS4gVGhlIGlkZWEgaXMgdG8ga2VlcCBib2FyZCBzcGVjaWZpYyBkZWZpbml0aW9ucyBz ZXBhcmF0ZQ0KZnJvbSB0aGUgY29yZSBCU1AuIFdoZW4gSSB0cmllZCB0byBidWlsZCB0aGUgcHJv amVjdCBiYXNlZCBvbiBzdWNoIGEgQlNQLA0KZmluYWwgcHJvamVjdCBNYWtlZmlsZSB3b3VsZCBp Z25vcmUgbWFjcm9zIHRoYXQgd2VyZSBpbmNsdWRlZCBpbnN0ZWFkIG9mDQppbmxpbmVkLiBEb2Vz IGFueW9uZSBrbm93IHdoYXQgaXMgZ29pbmcgb24/DQoNCkRyYWdhbi4NCg0KDQoNCi0tLS0tLS0t LS0tLS0tLS0tLS0tLS0tLS0tLQ0KDQpOZXdzZ3JvdXBzOiBjb21wLm9zLnZ4d29ya3MNClN1Ympl Y3Q6IFJlOiBbT1RdIFJlOiBSZWFsLXRpbWUgRGV2ZWxvcGVyDQpEYXRlOiBTYXQsIDA0IE1hciAy MDAwIDA5OjE5OjEzIEdNVA0KRnJvbTogcHdoaWNrZXJAaG9tZS5jb20gKFBhdWwgV2hpY2tlcikN Ck9yZ2FuaXphdGlvbjogQEhvbWUgTmV0d29yaw0KTWVzc2FnZS1JRDogPGx5NHc0LjM2MDckV0kx LjUyMjA4QG5ld3MxLmZybXQxLnNmYmEuaG9tZS5jb20+DQpSZWZlcmVuY2VzOiA8ZE1BdjQuNjAx NSRMMy4xNjgyMjg0NUBuZXdzLm1hZ21hLmNhPiA8M01JdjQuMTg4MCRXSTEuMjMxMDhAbmV3czEu Zg0Kcm10MS5zZmJhLmhvbWUuY29tPiA8dXUyaW9uaWQ0LmZzZl8tX0B5YWhvby5jb20+DQoNClJl YWx0aW1lLCB0byBtZSwgYWxzbyBtZWFucyBkZXRlcm1pbmlzdGljLiBJIHdhbnQgdG8ga25vdyB0 aGF0IHdoZW4gSSBleHBlY3QNCnNvbWV0aGluZyB0byBoYXBwZW4gdGhlbiBpdCB3aWxsIHRha2Ug YSBwYXJ0aWN1bGFyIHRpbWUuIEEgZGF0YWJhc2UganVzdCBpc24ndA0KcmVhbHRpbWUgYmVjYXVz ZSBpdCB3aWxsIHRha2UgaG93ZXZlciBsb25nIGl0IHRha2VzLCB0aGUgcGVyZm9ybWFuY2UgaXMg bm90DQpyZWxpYWJsZS4NCg0KcGF1bA0KDQpJbiBhcnRpY2xlIDx1dTJpb25pZDQuZnNmXy1fQHlh aG9vLmNvbT4sIEJpbGwgUHJpbmdsZW1laXINCjxicHJpbmdsZW1laXJAeWFob28uY29tPiB3cm90 ZToNCj4NCj5JdCBpcyByZWFsLXRpbWU7IGEgcmVhbGx5IGxvbmcgdGltZS4gIFJlYWwtdGltZSBn ZW5lcmFsbHkgbWVhbnMNCj50aGUgYWJpbGl0eSB0byBtZWV0IGEgZGVhZCBsaW5lIChjb3VsZCBi ZSBoYXJkIG9yIHNvZnQpLiAgSXQgaXMNCj5hIG11Y2ggbWFsaWduZWQgcGhyYXNlLiAgWW91IGNh biB1c2Ugd2l0aCBvdGhlciB3b3JkcyBvZiBpdHMgaWxrLA0KPmxpa2Ugb2JqZWN0IG9yaWVudGVk LCBSQUQsIHdpcmVsZXNzIEludGVybmV0IGFuZCBKYXZhIERPTS9TQVggWE1MDQo+cG9seW1vcnBo aWMgZGF0YSBkZXZpY2UuDQo+DQo+Q29sbGVjdCBtYW55IHBocmFzZXMgbGlrZSB0aGlzLCBwdXQg dGhlbSB0b2dldGhlciB3aXRoIGEgcmFuZG9tDQo+a2V5IGFuZCB5b3UgaGF2ZSBjcmVhdGVkIHlv dXIgYXZlcmFnZSBoaWdoIHRlY2ggbWFya2V0aW5nIGdlbml1cw0KPndpdGggYW4gTUJBLiAgVGhl eSBzYXkgdGhpbmdzIGxpa2UgJ0NhbGN1bHVzLCB3aGF0J3MgdGhhdD8nIGFuZA0KPidXaG8gY2Fy ZXMgYWJvdXQgcGh5c2ljcycuDQo+DQo+QmlsbA0KPg0KPj4+Pj4+ICJQYXVsIiA9PSBQYXVsIFdo aWNrZXIgPHB3aGlja2VyQGhvbWUuY29tPiB3cml0ZXM6DQo+DQo+ICAgIFBhdWw+IEdvb2RuZXNz ISBSZWFsLXRpbWUgc3liYXNlLCBpc24ndCB0aGF0IGFuIG94eW1vcm9uPyBXaGF0ZXZlcg0KPiAg ICBQYXVsPiBpcyB0aGUgd29ybGQgY29taW5nIHRvPw0KPg0KPj4+Pj4gIlRhbGVudExBQiBJbmMu IiAgPGluZm9AdGFsZW50bGFiLmNvbT4gd3JvdGU6DQo+ICAgID4+ICBPdXIgQ2xpZW50IGlzIGxv b2tpbmcgZm9yIGEgc29mdHdhcmUgZGV2ZWxvcGVyIHdobyBjYW4gd29yayBpbg0KPiAgICA+PiBh IHJlYWwtdGltZSBVTklYIGVudmlyb25tZW50LiBXb3JraW5nIHdpdGggdGhlIGV4aXN0aW5nDQo+ ICAgID4+IGRhdGFiYXNlIChzeWJhc2UpIHJ1bm5pbmcgb24gYSBVTklYIHBsYXRmb3JtLCB0aGUg c29mdHdhcmUNCj4gICAgPj4gZGV2ZWxvcGVyIHdpbGwgd29yayB0byBvcHRpbWl6ZSB0aGUgcGVy Zm9ybWFuY2Ugb2YgdGhlIGRhdGFiYXNlDQo+ICAgID4+IHVzaW5nIEMrKy4gVGhlIHN1Y2Nlc3Nm dWwgY2FuZGlkYXRlIHdpbGwgaGF2ZSB0aGUgZXhwZXJpZW5jZQ0KPiAgICA+PiB3aXRoIGVtYmVk ZGVkIFNRTCBhbmQgdW5kZXJzdGFuZCBsYXJnZSBzeXN0ZW1zLiBUaGUgcHJvZHVjdCBpcw0KPiAg ICA+PiBhIHNhdGVsbGl0ZSB0cmFja2luZyBkZXZpY2UgdXNlZCBpbiB0aGUgdHJ1Y2tpbmcgaW5k dXN0cnkuIFRoZQ0KPiAgICA+PiBkYXRhYmFzZSBpcyB1c2VkIGluIHRoZSBjb21tdW5pY2F0aW9u IG9mIGluZm9ybWF0aW9uLg0KPg0KDQotLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0NCg0KRW5k IG9mIE5ldy1OZXdzIGRpZ2VzdA0KKioqKioqKioqKioqKioqKioqKioqKg0KDQoqKioqKioqKioq DQoNCiAgICBUaGlzIGlzIGEgdXNlciBncm91cCBtYWlsaW5nIGxpc3QgZm9yIHZ4V29ya3MgcmVs YXRlZCB0b3BpY3MNCiAgICBzZWUgaHR0cDovL2NzZy5sYmwuZ292L3Z4d29ya3MgZm9yIHBvc3Rp bmcgYW5kIG90aGVyIGluZm8NCg0KLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLSBPcmlnaW5hbCBt ZXNzYWdlIGhlYWRlcjoNCj5NQUlMIEZST006PHZ4X2V4cGxvQGNzZy5sYmwuZ292Pg0KPlJDUFQg VE86PGpyb3dsZXlAdmlhc2F0LmNvbT4NCj5EQVRBDQo+UmVjZWl2ZWQ6IChmcm9tIHZ4X2V4cGxv QGxvY2FsaG9zdCkgYnkgY3NnLmxibC5nb3YgKDguOS4zLzguOS4zKSBpZCBFQUEyMDgyNSBmbw0K ciB2eGxpc3Q5Mzc1OTsgU2F0LCA0IE1hciAyMDAwIDA0OjEyOjAxIC0wODAwIChQU1QpDQo+RGF0 ZTogU2F0LCA0IE1hciAyMDAwIDA0OjEyOjAxIC0wODAwIChQU1QpDQo+TWVzc2FnZS1JZDo8MjAw MDAzMDQxMjEyLkVBQTIwODI1QGNzZy5sYmwuZ292Pg0KPkVycm9ycy1Ubzogdnh3ZXhwbG8tZXJy c0BsYmwuZ292DQo+VG86IHZ4d29ya3NfdXNlcnNAY3NnLmxibC5nb3YNCj5Gcm9tOiB2eHdleHBs b0BsYmwuZ292ICh0aGUgdnhXb3JrcyBVc2VycyBHcm91cCBFeHBsb2RlcikNCj5TdWJqZWN0OiBj b21wLm9zLnZ4d29ya3MgbmV3c2RpZ2VzdA0KPlByZWNlZGVuY2U6IGJ1bGsNCj5YLUNvbW1lbnQ6 IFZ4V29ya3MgVXNlcnMgR3JvdXAgRGlzY3Vzc2lvbiBFeHBsb2Rlcg0KLS0tLS0tLS0tLS0tLS0t LS0tLS0tLS0tLSBFbmQgb2YgbWVzc2FnZSBoZWFkZXIuDQp= --=_sb2000/03/06-03-04-01-538=_sb-- From vxwexplo-errs@csg.lbl.gov Mon Mar 6 04:03:09 2000 From: Vxworks Exploder Date: Mon Mar 6 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Mar 6 04:03:06 PST 2000 Subject: To hide password while typing on the command prompt Subject: óÏÔÒÕÄÎÉÞÅÓÔ×Ï × ÏÂÌÁÓÔÉ E-commerse Subject: Re: [OT] Re: Real-time Developer Subject: Re: fcntl () call in VxWorks Subject: Re: vxworks for dummies Subject: Re: Very simple question on open() and fopen() failure Subject: Re: changing the BSP Makefile Subject: WEBCAST MARCH 7: LAUNCH OF MAP-CA Subject: porting on sa1110 Subject: Re: vxworks for dummies : URGENT!! Subject: About parallel problem? Subject: Re: [OT] Re: Real-time Developer Subject: Re: Snmp NOTIFICATION-TYPE Subject: Re: How to set IP address on a running system Subject: Re: vxworks for dummies : URGENT!! ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: To hide password while typing on the command prompt Date: Sun, 05 Mar 2000 18:33:32 +0530 From: k m prasanna Organization: MIEL Message-ID: <38C25B24.351FCE73@miel.mot.com> hi, I have written my own application to run on the tatrget(860 board). To interact with this application user has to login (through telnet or serial port) with a valid password ( syntax is >>>login mypassword ). While typing the password i want just the ***** to appear or nothing at all so that others will not see it. How to do it? thanks and regards pras --------------------------- Newsgroups: comp.os.vxworks Subject: óÏÔÒÕÄÎÉÞÅÓÔ×Ï × ÏÂÌÁÓÔÉ E-commerse Date: Sun, 5 Mar 2000 15:47:32 +0200 From: "Lilya Galushchenko" Organization: Message-ID: <952264905.122694@ipt2.iptelecom.net.ua> óÏÚÄÁÎÉÅ ÓÏÂÓÔ×ÅÎÎÏÇÏ ÉÎÔÅÒÎÅÔ-ÍÁÇÁÚÉÎÁ Ó ÇÏÔÏ×ÏÊ ÓÉÓÔÅÍÏÊ ÒÁÓÞÅÔÏ× É ÄÏÓÔÁ×ËÉ ÔÏ×ÁÒÁ ÐÏ óîç. ðÏÄÒÏÂÎÏÓÔÉ ÐÏÓÌÅ ÒÅÇÉÓÔÒÁÃÉÉ ÎÁ www.intershop.txnet.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: [OT] Re: Real-time Developer Date: 05 Mar 2000 11:55:56 -0500 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <3MIv4.1880$WI1.23108@news1.frmt1.sfba.home.com> Sender: bpringle@DeadDuck A database can be 'deterministic'. That is the entire point of calling it real-time. However, the processing time may be large. You must specify the number of records in the data base to guarantee the same time for each operation. Use of heap-sort instead of quick sort will guarantee queries can be done in a 'deterministic' way. Actually everything on a computer is 'deterministic', strictly speaking. I think you want the upper and lower time for functions to be bound, or data independent. It must have worst case/best case of the same order... it may take seconds to execute though. Bill >>>>> "Paul" == Paul Whicker writes: Paul> Realtime, to me, also means deterministic. I want to know Paul> that when I expect something to happen then it will take a Paul> particular time. A database just isn't realtime because it Paul> will take however long it takes, the performance is not Paul> reliable. - -- Yes, Gnus does random signatures. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: fcntl () call in VxWorks Date: Sun, 05 Mar 2000 13:55:17 -0500 From: Joe Durusau Message-ID: <38C2AD95.F9B50B37@bellsouth.net> References: <89qgnv$bo33@overload.lbl.gov> The function does not exist in vxworks. You would have to write it yourself. There is something similar, I believe called fioctl which is in vxworks. You might see if that has what you want. An easy trick to seeing what is in your bsp is: get to a shell and type lkup "name" For example, lkup "fcntl" would tell you that your copy of vxworks doesn't have fcntl. You could get source for fcntl from linux or some such and see if you could adapt it to your situation. Good luck. Speaking only myself, Joe Durusau Navin Kumar Agrawal wrote: > > Tornado > > Hi Folks, > > How do we use fcntl () system call in a vxworks based application > program? > Do I have to include some specific include file to call this function? > > I used fcntl () call in one of my application program. While compiling > there > was no problem. But while downloading the .out file on the target, it > gave me > 'fcntl' symbol unresolved error. Can anybody help me in solving this? > > Regards, > Navin --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies Date: 5 Mar 2000 22:03:48 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <89ulk4$eld$1@news1.xs4all.nl> References: <89ot9u$oj9$1@nnrp1.deja.com> venksh@my-deja.com wrote: : I am a unix person and have no knowledge in realtime OS. Can anyone : give me pointers to tutorials/guides. Take a look at http://www.smartmodulartech.com/systems/library/docs/vxworks/vxintro/VXINTRO.HTM This gives an introduction into VxWorks for VxWorks for Alpha. But it is also a general introduction into realtime SW and VxWorks. P.S. This page did not work correct until recently. It seems to work again however. Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Mon, 6 Mar 2000 08:39:50 +0800 From: "joke" Organization: The news server of Changzhou TeleCom Message-ID: <89uund$jrc$1@news.cz.js.cn> References: <38C19191.5B7859E5@pacbell.net> you can use fopen("host:/YourFileDirectory/YourFile","r"). ntPassFileSystem is needed. type iosDevShow in the shell host = your disk that tornado installed . Marco Zandonadi wrote in message news:38C19191.5B7859E5@pacbell.net... > Hello, > I have VXWorks 5.4 and my target is VXSim for Windows. My problem is > that open() and fopen() always fail. Here's what I did. > > - I created a new project > - I double clicked "Create a bootable VXWorks image (custom configured)" > - I pressed Next > - I selected "An existing project" and pressed Next > - I pressed Finish > > - I double clicked on userAppInit.c and modified it as follows: > > #include > ... > void usrAppInit (void) > { > ... > FILE * f = fopen ("myfile", "r"); > if (!f) > printf ("File not found!\n"); > else > printf ("File found!\n"); > } > > - I selected Build (compilation works ok) > - I opened Object Modules, right clicked on vxWorks.exe and selected > Start 'vxWorks.exe' > > The simulator starts properly and prints out File not found! even though > myfile is in the same directory as vxWorks.exe. I also tried opening > other files like c:/autoexec.bat but the result is the same. Note that I > didn't touch the default VxWorks component configuration. > > Does anybody know what I'm missing? > > Thanks very much in advance. > > Marco Zandonadi > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: changing the BSP Makefile Date: Sun, 05 Mar 2000 18:02:30 -0700 From: James Scott Organization: Newsfeeds.com http://www.newsfeeds.com 60,000+ UNCENSORED Newsgroups. Message-ID: <38C303A5.DA026741@amug.org> References: <38C151F0.F251CC7A@amug.org> Sounds like what you need to do is make a new "BSP" for your specific hardware or configuration. You already have a base or reference BSP, so copy that whole directory to some other name, and modify the BSP specific *.h files for your hardware and/or configuration. For example I use the MBX860 BSP as a starting point for PowerPC 8xx designs. I copy and rename the BSP directory then I modify the MBX860.h file and config.h files as well as other files related to the changes I need. I have enabled the Tornado 1 type builds and use those when I build a new BSP, then when I get it to where I want it, I make the Tornado 2 project and work on applications from there. Tornado 2 seems very unforgiving when you modify configuration and build rules without using the configuration manager. Good Luck, James Dragan Ostojic wrote: > > I'm using Tornado 2. I'm sorry I was a bit vague in my description. What I > did is I modified BSP and then used poject manager to create project > Makefile. Project manager left those parameters that were included undifined > in the generated project Makefile. I would like to import through the BSP > automaticaly certain number of parameters before I start to change project > settings. > > -- > Dragan Ostojic > ostojic@sympatico.ca > > "James Scott" wrote in message > news:38C151F0.F251CC7A@amug.org... > > Are you using Tornado 1 or 2? Builds in Tornado 1 seem to > > follow "normal" make rules and such. Builds in Tornado 2 > > happen through the configuration manager, and the rules > > are setup when you define a project based upon a BSP. When > > you modify a BSP you need to re-create the project for the > > BSP to build properly. > > > > James > > > > Dragan Ostojic wrote: > > > > > > I moved portions of BSP Makefile into another file and then I included > it in > > > the same Makefile. The idea is to keep board specific definitions > separate > > > from the core BSP. When I tried to build the project based on such a > BSP, > > > final project Makefile would ignore macros that were included instead of > > > inlined. Does anyone know what is going on? > > > > > > Dragan. > > > > > > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- > > http://www.newsfeeds.com - The #1 Newsgroup Service in the World! > > -----== Over 80,000 Newsgroups - 16 Different Servers! =----- - -----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- http://www.newsfeeds.com - The #1 Newsgroup Service in the World! - -----== Over 80,000 Newsgroups - 16 Different Servers! =----- --------------------------- Newsgroups: comp.os.vxworks Subject: WEBCAST MARCH 7: LAUNCH OF MAP-CA Date: Sun, 05 Mar 2000 20:27:07 -0800 From: John Setel O'Donnell Organization: Equator Technologies, Inc. Message-ID: <38C3339B.1710A4E8@equator.com> Reply-To: map-ca@equator.com Greetings! Please join us on Tuesday, March 7th, for a webcast as Equator Technologies and Hitachi Ltd., proudly launch the MAP-CA Media Accelerated Processor for Consumer Appliances. The MAP-CA, a 30 GOP (Giga-Operations Per Second) processor, sets a new world record for DSP/Image Processing performance and price-performance. It supports software HDTV and software Personal Video Recorders, creating products that lead in value, flexibility and longevity. By moving the central functions of television and imaging products into software, the MAP-CA will usher in a revolution in television, and bring new media functions to the TV platform. Join us to see and hear why Cisco, Xerox, Canon, Videotron, Rogers, Snell&Wilcox, WindRiver, RealNetworks, Interval Research, and so many other leading companies support this new platform for Internet-connected imaging-intensive devices. What: Equator Technologies and Hitachi Ltd. webcast to launch the MAP-CA Media Accelerated Processor for Consumer Appliances. When: March 7th at 2 pm (Japan time); 10 am (Paris time); 10 am (Pacific Time) Where: http://www.equator.com John Setel O’Donnell, President of Equator Technologies, and Shigemichi Matsuka, Executive Vice President of Hitachi, will introduce the MAP-CA, followed by industry leaders describing how they plan to use the device in breakthrough products. Journalists who wish to set up one-on-one phone interviews, please contact Tom Mahon at tmahon@ncal.verio.com or 925.937.4921. If you unable to participate in the webcast at this time, the event will be archived at www.equator.com. An e-press kit is available there, as well. A minimum requirement to join the webcast is RealPlayer software, downloadable free from www.real.com/products/player/index.html, and at least a 28.8 Kbps connection to the Internet. If you experience problems during the webcast, please send an email to map-ca@equator.com. This same email address will be used for press Q&A immediately following the webcast. --------------------------- Newsgroups: comp.os.vxworks Subject: porting on sa1110 Date: Mon, 6 Mar 2000 13:43:05 +0900 From: "Kyoung doo Hong" Organization: ONSE Telecom Message-ID: <89vd1d$epl$1@nntp.shinbiro.com> Hi everybody, I'm trying to port vxworks on a customized board based on sa1110. Well, there are some problem running bootrom file on this board, First, when using bootrom.aif( compressed version ), the inflate routine always returns ERROR. i.e, inflate fails. Second, when using bootrom_res.aif( rom resident version ), the system hanged inside excVecInit() routine in usrInit(). When I skipped excVecInit(), the system hanged inside kernelInit(). I have 16MBytes SDRAM in bank 0, so the system has 16MBytes memory in contiguous block, It has mapped from address 0 in virtual memory. I worked based on the bsp of brutus board. There is a report about some bugs of SA1110 cpu, I think one of them is critical - - command LDRSB/LDRSH will fail after read buffer is allocated. But I'm not sure this bug causes the errors... Is there anybody who has some experience porting vxworks on sa1110 board? Thanks, Kyoung doo Hong --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies : URGENT!! Date: Mon, 06 Mar 2000 06:07:42 GMT From: ilike_dsps@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <89vhvc$561$1@nnrp1.deja.com> References: <89ot9u$oj9$1@nnrp1.deja.com> <89ulk4$eld$1@news1.xs4all.nl> In article <89ulk4$eld$1@news1.xs4all.nl>, johan.borkhuis@peektraffic.nl wrote: > Take a look at > http://www.smartmodulartech.com/systems/library/docs/vxworks/vxintro/VXI NTRO.HTM > This gives an introduction into VxWorks for VxWorks for Alpha. But it is > also a general introduction into realtime SW and VxWorks. > This page did not work correct until recently. It seems to work again > however. Thanx...it works...I'm also new to Vxworks...I want to know more about developing/tuning BSPs on PowerPC Vxworks platform - any help ? This is a bit URGETNT !! Thanx in adavnce.. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: About parallel problem? Date: Mon, 6 Mar 2000 15:02:28 +0800 From: "huge rock" Organization: The news server of Changzhou TeleCom Message-ID: <89vl29$1f16$1@news.cz.js.cn> Hi all, I am using the standard (WRS) lptDrv.c to drive a parallel printer on an 80486 target system. Everything appears to be working right,but the fact that the printer won't print ,Why? And the lptDrv not Work,The part program appear below: char *c= "123456789"; LPT_RESOURCE pResource; pResource.ioBase = 0x378; pResource.intVector = 0x08;(advisedly using 0x08) pResource.intLevel = (0x00+0x07); pResource.autofeed = TRUE; pResource.busyWait = 10000; pResource.strobeWait = 10000; pResource.retryCnt = 1; pResource.timeout = 1; if(lptDrv(0, &pResource) == ERROR) { printf("Init Printer ERROR!\n"); } if(lptDevCreate ("/lpt/0", 0) == ERROR) { printf("The Printer not Install!\n"); } fp = open ("/lpt/0", O_CREAT|O_WRONLY , 0); if(fp == ERROR) { printf("open IO Error!\n"); } if(write(fp,c,strlen(c)) == ERROR) { printf("Write Data to IO ERROR!\n"); close(fp); }... I am using lptShow to get the result forestall run lptDevCreate: controlReg = 0xff statusReg = 0xff created = FALSE autofeed = TRUE inservice = FALSE normalInt = 0 defaultInt = 0 retryCnt = 1 busyWait (loop) = 10000 strobeWait (loop) = 10000 timeout (sec) = 1 intLevel (IRQ) = 7 value = 22 = 0x16 Runing lptDevCreate,Why get hold of same result?Why isn't intLevel (IRQ) = 8? Why lptDrv never called? Thinks! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: [OT] Re: Real-time Developer Date: Mon, 06 Mar 2000 10:35:45 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38C36DE1.965FE5EE@bitband.com> References: <3MIv4.1880$WI1.23108@news1.frmt1.sfba.home.com> Reply-To: leonid@bitband.com Actually, for a database search to be deterministic, you must have a limit on number of seach keys and length of each key. True that key length puts a limit on total number of records, but that limit can be high enough for all practical purposes. More precisely, a tree-like structure of database index lets you traverse at most K elements, each of N bits in order to find an N*K bits long key. There are also formulae for calculating the size of the INDEX needed to perform such searches, whence the number of different keys is importantr. In other words, a database can be deterministic, and if RAM resident Index is used it can be very fast indeed. Examples of such fast data bases are Credit Card Verification, and SS7 switching node or ... backbone IP router ! Hope it helps, Leonid Bill Pringlemeir wrote: > A database can be 'deterministic'. That is the entire point of > calling it real-time. However, the processing time may be large. > You must specify the number of records in the data base to > guarantee the same time for each operation. Use of heap-sort > instead of quick sort will guarantee queries can be done in a > 'deterministic' way. > > Actually everything on a computer is 'deterministic', strictly > speaking. I think you want the upper and lower time for functions > to be bound, or data independent. It must have worst case/best > case of the same order... it may take seconds to execute though. > > Bill > > >>>>> "Paul" == Paul Whicker writes: > > Paul> Realtime, to me, also means deterministic. I want to know > Paul> that when I expect something to happen then it will take a > Paul> particular time. A database just isn't realtime because it > Paul> will take however long it takes, the performance is not > Paul> reliable. > > -- > Yes, Gnus does random signatures. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Snmp NOTIFICATION-TYPE Date: Mon, 6 Mar 2000 08:48:12 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <89vpog$ft0$1@ndsukns01.ndsuk.com> References: <89ov3b$q0m$1@nnrp1.deja.com> It means that you need to re-phrase your definition something like this: sdTraps OBJECT IDENTIFIER ::= { myproduct 4 } sdTrapsPrefix OBJECT IDENTIFIER ::= { sdTraps 0 } sysAccessDenied NOTIFICATION-TYPE STATUS current DESCRIPTION "." ::= { sdTrapsPrefix 1 } Don't ask me _why_ it's like this, it just _is_! (A bit like the apparently unnecessary 'TableEntry' node in a table definition). HTH Will rliu@inter-tel.com wrote in message <89ov3b$q0m$1@nnrp1.deja.com>... >I am using WindNet SNMPv1/v2c Agent and its mibcomp. Here is my mib for >notifications: > >sdTraps OBJECT IDENTIFIER ::= { myproduct 4 } >sysAccessDenied NOTIFICATION-TYPE > STATUS current > DESCRIPTION > "." > ::= { sdTraps 1 } > >When it is compiled, I got this warning >./snmp/superdsp.mib:854: Warning: The next-to-last sub-identifier in >the name of any newly-defined notification must have the value zero. > >Can anyone tell me what this means? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to set IP address on a running system Date: Mon, 6 Mar 2000 09:01:11 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <89vqgs$g68$1@ndsukns01.ndsuk.com> References: Øyvind Holmeide wrote in message ... > >Two questions: > >1.) How can the IP address be changed on a running system. See ifAddrSet(...) but note that if you want to set a subnet mask, you must do so with ifMaskSet(...) before setting the Address. >2.) How can the MAC address be extracted? I think this is BSP-specific. I'm using a PPC860SAR and the if_cpm Ethernet driver, which means that the BSP provides a function sysCpmEnetAddrGet(...) that returns the MAC address. See the help for your driver library. Alternatively, you could use the m2lib function m2IfTblEntryGet(...) to retrieve a pointer to the interfaces table entry for the interface. This contains the MAC address as ifPhysAddress. HTH, Will --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies : URGENT!! Date: 6 Mar 2000 11:55:36 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8a06bo$6qt$1@news1.xs4all.nl> References: <89ot9u$oj9$1@nnrp1.deja.com> <89ulk4$eld$1@news1.xs4all.nl> <89vhvc$561$1@nnrp1.deja.com> ilike_dsps@my-deja.com wrote: : Thanx...it works...I'm also new to Vxworks...I want to know more about : developing/tuning BSPs on PowerPC Vxworks platform - any help ? : This is a bit URGETNT !! If you want to do development on BSPs the BSP Development Kit is probably best. If you want to modify an existing BSP this can also be a good choice, but often it also works if you just start modifying an existing 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 === --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Mar 6 05:26:39 2000 From: "Blake, Don" Date: Mon Mar 6 05:26:41 PST 2000 Subject: RE: Get Socket Send Buffer Size Michael, VxWorks error 0x20 (decimal 32) is "Broken pipe" (EPIPE). It sounds like a problem on the receiving side. Error 0x23 (decimal 35) is "Unsupported value" (ENOTSUP). I don't understand this one. Don Blake Sr. Programmer Lockheed Martin Federal Systems Owego, NY -----Original Message----- Newsgroups: comp.os.vxworks Subject: Get Socket Send Buffer Size?l Date: Wed, 01 Mar 2000 18:23:24 GMT From: "Michael Murphy" Hi All, I'm send large buffers via TCP using send(). After a few MB or so, send will return an error 0x20. I'm assuming this is a Send Buffer Full Error. Does anyone know how to peek into send buffer to get its current size? All calls to send() afterwards return an error 0x23. Anyone know where I can get info on these error codes? Thanx, Michael From vxwexplo-errs@csg.lbl.gov Tue Mar 7 02:00:24 2000 From: susheel behl Date: Tue Mar 7 02:00:26 PST 2000 Subject: Lm4548 codec/ sound card driver We have bought & are using Vxworks. The hardwarespec: x86platform: cyrix/64mb orion refernce board. We are trying to play PCM file. Codec lm4548 requires SMI handling. But the SMI handler already exists. How do we attach the handler for it. Can we play PCM wave file at any sampling rate & shall be able to hear sound from Codec The audio busmaters are playing the auddiobuffers but no sound comes. We have choosen PCM file as test sample to play. Pleas provide solution. Regards susheel __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com From vxwexplo-errs@csg.lbl.gov Tue Mar 7 04:03:14 2000 From: Vxworks Exploder Date: Tue Mar 7 04:03:17 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Mar 7 04:03:11 PST 2000 Subject: getc(stdin) only returns when enter (\n) is pressed Subject: intConnect, intHandlerCreate and PPC Subject: What is the Interrupt Service Routines (ISR)? Subject: Errno 3d0004 observed Subject: Re: Very simple question on open() and fopen() failure Subject: Re: Free source code for any RTOS ? Subject: Re: getc(stdin) only returns when enter (\n) is pressed Subject: Re: booting from a SCSI harddrive Subject: Re: sysBusIntGen Subject: Any Zinc users out there? Subject: Re: intConnect, intHandlerCreate and PPC Subject: Re: Very simple question on open() and fopen() failure Subject: Re: Very simple question on open() and fopen() failure Subject: TCP socket send buffer Subject: Re: fcntl () call in VxWorks Subject: Re: intConnect, intHandlerCreate and PPC Subject: Problem booting bootROM Subject: Re: Very simple question on open() and fopen() failure Subject: Re: Very simple question on open() and fopen() failure Subject: Re: To hide password while typing on the command prompt Subject: Re: getc(stdin) only returns when enter (\n) is pressed Subject: Re: Very simple question on open() and fopen() failure Subject: about tornadoII Subject: Re: Problem booting bootROM Subject: Re: What is the Interrupt Service Routines (ISR)? Subject: Re: vxworks for dummies : URGENT!! Subject: WEB server Subject: Re: intConnect, intHandlerCreate and PPC Subject: Re: Snmp NOTIFICATION-TYPE Subject: Re: WEB server Subject: Re: vxworks for dummies : URGENT!! Subject: Re: getc(stdin) only returns when enter (\n) is pressed Subject: VME Interfacing ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: getc(stdin) only returns when enter (\n) is pressed Date: Mon, 06 Mar 2000 13:20:33 GMT From: "Øyvind Holmeide" Organization: Telia Internet Public Access Message-ID: I am using the getc(stdin) or getchar() routine in an application that expect input from the target shell. However, the routine(s) only returns when enter is pressed. How can this be avoided so that each character typed is receveid one by one? regards Øyvind Holmeide --------------------------- Newsgroups: comp.os.vxworks Subject: intConnect, intHandlerCreate and PPC Date: Mon, 06 Mar 2000 16:31:35 +0000 From: Dave Moore Organization: BAE SYSTEMS Message-ID: <38C3DD67.F814FEA3@gecm.com> Hi everyone, I'm experiencing difficulty changing ISRs on a MCP750 PPC Card. Basically, multiple calls to intConnect appear to add the referenced isr function to existing calls. For example calling intConnect twice with the same ISR means that when the interrupt is caught, the isr is executed twice. This is undesirable as re-running a piece of code which uses intConnect doubles up the calls to the isr. What I'm after is a intDisconnect call but there isn't one. I either need an equivalent of intConnect that can be run multiple times using the same isr without resultant multiple calls or a method of clearing the effect of intConnect. I've seen several instances of code using a combination of intHandlerCreate() and intVecSet() to get around this problem but the function intHandlerCreate() doesn't appear to be in my PPC bsp!. Anybody got any ideas - Thanks in advance. Dave. --------------------------- Newsgroups: comp.os.vxworks Subject: What is the Interrupt Service Routines (ISR)? Date: Mon, 6 Mar 2000 17:03:33 -0600 From: "Rao" Organization: Esat Net Message-ID: <38c3e4b1.0@seat.s3two.ie> Hi to everyone, I'm far away from the advanced features of RTOS and close to basics. I'm a beginner of VxWorks....please tell me what is ISR or what for it is? Is that routines provided by the device driver to perform low-level I/O functions to character oriented devices??? --------------------------- Newsgroups: comp.os.vxworks Subject: Errno 3d0004 observed Date: Mon, 06 Mar 2000 13:21:58 -0500 From: "Michael P. Card" Organization: Lockheed Martin Message-ID: <38C3F746.31C9F64F@lmco.com> This is a multi-part message in MIME format. - --------------27DBD1D200D0F41F20A3A27E Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello everyone! I was wondering if any of you have ever observed an ERRNO of 3d0004 in your task info window. I have a task which I am suspending for while using taskDelay, and when I execute the "ti" command in the WindSh to observe the timer running down, it shows the aforementioned mysterious errno. I looked this up and the module corresponding to 0x003d is objLib, which is not documented anywhere. Errno 4 for this module is "OBJ_TIMEOUT". Do any of you know what this means? I couldn't find anything in the documentation about it. - - Mike - --------------27DBD1D200D0F41F20A3A27E Content-Type: text/x-vcard; charset=us-ascii; name="michael.p.card.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Michael P. Card Content-Disposition: attachment; filename="michael.p.card.vcf" begin:vcard n:Card;Michael tel;fax:315-456-0441 tel;work:315-456-3022 x-mozilla-html:TRUE org:Lockheed Martin ;Ocean, Radar, and Sensor Systems version:2.1 email;internet:michael.p.card@lmco.com title:Principal Software Engineer adr;quoted-printable:;;Electronics Park=0D=0ABuilding 6, Room 201;Syracuse;NY;13221;USA fn:Michael Card end:vcard - --------------27DBD1D200D0F41F20A3A27E-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Mon, 06 Mar 2000 12:13:56 -0800 From: Marco Zandonadi Organization: SBC Internet Services Message-ID: <38C41184.437FB45@nospampacbell.net> References: <38C19191.5B7859E5@pacbell.net> <89uund$jrc$1@news.cz.js.cn> > you can use fopen("host:/YourFileDirectory/YourFile","r"). > > ntPassFileSystem is needed. > type iosDevShow in the shell > host = your disk that tornado installed . Thank you very much, that's what I needed. I still have one more problem, though. I can successfully create files with creat() but I can't write() to them. Every write() call on a file descriptor returned by creat() returns -1. Any suggestions will be greatly appreciated. Thank you Marco --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Free source code for any RTOS ? Date: Mon, 06 Mar 2000 21:49:04 +0100 From: Marcel van der Bilt Organization: A2000 Kabeltelevisie en Telecommunicatie Message-ID: <38C41986.10EFD0AD@fibre.A2000.nl> References: <38B5DD8B.6E89B2F9@blr.pin.philips.com> Reply-To: M.v.d.Bilt@fibre.A2000.nl look at the site of mike predo or at www.8051.com 0r www.8052.com M van der Bilt vikram khandpur wrote: > Hi, > I have to write a basic RTOs for 8051 platform. > Is there any free source code for a similar RTOs available for reference > on the web ? > > thanks > vikram --------------------------- Newsgroups: comp.os.vxworks Subject: Re: getc(stdin) only returns when enter (\n) is pressed Date: Mon, 06 Mar 2000 15:46:51 -0500 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <38C4193B.425D0004@lucent.com> References: Are you on a dedicated terminal or are you using a terminal emulation program? I am not sure if this is the problem, but some terminal emulators run in a 'cooked' mode that sends the input after you hit 'Enter'. I have used raw input from within vxWorks, bypassing the target shell, and single character I/O worked. Doug "=D8yvind Holmeide" wrote: > = > I am using the getc(stdin) or getchar() routine in an application that > expect input from the target shell. However, the routine(s) only retur= ns > when enter is pressed. How can this be avoided so that each character = typed > is receveid one by one? > = > regards > = > =D8yvind Holmeide --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting from a SCSI harddrive Date: Mon, 06 Mar 2000 15:08:37 -0600 From: Edsel Harrell Organization: Raytheon Company Message-ID: <38C41E54.71511189@raytheon.com> References: <89jtq0$4so5@overload.lbl.gov> This is a multi-part message in MIME format. - --------------104CCA5F87090A019557C3B9 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I had similar problem... added scsiReset or scsiBusReset call to reset the SCSI bus on each boot attempt... Things seemed to be more consistant... I was using a SCSI PMC card and only power-up reset reset the bus. (I don't have the code or manuals handy... but the call starts with "scsi" and has "reset" in it somewhere...) Edsel John_W_Cosgrove@res.raytheon.com wrote: > I hate followups to my own messages, but I have found out more. > > In scsiLoad(), I found that I reach the first scsiTestUnitRdy() call on > powerup. On the first warm start (^X or reset button), the system now > boots fine. On the second warm start (no matter whether ^X or reset > button) the system reaches the scsiPhysDevCreate() call (for > scsiTestUnitRdy() call) and stops... > > Does any of this sound familiar to anyone? > > John Cosgrove > (401)842-4167 (desk) > (401)842-5740 (lab) > > John Cosgrove writes: > > Hello, Folks, > > I am trying to boot my mvme2700 board from a Seagate ST32430N (2 gig SCSI) > hard drive. I am using vxWorks 5.3.1, BSP 1.1/4.2 (The Motorola patch to > the mv2604 BSP to make a 2700 BSP). > > I am encountering boot failures often with occasional success at the > booting. When it fails it looks like: > > Press any key to stop auto-boot... > 0 > auto-booting... > > boot device : scsi > processor number : 0 > host name : emperor > file name : /sd0/vxworks > inet on ethernet (e) : 192.168.56.113 > host inet (h) : 138.125.194.121 > gateway inet (g) : 192.168.56.1 > user (u) : jwc > flags (f) : 0x8 > target name (tn) : f100-hp-ray > other (o) : dc > > ID LUN VendorID ProductID Rev. Type Blocks BlkSize pScsiPhysDev > -- --- -------- ---------------- ---- ---- -------- ------- ------------ > 0 0 SEAGATE ST32430N 0640 0 4197405 512 0x01ffdcb0 > Can't load boot file!! > > and if it does "attaching to SCSI", everything is fine. The boot device of > scsi or scsi=0,0 both yeild this result! Using the boot device of scsi > sometimes the scsiConfig output is filled in, sometimes it is not. > > Has anyone had a similar problem, or does anyone know of a fix? > > John Cosgrove > (401)842-4167 (desk) > (401)842-5740 (lab) - -- Edsel A. Harrell, Jr (voice) (972) 575-6468 (email) eah@raytheon.com Senior Principle Systems Eng (pager) (972) 597-8874 Raytheon Systems Company (snail) 6600 Chase Oaks Blvd, MS 8454, Plano, Tx 75023 - ------------------------------------------------------------------------------------ "Proofread Carefully to See if You Any Words Out" DISCLAIMER: I speak only for myself, not my employer. EXCEPTION: If any of my ideas are worth more than $0.02, then they are property of my employer. - --------------104CCA5F87090A019557C3B9 Content-Type: text/x-vcard; charset=us-ascii; name="eah.