From mdcb@ing.iac.es Wed Jul 1 04:01:33 1998 From: Matthieu Bec Date: Wed Jul 1 04:01:37 PDT 1998 Subject: syslog'in with VxWorks VxWorks Hi, We're looking for some sort of equivalent of the syslog call under vxWorks. Has anyone already tried to implement such a scheme ? Our idea would be to modify the logLib so that it talks to syslogd on UNIX rather than sending its output to the console or a file. Any comments, suggestions, experience or wrapped-up package warmly welcome. Best regards, Matthieu -- Matthieu Bec Isaac Newton Group tel : +34 922 425427 Apartado de Correos 321 fax : +34 922 425401 38780 Santa Cruz de La Palma email: mdcb@ing.iac.es Canary Islands, Spain From prewitt@ticipa.Works.ti.com Wed Jul 1 08:58:41 1998 From: "James O. Prewitt" Date: Wed Jul 1 08:58:44 PDT 1998 Subject: How do you create VxWorks bootrom EPROMS for a MVME167? Date: 1 Jul 98 To: VxWorks Users' Group From: Jim Prewitt Subj: How do you create VxWorks bootrom EPROMS for a MVME167? I have created a bootrom code for a MVME167 board. However, this particular board has two 27C202 eproms. I understand that each is 16 bits wide. Does anyone know the procedure or utilities required to "split" the bootrom file into two (?) files that are suitable for preparing the two eproms from? Thanks for the help, Jim prewitt@works.ti.com 972-927-5771 From mikem@sdlabs.com Wed Jul 1 09:26:31 1998 From: Michael Morrison Date: Wed Jul 1 09:26:35 PDT 1998 Subject: Re: syslog'in with VxWorks the vxWorks Users Group Exploder wrote: > Submitted-by mdcb@ing.iac.es Wed Jul 1 04:01:33 1998 > Submitted-by: Matthieu Bec > > VxWorks > > Hi, > > We're looking for some sort of equivalent of the syslog call under > vxWorks. > Has anyone already tried to implement such a scheme ? > > Our idea would be to modify the logLib so that it talks to syslogd on > UNIX rather than sending its output to the console or a file. > > Any comments, suggestions, experience or wrapped-up package warmly > welcome. Open a UDP socket (port 514) to the syslogd running on your unix host and pass the socket's file descriptor to logInit() - no need to 'modify' logLib. From kkauper@draper.com Wed Jul 1 13:45:59 1998 From: Kris Kauper Date: Wed Jul 1 13:46:03 PDT 1998 Subject: MULTI tools tornado Has anyone had experience with Green Hills Software's "MULTI" development tools for VxWorks? I'm evaluating them for an upcoming project and would appreciate any feedback. Thanks. Kris Kauper C.S. Draper Laboratory 555 Technology Square, M/S 18 Cambridge, MA 02139 kkauper@draper.com 617-258-1590 617-258-3858 (fax) From chip@Cerent.com Wed Jul 1 15:30:48 1998 From: Chip Roberson Date: Wed Jul 1 15:30:51 PDT 1998 Subject: calendar time in VxWorks Hello, I was wondering what poeple are doing for maintaining calendar time in their systems? We are using the clock_settime() and clock_gettime() functions along with the TIMEZONE environment variable. The clock appears to generally work but it has a bug at the end of daylight savings time. It doesn't automatically fall back at 2am. We're continuing to test and they are other issues which we haven't isolated them but this problem already seems signficant. What are our options? We've opened a TSR with WRS. Any comments? Thanks, Chip -- Charles S. Roberson This message may contain two attachments: My Business vCard and My S/MIME Digital ID From rtp.co.uk!ihw@rtp.co.uk Thu Jul 2 00:45:58 1998 From: Ian Willats Date: Thu Jul 2 00:46:03 PDT 1998 Subject: How do you create VxWorks bootrom EPROMS for a MVME167? At RTP we use a general-purpose S-record mangling tool called "ssplit" which I wrote a while ago. Among other things, it can be made to split a single .hex file into two .hex files containing the high and low halves of 32-bit words in the original file, which is what you need for the 167. Incidentally it can also split one .hex file into two files which contain alternate 32-bit words, which is what you need for the flash on boards like the mv260x (IIRC). Let me know if you'd like me to send you a copy of the source. HTH Ian --------------------------------------------------------------- Ian Willats e-mail: mailto:ihw@rtp.co.uk Real-Time Products Ltd. direct: +44 (0) 121 234 6637 Chancery House, tel: +44 (0) 121 234 6600 8 Edward Street, Birmingham. fax: +44 (0) 121 234 6611 B1 2RX. England. web: http://www.rtp.co.uk --------------------------------------------------------------- From daemon@csg.lbl.gov Thu Jul 2 04:01:18 1998 From: daemon@csg.lbl.gov Date: Thu Jul 2 04:01:22 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 2 04:01:14 PDT 1998 Subject: dosFs with long file names Subject: i960 Flash File Systems ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: dosFs with long file names Date: Wed, 01 Jul 1998 16:22:01 GMT From: gmathew@my-dejanews.com Organization: Deja News - The Leader in Internet Discussion Message-ID: <6ndnn8$g2p$1@nnrp1.dejanews.com> I want to use dosFs with long file name support enabled (x86 PC target). Once this is done, the disk is no longer readable by DOS. So I can't use standard DOS tools like chkdisk any more. Are there any tools available for VxWorks that can be used to check the integrity of a dosFs disk? Thanks for any info. George Mathew gmathew@my-dejanews.com - -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum --------------------------- Newsgroups: comp.os.vxworks Subject: i960 Flash File Systems Date: Thu, 2 Jul 1998 09:56:41 +0100 From: "Mark Hambleton" Organization: EPL Ltd Message-ID: <6nfhsk$2t9$1@taliesin.netcom.net.uk> Does anybody know of a good (cheap) flash file system for vxWorks and the i960? --------------------------- End of New-News digest ********************** From ELowe@SYSTRAN.com Thu Jul 2 05:11:02 1998 From: Eric Lowe Date: Thu Jul 2 05:11:05 PDT 1998 Subject: HTTP server for VxWorks? This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------ =_NextPart_001_01BDA5B2.87E5C3F0 Content-Type: text/plain Does anyone know of an HTTP server for VxWorks sockets? For our uses, it only needs to be HTTP/1.0 compliant. Eric Lowe FibreXpress Software Engineer Co-op Systran Corporation elowe@systran.com ------ =_NextPart_001_01BDA5B2.87E5C3F0 Content-Type: text/html Content-Transfer-Encoding: quoted-printable HTTP server for VxWorks?

Does anyone know of an HTTP server for = VxWorks sockets?  For our uses, it only needs to be HTTP/1.0 = compliant.


Eric Lowe
FibreXpress Software Engineer = Co-op
Systran Corporation
elowe@systran.com