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Edsel Harrell Content-Disposition: attachment; filename="eah.vcf" begin:vcard n:A. Harrell, Jr;Edsel tel;pager:972-597-8874 tel;home:817-430-3903 tel;work:972-575-6468 x-mozilla-html:FALSE org:Raytheon Systems Company;Advanced Airborne EO version:2.1 email;internet:eah@raytheon.com title:Sr Principle Systems Engineer adr;quoted-printable:;;6600 Chase Oaks Blvd, MS 8454 (shipping)=0D=0Aor=0D=0AP.O.Box 660246, MS 8454 (mailing);Plano;TX;75066; x-mozilla-cpt:;0 fn:Edsel end:vcard - --------------104CCA5F87090A019557C3B9-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sysBusIntGen Date: Tue, 29 Feb 2000 09:29:15 -0500 From: Jeff Daly Organization: Sanders A Lockheed Martin Company Message-ID: <38BBD7BA.5A4BCBB9@lmco.com> References: <38BB73DC.54D2FE30@lin.foa.se> Reply-To: jeffrey.j.daly@lmco.com Tomas Larsson wrote: > Hi all! > > We have a configuration with > two MVME2700-boards + VIPC618 > and we generating interrups > through a DUAL/PIT IP module > sitting on the VIP- carrier. > > When trying to > -> sysBusIntGen(2,0x16) > > value = -1 = 0xffffffff > > We have nothing ,that we know > of, which could cause an > VME-bus interrrupt > if you have nothing that can cause a VME interrupt, then calling sysBusIntGen() will return -1 (ERROR). what you're seeing is correct operation. > > Any suggestion is appriciated > Tomas --------------------------- Newsgroups: comp.os.vxworks Subject: Any Zinc users out there? Date: Mon, 06 Mar 2000 16:19:07 -0600 From: Jeffrey Williams Organization: Verio Message-ID: <38C42EDB.93A0D3A7@dnaent.com> Hi. I've been directed to use Zinc on a project. The documentation provided was the "getting started guide" and the reference manual; both are useful, but they hardly constitute a complete set of documentation for a newbie. Does anyone know of additional documentation which I can buy (or get by screaming at Windriver - I feel like having a good scream)? LL&P Jeff --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Tue, 07 Mar 2000 10:27:00 +1100 From: Peter Brown Organization: Australian National University Message-ID: <38C43EC4.2FD15169@syseng.anu.edu.au> References: <38C3DD67.F814FEA3@gecm.com> Dave Moore wrote: > > Hi everyone, > I'm experiencing difficulty changing ISRs on a MCP750 PPC Card. > Basically, multiple calls to intConnect appear to add the referenced isr > function to existing calls. For example calling intConnect twice with > the same ISR means that when the interrupt is caught, the isr is > executed twice. This is undesirable as re-running a piece of code which > uses intConnect doubles up the calls to the isr. What I'm after is a > intDisconnect call but there isn't one. > > I either need an equivalent of intConnect that can be run multiple times > using the same isr without resultant multiple calls or a method of > clearing the effect of intConnect. > It appears that BSP writers are not obliged to include an intDisconnect function. By finding the source code for the BSP intConnect and writing an intDisconnect function we could disconnect ISRs from the list of handlers for a particular IRQ. That way I could repeatedly load, run, re-compile and re-load a device driver without leaving behind an ISR pointer to the wrong address. We're using a Force Pcore740-cpci board and Tornado I. > I've seen several instances of code using a combination of > intHandlerCreate() and intVecSet() to get around this problem but the > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > Anybody got any ideas - Thanks in advance. > > Dave. - -- Peter Brown, Software Engineer | Phone: +61 2 6279 8830 Robotic Systems Lab, RSISE | Fax: +61 2 6279 8660 Australian National University | Canberra ACT 0200 AUSTRALIA | - -------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Tue, 7 Mar 2000 08:53:32 +0800 From: "Harvey Tang" Organization: The news server of Changzhou TeleCom Message-ID: <8a1jt2$u0r$1@news.cz.js.cn> References: <38C19191.5B7859E5@pacbell.net> <89uund$jrc$1@news.cz.js.cn> <38C41184.437FB45@nospampacbell.net> You should create the file by window NT command at first, then you can use creat() and write() functions. I haven't used creat() and write(),I used fopen() and fprintf() etc. Marco Zandonadi wrote in message news:38C41184.437FB45@nospampacbell.net... > > you can use fopen("host:/YourFileDirectory/YourFile","r"). > > > > ntPassFileSystem is needed. > > type iosDevShow in the shell > > host = your disk that tornado installed . > > Thank you very much, that's what I needed. > > I still have one more problem, though. I can successfully create files > with creat() but I can't write() to them. Every write() call on a file > descriptor returned by creat() returns -1. > > Any suggestions will be greatly appreciated. > > Thank you > > Marco > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Mon, 06 Mar 2000 17:56:38 -0800 From: Marco Zandonadi Organization: SBC Internet Services Message-ID: <38C461D6.CBDD38A8@nospampacbell.net> References: <38C19191.5B7859E5@pacbell.net> <89uund$jrc$1@news.cz.js.cn> <38C41184.437FB45@nospampacbell.net> <8a1jt2$u0r$1@news.cz.js.cn> Hi Harvey, > You should create the file by window NT command at first, then you can use > creat() and write() functions. what do you mean by "create the file by window NT command"? If you mean that I should use a DOS shell I can't do it because I need to open several tens of files each time I run my program, so it definitely has to be done inside the program (also these files have different names each time I run my program). Thanks. Marco --------------------------- Newsgroups: comp.os.vxworks Subject: TCP socket send buffer Date: Tue, 07 Mar 2000 02:51:30 GMT From: Daniel Wong Organization: Deja.com - Before you buy. Message-ID: <8a1qri$q60$1@nnrp1.deja.com> Hi All, I have a TCP socket whose send buffer, under certain situations, will fill up to its maximum value. When I try to write to the TCP socket using write() when the send buffer is full, the task will hang while calling write(). The status of the task then becomes PEND+I. I would like to close the socket before its send buffer fills to its maximum. Any suggestions as to how I might go about closing the socket before it fills up? Is there a way to determine the size of the send buffer? ioctl() will tell you the size of the receive buffer but not the send buffer. Thanks in advance. - -- Daniel Wong daniel_k_wong@my-dejan.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: fcntl () call in VxWorks Date: Mon, 06 Mar 2000 13:22:50 -0500 From: Peter Colby Organization: Raytheon Systems Company C3I Message-ID: <38C3F77A.EDB777D0@res.raytheon.com> References: <89qgnv$bo33@overload.lbl.gov> <38C2AD95.F9B50B37@bellsouth.net> ioctl() (for the specific fd type) in vxWorks will cover most of what fcntl() does generically in POSIX Joe Durusau wrote: > > The function does not exist in vxworks. You would > have to write it yourself. There is something similar, > I believe called fioctl which is in vxworks. You might see if > that has what you want. [...] > > Navin Kumar Agrawal wrote: > > > > [...] > > How do we use fcntl () system call in a vxworks based application > > program? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Tue, 07 Mar 2000 04:09:08 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a1vd2$tes$1@nnrp1.deja.com> References: <38C3DD67.F814FEA3@gecm.com> <38C43EC4.2FD15169@syseng.anu.edu.au> Another approach, though perhaps not as clean as Peter's suggestion, is to connect your own wrapper to the interrupt, then call your own routine to attach or detach a new function to your wrapper. Something like this: void myIntRoutine(...) { if (myHandler != NULL) (*myHandler)(...); } void myIntConnect ( FUNCPTR newRoutine ) { myHandler = newRoutine; } Clearly, this has the disadvantage of adding extra instructions in the interrupt processing path, but it might be simpler than writing an intDisconnect() function (and 750s are pretty quick ;-) Regards, John... In article <38C43EC4.2FD15169@syseng.anu.edu.au>, Peter Brown wrote: > Dave Moore wrote: > > > > Hi everyone, > > I'm experiencing difficulty changing ISRs on a MCP750 PPC Card. > > Basically, multiple calls to intConnect appear to add the referenced isr > > function to existing calls. For example calling intConnect twice with > > the same ISR means that when the interrupt is caught, the isr is > > executed twice. This is undesirable as re-running a piece of code which > > uses intConnect doubles up the calls to the isr. What I'm after is a > > intDisconnect call but there isn't one. > > > > I either need an equivalent of intConnect that can be run multiple times > > using the same isr without resultant multiple calls or a method of > > clearing the effect of intConnect. > > > > It appears that BSP writers are not obliged to include an intDisconnect > function. By finding the source code for the BSP intConnect and writing an > intDisconnect function we could disconnect ISRs from the list of handlers for > a particular IRQ. That way I could repeatedly load, run, re-compile and > re-load a device driver without leaving behind an ISR pointer to the wrong > address. We're using a Force Pcore740-cpci board and Tornado I. > > > I've seen several instances of code using a combination of > > intHandlerCreate() and intVecSet() to get around this problem but the > > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > > > Anybody got any ideas - Thanks in advance. > > > > Dave. > > -- > Peter Brown, Software Engineer | Phone: +61 2 6279 8830 > Robotic Systems Lab, RSISE | Fax: +61 2 6279 8660 > Australian National University | > Canberra ACT 0200 AUSTRALIA | > - ------------------------------------------------------------------------ - -- > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Problem booting bootROM Date: Tue, 07 Mar 2000 04:23:44 GMT From: "Brad Ferguson" Organization: Another Optus Customer Message-ID: Hi all, We are developing a custom board based on the Motorola Power PC 860 processor. We have written our own board support package which we are confident is working correctly since we have it operating on a variant of our hardware board. We have traced the software using a disassembler tool and it appears to crash when it enters the kernel initialisation function kernelInit(...). It hangs and never calls the usrRoot() function. We have verified that all interrupts are masked and cleared before entry to kernelInit and that the RAM and ROM are set up correctly and can be accessed. Data caching is also disabled. We would greatly appreciate any ideas / explanations for this behaviour. Our specific details are as follows: Toolset: Tornado 1.0.1 RTOS: VxWorks 5.3.1 Processor: XPC860ENZP33C1 RAM: 4 Mbyte EDO ROM: 512 kbyte PLCC socketed flash Regards, Bradley Ferguson --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Tue, 7 Mar 2000 13:00:02 +0800 From: "Harvey Tang" Organization: The news server of Changzhou TeleCom Message-ID: <8a22bf$1gtm$1@news.cz.js.cn> References: <38C19191.5B7859E5@pacbell.net> <89uund$jrc$1@news.cz.js.cn> <38C41184.437FB45@nospampacbell.net> <8a1jt2$u0r$1@news.cz.js.cn> <38C461D6.CBDD38A8@nospampacbell.net> Yes, you must create the file by DOS shell command ,It seems a bug of VxSim.If your program running at real target board ,It should be ok ,but "host" need be replaced by your hostname that Tornado2 is running . Marco Zandonadi wrote in message news:38C461D6.CBDD38A8@nospampacbell.net... > Hi Harvey, > > > You should create the file by window NT command at first, then you can use > > creat() and write() functions. > > what do you mean by "create the file by window NT command"? If you mean > that I should use a DOS shell I can't do it because I need to open > several tens of files each time I run my program, so it definitely has > to be done inside the program (also these files have different names > each time I run my program). > > Thanks. > > Marco > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Mon, 6 Mar 2000 08:30:56 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8a29k5$12k$1@newsmaster01.nextra.at> References: <38C19191.5B7859E5@pacbell.net> hello, for some information (regarding VxWorks / VxSim bugs) to this topic see me about 5 days old reply to someone here to the subject: Pass Through Filesystem for VxSim you have to use the correct fixes and patches at least. - -- regards, Mario Semo. http://www.kirchnersoft.com Marco Zandonadi wrote in message <38C19191.5B7859E5@pacbell.net>... >Hello, >I have VXWorks 5.4 and my target is VXSim for Windows. My problem is >that open() and fopen() always fail. Here's what I did. > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: To hide password while typing on the command prompt Date: Tue, 7 Mar 2000 08:22:59 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <8a2am1$29e$1@crusher.de.colt.net> References: <38C25B24.351FCE73@miel.mot.com> for use over serial line switch off the ECHO and LINE mode of the UART via IOCTL at the time you expecting the password. Then echo each received character self with " * ". if you detect an CR/LF switch back the two modes. Achim k m prasanna schrieb in im Newsbeitrag: 38C25B24.351FCE73@miel.mot.com... > hi, > I have written my own application to run on the tatrget(860 board). > To interact with this application user has to > login (through telnet or serial port) with a valid > password ( syntax is >>>login mypassword ). > > While typing the password i want just the ***** to appear > or nothing at all so that others will not see it. > How to do it? > > > thanks and regards > pras > > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: getc(stdin) only returns when enter (\n) is pressed Date: Tue, 7 Mar 2000 08:19:19 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8a2and$25l$1@newsmaster01.nextra.at> References: <38C4193B.425D0004@lucent.com> Hello, how are you doing RAW input? i use the CRT getchar() routine, and i, too, see input only after a newline . i work either on the target at the command prompt or in telnet session. - -- regards, Mario Semo. http://www.kirchnersoft.com Douglas Fraser wrote in message <38C4193B.425D0004@lucent.com>... Are you on a dedicated terminal or are you using a terminal emulation program? I am not sure if this is the problem, but some terminal emulators run in a 'cooked' mode that sends the input after you hit 'Enter'. I have used raw input from within vxWorks, bypassing the target shell, and single character I/O worked. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Very simple question on open() and fopen() failure Date: Tue, 7 Mar 2000 08:21:07 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8a2anj$25l$2@newsmaster01.nextra.at> References: <38C19191.5B7859E5@pacbell.net> <89uund$jrc$1@news.cz.js.cn> <38C41184.437FB45@nospampacbell.net> isnt this exactly what i wrote? ok, here is the patch description: SPR 27918 IDE: Component: tornado-vxsim RTOS: Version: Product Name: Tornado Release#: 2.0 Release Status: FCS Products Affected: Tornado Host: Intel x86/Pentium HostOs: HostOSRev: 4.0 Arch Family: SIM NT Arch CPU: SIM NT BSP: Ver: DESCRIPTION Using Tornado 2 FCS on Windows NT, simpc target. The open flags values which VxWorks uses differ from those used by Windows NT, and ntPassFsLib does not convert between them. For instance, the O_CREAT flag has value 0x200 in VxWorks, but value 0x100 in Windows NT. This causes both open (...,O_CREAT | ..., ...) and creat() to fail on the host: ntPassFsLib device. WORKAROUND A patch for this SPR may be downloaded via the Web from the patches section on WindSurf, or by contacting WRS Customer Support. The patch title is, "27918 ntPassFsLib file creation patch". - -- regards, Mario Semo. http://www.kirchnersoft.com Marco Zandonadi wrote in message <38C41184.437FB45@nospampacbell.net>... >> you can use fopen("host:/YourFileDirectory/YourFile","r"). >> >> ntPassFileSystem is needed. >> type iosDevShow in the shell >> host = your disk that tornado installed . > >Thank you very much, that's what I needed. > >I still have one more problem, though. I can successfully create files >with creat() but I can't write() to them. Every write() call on a file >descriptor returned by creat() returns -1. > --------------------------- Newsgroups: comp.os.vxworks Subject: about tornadoII Date: Tue, 7 Mar 2000 15:37:47 +0800 From: "Jazz Lam" Organization: Alcatel Message-ID: <38c593ec.0@dnews.sbell.com.cn> who has dealed with the tornadoII licence? it seems not very convenient --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem booting bootROM Date: Tue, 07 Mar 2000 08:01:22 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a2d0g$6dn$1@nnrp1.deja.com> References: Hi Bradley, > We have traced the software using a disassembler tool and it appears to > crash when it enters the kernel initialisation function kernelInit(...). It > hangs and never calls the usrRoot() function. We have verified that all > interrupts are masked and cleared before entry to kernelInit and that the > RAM and ROM are set up correctly and can be accessed. Data caching is also > disabled. OK, you say that you traced it.. could you say exactly how far into kernelInit it gets? This is the most critical part of the kernel's initialisation since it is the time at which the switch from single thread to multi-tasking takes place. usrRoot is not called per-se; instead it is created as a task using taskInit and then activated with taskActivate. For this to work it is imperative that the parameters you supply to kernelInit are correct, and that you have enough stack space for the initial thread (there is a dummy TCB placed on the stack so you need enough space for that plus some other locals). Did you base your system on a WRS BSP? Feel free to email me at "john_94501 (at) yahoo dot com" [substitute anti-spam words in the obvious places] if you have any questions... I might be able to help! John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: What is the Interrupt Service Routines (ISR)? Date: 7 Mar 2000 08:10:21 GMT From: Geir Frode Raanes Organization: UNINETT news service Message-ID: <8a2dhd$3ee$1@snipp.uninett.no> References: <38c3e4b1.0@seat.s3two.ie> Sender: Geir Frode Raanes Rao wrote: : Hi to everyone, : I'm far away from the advanced features of RTOS and close to : basics. I'm a beginner of VxWorks....please tell me what is ISR or what for : it is? Is that routines provided by the device driver to perform low-level : I/O functions to character oriented devices??? Ehhhh... I now have to assume you do not have any understanding at all for the concept of hardware interrupts and how to service them. Assuming you do have some programming experience then you should know what "signal()" do to your (userspace) programs. Signals are the only asynchronous events or messages the operating system will _force_ upon you. Everything else you have to explicit ask for. The classic "signal 15"=TERMINATE is one example. A hardware interrupt is likewise the means by wich the hardware notifies the operating system of any asynchronous events like incoming data on the serial ports, counter wraparounds or simply timerticks. Every interrupt number has a corresponding input pin on the interrupt controller. You will experience an interrupt once one of this inputs is pulled low. Silicon in the CPU will then postprone the currently executing thread and instead run the code stub registered to correspond with the interrupt pin currently annoying the CPU. This code stub is your ISR. The means by wich ISRs are registered is by placing its enty point (address) into the correct position in the vector table. The vector table itself resides in a well defined spot of the complete CPU memory space - usually in the very low memory region. The responsibility of your ISR is to take care of the event that triggered the event and then _clear the pending interrupt._ Gotcha: VxWorks abstracts over CPUs and thus maintains its own vector table. Your job is to register your ISR in the vxWorks vector table and also notify VxWorks to wich hardware IRQ this vector corresponds. That means you will require some documentations on you hardware. VxWorks will take care of the actual CPU vector table and also place your ISR written as a standard C function into a wrapper that emulates C calling convention for your convenience. - -- ****************************************************** Never ever underestimate the power of human stupidity. -Robert Anson Heinlein GeirFRS@invalid.and.so.forth ****************************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies : URGENT!! Date: Tue, 7 Mar 2000 16:01:57 +0800 From: "Jazz Lam" Organization: Alcatel Message-ID: <38c59995.0@dnews.sbell.com.cn> References: <89ot9u$oj9$1@nnrp1.deja.com> <89ulk4$eld$1@news1.xs4all.nl> <89vhvc$561$1@nnrp1.deja.com> <8a06bo$6qt$1@news1.xs4all.nl> i am also starting to develope the BSP, where can i get the BSP Development Kit? thanx wrote in message news:8a06bo$6qt$1@news1.xs4all.nl... > ilike_dsps@my-deja.com wrote: > > : Thanx...it works...I'm also new to Vxworks...I want to know more about > : developing/tuning BSPs on PowerPC Vxworks platform - any help ? > : This is a bit URGETNT !! > > If you want to do development on BSPs the BSP Development Kit is probably > best. If you want to modify an existing BSP this can also be a good > choice, but often it also works if you just start modifying an existing > BSP. > > Groeten, > Johan > > -- > o o o o o o o . . . ___________________________________ > o _____ || Johan Borkhuis | > .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | > >(________|__|_[_________]_|________________________________| > _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` > === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: WEB server Date: Tue, 07 Mar 2000 09:04:31 GMT From: "Øyvind Holmeide" Organization: Telia Internet Public Access Message-ID: Any recommendation for a freeware/shareware web server for a VxWorks based platform? regards Øyvind Holmeide --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Tue, 07 Mar 2000 09:17:43 +0000 From: Dave Message-ID: <38C4C937.C484577A@gecm.com> References: <38C3DD67.F814FEA3@gecm.com> <38C43EC4.2FD15169@syseng.anu.edu.au> <8a1vd2$tes$1@nnrp1.deja.com> john_94501@my-deja.com wrote: > Another approach, though perhaps not as clean as Peter's suggestion, is > to connect your own wrapper to the interrupt, then call your own routine > to attach or detach a new function to your wrapper. > > Something like this: > > void myIntRoutine(...) > { > if (myHandler != NULL) > (*myHandler)(...); > } > > void myIntConnect > ( > FUNCPTR newRoutine > ) > { > myHandler = newRoutine; > } > > Clearly, this has the disadvantage of adding extra instructions in the > interrupt processing path, but it might be simpler than writing an > intDisconnect() function (and 750s are pretty quick ;-) Thanks for the suggestion John. I've already tried something similar for changing the called handler 'on the fly' or as the code is running but unfortunately this doesn't work if you want to completely stop your software and re-run it from scratch as this will enevitably result in a second execution of intConnect() and then I'm in the two ISR situation again. Perhaps if I could sense whether my ISR was already registered then I could refrain from executing intConnect a second time ie. when the code is rerun. Of course, I could just reboot every time I wanted to run the software but that seems over the top. Cheers, Dave Moore --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Snmp NOTIFICATION-TYPE Date: Tue, 7 Mar 2000 09:21:54 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <8a2g3o$s5t$1@ndsukns01.ndsuk.com> References: <89ov3b$q0m$1@nnrp1.deja.com> <89vpog$ft0$1@ndsukns01.ndsuk.com> <8a12v5$8av$1@nnrp1.deja.com> Glad the last one worked! This one is quite easy - move the definitions of 'mycompany' and 'products' to fall AFTER the MODULE-IDENTITY definition. Of course, this means that you cannot then define the OID 'myproduct' in terms of 'products', I know that this will work, though I don't know whether the bracketted node numbers are required or not. myproduct MODULE-IDENTITY ::= { enterprises(1) mycompany(9999) products(2) 1 } mycompany OBJECT IDENTIFIER ::= { enterprises 9999} products OBJECT IDENTIFIER ::= { mycompany 2 } The method I prefer is to create a separate definition for your company's top few nodes, something like this... eg. MY-SMI DEFINITIONS ::= BEGIN IMPORTS enterprises, MODULE-IDENTITY FROM SNMPv2-SMI; myCo MODULE-IDENTITY LAST-UPDATED "...Z" ORGANIZATION "My Co Ltd." CONTACT-INFO "Me me@somewhere.co.uk 44-123-45678" DESCRIPTION "Top of the MyCo MIB tree" ::= { enterprises 9999 } - -- Branch for OIDs for MyCo products myProducts OBJECT IDENTIFIER ::= { myCo 1 } myEmbeddedEquipment OBJECT IDENTIFIER ::= { myProducts 1 } myPcSystems OBJECT IDENTIFIER ::= { myProducts 2 } - --The root of the MyCo MIBs myMibs OBJECT IDENTIFIER ::= { myCo 2 } - --The root of the MyCo Agents Capabilities myAgentCapabilities OBJECT IDENTIFIER ::= { myCo 3 } END Throw this module at mibcomp ahead of your specific MIB module, which can now import these nodes, and you can hang your MIB beneath them cleanly. This definition should be re-used for all your company's MIBs, you just need to allocate a unique node number for each (use this beneath all/any of the (lowest-level) nodes defined here. HTH, Will rliu@inter-tel.com wrote in message <8a12v5$8av$1@nnrp1.deja.com>... >Will, > >It fixed the problem. Thank you very much. > >I have another mib compiling warning. You may be able to help me out. > >Here is the mib, which is based on the sample mib file >snmpTaskGroup.mib from WindNet SNMPv1/v2c Agent. > > MY-COMPANY-MIB > > --FORCE-INCLUDE > --FORCE-INCLUDE > --FORCE-INCLUDE > --FORCE-INCLUDE "mibsdhand.h" > > DEFINITIONS ::= BEGIN > > IMPORTS > MODULE-IDENTITY, OBJECT-TYPE, NOTIFICATION-TYPE, Integer32 > FROM SNMPv2-SMI > enterprises, IpAddress > FROM RFC1155-SMI > DisplayString > FROM RFC1213-MIB > RowStatus > FROM SNMPv2-TC > ; > mycompany OBJECT IDENTIFIER ::= { enterprises 9999} > products OBJECT IDENTIFIER ::= { mycompany 2 } > > myproduct MODULE-IDENTITY > LAST-UPDATED "0002110000Z" > ORGANIZATION "" > CONTACT-INFO "" > DESCRIPTION > " ." > ::= { products 1 } > > > -- Groups defined in this MIB > > configuration OBJECT IDENTIFIER ::= { myproduct 1 } > >I go this warning, > >../snmp/superdsp.mib:22: Warning: A MIB module must start with >exactly one invocation of the MODULE-IDENTITY macro immediately >following an optional IMPORTS clause. >../snmp/superdsp.mib:25: Warning: The MODULE-IDENTITY macro must be >invoked immediately after the optional IMPORTS clause > > >Thanks in advance. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: WEB server Date: Tue, 07 Mar 2000 09:33:41 GMT From: Computer Engineer Organization: Somewhere in the Earth Message-ID: <38C45C40.E9C4DCC9@netscape.net> References: WindRiver has just released a new version of Wind Web Server. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies : URGENT!! Date: Tue, 07 Mar 2000 09:50:45 GMT From: ilike_dsps@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a2jdk$bgm$1@nnrp1.deja.com> References: <89ot9u$oj9$1@nnrp1.deja.com> <89ulk4$eld$1@news1.xs4all.nl> <89vhvc$561$1@nnrp1.deja.com> <8a06bo$6qt$1@news1.xs4all.nl> > If you want to do development on BSPs the BSP Development Kit is probably > best. If you want to modify an existing BSP this can also be a good > choice, but often it also works if you just start modifying an existing > BSP. That's what !!! where can I find BSP dev. kit ? any exsiting [sample] BSP to start with - just to get idea initially ?? Thanx Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: getc(stdin) only returns when enter (\n) is pressed Date: Tue, 7 Mar 2000 11:13:34 +0100 From: "Jan Rosseel" Organization: UUNET-NL (http://www.nl.uu.net) Message-ID: <8a2k91$87o$1@newnews1.news.nl.uu.net> References: <38C4193B.425D0004@lucent.com> <8a2and$25l$1@newsmaster01.nextra.at> Most input is "line buffered". To have unbuffered input, I do the following (with fileno = 0 for stdin): oldttysettings = ioctl (fileno, FIOGETOPTIONS, 0); int newttysettings = oldttysettings & ~OPT_LINE; ioctl (fileno, FIOSETOPTIONS, newttysettings); Jan R. "Mario Semo" wrote in message news:8a2and$25l$1@newsmaster01.nextra.at... > Hello, > > how are you doing RAW input? i use the CRT getchar() routine, and i, too, > see input only after a newline . i work either on the target at the command > prompt or in telnet session. > > -- > regards, Mario Semo. http://www.kirchnersoft.com > Douglas Fraser wrote in message <38C4193B.425D0004@lucent.com>... > > Are you on a dedicated terminal or are you using a terminal > emulation program? I am not sure if this is the problem, but > some terminal emulators run in a 'cooked' mode that sends the > input after you hit 'Enter'. I have used raw input from within > vxWorks, bypassing the target shell, and single character I/O > worked. > > > --------------------------- Newsgroups: comp.os.vxworks Subject: VME Interfacing Date: Tue, 7 Mar 2000 13:26:00 +0200 From: =?iso-8859-1?Q?Morn=E9_Hartman?= Message-ID: <38c4e730$0$44101@hades.is.co.za> Hi all, I am using Tornado2 and would like to interface via the VME bus to other boards. My board is using the Tundra Universe II PCI to VME bridge. I would appreciate any information or examples of how this interfacing can be accomplished. Yours sincerely Morne Hartman mhartman@grintek.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Mar 7 07:28:24 2000 From: "Pratte, Kevin" Date: Tue Mar 7 07:28:26 PST 2000 Subject: General Micro Systems PMC QSIO Driver > I am working with a General Micro Systems PMC QSIO board running under > vxWorks 5.3.1 on a Motorola mv2604. So far, we have been unsuccessful > attempting to configure and communicate with the board. We are using the > QSIO driver purchased from General Micro Systems but once we successfully > get through the qsDrv initialization function, we are unable to access the > control/status registers on the QSIO board. All registers accessed (using > the driver qget function) return 0xffff indicating that the board is not > correctly configured and/or mapped. We are using the Extended VME > configuration for our vxWorks kernel memory map where: > PCI MEM space start = 0xfd000000 > PCI I/O space start = 0xfe000000 > > I am invoking the qsDrv initialization function as follows: > qsDrv(0xfe000000, 0xfd040000, 0); where: 0xfe000000 = > pci_base, 0xfd040000 = q_base > > When I check the PCI Configuration Header (using the vxWorks pciHeaderShow > function) following the qsDrv call, I can see that the QSIO board's base > address and command register have been modified with the specified > parameters but I still can not access the Zilog USC registers "Behind The > Bridge". We have tried the QSIO board directly on the 2604 PMC and also > on a PMC Span card with similar results. > > I have tried several different pci_base and q_base addresses and > referenced other drivers for PMC modules which we do have working on the > mv2604, but have not found the right answer. > Does anybody have experience with this particular board set and driver running vxWorks? Any other general words of wisdom regarding PCI configuration on a mv2604 may also be useful. > Any help would be greatly appreciated. > > > Thanks, > > Kevin > > From vxwexplo-errs@csg.lbl.gov Tue Mar 7 07:59:29 2000 From: "Larry Sinn" Date: Tue Mar 7 07:59:31 PST 2000 Subject: Re: intConnect, intHandlerCreate and PPC --------------9925733EE6FA33AEB1512190 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dave: Another way is to add an intconnect to your vxWorks image that does a semGive, then all your code has to do is a pend on the semaphore and you can reload as often as you like Modify the configAll.h parameter USER_APPL_INIT so it calls a routine that has a semBCreate and intConnet. Larry. john_94501@my-deja.com wrote: > Another approach, though perhaps not as clean as Peter's suggestion, is > to connect your own wrapper to the interrupt, then call your own routine > to attach or detach a new function to your wrapper. > > Something like this: > > void myIntRoutine(...) > { > if (myHandler != NULL) > (*myHandler)(...); > } > > void myIntConnect > ( > FUNCPTR newRoutine > ) > { > myHandler = newRoutine; > } > > Clearly, this has the disadvantage of adding extra instructions in the > interrupt processing path, but it might be simpler than writing an > intDisconnect() function (and 750s are pretty quick ;-) Thanks for the suggestion John. I've already tried something similar for changing the called handler 'on the fly' or as the code is running but unfortunately this doesn't work if you want to completely stop your software and re-run it from scratch as this will enevitably result in a second execution of intConnect() and then I'm in the two ISR situation again. Perhaps if I could sense whether my ISR was already registered then I could refrain from executing intConnect a second time ie. when the code is rerun. Of course, I could just reboot every time I wanted to run the software but that seems over the top. Cheers, Dave Moore -- Larry Sinn Spectral Dynamics (408)474-1746 voice 1983 Concourse Drive (408)474-1780 fax San Jose, Ca. 95131-1708 sinnl@sd-star.com --------------9925733EE6FA33AEB1512190 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Dave:

Another way is to add an intconnect to your vxWorks image that
does a semGive, then all your code has to do is a pend on the semaphore
and you can reload as often as you like
Modify the configAll.h parameter USER_APPL_INIT so it calls a
routine that has a semBCreate and intConnet.

Larry.
 
 

john_94501@my-deja.com wrote:

> Another approach, though perhaps not as clean as Peter's suggestion, is
> to connect your own wrapper to the interrupt, then call your own routine
> to attach or detach a new function to your wrapper.
>
> Something like this:
>
> void myIntRoutine(...)
>     {
>     if (myHandler != NULL)
>         (*myHandler)(...);
>     }
>
> void myIntConnect
>     (
>     FUNCPTR newRoutine
>     )
>     {
>     myHandler = newRoutine;
>     }
>
> Clearly, this has the disadvantage of adding extra instructions in the
> interrupt processing path, but it might be simpler than writing an
> intDisconnect() function (and 750s are pretty quick ;-)

Thanks for the suggestion John. I've already tried something similar for
changing the called handler 'on the fly' or as the code is running but
unfortunately this doesn't work if you want to completely stop your software
and re-run it from scratch as this will enevitably result in a second
execution of intConnect() and then I'm in the two ISR situation again.
Perhaps if I could sense whether my ISR was already registered then I could
refrain from executing intConnect a second time ie. when the code is rerun.
Of course, I could just reboot every time I wanted to run the software but
that seems over the top.

Cheers,
    Dave Moore
  

-- 
Larry Sinn
Spectral Dynamics               (408)474-1746  voice
1983 Concourse Drive            (408)474-1780  fax
San Jose, Ca. 95131-1708        sinnl@sd-star.com
  --------------9925733EE6FA33AEB1512190-- From vxwexplo-errs@csg.lbl.gov Tue Mar 7 09:37:49 2000 From: "Darnold, Joel R" Date: Tue Mar 7 09:37:51 PST 2000 Subject: RE: General Micro Systems PMC QSIO Driver Kevin, I've recently done work on several PMCs on a MVME2307 board which uses the same VxWorks BSP as the MVME2600s. I'm not familiar with the General Micro Systems PMC QSIO but my experience is that PCI devices are all very similar, at least in how you initialize them onto the PCI bus. You probably know this but there is no mention about how you are mapping your Zilog device onto the PCI bus. I'm wondering if the two addresses you mention are the mappings for the "bridge" chips internal registers. Typically with PCI devices, the first two PCI device mappings are for internal bridge registers, the first for memory mapping the internal registers, the second for I/O mapping the internal registers. In general (at least with the board I've worked with) you must also initialize a third (or more) mapping for devices on the backside the bridge. This usually also requires some setup of region descriptors for the mapping associated with the device on the backside of the bridge. The region descriptors are initialized in internal registers on the bridge chip. One thing that threw me for a little while is how the PCI is really mapped to the processor. You definitely have the base addresses correct for memory and I/O addresses from the processor point of view. The thing I didn't realize at first was that the PCI memory map has no translation offset. In other words, accessing 0xfd040000 from the processor accesses 0xfd040000 on the PCI bus. The lowest PCI memory address on the MVME2600 is 0xfd000000 (using Extended VME). So when you set up the memory map of your PCI device, be sure to put the full address in your Base address register. The I/O mapping is different. VxWorks translates 0xfe000000 from the processor down to 0x0 in PCI I/O space. So... if you want to map your device at PCI I/O 0x00400000, put 0x00400001 into the base address register then access the device by the processor address 0xfe040000. The "1" in the lsb of the address indicates that it is an I/O mapping and is not used during address decoding. Does your PMC use a PLX chip as it's bridge? If so I could probably give you some source that would help. Joel Darnold Member of Technical Staff Sandia National Laboratories Control Subsystems Dept. 2338 Voice: 505-844-2487 Fax: 505-844-2925 From vxwexplo-errs@csg.lbl.gov Tue Mar 7 10:51:39 2000 From: Alex Reznik Date: Tue Mar 7 10:51:41 PST 2000 Subject: FW: vxWorks download through serial > Hello, vxWorks experts, > I have got a simple question, but the correct answer seems not so > primitive. > > Starting point: > Tornado2, vxWorks v.5.4. Target - any PowerPC board. > > The bootROM supports "ppp" and "sl" boot devices. What I should do is: > download vxWorks using "ppp" or "sl" from PC. But there is no > Tornado on this PC! > The question is: how should I configure COM2-modem-network-internet > properties of this PC and what is the procedure of download. > I guess somebody has already got such the experience. Any help will > be appreciate with great. > > NOTE. May be I missed some "ppp" definitions when produced bootROM? > There are a lot of options there. > > Thanks in advance, > Alex Reznik > Team Leader, > Terayon Communication Systems > Terayon Broadband Voice > Tel: +972-3-5384-711 > E-mail: alexr@telegate.co.il > From vxwexplo-errs@csg.lbl.gov Tue Mar 7 13:50:44 2000 From: Dave Monachello Date: Tue Mar 7 13:50:46 PST 2000 Subject: RE: Raw Socket on Vxworks > Submitted-by vxwexplo-errs@csg.lbl.gov Wed Mar 1 15:09:32 2000 > Submitted-by: Navin Kumar Agrawal > > Tornado > > > Hi Folks, > > I would really appreciate if somebody can help me in solving a problem > of > mine which is related with the usage of Raw Socket on VxWorks. I have > put my effort in the following way. > > -------------------------------------------------------- > struct sockaddr_in serveraddr; > bzero ((char *)&serveraddr, sizeof(struct sockaddr_in)); > serveraddr.sin_family = AF_INET; > serveraddr.sin_len = sizeof(struct sockaddr_in); > serveraddr.sin_port = htons (6007); > serveraddr.sin_addr.s_addr = htonl (inet_addr ("192.168.100.140")); > > if ((fd = socket (AF_INET, SOCK_RAW, IPPROTO_ICMP)) == ERROR) > { > perror ("socket"); > return -1; > } > > if (bind (fd, (struct sockaddr *)&serveraddr, sizeof(struct > sockaddr_in)) == ERROR) > { > perror ("bind failed"); > return -1; > } > -------------------------------------------------------- > > My target system is based on PPC604 (PowerPC). On executing the above > program, > it is failing at bind () and the error code returned is > EADDRNOTAVAIL ("Can't > assign requested address"). > > Can anybody help me in sovling this problem? I assume you are doing the bind so that you can specify the source address of the IP packet. The IP address that you are binding must be the IP address of one of your IP interfaces. > From vxwexplo-errs@csg.lbl.gov Wed Mar 8 03:33:41 2000 From: Bob Date: Wed Mar 8 03:33:43 PST 2000 Subject: Building bootable images I have been building a vxworks downloadable project with a file structure shown below using Tornado 2. Is there a quick way to convert the project to build a bootable image. I.E. not having to cut and paste all of the entries from the makefile tool. c:gc12 |-----boot | I----default |-----ppc603gnu |-----mycodedir1 |-----mycodedir2 Also I am using the PRJ_LIBS macro to include an object file I.E $(PPC_S110)/Dskmv260x.o. Is it possible/easy to have the build process call the makefile that build the .o file. TIA Bob From vxwexplo-errs@csg.lbl.gov Wed Mar 8 04:03:16 2000 From: Vxworks Exploder Date: Wed Mar 8 04:03:18 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Mar 8 04:03:12 PST 2000 Subject: Re: vxworks for dummies : URGENT!! Subject: Re: WEB server Subject: where to get systemtime to put into set_clock Subject: Re: How to get current time & date for timestamps for DOS-FS files Subject: Re: What is the Interrupt Service Routines (ISR)? Subject: Re: WEB server Subject: Debuggers memory window Subject: Re: Get Socket Send Buffer Size Subject: changing the net device on a running system Subject: Re: How to get current time & date for timestamps for DOS-FS files Subject: RSVP - Resource Reservation Protocol on VxWorks Subject: Re: intConnect, intHandlerCreate and PPC Subject: Re: intConnect, intHandlerCreate and PPC Subject: Re: Snmp NOTIFICATION-TYPE Subject: VxFusion and I20 Subject: US-IL-Developer Needed Subject: Re: FW: vxWorks download through serial Subject: Re: Using a standard off the shelf GDB for debugging Subject: Re: where to get systemtime to put into set_clock Subject: VxWorks kernel tick on a Motorola Power PC 860 Subject: Re: Vxworks - fcntl() - tornado - GCC Subject: Re: VxWorks kernel tick on a Motorola Power PC 860 Subject: Re: builtin_new/builtin_delete Subject: Re: VxWorks kernel tick on a Motorola Power PC 860 Subject: VxWorks image Subject: Re: intConnect, intHandlerCreate and PPC Subject: rawFsLib (Raw Filesystem) Subject: TFFS (& Raw Filesystems) Subject: TFFS (& Raw Filesystems) Subject: rawFsLib (Raw Filesystem) ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxworks for dummies : URGENT!! Date: 7 Mar 2000 12:13:24 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8a2rp4$271$1@news1.xs4all.nl> References: <89ot9u$oj9$1@nnrp1.deja.com> <89ulk4$eld$1@news1.xs4all.nl> <89vhvc$561$1@nnrp1.deja.com> <8a06bo$6qt$1@news1.xs4all.nl> <8a2jdk$bgm$1@nnrp1.deja.com> ilike_dsps@my-deja.com wrote: :> If you want to do development on BSPs the BSP Development Kit is : probably :> best. If you want to modify an existing BSP this can also be a good :> choice, but often it also works if you just start modifying an : existing :> BSP. : That's what !!! where can I find BSP dev. kit ? any exsiting [sample] : BSP to start with - just to get idea initially ?? The BSP Development Kit is a product from WindRiver, so if you are interrested you can order it there. 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: WEB server Date: 7 Mar 2000 12:33:49 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8a2svd$271$2@news1.xs4all.nl> References: : Any recommendation for a freeware/shareware web server for a VxWorks based : platform? There is a freeware Webserver at GoAhead. It can be found at http://www.goahead.com/webserver. Another one can be found at ftp://www.pso.com/users/peacefulstar/ Groeten, Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan@borksoft.xs4all.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` === VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html === --------------------------- Newsgroups: comp.os.vxworks Subject: where to get systemtime to put into set_clock Date: Tue, 07 Mar 2000 05:27:01 -0800 From: LB Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <0ed1f350.6bf29dde@usw-ex0110-076.remarq.com> Using the simulator or a Intel based BSP we are trying to setup a standalone system. Therefor the system must be able to run without interference from any user. Thus when the system starts it must be able to get the system time. Because after a boot the system time is not set in VxWorks we want to receive the systemtime from the BIOS (or alike). Does anyone have a clue? Some one pointed to Linux sources but we were not able to find the right-stuff. thx. Leendert Biemans * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping. Smart is Beautiful --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get current time & date for timestamps for DOS-FS files Date: Tue, 07 Mar 2000 13:44:33 GMT From: m.fischer@steinbichler.de (Manfred Fischer) Organization: roNet GmbH Rosenheim Message-ID: <38c501af.17792784@news.ronet.de> References: <38bb9cf1.9562099@news.ronet.de> <38BC10F1.4F07@despam.pangea.ca> <89pb9i$t77$1@porthos.nl.uu.net> Thanks Harvey & Leendert > > Yes look at the dosFsLib docs. The function is called > dosFsDateTimeInstall(). > > It requires a handler function which typically looks like: > > STATUS DateTimeFunc(DOS_DATE_TIME * pDosDateTime) > { > time_t tNow = time(NULL); > struct tm* t = localtime(&tNow); > > pDosDateTime->dosdt_year = t->tm_year ; > pDosDateTime->dosdt_year += ( (t->tm_year > 70) ? 1900 : 2000); > pDosDateTime->dosdt_month = t->tm_mon + 1 ; /* Map 0..11 to 1..12 */ > pDosDateTime->dosdt_day = t->tm_mday ; > pDosDateTime->dosdt_hour = t->tm_hour ; > pDosDateTime->dosdt_minute = t->tm_min ; > pDosDateTime->dosdt_second = t->tm_sec ; > > return( OK ) ; > } > For the MV2700 PowerPC a solution for getting the current time & date will be: int getRTC(void) { struct tm currentTime; struct timespec ansiTime; /* read real time clock from nvram: */ currentTime.tm_year = BCDToInt(sysNvRead((ULONG) 0x1FFF)); currentTime.tm_year += (currentTime.tm_year < 70 ? 100 : 0); currentTime.tm_mon = BCDToInt(sysNvRead((ULONG) 0x1FFE)) - 1; currentTime.tm_mday = BCDToInt(sysNvRead((ULONG) 0x1FFD)); currentTime.tm_hour = BCDToInt(sysNvRead((ULONG) 0x1FFB)); currentTime.tm_min = BCDToInt(sysNvRead((ULONG) 0x1FFA)); currentTime.tm_sec = BCDToInt(sysNvRead((ULONG) 0x1FF9)); currentTime.tm_isdst = 0; /* convert in calendar time */ if ((ansiTime.tv_sec = mktime(¤tTime)) == ERROR) return(ERROR); ansiTime.tv_nsec = 0; /* set software clock */ clock_settime(CLOCK_REALTIME, &ansiTime); return(OK); } int BCDToInt(int BCD) { return ((BCD >> 4) * 10) + (BCD & 0x0F); } But now I have a new problem: How to set the RTC in nvram? vxWorks provides a sysNvWrite() function, but there is no example or advice how to use this function! I tried to do this with the following code but it does not work: int setRTC(void) { time_t ansiTime = time(NULL); struct tm *currentTime = localtime(&ansiTime); sysNvWrite((ULONG) 0x1FFF, (UCHAR)intToBCD(currentTime->tm_year % 100)); sysNvWrite((ULONG) 0x1FFE, (UCHAR)intToBCD(currentTime->tm_mon+1)); sysNvWrite((ULONG) 0x1FFD, (UCHAR)intToBCD(currentTime->tm_mday)); sysNvWrite((ULONG) 0x1FFB, (UCHAR)intToBCD(currentTime->tm_hour)); sysNvWrite((ULONG) 0x1FFA, (UCHAR)intToBCD(currentTime->tm_min)); sysNvWrite((ULONG) 0x1FF9, (UCHAR)intToBCD(currentTime->tm_sec)); return(OK); } int intToBCD(int n) { if ((n > 99) || (n < 0)) return(0); return ((n / 10) | (n % 10)); } Where is my problem? Manfred --------------------------- Newsgroups: comp.os.vxworks Subject: Re: What is the Interrupt Service Routines (ISR)? Date: Tue, 7 Mar 2000 14:08 +0000 (GMT Standard Time) From: rew@cp.cix.co.uk (Richard Wenham) Organization: CIX - Compulink Information eXchange Message-ID: References: <38c3e4b1.0@seat.s3two.ie> Reply-To: rew@cp.cix.co.uk ISR == Interrupt Service Routine. It's the code which responds to an interrupt from the device, and would collect characters from the device and put them in a buffer, or feed characters from a buffer to the device. Richard In article <38c3e4b1.0@seat.s3two.ie>, rao@s3group.com (Rao) wrote: > Hi to everyone, > I'm far away from the advanced features of RTOS and close to > basics. I'm a beginner of VxWorks....please tell me what is ISR or what > for > it is? Is that routines provided by the device driver to perform > low-level > I/O functions to character oriented devices??? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: WEB server Date: Tue, 07 Mar 2000 09:23:05 -0500 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <38C510C9.EF326A66@lucent.com> References: I have had reasonable behavior from: The webserver was easy to port to my StrongARM target. The method for supporting active pages is easy to use. I have encountered occasional page timeouts that I cannot track down. It is a rare occurance and may be attributed to our crummy subnet that our embedded devices are relegated to. I still need to find the spare cycles to figure it out. Doug --------------------------- Newsgroups: comp.os.vxworks Subject: Debuggers memory window Date: Tue, 7 Mar 2000 15:37:31 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <8a344k$dck$1@crusher.de.colt.net> has anyone realized, that the debugger window of T2 dosn't update every debugger step you do, independent if you have checked the flag in then options/debugger window ?? How to solf it ????? - -- Thanks Achim Zimmer R&D DICA Technologies Have a nice day ! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Get Socket Send Buffer Size Date: Tue, 07 Mar 2000 17:10:48 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38C51BF8.C0BC7398@bitband.com> References: <8a0eiu$jiq1@overload.lbl.gov> <8a0uc8$4ma$1@nnrp1.deja.com> Reply-To: leonid@bitband.com Unfirtunately, FIONWRITE is not supported in SENS, and this has been known for a while, perhaps there is an SPR open on this issue. Leonid Daniel Wong wrote: > I've tried using ioctl() with the FIONWRITE option to obtain the number > of bytes in the send buffer, but ioctl() returns an unreasonable number > of bytes. What's the correct method of retrieving the number of bytes > in the send buffer? > > Daniel > > In article <8a0eiu$jiq1@overload.lbl.gov>, > "Blake, Don" wrote: > > > > Michael, VxWorks error 0x20 (decimal 32) is "Broken pipe" (EPIPE). It > > sounds like a problem on the receiving side. > > > > Error 0x23 (decimal 35) is "Unsupported value" (ENOTSUP). I don't > > understand this one. > > > > Don Blake > > Sr. Programmer > > Lockheed Martin Federal Systems > > Owego, NY > > > > -----Original Message----- > > > > Newsgroups: comp.os.vxworks > > Subject: Get Socket Send Buffer Size?l > > Date: Wed, 01 Mar 2000 18:23:24 GMT > > From: "Michael Murphy" > > > > Hi All, > > > > I'm send large buffers via TCP using send(). After a few MB or so, > > send > > will return an error 0x20. I'm assuming this is a Send Buffer Full > > Error. > > Does anyone know how to peek into send buffer to get its current > > size? > > All calls to send() afterwards return an error 0x23. Anyone know > > where I can > > get info on these error codes? > > > > Thanx, > > > > Michael > > > > > > -- > Daniel Wong > daniel_k_wong@my-dejan.com > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: changing the net device on a running system Date: Tue, 07 Mar 2000 18:45:59 +0200 From: Ofer Goren Organization: nortelnetworks Israel Message-ID: <38C53247.1C200175@nortelnetworks.com> Hi. I have two different net deviced residing on my board. Only one of them is working at a specific time. In order to change between those two devices, we are rebooting the system, and change the boot device in the boot parameters. What do I need to do if I want to change the device (and initalize it correctly) on the fly (without rebooting the system)? Thanks, - -- Email : ogoren@nortelnetworks.com Phone number: +9723-6456023 Fax number : +9723-6479579 ESN : 826-6023 http://www.nortelnetworks.com - ---------------------------------------------------- If you lend someone $20, and never see that person again, it was probably worth the investment. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get current time & date for timestamps for DOS-FS files Date: Tue, 07 Mar 2000 11:49:31 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38C55D4B.16AB@despam.pangea.ca> References: <38bb9cf1.9562099@news.ronet.de> <38BC10F1.4F07@despam.pangea.ca> <89pb9i$t77$1@porthos.nl.uu.net> <38c501af.17792784@news.ronet.de> Manfred Fischer wrote: >[...] > But now I have a new problem: How to set the RTC in nvram? >[...] >sysNvWrite((ULONG) 0x1FFF, (UCHAR)intToBCD(currentTime->tm_year % 100)); > There may be some bit you have to set somewhere before you can write. What hardware exactly is located at 0x01FFF? Is it NVRAM or is it an RTC? NVRAM and RTC? Find out what kind of hardware it is. Glomm on over to that manufacturer's web site. Dig up the docs on that RTC. - -het - --- "Assembly of Japanese bicycle require great peace of mind." - Instruction manual mentioned by R. Pirsig Harvey Taylor mailto:het@despam.portal.ca http://www.pangea.ca/~het --------------------------- Newsgroups: comp.os.vxworks Subject: RSVP - Resource Reservation Protocol on VxWorks Date: Tue, 07 Mar 2000 13:19:04 -0500 From: John Sestak Organization: Raytheon, Falls Church Message-ID: <38C54818.252FF144@fallschurch.esys.com> Hi all, I'm trying to create an "RSVP-capable" socket. Since RSVP is a Transport layer protocol, I have to create a raw socket which is capable of receiving uni-cast and/or multi-cast traffic. The problem is "Identifying" this socket as an RSVP socket. Under Unix (I have some old code that someone else wrote), I believe you would do the following to create the RSVP socket: newsock = socket(AF_INET, SOCK_RAW, IPPROTO_RSVP); Unfortunately, VxWorks doesn't seem to recognize the constant "IPPROTO_RSVP" (I believe this tells the socket to look for incoming packets with the IP protocol field set = 46 (RSVP). So, does any know of a way to create an RSVP-capable socket? Thank you, John Sestak --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Tue, 07 Mar 2000 13:20:20 -0500 From: David Abbott Organization: Jefferson Lab Message-ID: <38C54863.35001656@jlab.org> References: <38C3DD67.F814FEA3@gecm.com> Dave Moore wrote: > Hi everyone, > I'm experiencing difficulty changing ISRs on a MCP750 PPC Card. > Basically, multiple calls to intConnect appear to add the referenced isr > function to existing calls. For example calling intConnect twice with > the same ISR means that when the interrupt is caught, the isr is > executed twice. This is undesirable as re-running a piece of code which > uses intConnect doubles up the calls to the isr. What I'm after is a > intDisconnect call but there isn't one. > > I either need an equivalent of intConnect that can be run multiple times > using the same isr without resultant multiple calls or a method of > clearing the effect of intConnect. > > I've seen several instances of code using a combination of > intHandlerCreate() and intVecSet() to get around this problem but the > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > Anybody got any ideas - Thanks in advance. > > Dave. I have attached an "intDisconnect()" routine that has served me well for many years. Just include it as an extra object when you are compiling your kernel. It takes one argument (the interrupt vector). David Abbott Jefferson Lab Data Acquisition Group abbottd@jlab.org /* * File: intDisconnect.c * Date: September 1996 * * Description: * ----------- * This function un-installs an interrupt service routine on a Tornado/PPC board * */ #include #include #include #include #include #include #include /* #include */ typedef struct intHandlerDesc /* interrupt handler desciption */ { VOIDFUNCPTR vec; /* interrupt vector */ int arg; /* interrupt handler argument */ struct intHandlerDesc * next; /* next interrupt handler & argument */ } INT_HANDLER_DESC; extern sysIntTbl[]; extern sysVmeIntTable[]; STATUS intDisconnect(); LOCAL int handlerDelete(); /****************************************************************************** * * intDisconnect - un-install an ISR * * This routine un-installs an ISR from the interrupt vector table used by * the local and VMEbus interrupt dispatchers. * *****************************************************************************/ STATUS intDisconnect(int vector) { INT_HANDLER_DESC *handler; if (vector < 16) { printf("ERROR: cannot disconnect a local bus ISR with intDisconnect\n"); return(ERROR); } if (vector > 255) { printf("ERROR: interrupt vector out of range\n"); return(ERROR); } handler = (INT_HANDLER_DESC *)sysIntTbl[(int)vector]; if (handler != NULL) { printf("disconnecting vector %d \n", vector); handlerDelete(handler); sysIntTbl[(int)vector] = NULL; } return(OK); } /****************************************************************************** * * handlerDelete() - recursively deletes ISRs from sysIntTbl vector * *****************************************************************************/ LOCAL int handlerDelete(INT_HANDLER_DESC *handler) { if (handler->next != NULL) { handlerDelete(handler->next); } free(handler); handler = NULL; return(OK); } --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Tue, 07 Mar 2000 18:29:17 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a3hpo$1et$1@nnrp1.deja.com> References: <38C3DD67.F814FEA3@gecm.com> <38C43EC4.2FD15169@syseng.anu.edu.au> <8a1vd2$tes$1@nnrp1.deja.com> <38C4C937.C484577A@gecm.com> Hi Dave, > Thanks for the suggestion John. I've already tried something similar for > changing the called handler 'on the fly' or as the code is running but > unfortunately this doesn't work if you want to completely stop your software > and re-run it from scratch as this will enevitably result in a second > execution of intConnect() and then I'm in the two ISR situation again. > Perhaps if I could sense whether my ISR was already registered then I could > refrain from executing intConnect a second time ie. when the code is rerun. > Of course, I could just reboot every time I wanted to run the software but > that seems over the top. You could always add the intConnect() call to the kernel's init sequence and treat your interrupt handling scheme as an extension of the kernel rather than part of your application. That would avoid multiple calls to intConnect(). Just a thought, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Snmp NOTIFICATION-TYPE Date: Tue, 07 Mar 2000 19:02:55 GMT From: rliu@inter-tel.com Organization: Deja.com - Before you buy. Message-ID: <8a3jor$30t$1@nnrp1.deja.com> References: <89ov3b$q0m$1@nnrp1.deja.com> <89vpog$ft0$1@ndsukns01.ndsuk.com> <8a12v5$8av$1@nnrp1.deja.com> <8a2g3o$s5t$1@ndsukns01.ndsuk.com> Will, Based on your message, I made the following change and the warning is gone. I am thinking about your second suggestion to create separate definiations for my company's top nodes. I really appreciate your help. Have a good day! myproduct MODULE-IDENTITY ::= { enterprises 9999 2 1 } mycompany OBJECT IDENTIFIER ::= { enterprises 9999} products OBJECT IDENTIFIER ::= { mycompany 2 } In article <8a2g3o$s5t$1@ndsukns01.ndsuk.com>, "Will Fookes" wrote: > Glad the last one worked! > > This one is quite easy - move the definitions of 'mycompany' and 'products' > to fall AFTER the MODULE-IDENTITY definition. Of course, this means that you > cannot then define the OID 'myproduct' in terms of 'products', I know that > this will work, though I don't know whether the bracketted node numbers are > required or not. > > myproduct MODULE-IDENTITY > > ::= { enterprises(1) mycompany(9999) products(2) 1 } > > mycompany OBJECT IDENTIFIER ::= { enterprises 9999} > products OBJECT IDENTIFIER ::= { mycompany 2 } > > The method I prefer is to create a separate definition for your company's > top few nodes, something like this... > > eg. > > MY-SMI DEFINITIONS ::= BEGIN > IMPORTS > enterprises, MODULE-IDENTITY > FROM SNMPv2-SMI; > > myCo MODULE-IDENTITY > LAST-UPDATED "...Z" > ORGANIZATION "My Co Ltd." > CONTACT-INFO > "Me > me@somewhere.co.uk > 44-123-45678" > DESCRIPTION > "Top of the MyCo MIB tree" > ::= { enterprises 9999 } > > -- Branch for OIDs for MyCo products > myProducts OBJECT IDENTIFIER ::= { myCo 1 } > myEmbeddedEquipment OBJECT IDENTIFIER ::= { myProducts 1 } > myPcSystems OBJECT IDENTIFIER ::= { myProducts 2 } > > --The root of the MyCo MIBs > myMibs OBJECT IDENTIFIER ::= { myCo 2 } > > --The root of the MyCo Agents Capabilities > myAgentCapabilities OBJECT IDENTIFIER ::= { myCo 3 } > > END > > Throw this module at mibcomp ahead of your specific MIB module, which can > now import these nodes, and you can hang your MIB beneath them cleanly. > This definition should be re-used for all your company's MIBs, you just need > to allocate a unique node number for each (use this beneath all/any of the > (lowest-level) nodes defined here. > > HTH, > Will > > rliu@inter-tel.com wrote in message <8a12v5$8av$1@nnrp1.deja.com>... > >Will, > > > >It fixed the problem. Thank you very much. > > > >I have another mib compiling warning. You may be able to help me out. > > > >Here is the mib, which is based on the sample mib file > >snmpTaskGroup.mib from WindNet SNMPv1/v2c Agent. > > > > MY-COMPANY-MIB > > > > --FORCE-INCLUDE > > --FORCE-INCLUDE > > --FORCE-INCLUDE > > --FORCE-INCLUDE "mibsdhand.h" > > > > DEFINITIONS ::= BEGIN > > > > IMPORTS > > MODULE-IDENTITY, OBJECT-TYPE, NOTIFICATION-TYPE, Integer32 > > FROM SNMPv2-SMI > > enterprises, IpAddress > > FROM RFC1155-SMI > > DisplayString > > FROM RFC1213-MIB > > RowStatus > > FROM SNMPv2-TC > > ; > > mycompany OBJECT IDENTIFIER ::= { enterprises 9999} > > products OBJECT IDENTIFIER ::= { mycompany 2 } > > > > myproduct MODULE-IDENTITY > > LAST-UPDATED "0002110000Z" > > ORGANIZATION "" > > CONTACT-INFO "" > > DESCRIPTION > > " ." > > ::= { products 1 } > > > > > > -- Groups defined in this MIB > > > > configuration OBJECT IDENTIFIER ::= { myproduct 1 } > > > >I go this warning, > > > >../snmp/superdsp.mib:22: Warning: A MIB module must start with > >exactly one invocation of the MODULE-IDENTITY macro immediately > >following an optional IMPORTS clause. > >../snmp/superdsp.mib:25: Warning: The MODULE-IDENTITY macro must be > >invoked immediately after the optional IMPORTS clause > > > > > >Thanks in advance. > > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxFusion and I20 Date: Tue, 7 Mar 2000 19:24:27 GMT From: David & Bella Aldrich Message-ID: <2000030719242783839@zetnet.co.uk> We are considering how to implement messaging between 2 PCI cards, each having a PowerPC and running VxWorks. VxFusion is a possibility for the application layer interface but has the disadvantage of being fairly expensive. We are also looking at I20 for the physical layer. We could just use I20 and write our own application layer interface. Here are some questions: 1) Has anyone experience of using VxFusion? 2) Is there an adapter layer available for PCI or I20? 3) Is an I20 driver available for PowerPC and VxWorks? I would welcome any comments or advice please. Thanks David Aldrich --------------------------- Newsgroups: comp.os.vxworks Subject: US-IL-Developer Needed Date: Tue, 7 Mar 2000 13:42:44 -0500 From: "Colleen Nielsen" Organization: EnterAct Corp. Message-ID: <8a3m83$amu$1@news.enteract.com> 12Month Contract in the Suburbs of Chicago Colleen Nielsen Senior Regional Recruiter Technisource 1400 Opus Place Suite 905 Downers Grove, IL 60515 (630)963-9401 As part of a small team, develop protocol code for wireless General Packet Radio Service (GPRS). Development will be done in C++ on Solaris and VxWorks. The protocol code will run on top of a TCP/IP stack. This position includes acting as our point of contact with another development team, so good communication and negotiation skills are needed in addition to strong technical skills. Candidate must have 2+ years experience with development of real-time software or firmware (at least one year using C++). Experience with development of protocol software (TCP/IP, SS7, ISDN, etc) is highly desirable. Experience with Rational Rose RT or ObjecTime is desirable but not mandatory. Please respond with resume to: Colleen Nielsen Senior Regional Recruiter Technisource 1400 Opus Place Suite 905 Downers Grove, IL 60515 (630)963-9401 or (800)330-3308 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: FW: vxWorks download through serial Date: 7 Mar 2000 19:48:45 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8a3met$je$1@news1.xs4all.nl> References: <8a3khd$jht3@overload.lbl.gov> Alex Reznik wrote: :> Hello, vxWorks experts, :> I have got a simple question, but the correct answer seems not so :> primitive. :> :> Starting point: :> Tornado2, vxWorks v.5.4. Target - any PowerPC board. :> :> The bootROM supports "ppp" and "sl" boot devices. What I should do is: :> download vxWorks using "ppp" or "sl" from PC. But there is no :> Tornado on this PC! :> The question is: how should I configure COM2-modem-network-internet :> properties of this PC and what is the procedure of download. :> I guess somebody has already got such the experience. Any help will :> be appreciate with great. You have to enable PPP in the bootROM, as you did. Then you have to start a PPP protocol on your PC. I have a description on how to do this on my page. 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 a standard off the shelf GDB for debugging Date: Wed, 08 Mar 2000 00:44:40 GMT From: "Roberta Cliff" Organization: Magma Communications Ltd. Message-ID: References: <88v71l$oov$1@nnrp1.deja.com> The version of gdb from WindRiver is a modifed version of GNU's. It is designed to access the target via tgtsvr. Are you trying to use a superior debugger over top of gnu? wrote in message news:88v71l$oov$1@nnrp1.deja.com... > Has anyone used an off the shelf GDB to debug with? I am trying to > debug a VxSim NT target and it always SIGSEGV's (seg fault), it doesn't > do it if I use the T2 gdb. For many reasons I want to use an off the > shelf GDB and not the one supplied by T2. Does anyone know why this is > happening? > > Thanks > > -Howard > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: where to get systemtime to put into set_clock Date: Wed, 08 Mar 2000 01:07:52 GMT From: snaphook@southwesternbell.net (Bruce) Organization: SBC Internet Services Message-ID: <38c5a7af.11379213@news.swbell.net> References: <0ed1f350.6bf29dde@usw-ex0110-076.remarq.com> In comp.os.vxworks LB wrote: >Using the simulator or a Intel based BSP we are trying to >setup a standalone system. >Therefor the system must be able to run without >interference from any user. Thus when the system starts it >must be able to get the system time. >Because after a boot the system time is not set in VxWorks >we want to receive the systemtime from the BIOS (or alike). > > Does anyone have a clue? This is board specific. Does your board HAVE a real time clock? Is it battery backed? You may not even be able to do this if the answers are no. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks kernel tick on a Motorola Power PC 860 Date: Wed, 08 Mar 2000 02:40:49 GMT From: "Brad Ferguson" Organization: Another Optus Customer Message-ID: Hi all, Can anyone tell me where vxWorks gets its kernel tick from when using the Motorola Power PC 860 processor? Does it use the decrementer timer? TIA, Brad Ferguson, Electronic Engineer Vision Abell Pty Ltd --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxworks - fcntl() - tornado - GCC Date: Fri, 25 Feb 2000 13:34:49 -0500 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <38B6CB49.C4EF4900@delphi.com> References: <896anv$c1m$3@reader1.fr.uu.net> You would have to build it into vxWorks. the compiler does not control which functions are available; the OS does. The easiest source I know of is in the linux distribution. I recently saw an add about Solaris 8 source beign available on Sun's website. In any case, such code would only be an example, since anything you write will have to be part of the vxWorks kernel. Speaking only for myself, Joe Durusau "Gaël Trinquart" wrote: > It seems that the fcntl() function is not implemented on GCC. > Does anyone have the source code from fcntl() ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks kernel tick on a Motorola Power PC 860 Date: Wed, 8 Mar 2000 08:13:12 +0100 From: "Achim Zimmer" Organization: Colt Telecom GmbH Message-ID: <8a4ufe$881$1@crusher.de.colt.net> References: In "sysClkInit.c" is a function called "usrClock", that is connected to the clock ISR. In this function the call to "tickAnnounce" informs the kernel. We have added there our counters for a "real" time clock, the only thing we have to do is do set the Clock variables at system startup. Achim Brad Ferguson schrieb in im Newsbeitrag: R4jx4.955$fw5.503291@news0.optus.net.au... > Hi all, > > Can anyone tell me where vxWorks gets its kernel tick from when using the > Motorola Power PC 860 processor? Does it use the decrementer timer? > > TIA, > > Brad Ferguson, > > Electronic Engineer > Vision Abell Pty Ltd > > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: builtin_new/builtin_delete Date: Wed, 08 Mar 2000 07:20:51 GMT From: magic_mag@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a4v0h$244$1@nnrp1.deja.com> References: <38BAA71C.3C96@fzi.de> In article <38BAA71C.3C96@fzi.de>, Torsten Rupp wrote: > Hi, > > currently I try to load some program code including C++ new()/delete() > calls on a MV2400/PPC 750 board. I have a problem with the functions > __builtin_new and __builtin_delete, which seems not to be included > in the PPC target lib/BSP anymore (it was included in the lib for > MC68040 targets). Are these functions now included in some other > (additional) lib? Thanx for some hint. > > Torsten > > -- > Computer Science Research Center (FZI) phone : +49-721-93372-13 > Mobility Management and Robotics fax : +49-721-93372-19 > Haid-und-Neu-Str. 10-14 e-mail: rupp@fzi.de > D-76131 Karlsruhe (Germany) www : www.fzi.de/robot > Hi Torsten, some days ago I was faced with the same problem, which is caused by an error in makeSymTbl.exe. This error is fixed with the “SPR 28187 Patch for PowerPC standalone symbol table 1/28/2000“ on the WRS WindSurf web page. Hope this helps Matthias ******************************************** * * * Matthias Guelck * * Research & Development * * * * Heidelberg Digital * * Siemenswall * * D-24107 Kiel * * * * Phone +49-431-386-1608 * * Fax +49-431-386-3882 * * E-Mail Matthias.Guelck@DE.Heidelberg.com * * * ******************************************** Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks kernel tick on a Motorola Power PC 860 Date: Wed, 08 Mar 2000 08:13:53 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a5241$3tq$1@nnrp1.deja.com> References: In article , "Brad Ferguson" wrote: > Hi all, > > Can anyone tell me where vxWorks gets its kernel tick from when using the > Motorola Power PC 860 processor? Does it use the decrementer timer? Yes, but there is the option of using the other timers for the aux clock. It all depends on the BSP... ads860 has this by default, mbx860 didn't, but it is easily changed. TTFN, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks image Date: Wed, 8 Mar 2000 11:11:40 +0200 From: "Vladimir" Organization: NetVision Israel Message-ID: <8a5569$t5v$1@news.netvision.net.il> Reply-To: "Vladimir" I have two problems (Tornado 2, x86 platform): 1. How to create a standalone image more than 512 KB (rule VxWorks_rom) - other rules such as VxWorks_romCompress don't work. Changing constants such as ROM_SIZE, ROW_LOW_ADDR... makes image stuck down. 2. Is it possible (if it is, how) to elliminate memory zeroing at the beginning (when the image is loaded). This is necessary, because start up time is critical. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Wed, 08 Mar 2000 10:14:33 +0000 From: Dave Moore Organization: BAE SYSTEMS Message-ID: <38C62809.E19F2695@gecm.com> References: <38C3DD67.F814FEA3@gecm.com> <38C54863.35001656@jlab.org> Hi everyone, Thanks for everyone's suggestions. In particular the intDisconnect code from David Abbott - this works a treat. Thanks. Now I can re-run my code without having to reboot!. Dave Moore David Abbott wrote: > > Dave Moore wrote: > > > Hi everyone, > > I'm experiencing difficulty changing ISRs on a MCP750 PPC Card. > > Basically, multiple calls to intConnect appear to add the referenced isr > > function to existing calls. For example calling intConnect twice with > > the same ISR means that when the interrupt is caught, the isr is > > executed twice. This is undesirable as re-running a piece of code which > > uses intConnect doubles up the calls to the isr. What I'm after is a > > intDisconnect call but there isn't one. > > > > I either need an equivalent of intConnect that can be run multiple times > > using the same isr without resultant multiple calls or a method of > > clearing the effect of intConnect. > > > > I've seen several instances of code using a combination of > > intHandlerCreate() and intVecSet() to get around this problem but the > > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > > > Anybody got any ideas - Thanks in advance. > > > > Dave. > > I have attached an "intDisconnect()" routine that has served me well for > many years. Just include it > as an extra object when you are compiling your kernel. It takes one > argument (the interrupt vector). > > David Abbott > Jefferson Lab Data Acquisition Group > abbottd@jlab.org > > /* > * File: intDisconnect.c > * Date: September 1996 > * > * Description: > * ----------- > * This function un-installs an interrupt service routine on a Tornado/PPC > board > * > */ > > #include > #include > #include > #include > #include > #include > #include > > /* > #include > */ > typedef struct intHandlerDesc /* interrupt handler desciption */ > { > VOIDFUNCPTR vec; /* interrupt vector */ > int arg; /* interrupt handler argument */ > struct intHandlerDesc * next; /* next interrupt handler & > argument */ > } INT_HANDLER_DESC; > > extern sysIntTbl[]; > extern sysVmeIntTable[]; > STATUS intDisconnect(); > LOCAL int handlerDelete(); > > /****************************************************************************** > > * > * intDisconnect - un-install an ISR > * > * This routine un-installs an ISR from the interrupt vector table used by > * the local and VMEbus interrupt dispatchers. > * > *****************************************************************************/ > > STATUS > intDisconnect(int vector) > { > > INT_HANDLER_DESC *handler; > > if (vector < 16) > { > printf("ERROR: cannot disconnect a local bus ISR with > intDisconnect\n"); > return(ERROR); > } > if (vector > 255) > { > printf("ERROR: interrupt vector out of range\n"); > return(ERROR); > } > > handler = (INT_HANDLER_DESC *)sysIntTbl[(int)vector]; > > if (handler != NULL) > { > printf("disconnecting vector %d \n", vector); > > handlerDelete(handler); > sysIntTbl[(int)vector] = NULL; > > } > return(OK); > > } > > /****************************************************************************** > > * > * handlerDelete() - recursively deletes ISRs from sysIntTbl vector > * > *****************************************************************************/ > > LOCAL int > handlerDelete(INT_HANDLER_DESC *handler) > { > if (handler->next != NULL) > { > handlerDelete(handler->next); > } > > free(handler); > handler = NULL; > > return(OK); > > } --------------------------- Newsgroups: comp.os.vxworks Subject: rawFsLib (Raw Filesystem) Date: Wed, 08 Mar 2000 10:39:48 +0000 From: Paul O'Reilly Organization: Nortel Networks Message-ID: <38C62DF4.CA00FF39@nortelnetworks.com> I have had trouble with File Corruption using the the DOS Filesystem libraries provided with vxWorks. I am looking to implement a small Raw Filesystem for two or three crucial configuration files that our system requires. The "Raw Block Device File System" provided by vxWorks is by it's very nature quite limited. I was wondering if 1. Anyone has seen (or written) a layer above this which facilitates such operations as open, close, read, write using standard filenames etc. 2. Any decent code examples using the "Raw Filesystem" (Assuming NO to 1. above) exist. Any replies would be appreciated, Paul O'Reilly. E-Mail : oreillyp@nortelnetworks.com --------------------------- Newsgroups: comp.os.vxworks Subject: TFFS (& Raw Filesystems) Date: Wed, 08 Mar 2000 10:54:01 +0000 From: Paul O'Reilly Organization: Nortel Networks Message-ID: <38C63149.DCE1A48E@nortelnetworks.com> I was wondering if anyone has used TFFS in conjunction with the "Raw Filesystem" libraries provided by vxWorks. Are the two compatible .... ?? According to the vxWorks documentation tffsDevCreate() returns a pointer to a Block Device Structure (BLK_DEV *) suitable for use with dosFs (see example below). IS THIS BLOCK DEVICE ALSO SUITABLE FOR RawFs .............. ?? RAW_VOL_DESC *rawFsDevInit ( char * volName, /* volume name */ BLK_DEV * pBlkDev /* pointer to block device info */ ) Any replies would be very much appreciated, Paul O'Reilly. E-Mail : oreillyp@nortelnetworks.com Example DOS Code Snippet - ------------------ BLK_DEV *pBlkDev; DOS_VOL_DESC *dVolDesc; /* Create a TrueFFS block device suitable for use with dosFs */ if((pBlkDev = tffsDevCreate(TFFS_FLASH_DRIVE_NO, REMOVEABLE_MEDIA_FLAG)) == (BLK_DEV *)NULL) { printf("dosFsCdriveTffsDrv : tffsDevCreate() FAILED\n"); return; } /* Associates a block device with dosFs file file system functions */ if((dVolDesc = dosFsDevInit("/C:", pBlkDev, (DOS_VOL_CONFIG *)NULL)) == (DOS_VOL_DESC *)NULL) { printf("dosFsCdriveTffsDrv : dosFsDevInit() FAILED\n"); return; } --------------------------- Newsgroups: comp.os.vxworks Subject: TFFS (& Raw Filesystems) Date: Wed, 08 Mar 2000 11:26:14 +0000 From: Paul O'Reilly Organization: Nortel Networks Message-ID: <38C638D6.F0730DCF@nortelnetworks.com> Hi everyone, I was wondering if anyone has used TFFS in conjunction with the "Raw Filesystem" libraries provided by vxWorks. Are the two compatible .... ?? According to the vxWorks documentation tffsDevCreate() returns a pointer to a Block Device Structure (BLK_DEV *) suitable for use with dosFs (see example below). IS THIS BLOCK DEVICE ALSO SUITABLE FOR RawFs .............. ?? RAW_VOL_DESC *rawFsDevInit ( char * volName, /* volume name */ BLK_DEV * pBlkDev) Any replies would be very much appreciated, Paul O'Reilly. E-Mail : oreillyp@nortelnetworks.com Example DOS Code Snippet ************************ BLK_DEV *pBlkDev; DOS_VOL_DESC *dVolDesc; /* Create a TrueFFS block device suitable for use with dosFs */ if((pBlkDev = tffsDevCreate(TFFS_FLASH_DRIVE_NO, REMOVEABLE_MEDIA_FLAG)) == (BLK_DEV *)NULL) { printf("dosFsCdriveTffsDrv : tffsDevCreate() FAILED\n"); return; } /* Associates a block device with dosFs file file system functions */ if((dVolDesc = dosFsDevInit("/C:", pBlkDev, (DOS_VOL_CONFIG *)NULL)) == (DOS_VOL_DESC *)NULL) { printf("dosFsCdriveTffsDrv : dosFsDevInit() FAILED\n"); return; } --------------------------- Newsgroups: comp.os.vxworks Subject: rawFsLib (Raw Filesystem) Date: Wed, 08 Mar 2000 11:25:33 +0000 From: Paul O'Reilly Organization: Nortel Networks Message-ID: <38C638AD.7A991F44@nortelnetworks.com> Hi everyone, I have had trouble with File Corruption using the the DOS Filesystem libraries provided with vxWorks. I am looking to implement a small Raw Filesystem for two or three crucial configuration files that our system requires. The "Raw Block Device File System" provided by vxWorks is by it's very nature quite limited. I was wondering if 1. Code exists for a layer above this which facilitates such operations as open, close, read, write using standard filenames etc. 2. Any decent code examples using the "Raw Filesystem" (Assuming NO to 1. above) exist. Any replies would be appreciated, Paul O'Reilly. E-Mail : oreillyp@nortelnetworks.com --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Mar 8 09:16:48 2000 From: David Laight Date: Wed Mar 8 09:16:50 PST 2000 Subject: re: changing the net device on a running system I had to resolve a similar problem. In my case there are drivers for several pcmcia network cards built into the vxWorks kernel (loading standalone from ROM, no bootrom code). If a card is plugged in the IP stack needs to come up on that interface, when the card is removed the stack should be torn down. At boot time a card may, or may not be present. To get this working I inverted the network startup code. The onus is now on the network card driver to initialise the IP stack, not on the system startup. Starting from the T2 network startup code (compiles fine for T1): 1) Rip out the code from usrNetInit that starts anything except "lo0" (and loads of other stuff that is not needed from this file so I could see what is going on...) 2) Ensure usrNetInit initialises the mux, dhcp and anything else you need 3) To add an interface: - call muxDevLoad() the first time an interface appears - call ipAttach() - call either usrNetIfConfig() (static config) or use dhcp for dynamic config I compare the interface name to that in sysBootParams, and use the static values if it matches. Note that the issued dhcp code will crash your system and not work... 4) To remove an interface: - ipDetach() does almost everything. However some routes get left behind and need to be deleted by hand. Also you need to wait until dhcp has finished playing with the lease. The static rtTblShow(0) will dump the entire table - it needs to have nothing left for your interface if you want to be able to start it again. You probably also need to: - remove the ioDefPathSet( devName ) call from usrNetInit. - always creat sysSymTbl - in usrRoot - but don't call netLoadSymTbl, maybe load a hand-crafted standTbl Good hacking. 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 Thu Mar 9 04:03:11 2000 From: Vxworks Exploder Date: Thu Mar 9 04:03:13 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Mar 9 04:03:08 PST 2000 Subject: RAW Socket Multicast Subject: Re: VxFusion and I20 Subject: Re: Free source code for any RTOS ? Subject: DVB Subject: Re: VxFusion and I20 Subject: Target Server Won't Connect Subject: Re: Building bootable images Subject: Tornado Debugger and Shell Symboltables out of sync Subject: Re: Target Server Won't Connect Subject: Re: STL list.h where are they? Subject: Re: dosFs and SCSI - HELP Subject: SCSI drive access problem Subject: Re: about tornadoII Subject: reading a message from pipe without removing it ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: RAW Socket Multicast Date: Wed, 08 Mar 2000 08:54:54 -0500 From: John Sestak Organization: Raytheon, Falls Church Message-ID: <38C65BAE.C3B3F19D@fallschurch.esys.com> Hi all, Is VxWorks capable of "Raw socket multicast sending/receiving"? Say I create a raw socket, is there a function call, i.e. setsockopt(), which I can call to put the socket in multicast mode so I can send and recv multicast data? Thank you, John Sestak --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxFusion and I20 Date: Wed, 08 Mar 2000 14:15:16 +0000 From: "Ag@Whd" Organization: White Horse Design Message-ID: References: <2000030719242783839@zetnet.co.uk> Reply-To: whd@catalpa.zetnet.zet On Tue, 7 Mar 2000 19:24:27 GMT, David & Bella Aldrich scribbled these words: >We are considering how to implement messaging between 2 PCI cards, >each having a PowerPC and running VxWorks. > >VxFusion is a possibility for the application layer interface but has >the disadvantage of being fairly expensive. We are also looking at >I20 for the physical layer. We could just use I20 and write our own >application layer interface. A possibility. >1) Has anyone experience of using VxFusion? >2) Is there an adapter layer available for PCI or I20? >3) Is an I20 driver available for PowerPC and VxWorks? No, don't know, and "possibly". >I would welcome any comments or advice please. Why not simply implement a shared memory interface on each card? Either your own implementation or use VxWorks shared memory support? Adrian WWW WWW Adrian Gothard WWW ww WWW White Horse Design WWWWWWWWWW WWWW WWWW rot13 to reply: juq@mrgarg.pb.hx - --- Designers of GPS-based satellite tracking systems for vehicles --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Free source code for any RTOS ? Date: Wed, 08 Mar 2000 15:13:24 +0000 From: "Ag@Whd" Organization: White Horse Design Message-ID: References: <38B5DD8B.6E89B2F9@blr.pin.philips.com> Reply-To: whd@catalpa.zetnet.zet On Thu, 24 Feb 2000 17:40:27 -0800, vikram khandpur scribbled these words: >I have to write a basic RTOs for 8051 platform. >Is there any free source code for a similar RTOs available for reference >on the web ? You can get uCOS-II free from www.ucos-ii.com, though I don't see a port for the 8051. Two important requirements are a hardware stack and can store / load the registers including stack pointer either on the stack or in memory. Processors like the 6805 don't meet this requirement so "cannot" run uCOS-II (easily). I suspect the 8051 is too limited, though there are other approaches that can be used, such as the tiny "RTOS" on the Arizona Microchip low end PIC processors. Investigate their application notes for the theory and code or obtain their CDROM. Adrian WWW WWW Adrian Gothard WWW ww WWW White Horse Design WWWWWWWWWW WWWW WWWW rot13 to reply: juq@mrgarg.pb.hx - --- Designers of GPS-based satellite tracking systems for vehicles --------------------------- Newsgroups: comp.os.vxworks Subject: DVB Date: Wed, 8 Mar 2000 16:22:36 -0600 From: "Rao" Organization: Esat Net Message-ID: <38c67e13.0@seat.s3two.ie> Can we need to use VxWorks for DVB device drivers or Open TV STBs? If yes/no then, What are the alternatives we can use for this other than VxWorks? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxFusion and I20 Date: Wed, 08 Mar 2000 16:26:11 GMT From: Robin Mitra Organization: Tektronix NewsReader Service Message-ID: <38C67F21.35A39B80@exgate.tek.com> References: <2000030719242783839@zetnet.co.uk> "Ag@Whd" schrieb: > > > Why not simply implement a shared memory interface on each card? > Either your own implementation or use VxWorks shared memory support? > shared memory won't work on pci. you need an atomic RMW cycle to do that. according to spec 2.1 LOCK won't pass pci bridges upstream. so unless you are using a very trivial pci bus system with only one bus vxmp can't be implemented effectivly. the idea in the pci world is to do multiprocessing on one board, where the processors talk to each others caches usin MESI or still advanced magic. multiple processors over a bus is considered at the most as 'distributed' computing using lose coupling mechanisms, out of which I2O makes the most sense. If you are using something proprietary, you can just use the HW FIFO's that are implemented in all I2O capable bridges. you still have the task of enumerating all I2O devices. and if you go for hot plug you might find yourself reinventing 'PLUG and PLAY'. it took MS a couple of years to get this working to some extent for Windows. Robin --------------------------- Newsgroups: comp.os.vxworks Subject: Target Server Won't Connect Date: Wed, 08 Mar 2000 19:45:03 GMT From: davidh@sncorp.com Organization: Deja.com - Before you buy. Message-ID: <8a6aju$2a5$1@nnrp1.deja.com> First Time VxWorks User. I have a problem getting the target server to connect. Using WRS BSP for my board (PPC555 beta). serial connection works between two PCs. I have a non-RTOS application working on the board and the serial port works in that mode, with all the same wiring. I get WDB_READY and it echoes characters from a terminal emulator when I set that debug option. If I have echoing off, when I try to launch the target server it just times out, failed to connect. If I leave echoing on, it complains 'Client can't decode result' Tornado launches it as: "tgtsvr.exe mpc555 -V -B wdbserial -Bt 5 -Br 5 -d COM1" According to the manual, I tried from the command line "tgtsvr -V mpc555 -B wdbserial -bps 9600 -d COM1" same results both times. Any tips on where to look next? david howard davidh@sncorp.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Building bootable images Date: 8 Mar 2000 20:54:07 GMT From: johan.borkhuis@peektraffic.nl Organization: XS4ALL Internet BV Message-ID: <8a6elf$aap$1@news1.xs4all.nl> References: <8a5ga2$jht5@overload.lbl.gov> Bob wrote: : I have been building a vxworks downloadable project with a file structure : shown below using Tornado 2. Is there a quick way to convert the project to : build a bootable image. I.E. not having to cut and paste all of the entries : from the makefile tool. There is an example in the Tornado getting started guide on how to build a downloadable project containing other projects. : Also I am using the PRJ_LIBS macro to include an object file I.E : $(PPC_S110)/Dskmv260x.o. Is it possible/easy to have the build process call : the makefile that build the .o file. That is possible. If you define a target (being the .o-file) and the files it depends on you can define a rule saying "cd <$(PPC_S110); make ". I used this mechanism to create the "sub-projects" build mechanism. See my homepage for 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: Tornado Debugger and Shell Symboltables out of sync Date: 8 Mar 2000 21:27:23 GMT From: "Klaus Popp" Message-ID: <01bf8945$57674e40$140101c0@popp> Hello, I have a problem using the Tornado II debugger and a PowerPC target. I can successfully load object modules from the tornado shell. I can also run the routines from the shell just loaded. But when I try to execute the same routine from the gdb I just get an error message "Symbol not found". But when I do a "load" from GDB then everything works fine (even the address of the routine is the same as within the shell). Note that I have enabled symbol table sync between target and Tornado shell and this works. Any hints? - -- Klaus Popp men GmbH, Software Development Klaus.Popp@men.de www.men.de --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Target Server Won't Connect Date: Wed, 08 Mar 2000 21:58:59 GMT From: davidh@sncorp.com Organization: Deja.com - Before you buy. Message-ID: <8a6iev$8ft$1@nnrp1.deja.com> References: <8a6aju$2a5$1@nnrp1.deja.com> never mind. I found that the build expects that RAM is zeroed out prior to loading the Vxworks load module. Since I am downloading with a BDM pod and debugger, not a VxWorks boot rom, I apparently have to zero it myself with the debugger before loading. Then it connects reliably. dave howard davidh@sncorp.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STL list.h where are they? Date: Wed, 8 Mar 2000 14:55:03 -0800 From: "Dave Craig" Organization: Verio Message-ID: References: <8a6i81$8dn$1@nnrp1.deja.com> list.h exists - remember STL is only supplied with Tornado2 and not Tornado 1.0.1. It is in the $(WIND_BASE)\target\h directory. NB: As per ANSI C++ #include replaces the use of #include Dave Craig Palo Alto, CA. wrote in message news:8a6i81$8dn$1@nnrp1.deja.com... > I can't seem to find the STL container class list.h, is this file > missing from the vxWorks STL implemenation? > > -- > Glen > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dosFs and SCSI - HELP Date: Thu, 09 Mar 2000 09:05:15 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38C74D2A.8C566A09@bitband.com> References: <38C6F2C1.CF85D5E8@adc.com> Reply-To: leonid@bitband.com Jenny, your problem is simply becuase the "cd" function you are trying to execute from WindSh runs on the HOST and not on the TARGET. Try this: -> @cd "SCSIDISK:" -> @ll Keep in mind that separation between target and host context, because WindSh is so nice it makes people forget this distinction. Leonid Jennifer Askew wrote: > I am trying to initialize the dosFs on my SCSI drive in the BSP code. > The SCSI drive seems to come up OK and the dosFs seems to be created OK > (I am assuming this because I get no errors in boot up indicating they > failed), but when I attach a target server and open a win-shell, I have > problems. If I type "devs", the device is listed (see below) > > from boot: > > Attached TCP/IP interface to dc0. > Subnet Mask: 0xffffff00 > Attaching network interface lo0... done. > Loading... 883936 > Starting at 0x30000... > > ID LUN VendorID ProductID Rev. Type Blocks BlkSize pScsiPhysDev > -- --- -------- ---------------- ---- ---- -------- ------- ------------ > 0 0 WDIGTL WDE4550 1.50 0 8891602 512 0x07ff1eb8 > Target Name: vxTarget > Attached TCP/IP interface to dc unit 0 > Attaching network interface lo0... done. > 0x7fffdf8 (tRootTask): 0 80 51 c2 1b ee > NFS client support not included. > > <..snip VxWorks logo..> > > from target shell: > > -> devs > drv name > 0 /null > 1 /tyCo/0 > 3 SCSIDISK: > 5 wst586: > 6 /pty/rlogin.S > 7 /pty/rlogin.M > 6 /pty/telnet.S > 7 /pty/telnet.M > 8 /vio > value = 0 = 0x0 > > But if I try to cd "SCSIDISK:", I get the following error: > > -> cd "SCSIDISK:" > couldn't change working directory to "SCSIDISK:": no such file or > directory > value = -1 = 0xffffffff > > Below is a snippit of the code I use to initialize the SCSI and dosFS. > Any help anyone could provide would be greatly appreciated!!! > > Thanks! > Jennifer > > Code is below > ================================== > > from usrConfig.c (unchanged by me): > void usrRoot > ( > char * pMemPoolStart, /* start of system memory > partition */ > unsigned memPoolSize /* initial size of mem pool */ > ) > <..snip..> > #ifdef INCLUDE_DOSFS > hashLibInit (); /* initialize hash table package > */ > dosFsInit (NUM_DOSFS_FILES); /* init dosFs filesystem */ > #endif /* INCLUDE_DOSFS */ > #ifdef INCLUDE_SCSI > > /* > * initialize either the SCSI1 or SCSI2 interface; initialize SCSI2 > when > * the SCSI2 interface is available. > */ > > #ifndef INCLUDE_SCSI2 > scsi1IfInit (); > #else > scsi2IfInit (); > #endif > > /* initialize SCSI controller */ > if (sysScsiInit () == OK) > { > usrScsiConfig (); /* configure SCSI > peripherals */ > } > else > { > printf ("sysScsiInit() Failed, SCSI system not initialized\n"); > } > > #endif /* INCLUDE_SCSI */ > ======================================== > > from sysScsi.c > > STATUS sysScsiConfig (void) > { > /* > * HARD DRIVE CONFIGURATION > * > * In order to configure a hard disk and initialise the dosFs > * file system. > */ > > /* configure a SCSI hard disk at busId = 0, LUN = 0 */ > > if ((gSpd20 = scsiPhysDevCreate (pSysScsiCtrl, 0, 0, 0, > NONE, 0, 0, 0)) > == (SCSI_PHYS_DEV *) NULL) > { > printErr ("usrScsiConfig: scsiPhysDevCreate failed.\n", > 0, 0, 0, 0, 0, 0); > } > else > { > > scsiShow (pSysScsiCtrl); > > /* create block devices */ > > if ((gSbd0 = scsiBlkDevCreate (gSpd20, 0, 0)) == NULL) > { > return (ERROR); > } > > if ((dosFsMkfs ("SCSIDISK:", gSbd0) == NULL) ) > { > return (ERROR); > } > > } > > return (OK); > } > > ============================= > from config.h > > /* SCSI driver configuration */ > > #if TRUE /* change FALSE to TRUE for SCSI > interface */ > #define INCLUDE_SCSI /* include SCSI interface */ > #define INCLUDE_SCSI_DMA /* include SCSI DMA interface */ > #undef INCLUDE_SCSI_BOOT /* include SCSI for boot option */ > #define INCLUDE_DOSFS /* DOS file system */ > #undef INCLUDE_TAPEFS /* tape file system */ > #undef INCLUDE_CDROMFS /* CD-ROM (iso9660) file system */ > #undef SCSI_AUTO_CONFIG /* Auto Config Scsi */ > #define SYS_SCSI_CONFIG /* BSP scsi config routine */ > #define SCSI_WIDE_ENABLE /* wide SCSI */ > #endif /* FALSE/TRUE */ > > /* Select either SCSI-1 or SCSI-2, independent of INCLUDE_SCSI */ > > #define INCLUDE_SCSI2 /* undef for SCSI-1 support */ --------------------------- Newsgroups: comp.os.vxworks Subject: SCSI drive access problem Date: Thu, 9 Mar 2000 08:37:11 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8a7krn$p5e$2@newsmaster01.nextra.at> Hello, Environment: Tornado 2, VxW 5.4 HW (Target) Intel x486 : board: ASUS PCI/I SP3G this has an onboard, PCI bus connected, NCR810 SCSI controller. now i want to access (NOT boot from) this device on a running system. anyone with an idea? - -- regards, Mario Semo. http://www.kirchnersoft.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: about tornadoII Date: Thu, 9 Mar 2000 08:31:04 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8a7krk$p5e$1@newsmaster01.nextra.at> References: <38c593ec.0@dnews.sbell.com.cn> what are your problems? - -- regards, Mario Semo. http://www.kirchnersoft.com Jazz Lam wrote in message <38c593ec.0@dnews.sbell.com.cn>... >who has dealed with the tornadoII licence? it seems not very convenient > > --------------------------- Newsgroups: comp.os.vxworks Subject: reading a message from pipe without removing it Date: Thu, 9 Mar 2000 11:09:13 +0100 From: "Markus MItterer" Organization: KPNQwest customer news service Message-ID: <0LKx4.21$xG6.1889@nreader1.kpnqwest.net> Hi VxWorkers, Currently I am using pipes and there I need the possibility to check a message without removing it from the pipe. When it is the expected message, I want to remove it. I couldn`t find such a ioctl function but I think I have seen it once. Does anybody know how to do it or any other ideas? Thank you in advance Greetings Markus Mitterer, STK --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Mar 9 06:01:39 2000 From: Joe Carissimi Date: Thu Mar 9 06:01:41 PST 2000 Subject: ORBIX and Tornado For an upcoming development effort (Power PC based running Tornado I and possibly upgrading to Tornado II), our system needs to be a CORBA client. IONA's ORBIX 3.0 has been recommended, but I am also looking at other vendors. I would like to hear from anyone who has had experience with ORBIX and Tornado (ease of integration, price, support... ). Thanks.. -- Joe Carissimi BBN Technologies Middletown, RI 02842 From vxwexplo-errs@csg.lbl.gov Thu Mar 9 06:56:47 2000 From: "Crozier, Bruce" Date: Thu Mar 9 06:56:49 PST 2000 Subject: SM networking over PCIbus Hi, SM networking works fine over PCI bus, using software T&S and either polling or mailbox interrupts (if available on both ends). Maybe not as efficient as running with hardware T&S, but often sufficient. We've run standard Tornado development tools for vxWorks over PCIbus to a PMC card, using its PCIbus host as the gateway. Basic SM networking (enough for socket I/O) does not require that you implement full vxMP. vxMP is what WRS says requires atomic RMW cycles (and even that sometimes works with software T&S). Regards, Bruce Bruce.Crozier@fci.com Force Computers Inc. -----Original Message----- Subject: Re: VxFusion and I20 Date: Wed, 08 Mar 2000 16:26:11 GMT From: Robin Mitra Organization: Tektronix NewsReader Service "Ag@Whd" schrieb: > > Why not simply implement a shared memory interface on each card? > Either your own implementation or use VxWorks shared memory support? shared memory won't work on pci. you need an atomic RMW cycle to do that. ---- Unless you are using a very trivial pci bus system with only one bus vxmp can't be implemented effectivly. Robin From vxwexplo-errs@csg.lbl.gov Thu Mar 9 08:17:11 2000 From: David Laight Date: Thu Mar 9 08:17:13 PST 2000 Subject: re: vxfusion and i20 - or why not shared memory Although PCI doesn't support RMW, software algorithms that have the desired effect do exist (see Cnuth - I think, I've never read it.). The code will have to do a back off and retry - so is not deterministic. Provided the duty cycle isn't too high there shouldn't be a problem. Whether the vxWorks shared memory can be made to use such an algorith is another matter... David (The basic idea behind s/w mutual exclusion is that you allocate a single location to each posible owner. You own the resource if you succeed in READING your location non-zero and everone elses as zero.) ---------------------------------------------------------------- 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 Fri Mar 10 04:03:27 2000 From: Vxworks Exploder Date: Fri Mar 10 04:03:29 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Mar 10 04:03:09 PST 2000 Subject: Re: reading a message from pipe without removing it Subject: Re: Errno 3d0004 observed Subject: Re: dosFs and SCSI - HELP Subject: I/O reverts to serial port Subject: Re: Q: Locating Multiple Code Segments with T2 Subject: How do I get a reboot(2) with Ctrl-X? Subject: NFS on vxWorks Subject: Re: porting on sa1110 Subject: Re: NFS on vxWorks Subject: Problem Munching on PPC gcc 2.95 Subject: Re: TFFS (& Raw Filesystems) Subject: I/O system "locks up" Subject: STL list.h where are they? Subject: Re: NFS on vxWorks Subject: Re: Tornado Debugger and Shell Symboltables out of sync Subject: Re: fatal WTX error 0x10136, 0x1012f Subject: cpu usage/idle time as value Subject: cpu usage/idle time as value Subject: including Universe.h Subject: Re: GUIs for Embedded Systems using VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: reading a message from pipe without removing it Date: Thu, 09 Mar 2000 17:28:51 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38C7C333.3C43925F@bitband.com> References: <0LKx4.21$xG6.1889@nreader1.kpnqwest.net> Reply-To: leonid@bitband.com You can not do this, unless you write your own version of the pipe driver, and add some sort of ioctl() command to do that, as standard read() arguments do not leave room for "peeking". Markus MItterer wrote: > Hi VxWorkers, > Currently I am using pipes and there I need the possibility to check a > message without removing it from the pipe. When it is the expected message, > I want to remove it. > I couldn`t find such a ioctl function but I think I have seen it once. Does > anybody know how to do it or any other ideas? > > Thank you in advance > Greetings > Markus Mitterer, STK --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Errno 3d0004 observed Date: Thu, 09 Mar 2000 10:33:11 -0500 From: "Michael P. Card" Organization: Lockheed Martin Message-ID: <38C7C437.7391CD62@lmco.com> References: <38C3F746.31C9F64F@lmco.com> <8a1v06$t4e$1@nnrp1.deja.com> This is a multi-part message in MIME format. - --------------57893C8265664A62234057BD Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit John- Thanks so much for your reply. Based on your information, I have found the culprit! What I was doing was running two programs on a PPC 604 board running Tornado 1.0.1. These programs were applications which used our real-time embedded object database manager to create, organize & otherwise manage our objects. Our real-time object database manager is written in Ada95, as are the objects we are storing in it. The first program creates a System Of Record (SOR, i.e. writable) database and populates it with some objects. It then suspends itself for a while with a taskDelay call. The second program opens a replicant (i.e. mirror-image but read-only) copy of this database and runs some retrieval queries against it and then terminates. The first program later wakes up from the taskDelay and terminates. The problem I was seeing was that after the second program terminated, I no longer saw any I/O from the first program after it woke up from the taskDelay. When I entered "i" at the Windsh prompt, I would see the mysterious 3d0004 for the first program in its ERRNO column. As it turns out, this errno value simply reflected the effect of an Ada "delay" statement in one of our Ada tasks. Our Ada compiler vendor implements the Ada "delay" statement (analogous to the VxWorks taskDelay statement) by doing a semTake() with a timer value. The timer always expires at the end of the delay period, hence the errno. The missing I/O was due to an unrelated problem which I will address in another post. - - Mike john_94501@my-deja.com wrote: > S_objLib_OBJ_TIMEOUT is set when a task is waiting for an object to > become available (eg a semaphore or a message from a message queue) with > a timeout and the timeout occurs before the object is made available. > > The documentation for semTake() does mention this errno, but, it seems, > omits to include in the explicit list of errnos at the end of the > entry... msgQReceive()'s manual does include it in the list of possible > errnos for the routine. > > taskDelay() will not cause this value to be set (nor taskUndelay(), > which sets errno to EINTR for the task that was undelayed). Are you > using semTake() or msgQReceive() with a timeout somewhere in your code > before the taskDelay() call? > > Regards, > > John... > > In article <38C3F746.31C9F64F@lmco.com>, > "Michael P. Card" wrote: > > This is a multi-part message in MIME format. > > --------------27DBD1D200D0F41F20A3A27E > > Content-Type: text/plain; charset=us-ascii > > Content-Transfer-Encoding: 7bit > > > > Hello everyone! > > > > I was wondering if any of you have ever observed an ERRNO of 3d0004 in > > your task info window. I have a task which I am suspending for while > > using taskDelay, and when I execute the "ti" command in the WindSh to > > observe the timer running down, it shows the aforementioned mysterious > > errno. > > > > I looked this up and the module corresponding to 0x003d is objLib, > which > > is not documented anywhere. Errno 4 for this module is "OBJ_TIMEOUT". > Do > > any of you know what this means? I couldn't find anything in the > > documentation about it. > > > > - Mike > > > > --------------27DBD1D200D0F41F20A3A27E > > Content-Type: text/x-vcard; charset=us-ascii; > > name="michael.p.card.vcf" > > Content-Transfer-Encoding: 7bit > > Content-Description: Card for Michael P. Card > > Content-Disposition: attachment; > > filename="michael.p.card.vcf" > > > > begin:vcard > > n:Card;Michael > > tel;fax:315-456-0441 > > tel;work:315-456-3022 > > x-mozilla-html:TRUE > > org:Lockheed Martin ;Ocean, Radar, and Sensor Systems > > version:2.1 > > email;internet:michael.p.card@lmco.com > > title:Principal Software Engineer > > adr;quoted-printable:;;Electronics Park=0D=0ABuilding 6, Room > 201;Syracuse;NY;13221;USA > > fn:Michael Card > > end:vcard > > > > --------------27DBD1D200D0F41F20A3A27E-- > > > > > > Sent via Deja.com http://www.deja.com/ > Before you buy. - --------------57893C8265664A62234057BD Content-Type: text/x-vcard; charset=us-ascii; name="michael.p.card.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Michael P. Card Content-Disposition: attachment; filename="michael.p.card.vcf" begin:vcard n:Card;Michael tel;fax:315-456-0441 tel;work:315-456-3022 x-mozilla-html:TRUE org:Lockheed Martin ;Ocean, Radar, and Sensor Systems version:2.1 email;internet:michael.p.card@lmco.com title:Principal Software Engineer adr;quoted-printable:;;Electronics Park=0D=0ABuilding 6, Room 201;Syracuse;NY;13221;USA fn:Michael Card end:vcard - --------------57893C8265664A62234057BD-- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: dosFs and SCSI - HELP Date: Thu, 09 Mar 2000 09:38:20 -0600 From: Jennifer Askew Organization: ADC Telecommunications Message-ID: <38C7C56C.CD92059B@adc.com> References: <38C6F2C1.CF85D5E8@adc.com> <38C74D2A.8C566A09@bitband.com> Thanks! Leonid Rosenboim wrote: > > Jenny, > > your problem is simply becuase the "cd" function you are trying to execute > from WindSh runs on the HOST and not on the TARGET. > > Try this: > > -> @cd "SCSIDISK:" > > -> @ll > > Keep in mind that separation between target and host context, because WindSh > is so nice it makes people forget this distinction. > > Leonid --------------------------- Newsgroups: comp.os.vxworks Subject: I/O reverts to serial port Date: Thu, 09 Mar 2000 11:03:58 -0500 From: "Michael P. Card" Organization: Lockheed Martin Message-ID: <38C7CB6E.8CDC2D33@lmco.com> This is a multi-part message in MIME format. - --------------A2B886BEF995AABA6557B7E1 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello everyone! Recently, I was running a test on a PPC 604 board running Tornado 1.0.1. I was using a TCL script in my .wind directory which would redirect the I/O from its default destination (a serial port on the board) to my WindSh or Virtual Console. My test involved two programs running on the board. The first program would run a while, then suspend itself for a while using a taskDelay() call. The second program would run while the first was suspended on the taskDelay() and then terminate. When the first program woke up from the taskDelay(), it ran normally and completed but all of the I/O it was writing went back to the serial port for some reason; it was as if the results of my TCL script had been forgotten. What I did as a test was to put my WindSh in TCL mode during the time the first program was asleep on the taskDelay(). I then reran the TCL script that was in my .wind directory, i.e. tcl> Organization: Lucent Technologies Message-ID: <38C7EA05.E25D238@lucent.com> References: MonkeyMan wrote: > > Does anyone know an "easy" way to locate multiple code and data segments > with T2. I need some code to run out of flash (ram test code) and some to > run out of ram (flash burn code) and the bulk will run out of ram after > being copied. Creating a custom link script is what I've been told, but > T2's use of partial links seems to be getting in the way. Where I can find > information about things like .sdata, sdata2, etc... I haven't hd to do this, yet.... > > Has anyone seen a problem with elfhex in that the resultant hex file is > infinite in size with nothing but zeros for data? It is not infinite, it is just HUGE! If you have your ROM linked at something like 0x41000000 and your RAM linked at someplace like 0x1000, it will try to write out a hex file that starts at 0x1000, filled with .data,.bss, then it will fill in all the space from there until the start of rom with zeroes. Like I said, HUGE, but not infinite. Run 'make vxworks.res_rom.hex' to see the steps that WRS uses to get around this behavior. That or 'make bootrom_res.hex'. Doug. > > Regards, > Russ --------------------------- Newsgroups: comp.os.vxworks Subject: How do I get a reboot(2) with Ctrl-X? Date: Thu, 09 Mar 2000 18:06:19 GMT From: proggoddess@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a8p6o$ro2$1@nnrp1.deja.com> Hello! I've noticed that in our system, a power cycle will clear all of the memory while a reboot using Ctrl-X from the terminal will not. Digging into the VxWorks documentation, I find that Ctrl-X will call a normal reboot, or reboot(0). I'd like to make it a boot clear, reboot(2). How can I change the Ctrl-X "handler" so that it does this? Any suggestions are welcome. You can post here or email me at batanacio@clearwire.com. Thanks! - --Mina Atanacio Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: NFS on vxWorks Date: Thu, 9 Mar 2000 14:33:53 -0500 From: Burt Bicksler Organization: BBB Message-ID: Hi, Are there any known issues with NFS on vxWorks? (i386). I've built with the NFS client included and verified that it is present via nfsDrvNumGet(). I have my NFS Server running on WinNT, and that appears to be setup OK but when I attempt to mount the remote file system it fails. Any suggestions on what to look for? Thanks, Burt Bicksler P.S. If responding via e-mail remove the spam bait. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: porting on sa1110 Date: 09 Mar 2000 14:43:37 -0500 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: References: <89vd1d$epl$1@nntp.shinbiro.com> Sender: bpringle@DeadDuck >>>>> "Kyoung" == Kyoung doo Hong writes: Kyoung> Hi everybody, I'm trying to port vxworks on a customized Kyoung> board based on sa1110. Well, there are some problem Kyoung> running bootrom file on this board, Kyoung> First, when using bootrom.aif( compressed version ), the Kyoung> inflate routine always returns ERROR. i.e, inflate fails. Kyoung> Second, when using bootrom_res.aif( rom resident version Kyoung> ), the system hanged inside excVecInit() routine in Kyoung> usrInit(). When I skipped excVecInit(), the system hanged Kyoung> inside kernelInit(). Kyoung> I have 16MBytes SDRAM in bank 0, so the system has Kyoung> 16MBytes memory in contiguous block, It has mapped from Kyoung> address 0 in virtual memory. Kyoung> I worked based on the bsp of brutus board. Kyoung> There is a report about some bugs of SA1110 cpu, I think Kyoung> one of them is critical - command LDRSB/LDRSH will fail Kyoung> after read buffer is allocated. Kyoung> But I'm not sure this bug causes the errors... Kyoung> Is there anybody who has some experience porting vxworks Kyoung> on sa1110 board? I have. Also, WRS is offering an abbaset BSP for the SA1110. excVecInit() is using memory at zero to install the vector routines. Have you modified the routines in sysLib.c that setup the MMU? Take a look there. I think that you will have to change them from the Brutus BSP. hth, Bill - -- Brain surgery is more of a dextrous skill than cerebral. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: NFS on vxWorks Date: Thu, 9 Mar 2000 14:03:48 -0600 From: "Tim Berger" Organization: Motorola Message-ID: <8a9016$m048@nntp.cig.mot.com> References: Are you typing these commands into the debug shell? hostAdd "blah", "10.0.0.1" nfsMount "blah", "d:/root", "/" cd "/" Also, some tips that have helped us out: try putting your exported directory on a FAT partition. use lowercase letters only for the exported directory. Burt Bicksler wrote in message news:MPG.1331c6b06820ece1989695@nntp.ix.netcom.com... > Hi, > > Are there any known issues with NFS on vxWorks? (i386). I've built with > the NFS client included and verified that it is present via > nfsDrvNumGet(). > > I have my NFS Server running on WinNT, and that appears to be setup OK > but when I attempt to mount the remote file system it fails. > > Any suggestions on what to look for? > > Thanks, > Burt Bicksler > > P.S. If responding via e-mail remove the spam bait. --------------------------- Newsgroups: comp.os.vxworks Subject: Problem Munching on PPC gcc 2.95 Date: Thu, 09 Mar 2000 20:28:14 GMT From: wing@praxon.com Organization: Deja.com - Before you buy. Message-ID: <8a91gn$2kv$1@nnrp1.deja.com> Has anyone encountered problems munching on a PPC target using a gcc 2.95 compiler? We've tracked the problem down to the compiler NOT tagging the static initializers to global, but haven't a clue as to how to turn on the global tag. The symptoms of the problem are error messages from the linker, like this.. ctdt.o(.data+0xc)://D/swdevppc/Denali/target/config/est8260/ctdt.c: undefined reference to `global constructors keyed to myglobalobject' The code which produced this error is.. - --------------------------------- class CJunk { public: CJunk(void); int m_j; }; CJunk myglobalobject; - --------------------------------- Any help is appreciated. wing wing@praxon.com Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: TFFS (& Raw Filesystems) Date: Thu, 09 Mar 2000 22:37:48 GMT From: Andrey Kaganovsky Organization: @Home Network Message-ID: <38C827D3.9F8F6AE2@home.com> References: <38C63149.DCE1A48E@nortelnetworks.com> <38C66E59.A52187E8@bitband.com> The MS-DOS specific code in TrueFFS could be disabled by calling the routine tffsDevOptionsSet() right after the call to tffsDevCreate() (void) tffsDevOptionsSet(pBlkDev); Once you have done this, you should be able to use TrueFFS with raw filesystem. regards, Andray Leonid Rosenboim wrote: > > In principle, TFFS provides a block device so it could interoperate with any > file system, but in practice, TFFS is aware of the MS-DOS file system > structure in order to implement its wear leveling, hence it can result in > data corruption if the MS-DOS file system format is even slightly changed, > for example if the VxWorks proprietary 40-chars long file names are used. > > In other word, I don't think this will work. > > Leonid > > Paul O'Reilly wrote: > > > I was wondering if anyone has used TFFS in conjunction with the "Raw > > Filesystem" > > libraries provided by vxWorks. Are the two compatible .... ?? > > > > According to the vxWorks documentation tffsDevCreate() returns a pointer > > to a Block > > Device Structure (BLK_DEV *) suitable for use with dosFs (see example > > below). > > > > IS THIS BLOCK DEVICE ALSO SUITABLE FOR RawFs .............. ?? > > > > RAW_VOL_DESC *rawFsDevInit ( char * volName, /* volume name > > */ > > BLK_DEV * > > pBlkDev /* pointer to block device info */ ) > > > > Any replies would be very much appreciated, > > Paul O'Reilly. > > > > E-Mail : oreillyp@nortelnetworks.com > > > > Example DOS Code Snippet > > ------------------ > > BLK_DEV *pBlkDev; > > DOS_VOL_DESC *dVolDesc; > > > > /* Create a TrueFFS block device suitable for use with dosFs > > */ > > > > if((pBlkDev = tffsDevCreate(TFFS_FLASH_DRIVE_NO, > > REMOVEABLE_MEDIA_FLAG)) > > == (BLK_DEV *)NULL) > > { > > printf("dosFsCdriveTffsDrv : tffsDevCreate() FAILED\n"); > > return; > > } > > > > /* Associates a block device with dosFs file file system functions > > */ > > > > if((dVolDesc = dosFsDevInit("/C:", pBlkDev, (DOS_VOL_CONFIG *)NULL)) > > > > == (DOS_VOL_DESC *)NULL) > > { > > printf("dosFsCdriveTffsDrv : dosFsDevInit() FAILED\n"); > > return; > > } - -- Andray Kaganovsky M-Systems andrayk@m-sys.com --------------------------- Newsgroups: comp.os.vxworks Subject: I/O system "locks up" Date: Thu, 09 Mar 2000 17:17:48 -0500 From: "Lawrence T. Hoff" Organization: Brookhaven National Laboratory Message-ID: <38C8230C.637EE184@bnl.gov> Recently, some of our VxWorks systems have started failing with a very distinctive "signature". I'm at a loss as to where to start looking for the problem, and I'm hoping for some suggestions. The problem is that the I/O system seems to "lock up". When this happens, then it is no longer possible to open or create files, or to create new network connections. I can "cure" the problem by executing the command iosUnlock(), but I'd rather understand and fix the problem, rather than cure it when it shows up. Here are some task traces from when the condition occurs (in case they are illuminating) : The telnet task : - -> tt 0x1bd4078 tt 0x1bd4078 190ec8 vxTaskEntry +60 : telnetd () 13f704 telnetd +130: accept () 13efe8 accept +50 : bsdAccept () 143afc bsdAccept +1c8: iosFdNew () 1539a8 iosFdNew +2c : 153f8c () 153fa8 iosIoctl +154: semTake () 187768 semTake +50 : semBTake () value = 0 = 0x0 A task writing to the DOS file-system based RAMdisk : - -> tt 0xb7b808 tt 0xb7b808 190ec8 vxTaskEntry +60 : helper__20backgroundClockEventP20backgroundClockEventUl (b7ba88, 5) 1a8b2c0 helper__20backgroundClockEventP20backgroundClockEventUl+3c : doIt__20backgroundClockEve ntP20backgroundClockEvent (b7ba88) 1a8b300 doIt__20backgroundClockEventP20backgroundClockEvent+24 : eventOccurred__8fecEventPv () 1a885a4 eventOccurred__8fecEventPv+9c : adoWriteArchive__3ADOP3ADO () 1a71a18 adoWriteArchive__3ADOP3ADO+28 : WriteArchive__3ADO () 1a6ad08 WriteArchive__3ADO+b4 : creat () 152488 creat +8c : ioFullFileNameGet () 152bdc ioFullFileNameGet+44 : pathCat () 156948 pathCat +74 : iosDevFind () 153580 iosDevFind +24 : 1535e4 () 15360c iosDevFind +b0 : 153f8c () 153fa8 iosIoctl +154: semTake () 187768 semTake +50 : semBTake () value = 0 = 0x0 A task trying to open "/null" : - -> tt t130 tt t130 190ec8 vxTaskEntry +60 : open () 152690 open +a4 : ioFullFileNameGet () 152bdc ioFullFileNameGet+44 : pathCat () 156948 pathCat +74 : iosDevFind () 153580 iosDevFind +24 : 1535e4 () 15360c iosDevFind +b0 : 153f8c () 153fa8 iosIoctl +154: semTake () 187768 semTake +50 : semBTake () value = 0 = 0x0 I have more, but you probably get the idea. TIA -- Larry - -- Building 911-C, Brookhaven National Laboratory PO Box 5000, Upton, NY, 11973 http://www.rhichome.bnl.gov/~hoff, mailto:hoff@bnl.gov Phone: (516) 344-2194 FAX: (516) 344-5443 --------------------------- Newsgroups: comp.os.vxworks Subject: STL list.h where are they? Date: Wed, 08 Mar 2000 21:55:16 GMT From: glenrm@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8a6i81$8dn$1@nnrp1.deja.com> I can't seem to find the STL container class list.h, is this file missing from the vxWorks STL implemenation? - -- Glen Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: NFS on vxWorks Date: Fri, 10 Mar 2000 09:04:13 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8aaak5$8a4$3@newsmaster01.nextra.at> References: a) q question. what NFS Server are you using under NT? is there a free one available? b) we have an NFS Server setup on LINUX and use it under VxW2.0 (nfsClient+mount). i will ask my team how they setup the nfs. - -- regards, Mario Semo. http://www.kirchnersoft.com Burt Bicksler wrote in message ... >Hi, > >Are there any known issues with NFS on vxWorks? (i386). I've built with >the NFS client included and verified that it is present via >nfsDrvNumGet(). > >I have my NFS Server running on WinNT, and that appears to be setup OK >but when I attempt to mount the remote file system it fails. > >Any suggestions on what to look for? > >Thanks, >Burt Bicksler > >P.S. If responding via e-mail remove the spam bait. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tornado Debugger and Shell Symboltables out of sync Date: Fri, 10 Mar 2000 08:25:37 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8aaak4$8a4$1@newsmaster01.nextra.at> References: <01bf8945$57674e40$140101c0@popp> juast in case. the following tip solved the problem (exactly this problem) on WinNT: - -----Original Message----- From: Lori Fraleigh [mailto:lori@rti.com] Sent: Wednesday, January 12, 2000 8:01 PM To: mario.semo@kirchnersoft.com Subject: Re: fatal WTX error 0x10136, 0x1012f In Tornado, go to Tools->Options. Then scroll over to the right tab to the Tornado Registry. Select remote registry instead of local registry and type in the name of your host (mariosemo). Then restart the target server and reload your modules. There is a bug in the WRS that requires this to get symSyncLib to work on Windows. Regards, Lori Mario Semo wrote: > > Hello, > > what are the effects you had? > > i have a registry running (from installation) and in the regsitry i see: > > wtxregd (wtxregd@mariosemo): Wed Jan 12 08:07:19 2000 > Wind River Systems Tornado Registry: NT/Win95 version > Tornado Registry initializing ... succeeded. > Tornado Registry restoring data base from file ... succeeded. > Tornado Registry Server Installing ... succeeded. > > and i can load the target server and i can start tornado and can run > applications in tornado - in the shell and the apps are running on the > target machine. > > the problems i have are: > > when i load an application direct on the server (in the shell on the server > machine - e.g. at boot time with a boot script) the Tornado Shell is not > aware of the symbols an of the modules loaded. when i type: > "moduleShow " > i see nothing. even when something is loaded. (i see what is loaded when i > type @moduleShow). > > ok, so i thought that symbol table synchronization would be a good idea. > > but thats what i cannot do. either the target or the host reports an error. > > PS: i just have one network card. and i use Tornado 2. > > -- > regards, Mario Semo. http://www.kirchnersoft.com > Markus MItterer wrote in message ... > >Hello Mario, > >I suppose that you use Tornado2. There you must set up your registry (last > >tab in the option-box). You have to specify whether the registry is running > >on your local or on a remote host. My problem was, that I had two network > >cards at host side and the registry could not work with the local entry - -- Lori Fraleigh Real-Time Innovations http://www.rti.com lori@rti.com Phone: 408-734-4200 Fax: 408-734-5009 Please send all support related email to support@rti.com - -- regards, Mario Semo. http://www.kirchnersoft.com Klaus Popp wrote in message <01bf8945$57674e40$140101c0@popp>... >Hello, > >I have a problem using the Tornado II debugger and a PowerPC target. > >I can successfully load object modules from the tornado shell. I can also >run the routines from the shell just loaded. But when I try to execute the >same routine from the gdb I just get an error message "Symbol not found". >But when I do a "load" from GDB then everything works fine (even the >address of the routine is the same as within the shell). >Note that I have enabled symbol table sync between target and Tornado shell >and this works. > >Any hints? > >-- >Klaus Popp > >men GmbH, Software Development >Klaus.Popp@men.de >www.men.de > --------------------------- Newsgroups: comp.os.vxworks Subject: cpu usage/idle time as value Date: Fri, 10 Mar 2000 08:30:13 GMT From: juno_kampstra@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8aabqm$15r$1@nnrp1.deja.com> I want to view my RealTime idle time/cpu usage as an graph in my application (like the cpu usage in the NT taskmanager). Therefore I have to send the idle time/cpu usage value to my application, but how do I get the value in my RealTime (vxworks?) In the Shell there is no idle task, so I cannot use taksLib (or do I forget something?) Juno. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: cpu usage/idle time as value Date: Fri, 10 Mar 2000 08:36:20 GMT From: juno@fokkercontrol.com Organization: FCS Message-ID: <8aac64$18p$1@nnrp1.deja.com> using the spy tool I can get the following result, I however just want to have the IDLE total % as an integer number. NAME ENTRY TID PRI total % (ticks) delta % (ticks) - -------- -------- ----- --- --------------- --------------- tExcTask _excTask fbb58 0 0% ( 0) 0% ( 0) ... IDLE 95% ( 7990) 95% ( 1998) TOTAL 99% ( 8337) 98% ( 2083) In the spylib.h there are function for customised print funcitons: extern void spyLibInit (void); extern STATUS spyClkStartCommon (int intsPerSec, FUNCPTR printRtn); extern void spyCommon (int freq, int ticksPerSec, FUNCPTR printRtn); extern void spyClkStopCommon (void); extern void spyReportCommon (FUNCPTR printRtn); extern void spyStopCommon (void); extern void spyComTask (int freq, FUNCPTR printRtn); but the definition of this print FUNCPTR is not described. Could somebody give me a clue (or solution)? Juno Kampstra (juno@fokkercontrol.com) Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: including Universe.h Date: Fri, 10 Mar 2000 10:51:51 +0200 From: =?iso-8859-1?Q?Morn=E9_Hartman?= Message-ID: <38c8b783$0$35282@helios.is.co.za> Hi all, I am trying to interface to the VME bus using Tornado2 and a PcPentProVme BSP. I am trying to include the Universe.h file into my project. The build tells me that there is no rule to build the universe.h file. It is looking in my project directory though. How do I route the path to the /target/h/drv/vme directory? I have tried copying the file into my project directory and the builder accepts it, but the functions within this file are now not defined. Please Help Morne Hartman mhartman@grintek.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GUIs for Embedded Systems using VxWorks Date: Fri, 25 Feb 2000 14:15:03 -0500 From: Thomas Davis Organization: Interpath Communications, Inc. Message-ID: <38B6D4B7.7B2777BF@galois.net> References: <896dc8$q2o3@overload.lbl.gov> "Fitch, Chris" wrote: > > Fellow VxWorkers, > > We are trying to find a replacement for our current method of building GUIs: > RTX Windows and an X-Windows based GUI Builder. What other products are > available? For additional background: our system employs a Power PC Single > Board Computer with an attached Video Card. > > Thank you in advance, > Chris Fitch > LM NE&SS If you are networked, consider, ditching the video and go with an embedded web server like WindWeb. The benefits are incredible. That is the best tip I have for you. Thomas Davis - -- Galois, Inc. http://www.galois.net We color the world! --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Fri Mar 10 07:01:48 2000 From: Marco Buro Date: Fri Mar 10 07:01:51 PST 2000 Subject: Motorola 8260 RTOS (!=VxWorks) Hi All, we are planning to use Motorola 8260 processor for our new equipment, so we would like to investigate on a VxWorks alternative RTOs. At the moment doesn't matter if It's a Free Source code Os or not given that we are looking around for it. Thank you very much, Marco -- ==================================================== Marco Buro SIEMENS Information and Communication Networks S.p.A. Centro Direzionale Lombardo - Palazzo CD1 Via Roma 108, 20060 Cassina de' Pecchi (MI), Italy Tel: +39-02-43886.035 Fax: +39-02-43886.088 Marco.Buro@icn.siemens.it ==================================================== ( tornado vxworks ) From vxwexplo-errs@csg.lbl.gov Fri Mar 10 12:29:19 2000 From: John Moore Date: Fri Mar 10 12:29:21 PST 2000 Subject: MV2433 is running 10 times slower than Force P750 ? Hello, I'm running some FFT's on an MVME2433-1 which is running at 350MHz, and comparing the times to a FORCE Powercore 6750 running at 300MHz. The FORCE board is completing 10 times faster for all sizes ranging from 256 points to 128k points, using the same .o file. I'm guessing the cache isn't turned on on the Motorola board, is there anything unusual to enable the cache on these boards. Thanks, John Moore vxWorks From vxwexplo-errs@csg.lbl.gov Sat Mar 11 04:03:10 2000 From: Vxworks Exploder Date: Sat Mar 11 04:03:12 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Mar 11 04:03:06 PST 2000 Subject: Re: NFS on vxWorks Subject: Interaction between network driver and m2lib Subject: Re: Using a standard off the shelf GDB for debugging Subject: Re: I/O reverts to serial port Subject: re: 64bit support Subject: Re: I/O system "locks up" Subject: NE2000 on SA1110 Subject: Special Workshop on Intelligence at the Network Edge - Final Notice Subject: Re: NE2000 on SA1110 Subject: Re: RAW Socket Multicast Subject: Re: VxWorks not finding network interface Subject: Querying routing tables ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: NFS on vxWorks Date: Fri, 10 Mar 2000 13:29:53 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8aapsh$4k8$1@newsmaster01.nextra.at> References: Hello, here is the response i got: > nfsMount "linux","/home/vxworks","/nfsdir/" more Infos see doc "nfsHelp". < - -- regards, Mario Semo. http://www.kirchnersoft.com Burt Bicksler wrote in message ... >Hi, > >Are there any known issues with NFS on vxWorks? (i386). I've built with >the NFS client included and verified that it is present via >nfsDrvNumGet(). > --------------------------- Newsgroups: comp.os.vxworks Subject: Interaction between network driver and m2lib Date: Fri, 10 Mar 2000 13:53:48 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <8aat60$693$1@ndsukns01.ndsuk.com> I'm using VxWorks 5.3.1 and the pre-SENS BSD IP stack. I have added a new IP interface (over ATM) a short while after startup - ie. not from usrNetwork. I am managing this system via SNMP, and I notice that there seem to be some problems in getting some of the status info reported properly in the interfaces ifTable. I can arrange for the counters to be properly reported, because I get a call to my driver's ioctl case SIOCGMIB2CNTRS to retrieve them; configuration is similarly retrieved via ioctl case SIOCGMIB2CNFG. But there doesn't appear to be a call to retrieve ifLastChange (the time of last change to the interface's operational state). I can see the table NET_IF_TBL * pm2IfTable declared in m2LibP.h, but I can find nothing in the BSP Porting kit or online manuals that offers any advice on what the driver needs to do to keep it up-to-date. Has anybody worked out how to do this? Will --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Using a standard off the shelf GDB for debugging Date: Fri, 10 Mar 2000 08:53:02 -0700 From: Matthew Majka Organization: Honeywell, Inc. (Defense Avionics Systems) Message-ID: <38C91A5E.D1974DBD@honeywell.com> References: <88v71l$oov$1@nnrp1.deja.com> > wrote in message news:88v71l$oov$1@nnrp1.deja.com... > > Has anyone used an off the shelf GDB to debug with? I am trying to > > debug a VxSim NT target and it always SIGSEGV's (seg fault), it doesn't > > do it if I use the T2 gdb. For many reasons I want to use an off the > > shelf GDB and not the one supplied by T2. Does anyone know why this is > > happening? If you ran VxWorks 5.3 in RDB-mode, you could use the "standard" GDB distribution to do your debugging. Presuming WRS hasn't removed support for RDB-mode (aka 5.2-mode) debugging in T2, then you should be able to do it. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I/O reverts to serial port Date: Fri, 10 Mar 2000 18:49:26 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8abg3j$rjh$1@nnrp1.deja.com> References: <38C7CB6E.8CDC2D33@lmco.com> Hi, If I remember correctly, the way that the script works is to change the global standard I/O channels to connect to the VIO channel. Now, if you have something else in your system also modifying them that will cause the change. You mentioned that you are using Ada - do you know if your Ada runtime system tries to attach to the standard I/O channels at all? You can overcome the problem by getting your task's to attach to the VIO independently (take a look at ioTaskStdSet() which is analogous to the ioGlobalStdSet() that I suspect your TCL script is using). Hope that helps, John... In article <38C7CB6E.8CDC2D33@lmco.com>, "Michael P. Card" wrote: > This is a multi-part message in MIME format. > --------------A2B886BEF995AABA6557B7E1 > Content-Type: text/plain; charset=us-ascii > Content-Transfer-Encoding: 7bit > > Hello everyone! > > Recently, I was running a test on a PPC 604 board running Tornado 1.0.1. > I was using a TCL script in my .wind directory which would redirect the > I/O from its default destination (a serial port on the board) to my > WindSh or Virtual Console. > > My test involved two programs running on the board. The first program > would run a while, then suspend itself for a while using a taskDelay() > call. The second program would run while the first was suspended on the > taskDelay() and then terminate. When the first program woke up from the > taskDelay(), it ran normally and completed but all of the I/O it was > writing went back to the serial port for some reason; it was as if the > results of my TCL script had been forgotten. > > What I did as a test was to put my WindSh in TCL mode during the time > the first program was asleep on the taskDelay(). I then reran the TCL > script that was in my .wind directory, i.e. > > tcl> > Then, when the first program came back from the taskDelay(), all of the > I/O was going to the Virtual Console just as I would expect. > > Does anyone know what can cause the I/O settings to revert back to the > board's serial port so that you have to re-run the TCL script to get the > I/O back into the Virtual Console? > > - Mike > > --------------A2B886BEF995AABA6557B7E1 > Content-Type: text/x-vcard; charset=us-ascii; > name="michael.p.card.vcf" > Content-Transfer-Encoding: 7bit > Content-Description: Card for Michael P. Card > Content-Disposition: attachment; > filename="michael.p.card.vcf" > > begin:vcard > n:Card;Michael > tel;fax:315-456-0441 > tel;work:315-456-3022 > x-mozilla-html:TRUE > org:Lockheed Martin ;Ocean, Radar, and Sensor Systems > version:2.1 > email;internet:michael.p.card@lmco.com > title:Principal Software Engineer > adr;quoted-printable:;;Electronics Park=0D=0ABuilding 6, Room 201;Syracuse;NY;13221;USA > fn:Michael Card > end:vcard > > --------------A2B886BEF995AABA6557B7E1-- > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: re: 64bit support Date: Fri, 10 Mar 2000 22:59:37 GMT From: gbabaeff@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8abuon$6sp$1@nnrp1.deja.com> There is some 64-bit support in the gcc compiler (cygnus 2.7.2) shipped with Tornado 1.0.1. I'm working with i960, and to enable access to 64-bit data, I use this: typedef unsigned long long UINT64; then UINT64 *p = ; *p = 0x0123456789abcdef; Note, however, that tools such as windsh, do NOT support 64-bit data. Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: I/O system "locks up" Date: Fri, 10 Mar 2000 16:26:40 -0700 From: "Mike Lechner" Organization: Dimensional Communications Message-ID: References: <38C8230C.637EE184@bnl.gov> Hello Lawrence, I see the same behavior as you're having. I'd like to see if iosUnlock fixes my hang, but I don't see that function anywhere. Where'd you find it? I'm using a PPP serial connection to my board (MIPS IDT 79RV4650). Sockets and telnet die within ~10 minutes. We're building a network processor chip. On our board, when we start packet forwarding, VxWorks dies in seconds. If I first blow away telnet, ppp, tNetwork, I can run for hours without problems. It feels like a different problem than the io hang, but I'm not sure yet. I've stepped through the postmortem logic analyzer code trace, and it looks pretty similar to yours. Any luck on your stuff yet? - --Mike Lawrence T. Hoff wrote in message news:38C8230C.637EE184@bnl.gov... > > Recently, some of our VxWorks systems have started > failing with a very distinctive "signature". I'm at a loss > as to where to start looking for the problem, and I'm hoping > for some suggestions. > > The problem is that the I/O system seems to "lock up". > When this happens, then it is no longer possible to open or > create files, or to create new network connections. > > I can "cure" the problem by executing the command > iosUnlock(), but I'd rather understand and fix the problem, > rather than cure it when it shows up. > > Here are some task traces from when the condition > occurs (in case they are illuminating) : > > The telnet task : > > > -> tt 0x1bd4078 > tt 0x1bd4078 > > 190ec8 vxTaskEntry +60 : telnetd () > 13f704 telnetd +130: accept () > 13efe8 accept +50 : bsdAccept () > 143afc bsdAccept +1c8: iosFdNew () > 1539a8 iosFdNew +2c : 153f8c () > 153fa8 iosIoctl +154: semTake () > 187768 semTake +50 : semBTake () > value = 0 = 0x0 > > > A task writing to the DOS file-system based RAMdisk : > > -> tt 0xb7b808 > tt 0xb7b808 > > 190ec8 vxTaskEntry +60 : > helper__20backgroundClockEventP20backgroundClockEventUl (b7ba88, 5) > 1a8b2c0 helper__20backgroundClockEventP20backgroundClockEventUl+3c : > doIt__20backgroundClockEve > ntP20backgroundClockEvent (b7ba88) > 1a8b300 doIt__20backgroundClockEventP20backgroundClockEvent+24 : > eventOccurred__8fecEventPv () > 1a885a4 eventOccurred__8fecEventPv+9c : adoWriteArchive__3ADOP3ADO () > 1a71a18 adoWriteArchive__3ADOP3ADO+28 : WriteArchive__3ADO () > 1a6ad08 WriteArchive__3ADO+b4 : creat () > 152488 creat +8c : ioFullFileNameGet () > 152bdc ioFullFileNameGet+44 : pathCat () > 156948 pathCat +74 : iosDevFind () > 153580 iosDevFind +24 : 1535e4 () > 15360c iosDevFind +b0 : 153f8c () > 153fa8 iosIoctl +154: semTake () > 187768 semTake +50 : semBTake () > value = 0 = 0x0 > > > A task trying to open "/null" : > > -> tt t130 > tt t130 > > 190ec8 vxTaskEntry +60 : open () > 152690 open +a4 : ioFullFileNameGet () > 152bdc ioFullFileNameGet+44 : pathCat () > 156948 pathCat +74 : iosDevFind () > 153580 iosDevFind +24 : 1535e4 () > 15360c iosDevFind +b0 : 153f8c () > 153fa8 iosIoctl +154: semTake () > 187768 semTake +50 : semBTake () > value = 0 = 0x0 > > > I have more, but you probably get the idea. > > TIA -- Larry > > > -- > Building 911-C, Brookhaven National Laboratory > PO Box 5000, Upton, NY, 11973 > http://www.rhichome.bnl.gov/~hoff, mailto:hoff@bnl.gov > Phone: (516) 344-2194 FAX: (516) 344-5443 --------------------------- Newsgroups: comp.os.vxworks Subject: NE2000 on SA1110 Date: 10 Mar 2000 19:57:53 -0500 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: Sender: bpringle@DeadDuck Hello, I just got an NE2000 driver working on an SA1110. Or it appears so. Here is the text from the console, Attached TCP/IP interface to ene unit 0 Attaching network interface lo0... done. NFS client support not included. VxWorks Copyright 1984-1998 Wind River Systems, Inc. CPU: TekYak - ARMSA1110 VxWorks: 5.4 BSP version: 1.2/1 Creation date: Mar 10 2000 WDB: Ready. 0x1f427c4 (tNetTask): arptnew failed on a800519 0x1f427c4 (tNetTask): arptnew failed on a800519 When I look at the network with a sniffer, the board is sending ICMP redirects when I attempt to ping it. a800519 is 10.128.5.25, which is my computer (local class IP subnet). The box attempts to send to a gateway. All of the arps are completing fine. The redirects report that an IP checksum has failed. However, when I look at the previous packets, everything is fine. Here is my bootline, #define DEFAULT_BOOT_LINE \ "ene(0,0)host:vxWorks " \ "h=10.128.5.25 e=10.128.68.69:128.128.0.0 " \ "b=10.128.68.69 g=10.128.0.1 u=bpringle tn=bill f=40" The board doesn't appear to do DHCP (or bootp). Is the syntax f=0x40 or f=64? I don't know if the subnet mask is correct either. However, I don't think the IP checksum was wrong. I have this line in configNet.h, #define NE2000_LOAD_STRING "0x20000300:0x0a:0x0a:1:0:0x00" I am unsure what the offset is suppose to be in the NE2000 init string. Does this account for the extra 4 bytes that the NE2000 appends to receive data? Maybe the MBUFs are confuse on the way to the ICMP layer? It looks like the NE2000 code takes care of this with a header struct. I think this is something stupid, like usual. tia, Bill - -- That's me herc, that's me... --------------------------- Newsgroups: comp.arch.embedded,comp.realtime,comp.robotics,sci.engr.control,comp.home.automation,comp.os.vxworks,com.os.qnx,comp.dcom.cell-relay,comp.dcom.fax Subject: Special Workshop on Intelligence at the Network Edge - Final Notice Date: Sat, 11 Mar 2000 01:48:25 GMT From: moun@usenix.org (Moun Chau) Organization: USENIX Association Keywords: Embedded Systems, QoS, SmartBox, Security, Networks, Networking, USENIX, workshops, seminar, panel, addressing, routing, infrastructure, applications, software, internet, middleware, caches, USENIX, conference, workshop, invited talks, presentation, 3Com, RSIP, servers, systems Message-ID: SPECIAL WORKSHOP ON INTELLIGENCE AT THE NETWORK EDGE Sponsored by USENIX, The Advanced Computing Systems Association with support from 3Com. March 20, 2000 Holiday Inn Fisherman's Wharf San Francisco, California, USA This event is intended to be a workshop, not a short conference or symposium. Session and panel chairs will engage the audience as active participants in the discussions. Those who are interested in identifying new opportunities in networking technology would benefit most from attending this workshop. Regardless of your level of expertise, you'll come away from this workshop with new information and practical solutions. TOPICS include: * SmartBox: An Add-on solution for Guaranteed QoS * Integrating Active Networking and Commercial Grade Routing * QoS provisioning at the network edge * RSIP: Address Sharing with End-to-End Security * Internet Addressing at the Edge * Proxy Caches Move to Middleware Pillars * Trends and Forecasts in Internet Publishing Infrastructure * Open Session on Intelligence at the Network Edge ====================================================================== Register online at: http://www.usenix.org/events/ine2000 Hotel Reservation special rates end February 25, 2000 ====================================================================== USENIX is the Advanced Computing Systems Association, an international not-for-profit society of scientists, engineers, and system administrators working on the cutting edge of systems and software. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: NE2000 on SA1110 Date: Sat, 11 Mar 2000 03:34:37 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8acesd$i4l$1@nnrp1.deja.com> References: Hi Bill, > #define DEFAULT_BOOT_LINE \ > "ene(0,0)host:vxWorks " \ > "h=10.128.5.25 e=10.128.68.69:128.128.0.0 " \ > "b=10.128.68.69 g=10.128.0.1 u=bpringle tn=bill f=40" > > The board doesn't appear to do DHCP (or bootp). Is the syntax > f=0x40 or f=64? I would go with f=64 - "f=40" will probably mean decimal 40, which is probably not what you wanted... To check, use p from the boot prompt to tell you what it thinks you entered! That will use the '0x' prefix for hex numbers in the options field. > I don't know if the subnet mask is correct > either. Not the correct format... Here's an example: 146.10.50.76:ffffff00 The IP is standard dot formatted decimal, the subnet mask is hex. The above would be for a network where (using the example above) the network address is 146.10.50.0, and the machine is address 76 on that network. If your subnet format follows the class of your IP address then you can safely leave it off. If unsure, check what your host machine is using. For Solaris use /sbin/ifconfig -a to see, under windows it should be in the network properties somewhere... Hope that's some help, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RAW Socket Multicast Date: Fri, 10 Mar 2000 21:33:03 -0800 From: brooks johnston Organization: SBC Internet Services Message-ID: <38C9DA8F.A9DF7C21@pacbell.net> References: <38C65BAE.C3B3F19D@fallschurch.esys.com> You can use setsockopt() with optname of IP_MULTICAST_IF to select your multicast interface and IP_ADD_MEMBERSHIP to (you guessed it) add membership to a multicast group. Brooks John Sestak wrote: > Hi all, > Is VxWorks capable of "Raw socket multicast sending/receiving"? Say > I create a raw socket, is there a function call, i.e. setsockopt(), > which I can call to put the socket in multicast mode so I can send and > recv multicast data? > > Thank you, > John Sestak --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks not finding network interface Date: 11 Mar 2000 07:37:01 GMT From: Diamand@btinternet.com (Luke Diamand) Organization: Customer of Planet Online Message-ID: References: <8abb2r$nhb$1@nnrp1.deja.com> This is just a wild guess, but my understanding is that vxworks will load all the devices it finds in configNet.h in your BSP. So what devices have you defined in there? It should also attach the ip stack to the interface specified in your bootline (wherever that might be, but there's also a default bootline somewhere in the BSP). Also, it won't do any of this automatic device loading unless you've enabled it when configuring your image. Can you get a console onto your target? If so, you could try loading up the device manually (muxDevLoad). I don't think you need to tweak config.h variables directly - you can adjust them from the project configuration thingy. In fact, I think the changes made to config.h in your BSP will certainly not be reflected in any projects you've got. And you probably could do worse than to try building a new project (bootable, custom configured probably) from scratch, which will certainly pick up your config.h changes. HTH! Luke Diamand In article <8abb2r$nhb$1@nnrp1.deja.com>, sue@loverso.southborough.ma.us wrote: >I am porting a product to VxWorks and am having newbie problems getting >my target available (I already have it ported to the simulator). I >am running Tornado 2.0 and VxWorks 5.4. > >I have a Sparc Ultra 2 running Solaris 2.6 as my host. >I am using the x86 BSP. Basically, I just want to get VxWorks booted >(I don't even need to use the disk on the PC) and talking to my host. > >The PC where I am running VxWorks has an Intel 82557 Ethernet card. >According to the WRS docs, I need to use the if_fei driver. However, >according to the docs, it should automatically find the network (ha). >I also have available to me a different PC with a 3Com 3c905B and I >need to use the el3c90x driver if that is available. > >I have some newbie (to VxWorks) questions. I am sure these are simple >but after a few days of reading various docs, the details are just >not there and I am sure someone can give me the answer. I suspect >there is just a step or two that I am missing: > >1. I am booting from a floppy on the PC, but when it boots it isn't >finding nor configuring my network interface. The Tornado Users Guide >v2.0 (Unix version) on pages 51-52 say that it should try to attach >to the network interface. When I boot, it says: >Attaching to floppy disk device (as I'd expect) >Loading.... >Starting at ... >[Docs say I should get an Attaching interface message now, but I don't] >I then get the logo screen and prompt. >I am using the vxworks.st image from the original boot floppy sent. >Why is it not finding my ethernet? How can I get it to do so? >The Ethernet itself is just fine as I can boot Windows from the disk >and it comes up on my Ethernet. > >2. I've modified config.h and built a new bootrom image that has >INCLUDE_FEI and INCLUDE_EL_3C90X_END defined. However, that only >makes me a new bootrom. Do I have to build a new VxWorks image? >If so, how do I do that? Is there a master project file somewhere? > >Help! Anyone, please! I have spent more time with just trying to >boot so far than I did porting the complete product. > >Susan LoVerso >Sleepycat Software >sue@loverso.southborough.ma.us > --------------------------- Newsgroups: comp.os.vxworks Subject: Querying routing tables Date: 11 Mar 2000 07:37:01 GMT From: Diamand@btinternet.com (Luke Diamand) Organization: Customer of Planet Online Message-ID: Hi! Does anyone know how to query the routing tables in vxWorks to find the next-hop IP address from the destination IP address? I'm trying to send out an IP packet on an interface without using the regular IP stack, so I need to get the next-hop IP address to do my ARPing properly. I've got as far as calling rtalloc1( paddr, 0 ) but I can't seem to make any sense of the result I get back - when I look inside the rt_entry->rt_gateway field, it just seems to be garbage (and certainly not an IP address that I recognise). Any pointers would be gratefully received! Rgrds! Luke Diamand --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Thu Mar 9 23:29:47 2000 From: Fred Roeber Date: Thu Mar 9 23:29:49 PST 2000 Subject: Re: How do I get a reboot(2) with Ctrl-X? Mina asked: > I've noticed that in our system, a power cycle will clear all of the memory > while a reboot using Ctrl-X from the terminal will not. Digging into the > VxWorks documentation, I find that Ctrl-X will call a normal reboot, or > reboot(0). I'd like to make it a boot clear, reboot(2). How can I change > the Ctrl-X "handler" so that it does this? > > Any suggestions are welcome. You can post here or email me at > batanacio@clearwire.com. Thanks! Mina, Unfortunately, the ^x functionality is hardwired to do a "reboot (BOOT_WARM_AUTOBOOT)" call when the reboot character is recognized. There isn't any way to change what type of reboot is done. Some possible ways around this (none pretty) are to: a) Change the sysToMonitor function that gets called as the last step of a reboot to override the "start type" used. This function is normally provided in sysLib.c as part of your BSP. b) Use the abort function capability of the terminal driver to do the type of reboot you want. The abort function is normally tied to ^C and is used to restart the "shell". If you can live without that capability, you could use tyAbortSet() to change the abort character to ^X and use tyAbortFuncSet() to install a function that would do the proper type of reboot (i.e. call reboot(BOOT_COLD)). You would need to make sure OPT_ABORT option was enabled for the serial connection. c) Install your own "reboot" handler that "took over" and finished the reboot process. You could use rebootHookAdd() to put in a function that did what reboot does. As I remember from when I decoded the routine it did some cache disabling and then called sysToMonitor() with the "start type" parameter. You could do those same things with your functions; calling sysToMonitor(BOOT_COLD). The first option is the easiest by far. Good luck. Fred | Fred J Roeber, BBN Systems & Technologies | | 4 John Clarke Road Middletown, RI 02842-5202 | | froeber@bbn.com 401-848-3548 | | TraceMaker product manager | | http://www.tracemaker.bbn.com | From vxwexplo-errs@csg.lbl.gov Fri Mar 10 04:03:05 2000 From: Vxworks Exploder Date: Fri Mar 10 04:03:08 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Mar 10 04:03:03 PST 2000 Subject: Re: including Universe.h ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: including Universe.h Date: Sat, 11 Mar 2000 20:47:27 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8afsor$jgb$1@newsmaster01.nextra.at> References: <38c8b783$0$35282@helios.is.co.za> i am not sure if it helps. so just in case. i had a lot of "no rule to build xxx.h" and then someone told me that i had to set a VPATH. here is the final mail. maybe it helps: - -- regards, Mario Semo. http://www.kirchnersoft.com Morné Hartman wrote in message <38c8b783$0$35282@helios.is.co.za>... >Hi all, > >I am trying to interface to the VME bus using Tornado2 and a >PcPentProVme BSP. > >I am trying to include the Universe.h file into my project. The build >tells me that there is no rule to build the universe.h file. It is >looking in my project directory though. How do I route the path to the >/target/h/drv/vme directory? > begin 666 Re Tornado header files.nws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rganization: Colt Telecom GmbH Message-ID: <8ai3ad$7jg$1@crusher.de.colt.net> References: <8afjo1$761@overload.lbl.gov> switch the CTRL-X function of for the terminal device wit hioctl function and make self checking for the keys then you can reboot how you will. Fred Roeber schrieb in im Newsbeitrag: 8afjo1$761@overload.lbl.gov... > Mina asked: > > > I've noticed that in our system, a power cycle will clear all of the memory > > while a reboot using Ctrl-X from the terminal will not. Digging into the > > VxWorks documentation, I find that Ctrl-X will call a normal reboot, or > > reboot(0). I'd like to make it a boot clear, reboot(2). How can I change > > the Ctrl-X "handler" so that it does this? > > > > Any suggestions are welcome. You can post here or email me at > > batanacio@clearwire.com. Thanks! > > Mina, Unfortunately, the ^x functionality is hardwired to do a > "reboot (BOOT_WARM_AUTOBOOT)" call when the reboot character is recognized. > There isn't any way to change what type of reboot is done. Some possible > ways around this (none pretty) are to: > > a) Change the sysToMonitor function that gets called as the last step > of a reboot to override the "start type" used. This function is normally > provided in sysLib.c as part of your BSP. > > b) Use the abort function capability of the terminal driver to do the > type of reboot you want. The abort function is normally tied to ^C > and is used to restart the "shell". If you can live without that > capability, you could use tyAbortSet() to change the abort character > to ^X and use tyAbortFuncSet() to install a function that would > do the proper type of reboot (i.e. call reboot(BOOT_COLD)). You > would need to make sure OPT_ABORT option was enabled for the serial > connection. > > c) Install your own "reboot" handler that "took over" and finished the > reboot process. You could use rebootHookAdd() to put in a function > that did what reboot does. As I remember from when I decoded the > routine it did some cache disabling and then called sysToMonitor() > with the "start type" parameter. You could do those same things with > your functions; calling sysToMonitor(BOOT_COLD). > > The first option is the easiest by far. Good luck. Fred > > | Fred J Roeber, BBN Systems & Technologies | > | 4 John Clarke Road Middletown, RI 02842-5202 | > | froeber@bbn.com 401-848-3548 | > | TraceMaker product manager | > | http://www.tracemaker.bbn.com | > --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Querying routing tables Date: Mon, 13 Mar 2000 08:58:30 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <8ai905$v7e$1@ndsukns01.ndsuk.com> References: Luke, How about the m2Lib function m2IpRouteTblEntryGet() ? Will Luke Diamand wrote in message ... >Hi! > >Does anyone know how to query the routing tables in vxWorks >to find the next-hop IP address from the destination IP >address? > >I'm trying to send out an IP packet on an interface without >using the regular IP stack, so I need to get the next-hop >IP address to do my ARPing properly. > >I've got as far as calling rtalloc1( paddr, 0 ) but I can't >seem to make any sense of the result I get back - when I look >inside the rt_entry->rt_gateway field, it just seems to be >garbage (and certainly not an IP address that I recognise). > >Any pointers would be gratefully received! > >Rgrds! >Luke Diamand > --------------------------- Newsgroups: comp.os.vxworks Subject: VxMP without VME_EXTENDED Date: Mon, 13 Mar 2000 10:26:53 GMT From: kim_jh@khgw.info.samsung.co.kr (kim, Jeong hee) Organization: Unitel and AT&T Message-ID: <38ccc0e9.10186297@88.4.1.80> Hello everyone, Does anyone have a experience to use VxMP products without VME_EXTENDED mode on MVME2700 ?? Please give me some information. Best regards, Jeong Hee Kim --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Sat Mar 11 23:00:30 2000 From: Anurag Srivastava Date: Sat Mar 11 23:00:32 PST 2000 Subject: NAT Hi, I have to implement NAT on VxWorks. I need information about its available implementation on VxWorks. How much effort is needed to port linux implementation of this ? Is there any freeware available for BSD ? Thanks, Anurag From vxwexplo-errs@csg.lbl.gov Sun Mar 12 22:40:08 2000 From: Rainer Stelzer Date: Sun Mar 12 22:40:10 PST 2000 Subject: BIOSless vxworks BSP for AMD's SC400 or SGS's STPC Hi, all BSP for Intel x86/Pentium we found until now, are BIOS based. For several reasons we do not want to use a BIOS rather than writing our own initializiation code (like we do for our 68K Systems) . Here are some of the reasons: - we do not need the functionality of changing chipset features by the user - we do not want the system start time to be increased by the BIOS - important system (chipset) settings must not be stored in a cmos-RAM - the power on/down behaviour is not changeable in the way we like - we must not handle different hardware options - .... Is a BIOSless (free or commercial) BSP for AMD's SC400 or SGS's STPC (or any other controller with intel core) already avail ? Every hint is welcom. Thanx in advance. regards Rainer ===================================================== Gebrüder Stoye GmbH department: traffic control systems name: Rainer Stelzer adr: Longericher Strasse 177 city: Cologne, Germany zip: 50739 phone: ++49/221/2616-551 mailto: r.stelzer@stoye.de www: http://www.stoye.de ===================================================== vxworks Tornado From vxwexplo-errs@csg.lbl.gov Mon Mar 13 04:03:08 2000 From: Vxworks Exploder Date: Mon Mar 13 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Mar 13 04:03:05 PST 2000 Subject: Networking issues Subject: xxx Subject: volatile? in vxWorks Subject: Re: volatile? in vxWorks Subject: Looking for MVME1604 100Mhz board or memory Subject: DMA setup latency for i960 Subject: Does WindView support PPC860? Subject: Re: Intel 82559ER driver Subject: Re: volatile? in vxWorks Subject: RtX windows config Subject: Semaphores Subject: Re: RtX windows config Subject: Re: Semaphores Subject: Re: Does WindView support PPC860? Subject: Embedded Web Strategy Subject: Dual P3 processors and Dual port Intel Enet Subject: Re: Does WindView support PPC860? Subject: ICMP redirects. Subject: Re: volatile? in vxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Networking issues Date: Mon, 13 Mar 2000 17:55:01 +0530 From: N K Jain Organization: Hughes Software Systems Message-ID: <38CCDE1C.613538D6@hss.hns.com> Hi All I am a newbie on VxWorks and embedded systems. I am having a couple of problems. 