------ =_NextPart_001_01BDA5B2.87E5C3F0-- From jhend@primenet.com Thu Jul 2 17:23:05 1998 From: JRH Date: Thu Jul 2 17:23:09 PDT 1998 Subject: Multiple RTOS ethernet download attempts We have a problem were we get the error "d0003" after attempting to download VxWorks over ethernet. After three to ten times of pressing the reset buton the RTOS finally gets read over the ethernet and all is well. I believe "d0003" is equivalent to an "invalid file handle" error. We are using an MV2603 board, an NT 4.0 host, FTP, no gateway is used or specified, and the VxWorks 5.3 RTOS. We don't have any tools to sniff the ethernet to look at the TCP/IP traffic. Any ideas? Thank You, jhend@primenet From ptp@mclean.sparta.com Thu Jul 2 20:08:04 1998 From: Patrick Pinkowski Date: Thu Jul 2 20:08:07 PDT 1998 Subject: VxWorks PCMCIA ATA Device on Mbx860 I have a Motorola MBX 860 board and BSP. I want to get a PCMCIA ATA Flash Disk and PCMCIA ATA Hard Disk working on the MBX 860 under VxWorks. The BSP supports PCMCIA ram cards, but nothing else. Has anyone extended the PCMCIA support in the BSP? Any PCMCIA ATA Disk Support? Thanks in advance. Regards Patrick T. Pinkowski ptp@mclean.sparta.com From ptp@mclean.sparta.com Thu Jul 2 20:33:29 1998 From: Patrick Pinkowski Date: Thu Jul 2 20:33:32 PDT 1998 Subject: VxWorks MBX-860 Floppy Support Has anyone gotten the floppy working on the MBX-860 under VxWorks? Thanks in Advance Patrick Pinkowski ptp@mclean.sparta.com From patrick.keliher@wrs.com Fri Jul 3 17:14:52 1998 From: Patrick Keliher Date: Fri Jul 3 17:14:55 PDT 1998 Subject: Re: VxWorks PCMCIA ATA Device on Mbx860 Patrick, Talk to your local WRS FAE. A special projects consulting group inside Wind River did exactly this project. Let me know if you don't know your local guys name. Hope this helps, Pat Keliher At 08:08 PM 7/2/98 PDT, you wrote: >Submitted-by ptp@mclean.sparta.com Thu Jul 2 20:08:04 1998 >Submitted-by: Patrick Pinkowski > >I have a Motorola MBX 860 board and BSP. I want to get a PCMCIA ATA >Flash Disk and PCMCIA ATA Hard Disk working on the MBX 860 under >VxWorks. The BSP supports PCMCIA ram cards, but nothing else. Has >anyone extended the PCMCIA support in the BSP? Any PCMCIA ATA Disk >Support? > >Thanks in advance. > >Regards > >Patrick T. Pinkowski > >ptp@mclean.sparta.com Patrick Keliher, PE MAILTO:patrick.keliher@wrs.com Senior Field Application Engineer Wind River Systems http://www.wrs.com Phone: (972)776-3555 4201 Spring Valley Road, Suite 1400 Fax: (972)776-3556 Dallas, TX 75244 Pager: (800)409-8107 Pager: MAILTO:4098107@skytel.com From daemon@csg.lbl.gov Sat Jul 4 04:03:21 1998 From: daemon@csg.lbl.gov Date: Sat Jul 4 04:03:24 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 4 04:03:18 PDT 1998 Subject: Re: Cryptic Errors ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Cryptic Errors Date: 3 Jul 98 19:11:35 GMT From: georg@sgl.crestech.ca (Georg Feil) Organization: Centre for Research in Earth and Space Technology Message-ID: References: Michelle Ernst writes: >Access Fault >Program Counter: 0x01e78090 >Status Register: 0x3000 >Access Address : 0x01e78090 >Special Status : 0x01e6 Get the MC68040 User's Manual from Motorola. This describes architectural issues such as status register layout, special exception status, and exception types (Access Fault being one of them). Good luck, Georg Feil | http://www.sgl.crestech.ca/ Space Geodynamics Laboratory | Email: georg@sgl.crestech.ca CRESTech | Phone: (416) 665-5458 4850 Keele St./North York/Ont/Canada/M3J 3K1 | Fax: (416) 665-1815 - -- Georg Feil | http://www.sgl.crestech.ca/ Space Geodynamics Laboratory | Email: georg@sgl.crestech.ca CRESTech | Phone: (416) 665-5458 4850 Keele St./North York/Ont/Canada/M3J 3K1 | Fax: (416) 665-1815 --------------------------- End of New-News digest ********************** From Igal@tadirantele.com Sat Jul 4 23:36:06 1998 From: Chernobelsky Igal Date: Sat Jul 4 23:36:28 PDT 1998 Subject: i960 Flash File System Hello VxWorks/Tornado users, Mark Hambleton asked: >Does anybody know of a good (cheap) flash file system for vxWorks and the >i960? Yes I do. I am using Flash File System called FlashLIb from RST (now Wind River System Israel) for VxWorks/i960 for performance monitoring purpose in Wireless Multigain System. Our system hardware includes i960JA adn 2 Intel 28F016 Flashes that provide us 4 Mbyte disk space. Software collects and saves statistics in the files every hour. FlashLib includes drivers for wide-used flash chips and is provided with full C sources. We have made some minor changes to manage "disk squeeze" procces and it works fine. If you need I can provide you more information. Best Regards =============================== Chernobelsky Igal InnoWave Wireless Systems Tel: 972-3-9263600 Fax: 972-3-9263678 =============================== From ksamavedam@hns.com Sun Jul 5 17:15:51 1998 From: ksamavedam@hns.com (Krishna Samavedam) Date: Sun Jul 5 17:15:54 PDT 1998 Subject: VXWorks ppl driver problems.. Hi, I m using the vxWorks simulator "vxSim" on HP platform(HP-UX 10.20). There is always a problem with the tcp/ip connection between the vxSim process and any client on the Host machine. The vxSim NW is emulated using PPL drivers. Any Help ? regards krishna From MSchrape@atomika.com Mon Jul 6 03:34:06 1998 From: Martin Schrape Date: Mon Jul 6 03:34:09 PDT 1998 Subject: max. # of fds Hi, does anyone know what the maximum number of file descriptors is. The standard VxWorks configuration calls iosInit(int max_drivers, int max_files, char* nullDevice) with max_files = NUM_FILES defined as 50. So, what about max_files = INT_MAX ? And how many memory is allocated in the fd table for each fd? 8 bytes ? -martin --------------------------------------------------------------- Dipl.-Phys. Martin Schrape schrape@atomika.com Senior Software Engineer Tel. +49 89 315 891 34 Atomika Instruments GmbH Fax +49 89 315 59 21 Bruckmannring 40 www.atomika.com 85764 Oberschleissheim/Munich, Germany --------------------------------------------------------------- "The technology brings a certain thrill to simple tasks."(Bill Gates) From daemon@csg.lbl.gov Mon Jul 6 04:07:12 1998 From: daemon@csg.lbl.gov Date: Mon Jul 6 04:07:15 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 6 04:07:09 PDT 1998 Subject: Re: tcl scripts to build a target application outside of Tornado Subject: Re: Does VxWorks works on Windows NT 4.0? Subject: Re: VxWorks PCMCIA ATA Device on Mbx860 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: tcl scripts to build a target application outside of Tornado Date: Sat, 4 Jul 1998 10:01:27 +0200 From: Philippe Le Foll Organization: Wanadoo - (Client of French Internet Provider) Message-ID: <98070410060100.24345@fridu> References: <6nii37$hp5$1@news.cegelecproj.co.uk> Followup-To: comp.os.vxworks On Fri, 03 Jul 1998, Kevin Taylor wrote: >Has anyone successfully added tcl scripts for menu options to >Tornado 1.0.1 (Win NT) in order to perform the build of a >target outside of the main c:\Tornado file tree ? > I builded a custum TCLsh for WTX under Tornado 1.X outside of any Wind River tree and makefile [Unix&WIN32], you can find my work on www.fridu.com makefile skeleton are in Common and procuded program is jWrap. jWrap is a program that parse your target C files and produce a TCL dll that add in tcl8.x one entry for each of your procedure. This mecanism allows developper to acces from the host a target routine without having to handle wtx. hope it can help Philippe --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Does VxWorks works on Windows NT 4.0? Date: Mon, 06 Jul 1998 09:22:24 +0100 From: David Rigg Organization: GEC Marconi Avionics Message-ID: <35A08940.1A8A@gecm.NOSPAM.com> References: <01bda64c$7bf570d0$f561460f@qdrd05> Reply-To: david.rigg@gecm.NOSPAM.com Peter LIU wrote: > > Hello, > > We wnat to buy a development environment running on windows NT 4.0. Who can > tell me if VxWorks runs on Windows NT? > -- > Peter LIU Certainly does Peter - -- David Rigg \ _ / Mission Avionics Division, x__________\_(0)_/__________x GEC Marconi Avionics, Scotland @ @ (](_o_)[) @ @ Tel : (+44) (0)131 314 8223 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks PCMCIA ATA Device on Mbx860 Date: 6 Jul 98 10:27:16 GMT From: "Franke" Organization: men Message-ID: <01bda8c8$92334100$d30101c0@Franke.men.de> References: <359C4997.335D0849@mclean.sparta.com> Patrick Pinkowski schrieb im Beitrag <359C4997.335D0849@mclean.sparta.com>... > I have a Motorola MBX 860 board and BSP. I want to get a PCMCIA ATA > Flash Disk and PCMCIA ATA Hard Disk working on the MBX 860 under > VxWorks. The BSP supports PCMCIA ram cards, but nothing else. Has > anyone extended the PCMCIA support in the BSP? Any PCMCIA ATA Disk > Support? > > Thanks in advance. > > Regards > > Patrick T. Pinkowski > > ptp@mclean.sparta.com > > Hello Patrick, the pc486 standard WRS BSP includes support for ATA on PCMCIA. Ulf ================================ men mikro elektronik nuernberg software development Phone +49/911-99335-67 Fax +49/911-99335-95 Email: Ulf.Franke@men.de WWW: http://www.men.de ================================ --------------------------- End of New-News digest ********************** From jrm@mclean.sparta.com Mon Jul 6 04:54:28 1998 From: "John R. Moore III" Date: Mon Jul 6 04:54:35 PDT 1998 Subject: Re: max. # of fds the vxWorks Users Group Exploder wrote: > > Submitted-by MSchrape@atomika.com Mon Jul 6 03:34:06 1998 > Submitted-by: Martin Schrape > > Hi, > > does anyone know what the maximum number of file descriptors is. The > standard VxWorks configuration calls > > iosInit(int max_drivers, int max_files, char* nullDevice) > > with max_files = NUM_FILES defined as 50. > > So, what about max_files = INT_MAX ? > And how many memory is allocated in the fd table for each fd? 8 bytes > ? The maximum number of file descriptors is 256, and each requires 12 bytes. -- ------------------------------------------------------------ |John R. Moore | Voice -> 703 448 1683 x229 | |Principal Engineer | FAX -> 703 893 5494 | |EMAIL: | On the WEB! | |jrm@mclean.sparta.com | http://www.mclean.sparta.com | |___________________________|______________________________| | | | | /\\ SPARTA, Inc. | | | ///\\ 7926 Jones Branch Drive | | | /////\\ Suite 900 | | | ///// \\ McLean, VA 22102 | | | / \\\\\ \\ 703-448-0120 | | | \ \\\\\ // | | | \ ////// | | | \////// SPARTA | | | \//// Pride In Performance | | | \// -------------------- | | ------------------------------------------------------------ From rtp.co.uk!ihw@rtp.co.uk Mon Jul 6 06:47:49 1998 From: Ian Willats Date: Mon Jul 6 06:47:54 PDT 1998 Subject: max. # fds (VxWorks) John R. Moore III wrote: > The maximum number of file descriptors is 256, To clarify my earlier post (and disagree with the above ;-) it is easily possible to have more than 256 open files (by just increasing NUM_FILES to as big a number as you need) but just don't expect them to work with select(). (I know because I've tried.) Ian From jrm@mclean.sparta.com Mon Jul 6 07:53:54 1998 From: "John R. Moore III" Date: Mon Jul 6 07:53:57 PDT 1998 Subject: Re: max. # fds the vxWorks Users Group Exploder wrote: > > Submitted-by rtp.co.uk!ihw@rtp.co.uk Mon Jul 6 06:47:49 1998 > Submitted-by: Ian Willats > > (VxWorks) > > John R. Moore III wrote: > > > The maximum number of file descriptors is 256, > > To clarify my earlier post (and disagree with the above ;-) it is > easily > possible to have more than 256 open files (by just increasing > NUM_FILES to > as big a number as you need) but just don't expect them to work with > select(). (I know because I've tried.) Oh, I'm sorry, I made the assumption that the requestor wanted information that would leave him with a fully functioning kernel. If that is not the case, then I agree that there is no limit. The same applies to priorities, negative numbers can be used if you want a really high priority, and numbers greater that 255 if you want a really low one. But to paraphrase the above, "just don't expect them to work" :-) ------------------------------------------------------------ |John R. Moore | Voice -> 703 448 1683 x229 | |Principal Engineer | FAX -> 703 893 5494 | |EMAIL: | On the WEB! | |jrm@mclean.sparta.com | http://www.mclean.sparta.com | |___________________________|______________________________| | | | | /\\ SPARTA, Inc. | | | ///\\ 7926 Jones Branch Drive | | | /////\\ Suite 900 | | | ///// \\ McLean, VA 22102 | | | / \\\\\ \\ 703-448-0120 | | | \ \\\\\ // | | | \ ////// | | | \////// SPARTA | | | \//// Pride In Performance | | | \// -------------------- | | ------------------------------------------------------------ From MSchrape@atomika.com Mon Jul 6 08:07:14 1998 From: Martin Schrape Date: Mon Jul 6 08:07:18 PDT 1998 Subject: Re: max. # of fds At 04:54 7/6/98 PDT, you wrote: >Submitted-by jrm@mclean.sparta.com Mon Jul 6 04:54:28 1998 >Submitted-by: "John R. Moore III" > >the vxWorks Users Group Exploder wrote: >> >> Submitted-by MSchrape@atomika.com Mon Jul 6 03:34:06 1998 >> Submitted-by: Martin Schrape >> >> Hi, >> >> does anyone know what the maximum number of file descriptors is. The >> standard VxWorks configuration calls >> >> iosInit(int max_drivers, int max_files, char* nullDevice) >> >> with max_files = NUM_FILES defined as 50. >> >> So, what about max_files = INT_MAX ? >> And how many memory is allocated in the fd table for each fd? 8 bytes >> ? > > >The maximum number of file descriptors is 256, iosInit() does not fail with SHRT_MAX/2 and I opened all of them. > and each requires 12 bytes. uups. 4 bytes or 2bytes for the drvnum and 4 bytes for the return value of my xxopen is just 6 or 8 bytes, but not 12. -martin From rtp.co.uk!ihw@rtp.co.uk Mon Jul 6 09:59:37 1998 From: Ian Willats Date: Mon Jul 6 09:59:42 PDT 1998 Subject: max. # of fds (VxWorks) > does anyone know what the maximum number of file descriptors is As far as I know this is only limited by: 1) memory taken up by the fd table (fixed independent of how many you actually use, so setting NUM_FILES to INT_MAX is not a good plan) 2) memory taken up by the device drivers as you open files I don't offhand know figures for either of these - in any case (2) will vary according to the driver. One other point to watch is that if you're using select() you will potentially run into trouble once your fds get bigger than 255, because that's the maximum fd you can use in conjunction with an fd set. There is no way round this (that I know of) other than to rebuild the select() library from source. HTH Ian --------------------------------------------------------------- Ian Willats e-mail: mailto:ihw@rtp.co.uk Real-Time Products Ltd. direct: +44 (0) 121 234 6637 Chancery House, tel: +44 (0) 121 234 6600 8 Edward Street, Birmingham. fax: +44 (0) 121 234 6611 B1 2RX. England. web: http://www.rtp.co.uk --------------------------------------------------------------- From mumcu@venus.aselsan.com.tr Mon Jul 6 10:23:58 1998 From: Ahmet Mumcu Date: Mon Jul 6 10:24:03 PDT 1998 Subject: PROBLEM WITH NFS Hi, I have a problem with NFS. I make the following NFS mount successfully. ->nfsMount "toraman","/export/home0","/home0/" But when I try to access the home0 directory I receive an error as shown below. -> ls "/home0/" /home0/:No such file or directory. With nfsDevShow it seems that the directory was successfully mounted. However the same method works very well when I select to use 5.2 shell. But fails as shown above with Tornado shell. My target is PowerPC and I use VxWorks 5.3.1 with Sun Solaris host. I will be very pleased for your recomendations. Ahmet Mumcu ASELSAN Inc. Ankara-TURKEY mumcu@mst.aselsan.com.tr From jrm@mclean.sparta.com Mon Jul 6 10:55:41 1998 From: "John R. Moore III" Date: Mon Jul 6 10:55:44 PDT 1998 Subject: Re: max. # of fds the vxWorks Users Group Exploder wrote: > > Submitted-by MSchrape@atomika.com Mon Jul 6 08:07:14 1998 > > > >The maximum number of file descriptors is 256, > > iosInit() does not fail with SHRT_MAX/2 and I opened all of them. > > > and each requires 12 bytes. > > uups. 4 bytes or 2bytes for the drvnum and 4 bytes for the return > value of > my xxopen is just 6 or 8 bytes, but not 12. > > -martin > The space used is for the structure FD_ENTRY, whose size used to be 12 but I haven't checked lately. The table is created in iosInit. Any space you use to store the fd is dynamically allocated and would be over and above the 12 bytes/entry used by the table. ------------------------------------------------------------ |John R. Moore | Voice -> 703 448 1683 x229 | |Principal Engineer | FAX -> 703 893 5494 | |EMAIL: | On the WEB! | |jrm@mclean.sparta.com | http://www.mclean.sparta.com | |___________________________|______________________________| | | | | /\\ SPARTA, Inc. | | | ///\\ 7926 Jones Branch Drive | | | /////\\ Suite 900 | | | ///// \\ McLean, VA 22102 | | | / \\\\\ \\ 703-448-0120 | | | \ \\\\\ // | | | \ ////// | | | \////// SPARTA | | | \//// Pride In Performance | | | \// -------------------- | | ------------------------------------------------------------ From jrm@mclean.sparta.com Mon Jul 6 11:04:02 1998 From: "John R. Moore III" Date: Mon Jul 6 11:04:06 PDT 1998 Subject: Re: PROBLEM WITH NFS the vxWorks Users Group Exploder wrote: > > Submitted-by mumcu@venus.aselsan.com.tr Mon Jul 6 10:23:58 1998 > Submitted-by: Ahmet Mumcu > > Hi, > > I have a problem with NFS. > > I make the following NFS mount successfully. > > ->nfsMount "toraman","/export/home0","/home0/" > > But when I try to access the home0 directory I receive an error as > shown > below. > > -> ls "/home0/" > /home0/:No such file or directory. > > With nfsDevShow it seems that the directory was successfully mounted. > > However the same method works very well when I select to use 5.2 > shell. > But fails as shown above with Tornado shell. > > My target is PowerPC and I use VxWorks 5.3.1 with Sun Solaris host. > > I will be very pleased for your recomendations. > The shell in Tornado is trying to access the filesystem from your host. Put an @ symbol in front of the ls to force the command to the target. This should help. -- ------------------------------------------------------------ |John R. Moore | Voice -> 703 448 1683 x229 | |Principal Engineer | FAX -> 703 893 5494 | |EMAIL: | On the WEB! | |jrm@mclean.sparta.com | http://www.mclean.sparta.com | |___________________________|______________________________| | | | | /\\ SPARTA, Inc. | | | ///\\ 7926 Jones Branch Drive | | | /////\\ Suite 900 | | | ///// \\ McLean, VA 22102 | | | / \\\\\ \\ 703-448-0120 | | | \ \\\\\ // | | | \ ////// | | | \////// SPARTA | | | \//// Pride In Performance | | | \// -------------------- | | ------------------------------------------------------------ From MSchrape@atomika.com Mon Jul 6 14:49:28 1998 From: Martin Schrape Date: Mon Jul 6 14:49:31 PDT 1998 Subject: Re: max. # of fds At 09:59 7/6/98 PDT, you wrote: >Submitted-by rtp.co.uk!ihw@rtp.co.uk Mon Jul 6 09:59:37 1998 >Submitted-by: Ian Willats > >(VxWorks) > >> does anyone know what the maximum number of file descriptors is > >As far as I know this is only limited by: > >1) memory taken up by the fd table (fixed independent of how many you >actually use, > so setting NUM_FILES to INT_MAX is not a good plan) > >2) memory taken up by the device drivers as you open files > >I don't offhand know figures for either of these - in any case (2) will >vary according to the driver. > >One other point to watch is that if you're using select() you will >potentially run into trouble once your fds get bigger than 255, because >that's the maximum fd you can use in conjunction with an fd set. There is >no way round this (that I know of) other than to rebuild the select() >library from source. > Well, how can I use the fds > 256 for fd that does not need select and reserve the others. Should I manipulate fdTable directly? -martin --------------------------------------------------------------- Dipl.-Phys. Martin Schrape schrape@atomika.com Senior Software Engineer Tel. +49 89 315 891 34 Atomika Instruments GmbH Fax +49 89 315 59 21 Bruckmannring 40 www.atomika.com 85764 Oberschleissheim/Munich, Germany --------------------------------------------------------------- "The technology brings a certain thrill to simple tasks."(Bill Gates) From daemon@csg.lbl.gov Tue Jul 7 04:04:01 1998 From: daemon@csg.lbl.gov Date: Tue Jul 7 04:04:04 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 7 04:03:59 PDT 1998 Subject: UART/SMC interrupts Subject: Re: PPC860 target doesn't boot with WindView instrumentation ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: UART/SMC interrupts Date: Mon, 06 Jul 1998 08:26:45 -0400 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <35A0C285.4BB07B1B@lucent.com> Has anyone seen SMC1 interrupt failures on the MPC860EN at 50Mhz? Everything is fine until we run both I2C (with the patch) and SMC1 together. Eventually the SMC ends up with an unresolved interrupt. The end result is that the SIPEND register is zero, the CIPR is 0x00000010, and the CISR is 0x00000010, with the proper masks. This seems to be an impossible state for the processor. I am in contact with Motorola also, but was wondering if anyone else in THIS newsgroup has experienced this failure. We have seen this on both the MBX860 at 50Mhz and our own controller using the same chip and mask version, so I believe it may be the chip. The other thing both boards have in common are the drivers and interrupt routines from WRS. We have the BSP porting kit and the MBX860 BSP, and I have instrumented both the ISRs as well as the SIU interrupt demux and the CPM interrupt demux. If anyone else has hit this, I would appreciate a honk back. TIA - -- Doug Fraser --------------------------- Newsgroups: comp.os.vxworks Subject: Re: PPC860 target doesn't boot with WindView instrumentation Date: Mon, 06 Jul 1998 10:25:46 -0400 From: Douglas Fraser Organization: Lucent Technologies Message-ID: <35A0DE6A.A321D161@lucent.com> References: <6nqgcn$k1c$1@galaxy.mchh.siemens.de> Andy, We have used WindView with both the Motorola MBX860 and our locally developed MPC860 target. I can't recall having to do anything special to get WindView running on the MBX860. Is this an ADS860 or your own board that you based on the ADS860 BSP? I ask only because we are seeing interrupt problems with SMC1 on both the MBX860 and our local board. See my previous post, just before yours. The problem causes the local terminal to stop printing. Any task that subsequently calls printf() PENDS forever. - -- Doug Fraser --------------------------- End of New-News digest ********************** From hakan.markor@metric.se Tue Jul 7 05:51:52 1998 From: hakan.markor@metric.se Date: Tue Jul 7 05:51:56 PDT 1998 =20 Dear VxWorkers, =20 I will soon get a FORCE Powercore ppc603 and 2 PMC modules to powerup=2E One is a 1553 from Radstone and the other is a WAN-4534 from=20 Interphase with a ppc860 onboard=2E =20 Does any one have any experience in this setup regaring getting=20 VxWorks bsps:s with drivers up and running or any info that they like=20 to share I would really appreciate it=2E =20 Regards Hakan =20 *********************************************************** | | | Hakan Markor M=2ESc=2E Hakan=2EMarkor@metric=2Ese | | Senior Software Engineer | | Metric Tel: +46 8 629 03 68 | | Box 1494 Fax: +46 8 629 03 01 | | S-171 29 Solna http://www=2Emetric=2Ese | | Sweden | | | ************************************************************ From rtp.co.uk!jge@rtp.co.uk Tue Jul 7 06:46:33 1998 From: John Efstathiades Date: Tue Jul 7 06:46:38 PDT 1998 Subject: RE: PROBLEM WITH NFS Ahmet Mumcu wrote: > I have a problem with NFS. > > I make the following NFS mount successfully. > > ->nfsMount "toraman","/export/home0","/home0/" > > But when I try to access the home0 directory I receive an error as shown > below. > > -> ls "/home0/" > /home0/:No such file or directory. > > With nfsDevShow it seems that the directory was successfully mounted. > > However the same method works very well when I select to use 5.2 shell. > But fails as shown above with Tornado shell. The reason that you cannot see the nfs directory is that the ls command is built into the windSh tool, and can only see file systems mounted on your host. In order to view file systems that are local to, or mounted on, your target, you have to use the ls command found in the usrLib.c file in the WIND_BASE/target/src/usr directory. If you define a VxWorks 5.2 style configuration this file is automatically included. To use the target based ls, compile the usrLib.c file and load the object file onto your target. The target ls is invoked by typing @ls, or just use ll, to get a long listing. John Efstathiades ---------------------------------------------------------------- John Efstathiades Real Time Products Ltd e-mail: jge@rtp.co.uk Chancery House, Tel: +44 (0)121 234 6600 8 Edward Street, Fax: +44 (0)121 234 6611 Birmingham. B1 2RX. England. Web: www.rtp.co.uk ---------------------------------------------------------------- From Johan.van.der.Werf@asml.nl Wed Jul 8 02:38:43 1998 From: Johan.van.der.Werf@asml.nl (Johan van der Werf) Date: Wed Jul 8 02:38:47 PDT 1998 Subject: File-descriptor type Dear fellow VxWorks (5.3.1) users, To solve the misuse of getsockopt() on a non socket fd, I am in need of a surefire way to determine if an fd is related to a socket, or to some other driver. I tried using iosFdValue(). But unfortunately, for a socket fd, this function does not return a pointer to a DEV_HDR structure as for most drivers. Instead it seems to return a pointer to the socket structure. I can get the DEV_HDR pointer from the fdTable, but that would mean including private/ioLibP.h, which I would only do if there is no other way. How do you determine the type of driver connected to an fd? regards, -- |Johan van der Werf | tel: ++31 (0)40 2304280 |ハンス | |ASM Lithography | fax: ++31 (0)40 2304455 |ヴァン | |Dept: SEAL | email: ujw@asml.nl |デル | |Room: 5C023 | vdwerf@xs4all.nl |ウエルフ| From daemon@csg.lbl.gov Wed Jul 8 04:02:03 1998 From: daemon@csg.lbl.gov Date: Wed Jul 8 04:02:06 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 8 04:02:00 PDT 1998 Subject: Interrupt handling with PPC603 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Interrupt handling with PPC603 Date: Tue, 07 Jul 1998 11:34:34 GMT From: markus_mitterer@sbu1.storkgroup.com (Markus Mitterer) Organization: STK Message-ID: <35a2080b.0@mailserv.stk.co.at> Reply-To: "Markus Mitterer" Hi, I am using VxWorks with a mv2303-BSP (PPC603). I have some problems with the interrupt handling. I used the functions: sysUniverseInit sysIntEnable intConnect There is a SystemVector- and an UniverseVector Table. When do I use the UniverseVector Table ? When I reset the interrupt source at the end of my ISR, the CPU tries to read a second time an interrupt vector and I get the message: "bad vme interrupt 0" (I already watched the IRQ-Line => the interrupt occured only once) If the interrupt frequency is low (10Hz), I need more time for my ISR. Has anybody some experience with this troubles ??? Some source code examples would be really helpful. Thanks in advance, Mark --------------------------- End of New-News digest ********************** From ThanosJH@Npt.NUWC.Navy.Mil Wed Jul 8 05:10:57 1998 From: Thanos John H NUWCDIVNPT Date: Wed Jul 8 05:11:01 PDT 1998 Subject: How do I install a DOS fs on an Iomega 1 GB Jaz drive? I have just recently requested to receive messages posted to this interest group, so I apologize if the answer to this question has already been posted. I am using VxWorks 5.2 on a Motorola MVME167 single board computer. I have three peripheral devices connnected to the SCSI controller on the MVME167 board: a hard disk drive, a floppy disk drive, and an Iomega 1 GB external Jaz drive. I am able to modify the VxWorks file, userScsi.c, to correctly initialize a DOS file system on the hard disk drive and floppy disk drive. However, I am unable to do the same for the Jaz drive. The VxWorks functions, scsiPhysDevCreate() and scsiBlkDevCreate() complete successfully for the Jazz drive, and I can enter scsiShow at the VxWorks shell prompt to display information about the Jaz drive. However, the VxWorks function, dosFsDevInit() fails for the Jaz drive during VxWorks initialization. I am treating the Jaz drive as another floppy drive, since both drives employ removable disks. I suspect that the reason that dosFsDevInit() fails for the Jaz drive is because of one or more of the following reasons: 1.) when the userScsiConfig() function invokes scsiModeSelect(), the modeData argument (i.e., modeData[1]] = 0x1b) to the scsiModeSelect() function is wrong for this Jaz drive. Does anyone know what modeData[1] should be? 2.) when the userScsiConfig() function invokes scsiBlkDevInit(), the disk geometry arguments are wrong for this Jaz drive. Information at the Iomega web site seemed to indicate that the number of sectors per track is 32, and the number of heads is 64, but I don't know if this is really true. Does anyone know what those arguments should be set to? I would be interested in hearing from anyone who has successfully initialized a DOS file system on a 1 GB Jaz drive using VxWorks. Thank you. John Thanos thanosjh@nuwc.npt.navy.mil From brett.smith@wg.com Wed Jul 8 06:01:48 1998 From: brett.smith@wg.com Date: Wed Jul 8 06:01:52 PDT 1998 Subject: DosFs file date and time I am using DosFs and do a lot of renaming and copying of test/good files at bootup time. I want files that are renamed or copied to retain the date/time of the original file. Does anyone know how I can copy/rename without modifying the date/time? Or, how to set the date/time for a file? Thanks, BWS vxworks From MSchrape@atomika.com Wed Jul 8 07:13:57 1998 From: Martin Schrape Date: Wed Jul 8 07:14:02 PDT 1998 Subject: Re: File-descriptor type At 02:38 7/8/98 PDT, you wrote: >Submitted-by Johan.van.der.Werf@asml.nl Wed Jul 8 02:38:43 1998 >Submitted-by: Johan.van.der.Werf@asml.nl (Johan van der Werf) > > > > Dear fellow VxWorks (5.3.1) users, > > To solve the misuse of getsockopt() on a non socket fd, I am in need of > a surefire way to determine if an fd is related to a socket, or to some > other driver. I tried using iosFdValue(). But unfortunately, for a > socket fd, this function does not return a pointer to a DEV_HDR > structure as for most drivers. Instead it seems to return a pointer to > the socket structure. I can get the DEV_HDR pointer from the fdTable, > but that would mean including private/ioLibP.h, which I would only do > if there is no other way. > > How do you determine the type of driver connected to an fd? 1st idea. Call getsockname() or ioctl() with FIONBIO. If they return OK than it's a socket fd. 2nd idea. Use the not documented function iosFdDevFind() in iosLib.h -martin From ensslin@3SOFT.de Wed Jul 8 08:29:18 1998 From: Fritz Ensslin Date: Wed Jul 8 08:29:22 PDT 1998 Subject: longjmp in signal handler ? Hi, VxWorkers can anyone of you help me with the following question: Is it allowed to leave a VxWorks signal handler using longjmp ? Here is a short code example for what I tried: jmp_buf env; void fooTask (void) { setjmp (&env); signal (SIGXYZ, myHandler); while (TRUE) { /* do some processing */ } } void myHandler (int i) { longjmp (env, 1); } =20 What I observed is, that the first occuring signal is handled as = expected - the task goes back to the point after the setjmp call. But = all following signals then have no effect, the signal handler isn't even = entered.=20 If longjmp isn't allowed in signal handlers, what else could I do, to = transfer control back to a specific point during startup, when a signal = arrives ? TIA, Friedrich ----------------------------------------------------- Friedrich Ensslin 3SOFT GmbH Wetterkreuz 19a , D-91058 Erlangen , Germany phone: +49-9131-7701-113 , fax: +49-9131-7701-333 email: ensslin@3SOFT.de From cgrames@mdc.com Wed Jul 8 10:22:38 1998 From: Charlie Grames Date: Wed Jul 8 10:22:41 PDT 1998 Subject: Problems with vxWorks.st_rom VxWorks 5.3.1 MVME2604 1.1/4 Within the last few weeks, I have complained about problems trying to get VxWorks in ROM (using vxWorks.st_rom). The last problem I ran into is a legitimate problem with VxWorks that you may want to be aware of. The details can be found in the attached message I sent to Wind River Support. If anyone wants more details, feel free to contact me. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com From: Charlie Grames To: internet:carolina.fox@wrs.com Date: 7/8/98 12:08pm Subject: TSR 111880 Carolina, Following is the information for TSR 111880: SYMPTOMS: The vxWorks.st_rom image does not initialize correctly. After powering on the target, the following is displayed at the console: Error in boot line: [various garbage] Because the boot line does not contain valid network addresses, no network interfaces are initialized. PROBLEM: The following code from usrNetwork.c is responsible for copying the boot line from nonvolatile RAM to the boot line location defined in configAll.h (for PowerPC, 0x4200): void usrBootLineInit ( int startType ) { if ((startType & BOOT_CLEAR) || (* BOOT_LINE_ADRS == EOS)) { /* either cold boot or empty boot line -- initialize boot line */ if ((sysNvRamGet (BOOT_LINE_ADRS, BOOT_LINE_SIZE, 0) == ERROR)) { /* no non-volatile RAM -- use default boot line */ strcpy (BOOT_LINE_ADRS, DEFAULT_BOOT_LINE); } } } Unfortunately, with vxWorks.st_rom, startType is BOOT_WARM_AUTOBOOT (as defined in _sysInit [sysALib.s]), and *BOOT_LINE_ADRS is not guaranteed to be zero (the definition of RESERVED in configAll.h causes it to be excluded from memory clearing in romStart [bootInit.c]). WORKAROUND: The following change to romStart in bootInit.c corrects the problem: Original: #endif /* UNCOMPRESS */ #endif /* ROM_RESIDENT */ } #endif /* (CPU_FAMILY != MIPS) && (!defined (BOOTCODE_IN_RAM)) */ Revised: #endif /* UNCOMPRESS */ #endif /* ROM_RESIDENT */ /* ** Ensure the boot line is null. This is required for targets whose boot ** lines are contained in memory which is excluded from clearing. */ *BOOT_LINE_ADRS = '\0'; } #endif /* (CPU_FAMILY != MIPS) && (!defined (BOOTCODE_IN_RAM)) */ Please let me know if you have any questions. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com From joew@wtc.gov Wed Jul 8 11:44:57 1998 From: "Joe W." Date: Wed Jul 8 11:45:01 PDT 1998 Subject: VxWorks: Task trace / Socket Problem Hello, I've come across a strange problem which maybe someone else out there has seen. A task has a TCP/IP socket open to another task. The first task gets stuck in a write to the socket file descriptor, and never returns (at least not for many hours.) The second task, the one which should be receiving data across the socket, appears to be running just fine, doing everything as expected, except that each time it used FD_ISSET keeps returning false, indicating that the second task doesn't see anything sitting on the socket to read. The really wierd thing is that when I try to task trace the second task, I get an access fault in the shell. Then the second task suspends itself. The access address for the access fault is 0xfffffffe, although the PC and SP shown in the "i" command seem to be correct. taskRegsShow does not show anything unusual, nor does checkStack. I'm doubly confused by this... I'm assuming that the reason the write hangs is that the socket has dropped, but why can't the task be traced? And, I've tried to see if the system knows about the socket with inetstatShow, but I can't figure out how to correlate the data there with particular socket file descriptors. Any help which anyone could provide me would be greatly appreciated. --Joe W. (joew@wtc.gov) From Jim.Rosenow@mts.com Wed Jul 8 12:27:01 1998 From: "Rosenow, Jim" Date: Wed Jul 8 12:27:04 PDT 1998 Subject: Display FP regs with ti command with windsh on Tornado 1.0.1 vxWo How do I get windsh, the tornado shell, to display the FP regs for a VX_FP_TASK on PowerPC? Any help would be appreciated. Thanks. Jim Rosenow MTS System Corp. 14000 Technology Drive Eden Prairie, MN 55344 From Bob.Ferrara@digital.com Wed Jul 8 13:43:08 1998 From: Bob Ferrara Date: Wed Jul 8 13:43:11 PDT 1998 Subject: VxWorks driver for the QLogic ISP 1020 SCSI controller chip This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------ =_NextPart_000_01BDAA8F.780C01E0 Content-Type: text/plain Fellow VxWorks programmers, Does anyone out there have a VxWorks device driver for the QLogic ISP 1020 SCSI controller chip that we could obtain? Thanks in advance, Bob Ferrara Compaq Computer Corporation CustomSystems Real-time Software Engineering Manager Phone: 603-884-3094 Fax: 603-884-5191 E-mail: Bob.Ferrara@Digital.com ------ =_NextPart_000_01BDAA8F.780C01E0 Content-Type: application/ms-tnef Content-Transfer-Encoding: base64 eJ8+IgEUAQaQCAAEAAAAAAABAAEAAQeQBgAIAAAA5AQAAAAAAADoAAEIgAcAGAAAAElQTS5NaWNy b3NvZnQgTWFpbC5Ob3RlADEIAQWAAwAOAAAAzgcHAAgAEAArACMAAwBFAQEggAMADgAAAM4HBwAI ABAAKgA7AAMAXAEBCYABACEAAABBMDIwM0NCRTc5MTZEMjExQjJDQzAwMDBGODIyM0JDMgAFBwEE gAEAPAAAAFZ4V29ya3MgZHJpdmVyIGZvciB0aGUgUUxvZ2ljIElTUCAxMDIwIFNDU0kgY29udHJv bGxlciBjaGlwACAUAQ2ABAACAAAAAgACAAEDkAYA/AUAACAAAAACAQkQAQAAAAICAAD+AQAANgMA AExaRnXSWedtAwAKAHJjcGcxMjX+MgD/AgYCpAPkBesCgwBQEwNUAgBjaArAc2V0/jIGAAbDAoMO UAPVBxMCg+IzEedwcnES4AdtAoOuNAPFFWUUSDUTzX0KgL8IzwnZAoAKgQ5xC2BuDhAUMDMU4GgF sHpkb2pjAAAqElUgApEdMGwvHWUK+xiiAdAgCoVGZUZsGhAH4FZ4VwWwa8sEIBWAbwnAYW0HgBKg ZiwKhQqFRG8HkQBweZkCIGUgCGAFQHRoBJBrI3ASgHYjcGEgpw5wdvsN4CNwZAUQJFAFwAIQBcAz I9EfxlFMIVAN4CBJGFNQIBxQH6FTQ1P8SSAFoAIwA2AgYCXREnBrBSAjwWEFQHcjcAWgdbRsZCOA YgGQC4A/Ifw6VBKAbiEBC4AjEGR2vwBwJWAh7Qr7FxIfoUIqIL0gMXIhcCFwIfwIUG0tMJ5xMDMj oCXRCFBycAWwfSlgaQIgLQcLZBThC/Jp8TAwdXN0A3AtzAogEvE5M5BlbReALdwtBVJlXQdALTGw B4AGAG8BgHf7CsAjcEUcMAuACeAFEBwwswXQAHBhZwSQLQVQHKApI2A6IDnjNhxgLTiAODQtMzA5 NB/WDGF4OdU6CjUxOTF5LQVFLQDAAxA50y6xLjEu9UBEaScgAZBsLi0FoG0tBRmhAEBQAAAeAHAA AQAAADwAAABWeFdvcmtzIGRyaXZlciBmb3IgdGhlIFFMb2dpYyBJU1AgMTAyMCBTQ1NJIGNvbnRy b2xsZXIgY2hpcAACAXEAAQAAABYAAAABvaqw26O3FdOBFN0R0rP2AAD4ebrTAABAADkAP2WqE7Gq vQEDAPE/CQQAAB4AMUABAAAACQAAAEZFUlJBUkFCAAAAAAMAGkAAAAAAHgAwQAEAAAAJAAAARkVS UkFSQUIAAAAAAwAZQAAAAAADAP0/5AQAAAMAJgAAAAAAAwA2AAAAAAACAUcAAQAAADYAAABjPVVT O2E9IDtwPURpZ2l0YWw7bD1DU1MtQU1BWktPMkEtOTgwNzA4MjA0MzM1Wi01MDc2NQAAAAIB+T8B AAAATgAAAAAAAADcp0DIwEIQGrS5CAArL+GCAQAAAAAAAAAvTz1ESUdJVEFML09VPVNCVUFNQUVO Ry9DTj1SRUNJUElFTlRTL0NOPUZFUlJBUkFCAAAAHgD4PwEAAAAMAAAAQm9iIEZlcnJhcmEAHgA4 QAEAAAAJAAAARkVSUkFSQUIAAAAAAgH7PwEAAABOAAAAAAAAANynQMjAQhAatLkIACsv4YIBAAAA AAAAAC9PPURJR0lUQUwvT1U9U0JVQU1BRU5HL0NOPVJFQ0lQSUVOVFMvQ049RkVSUkFSQUIAAAAe APo/AQAAAAwAAABCb2IgRmVycmFyYQAeADlAAQAAAAkAAABGRVJSQVJBQgAAAABAAAcwBwmsLrCq vQFAAAgwf4aB/rCqvQEeAD0AAQAAAAEAAAAAAAAAHgAdDgEAAAA8AAAAVnhXb3JrcyBkcml2ZXIg Zm9yIHRoZSBRTG9naWMgSVNQIDEwMjAgU0NTSSBjb250cm9sbGVyIGNoaXAAHgA1EAEAAABDAAAA PEY0NkNEN0MxQzhCRUQxMTFCMkM3MDAwMEY4MjIzQkMyNjZGNDlEQGNzcy1hbWF6a28yYWMuemtv LmRlYy5jb20+AAALACkAAAAAAAsAIwAAAAAAAwAGEHRFXtMDAAcQGwEAAAMAEBAAAAAAAwAREAAA AAAeAAgQAQAAAGUAAABGRUxMT1dWWFdPUktTUFJPR1JBTU1FUlMsRE9FU0FOWU9ORU9VVFRIRVJF SEFWRUFWWFdPUktTREVWSUNFRFJJVkVSRk9SVEhFUUxPR0lDSVNQMTAyMFNDU0lDT05UUk9MTEVS AAAAAAIBfwABAAAAQwAAADxGNDZDRDdDMUM4QkVEMTExQjJDNzAwMDBGODIyM0JDMjY2RjQ5REBj c3MtYW1hemtvMmFjLnprby5kZWMuY29tPgAAV38= ------ =_NextPart_000_01BDAA8F.780C01E0-- From espin@idiom.com Wed Jul 8 17:12:57 1998 From: Geoffrey Espin Date: Wed Jul 8 17:13:00 PDT 1998 Subject: Re: VxWorks driver for the QLogic ISP 1020 SCSI controller chip Bob, > Does anyone out there have a VxWorks device driver for the > QLogic ISP 1020 SCSI controller chip that we could obtain? > Thanks in advance, > Bob Ferrara > Compaq Computer Corporation > CustomSystems > Real-time Software Engineering Manager > Phone: 603-884-3094 > Fax: 603-884-5191 > E-mail: Bob.Ferrara@Digital.com Contact Shishir Shah of QLogic at s_shah@qlc.com. Geoff -- Geoffrey Espin espin@idiom.com From cwcgoisp@leonis.nus.edu.sg Wed Jul 8 23:55:09 1998 From: Goi Date: Wed Jul 8 23:55:13 PDT 1998 Subject: Help with ether_output routine I am writing a 4.3 BSD-style network interface driver but currently using SENS for Tornado 1.0 . The arguments to the ether_output routine seems to be different from that of documented in VxWorks BSP Porting Kit 1.1 . Does anyone know what is the difference between them? Pls advice. Thanks! Goi, CWC From daemon@csg.lbl.gov Thu Jul 9 04:02:48 1998 From: daemon@csg.lbl.gov Date: Thu Jul 9 07:45:06 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 9 04:02:45 PDT 1998 Subject: Re: Max number of TCP connections ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max number of TCP connections Date: Wed, 08 Jul 1998 16:00:34 +0200 From: Philippe Klein Organization: Vocaltec Communications Ltd Message-ID: <35A37B82.E19F8E98@vocaltec.com> References: <35A354F9.4A43EE64@vocaltec.com> I found the solution: NUM_FILES which sets the max number of files open simultaneously in VxWorks (target/config/all/configAll.h) as a default value of 50. Incrementing this number fixes the limited number of simultaneous TCP connections problem. /Philippe Philippe Klein wrote: > I wrote 2 small TCP Server and TCP client ACE programs to check how many > > simultaneous TCP connections I can open on a single port. > > I run both Client and Server on VxWorks and NT. Every Client-Server > combination on Nt & Vxworks works fine (opening 256 connections) > except for the Vxwork client which refuses to open more than 44 > simultaneous connections. > > I did not find any parameter fixing the max number of connections. Any > hint will be appreciated. > > Thank you /Philippe --------------------------- End of New-News digest ********************** From denise@greenspring.com Thu Jul 9 09:00:19 1998 From: Denise Denney Date: Thu Jul 9 10:19:23 PDT 1998 Subject: MPC860T Does anyone out there have experience with the MPC860T? I'm looking for a vxworks driver for the Fast Ethernet Controller or for someone who might be interested in working on one. Any help/advice would be much appreciated. Thanks, Denise Denise Denney Software Engineer SBS GreenSpring Modular I/O 181 Constitution Drive Menlo Park, CA 94025 Phone: 650-327-1200 Fax: 650-327-3808 denise@greenspring.com ---------------------- From bdf@amherst.com Thu Jul 9 12:58:17 1998 From: Bruce Friedman Date: Thu Jul 9 12:58:26 PDT 1998 Subject: Re: comp.os.vxworks newsdigest VxWorkers - I hope someone can help me with this mysterious problem - Problem: I am running into difficulties when I try to get my Tulip interface on my AlphaVME card to accept a *UDP* packet destined to a multicast ip addresses (224.0.0.0 - 239.255.255.255). (See below for my complete system configuration details). I would like to know if there is any way to setup vxWorks to accept a *UDP* packet destined to a multicast address [at the UDP level - not the ethernet level]. Situation: I am sending some UDP packets from an Alpha workstation [that is on the same network as our AlphaVME card] to IP address 224.0.1.1 [a well known multicast address]. I have a number of UDP receivers listening for these packets - one of which is on another Alpha workstation and the other is on our vxWorks AlphaVME box. The receiver on the Alpha workstation originally "joined" the group 224.0.1.1 and hence is able to receive the UDP multicast packets successfully. However, this is not the case on our AlphaVME board running vxWorks. Even after using the function "tuFilterSetup" to allow the AlphaVME tu interface to accept multicast addresses, the packets only make it to the "ethernet level"; the UDP level never sees the packets. SYSTEM: VxWorks version V5.2D KERNEL: WIND version 2.4 Copyright Wind River Systems, Inc., 1984-1997 CPU: Digital Alpha VME 4/288. Processor #0 Memory 64 Meg. DEC Tulip Ethernet Drivers Thanks Bruce Friedman -------------------------------------------------------------------------- _/ _/_/_/_/_/_/ Bruce Friedman | Amherst Systems, Inc FROM _/_/_/_/ bdf@amherst.com | 30 Wilson Rd. _/_/_/_/_/_/ Voice 716-631-0610 x213 | Buffalo, NY 14221 _/ Fax 716-631-0629 | From gnn@wrs.com Thu Jul 9 13:55:53 1998 From: "George V. Neville-Neil" Date: Thu Jul 9 13:55:56 PDT 1998 Subject: Re: comp.os.vxworks newsdigest At 12:58 PM 7/9/98 PDT, the vxWorks Users Group Exploder wrote: >SYSTEM: VxWorks version V5.2D >KERNEL: WIND version 2.4 > Copyright Wind River Systems, Inc., 1984-1997 >CPU: Digital Alpha VME 4/288. Processor #0 >Memory 64 Meg. >DEC Tulip Ethernet Drivers VxWorks 5.2 has a BSD 4.3 network stack which is not enabled for Multicast. You need to use the new SENS stack for that. Later, George ---------------------------------------------------------------------------- George V. Neville-Neil gnn@wrs.com "I have crawled broke and desperate thorugh the dumpsters of the lord." --- Firewater from The Ponzi Scheme From daemon@csg.lbl.gov Fri Jul 10 04:06:50 1998 From: daemon@csg.lbl.gov Date: Fri Jul 10 04:06:54 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 10 04:06:48 PDT 1998 Subject: Re: DosFs file date and time Subject: Re: how to unload *.o ? Subject: Re: how to unload *.o ? Subject: dosFs file truncate ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: DosFs file date and time Date: 9 Jul 1998 13:30:51 -0600 From: bgeer@xmission.xmission.com (bgeer) Organization: XMission http://www.xmission.com Message-ID: <6o35pb$srv$1@xmission.xmission.com> References: <9807088999.AA899902829@smtp.wg.com> chc@nasa2.ksc.nasa.gov (Charles H. Chapman) writes: >On Wed, 08 Jul 98 09:02:15 -0500, brett.smith@wg.com wrote: >> >> I am using DosFs and do a lot of renaming and copying of test/good >> files at bootup time. I want files that are renamed or copied to >> retain the date/time of the original file. >> >> Does anyone know how I can copy/rename without modifying the >> date/time? Or, how to set the date/time for a file? >Check out the "utime" and/or "dosFsDateSet" and/or "dosFsDateTimeInstall" >functions in the VxWorks Reference Manual. Here's code that works for me. A footnote, tho: I haven't encountered a BSP yet that set VxWorks time to current clock time. One must roll their own code to read [& interpret, & offset] the values in a RTC chip to set the system time & data. /* * dosFsDate.c * * Provide date to DOS filesystem -- see "man -M /man dosFsLib" * * STATUS dosFsDateTimeRetrieve( DOS_DATE_TIME* d ) * */ #include "vxWorks.h" #include "dosFsLib.h" #include "time.h" /*______*/ STATUS dosFsDateTimeRetrieve( DOS_DATE_TIME* d ) { /* See "man -M /man dosFsLib" for dosFsDateTimeInstall() info: */ extern int tod_century ; const time_t seconds = (const) time( (time_t) NULL ) ; struct tm* t = localtime( &seconds ) ; /* * Per time.h: * map years since 1900 to current year * map months range 0..11 to 1..12 * assume day of week not used. * See /h/time.h for struct tm definition; * See /h/dosFsLib for DOS_DATE_TIME [struct] definition. */ d->dosdt_year = t->tm_year ; if (tod_century != ERROR) /* If we already have century recorded in tod.c: */ d->dosdt_year += tod_century ; else /* This should work until 2097! */ d->dosdt_year += ( (t->tm_year > 96) ? 1900 : 2000 ) ; d->dosdt_month = t->tm_mon + 1 ; /* Map 0..11 to 1..12 */ d->dosdt_day = t->tm_mday ; d->dosdt_hour = t->tm_hour ; d->dosdt_minute = t->tm_min ; d->dosdt_second = t->tm_sec ; return( OK ) ; } /* dosFsDateTimeRetrieve */ /*______*/ STATUS dosFsDateInstall() { /* Provide a function for the file system to get current time & date: */ dosFsDateTimeInstall( dosFsDateTimeRetrieve ) ; return( OK ) ; } /* dosFsDateInstall() */ - -- <> Robert Geer & Donna Tomky / * <> <> bgeer@xmission.com | _o * o * o <> <> Salt Lake City, Utah | -\<, * <\ <> U S A | O/ O __ /__, /> <> --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to unload *.o ? Date: Thu, 09 Jul 1998 20:56:44 GMT From: jirie@my-dejanews.com Organization: Deja News - The Leader in Internet Discussion Message-ID: <6o3aqc$d2t$1@nnrp1.dejanews.com> References: <35A50574.4656@crhc.uiuc.edu> In article <35A50574.4656@crhc.uiuc.edu>, Lan Chen wrote: > Hi, I need to unload a module which I loaded with: > > > ld < xxx.o > > Can you tell me how to unload this module? > Thanks! > unld "xxx.o" should work. Jiri - -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum --------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to unload *.o ? Date: Fri, 10 Jul 1998 09:10:16 +0200 From: grave Organization: I.P.N Orsay Message-ID: <35A5BE57.8E@ipnsun5.in2p3.fr> References: <35A50574.4656@crhc.uiuc.edu> <6o3aqc$d2t$1@nnrp1.dejanews.com> jirie@my-dejanews.com wrote: > > In article <35A50574.4656@crhc.uiuc.edu>, > Lan Chen wrote: > > Hi, I need to unload a module which I loaded with: > > > > > ld < xxx.o > > > > Can you tell me how to unload this module? > > Thanks! > > > > unld "xxx.o" > should work. the function have to be included in the kernel. xavier --------------------------- Newsgroups: comp.os.vxworks Subject: dosFs file truncate Date: Fri, 10 Jul 1998 11:38:18 +0200 From: "digitek" Organization: digitek s.r.l. Message-ID: <6o4n80$nj9$1@server-b.cs.interbusiness.it> My problem is to truncate a dosFs file, but at the beginning of the file instead at the end . Is this possible in some way ? Can I access to the FAT to do thath ? thanks very mutch Paolo MONARI digitek@arcanet.it --------------------------- End of New-News digest ********************** From amolkho@lannet.com Fri Jul 10 05:11:22 1998 From: "Adi molkho DVP-TA" Date: Fri Jul 10 05:11:26 PDT 1998 Subject: Drivers in vxWorks 5.3.1 vxWorks users We have wrote several drivers for vxWorks 5.2 and 5.3 and we are now starting to develop in 5.3.1 . We understand that there are some differences in the tcp/ip stack of this new version Does anyone know if this changes might influence our drivers (interfaces etc..) Thanks Adi +---------------------------------+ Adi Molkho Lannet a company of Madge Networks amolkho@lannet.com 972-3-6458351 From sl065st@uni-duisburg.de Fri Jul 10 07:53:55 1998 From: stevens Date: Fri Jul 10 07:53:58 PDT 1998 Subject: CAN Open driver for vxworks Hi VxWorkers ! We need a driver for CAN Open. The target is a CPU32. Has anyone experience writing a driver for CAN in VxWorks ? Where can we get code for this driver ( source code) ? Thanks in advance, Torsten -- ************************************************************************** fraunhofer institute of microelectronic circuits and systems department sat name torsten stevens mailto:stevens@ims.fhg.de address finkenstrasse 61 phone +0049/203/3783-238 47057 duisburg fax +0049/203/3783-266 germany http://www.fhg.de/depts/ims-du-e.html ************************************************************************** From william.j.sheehan@lmco.com Fri Jul 10 11:25:11 1998 From: Bill Sheehan Date: Fri Jul 10 11:25:15 PDT 1998 Subject: Parallel Port Interrupt handler for VxWorks > Hi VxGroupLeader ! > I don't have a CAN opener, But does anyone have a parallel port interrupt handler that removes unnecessaryINT7 interrupts? Platform: PowerPC 604e Manufacturer: I won't say who they are, they should know better than to say they provide a "Centronics Parallel Port that can be bidirectional with simple program drivers" -- they did't provide any simple drivers. Parallel Port: Exar ST16C552 PIC: Intel 82378ZB > Where can we get code for this driver ( source code) ? > > Thanks in advance, > Bill Sheehan From KARTHUR@prodigy.net Fri Jul 10 18:43:43 1998 From: "Kurt A Lambrecht" Date: Fri Jul 10 18:43:46 PDT 1998 Subject: VxWorks Target Server Connect Problems This is a multi-part message in MIME format. ------=_NextPart_000_0027_01BDAC4B.551F9620 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I am having problems in connecting a Compact PCI Gespac PCIPPC-3 board = to a Sun/Solaris as a target. =20 =20 Tgtsvr is returns the following messages:=20 tgtsvr (vxTarget@elcapitan): <--- not exactly sure what i saw here) License request... authorized on host 'elcapitan'. Attaching backend... succeeded. Connecting to target agent...=20 Error: rpccore backend client RPC: Timed out=20 =20 However, I have included "vxTarget" lines in the /etc/hosts and .rhosts = files to look just like the lines for the other targets in the lab.=20 =20 Boot messsages include: =20 Attaching network interface dc0... done=20 Attaching network interface lo0... done =20 Loading 494580... done Starting at 0x30000 =20 Attaching network interface dc0... done =20 Attaching network interface lo0... done =20 NFS client support not included =20 =20 The boot messages eventually return the "WDB Ready" and it looks like it = is ready to go. =20 =20 But I cannot seem to get a target server to connect. Anyone got any = suggestions on how to debug this problem? =20 Thanx, Kurt Lambrecht=20 LTS Corporation=20 ------=_NextPart_000_0027_01BDAC4B.551F9620 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I am having problems in connecting a = Compact PCI=20 Gespac PCIPPC-3 board to a Sun/Solaris as a target. 
 