1. Im using the SLIP on my VxWorks system. The local IP address I specify using the console. (The remote IP address is negotiated by the VxWorks is negotiated by the VxWorks using the SLIP itself). In fact this application I had on Solaris, and could use the gethostname/gethostbyname pair of system calls to find my own IP address as I had a name server on the network. The VxWorks is on a private test network without any name server and so the gethostbyname call fails for me here. Is there a way out to query the VxWorks for its own IP address? If yes, how. 2. My application runs above the transport layer i.e. I use the UDP/TCP to communicate with other machines. After receiving a few packets over the network, the VxWorks prints : Interrupt: panic: netJobAdd: ring buffer overflow. After searching a while on the net, and going thru the resources pointed to by list, I find that this may be a problem if the ring buffer size mentioned while creating is very small. My query is: How do I find what ring buffer was created for this purpose. Or is there an alternate way to the problem? Im not sure if these have more to do with networking or with VxWorks? Regards Narender --------------------------- Newsgroups: comp.os.vxworks Subject: xxx Date: Mon, 13 Mar 2000 15:51:26 -0600 From: "ravi" Organization: Esat Net Message-ID: <38cd0e36.0@seat.s3two.ie> Reply-To: "ravi" --------------------------- Newsgroups: comp.os.vxworks Subject: volatile? in vxWorks Date: Tue, 14 Mar 2000 09:19:24 +0900 From: "À̵¿±Ô" Organization: Korea Telecom Message-ID: <8ak0c4$kpd$1@news1.kornet.net> Reply-To: "À̵¿±Ô" I am a embeded S/W programmer. But I am not sure that when I must use keyword "volatile". tell me.... --------------------------- Newsgroups: comp.os.vxworks Subject: Re: volatile? in vxWorks Date: Mon, 13 Mar 2000 17:56:31 -0800 From: "Michael R. Kesti" Organization: MK Associates Message-ID: <38CD9C4F.3551@gv.net> References: <8ak0c4$kpd$1@news1.kornet.net> Reply-To: mkesti@gv.net À̵¿±Ô wrote: >I am a embeded S/W programmer. >But I am not sure that when I must use keyword "volatile". >tell me.... The volatile keyword tells the compiler that it should not attempt to perform optimizations concerning the memory so declared. These optimizations are based on what the compiler can see regarding that memory's use in the file currently being compiled. One place to use this, then, is when the variable is genuinely subject to change outside of the compiler's scope. This is typically due to that memory being subject to change via hardware, such as memory mapped peripherals. A UART's registers, for example, may appear to the system as memory, but may change based on the traffic they send and receive as well as the program's flow. The programmer might use a status register to terminate a while loop, but if the optimizer sees that the loop body never alters the value of that location, will generate code that either never enters or never exits the loop. Volatile can be used in other ways, too. A fully enabled optimizer might compeltely remove this code: int index; for(index = 0 ; index < 100 ; index++); This loop might be used to generate a short delay (I know! I know! There are better ways to generate short delays, but they don't suit my example!) but because it is an empty loop and index is never used except as the loop index, the compiler is "correct" to completely remove index and the loop. Declare index as a volatile int, however, and the loop is not discarded, as the compiler will no longer make assumptions about how index might change in ways it cannot see. - -- ======================================================================== 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: Looking for MVME1604 100Mhz board or memory Date: Mon, 13 Mar 2000 19:51:56 -0800 From: "Marty" Organization: MindSpring Enterprises Message-ID: <8ak9mb$5u8$1@slb7.atl.mindspring.net> Anyone have a spare/excess MVME1604-003 PPC604 100Mhz board or a memory card 16MB or 32MB that goes on it? The tab on the board says MVME1600-001 and the label sticker says MVME1604-003. --------------------------- Newsgroups: comp.os.vxworks Subject: DMA setup latency for i960 Date: Tue, 14 Mar 2000 14:42:01 +0800 From: "Mark Evensen" Organization: IMS Netvigator Message-ID: <8akn18$5je1@imsp212.netvigator.com> We're using IxWorks for I2O development and have noticed considerable latency for DMA setup and chaining using the DMA driver supplied with Tornado/IxWorks for the i960Rx. Is this a known problem that is fixed with a particular driver and/or application note? Rgds, Mark - -- - ------------------------------------------------------------------------ Entone Technologies Ltd. Tel: +852 2824 8114 2/F, Shui On Centre Direct: +852 2244 5429 8 Harbour Road Mobile: +852 9030 7420 Wanchai, Hong Kong Fax: +852 2244 5425 --------------------------- Newsgroups: comp.os.vxworks Subject: Does WindView support PPC860? Date: Tue, 14 Mar 2000 06:52:57 GMT From: zhouxiaochen@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8aknk8$65l$1@nnrp1.deja.com> Hi all, When I add "#define INCLUDE_INSTRUMENTATION" in config.h, the target (PPC860) dead. If not, the BSP is work very well. Does WindView support PPC860 or I made some mistakes? I using Tornado1.0.1 Thanks! Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Intel 82559ER driver Date: Tue, 14 Mar 2000 11:19:18 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38CE0415.B558151F@bitband.com> References: <38CD4039.F11867FE@t-online.de> Reply-To: leonid@bitband.com Intel has a ZIP which contains drivers for the chip, including a presumably BSD driver for VxWorks (object code, not source). I have not used it but it may be worth a try. http://developer.intel.com/design/network/drivers/82559erdrivers.htm HTH, Leonid Heribert Volkhausen wrote: > Hi, > > does anyone is using an Intel 82559ER (not 82559). > I still haven't found a working driver. > > Regards > Heribert Volkhausen --------------------------- Newsgroups: comp.os.vxworks Subject: Re: volatile? in vxWorks Date: Tue, 14 Mar 2000 15:08:45 +0200 From: Leonid Rosenboim Organization: BitBand Technologies Ltd. http://www.bitband.com Message-ID: <38CE39DC.8D11DABD@bitband.com> References: <8ak0c4$kpd$1@news1.kornet.net> <38CD9C4F.3551@gv.net> <8akunn$bk9$1@nnrp1.deja.com> Reply-To: leonid@bitband.com john_94501@my-deja.com wrote: > Hi, > > > One place to use this, then, is when the variable is genuinely subject > > to change outside of the compiler's scope. This is typically due to > > that memory being subject to change via hardware, such as memory > > mapped peripherals. > > Important to remember that other tasks that may change the variable are > outside the compiler's scope too... it's not just hardware devices, > though they are the classic example since they cause the problem in a > single thread of execution. > But here, with a variable which is sahred among tasks, some sort of Mutex must be used to protect simultaneous modification of the variable, and the mutex naturally means that the variable access will be enclosed with function calls. The compiler will allways make sure the in0memory copy of the variable is updated (volatile or not) before any function call, in case there could be a side-effect of the funciton call modifying the variable. This in this case: static int sharedCounter ; SEM_ID sharedCounterMutex; ... semTake(sharedCounterMutexutex, WAIT_FOREVER); sharedCounter ++; semGive(sharedCounterMutex); the counter needs not be define "vilatile". Leonid --------------------------- Newsgroups: comp.os.vxworks Subject: RtX windows config Date: Tue, 14 Mar 2000 14:31:40 +0000 From: Brian Genisio Organization: Michigan State University Message-ID: <38CE4D4C.8F899A8A@orion-sim.com> I am trying to configure RtX Windows for VxWorks, but unfortunately, the card I am using is not directly supported, and the chip is not directly included. I am working on a DY4 760 Graphics board, which I THINK has a TI 3032 processor on it. Is this compatable with any board? How might I configure something like this? (This is a PPC 603 arch) Thanks, Brian Genisio --------------------------- Newsgroups: comp.os.vxworks Subject: Semaphores Date: Tue, 14 Mar 2000 14:36:59 -0600 From: "Satyannarayana" Organization: Esat Net Message-ID: <38ce4e42.0@seat.s3two.ie> Reply-To: "Satyannarayana" It is given that the ISRs can give the semaphores but can't take. How it is possible? How the semaphore initially will be given to ISR to release for a blocked task? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RtX windows config Date: Tue, 14 Mar 2000 16:35:51 GMT From: Robert Klinkhammer Organization: Road Runner Message-ID: <38CE6DF5.FC06BD24@mediaone.net> References: <38CE4D4C.8F899A8A@orion-sim.com> Reply-To: klink@mediaone.net It has a TMS320C60 on it, as I recall. (ex DY4 employee) Brian Genisio wrote: > > I am trying to configure RtX Windows for VxWorks, but unfortunately, the > card I am using is not directly supported, and the chip is not directly > included. I am working on a DY4 760 Graphics board, which I THINK has a > TI 3032 processor on it. Is this compatable with any board? How might > I configure something like this? (This is a PPC 603 arch) > > Thanks, > Brian Genisio --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Semaphores Date: Tue, 14 Mar 2000 12:34:15 -0500 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <38CE7817.F2A11C1D@delphi.com> References: <38ce4e42.0@seat.s3two.ie> When the semaphore is created, it is created empty. Note that you should use a binary semaphore, (semBCreate, etc) NOT a MUTEX. Any code module that has knowledge of the semaphore can give it, and any code module that has knowledge can take it, except for an ISR. The application must ensure that only the ISR gives the semaphore and that only the desired handler takes it. Note that the general case for an ISR is that it can never block for any reason. This means no printf, most (or all) sysLib functions not allowed, etc. Speaking only for myself, Joe Durusau Satyannarayana wrote: > > It is given that the ISRs can give the semaphores but can't take. > How it is possible? How the semaphore initially will be given to ISR to > release for a blocked task? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Does WindView support PPC860? Date: Tue, 14 Mar 2000 13:07:51 -0500 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <38CE7FF7.EFF09D0D@lucent.com> References: <8aknk8$65l$1@nnrp1.deja.com> I used WindView on PPC860 back in 1998. It worked fine. Using a different processor now, so I can not be sure, but I have to assume that if it worked in Tornado I that it will still work in Tornado II. Doug zhouxiaochen@my-deja.com wrote: > > Hi all, > > When I add "#define INCLUDE_INSTRUMENTATION" in config.h, the target > (PPC860) dead. > > If not, the BSP is work very well. > > Does WindView support PPC860 or I made some mistakes? > > I using Tornado1.0.1 > > Thanks! > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Embedded Web Strategy Date: Tue, 14 Mar 2000 18:13:21 -0500 From: Yibing Fan Organization: UUNET Canada News Reader Service Message-ID: <38CEC790.EF98D38A@lynx.tsb.ca> Hi Vxworkers, We have an application that has command line interface via socket (telnet). These commands basiclly calls c-functions in application. We need to add Web enabling to use those commands. Now let's consider two ways to achieve this: 1. Use Java applets. In this case, the java applets running on host giving user a GUI to the Command Line Interface. Java applets are downloaded to user's host when user first time access target, which has embedded web server (in our case GoAhead WebServer). We open additional socket to communicate with target application. The advantage of this approach is we can use the Command Line Interface as it is. It is also great to display stream data. Drawback is a lot of Java applets programming and memory space requirement for those applets on the target. 2. Use embedded HTML c-language integration, Agranat System product in my mind now. Advantage is small, effecient, fast develop as they advertised (Is that true?). It does not require a java capable browser. Disadvantage is need to change firmware and to learn their stuff. Which one would you use? What's your experience? Any comments and suggestions? Many thanks in advance. Yibing - ----------- yibing_fan@tsb.ca --------------------------- Newsgroups: comp.os.vxworks Subject: Dual P3 processors and Dual port Intel Enet Date: Tue, 14 Mar 2000 15:43:21 -0800 From: "Stephen Philibert" Message-ID: <44Az4.34$JM4.120@bunson.tor.sfl.net> Is anybody using dual processors on the Intel 440BX chipset or Intels Dual port Pro/100 server ethernet cards ? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Does WindView support PPC860? Date: Tue, 14 Mar 2000 17:39:29 -0800 From: Robert Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here Message-ID: <02de98b0.629c13cc@usw-ex0102-015.remarq.com> Yes, they do support the 860 now. I use the 860 BSP for an 823. * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free! --------------------------- Newsgroups: comp.os.vxworks Subject: ICMP redirects. Date: 14 Mar 2000 22:59:02 -0500 From: Bill Pringlemeir Organization: Factory of the mind Message-ID: Sender: bpringle@DeadDuck I am still having problems with the TCP/IP setup on VxWorks. I have no bootprom on my board. I have setup to always use the hardcoded bootline in config.h. Ie, usrConfig.c, usrBootLineInit (BOOT_CLEAR /* sysStartType */); /* crack the bootline */ config.h #define DEFAULT_BOOT_LINE \ "ene(0,0)host:/usr/target/config/sa11x0/vxWorks " \ "f=64 h=10.128.5.69 e=10.128.5.69 g=10.128.0.1" This should always copy the DEFAULT_BOOT_LINE to the BOOT_LINE_ADRS. I have printed this out several times and it is whatever I define the DEFAULT to be. bootParamsShow(bootString) gives, boot device : ene unit number : 0 processor number : 0 host name : host file name : /usr/target/config/sa11x0/vxWorks inet on ethernet (e) : 10.128.5.69 host inet (h) : 10.128.5.69 gateway inet (g) : 10.128.0.1 flags (f) : 0x40 I had the host set to my own computer. However, when I ping the board it responds with an _ICMP request_ with the source as my host computers IP and the destination as the boards IP (10.128.5.69). The MAC addresses are set like an _ICMP reply_. Also, the board sends an ICMP redirect with the address set correctly. Why would the TCP/IP stack send a _ICMP request_ with a source not it's own and the destination set to its own. I have added several 'show' routines in the demo task. The arp table seems to act as expected. Replies are added to the arp table. I have tried several host and back plane addresses. I don't think that the 'flags' are documented any where. I am now thinking of trying the BSD driver for the NE2000 instead of the END driver. Arrgh! thanks, Bill *** **** ** Device: ene Unit: 0 Description: ne2000 Enhanced Network Driver Protocol: etherInputHook Type: 257 Recv 0x8056ccc Shutdown 0x0 Protocol: IP 4.4 ARP Type: 2054 Recv 0x8029a34 Shutdown 0x8029cb8 Protocol: IP 4.4 TCP/IP Type: 2048 Recv 0x8029a34 Shutdown 0x8029c0c *** **** ** ene (unit number 0): Flags: (0x8063) UP BROADCAST MULTICAST ARP RUNNING Type: ETHERNET_CSMACD Internet address: 10.128.5.69 Broadcast address: 10.255.255.255 Netmask 0xff000000 Subnetmask 0xff000000 Ethernet address is 00:c0:1b:00:b6:81 Metric is 0 Maximum Transfer Unit size is 1500 41 packets received; 16 packets sent 38 multicast packets received 3 multicast packets sent 0 input errors; 0 output errors 0 collisions; 0 dropped *** **** ** Active Internet connections (including servers) PCB Proto Recv-Q Send-Q Local Address Foreign Address (state) - -------- ----- ------ ------ ------------------ ------------------ ------- *** **** ** total 15 badsum 0 tooshort 0 toosmall 0 badhlen 0 badlen 0 infragments 0 fragdropped 0 fragtimeout 0 forward 3 cantforward 12 redirectsent 3 unknownprotocol 0 nobuffers 0 reassembled 0 outfragments 0 noroute 0 *** **** ** LINK LEVEL ARP TABLE destination gateway flags Refcnt Use Interface - -------------------------------------------------------------------------- 10.128.0.1 00:30:96:1a:9c:9f 405 1 0 ene0 10.128.5.25 00:50:da:5d:94:68 405 1 11 ene0 - -------------------------------------------------------------------------- *** **** ** ROUTE NET TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 0.0.0.0 10.128.0.1 3 0 0 ene0 10.0.0.0 10.128.5.69 101 0 0 ene0 - ---------------------------------------------------------------------------- ROUTE HOST TABLE destination gateway flags Refcnt Use Interface - ---------------------------------------------------------------------------- 127.0.0.1 127.0.0.1 5 0 0 lo0 - ---------------------------------------------------------------------------- *** **** ** ICMP: 0 call to icmp_error 0 error not generated because old message was icmp Output histogram: routing redirect: 3 0 message with bad code fields 0 message < minimum length 0 bad checksum 0 message with bad length 0 message response generated Alot - -- Little girls, like butterflies, need no excuse. Robert Heinlein --------------------------- Newsgroups: comp.os.vxworks Subject: Re: volatile? in vxWorks Date: Wed, 15 Mar 2000 08:38:07 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ani5f$7km$1@nnrp1.deja.com> References: <8ak0c4$kpd$1@news1.kornet.net> <38CD9C4F.3551@gv.net> <8akunn$bk9$1@nnrp1.deja.com> <38CE39DC.8D11DABD@bitband.com> In article <38CE39DC.8D11DABD@bitband.com>, leonid@bitband.com wrote: > But here, with a variable which is sahred among tasks, some sort of Mutex > must be used to protect simultaneous modification of the variable, and the > mutex naturally means that the variable access will be enclosed with > function calls. The compiler will allways make sure the in0memory copy of > the variable is updated (volatile or not) before any function call, in case > there could be a side-effect of the funciton call modifying the variable. > This in this case: You are making the assumption that the mutual exclusion is function call based... there are many examples where this is not the case, but the code is quite legitimate, and often more efficient... TTFN, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Mar 13 07:04:09 2000 From: Morris John M DLVA Date: Mon Mar 13 07:04:12 PST 2000 Subject: Printing From VxWorks via Ethernet Hi, I'm pretty new to VxWorks, so I apologize if I ask a stupid question or leave out any relevant data. I was wondering if anyone has had any experience printing from VxWorks via ethernet. (I'm currently using Tornado 2.) Are there drivers (or other necessary software) that exist commercially to implement this? I also just started looking at HP's JetSend protocol for accomplishing the same thing. Has anyone had any experience with it? Would it be overkill for this? Thanks In Advance, John From vxwexplo-errs@csg.lbl.gov Mon Mar 13 18:43:04 2000 From: Fred Roeber Date: Mon Mar 13 18:43:06 PST 2000 Subject: Re: ICMP redirects. On Mon, 13 Mar 2000, Bill Pringlemeir wrote: > > I am still having problems with the TCP/IP setup on VxWorks. > I have no bootprom on my board. I have setup to always use > the hardcoded bootline in config.h. Ie, > > usrConfig.c, > usrBootLineInit (BOOT_CLEAR /* sysStartType */); /* crack the bootline */ > > config.h > #define DEFAULT_BOOT_LINE \ > "ene(0,0)host:/usr/target/config/sa11x0/vxWorks " \ > "f=64 h=10.128.5.69 e=10.128.5.69 g=10.128.0.1" > > This should always copy the DEFAULT_BOOT_LINE to the BOOT_LINE_ADRS. > I have printed this out several times and it is whatever I define > the DEFAULT to be. bootParamsShow(bootString) gives, > > boot device : ene > unit number : 0 > processor number : 0 > host name : host > file name : /usr/target/config/sa11x0/vxWorks > inet on ethernet (e) : 10.128.5.69 > host inet (h) : 10.128.5.69 > gateway inet (g) : 10.128.0.1 > flags (f) : 0x40 You were having a problem pinging your board. A large part of the problem is that you have the boards address and the host address set the same value. These should be different. Normally, "host inet" is the network address of the machine you want to boot your target from. I also notice that the flag value you have set (based on info in sysLib.h) is: #define SYSFLG_AUTOCONFIG 0x40 /* Use bootp or DHCP to get boot */ /* parameters */ Are you sure you want to use that flag? I would suggest fixing the IP address values. Fred | Fred J Roeber, BBN Systems & Technologies | | 4 John Clarke Road Middletown, RI 02842-5202 | | froeber@bbn.com 401-848-3548 | | TraceMaker product manager | | http://www.tracemaker.bbn.com | From vxwexplo-errs@csg.lbl.gov Tue Mar 14 04:03:09 2000 From: Vxworks Exploder Date: Tue Mar 14 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Mar 14 04:03:05 PST 2000 Subject: Frame grabber Subject: Re: VxWorks not finding network interface Subject: Re: VxFusion Subject: Re: BIOSless vxworks BSP for AMD's SC400 or SGS's STPC Subject: Add Device Subject: FTP Client? Subject: Anyone using Force CPU30 for networking with tornado 2 ? Subject: Re: cpu usage/idle time as value Subject: Re: intConnect, intHandlerCreate and PPC Subject: Long delay before loading symbol table Subject: Re: Long delay before loading symbol table Subject: PSO stack (NetBSD port to vxWorks) Subject: Hosting Tornado: Solaris vs. WinNT Subject: Batch Build in Tornado? Subject: Re: Printing From VxWorks via Ethernet Subject: Re: Embedded Web Strategy Subject: multi boot? Subject: Re: multi boot? Subject: Re: intConnect, intHandlerCreate and PPC Subject: Re: Long delay before loading symbol table ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Frame grabber Date: Wed, 15 Mar 2000 13:28:42 +0100 From: Xavier =?iso-8859-1?Q?Armangu=E9?= Quintana Organization: CESCA News service. Catalunya Message-ID: <38CF81FA.F5D3B32@eia.udg.es> Does any one of you know if there is a frame grabber for Intel-based platfforms? We are interested in the driver for VxWorks and the enterprise which distributes the frame grabber. Thanks - -- *********************************************** Xavier Armangué Quintana Dep. Electrònica, Informàtica i Automàtica Institut d'Informàtica i Aplicacions Universitat de Girona Avda. Lluis Santaló s/n 17003 GIRONA (Spain) Tel: 34 972 418474 Fax: 34 972 418098 email: armangue@eia.udg.es *********************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks not finding network interface Date: Wed, 15 Mar 2000 13:17:39 GMT From: csteinman8302@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ao2hh$i8i$1@nnrp1.deja.com> References: <8abb2r$nhb$1@nnrp1.deja.com> Sue (et.all); Were you able to resolve your network problem? I am in a similar situation... I am using a x86 target, with the "PC386" BSP, and cannot get VxWorks to attach to the network host. I have a dedicated Ethernet card on a Pentium PC for the host. The target has a Realtek RTL8029AS Ethernet chip on the board. I also have a serial link attached between the two, which has been verified as functional. I made the necessary changes to the Config.H file to select a compatible network device (the Realtek chip is NE2000 compatible), and added those changes in the Tornado GUI. I have tried everything I can think of, but have not made any progress on this. Any suggestions would be appreciated. Chuck Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxFusion Date: Wed, 15 Mar 2000 15:51:36 +0100 From: Gerald van Kampen Organization: =?iso-8859-1?Q?Oc=E9=2DTechnologies?= B.V. Message-ID: <38CFA378.C4B110C7@oce.nl> References: <2000030120241083839@zetnet.co.uk> We have some prototype experience using a CAN adapter layer with VxFusion. What do you like to know? Gerald David & Bella Aldrich wrote: > > Hi > > I wonder if anyone has experience of using VxFusion? It is expensive > (6k pounds) and I would like some opinions before our company goes > ahead with it. Does anyone know of the existance of a VxFusion > Adapter layer for PCI? > > Thank you > > David Aldrich --------------------------- Newsgroups: comp.os.vxworks Subject: Re: BIOSless vxworks BSP for AMD's SC400 or SGS's STPC Date: Wed, 15 Mar 2000 16:07:40 +0100 From: Gerald van Kampen Organization: =?iso-8859-1?Q?Oc=E9=2DTechnologies?= B.V. Message-ID: <38CFA73C.3949B92A@oce.nl> References: <8andbo$7f81@overload.lbl.gov> Have a look at: ftp://ftp.amd.com/pub/epd/e86/SC4x0/vxworks/elan.tar It includes a biosless startup source based on their (AMD) evaluation board. Gerald Rainer Stelzer wrote: > > Hi, > > all BSP for Intel x86/Pentium we found until now, are BIOS based. > For several reasons we do not want to use a BIOS rather than writing our > > own initializiation code (like we do for our 68K Systems) . > Here are some of the reasons: > - we do not need the functionality of changing chipset features by the > user > - we do not want the system start time to be increased by the BIOS > - important system (chipset) settings must not be stored in a cmos-RAM > - the power on/down behaviour is not changeable in the way we like > - we must not handle different hardware options > - .... > > Is a BIOSless (free or commercial) BSP for AMD's SC400 or SGS's STPC > (or any other controller with intel core) already avail ? > > Every hint is welcom. > > Thanx in advance. > > regards > > Rainer > > ===================================================== > Gebrüder Stoye GmbH > department: traffic control systems > name: Rainer Stelzer > adr: Longericher Strasse 177 > city: Cologne, Germany > zip: 50739 > phone: ++49/221/2616-551 > mailto: r.stelzer@stoye.de > www: http://www.stoye.de > ===================================================== > vxworks Tornado --------------------------- Newsgroups: comp.os.vxworks Subject: Add Device Date: Wed, 15 Mar 2000 15:10:39 -0600 From: "Satyannarayana" Organization: Esat Net Message-ID: <38cfa7a3.0@seat.s3two.ie> Reply-To: "Satyannarayana" Is this wrong syntax: iosDevAdd (Device_Descriptor_Address, Driver_Number); --------------------------- Newsgroups: comp.os.vxworks Subject: FTP Client? Date: Wed, 15 Mar 2000 10:32:26 -0500 From: "John R." Message-ID: Does the FTP client use port numbers above 5000? If so can I limit it so it does not? - -JR --------------------------- Newsgroups: comp.os.vxworks Subject: Anyone using Force CPU30 for networking with tornado 2 ? Date: Wed, 15 Mar 2000 16:40:07 +0100 From: Leif Thuresson Organization: TSL Message-ID: <38CFAED7.4F9DF2F4@tsl.uu.se> Reply-To: leif.thuresson@tsl.uu.se Hi all, I'm using Force CPU30 (BSP frc30), Motorola MVME162 (BSP mv162) boards. After upgrading to vxWorks tornado 2 the network communication is unreliable on the CPU30 boards. If I run an application with intense networking the network dies after a few minutes it doesn't even answer to ping ! The Motorola boards run the same application without any problems. I have tried both END and BSD driver with the same result. I guess the CPU30 is pretty ancient to day, but If anyone is running T2 networking on CPU30 successfully or unsuccessfully it would be interesting to know. - --leif - --------------------------------------------------------- Leif Thuresson Tel: +46-18-4713860 Uppsala University Fax: +46-18-4713833 The Svedberg Laboratory E-mail: leif.thuresson@tsl.uu.se P.O.Box 533 S-751 21 Uppsala Sweden --------------------------- Newsgroups: comp.os.vxworks Subject: Re: cpu usage/idle time as value Date: Wed, 15 Mar 2000 09:32:00 -0600 From: "Bo Yu" Organization: Motorola CIG Message-ID: <8ao9pr$gm01@nntp.cig.mot.com> References: <8aabqm$15r$1@nnrp1.deja.com> Why not create a task by yourself with a lowest priority (255). The task should run infinite loops. wrote in message news:8aabqm$15r$1@nnrp1.deja.com... > I want to view my RealTime idle time/cpu > usage as an graph in my application > (like the cpu usage in the NT taskmanager). > Therefore I have to send the idle time/cpu > usage value to my application, but how > do I get the value in my RealTime (vxworks?) > In the Shell there is no idle task, so I > cannot use taksLib (or do I forget something?) > > Juno. > > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Wed, 15 Mar 2000 15:02:12 -0500 From: Joe Durusau Organization: Lockheed Martin Corporation Message-ID: <38CFEC44.3E7C2CDA@delphi.com> References: <38C3DD67.F814FEA3@gecm.com> <38CF8BBC.A6DA0EBE@mail.ford.com> The general idea of an ISR is that it does the absolute minimum necessary to get the interrupt released (a strobe or some such) and give a semaphore or something similar that will allow the routine that does the real work to be done. This should mean that the ISR will talk at most a few microseconds to execute. The routines that are waiting for the semaphores given by the ISRs should each have a proper priority. If you have a lot of interrupts, you will lose some time in the ISRs, but the highest priority routine will get a VERY large share of available resoureces. If the highest pri routine needs even more, it can disable interrupts completely. If your system is so bogged down in interrupts that you can't tolerate normal operation, you need to consider switching to a faster processor. (Or, maybe get those who write your ISRs to rethink what they do in them). Speaking only for myself, Joe Durusau "Connolly, F. T. (Francis)" wrote: > > Just to point out what was not explicitly mentioned, if you look at > sysIntTbl[] an ISR handler element in the table has a "next handler" > which is itself an ISR handler structure. So intConnect() seems to > turn this into I think what is called a linked list if you call it > again to connect a user ISR to an interrupt vector that is already > used (i.e., the sysIntTbl[] entry to attaching to is not NULL). You > wind up with the user's ISR executing more than once. I have heard > people say that Wind River claims you can not make an intDisconnect(). > > What's sysVmeIntTbl[]? > > Now that I have your attention, have you ever tried to have a higher > priority VME interrupt preempt (or, interrupt) a lower priority one on > a PPC board? (I put something about this up on the newsgroup earlier, > but didn't get many bites.) This seems to happen as a matter of course > on 68K boards, but without a long winded story, PPC boards in general > use one processor (PCI) interrupt for all VME interrupts. On the board > we are using (a Force Pcore604 which is installed in a Kinetic Systems > V151 VXI slot-0 controller; VXI doesn't do anything to way the > interrupts function on a lower level; one can actually use a VXI > chassis as a VME chassis in most regards.), a "higher" priority VME > interrupt can not interrupt a lower priority one. This is a big > problem for our application, which we would like to port a PPC board. > I'm not an expert or terribly experienced at this, but have gotten a > bit of a confusing story about this from various places. Some people > mentioned using the task scheduler in the VME system interrupt handler > to switch user ISR "tasks", but my colleague thinks this would add > milliseconds to latency times. We need latency on the microsecond > level, as normally happens with VME interrupts. Force wants big $ (of > course) to make this happen in a custom BSP, but they might even be > talking about physically changing the board. Either way, they seem to > be talking about a custom BSP that is an order of magnitude more > expensive than the "normal" BSP, for something it seems it should do > anyway. WRS says that on some PPC boards you can change it in > firm/software. It just seems pretty absurd to me that a higher > priority VME interrupt can not interrupt a lower priority one. > > Frank Connolly > > Dave Moore wrote: > > > Hi everyone, > > I'm experiencing difficulty changing ISRs on a MCP750 PPC > > Card. > > Basically, multiple calls to intConnect appear to add the referenced > > isr > > function to existing calls. For example calling intConnect twice > > with > > the same ISR means that when the interrupt is caught, the isr is > > executed twice. This is undesirable as re-running a piece of code > > which > > uses intConnect doubles up the calls to the isr. What I'm after is a > > > > intDisconnect call but there isn't one. > > > > I either need an equivalent of intConnect that can be run multiple > > times > > using the same isr without resultant multiple calls or a method of > > clearing the effect of intConnect. > > > > I've seen several instances of code using a combination of > > intHandlerCreate() and intVecSet() to get around this problem but > > the > > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > > > Anybody got any ideas - Thanks in advance. > > > > Dave. > > David Abbott wrote: > > > Dave Moore wrote: > > > > > Hi everyone, > > > I'm experiencing difficulty changing ISRs on a MCP750 PPC > > Card. > > > Basically, multiple calls to intConnect appear to add the > > referenced isr > > > function to existing calls. For example calling intConnect twice > > with > > > the same ISR means that when the interrupt is caught, the isr is > > > executed twice. This is undesirable as re-running a piece of code > > which > > > uses intConnect doubles up the calls to the isr. What I'm after is > > a > > > intDisconnect call but there isn't one. > > > > > > I either need an equivalent of intConnect that can be run multiple > > times > > > using the same isr without resultant multiple calls or a method of > > > > > clearing the effect of intConnect. > > > > > > I've seen several instances of code using a combination of > > > intHandlerCreate() and intVecSet() to get around this problem but > > the > > > function intHandlerCreate() doesn't appear to be in my PPC bsp!. > > > > > > Anybody got any ideas - Thanks in advance. > > > > > > Dave. > > > > I have attached an "intDisconnect()" routine that has served me > > well for > > many years. Just include it > > as an extra object when you are compiling your kernel. It takes one > > argument (the interrupt vector). > > > > David Abbott > > Jefferson Lab Data Acquisition Group > > abbottd@jlab.org > > > > /* > > * File: intDisconnect.c > > * Date: September 1996 > > * > > * Description: > > * ----------- > > * This function un-installs an interrupt service routine on a > > Tornado/PPC > > board > > * > > */ > > > > #include > > #include > > #include > > #include > > #include > > #include > > #include > > > > /* > > #include > > */ > > typedef struct intHandlerDesc /* interrupt handler > > desciption */ > > { > > VOIDFUNCPTR vec; /* interrupt vector */ > > int arg; /* interrupt handler > > argument */ > > struct intHandlerDesc * next; /* next interrupt handler & > > argument */ > > } INT_HANDLER_DESC; > > > > extern sysIntTbl[]; > > extern sysVmeIntTable[]; > > STATUS intDisconnect(); > > LOCAL int handlerDelete(); > > > > /************************* > > **************************************************** > > > > * > > * intDisconnect - un-install an ISR > > * > > * This routine un-installs an ISR from the interrupt vector table > > used by > > * the local and VMEbus interrupt dispatchers. > > * > > * > > ***************************************************************************/ > > > > STATUS > > intDisconnect(int vector) > > { > > > > INT_HANDLER_DESC *handler; > > > > if (vector < 16) > > { > > printf("ERROR: cannot disconnect a local bus ISR with > > intDisconnect\n"); > > return(ERROR); > > } > > if (vector > 255) > > { > > printf("ERROR: interrupt vector out of range\n"); > > return(ERROR); > > } > > > > handler = (INT_HANDLER_DESC *)sysIntTbl[(int)vector]; > > > > if (handler != NULL) > > { > > printf("disconnecting vector %d \n", vector); > > > > handlerDelete(handler); > > sysIntTbl[(int)vector] = NULL; > > > > } > > return(OK); > > > > } > > > > / > > ***************************************************************************** > > > > * > > * handlerDelete() - recursively deletes ISRs from sysIntTbl vector > > * > > * > > ***************************************************************************/ > > > > LOCAL int > > handlerDelete(INT_HANDLER_DESC *handler) > > { > > if (handler->next != NULL) > > { > > handlerDelete(handler->next); > > } > > > > free(handler); > > handler = NULL; > > > > return(OK); > > > > } > > -- Francis Connolly Technical Specialist > FMC-FAO-PD-R&VT-P&AE-CAPE-APPTE-CEFET-P/T ET 313-594-4081 office > (1115 GTL) fconnoll@ford.com 313-317-7404 fax CDS ID: > fconnoll --------------------------- Newsgroups: comp.os.vxworks Subject: Long delay before loading symbol table Date: Wed, 15 Mar 2000 16:02:22 -0500 From: "Mark P. Fishman" Message-ID: When booting from a Sun workstation using ftp, after loading vxWorks there's a long delay before loading the vxWorks.sym symbol file. It looks like the port or socket is still in use and cannot be re-used for about 90 seconds or 2 minutes. Is there a way to reduce the time out period to something manageable, say 15 seconds, during this bootup process? Thanks -- Mark F. - -- reply to: mfishman at el el dot em eye tee dot ee dee ewe OPINIONS EXPRESSED ARE MY OWN AND DO NOT REFLECT THOSE OF MY EMPLOYER OR ANY OTHER ENTITY IN THE KNOWN WORLD. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Long delay before loading symbol table Date: Wed, 15 Mar 2000 21:04:16 GMT From: Curt McDowell Organization: Broadcom Corp. Message-ID: <38CFFACF.2B67DD51@broadcom.com> References: Reply-To: csm@broadcom.com "Mark P. Fishman" wrote: > > When booting from a Sun workstation using ftp, after loading > vxWorks there's a long delay before loading the vxWorks.sym > symbol file. It looks like the port or socket is still in use > and cannot be re-used for about 90 seconds or 2 minutes. > > Is there a way to reduce the time out period to something > manageable, say 15 seconds, during this bootup process? From an earlier posting: > >Leonid Rosenboim wrote in message news:381C0827.CC76A659@bitband.com... > > > > ndd -set /dev/tcp tcp_close_wait_interval 1000 --------------------------- Newsgroups: comp.os.vxworks Subject: PSO stack (NetBSD port to vxWorks) Date: Thu, 16 Mar 2000 00:52:42 +0300 From: "Elena A. Vengerova" Organization: TEPKOM Message-ID: <38D00629.814B0895@tepkom.ru> Reply-To: helen@tepkom.ru - --------------DA0B1A1560CECABDA229497B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi! May be, anybody knows how global variables in netinet/ip_input.c are protected when several driver threads call ipintr simultaneously? (This problem takes place only in PSO's port). Thanks, Helen - -- Elena A. Vengerova TEPKOM State Enterprise Address: Botanicheskaya 70/4-70, St.-Petersburg, Russia Office phone: 7-812-4284659 - --------------DA0B1A1560CECABDA229497B Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit 
Hi!