Tgtsvr is returns the following messages: =
tgtsvr (vxTarget@elcapitan):   <--- not = exactly=20 sure what i saw here)
    License request... authorized on = host=20 'elcapitan'.
    Attaching backend...=20 succeeded.
    Connecting to target agent...=20
    Error: rpccore backend client RPC: Timed=20 out 
 
However, I have included "vxTarget" lines = in the=20 /etc/hosts and .rhosts files to look just like the lines for the other = targets=20 in the lab. 
 
Boot messsages include:
 
Attaching network interface dc0... done=20
Attaching network interface lo0...=20 done
 
    Loading 494580... = done
    Starting=20 at 0x30000
 
    Attaching network interface = dc0...=20 done       
    Attaching=20 network interface lo0... = done        =20
    NFS client support not=20 included           = ;       =20
 
The boot messages eventually return the "WDB = Ready"=20 and it looks like it is ready to go. 
 
But I cannot seem to get a target = server to=20 connect.  Anyone got any suggestions on how to debug this=20 problem?
 
Thanx,
Kurt Lambrecht 
LTS = Corporation 
------=_NextPart_000_0027_01BDAC4B.551F9620-- From daemon@csg.lbl.gov Sat Jul 11 04:05:22 1998 From: daemon@csg.lbl.gov Date: Sat Jul 11 04:05:25 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 11 04:05:20 PDT 1998 Subject: Re: how to unload *.o ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to unload *.o ? Date: Fri, 10 Jul 1998 20:42:44 GMT From: bruce@CUTMEOUTscimisys.com (Bruce Wedding) Organization: NeoSoft, Inc. Message-ID: <35a67c72.10177265@news.neosoft.com> References: <35A50574.4656@crhc.uiuc.edu> <35A640BB.B59F7017@cadence.com> On Fri, 10 Jul 1998 09:26:36 -0700, James Short wrote: > unld xxx.o will work but if you just want to load an updated version of >xxx.o, the old verison is automatically unloaded by doing another ld > References: <35A50574.4656@crhc.uiuc.edu> <01bdabd3$6462fd20$43b3fcc1@850155517> You also need to build VxWorks with "INCLUDE_UNLOAD" defined. Otherwise, you won't get 'unld'. Carlos Sham csham@ibm.net - -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum --------------------------- End of New-News digest ********************** From iliab@telegate.co.il Sun Jul 12 09:26:50 1998 From: Ilia Bosis Date: Sun Jul 12 09:26:54 PDT 1998 Subject: Re: Drivers in vxWorks 5.3.1 This is a multi-part message in MIME format. --------------DD2BCB93EFBC0CD33F08294C Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > Submitted-by: "Adi molkho DVP-TA" Fri Jul 10 05:11:22 1998 > > vxWorks users > We have wrote several drivers for vxWorks 5.2 and 5.3 and we are now > starting to develop > in 5.3.1 . > We understand that there are some differences in the tcp/ip stack of this > new version > Does anyone know if this changes might influence our drivers (interfaces > etc..) > Thanks > Adi Hi! There are 3 (Three) different network driver models in the VxWorks 5.3.1. 1. The "regular" VxWorks 5.3.1 implements the BSD4.3 network driver model. As far, as I know it's the same, as in the previous versions, so if You use the mentioned "regular" VxWorks 5.3.1, no changes seem needed. The new SENS networking stack for VxWorks adds 2 more driver models: 2. BSD4.4 model. It's need a relatively minor changes to the existent BSD4.3 drivers. I have some (abeit little) experience in that, and it's not a big deal. 3. END/MUX model. This is a new model, that's very different from the BSD models, promising some appealing features. Porting to the END/MUX model seem to bring a substantial re-writing of the existing driver code, but, hopefully it isn't necessary if You doesn't need the advanced features, this model brings. You may choose any model for each driver in Yours system (the BSD4.3 and END/MUX live peacefully together). For more details, look at the the VxWorks documentation book: SENS for Tornado 1.0 Component Release Supplement chapter Upgrading 4.3BSD network drivers. Hope, this help Ilia You may pick one of them, drivers --------------DD2BCB93EFBC0CD33F08294C Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Ilia Bosis Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Ilia Bosis n: Bosis;Ilia org: Telegate adr: Haplada 7;;;Or-Yehuda;;60218;Israel email;internet: iliab@telegate.co.il title: S/W engineer tel;work: 972-3-5388-685 tel;fax: 972-3-5335877 x-mozilla-cpt: ;0 x-mozilla-html: TRUE version: 2.1 end: vcard --------------DD2BCB93EFBC0CD33F08294C-- From Thomas.Arand@oen.siemens.de Mon Jul 13 05:45:22 1998 From: Arand Thomas Date: Mon Jul 13 05:45:26 PDT 1998 Subject: Memory leaks and similar problems Hi VxWorks-World, does anyone know about a tool which checks the memory heap for various violations (memory leaks, array boundary, ...). I'm looking for another tool than CodeTest. Thanks and Ciao, Thomas ---------------------------------------------------------------------- Thomas Arand Siemens AG Tel. +49 - 89 - 722 42296 Fax. +49 - 89 - 722 26572 thomas.arand@oen.siemens.de From Thomas.Arand@oen.siemens.de Mon Jul 13 05:48:48 1998 From: Arand Thomas Date: Mon Jul 13 05:48:51 PDT 1998 Subject: disk mirroring Hi VxWorks world, I'm trying to implement a disk mirroring process on VxWorks-platforms: Two platforms are connected via a ethernet interface and have a file system on Flash. One of them (slave) should always be a mirror of the other (master). Does anyone have similar problems? Does anyone have a solution? Help is greatly appreciated! Thanks and Ciao, Thomas ---------------------------------------------------------------------- Thomas Arand Siemens AG Tel. +49 - 89 - 722 42296 Fax. +49 - 89 - 722 26572 thomas.arand@oen.siemens.de From cgrames@mdc.com Mon Jul 13 06:57:02 1998 From: Charlie Grames Date: Mon Jul 13 06:57:06 PDT 1998 Subject: comp.os.vxworks newsdigest -Reply It depends on whether you are loading from WindSh or from the target shell. With WindSh, each load of a same-named module causes the previous copy to be unloaded. With the target shell, you have to unload the previous version manually with unld() or reld(). Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/11/98 06:05am >>> Submitted-by daemon@csg.lbl.gov Sat Jul 11 04:05:22 1998 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 11 04:05:20 PDT 1998 Subject: Re: how to unload *.o ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: how to unload *.o ? Date: Fri, 10 Jul 1998 20:42:44 GMT From: bruce@CUTMEOUTscimisys.com (Bruce Wedding) Organization: NeoSoft, Inc. Message-ID: <35a67c72.10177265@news.neosoft.com> References: <35A50574.4656@crhc.uiuc.edu> <35A640BB.B59F7017@cadence.com> On Fri, 10 Jul 1998 09:26:36 -0700, James Short wrote: > unld xxx.o will work but if you just want to load an updated version of >xxx.o, the old verison is automatically unloaded by doing another ld > Submitted-by Thomas.Arand@oen.siemens.de Mon Jul 13 05:45:22 1998 > Submitted-by: Arand Thomas > > Hi VxWorks-World, > > does anyone know about a tool which checks the memory heap for various > violations (memory leaks, array boundary, ...). > I'm looking for another tool than CodeTest. > > Thanks and Ciao, > > Thomas RTI includes a utility in its RTILib product called HeapTrace. HeapTrace replaces the malloc, calloc, dealloc, and free calls with routines that manage data structures that allow HeapTrace to keep track of memory usage. It can report where (by function name) a block was allocated, how many have been allocated and not freed, etc. This helps you track down memory leaks. HeapCheck helps you find memory problems that corrupt the heap. These can include writing off the end of an array, using an orphan pointer, etc. RTILib is shipped along with StethoScope. If you have purchased StethoScope, then you already have RTILib. There are more details on our website. If you have any questions, please let me know. Regards, -- Brett ============================================================================== = = = = Brett Murphy = Product Manager = = Real-Time Innovations, Inc. = email: brett@rti.com = = 155A Moffett Park Drive, Suite 111 = Phone: (408) 720-8312 = = Sunnyvale, CA 94089 = Fax: (408) 734-5009 = = = = ============================================================================== From tong@nortel.ca Mon Jul 13 10:40:20 1998 From: "Tong Luo" Date: Mon Jul 13 10:40:24 PDT 1998 Subject: Porting from BSD to VxWorks Hi, We are porting our system from BSD to VxWorks. I am wondering whether there is anybody out there has done the same experience before. If you do, I really like to get in touch with you to learn your experience and get some tips. My e-mail address: tong@nortel.ca Thanks in advance. Tong From Scott.Peters@AGAI.com Mon Jul 13 12:18:32 1998 From: Scott Peters Date: Mon Jul 13 12:18:35 PDT 1998 Subject: Looking for a A to D device with VxWorks enviornment Hello vx-ers: Can some one point me in the right direction for a high resolution A to D device, that maybe uses vxworks as a development environment? Please let me know of a company, catalog or website. We need a least 16-bit resolution and 30 - 50 khz sampling sampling rate. We will need to do some accumulations on the device and then interrupt another host VxWorks board with the accumulations. Thanks Scott Peters AG Associates From msaunders@taz.dera.gov.uk Tue Jul 14 01:51:10 1998 From: "Land Systems 4" Date: Tue Jul 14 01:51:14 PDT 1998 Subject: Re: VxWorks Target Server Connect Problems vxWorks VxWorks Sometimes WDB won't attach if your target is thrashing, have you added application code to VxWorks which is thrashing the CPU. Pete Gardiner DERA (UK) ls4@taz.dra.hmg.gb From STH@pentland.co.uk Tue Jul 14 03:03:32 1998 From: "Steven Hardie" Date: Tue Jul 14 03:03:35 PDT 1998 Subject: Re: Looking for a A to D device with VxWorks enviornment check out our website for various solutions to your requirements. www.pentlandsys.com or email me directly and I'll put you in touch with our sales / applications people who'll be able to help you. Best Regards Steven Hardie Pentland Systems Ltd Voice +44(0) 1506 464666 Fax +44(0) 1506 463030 Steven.Hardie@pentland.co.uk Steven Hardie Pentland Systems Ltd 1 Cochrane Square Brucefield Industrial Park Livingston West Lothian EH54 9DR tel +44 1506 464666 fax +44 1506 463030 email sth@pentland.co.uk From chuckstack@yahoo.com Tue Jul 14 07:13:06 1998 From: Charles Stack Date: Tue Jul 14 07:13:09 PDT 1998 Subject: web servers for VxWorks Hi everyone, I'm investigating alternatives for web servers for VxWorks. I'd appreciate any comments or recommendations from those experienced with using commercial versions for VxWorks as well as those who have attempted and/or successfully ported freeware versions (Apache, etc.). Thanks much. Chuck Stack ChuckStack@yahoo.com _________________________________________________________ DO YOU YAHOO!? Get your free @yahoo.com address at http://mail.yahoo.com From McGarry_Mike@timeplex.com Wed Jul 15 07:15:19 1998 From: "Mike McGarry" Date: Wed Jul 15 07:15:23 PDT 1998 Subject: Target Shell Problem This is a multi-part message in MIME format. --------------05D3686CAACA844D6C54AAFA Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello, fellow VxWorks users: I was wondering if anyone out there has had any problems running apps on their target that accept character input from the Target Shell. I am trying to run ScopeProfile from RTI through my Target Shell and there is a routine called ProfilePrompt that accepts input from the Target Shell. Once the routine tries to accept input, the task it spawned gets SUSPENDED. I have redirected stdin, stdout, and stderr to "/vio/0". So, in theory, it should work. BTW, the problem also occurs if I just call "getchar()" from the Target Shell. So, the problem is not related to ScopeProfile. Here are the routines I am using to tie stdin, stdout, and stderr to "/vio/0": # set virtual I/O vf0 = open("/vio/0",2,0) ioGlobalStdSet 0,vf0 ioGlobalStdSet 1,vf0 ioGlobalStdSet 2,vf0 Any help would be very much appreciated! Thank you, --------------05D3686CAACA844D6C54AAFA Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Michael McGarry Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Michael McGarry n: McGarry;Michael org: TimePlex Group adr: 400 Chestnut Ridge Road;;;Woodcliff Lake;NJ;07675;USA email;internet: mcgarry_mike@timeplex.com title: Engineer II tel;work: (201) 391-1111 x4693 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------05D3686CAACA844D6C54AAFA-- From jshaw_shade@compuserve.com Wed Jul 15 09:07:22 1998 From: John Shaw Date: Wed Jul 15 09:07:27 PDT 1998 Subject: Microsoft Windows DDE support in VxWorks? Hello, Does anyone know if VxWorks has any native support for Microsoft Windows Dynamic Data Exchange protocol across Ethernet (used by Microsoft Office via NET-DDE)? If there is no native support (very probable), are there any third party software packages that could be run under VxWorks and provide that support? John Shaw, Shade (Computer Services) Ltd., Calne, Wiltshire, England, UK From Richard.Boyd@COMPAQ.com Wed Jul 15 10:56:55 1998 From: "Boyd, Richard (HW Stor)" Date: Wed Jul 15 10:57:02 PDT 1998 Subject: RE: Target Shell Problem I had the same problem. This is the solution that I got from Wind River. It works, but now I have to wrap all of my stdin functions to suspend the shell when I want input. Richard Boyd richard.boyd@compaq.com /*getc.c - demonstrates the usage of getting standard input from the keyboard*/ /* DESCRIPTION The Shell is running at a higher priority than the application tasks. Hence serial input goes to the shell first. One possible solution to get the serial input to your application task is to suspend the shell while you're running your task and then resume the shell after your tasks get the required standard input. This program uses this solution to demonstrate the usage of getting standard input from the keyboard. To run this program from the VxWorks shell do as follows: -> ld < getc.o -> stdinDemo */ #include "vxWorks.h" #include "taskLib.h" /*************************************************************************** * taskStdinGet - Gets the required standard input from the keyboard and * then resume the shell task. */ void taskStdinGet (void) { int ch = 0; int status = OK; int num_of_space = 0; /* Get the required standard input from the keyboard */ printf ("\nThis demo program computes the number of spaces in a sentence\n"); printf ("Enter a sentence here followed by key:"); ch = getchar(); while (ch != '\n') { if (ch == ' ') num_of_space++; ch = getchar (); } printf ("The number of space(s) in your sentence is: %d\n",num_of_space); /* Resume the shell task */ status = taskResume (taskIdFigure ("tShell")); if (status == ERROR) perror ("Problem in resuming tShell task "); } /*************************************************************************** * stdinDemo - Spawns getStdinTask and suspends the shell task to get stdin * from the key board. */ void stdinDemo () { int status; status = taskSpawn ( "stdinGetTask", 100, VX_SUPERVISOR_MODE | VX_STDIO, 20000, (FUNCPTR) taskStdinGet, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0); if ( status == ERROR) perror ("Spawning taskStdinGet task failed"); status = taskSuspend (taskIdFigure ("tShell")); if (status == ERROR) perror ("Suspending the Shell task failed"); } /*** END OF FILE ***/ From msaunders@taz.dera.gov.uk Thu Jul 16 02:37:50 1998 From: "Land Systems 4" Date: Thu Jul 16 02:37:54 PDT 1998 Subject: Re: Target Shell Problem vxWorks VxWorks the task which uses getchar() can redirect stdin to another tty device, this will allow the target based shell to continue as normal while you interact with the task via a second terminal. ioTaskStdSet() Pete Gardiner DERA Chertsey (UK) ls4@taz.dra.hmg.gb From MSchrape@atomika.com Thu Jul 16 03:44:28 1998 From: Martin Schrape Date: Thu Jul 16 03:44:33 PDT 1998 Subject: Re: Target Shell Problem Hi Michael, some month ago the WRS support told me to use INCLUDE_CONFIGURATION_5_2, but now I'm connecting to my interactive program via telnet (but not port 23) with a modified telnet deamon. The most important function is telnetServiceAdd: /*************************************************************************** **** * * telnetServiceAdd - add telnet service * * This routine spawn the new telnet demon to service telnet connection requests * arrives to the port. It takes the following arguments: * .IP * a string constant that specifies the name of the desired service (don't use a * variable value); * .IP * the protocol port number assigned to the service; * .IP * priority of the telnet demon ; * .IP * a pointer to the user supplied routine which connect the command interpreter * with the pty device created by the telent demon to service the telent * connection. * .IP * an argument that telnet demon passes to the interpConn(); * .IP * a pointer to the user supplied routine which disconnect the command * interpreter from the pty device. * * RETURNS: OK or ERROR, if the service can't be added. */ STATUS telnetServiceAdd ( char * serviceName, /* name of the service */ int serviceNum, /* port number of the service */ int servicePriority, /* priority of telnet demon */ INTERP_CON_PROC interpConn, /* connects comm. interp. with telnet */ int interpConnArg, /* argument for interpConn () */ INTERP_DIS_PROC interpDisconn /* disconnects comm. interp. from telnet */ ) -martin At 07:15 7/15/98 PDT, you wrote: >Submitted-by McGarry_Mike@timeplex.com Wed Jul 15 07:15:19 1998 >Submitted-by: "Mike McGarry" > >This is a multi-part message in MIME format. >--------------05D3686CAACA844D6C54AAFA >Content-Type: text/plain; charset=us-ascii >Content-Transfer-Encoding: 7bit > >Hello, fellow VxWorks users: > >I was wondering if anyone out there has had any problems running apps on >their target that accept character input from the Target Shell. > >I am trying to run ScopeProfile from RTI through my Target Shell and >there is a routine called ProfilePrompt that accepts input from the >Target Shell. Once the routine tries to accept input, the task it >spawned gets SUSPENDED. I have redirected stdin, stdout, and stderr to >"/vio/0". So, in theory, it should work. > >BTW, the problem also occurs if I just call "getchar()" from the Target >Shell. So, the problem is not related to ScopeProfile. > >Here are the routines I am using to tie stdin, stdout, and stderr to >"/vio/0": > ># set virtual I/O >vf0 = open("/vio/0",2,0) >ioGlobalStdSet 0,vf0 >ioGlobalStdSet 1,vf0 >ioGlobalStdSet 2,vf0 > >Any help would be very much appreciated! > >Thank you, > > >--------------05D3686CAACA844D6C54AAFA >Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" >Content-Transfer-Encoding: 7bit >Content-Description: Card for Michael McGarry >Content-Disposition: attachment; filename="vcard.vcf" > >begin: vcard >fn: Michael McGarry >n: McGarry;Michael >org: TimePlex Group >adr: 400 Chestnut Ridge Road;;;Woodcliff Lake;NJ;07675;USA >email;internet: mcgarry_mike@timeplex.com >title: Engineer II >tel;work: (201) 391-1111 x4693 >x-mozilla-cpt: ;0 >x-mozilla-html: FALSE >version: 2.1 >end: vcard > > >--------------05D3686CAACA844D6C54AAFA-- From kimberly@spk.hp.com Thu Jul 16 12:24:20 1998 From: Kim Middleton Date: Thu Jul 16 12:24:24 PDT 1998 Subject: crossing the 32MB boundary Hi! We recently moved to a board that has 64 MB of memory thinking that would be plenty of memory for our application code to grow. Since then, we've learned that VxWorks cannot support offsets greater than 32 MB, so even though we should have 64 MB, we're limited to 32. Has anyone figured out a way around this? Thanks in advance! Kim -- Kim Middleton (formerly McCulla) Hewlett-Packard Company kimberly@spk.hp.com 24001 E. Mission Ave Voice: (509) 921-3817 Liberty Lake, WA 99019 FAX: (509) 921-3991 From jieming@Brocade.COM Thu Jul 16 18:04:18 1998 From: jieming@Brocade.COM (Jieming Zhu) Date: Thu Jul 16 18:04:23 PDT 1998 Subject: VxWorks 5.3 issue For some reason, after moving to 5.3, "tt taskname" only returns one entry of the stack, instead of the whole stack of entries for the task, which we got used to in 5.2. Any insights? Many Thanks, ------------------------------------------------------------------------ Jieming Zhu jieming@brocade.com Brocade Communications Systems Tel: (408) 487-8113 1901 Guadalupe Parkway Fax: (408) 487-8090 San Jose, CA 95131 http://www.brocade.com ------------------------------------------------------------------------ From Friedrich.Scheurer@t-online.de Fri Jul 17 01:03:46 1998 From: Friedrich.Scheurer@t-online.de (Friedrich Scheurer / Goeppingen) Date: Fri Jul 17 01:03:50 PDT 1998 Subject: Setting up BSP VxWorks 5.3 MVME2700 Hi VxWorkers, i got the problem, that i can't build a BSP for our new MVME2700 target. 1. WRS sent me a new CD, but on the new revision i got once again only the BSP MV2604 though they told me the opposite. At the momenent i don't know whether it's a problem of the license or not. 2. I installed the contents of the new CD once again. After that i created ../target/config-Directory mv2700, in which i copied all files from mv2604. Afterwards i copied all Files from the Motorola Early-Access-Package(Feb. or March 98) to the mv2700-Directory. The following Files have been substituted: zfssun7:windadm:48: cp -rp mv2700.subset.mot/* mv2700.mot2 cp: overwrite mv2700.mot2/Makefile (yes/no)? y cp: overwrite mv2700.mot2/README (yes/no)? y cp: overwrite mv2700.mot2/config.h (yes/no)? y cp: overwrite mv2700.mot2/mv2600.h (yes/no)? y cp: overwrite mv2700.mot2/romInit.s (yes/no)? y cp: overwrite mv2700.mot2/sysALib.s (yes/no)? y cp: overwrite mv2700.mot2/sysCache.c (yes/no)? y cp: overwrite mv2700.mot2/sysLib.c (yes/no)? y cp: overwrite mv2700.mot2/target.nr (yes/no)? y cp: overwrite mv2700.mot2/target.txt (yes/no)? y cp: overwrite mv2700.mot2/universe.c (yes/no)? y cp: overwrite mv2700.mot2/universe.h (yes/no)? y cp: overwrite mv2700.mot2/vxWorks.st (yes/no)? y zfssun7:windadm:49: cd .. Without changing the default-configuration i tried to build the new BSP. Neither with WindCFG nor with command-line i got it. The following error occurs: zfssun7:windadm:75: windmake vxWorks ccppc -B/usr/local/wind/host/sun4-solaris2/lib/gcc-lib/ -mstrict-align -ansi -nostdinc -O2 -fvolatile -fno-builtin -fno-for-scope -Wall -I/h -I. -I/usr/local/wind/target/config/all -I/usr/local/wind/target/h -I/usr/local/wind/target/src/config -I/usr/local/wind/target/src/drv -DCPU=PPC604 -DMV2700 -DTARGET_DIR="\"mv2700\"" -c sysLib.c sysLib.c: In function `sysHwInit': sysLib.c:825: `_func_vxMemProbeHook' undeclared (first use this function) sysLib.c:825: (Each undeclared identifier is reported only once sysLib.c:825: for each function it appears in.) sysLib.c: In function `sysProbeErrClr': sysLib.c:1959: warning: implicit declaration of function `vxDarSet' sysLib.c:1960: warning: implicit declaration of function `vxDsisrSet' sysLib.c:1961: warning: implicit declaration of function `vxSrr0Set' sysLib.c:1962: warning: implicit declaration of function `vxSrr1Set' sysLib.c: In function `sysBusProbe': sysLib.c:2189: warning: implicit declaration of function `vxMemArchProbe' make: *** [sysLib.o] Error 1 zfssun7:windadm:76: grep vxDarSet *.c symTbl.c:IMPORT vxDarSet (); symTbl.c: {{NULL},"vxDarSet", (char*) vxDarSet, 0, N_EXT | N_TEXT}, sysLib.c: vxDarSet (0); zfssun7:windadm:77: grep _func_vxMemProbeHook *.c symTbl.c:IMPORT int _func_vxMemProbeHook; symTbl.c: {{NULL},"_func_vxMemProbeHook", (char*) &_func_vxMemProbeHook, 0, N_EXT | N_DATA}, sysLib.c: _func_vxMemProbeHook = (FUNCPTR)sysBusProbe; zfssun7:windadm:78: grep _func_vxMemProbeHook *.h It seems to be a problem of mis-set Defines in the configuration. Any help appreciated! TIA. Best regards! F. Scheurer -- +-------------------------------------------------+ | | | Dipl.-Ing. F. Scheurer | | | | c/o Zentrum Fertigungstechnik Stuttgart | | Nobelstrasse 15 | | D-70569 Stuttgart | | | | Telefon +49 711 13162-41(-0) | | Fax +49 711 13162-11 | | E-Mail scheurer@zfs.rus.uni-stuttgart.de | | zfssrr@zfssun6.tz.rus.uni-stuttgart.de | | WWW http://zfs.rus.uni-stuttgart.de | | | +-------------------------------------------------+ | | | privat: | | Telefon +49 7161 686362 | | Fax +49 7161 686301 | | E-Mail Friedrich.Scheurer@t-online.de | | | +-------------------------------------------------+ From auccello@asec-md.com Fri Jul 17 08:04:54 1998 From: "Uccello, Adam" Date: Fri Jul 17 08:05:01 PDT 1998 This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------ =_NextPart_001_01BDB194.36B4CB60 Content-Type: text/plain Has anyone had any luck in finding driver support for SCSI devices using the Compact PCI bus? Thanks, -Adam Uccello vxWorks / Tornado ________________________________________________________ Adam P. Uccello ASEC - Columbia, MD v-mail: 410.309.9720 x257 e-mail: auccello@asec-md.com ------ =_NextPart_001_01BDB194.36B4CB60 Content-Type: text/html