May be, anybody knows how global variables in netinet/ip_input.c are protected
when several driver threads call ipintr simultaneously? (This problem takes place
only in PSO's port).
Thanks,
  Helen
-- 
Elena A. Vengerova
TEPKOM State Enterprise
Address: Botanicheskaya 70/4-70, St.-Petersburg, Russia
Office phone: 7-812-4284659
  - --------------DA0B1A1560CECABDA229497B-- --------------------------- Newsgroups: comp.os.vxworks Subject: Hosting Tornado: Solaris vs. WinNT Date: Wed, 15 Mar 2000 22:12:43 GMT From: Marcelo Mourier Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: <38D00A57.B481DAD7@yahoo.com> Hi There! I've used Tornado in several projects and I'm considering using it again in my next project. I always used it under Solaris, but now I'm considering using it under WinNT. I was wondering if anybody out there has gone through this transition and would like to share his/her comments (especially the bad ones!) I'm mostly interested in the following issues: o Productivity: all our programmers are very used to Unix as their software development environment. How did the transition to WinNT affect their productivity? o Robustness: the Solaris-based Tornado tools are mature and robust. What about the WinNT tools? o Maintenence: the Solaris-based Tornado tools can be installed on an NFS server where they can be shared by all the developers. What about the WinNT version? Does each developer need to have a private installation on his/her PC, like with most IDE's? o Pricing: is there any significant price difference between the Solaris and WinNT versions? Thanks a bunch, Marcelo. --------------------------- Newsgroups: comp.os.vxworks Subject: Batch Build in Tornado? Date: Wed, 15 Mar 2000 22:54:47 GMT From: liuja@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8ap4bl$dij$1@nnrp1.deja.com> Is there a way to do batch build in Tornado for all projects within the same work space? If not, is there a way to derive a top level workspace-wide Makefile from the .wsp file? Many Thanks. JL Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Printing From VxWorks via Ethernet Date: Wed, 15 Mar 2000 17:20:40 -0600 From: Frank Minich Organization: NCP Message-ID: <38D01AC8.63D5@ti.com> References: <8ao9fo$7f72@overload.lbl.gov> Reply-To: fminich@ti.com Morris John M DLVA wrote: > > Hi, I'm pretty new to VxWorks, so I apologize if I ask a stupid question or > leave out any relevant data. I was wondering if > anyone has had any experience printing from VxWorks via ethernet. Use the lpr protocol (rfc 1179). I think you get to write it from scratch, but it's not too tough. Good luck, Frank Minich fminich@ti.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Embedded Web Strategy Date: Wed, 15 Mar 2000 14:59:02 -0500 From: Thomas Davis Organization: Interpath Communications, Inc. Message-ID: <38CFEB86.438783BD@galois.net> References: <38CEC790.EF98D38A@lynx.tsb.ca> Yibing Fan wrote: > > Hi Vxworkers, > > We have an application that has command line interface via socket > (telnet). These commands basiclly calls c-functions in application. We > need to add Web enabling to use those commands. Now let's consider two > ways to achieve this: > > 1. Use Java applets. In this case, the java applets running on host > giving user a GUI to the Command Line Interface. Java applets are > downloaded to user's host when user first time access target, which has > embedded web server (in our case GoAhead WebServer). We open additional > socket to communicate with target application. The advantage of this > approach is we can use the Command Line Interface as it is. It is also > great to display stream data. Drawback is a lot of Java applets > programming and memory space requirement for those applets on the > target. > > 2. Use embedded HTML c-language integration, Agranat System product in > my mind now. > Advantage is small, effecient, fast develop as they advertised (Is that > true?). It does not require a java capable browser. Disadvantage is > need to change firmware and to learn their stuff. > > Which one would you use? What's your experience? Any comments and > suggestions? > > Many thanks in advance. > > Yibing > ----------- > yibing_fan@tsb.ca Just started using WindWeb. In only a month and a half we have integrated into the kernel BSP and the Applications. We have replaced some old Motif apps and turned a very mature product in a totally different direction. It realy made things simple to do and development time droped like a rock. We even had time to add tons of new features never before realized as possible. We are not using Java yet but we intend to. For our first release we just use HTML/JavaScript. We are most impressed. We have already saved time and money over the old way. Thomas - -- Galois, Inc. http://www.galois.net We color the world! --------------------------- Newsgroups: comp.os.vxworks Subject: multi boot? Date: Thu, 16 Mar 2000 14:26:31 +0900 From: "À̵¿±Ô" Organization: Korea Telecom Message-ID: <8aprcq$t7b$1@news2.kornet.net> Reply-To: "À̵¿±Ô" My system must boot from ATM or Ethernet? like DOS, IS VxWorks able to multi-boot? --------------------------- Newsgroups: comp.os.vxworks Subject: Re: multi boot? Date: Thu, 16 Mar 2000 06:47:46 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8aq02h$16l$1@nnrp1.deja.com> References: <8aprcq$t7b$1@news2.kornet.net> In article <8aprcq$t7b$1@news2.kornet.net>, "À̵¿±Ô" wrote: > My system must boot from ATM or Ethernet? > like DOS, IS VxWorks able to multi-boot? I don't think it could do it out of the box, but since you get the source for the boot code you can customise it to do whatever you want... I modified one once to determine all sorts of information from the slot it was plugged into and boot the appropriate image from a set of available ones... you should have no problems trying a sequence of locations for booting from... Good luck, John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: intConnect, intHandlerCreate and PPC Date: Thu, 16 Mar 2000 07:19:51 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8aq1uk$2c9$1@nnrp1.deja.com> References: <38C3DD67.F814FEA3@gecm.com> <38CF8BBC.A6DA0EBE@mail.ford.com> Hi all, > Now that I have your attention, have you ever tried to have a higher priority VME > interrupt preempt (or, interrupt) a lower priority one on a PPC board? (I put > something about this up on the newsgroup earlier, but didn't get many bites.) This > seems to happen as a matter of course on 68K boards, but without a long winded > story, PPC boards in general use one processor (PCI) interrupt for all VME > interrupts. On the board we are using (a Force Pcore604 which is installed in a > Kinetic Systems V151 VXI slot-0 controller; VXI doesn't do anything to way the > interrupts function on a lower level; one can actually use a VXI chassis as a VME > chassis in most regards.), a "higher" priority VME interrupt can not interrupt a > lower priority one. This is a big problem for our application, which we would like > to port a PPC board. I'm not an expert or terribly experienced at this, but have > gotten a bit of a confusing story about this from various places. Some people > mentioned using the task scheduler in the VME system interrupt handler to switch > user ISR "tasks", but my colleague thinks this would add milliseconds to latency > times. We need latency on the microsecond level, as normally happens with VME > interrupts. Force wants big $ (of course) to make this happen in a custom BSP, but > they might even be talking about physically changing the board. Either way, they > seem to be talking about a custom BSP that is an order of magnitude more expensive > than the "normal" BSP, for something it seems it should do anyway. WRS says that > on some PPC boards you can change it in firm/software. It just seems pretty absurd > to me that a higher priority VME interrupt can not interrupt a lower priority one. OK, without knowing the arrangement of the hardware it's going to be a little difficult to say anything for certain here, but assuming that there is nothing in the hardware that prevents you from doing so, surely all you need is a software dispatcher connected to the VME interrupt at PCI level which re-enables the single PCI interrupt level assigned to VME. This "dispatching" handler will catch the interrupt, mask lower ones out (ideally in hardware in the VME bridge if it supports it), re-enable the VME interrupt at the PCI layer and call the registered handler. If a higher priority interrupt occurs then the running handler will be interrupted again and the bar raised to the new level... If hardware masking is not supported by the bridge (and I have no idea if this is likely to be available, although it does seem sensible to me...), then the dispatch software must be written a little more carefully. It has to check when an interrupt arrives whether a higher priority one is already being processed. If it is, then the new one should be queued somehow for when the current one ends. Without hardware support it is a little tougher, and the performance may be lower since even the high priority handlers will be interrupted briefly by lower ones, but still possible. In both cases though you should have the choice of what is higher and lower in priority by masking the appropriate things out... Does that make any sense? John... Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Long delay before loading symbol table Date: Thu, 16 Mar 2000 10:42:24 +0100 From: "Borkhuis, Johan" Organization: UUNET-NL (http://www.nl.uu.net) Message-ID: References: > ---------- > From: Mark P. Fishman[SMTP:mfishman@ll.mit.edu] > Posted At: woensdag 15 maart 2000 22:02 > Posted To: vxworks > Conversation: Long delay before loading symbol table > Subject: Long delay before loading symbol table > > When booting from a Sun workstation using ftp, after loading > vxWorks there's a long delay before loading the vxWorks.sym > symbol file. It looks like the port or socket is still in use > and cannot be re-used for about 90 seconds or 2 minutes. > > Is there a way to reduce the time out period to something > manageable, say 15 seconds, during this bootup process? > Take a look at: http://www.xs4all.nl/~borkhuis/vxworks/vxw_pt2.html#2.3-G. Here you can find a couple of possible solutions for this problem. Groeten Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan.borkhuis@peektraffic.nl | >(________|__|_[_________]_|________________________________| _/oo OOOOO oo` ooo ooo 'o!o!o o!o!o` = VxWorks page: http://www.xs4all.nl/~borkhuis/vxworks/vxworks.html = --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Tue Mar 14 06:59:30 2000 From: "Mike Anderson" Date: Tue Mar 14 06:59:33 PST 2000 Subject: RE: multi boot? VxWorks Greetings! >From: "À̵¿±Ô" >Organization: Korea Telecom > >My system must boot from ATM or Ethernet? >like DOS, IS VxWorks able to multi-boot? Yes, VxWorks is quite capable of multi-booting. However, not without some work on your part and then reburning the boot ROMs (or rebuilding the boot image if you're using BOOTP). What you have to do is to modify the bootConfig.c file in the area of the routine called the bootCmdLoop. You will need to define an alternate bootline and store it somewhere. You'll find an example in /config.h. It should look something like this: #define DEFAULT_BOOT_LINE \ "ei(0,0)host:/usr/vw/config/mv162/vxWorks h=90.0.0.3 e=90.0.0.50 u=target" #define ALTERNATE_BOOT_LINE1 \ "scsi=0,0(0,0)intructor:/u/team7/vxWorks h=192.168.32.50 e=192.168.32.17 u=team7 tn=vx7 s=/sd0/startup.cmd o=ei" #define ALTERNATE_BOOT_LINE2 \ "ei(0,0)intructor:/u/team7/vxWorks h=192.168.32.50 e=192.168.32.17 u=team7 tn=vx7 s=/u/team7/startup.cmd" ======================================================================== Then, in bootConfig.c add the following global: char secondChance = 2; Then, in bootConfig.c (bootCmdLoop) you would add something like this: /* start autoboot, unless no-autoboot specified */ bootStringToStruct (BOOT_LINE_ADRS, ¶ms); sysFlags = params.flags; if (!(sysStartType & BOOT_NO_AUTOBOOT) && !(sysFlags & SYSFLG_NO_AUTOBOOT)) { int timeout = TIMEOUT; if ((sysStartType & BOOT_QUICK_AUTOBOOT) || (sysFlags & SYSFLG_QUICK_AUTOBOOT)) { timeout = 1; } key = autoboot (timeout); /* doesn't return if successful */ } else { int timeout = TIMEOUT; /* we only get here if the autoboot above fails */ switch (secondChance) { case 2 : strcpy (BOOT_LINE_ADRS, ALTERNATE_BOOT_LINE2); secondChance--; key = autoboot (timeout); /* doesn't return if successful */ break; case 1 : timeout = TIMEOUT; strcpy (BOOT_LINE_ADRS, ALTERNATE_BOOT_LINE1); secondChance--; key = autoboot (timeout); /* doesn't return if successful */ break; } } /* If we're here, either we aren't auto-booting, or we got an error * auto-booting, or the auto-booting was stopped. */ ========================================================================= Make sure that the boot lines are configured properly and that all of the devices that you are trying to boot from are known to both the bootConfig.c (bootROM) and the usrNetwork.c and usrLoadSym.c code if you're using the two step boot process (load the O/S then the symbol table). You will have to rebuild the bootROM image for this to work. This code snippet actually supports booting from 3 different devices as "failsoft" devices. It trys the default NVRAM bootline, then ALTERNATE_BOOT_LINE2 then ALTERNATE_BOOT_LINE1. I've even had this default to dialing up via SLIP/PPP and booting via modem -- the ultimate phone home ;-) -- if all else fails. HTH, ----------------------------------------------------------------- Michael E. Anderson | Integrated Chipware Chief Scientist | 1861 Wiehle Ave. #300 | Reston, VA 20190 mike@chipware.com | (703) 736-3504 www.chipware.com (888)430-CHIP | (703) 736-3556 FAX ----------------------------------------------------------------- "Software development is like making a baby. You can't make a baby in one month by impregnating nine women. Some things just take time..." From vxwexplo-errs@csg.lbl.gov Tue Mar 14 14:06:20 2000 From: Charles_Marggraf@tsrc.net Date: Tue Mar 14 14:06:22 PST 2000 Subject: Great Opportunity at Lucent in Illinois!!!! There are many current openings for individuals with VxWorks experience at the Naperville Lucent facility. Lucent is looking to bring on over 15 individuals immidiatley. If anyone is interested in hearing more about these opportunities and have all or some of the following skills, please call me. The ideal candidates will have experience programming in C++, on Solaris and VxWorks. A background in TCP/IP, SS7, or ISDN would be a huge plus, and any experience with Rational Rose or ObjecTime is desireable. Please contact me if you feel you might be a match for this position. This is a year long contract in Naperville, IL. I look forward to speaking to you soon. Thanks, Charly Marggraf Technisource 800.330.3308 cmarggraf@tsrc.net From vxwexplo-errs@csg.lbl.gov Wed Mar 15 03:33:04 2000 From: "Marco Penso" Date: Wed Mar 15 03:33:07 PST 2000 Subject: Help on VxWorks newsgroup. This is a multi-part message in MIME format. ------=_NextPart_000_001F_01BF8FF6.513F6860 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I don't know if you are able to help me. I tried to subscribe at the newsgroup comp.os.vxworks, but I received = the message "impossible find server". Maybe I made a mistake in the name of the NNTP sever; what is its name? Thank you and best regards. _____________________________________________ Marco Penso Elettromeccanica PARIZZI S.p.A. via Fosse Ardeatine, 120 20099 Sesto S. Giovanni (MI) - Italy e-mail: marco.penso@parizzi.it Tel.: 02-244.23.403 _____________________________________________ ------=_NextPart_000_001F_01BF8FF6.513F6860 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I don't know if you are able to help = me.
 
I tried to subscribe at the newsgroup=20 comp.os.vxworks, but I received the message = "impossible find server".
Maybe I made a mistake in the name of the NNTP = sever;=20 what is its name?
 
Thank you and best regards.
_____________________________________________
 
Marco Penso
Elettromeccanica PARIZZI = S.p.A.
via Fosse=20 Ardeatine, 120
20099 Sesto S. Giovanni (MI) - Italy
e-mail: marco.penso@parizzi.it
Tel.= :=20 02-244.23.403
_____________________________________________ ------=_NextPart_000_001F_01BF8FF6.513F6860-- From vxwexplo-errs@csg.lbl.gov Wed Mar 15 04:03:08 2000 From: Vxworks Exploder Date: Wed Mar 15 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Mar 15 04:03:05 PST 2000 Subject: Re: Frame grabber Subject: Writing Protocols that use the MUX API Subject: Only (16MB) PCI memory space ? Subject: PerformancePak: StethoScope, TraceScope Subject: Re: Add Device Subject: Re: VxWorks not finding network interface Subject: "interrupt: bad vme interrupt 0" on MVME2604 Subject: memPartFree() with an invalid pointer Subject: Re: Hosting Tornado: Solaris vs. WinNT Subject: Re: Batch Build in Tornado? Subject: Great Opportunitie for VxWorks Developers!!!! Subject: Looking for real time system programmers Subject: Re: STL and VxWorks? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Frame grabber Date: Thu, 16 Mar 2000 12:46:02 -0000 From: "Dr. Pye" Organization: Customer of Planet Online Message-ID: References: <38CF81FA.F5D3B32@eia.udg.es> Xavier Armangué Quintana wrote in message news:38CF81FA.F5D3B32@eia.udg.es... > Does any one of you know if there is a frame grabber for Intel-based > platfforms? > We are interested in the driver for VxWorks and the enterprise which > distributes > the frame grabber. > http://www.datacell.co.uk/products/Components/manufac/datacell/snapper/snapp er_framegrabbers1.htm --------------------------- Newsgroups: comp.os.vxworks Subject: Writing Protocols that use the MUX API Date: Thu, 16 Mar 2000 12:58:18 GMT From: "Laszlo Szebelledy" Organization: EarthLink Inc. -- http://www.EarthLink.net Message-ID: Hi! Has anyone here had any experience writing a custom protocol for VxWorks that interfaces to the MUX? I'm looking for examples or some kind of template I could follow. Thanks, Laszlo Szebelledy --------------------------- Newsgroups: comp.os.vxworks Subject: Only (16MB) PCI memory space ? Date: Thu, 16 Mar 2000 09:25:41 -0500 From: Dan Lucas Organization: Applied Research Lab Penn State Univ. Message-ID: <38D0EEE5.92579FAA@psu.edu> I am attempting to install a PMC module on my MVME2700, the PMC module requires 16MB + 256 bytes of PCI mem space. After reading through the BSP reference I find that there is only 16MB of PCI mem space available, am I reading this right ? Eventually I would like to install a PMC_SPAN module with two PMC modules, each would require >16MB of PCI mem space. Has anyone been down this road? thanks, - -- Daniel K. Lucas email: dlucas@psu.edu Applied Research Lab phone: (814) 865-2907 Penn State Univ. fax: (814) 863-7270 P.O. Box 30 State College, Pa. 16804 --------------------------- Newsgroups: comp.os.vxworks Subject: PerformancePak: StethoScope, TraceScope Date: 16 Mar 2000 14:53:04 GMT From: hbchen@gossamer.wh.lucent.com (Hsing-bung(HB) Chen) Organization: Lucent Technologies, Whippany, New Jersey Message-ID: <8aqsgg$17s@nntpa.cb.lucent.com> Reply-To: hbchen@drmail.dr.lucent.com I am looking for information about StethoScope and TraceScope. Experience and suggestion are welcome. Can I use them for task level or procedure(function) level run-time performance monitoring? Thanks. HB Chen - -- Hsing-bung(HB) CHEN, MTS Wireless Network Group, 67 Whippany Rd., Whippany, NJ 07981 (973)386-2256 hbchen@lucent.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Add Device Date: Thu, 16 Mar 2000 17:36:33 -0000 From: "Dave Korn" Organization: (Posted via) U-NET Internet Ltd. Message-ID: References: <38cfa7a3.0@seat.s3two.ie> Satyannarayana wrote in message <38cfa7a3.0@seat.s3two.ie>... >Is this wrong syntax: > > iosDevAdd (Device_Descriptor_Address, Driver_Number); Yes. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks not finding network interface Date: Wed, 15 Mar 2000 12:18:53 -0500 From: Susan LoVerso Message-ID: <38CFC5FC.9E73E7CF@loverso.southborough.ma.us> References: <8abb2r$nhb$1@nnrp1.deja.com> <8ao2hh$i8i$1@nnrp1.deja.com> csteinman8302@my-deja.com wrote: > Sue (et.all); > > Were you able to resolve your network problem? Yes, I finally was thanks to all the various pointers that the people responding gave! My final error (as with all those that really confound) was really trivial, as I expected it would be. > I am in a similar situation... > > I made the necessary changes to the Config.H file to select a > compatible network device (the Realtek chip is NE2000 compatible), and > added those changes in the Tornado GUI. I have tried everything I > can think of, but have not made any progress on this. The change I had to make to config.h to get me further was to make sure I included the interface name in the 'other' field "o=fei". I, in all my Unix wisdom, was assigning it a number as well, "o=fei0" and it (obviously) didn't work. I see it using fd0, attaching the loopback at lo0, I figured 'fei0' was what I needed. I put too much thought into it, I guess!! :-) It also took me a while to build a correctly configured vxWorks image. I wasn't clear on whether I needed a vxWorks, vxWorks_rom, vxWorks.st or what. I ended up using just the vxWorks. Make sure you have your interface correctly configured. Under the 'vxWorks' tab in your project to build your own vxworks image, check the values of all the stuff configured under the ''network components' area. In particular, make sure it included 'network components > basic network initialization components > bootline processing components > network device name selection' so that it looks in the "other" field for the interface name. Also, make sure you have included the right drivers, I had to add the drivers from 'network components > network devices > BSD Ethernet drivers' . My target is now booted. Now I am seeing 'rpccore backend timeouts' when I attempt to start my target server. :-( One baby step at a time. That is a topic for another post, as my search of Deja and Windsurf shows rpccore coming up again and again... Sue LoVerso sue@loverso.southborough.ma.us --------------------------- Newsgroups: comp.os.vxworks Subject: "interrupt: bad vme interrupt 0" on MVME2604 Date: Wed, 15 Mar 2000 12:21:59 -0800 From: Ken Norlin Organization: NASA Dryden FRC Message-ID: <38CFF0E7.E041D527@dfrc.nasa.gov> Hi, I keep getting the error message 'interrupt: bad vme interrupt 0' on my shell window. There are 3 interrupting boards in my VME chassis. From a bus analyzer trace, I see the following sequence when I get the error message: Time BgL AM Address Data Size Cycle Stat IRQ7:1* Iack -5 310.45us ---- 29 IACK=2 FFFFFF8C LBYTE RD OK .....2. IACK2 -4 5.61us ---- 29 ....C105 ......7D LBYTE RD OK ....... ---- -3 1.39us ---- 29 ....C107 ......50 LBYTE RD OK ....... ---- -2 1.37us ---- 29 ....C105 ......38 LBYTE WRI OK ....... ---- -1 11.73us ---- 29 IACK=4 FFFFFF64 LBYTE RD OK ...4.2. IACK4 TRIG 253.49us ---- 29 IACK=2 FFFFFFFF LBYTE RD BERR .....2. IACK2 1 530ns ---- 39 ..06000A ....0021 WORD RD OK .....2. ---- 2 1.07us ---- 39 ..06000C ....0002 WORD RD OK .....2. ---- 3 1.07us ---- 39 ..06000E ....0003 WORD RD OK .....2. ---- 4 18.05us ---- 29 IACK=2 FFFFFF84 LBYTE RD OK .....2. IACK2 5 5.61us ---- 29 ....C101 ......6D LBYTE RD OK ....... ---- 6 1.37us ---- 29 ....C103 ......50 LBYTE RD OK ....... ---- 7 1.37us ---- 29 ....C101 ......38 LBYTE WRI OK ....... ---- As you can see at the TRIG line, the interrupt acknowledge cycle initially returns with an interrupt vector of all F's. This error always occurs when two IRQ lines become active at approximately the same time. Have you seen this? I'm wondering if it's my code, VxWorks, Universe II, or the board that is generating IRQ2 (a Greensprings IP-OctalPlus-485 on VIPC310). I'm using VxWorks BSP for mv2604 v1.1/5. Thanks for any suggestions, Ken N. - -- Ken Norlin --- Simulation Engineer NASA Dryden Flight Research Center ken.norlin@dfrc.nasa.gov (661) 258-2046 FAX (661) 258-2792 --------------------------- Newsgroups: comp.os.vxworks Subject: memPartFree() with an invalid pointer Date: Thu, 16 Mar 2000 15:37:53 -0500 From: Jason Lu Organization: Hitachi Telecom Message-ID: <38D14621.B670E82@hitel.com> Hi all, I am using vxWorks and memory partitions. My question is what would happen if I would try to free the memory, using memPartFree(), within one partition with an invalid pointer (say a pointer to a memory block in another partition). Would this operation just return an ERROR, or would it damage the partition. I remember I read or heard somewhere that some operations may trash the entire memory management system. Thanks a lot in advance. Jason --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Hosting Tornado: Solaris vs. WinNT Date: Thu, 16 Mar 2000 16:33:08 +0100 From: "Borkhuis, Johan" Organization: UUNET-NL (http://www.nl.uu.net) Message-ID: References: <38D00A57.B481DAD7@yahoo.com> > ---------- > From: Marcelo Mourier[SMTP:marcelo_mourier@yahoo.com] > Posted At: woensdag 15 maart 2000 23:12 > Posted To: vxworks > Conversation: Hosting Tornado: Solaris vs. WinNT > Subject: Hosting Tornado: Solaris vs. WinNT > > o Maintenence: the Solaris-based Tornado tools can be installed on an > NFS server where they can be shared by all the developers. What about > the WinNT version? Does each developer need to have a private > installation on his/her PC, like with most IDE's? > That is possible. One of the options during installation is to only install the program groups in your start menu. This will refer to a central installation. Groeten Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan.borkhuis@peektraffic.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: Batch Build in Tornado? Date: Thu, 16 Mar 2000 16:40:13 +0100 From: "Borkhuis, Johan" Organization: UUNET-NL (http://www.nl.uu.net) Message-ID: References: <8ap4bl$dij$1@nnrp1.deja.com> > ---------- > From: liuja@my-deja.com[SMTP:liuja@my-deja.com] > Posted At: woensdag 15 maart 2000 23:54 > Posted To: vxworks > Conversation: Batch Build in Tornado? > Subject: Batch Build in Tornado? > > Is there a way to do batch build in Tornado for all projects within > the same work space? If not, is there a way to derive a top level > workspace-wide Makefile from the .wsp file? > No, that is not possible. What you can do is define an extra project to compile all the other projects. On my home page there is an example on how to do this, this example is derived from an example in the Tornado getting started guide. Another possibility is to write a batch-file that calls "make" with the right parameters in the different directories. Groeten Johan - -- o o o o o o o . . . ___________________________________ o _____ || Johan Borkhuis | .][__n_n_|DD[ ====_____ | johan.borkhuis@peektraffic.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: Great Opportunitie for VxWorks Developers!!!! Date: Wed, 15 Mar 2000 15:35:33 -0500 From: "Charles Maggraf" Organization: EnterAct Corp. Message-ID: <8armcd$1o31$1@news.enteract.com> Lucent Technologies is looking to bring on a number of VxWorks Developers with TCP/IP, SS7, and C++ experience. I have a number of year long contracts available for the Naperville, IL location. If anyone is interested in hearing more about these opportunities please either call or e-mail me. Thanks, Charly Marggraf 800.330.3308 cmarggraf@tsrc.net --------------------------- Newsgroups: comp.os.vxworks Subject: Looking for real time system programmers Date: 17 Mar 2000 05:36:17 GMT From: "Bala P. Kumar" Organization: Prodigy Internet http://www.prodigy.com Message-ID: <01bf8fd2$3fc459c0$fc4dfd3f@default> Greetings! We are looking for real time OS programmers. If you are interested, send your resumes to kpadma@hotmail.com. Best Wishes Bala --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STL and VxWorks? Date: Fri, 17 Mar 2000 08:16:16 GMT From: Computer Engineer Organization: Somewhere in the Earth Message-ID: <38D1791C.39BB46AD@netscape.net> References: <38D0E59E.FEF30E79@americasm01.nt.com> Make sure you are using Tornado 2.0. Tornado 1.0.1 doesn't support STL. "Long, Jeanette [TOR:1V75:EXCH]" wrote: > > Hi > > I am pretty new to VxWorks and I'd like to program my application using > C++, and in particular to make use of the STL. I was wondering if > anyone is using STL in VxWorks and if so, > > 1) What implementation of STL are you using? > 2) Have you encountered any other issues using STL with VxWorks? > > Any other relevant comments welcome too. > Thanks > Jeanette --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Wed Mar 15 10:55:22 2000 From: "Terence Kelleher" Date: Wed Mar 15 10:55:25 PST 2000 Subject: Firmware Development positions available Pathlight Technology has positions available for firmware engineers, for development of high-speed I/O SAN attach systems. Experience with VxWorks and storage attach protocols (SCSI, FC) preferred. Other development and support positions are also available. See www.pathlight.com/employment.html. =============== = Terence M. Kelleher ============= == Principal Engineer - Embedded Software Group =========== === Pathlight Technology, Inc. ========= ==== 9 Brown Road ======= ===== Ithaca, New York 14850 ===== ====== Tel : (607) 266-4000 Ext. 424 === ======= Fax : (607) 266-0352 P A T H L I G H T ========= email: terryk@pathlight.com ========== website http://www.pathlight.com From vxwexplo-errs@csg.lbl.gov Thu Mar 16 04:03:08 2000 From: Vxworks Exploder Date: Thu Mar 16 04:03:11 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Mar 16 04:03:05 PST 2000 Subject: timer_connect Subject: testing Subject: Re: STL and VxWorks? Subject: Fast Ethernet, End Drivers, PowerPC 860T Subject: Tornado Training ? Subject: Re: timer_connect Subject: clock on CPU30 and NVRAM Subject: clock on CPU30 and NVRAM Subject: Re: Hosting Tornado: Solaris vs. WinNT Subject: Trouble compiling TAO for VxWorks simulator Subject: arp cache problem Subject: multiple instances of VxSim Subject: Problem about BSP Subject: ISR or highest priority task Subject: Re: ISR or highest priority task ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: timer_connect Date: Fri, 17 Mar 2000 10:41:38 -0500 From: "Brochu, Patrick" Organization: Bell Northern Research Message-ID: <8atji4$3no$1@bcrkh13.ca.nortel.com> Hi all: Anybody has a working example of using the timer_create/timer_connect functions? I am basically trying to do something as dumb as creating a timer, connecting a function handler to it, setting the interval for the timer and then waiting for my handler function to be called. It seems to do the 'timing' properly, but does a GPF when it should call my handler functioni (i.e., it will take 5 seconds to crash my program if I set my timer to 5 seconds, etc.) . All i'm doing in the handler is increment a global variable. Any ideas or examples? thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: testing Date: Fri, 17 Mar 2000 10:53:15 -0500 From: "Michel Lepine" Message-ID: --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STL and VxWorks? Date: Fri, 17 Mar 2000 16:03:05 +0000 From: Luke Diamand Organization: Virata Message-ID: <38D25739.5F7B0C56@virata.com> References: <38D0E59E.FEF30E79@americasm01.nt.com> I've not used STL on VxWorks, but I've used it on Linux and HP-UX. Beware code bloat! HTH! Luke Diamand "Long, Jeanette [TOR:1V75:EXCH]" wrote: > > Hi > > I am pretty new to VxWorks and I'd like to program my application using > C++, and in particular to make use of the STL. I was wondering if > anyone is using STL in VxWorks and if so, > > 1) What implementation of STL are you using? > 2) Have you encountered any other issues using STL with VxWorks? > > Any other relevant comments welcome too. > Thanks > Jeanette - -- Virata http://www.virata.com Cambridge Tel: +44 1223 566919 Fax: +44 1223 566915 --------------------------- Newsgroups: comp.os.vxworks Subject: Fast Ethernet, End Drivers, PowerPC 860T Date: Fri, 17 Mar 2000 16:01:16 GMT From: rfeller@pgaero.co.uk Organization: Deja.com - Before you buy. Message-ID: <8atks8$lf9$1@nnrp1.deja.com> Reference BSP :- ADS860 - Motorola MPC821/860ADS and MPC860TFADS The Problem..... We are using the MPC860FADS hardware / BSP. We have configured the system to use the on board Fast Ethernet Controller. This hardware uses a external Fast Ethernet Transceiver LXT970. For our final hardware the Fast Ethernet Transceiver shall be a LXT971 device which has a slightly different register set-up, and I believe this is not directly compatible with the previous device. ( Both devices are MII compliant ( Note MII stands for Media Independent Interface ) ) From the BSP it states that the platform complies to the MII specification. (I am not fully aware of the MII specification etc ) Therefore any device that is MII compliant should work correctly, is this correct ??. Therefore I assume that replacing the LTX970 with the LTX971 should work ??. (Assuming the hardware is set-up correctly) I ask because we do not have the source code, only an object module- as follows :- motFecEnd.o - END-style Fast-Ethernet Controller communication library Has anyone written an End driver for this type of device ? I am not familiar with Fast Ethernet or this FEC, your help would very much be appreciated. Riccardo Feller Software Engineer rfeller@pgaero.co.uk Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado Training ? Date: Fri, 17 Mar 2000 11:16:23 -0500 From: "Michel Lepine" Message-ID: Hi, I am a new user with VxWorks and associated tools and I was wondering if anybody have already assist to any tornado training. If so, could somebody tell me a bit about his experience.... Thanks M.Lepine --------------------------- Newsgroups: comp.os.vxworks Subject: Re: timer_connect Date: Fri, 17 Mar 2000 16:13:46 -0000 From: "Will Fookes" Organization: NDS UK Message-ID: <8atjqj$is2$1@ndsukns01.ndsuk.com> References: <8atji4$3no$1@bcrkh13.ca.nortel.com> Patrick, One thing to watch out for (it got me) is that your handler for the timer is executed IN THE CONTEXT OF THE TASK THAT CREATED IT. If the timer was created by some startup task which has since died and gone away, then this will quite likely crash your system the first time that the timer callback function is activated! HTH, Will Brochu, Patrick wrote in message <8atji4$3no$1@bcrkh13.ca.nortel.com>... >Hi all: > >Anybody has a working example of using the timer_create/timer_connect >functions? > >I am basically trying to do something as dumb as creating a timer, >connecting a function handler to it, setting the interval for the timer and >then waiting for my handler function to be called. It seems to do the >'timing' properly, but does a GPF when it should call my handler functioni >(i.e., it will take 5 seconds to crash my program if I set my timer to 5 >seconds, etc.) . All i'm doing in the handler is increment a global >variable. > >Any ideas or examples? > >thanks. > > --------------------------- Newsgroups: comp.os.vxworks Subject: clock on CPU30 and NVRAM Date: Fri, 17 Mar 2000 17:22:51 +0100 From: Gerard Henry Organization: LATP Message-ID: <38D25BDB.42882CFD@cmi.univ-mrs.fr> Hello, we have a problem with clock of our card: CPU30ZBE rev2 this is an old card, and in the beginning of our program, we do test to verify clocking. But when we changed battery, it appears taht our tests failed because clocking is different. We think that the programmer who did the program had modified "something"?? in NVRAM but we don't know what. Anyone knows? Is there documented anywhere? Thanks for your help. - -- Gerard HENRY Ingenieur systeme Laboratoire d'Analyse de Topologie et de Probabilites Centre de Mathématiques et d'Informatique 39, rue F. Joliot Curie 13453 MARSEILLE Cedex 13 Tel.: 04 91 11 35 53 Fax.: 04 91 11 35 52 email: gerard.henry@cmi.univ-mrs.fr --------------------------- Newsgroups: comp.os.vxworks Subject: clock on CPU30 and NVRAM Date: Fri, 17 Mar 2000 17:23:28 +0100 From: Gerard Henry Organization: LATP Message-ID: <38D25C00.F744B33E@cmi.univ-mrs.fr> Hello, we have a problem with clock of our card: CPU30ZBE rev2 this is an old card, and in the beginning of our program, we do test to verify clocking. But when we changed battery, it appears taht our tests failed because clocking is different. We think that the programmer who did the program had modified "something"?? in NVRAM but we don't know what. Anyone knows? Is there documented anywhere? Thanks for your help. - -- Gerard HENRY Ingenieur systeme Laboratoire d'Analyse de Topologie et de Probabilites Centre de Mathématiques et d'Informatique 39, rue F. Joliot Curie 13453 MARSEILLE Cedex 13 Tel.: 04 91 11 35 53 Fax.: 04 91 11 35 52 email: gerard.henry@cmi.univ-mrs.fr --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Hosting Tornado: Solaris vs. WinNT Date: Fri, 17 Mar 2000 16:22:57 +0000 From: Luke Diamand Organization: Virata Message-ID: <38D25BE1.F23212F@virata.com> References: <38D00A57.B481DAD7@yahoo.com> <38D0E26B.1DFEDFA9@lucent.com> We've just gone the other way - we started out with NT (for various reasons) and are now switching to a Sun server. Life is definitely easier on a Sun, especially if, like us, you're used to a UNIX/Linux environment. Main nice points: * single installation on Sun vs multiple installation on NT cuts down on installation time - this is actually quite significant (one to three days per developer) once you factor in setting up source code access, sensible tools and editors, and so-on (Tornado isn't the only tool you need). * Ease of patching - when a patch comes out from WRS, we can apply it one place rather than many (although I've noticed some of the patches struggle to work properly on Solaris). * remote development - I can work at home via a dial-up line when my children are ill. * quicker - our single central Sun Ultrasparc seems much faster (local objects, source code via 100BaseT NFS) than the fastest NT machines we have (source code and objects all local). * vi works properly! Of course all these things can be sorted on NT if you have the time and energy to devote to it. HTH! Luke Diamand Douglas Fraser wrote: > > Marcelo Mourier wrote: > > > > Hi There! > > > > I've used Tornado in several projects and I'm considering using it again > > in my next project. I always used it under Solaris, but now I'm > > considering using it under WinNT. I was wondering if anybody out there > > has gone through this transition and would like to share his/her > > comments (especially the bad ones!) > > > > I'm mostly interested in the following issues: > > > > o Productivity: all our programmers are very used to Unix as their > > software development environment. How did the transition to WinNT affect > > their productivity? > > I took a huge 'ding' going from Unix to NT. Avoid it if you can. > Not just because of weirdness in the NT version of Torndao, just the > whole difference between Unix and Windows. Quite frankly, Windows stinks. > It may be more work to become productive in Unix, but once you achieve > that, Windows forces a _HUGE_ dumbing down. I could not survive at all > without UWIN installed on this beast. Plan on a copy of SOME Unix layer > for everyone moving to NT from Unix. UWIN, or CYGWIN, or MKS,... anything! > > > > > o Robustness: the Solaris-based Tornado tools are mature and robust. > > What about the WinNT tools? > > They suffer some weirdness. Fortunately, Tornado II lets you use your > own editor! (Try GVIM, or better yet, Visual SlickEdit) The project facility > is a joke for anything more than a few files. The target server and > the registry are a pain. The Unix version of Tornado works better for > both. In my prior organization we had Tornado on Unix for the software > developers and Tornado on Windows for a few of the hardware people so that > they could load our drivers. They constantly complained of long connect > times for their targets, yet I could connect to them from my Solaris > workstation in a heartbeat. Now that I am on NT, I understand what they > were going through. > > > > > o Maintenence: the Solaris-based Tornado tools can be installed on an > > NFS server where they can be shared by all the developers. What about > > the WinNT version? Does each developer need to have a private > > installation on his/her PC, like with most IDE's? > > Yes! Tornado on NT buries all sorts of stuff in the registry. Ahhhrrrgghhh! > And don't even get me started about Tornado on Win98.... > > > > > o Pricing: is there any significant price difference between the > > Solaris and WinNT versions? > > Nope! > > Save yourself the grief. If you have experienced Unix developers, keep > that environment and let them do their work. My current organization > does a lot of Windows driver development also, hence the Windows-centric > environment. > > BTW: I really like working with vxWorks. I do. I just miss my Unix host. :-( > > Doug Fraser > > > > > Thanks a bunch, > > > > Marcelo. - -- Virata http://www.virata.com Cambridge Tel: +44 1223 566919 Fax: +44 1223 566915 --------------------------- Newsgroups: comp.soft-sys.ace,comp.os.vxworks Subject: Trouble compiling TAO for VxWorks simulator Date: Fri, 17 Mar 2000 12:54:35 -0600 From: Lee Graba Organization: Honeywell Technology Center Message-ID: <38D27F6A.4EA7E855@htc.honeywell.com> I am trying to compile TAO (from Washington University), targeting the VxWorks simulator under Windows NT. I was able to compile and build libACE.a fine, but TAO fails on the very first file with the following messages: bash-2.02$ make ccsimpc -fno-implicit-templates -U__WINNT -UWIN32 -U__WINNT__ -UWINNT - -U__MINGW32__ -U_WIN32 -U__WIN 32 -U__WIN32__ -U__WIN32 - -B/Tornado/host/x86-win32/lib/gcc-lib/i386-pc-mingw32/egcs-2.90.29 - -DCPU=SI MNT -DVXWORKS=1 -D_REENTRANT -ansi -fno-builtin -fno-defer-pop - -fvolatile -nostdinc -nostdlib -W -W all -DACE_VXWORKS=0x540 -Wno-uninitialized -O -g -fno-rtti - -DACE_LACKS_RTTI -fno-exceptions -fchec k-new -I/Tornado/target/h - -I/d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full -DTAO_HAS_CORBA_MESSAGING - -I/d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full /tao -DTAO_PLATFORM_SVC_CONF_FILE_NOTSUP -c -o .obj/Object_Adapter.o Object_Adapter.cpp (C:\Tornado\host\x86-win32\lib\gcc-lib\i386-pc-mingw32\egcs-2.90.29\cc1plus.exe 1002) commit_and_inc : VirtualAlloc failed /d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full/tao/tao/POAManager.i: In met hod `void TAO_POA_Manager::activate(class CORBA_Environment & = TAO_default_environment()())': In file included from /d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full/tao/ta o/POAManager.h:103, from /d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full/tao/ta o/POA.h:46, from Object_Adapter.cpp:4: /d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full/tao/tao/POAManager.i:18: vir tual memory exhausted ccsimpc: file path prefix `/Tornado/host/x86-win32/lib/gcc-lib/i386-pc-mingw32/egcs-2.90.29' never u sed make: *** [.obj/Object_Adapter.o] Error 1 bash-2.02$ Environment variables of note are: export WIND_BASE=/Tornado export WIND_HOST=x86-win32 export WIND_HOST_TYPE=x86-win32 export ACE_ROOT=/d/Projects/ASTD/COFP/Infrastructure/PlatformServices/ACE_wrappers_Full export TAO_ROOT=$ACE_ROOT/tao export GCC_EXEC_PREFIX=$WIND_BASE/host/$WIND_HOST_TYPE/lib/gcc-lib/ #export VXWORKS_COMP=powerpc-wrs-vxworks #export COMP_VERSION=cygnus-2.7.2-960126 export VXWORKS_COMP=i386-pc-mingw32 export COMP_VERSION=egcs-2.90.29 export CPU=SIMNT export CYG_PATH=/cygnus/cygwin-b20 PATH=$WIND_BASE/host/$WIND_HOST_TYPE/bin PATH=$PATH:$WIND_BASE/host/$WIND_HOST_TYPE/lib/gcc-lib/$VXWORKS_COMP/cygnus-2.7.2-960126 PATH=$PATH:$WIND_BASE/host/$WIND_HOST_TYPE/$VXWORKS_COMP/bin PATH=$PATH:/d/Projects/ASTD/ACE_wrappers.NT/bin PATH=$PATH:$CYG_PATH/H-i586-cygwin32/bin:$CYG_PATH/tcl/bin PATH=$PATH:/WINNT/system32:/WINNT:/WINNT/system32/nls/ENGLISH PATH=$PATH:/Perl/bin export PATH One of the messages is about running out of virtual memory. This would surprise me if true, since I have 96 M of RAM, and a total of up to 200-400 M available for swap. Is this really a virtual memory problem. Another question: I am using the i386-pc-mingw32 compiler supplied with VxWorks Tornado in order to target the simulator. Is this correct? - -- - -------------------------------------------------------------------------------- Lee Graba Honeywell Technology Center graba_lee@htc.honeywell.com - -------------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: arp cache problem Date: Fri, 17 Mar 2000 19:03:31 GMT From: dhanamjay@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8atvhv$tv9$1@nnrp1.deja.com> problem is vxworks arp cache fills with around 63-70 entries and it does not respond to any more new arp entries. it seems to me that arp allocates dynamically for its entries I did not understand the cause for this problem does anybody know this problem??? that will be great help for me Regards Dhanamjay Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: multiple instances of VxSim Date: Fri, 17 Mar 2000 15:32:57 -0500 From: "Chen, Changbiao [WDLN2:2O41:EXCH]" Organization: Nortel Networks Message-ID: <38D29679.825B1D1A@americasm01.nt.com> Hi,folks: I am tring to run VxSim with multiple instances. The Back end I used is WDBpipe. When I invoke second instance of VxSim, I was informed that the pipe has already been used. So, I failed when I tried to connect to that VxSim. Is there anybody who have experiences with this? Thanks. James --------------------------- Newsgroups: comp.os.vxworks Subject: Problem about BSP Date: Fri, 17 Mar 2000 15:37:11 -0500 From: "Chen, Changbiao [WDLN2:2O41:EXCH]" Organization: Nortel Networks Message-ID: <38D29777.52301445@americasm01.nt.com> Hi,folks: I just got Force computers BSP run on the target board of Force computers. I happenly found that I cannot download VxWorks from the UNIX server to the target board smoothly. In order to be connected, I have to do the @ command twice without interval. Could you please give me some idea to deal with such a kind of thing? Regards. James --------------------------- Newsgroups: comp.os.vxworks Subject: ISR or highest priority task Date: Fri, 17 Mar 2000 15:40:45 -0500 From: "Chen, Changbiao [WDLN2:2O41:EXCH]" Organization: Nortel Networks Message-ID: <38D2984D.EF9142D8@americasm01.nt.com> Hi,there: To my understanding, the best way to achieve deterministic is to use ISR. But someone said the highest priority task is in fact the same thing as ISR. I am not sure. Could someone explain the difference between ISR and highest priority task. Cheers. James --------------------------- Newsgroups: comp.os.vxworks Subject: Re: ISR or highest priority task Date: Fri, 17 Mar 2000 21:01:09 -0800 From: Harvey Taylor Organization: Organization? What organization?! Message-ID: <38D30D95.FBD@despam.pangea.ca> References: <38D2984D.EF9142D8@americasm01.nt.com> Chen, Changbiao [WDLN2:2O41:EXCH] wrote: >[...] > Could someone explain the difference between ISR and highest priority > task. > The question is off slightly. ISR are not tasks; most notably because there are some things you cannot do in an ISR (any VxWorks library call which may block). Tasks may have priority 0 to 255. ISRs run at priority -1, so to speak. ie. they run in interrupt context. - -het - -- "If the code and the comments disagree, then both are probably wrong." -- Norm Schryer Harvey Taylor mailto:het@despam.pangea.ca http://www.pangea.ca/~het --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Fri Mar 17 04:03:05 2000 From: Vxworks Exploder Date: Fri Mar 17 04:03:07 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Mar 17 04:03:02 PST 2000 Subject: Re: Fragmented memory after initialization on MBX860 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Fragmented memory after initialization on MBX860 Date: Sun, 19 Mar 2000 05:39:08 GMT From: john_94501@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8b1p5r$uk2$1@nnrp1.deja.com> References: <38D40B9B.B2DA3683@ObjecTime.com> This sounds like it could be due to MMU page tables. Do you have a large area of the PCI address space mapped (i.e. described in your sysPhysMemDesc structure)? If you do, and you don't need it all, I would recommend shrinking it to match your requirements. John... In article <38D40B9B.B2DA3683@ObjecTime.com>, Emeka Nwafor wrote: > Has anyone experienced fragmented memory on VxWorks kernels built for > the MBX860 boards? After initialization, I have 255 4088 byte blocks in > the my heap's free list. I'm experiencing this problem with VxWorks > 5.3.1 and VxWorks 5.4. > > - Emeka. > > Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- End of New-News digest ********************** From vxwexplo-errs@csg.lbl.gov Mon Mar 20 10:05:25 2000 From: "Marco Penso" Date: Mon Mar 20 10:05:28 PST 2000 Subject: Setting parity bit to even This is a multi-part message in MIME format. ------=_NextPart_000_0063_01BF9295.79A67E00 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi everybody, I'm trying to receive and transmit data via serial line. The default for = VxWorks is no parity, 8 data bit, 1 stop bit. Can someone tell me where (and how) must I specify parameters to modify = parity to EVEN. Thank you! ------=_NextPart_000_0063_01BF9295.79A67E00 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi everybody,
 
I'm trying to receive and transmit data via serial = line. The=20 default for VxWorks is no parity, 8 data bit, 1 stop = bit.
 
Can someone tell me where (and how) must I specify = parameters=20 to modify parity to EVEN.
 
Thank you!
------=_NextPart_000_0063_01BF9295.79A67E00-- From vxwexplo-errs@csg.lbl.gov Mon Mar 20 12:58:48 2000 From: "Mike Anderson" Date: Mon Mar 20 12:58:50 PST 2000 Subject: RE: Setting parity bit to even VxWorks Greetings! > Submitted-by: "Marco Penso" > > > I'm trying to receive and transmit data via serial line. The default for = > VxWorks is no parity, 8 data bit, 1 stop bit. > > Can someone tell me where (and how) must I specify parameters to modify = > parity to EVEN. > Marco, Typically, that would be done in the "ioctl" section of your serial driver. There is no standard way that WRS implements this feature because it is typically not implemented at all in the shipped drivers. What you will have to do is to add an ioctl call to your serial driver and using the data sheets for the U(S)ART in question, set it up to support setting the parity. For many of the typical U(S)ARTS, someone on the net will generally have the code for setting the parity already put together. They may even be willing to share that with you :-). But, you'll have to supply more specific information about which serial device you're trying to support and which version of VxWorks/Tornado you're using. HTH, Mike ----------------------------------------------------------------- Michael E. Anderson | Integrated Chipware Chief Scientist | 1861 Wiehle Ave. #300 | Reston, VA 20190 mike@chipware.com | (703) 736-3504 www.chipware.com (888)430-CHIP | (703) 736-3556 FAX ----------------------------------------------------------------- "Software development is like making a baby. You can't make a baby in one month by impregnating nine women. Some things just take time..." From vxwexplo-errs@csg.lbl.gov Tue Mar 21 00:58:12 2000 From: "Marco Penso" Date: Tue Mar 21 00:58:14 PST 2000 Subject: I: Setting parity bit to even Hi everybody, in the previous message i forgot some informations about my system. I'm trying to receive and transmit data via serial line. The default for VxWorks is no parity, 8 data bit, 1 stop bit. Can someone tell me where (and how) must I specify parameters to modify parity to EVEN. The target on which I use VxWorks 5.4 is a PC Pentium with a UART 16550 (it works correctly even with the default driver i8250). Thank you! From vxwexplo-errs@csg.lbl.gov Tue Mar 21 01:24:00 2000 From: David Laight Date: Tue Mar 21 01:24:02 PST 2000 Subject: RE: Setting parity bit to even > > Typically, that would be done in the "ioctl" section of your serial > driver. There is no standard way that WRS implements this feature > because it is typically not implemented at all in the shipped drivers. > What you will have to do is to add an ioctl call to your serial driver > and using the data sheets for the U(S)ART in question, set it up to > support setting the parity. For many of the typical U(S)ARTS, > someone on the net will generally have the code for setting the parity > already put together. They may even be willing to share that with > you :-). But, you'll have to supply more specific information about > which serial device you're trying to support and which version of > VxWorks/Tornado you're using. I took a slightly different approach to setting serial line options, allowing any program to specify the required serial line options to open - without needing any ioctl (or termios) calls. I make use of the VxWorks 'feature' that the unmatched part of the users filename is passed to the device drive open routine. My serial driver parses the filename passed to it looking for any of: / baud rate nnn /p[oe10] parity odd, even, force 1, force 0 (default none) /-dtr do not assert DTR /-rts do not assert RTS /-flow do not do hardware (CTS) tx flow control /-sig do not take mode signal interupts /nmea GPS nmea messages /rwib GPS Rockwell binary messages eg to read GPS data one might open /serial_0/-flow/-sig/4800/nmea 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 Tue Mar 21 04:03:13 2000 From: Vxworks Exploder Date: Tue Mar 21 04:03:15 PST 2000 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Mar 21 04:03:09 PST 2000 Subject: Modem Support Subject: explicit template instantiation Subject: Fragmented memory after initialization on MBX860 Subject: Write my own Mblk to buffer copy routine in END Subject: Re: STL and VxWorks? Subject: Tornado 1.0.1 BSP compatibility problem in Tornado 2.0 Subject: Re: Tornado 1.0.1 BSP compatibility problem in Tornado 2.0 Subject: vxworks & add one serialport Subject: Re: STL and VxWorks? Subject: Re: Hosting Tornado: Solaris vs. WinNT Subject: Re: Help on VxWorks newsgroup. Subject: Linking code to image Subject: Re: Help on VxWorks newsgroup. Subject: RAID support on vxworks Subject: Error 3d0004 (reprise for me) Subject: Re: Trouble compiling TAO for VxWorks simulator Subject: Re: Hosting Tornado: Solaris vs. WinNT Subject: Re: Trouble compiling TAO for VxWorks simulator Subject: Re: volatile? in vxWorks Subject: C++ exceptions & GCC 2.95.2 Subject: Reset signal Subject: Re: Reset signal Subject: Re: Linking code to image ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Modem Support Date: Sun, 19 Mar 2000 12:53:44 GMT From: nlm1@my-deja.com Organization: Deja.com - Before you buy. Message-ID: <8b2ikl$f22$1@nnrp1.deja.com> Does anyone know where I can get a modem support package or application that will run on vxWorks? I need it to do call answer and if possible to do call out. Thanks Norm Magnan Sent via Deja.com http://www.deja.com/ Before you buy. --------------------------- Newsgroups: comp.os.vxworks Subject: explicit template instantiation Date: Sat, 18 Mar 2000 16:30:19 -0800 From: "Kevin Rea" Message-ID: <38d41d97$0$20821@nntp1.ba.best.com> We are using the STL for our embedded development. We have ported the RogueWave STL implementation and also used the VxWorks STL support in Tornado 2.0. Controlling template instantiation to control code bloat is an issue. The -frepo flag for the GNU compiler that WRS tells you to use and is supposed to automatically sort this out flat out does not work (WRS has not been very responsive on this). We have been successful using the -fno-implicit-templates flag with the RogueWave STL and creating a library to explicitly instantiate the templates we are using. We have not been able to make this work for the VxWorks STL. The problem is that the VxWorks STL implementation seems to be considerably more space efficient than the RW STL, so we would prefer to use it if possible. Anybody had any experience getting explicit template instantiation to work with the VxWorks STL implementation? Kevin Rea krea@alidian.com --------------------------- Newsgroups: comp.os.vxworks Subject: Fragmented memory after initialization on MBX860 Date: Sat, 18 Mar 2000 18:04:59 -0500 From: Emeka Nwafor Organization: UUNET Canada News Reader Service Message-ID: <38D40B9B.B2DA3683@ObjecTime.com> Has anyone experienced fragmented memory on VxWorks kernels built for the MBX860 boards? After initialization, I have 255 4088 byte blocks in the my heap's free list. I'm experiencing this problem with VxWorks 5.3.1 and VxWorks 5.4. - - Emeka. --------------------------- Newsgroups: comp.os.vxworks Subject: Write my own Mblk to buffer copy routine in END Date: Mon, 20 Mar 2000 12:17:24 +0800 From: "Cyclone" Organization: Alcatel Message-ID: <38d68892.0@dnews.sbell.com.cn> Hi, all Because my device cannot access network buffer pool directly, I have to write its own copy routine. Manual of netMblkToBufCopy( ) doesn't describe this very clear. Does anyone know how to do it? Thanks, Wang Jian --------------------------- Newsgroups: comp.os.vxworks Subject: Re: STL and VxWorks? Date: Sun, 19 Mar 2000 00:26:12 +0100 From: "Mario Semo" Organization: Nextra Telekom GmbH Message-ID: <8b4hpc$eqg$1@newsmaster01.nextra.at> References: <38D0E59E.FEF30E79@americasm01.nt.com> we use STL as shipped with Tornado 2.0. this is a VERY old version of SGL-STL. but we also use a 3.01 SGL-STL for other development (we compile our coede not just for VxWorks, but for other HW platoforms too). and the STL shipped with tornado is (as far as we used it) compatible to SGL-STL 301. i have no idea which was exactly the tornaod SGL-STL is. but when you go to SGL-STL homepage, you can, as far as i remember, download STLs for VxWorks. http://www.stlport.org - -- regards, Mario Semo. http://www.kirchnersoft.com Long, Jeanette [TOR:1V75:EXCH] wrote in message <38D0E59E.FEF30E79@americasm01.nt.com>... >Hi > >I am pretty new to VxWorks and I'd like to program my application using >C++, and in particular to make use of the STL. I was wondering if >anyone is using STL in VxWorks and if so, > >1) What implementation of STL are you using? >2) Have you encountered any other issues using STL with VxWorks? > >Any other relevant comments welcome too. >Thanks >Jeanette --------------------------- Newsgroups: comp.os.vxworks Subject: Tornado 1.0.1 BSP compatibility problem in Tornado 2.0 Date: Sun, 19 Mar 2000 13:58:49 +0800 From: "TonyHuang" Organization: BST Communication Technology Ltd. Message-ID: <8b1q7v$v60$1@news.cz.js.cn> We are developing our BSP on the MPC860 based on a third party vendor's BSP(ver1.1 from Plx Technology CPCI9054-860 board ), everything is OK when we use "standard BSP build" option in Tornado2.0 environment to build our BSP(vxworks_rom.hex image), because it's "Tornado1.0.1 compatible". After we create a new project based on this ver1.1 BSP, or we create a default project according to "Tornado BSP Developer's Kit for VxWorks User's Guide"Appendix A-"Upgrading a BSP for Tornado2.0", we got problem in booting vxworks_rom.hex image. we trace to RAM_LOW_ADRS, and found that it's not any opcode for sysinit(), but the vxWorks image starts with "-ELF---------------". we know that the Makefile for projects or v1.2 BSP is quite different from the Makefile for ver1.1 BSP. IN T1: the defs.bsp file has the follow definitions: LD_LOW_FLAGS = -Ttext $(RAM_LOW_ADRS) LD_HIGH_FLAGS = -Ttext $(RAM_HIGH_ADRS) RES_LOW_FLAGS = -Ttext $(ROM_TEXT_ADRS) -Tdata $(RAM_LOW_ADRS) RES_HIGH_FLAGS = -Ttext $(ROM_TEXT_ADRS) -Tdata $(RAM_HIGH_ADRS) ROM_ENTRY = _romInit SYS_ENTRY = _sysInit USR_ENTRY = _usrInit LDFLAGS = -X -N LD_PARTIAL_FLAGS = -X -r and the rules.bsp file has the rules: vxWorks_rom : depend.$(BSP_NAME) bootInit_uncmp.o romInit.o \ dataSegPad.o usrConfig.o $(MACH_DEP) $(LDDEPS) $(LIBS) - @ $(RM) $@ - @ $(RM) $(IMI) - @ $(RM) $(IBR) - @ $(RM) version.o - @ $(RM) ctmp.o ctdt.c ctdt.o $(CC) -c $(CFLAGS) -o version.o $(CONFIG_ALL)/version.c $(LD_PARTIAL) \ -o ctmp.o usrConfig.o \ $(MACH_DEP) version.o $(LIBS) $(NM) ctmp.o | $(MUNCH) > ctdt.c $(MAKE) CC_COMPILER="-fdollars-in-identifiers" ctdt.o $(LD) $(LDFLAGS) -e $(ROM_ENTRY) $(LD_LOW_FLAGS) \ -o $@ romInit.o bootInit_uncmp.o dataSegPad.o \ ctmp.o ctdt.o - @ $(RM) ctmp.o $(LDOUT_HOST) $@ $(ROMSIZEPROG) -b $(ROM_SIZE) $@ $(LDOUT_CONV) $@ vxWorks_rom.hex : depend.$(BSP_NAME) vxWorks_rom $(IMI) - @ $(RM) $@ $(BINHEX) $(HEX_FLAGS) $(SECT_SPEC)vxWorks_rom$(TXT_OFFSET) \ $(IMI_SPEC)$(IMI)$(IMI_OFFSET) > $@$(MAP_EXT) -@ $(MV_BIN) IN T2: the defs.project and defs.bsp file has the follow definitions: ROM_ENTRY = _romInit SYS_ENTRY = _sysInit USR_ENTRY = _usrEntry LDFLAGS = -X -N LD_PARTIAL_FLAGS = -X -r LD_RAM_FLAGS = -Ttext $(RAM_LOW_ADRS) LD_ROM_RES_FLAGS= $(ROM_LDFLAGS) -Ttext $(ROM_TEXT_ADRS) \ -Tdata $(RAM_LOW_ADRS) CC_ROM_RES_FLAGS= -DROM_RESIDENT -DRAM_DATA_ADRS=0x$(RAM_LOW_ADRS) \ -DRAM_DST_ADRS=0x$(RAM_LOW_ADRS) LD_ROM_CPY_FLAGS= $(ROM_LDFLAGS) -Ttext $(ROM_TEXT_ADRS) \ -Tdata $(RAM_HIGH_ADRS) CC_ROM_CPY_FLAGS= -DROM_COPY -DRAM_DATA_ADRS=0x$(RAM_HIGH_ADRS) \ -DRAM_DST_ADRS=0x$(RAM_LOW_ADRS) LD_ROM_CMP_FLAGS= $(ROM_LDFLAGS) -Ttext $(ROM_TEXT_ADRS) \ -Tdata $(RAM_HIGH_ADRS) CC_ROM_CMP_FLAGS= -DROM_COMPRESS -DRAM_DATA_ADRS=0x$(RAM_HIGH_ADRS) \ -DRAM_DST_ADRS=0x$(RAM_LOW_ADRS) and the rules.bsp file has the rules: vxWorks: partialImage.o $(POSTLINK_MODULES) - @ $(RM) $@ $(CC) $(CFLAGS) -c $(DATA_SEG_PAD_C) $(LD) $(LDFLAGS) -e $(SYS_ENTRY) $(LD_RAM_FLAGS) \ dataSegPad.o partialImage.o $(POSTLINK_MODULES) -o $@ $(LDOUT_CONV) $@ $(VXSIZEPROG) -v $(RAM_HIGH_ADRS) $(RAM_LOW_ADRS) $@ $(MAKE_SYM) $(MAKE_SYM_CVT) $(POST_BUILD_RULE) vxWorks_rom: vxWorks $(BOOT_OBJS) $(COMPONENT_LIBS) $(LIBS) $(IMI) - @ $(RM) $@ $<.o tmp.Z $<.Z.s $<.Z.o $(CP) $< $<.o $(LDOUT_HOST) $<.o $(EXTRACT_BIN) < $<.o > tmp.Z $(BINTOASM) tmp.Z > $<.Z.s $(CC) -c $(CFLAGS) $<.Z.s -o $<.Z.o $(CC) $(CFLAGS) -c $(VERSION_C) $(LD) $(LDFLAGS) -e $(ROM_ENTRY) $(LD_ROM_CPY_FLAGS) \ $(BOOT_OBJS) version.o $(COMPONENT_LIBS) $(LIBS) $<.Z.o -o $@ - @ $(RM) $<.o tmp.Z $<.Z.s $<.Z.o $(ROMSIZEPROG) -b $(ROM_SIZE) $@ $(LDOUT_CONV) $@ $(BINHEX) $(HEX_FLAGS) $(SECT_SPEC) $@ $(TXT_OFFSET) \ $(IMI_SPEC)$(IMI)$(IMI_OFFSET) > $@.hex $(MAP_EXT) $(POST_BUILD_RULE) Here is the actual operation of the make in T1 and T2: IN T1: ldppc -X -N -e _romInit -Ttext 00100100\ -o vxWorks_rom romInit.o bootInit_uncmp.o dataSegPad.o\ ctmp.o ctdt.o IN T2: ldppc -X -N -e _sysInit -Ttext 00100100\ dataSegPad.o partialImage.o ctdt.o -o vxWorks ldppc -X -N -e _romInit -Ttext FFF00100 -Tdata 00800100\ romInit.o romStart.o version.o C:\TORNADO\target\lib\libPPC860gnuvx.a vxWorks.Z.o -o vxWorks_rom How can we port the BSP from ver1.1 to ver1.2 ? any suggestion is appreciate. Our specific details are as follows: Toolset: Tornado 2.0 RAM: 16 Mbytes SDRAM ROM: 512 kbyte PLCC socketed flash ROM_BASE_ADRS=0xFFF00000 ROM_TEXT_A