Has anyone had any luck in finding driver support for SCSI devices using the Compact PCI bus?

Thanks,
-Adam Uccello

vxWorks / Tornado
________________________________________________________
Adam P. Uccello
ASEC - Columbia, MD
v-mail: 410.309.9720 x257
e-mail: auccello@asec-md.com


------ =_NextPart_001_01BDB194.36B4CB60-- From krueger@comsys.rockwell.com Fri Jul 17 08:55:12 1998 From: Bill Krueger Date: Fri Jul 17 08:55:19 PDT 1998 Subject: Re: crossing the 32MB boundary We run simulations of large wireless radio networks on PCs running our exact software replicated many times. The OS is VxWorks, and I have 384 MB which at one time was almost completely used. The one change I had to make was to comment out LOCAL_MEM_AUTOSIZE in config.h and define LOCAL_MEM_SIZE to be the size of the memory in the machine. Bill > We recently moved to a board that has 64 MB of memory > thinking that would be plenty of memory for our application > code to grow. Since then, we've learned that VxWorks > cannot support offsets greater than 32 MB, so even > though we should have 64 MB, we're limited to 32. Has > anyone figured out a way around this? > > Thanks in advance! > > Kim > From cgrames@mdc.com Fri Jul 17 14:13:22 1998 From: Charlie Grames Date: Fri Jul 17 14:13:26 PDT 1998 Subject: crossing the 32MB boundary -Reply Kim, I'm assuming you are using PowerPC. It's not VxWorks that is causing the 32 MB limit, but the PowerPC architecture and corresponding Embedded ABI. There are two workarounds to the 32 MB problem: 1) Set USER_RESERVED_MEM in config.h to include all memory except 32 MB. After VxWorks boots, load all your objects, and then use memAddToPool() to add the remaining memory to the system memory pool. Unfortunately, this causes sysMemTop() to return that you have only 32 MB of memory configured; we modified the BSP to work around this problem. 2) Use a compiler that provides an option for "far calls." I believe the GNU compilers delivered with Tornado 1.0.1 do not have such an option, but the newer GNU compilers do. We are using the Green Hills compilers, which have a "far calls" option. We are actually using a combination of the two workarounds, because we have some existing objects that were not compiled with the "far calls" option. HTH Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/16/98 02:24pm >>> Submitted-by kimberly@spk.hp.com Thu Jul 16 12:24:20 1998 Submitted-by: Kim Middleton Hi! We recently moved to a board that has 64 MB of memory thinking that would be plenty of memory for our application code to grow. Since then, we've learned that VxWorks cannot support offsets greater than 32 MB, so even though we should have 64 MB, we're limited to 32. Has anyone figured out a way around this? Thanks in advance! Kim From cgrames@mdc.com Fri Jul 17 14:19:08 1998 From: Charlie Grames Date: Fri Jul 17 14:19:11 PDT 1998 Subject: Setting up BSP VxWorks 5.3 MVME2700 -Reply Did you install the PowerPC Collective 1 patch also on the CD? If not, install it, and then recombine the 2700 patch with the 2604 BSP. HTH Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/17/98 03:03am >>> Submitted-by Friedrich.Scheurer@t-online.de Fri Jul 17 01:03:46 1998 Submitted-by: Friedrich.Scheurer@t-online.de (Friedrich Scheurer / Goeppingen) Hi VxWorkers, i got the problem, that i can't build a BSP for our new MVME2700 target. 1. WRS sent me a new CD, but on the new revision i got once again only the BSP MV2604 though they told me the opposite. At the momenent i don't know whether it's a problem of the license or not. 2. I installed the contents of the new CD once again. After that i created ../target/config-Directory mv2700, in which i copied all files from mv2604. Afterwards i copied all Files from the Motorola Early-Access-Package(Feb. or March 98) to the mv2700-Directory. It seems to be a problem of mis-set Defines in the configuration. Any help appreciated! TIA. Best regards! F. Scheurer -- +-------------------------------------------------+ | | | Dipl.-Ing. F. Scheurer | | | | c/o Zentrum Fertigungstechnik Stuttgart | | Nobelstrasse 15 | | D-70569 Stuttgart | | | | Telefon +49 711 13162-41(-0) | | Fax +49 711 13162-11 | | E-Mail scheurer@zfs.rus.uni-stuttgart.de | | zfssrr@zfssun6.tz.rus.uni-stuttgart.de | | WWW http://zfs.rus.uni-stuttgart.de | | | +-------------------------------------------------+ | | | privat: | | Telefon +49 7161 686362 | | Fax +49 7161 686301 | | E-Mail Friedrich.Scheurer@t-online.de | | | +-------------------------------------------------+ From mfischer@qualcomm.com Fri Jul 17 16:00:53 1998 From: Mark Fischer Date: Fri Jul 17 16:01:01 PDT 1998 Subject: Re: crossing the 32MB boundary -Reply According to the vxWorks Users Group Exploder: > I'm assuming you are using PowerPC. It's not VxWorks that is > causing the 32 MB limit, but the PowerPC architecture and > corresponding Embedded ABI. I beg to differ. It IS VxWorks that is causing the problem; they just excuse it with a reference to the EABI. The PowerPC Runtime Architecture specifies that all code is read-only, and position independent. Any references WITHIN A FRAGMENT must be within 24 bits. References outside of that fragment (eg, to other fragments) are resolved at link time by the 'Code Fragment Manager' with reference to the Table of Contents generated for each fragment by the compiler. The existing Cygnus compiler that VxWorks uses does NOT implement the Table of Contents (how could it? it was released in 1996), and their linker performs NO code fragment management. Generated code DOES comply to the EABI, but without the rest of the PowerPC Runtime Architecture, has no way to resolve references beyond 24 bits, hence the problem. Wind River supposedly has a 'patch' that allows 'far calls', but I have yet to see any effort made to comply with the PPC runtime spec. That is my understanding, having spent months trying to nail down the nature of the problem. If I am in error, please feel free to post corrections to the list. Mark ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ps. By the way, LynxOS uses a newer version of the Cygnus compiler that does support the PowerPC runtime spec. Why Wind River does not get a copy, I have no clue. From kkoshimizu@cup.teainet.com Fri Jul 17 17:42:13 1998 From: "TEM Koshimizu, Kaz" Date: Fri Jul 17 17:42:16 PDT 1998 Subject: Drivers for PC87308(vxWorks 5.3.1) I am looking for the device drivers for PC87308 Super I/O chip on MVME1603. Parallel Port Driver and FDC Driver in perticular. We are using VxWorks 5.3.1 on Sun Hosts. Thank you in advance. Kazu Koshimizu Tokyo Electron America. From vyas@teil.soft.net Sat Jul 18 05:21:36 1998 From: Vyasaraj S Date: Sat Jul 18 05:21:40 PDT 1998 Subject: Tornado.. crosswind debugger hi vxworkers, I am new to Tornado development environment and vxWorks. I have troubles with Vxsim Target server. When I am debugging my application, my target server gets killed, and I am not able to attach again my target server to the target. Trying to start the target gives Ulip Init failed. error 0x00 The manual says, that the target was killed with the Ulip device some unknown state.. It recommends the command... ulip stop and ulip start again.. I am interested in knowing whether there is any way to restart one particular ulip device.?? Further, I am facing problems with the crosswind. I am receiving errors like "Tag overflow ". Can anybody help me out to understand why is this happening , and how to overcome.. The vxSim process ( the target ) gets killed when my code branches to bcopy .. Thanks in advance. vyas From mumcu@venus.aselsan.com.tr Sun Jul 19 23:22:54 1998 From: Ahmet Mumcu Date: Sun Jul 19 23:23:06 PDT 1998 Subject: Problem with Shared Memory Networking Hi friends, I have a problem with shared memeory networking. In my configuration a MVME1604 PowerPC board is connected to our network and MVME162 680LC040 board is configured to boot over the VME backplane. I use Tornado 5.3.1 and I use SUN host running Solaris 2.6. MVME162 boots and starts VxWorks and I saw everyhing to be pretty well fom its console. But I cannot ping to the MVME162's IP address on the VME backplane although I can ping MVME1604's both backplane address and ethernet address. Also I noticed that MVME162 seems to be fail to boot sometime. Is this a sign of a timing problem. In fact both boards use the same VME I/F chip. Has anyone else faced with a similar problem using these 2 boards? In our company we use the backplane networking with two similar PowerPC boards without problem. I will be very pleased for your recomendations. Thank You. ---------------------------- Ahmet Mumcu Technical Leader ASELSAN Inc. MST Group/YMM Ankara 06172 TURKEY Tel: 90-312-3851900 ext 2633 Fax: 90-312-3545205 e-mail: mumcu@mst.aselsan.com.tr http://www.aselsan.com.tr ---------------------------- From blohm@vs.dasa.de Mon Jul 20 02:39:46 1998 From: "Gundula Blohm" Date: Mon Jul 20 02:39:50 PDT 1998 Subject: Re: Target Shell Problem On 15 Jul 98 at 7:15, the vxWorks Users Group Explo wrote: > I was wondering if anyone out there has had any problems running apps on > their target that accept character input from the Target Shell. > I did an redirection on the targetserver window. If I type my characters in that window, they are accepted and put trough to the application. Note: the targetserver window is not a child window of my Tornado application. =================================================== Gundula Blohm Tel. +49 (731) 392-4731 Fax +49 (731) 392-4958 mailto:blohm@vs.dasa.de =================================================== From daemon@csg.lbl.gov Mon Jul 20 04:00:34 1998 From: daemon@csg.lbl.gov Date: Mon Jul 20 04:00:37 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 20 04:00:31 PDT 1998 Subject: want register address? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: want register address? Date: Fri, 17 Jul 1998 15:51:37 -0500 From: Lan Chen Organization: University of Illinois at Urbana-Champaign Message-ID: <35AFB959.5DAA@crhc.uiuc.edu> Hi, I want to read/write the registers of the vxworks processor (PowerPC603), maybe they are memory mapped in some address space, and this information should be defined in some header files, but I just cannot find them? Any suggestion regard this is highly appreciated. Lan - -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Dept of Electrical & Computer Engineering Univ of Illinois at Urbana-Champaign Phone: (217)344-4728 (H) (217)244-8288 (O) Email: lanchen1@students.uiuc.edu ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --------------------------- End of New-News digest ********************** From cgrames@mdc.com Mon Jul 20 07:51:54 1998 From: Charlie Grames Date: Mon Jul 20 07:51:57 PDT 1998 Subject: Re: crossing the 32MB boundary -Reply -Reply In a previous note, I wrote: > I'm assuming you are using PowerPC. It's not VxWorks that is > causing the 32 MB limit, but the PowerPC architecture and > corresponding Embedded ABI. Mark Fischer responded: >I beg to differ. > >It IS VxWorks that is causing the problem; they just excuse it with >a reference to the EABI. Much as I hate to admit it, VxWorks does comply with the EABI. In chapter 3, "Low-Level System Information", in the "SVR4 ABI Execution Support" section clearly states: " A conforming entity, such as an application or a static linker, shall not have requirements pertaining to: o dynamic linking o global offset tables o procedure linkage tables o shared objects" Admittedly, this is for BASE conformance. It goes on to say that EXTENDED conformance requires implementation of the mechanisms specified in the SVR4 ABI (UNIX System V Release 4 Application Binary Interface, including the PowerPC supplement). Unfortunately, VxWorks makes no claim to extended conformance. I understand the annoyance generated by this problem (BELIEVE me, I do), and I do agree wholeheartedly that Wind River should provide SOME means to work around this problem (other than limiting memory size to 32 MB). However, I don't know that it is reasonable to expect them to adopt the SVR4 ABI philosophy in this matter. Anyone who believes otherwise, feel free to blast me. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com From James_M_Flynn@res.raytheon.com Mon Jul 20 14:23:23 1998 From: James_M_Flynn@res.raytheon.com Date: Mon Jul 20 14:23:26 PDT 1998 Subject: ROMBoot of vxWorks Boot Kernel from MV2604 Could anyone save me some heartache and tell me how to boot the vxWorks boot kernel from the PPCBug using ROMBoot on the MV2604? I have both banks of flash enabled, with the PPCBug in bank A and the vxWorks boot in bank B. Using the external jumper, I can switch between PPCBug and vxWorks boot. However, when I try to use the ROMBoot (adding the proper 'BOOT" header and checksum, etc), the processor never returns after passing control from the PPCBug. Any help would be appreciated. Jim From altan@venus.aselsan.com.tr Tue Jul 21 02:08:45 1998 From: Cagatay ALTAN Date: Tue Jul 21 02:08:55 PDT 1998 Subject: bootrom and bootrom.hex files Dear colleagues, I have a problem with the object files that VxWorks produce. Here is the situation: I boot from the flash on the mvme162. There is a boot code on the flash, currently; that is I boot from the flash and dynamically load the object code to the flash from the vxWorks shell. I have a code in that object file that erase/program the flash. After I boot up the system, I want to program the bootrom file to the flash so that I test if I can succesfully program the flash. But programming the flash with the "bootrom" file does not help. I compared the code that is currently on the flash(which was formed by programming the "bootrom.hex" file onto the flash by the pflash utility of the 162bug) with the "bootrom" file which, I assume, is the binary form of the "bootrom.hex" file, as I understand from the make file. But the hex utility that converts the "bootrom" file to "bootrom.hex" file during the make process does something more. I realised that, there is an 8-byte extra code in the flash now(which is "0002 0000 ff80 0008", supposedly the entry address). And there is an extra 32-byte code(which is "0102 0107 0000 06EC 0002 0178 0001 B12C 0000 02A0 0002 0000 0000 0000 0000 0000") at the beginning of the "bootrom" file. If a ignore the 8-byte extra in the flash and the 32-byte extra from the "bootrom" file I want to program onto the flash, the rest of the code seems to be the same, with an exception of 1256 bytes extra at the very end of the "bootrom" file, which seems to be some text such as: "HromInit.o.cold.warm.cpu25mhz.cpu33mhz.sramHigh.eccDRAM.noECC. readRam.start.bootInit.o.gcc2_compiled.._bcopyLongs._bfillLongs. version.o.gcc2_compiled..copyright.o.gcc2_compiled..uncompress.o. gcc2_compiled.._stack.6._binEnd._binArray._offset._size._block_compress ._fsize._clear_flg._maxbits._maxmaxcode._rmask._hsize._n_bits._maxcode. _tab_suffix._codetab._free_e" What must I perform to convert the "bootrom" file to a file that I can write on to the flash and it works? As another solution, "bootrom.hex" can be converted to object code before programming onto the flash as pflash utility does, but I do not have the code to perform it. Can anyone help me in either of the approaches(or perhaps in another approach I missed.) Thanks in advance. Tevfik Cagatay ALTAN, BScEE mail: altan@mst.aselsan.com.tr ***************************************************************** * "No questions I'm not alone, somehow I'll find my way home." * ***************************************************************** Microwave and Systems Technologies / Tel: (90 312) 3851900 / 2940 Software Engineering Dept. Fax: (90 312) 3545205 ASELSAN web: www.aselsan.com.tr From daemon@csg.lbl.gov Tue Jul 21 04:01:08 1998 From: daemon@csg.lbl.gov Date: Tue Jul 21 04:01:12 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 21 04:01:05 PDT 1998 Subject: Re: RTTI support ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: RTTI support Date: Tue, 21 Jul 1998 11:26:19 +0200 From: Fredrik Jonsson Organization: Ericsson Message-ID: <35B45EBB.676B6DE5@ericsson.com> References: <35AFB8D3.9786BD86@j-works.com> <35B3A52C.8B505348@aplcomm.jhuapl.edu> Fran Horan wrote: > Jerry J. Mikos wrote: > > > Does anyone know if GNU version 2.7 support RTTI > > Thanks > > No. > See http://www.wrs.com/csdocs/SortedFAQs/Main.html for > other unsupported features. (you need an account to login) Even though it IS unsupported, it is still possible to test the "beta" functionality of RTTI with GNU 2.7.2 on VxWorks (Tornado 1.0.1) Compile with the -frtti option. You also have to bring in typeinfo from another GNU 2.7.2 distribution, and strip that header file from all exceptions. (You only need the type_id struct and it's inline implementation) That's it. I managed to do a dynamic_cast<> on a VxWorks 68MC360 target with that hack! :) Still, you couldn't use it in practical. RTTI requires that all code in the process is compiled with RTTI support to work. Iostream, Tool.h++, WFC are all compiled without RTTI and without exception handling compilation flags. Therefor it makes little sense bringing in those language features on your own. Also, RTTI in GNU 2.7.2 is very incomplete according to FAQs and release notes. What about the jungledrum? Has anybody heard rumours on upcoming Tornado 2.0 releases? - -- Fredrik Jonsson --------------------------- End of New-News digest ********************** From ptp@mclean.sparta.com Tue Jul 21 08:44:37 1998 From: "Patrick T. Pinkowski" Date: Tue Jul 21 08:44:41 PDT 1998 Subject: X86 VxWorks SCSI Controller Device Drivers I am interested in finding a VxWorks Device Driver for Adaptec or Symbios SCSI controllers for an x86 platform. TIA Pat -- -------------------------------------------------- /\ Patrick T. Pinkowski, Senior Engineer /##\ ptp@mclean.sparta.com /####\ 703.448.1683 x228 /####/-- /####/---- SPARTA, Inc. -\####\__--- 7926 Jones Branch Drive, Suite 900 ---\##\ /---- McLean, Virginia 22102 ----\##\/\---- (703)448-0210 (Main) ----\####\-- (703)893-5494 (Facsimile) ----\####/ http://www.mclean.sparta.com ---/###/ -/###/ SPARTA \##/ ~~~~~~ \/ Pride In Performance From matthew.brown@lmco.com Tue Jul 21 10:44:45 1998 From: "Brown, Matthew" Date: Tue Jul 21 10:44:48 PDT 1998 Subject: Deleting VxWorks SCSI_BLK_DEVs Does anyone know how to delete a SCSI_BLK_DEV? We are running VxWorks 5.3 with a PowerPC and have two Magneto Optical drive running over a SCSI interface. We would like to be able to delete the SCSI_PHYS_DEV when the MO is removed from the drive. scsiPhysDevDelete () will not allow us to do so, if any SCSI_BLK_DEVs have been created on the physical device. I haven't seen any function calls in scsiLib that will delete SCSI_BLK_DEVs. Thanks -Matt ------------------------------------------- * Matt Brown - Software Engineer * * Lockheed Martin Naval Launching Systems * ------------------------------------------- ------------------------------------------- * Matt Brown - Software Engineer * * Lockheed Martin Naval Launching Systems * * 410-682-0823 * ------------------------------------------- From cgrames@mdc.com Tue Jul 21 10:47:43 1998 From: Charlie Grames Date: Tue Jul 21 10:47:54 PDT 1998 Subject: bootrom and bootrom.hex files -Reply I'm somewhat confused by your description. .hex files are typically in Motorola S-record format and are interspersed with control codes and checksums for ROM burners. The original "bootrom" file you referred to should be in the native object format for the MVME162 (I don't know whether the object format is a.out, COFF, or ELF). Neither is suitable (or so I thought) for direct loading into flash with the PFLASH command. Instead, you need to use the elfToBin utility to create a pure binary. This is done from the command line with the command elfToBin boot.bin where boot.bin is whatever name you choose for your binary. This binary can then be loaded into flash directly. Again, I don't claim expertise with the MVME162 platform. If I'm wrong, someone more knowledgeable please correct me. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/21/98 04:08am >>> Submitted-by altan@venus.aselsan.com.tr Tue Jul 21 02:08:45 1998 Submitted-by: Cagatay ALTAN Dear colleagues, I have a problem with the object files that VxWorks produce. Here is the situation: I boot from the flash on the mvme162. There is a boot code on the flash, currently; that is I boot from the flash and dynamically load the object code to the flash from the vxWorks shell. I have a code in that object file that erase/program the flash. After I boot up the system, I want to program the bootrom file to the flash so that I test if I can succesfully program the flash. But programming the flash with the "bootrom" file does not help. I compared the code that is currently on the flash(which was formed by programming the "bootrom.hex" file onto the flash by the pflash utility of the 162bug) with the "bootrom" file which, I assume, is the binary form of the "bootrom.hex" file, as I understand from the make file. But the hex utility that converts the "bootrom" file to "bootrom.hex" file during the make process does something more. I realised that, there is an 8-byte extra code in the flash now(which is "0002 0000 ff80 0008", supposedly the entry address). And there is an extra 32-byte code(which is "0102 0107 0000 06EC 0002 0178 0001 B12C 0000 02A0 0002 0000 0000 0000 0000 0000") at the beginning of the "bootrom" file. If a ignore the 8-byte extra in the flash and the 32-byte extra from the "bootrom" file I want to program onto the flash, the rest of the code seems to be the same, with an exception of 1256 bytes extra at the very end of the "bootrom" file, which seems to be some text such as: "HromInit.o.cold.warm.cpu25mhz.cpu33mhz.sramHigh.eccDRAM.noECC. readRam.start.bootInit.o.gcc2_compiled.._bcopyLongs._bfillLongs. version.o.gcc2_compiled..copyright.o.gcc2_compiled..uncompress.o. gcc2_compiled.._stack.6._binEnd._binArray._offset._size._block_compress ._fsize._clear_flg._maxbits._maxmaxcode._rmask._hsize._n_bits._maxcode. _tab_suffix._codetab._free_e" What must I perform to convert the "bootrom" file to a file that I can write on to the flash and it works? As another solution, "bootrom.hex" can be converted to object code before programming onto the flash as pflash utility does, but I do not have the code to perform it. Can anyone help me in either of the approaches(or perhaps in another approach I missed.) Thanks in advance. Tevfik Cagatay ALTAN, BScEE mail: altan@mst.aselsan.com.tr ***************************************************************** * "No questions I'm not alone, somehow I'll find my way home." * ***************************************************************** Microwave and Systems Technologies / Tel: (90 312) 3851900 / 2940 Software Engineering Dept. Fax: (90 312) 3545205 ASELSAN web: www.aselsan.com.tr From norgren@mate01.nc.tekelec.com Wed Jul 22 05:50:45 1998 From: norgren@mate01.nc.tekelec.com (Eric Norgren) Date: Wed Jul 22 05:50:49 PDT 1998 Subject: intel 28F008SC SmartVoltage FlashFile Memory driver Hello, I am searching for a vxWorks driver to support the intel 28F004SC/ 28F008SC/28F016SC byte-wide smartVoltage flashFile memory family devices. If anyone has a driver they would be willing to share I would be most appreciative. Also, any suggestions as to where I might obtain this vxWorks driver would be most helpful? Thank you for your help, Eric Norgren Tekelec Software Engineer email: eric.norgren@tekelec.com voice: (919)380-3813 From jhend@primenet.com Wed Jul 22 20:24:08 1998 From: JRH Date: Wed Jul 22 20:24:12 PDT 1998 Subject: Disabling LAN CRC error messages To all, Using a 2603 board w/ Tornado 1.0 and an NT host we occassionally see a CRC error displayed on the serial command shell during heavy traffic on the LAN that is shared with the above system. The overall throughput of the LAN during this time is acceptable with respect to the execution of the application on the 2603. I JUST WANT TO DISABLE THE WARNING. How do I do this? Thank you in advance, John R Henderson S/W Guy From mfischer@qualcomm.com Wed Jul 22 21:43:05 1998 From: Mark Fischer Date: Wed Jul 22 21:43:09 PDT 1998 Subject: Re: Disabling LAN CRC error messages Dear list, Recently someone posted a response concerning the vxworks eabi issue. I seem to have lost it, but I am interested in the contents. Would the poster please bounce me a copy? Thanks, Mark Fischer Qualcomm, Inc. From daemon@csg.lbl.gov Thu Jul 23 04:03:15 1998 From: daemon@csg.lbl.gov Date: Thu Jul 23 04:03:18 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 23 04:02:17 PDT 1998 Subject: Re: Problem with Shared Memory Networking Subject: Re: Vxworks kernel NOT downloading symbol table ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem with Shared Memory Networking Date: Tue, 21 Jul 1998 16:47:21 GMT From: fudath@ct.essd.northgrum.com (Thomas Fuda) Organization: Northrop Grumman Corporation, Baltimore, MD Message-ID: <35b4c48c.24363845@155.104.79.99> References: <35B2E268.9E36A5D@mst.aselsan.com.tr> Sender: news%mother@mdnews.md.essd.northgrum.com On Mon, 20 Jul 1998 09:23:37 +0300, Ahmet Mumcu wrote: >Hi friends, > >I have a problem with shared memeory networking. > >In my configuration a MVME1604 PowerPC board is connected to our network >and MVME162 680LC040 board is configured to boot over the VME backplane. > >I use Tornado 5.3.1 and I use SUN host running Solaris 2.6. > >MVME162 boots and starts VxWorks and I saw everyhing to be pretty well >fom its console. But I cannot ping to the MVME162's IP address on the >VME backplane although I can ping MVME1604's both backplane address and >ethernet address. > >Also I noticed that MVME162 seems to be fail to boot sometime. Is this >a sign of a timing problem. In fact both boards use the same VME I/F >chip. > >Has anyone else faced with a similar problem using these 2 boards? In >our company we use the backplane networking with two similar PowerPC >boards without problem. > >I will be very pleased for your recomendations. > >Thank You. > >---------------------------- > Ahmet Mumcu > Technical Leader > ASELSAN Inc. > MST Group/YMM > Ankara 06172 > TURKEY > > Tel: 90-312-3851900 ext 2633 > Fax: 90-312-3545205 > e-mail: mumcu@mst.aselsan.com.tr > http://www.aselsan.com.tr >---------------------------- > You mentioned that you are unable to ping the MVME162 board on the backplane. Do you mean from the Sun host or the MVME1604? If you are trying to ping the MVME162 from the Sun host you probably have to add a route to both the Sun host and the MVME162's routing table to define the gateway that routes from the ethernet to the VME bus. In this case, your shared memory master (the MVME1604) is the gateway. On the MVME162 board you would do this with the "routeAdd" command. On the Sun host you would use the "route" command. Maybe you've already done this, in which case I don't really know what is wrong. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Vxworks kernel NOT downloading symbol table Date: Thu, 23 Jul 1998 17:03:33 +0900 From: "@S 4vAV" Organization: Hanyang Univ. InterNetNews site Message-ID: <6p6qbm$9cd$1@news.hanyang.ac.kr> References: <35B3994A.AB77DDF7@ti.com> Hi. I had met problem like yours... I have Heurikon Laguna board, VxWorks ver. 5.2, kernel(Wind Version 2.4... My problem was following.. ========================================================== Attaching network interface ei0... done. Attaching network interface lo0... done. Loading... 643954 + 62680 Starting at 0x80010000... NCR 53c720/825 SCSI-II driver version 02b - Copyright 1994-1995 Heurikon Corp. Auto-configuring SCSI bus... ID LUN VendorID ProductID Rev. Type Blocks BlkSize pScsiPhysDev - -- --- -------- ---------------- ---- ---- -------- ------- ------------ Attaching network interface ei0... done. Attaching network interface lo0... done. NFS client support not included. Loading symbol table from charon:/e/vxWorks.sym ... Error opening charon:/e/vxWorks.sym: status = 0x1a9 =========================================================== I solved this problem by installing ftp server("WAR-FTPD") in my PC , installing "SAMBA" into the Workstation installed VxWorks , copying vxWorks and vxWorks.sym to my PC and booting from my PC In my view, there is a problem in ftp server... I want that my experience helps you... Good Luck.. Doug Lewis @L(0!) <35B3994A.AB77DDF7@ti.com> 8^=CAv?!<- @[<:GO?4=@4O4Y... >I have a version of the VxWorks kernel that our customer supplied us >that the kernel does not download the symbol table file from the host. >I've checked the BSP configuration and the neccessary symbol tables >INCLUDE's are in the BSP. I've even compared the configuration to our >own versions of the kernel and they match. Any ideas about how I can >get the kernel to download the symbol table? > >-- >Doug Lewis ---------------------------------- >Senior Software Eng - The views expressed are my own & >Raytheon Systems Co. - do not necessarily reflect those > - of Raytheon Systems Company > ---------------------------------- > > --------------------------- End of New-News digest ********************** From campbell@appliedmicro.ns.ca Thu Jul 23 05:17:29 1998 From: "John Campbell" Date: Thu Jul 23 05:17:32 PDT 1998 Subject: RE: Parallel Port Interrupt handler for VxWorks From jbhawth@aptec.com Thu Jul 23 06:55:51 1998 From: "James B. Hawthorne" Date: Thu Jul 23 06:55:55 PDT 1998 Subject: ATTENTION: Realtime Systems Engineers Wanted!!! VxWorkers- Applied Technology Associates (ATA), a 23 year-old company leading research and development in the areas of data systems, optics and imaging, and control system/stabilization technologies is looking for 2-3 Realtime Systems Engineers. Responsibilities will include realtime computing, including requirements definition, specification, integration, testing, and maintenance of hardware and software systems used for data acquisition, storage, display, and peripheral control. Qualifications include BSEE, BSCE, or BSCS plus 5 years experience, or MS plus 3 years experience and knowledge of VME, PC, C/C++, Fortran, and realtime programming in VxWorks RTOS. Experience with network routines using TCP/IP for task to task communication, writing multi-tasking or parallel processor applications and the ability to write VxWorks device drivers are a plus. The successful candidate will work as part of a dedicated team here in Albuquerque on a project for the High Energy Laser Systems Test Facility located at the White Sands Missle Range in southern New Mexico. Selected candidates may be subject to a government security investigation and should meet eligibility requirements for access to classified information. ATA offers a competitive salary and benefits package as well as an incomparable environment for personal and professional growth. Qualified individuals should send their resumes to Tim Tamerler (tamerler@aptec.com) or myself (jbhawth@aptec.com) ASAP. Jim Hawthorne Realtime Systems Engineer Applied Technology Associates 1900 Randolph Road SE Albuquerque, New Mexico 87106 Phone: (505)767-1228 Fax: (505)768-1379 Be sure to visit our web page at: http://www.aptec.com From dencker@aonix.de Thu Jul 23 08:05:41 1998 From: Peter Dencker Date: Thu Jul 23 08:05:44 PDT 1998 Subject: PowerPC Multiprocessor application with VxWorks Has anybody written down his/her experience with a PowerPC Multiprocessor application using VxWorks? Or do you have a description of an example? If you reply, please copy to my e-mail address as well, as I am not yet subscribed to this list. Thank you! Dr. Peter Dencker, email: dencker@aonix.de Aonix GmbH, phone: +49(721)98653-22, mobile: +49(172)9509944 Kleinoberfeld 7, fax: +49(721)98653-98 D-76135 Karlsruhe, http://www.aonix.de From william.j.sheehan@lmco.com Thu Jul 23 10:15:03 1998 From: Bill Sheehan Date: Thu Jul 23 10:15:07 PDT 1998 Subject: Re: Parallel Port Interrupt handler for VxWorks I have solved my Parallel Port problem. The BSP was Level triggered Int7. It should have been Egde Triggered. Thanks To All. Bill Sheehan From Mitch_Hendrickson@Telestream.net Thu Jul 23 11:35:03 1998 From: Mitch Hendrickson Date: Thu Jul 23 11:35:07 PDT 1998 Subject: SENS and i960 (troubles) [I think the dog ate the first one since I didn't say vxWorks. If not, my apologies...] Hello vxWorkers- Can anyone out there share with me whether you've had success (or lack thereof) with SENS on an i960 target (especially if you're using the if_fei driver)? We're having a hack of a time with it.... -Mitch From chuckvb@ix.netcom.com Thu Jul 23 12:40:42 1998 From: Chuck Van Blaricom Date: Thu Jul 23 12:40:49 PDT 1998 Subject: WindView over Serial Backend with Tornado I need help in getting WindView to work in Tornado over a Serial Backend. Even the Post Mortem mode would be of great help. I have configured VxWorks for WindView in Post Mortem mode per the help manuals, and it seems to be collecting events on the i960 target. Does anyone know how to upload the raw data over the serial backend so that WindView can display it? I was able to get a memory dump into a PC Host file, but WindView could not read it. Any help is appreciated. Thanks in advance Chuck -- Chuck Van Blaricom ComTier, 2525 Walsh Ave, Santa Clara, CA 95051 (408) 235-7039 FAX: (408) 235-7085 EMail: chuckvb@ix.netcom.com From rvanallen@svsinc.com Thu Jul 23 13:10:50 1998 From: rvanallen@svsinc.com (Robert Van Allen) Date: Thu Jul 23 13:10:53 PDT 1998 Subject: Re: Interesting ATA post to VxWorks Newsgroup let's find out and attack...bva ************************************************************ SVS, Inc. 6207 Pan American Freeway NE, Albuquerque, NM 87109-3425 505-342-4600 505-345-8303 (fax) ************************************************************ From cgrames@mdc.com Thu Jul 23 14:38:50 1998 From: Charlie Grames Date: Thu Jul 23 14:38:53 PDT 1998 Subject: 32 MB Boundary on PowerPC VxWorks 5.3.1 MVME2604 1.1/4 I thought those of you following the discussion about the 32 MB boundary on PowerPC might be interested in the following e-mail I received from Wind River on this topic. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com Received: from blv-hub-01.boeing.com ([192.48.21.11]) by mime.mdc.com (GroupWise SMTP/MIME daemon 4.1 v3) ; Mon, 20 Jul 98 14:14:13 CDT Return-Path: Received: from blv-smtpin-01.boeing.com by blv-hub-01.boeing.com with ESMTP; Mon, 20 Jul 1998 09:28:05 -0700 Received: from mail.wrs.com ([147.11.1.11]) by blv-smtpin-01.boeing.com (8.9.0/8.8.5) with ESMTP id JAA28513 for ; Mon, 20 Jul 1998 09:27:53 -0700 (PDT) Received: from wrsz7484398.wrs.com (gordonu-pc.wrs.com [147.11.31.122] (may be forged)) by mail.wrs.com (8.8.6/8.8.5) with SMTP id JAA08615 for ; Mon, 20 Jul 1998 09:27:48 -0700 (PDT) Message-Id: <3.0.5.32.19980720114837.008b24c0@mail.wrs.com> X-Sender: gordonu@mail.wrs.com X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32) Date: Mon, 20 Jul 1998 10:48:37 -0500 From: Gordon Uchenick To: charles.r.grames@boeing.com Subject: PPC > 32 MB Hi Charles, I've been observing the dialogue on the exploder on this subject. I'd like to inform you that a patch is now available that gets past the restrictions of EABI. You will now be able to straightforwardly load object modules into mv2604s with more than 32MB of DRAM. It does require some work on your part to modify your makefiles, however. The patch is available for download from our web site. To find it, click on Support from the home page. Then, follow the link to Windsurf. You will need a Windsurf username and password. If you don't already have one, Customer Support will give you one either over the phone or by return email from support@wrs.com. Click on What's New on WindSurf. You are interested in the Jul 15 entry, Patch to Fix SPR 20257. Customer support (1-800-USA-4WRS) should be able to help you if you have any further questions. Hope this helped, Gordon ================================================================ Gordon M. Uchenick Phone : 410-256-7102 Sr. Field Application Engineer Fax : 410-256-7104 4132 E. Joppa Road #110-966 Email : gordon.uchenick@wrs.com Baltimore, MD 21236 http://www.wrs.com ================================================================ From cgrames@mdc.com Thu Jul 23 14:45:23 1998 From: Charlie Grames Date: Thu Jul 23 14:45:27 PDT 1998 Subject: Disabling LAN CRC error messages -Reply John, We used to see this problem with 1.0, also. It was corrected in the BSP in one of the later 1.0.1 releases. Short of upgrading to 1.0.1, you might want to see what Wind River can do for you individually. Good luck. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/22/98 10:24pm >>> Submitted-by jhend@primenet.com Wed Jul 22 20:24:08 1998 Submitted-by: JRH To all, Using a 2603 board w/ Tornado 1.0 and an NT host we occassionally see a CRC error displayed on the serial command shell during heavy traffic on the LAN that is shared with the above system. The overall throughput of the LAN during this time is acceptable with respect to the execution of the application on the 2603. I JUST WANT TO DISABLE THE WARNING. How do I do this? Thank you in advance, John R Henderson S/W Guy From ThanosJH@Npt.NUWC.Navy.Mil Fri Jul 24 09:54:06 1998 From: Thanos John H NUWCDIVNPT Date: Fri Jul 24 09:54:09 PDT 1998 Subject: DOS fs compatibility for 1 GByte Jaz drive I am using VxWorks 5.2 on a Motorola MVME167 single board computer. I have successfully modified the function usrScsiConfig() in the file usrScsi.c to install a DOS filesystem on two SCSI devices attached to the MVME167 SCSI controller: a hard disk drive ("/sd0/") and a 1.44 MByte floppy disk drive ("/fd0/"). I am able to copy files between these two SCSI devices, using the copy command at the VxWorks target shell prompt. Also, the DOS filesystem on the floppy disk is compatible with both VxWorks and MS-DOS on a PC. However, I also have a 1 GByte SCSI Jaz disk drive from Iomega on this SCSI bus. I have modified the source code in the file usrScsi.c to install a DOS filesystem under VxWorks that can access a pre-formatted (on a PC) Jaz disk. I use the function call, dosFsDevInit("/sd1/", pSbdJ0, NULL) in usrScsiConfig(), to initialize a volume descriptor for the Jaz drive, using the information in the boot sector on the pre-formatted Jaz disk. When I enter the command dosFsConfigShow "/sd1/" at the VxWorks target shell prompt, the displayed parameter values in the DOS_VOL_CONFIG structure do not make sense: device name: /sd1/ total number of sectors: 28489 bytes per sector: 8293 media byte: 0x6d # of sectors/cluster: 48 # of reserved sectors: 14640 # of sectors per FAT: 26469 max # of root dir entries: 8237 # of hidden sectors: 1919905906 removable medium: TRUE disk change w/out warning: not enabled auto-sync mode: not enabled long file names: not enabled exportable file system: not enabled lowercase-only filenames: not enabled volume_mode: O_RDWR (read/write) available space: 88564224 bytes The VxWorks target shell restarts while trying to determine the maximum available contiguous space. Also, I cannot use the ls command or the copy command to access the device /sd1/. 1. Does anyone know what are the correct DOS_VOL_CONFIG parameters for an Iomega 1 GByte SCSI Jaz disk drive, so that the DOS filesystem will be compatible with both VxWorks and MS-DOS? 2. Is it necessary to treat a removable Jaz drive exactly like a removable floppy drive, in the usrScsiConfig() function, i.e., is it necessary to call these functions in the order listed?: scsiPhysDevCreate scsiBlkDevCreate scsiModeSelect scsiPhysDevDelete scsiPhysDevCreate scsiBlkDevCreate scsiBlkDevInit dosFsDevInit 3. If the answer to question #2 is YES, what are the correct values for the following: a. modeData[1] (This is the second argument to the scsiModeSelect() call.) b. sectors per track (This is the first argument to the scsiBlkDevInit() call.) c. number of heads (This is the second argument to the scsiBlkDevInit() call.) Any assistance in this area would be greatly appreciated. John Thanos, Naval Undersea Warfare Center, Newport, RI 401-832-8551 thanosjh@npt.nuwc.navy.mil From scheurer@terra1.tz.rus.uni-stuttgart.de Fri Jul 24 13:19:30 1998 From: "F. Scheurer" Date: Fri Jul 24 13:19:34 PDT 1998 Subject: VxWorks 5.3: Problems with Booting SPARC Hi VxWorks community! Is there anyone having experience in SPARC on VxW5.3. Our problem with Themis SPARC 10HS (16Meg) still exists (just posted it)! I built a new vxworks-Image and a bootrom_uncmp using the attached configuration. With the 5.3 bootrom_uncmp i get the following output after completely having load the bootfile (snoop says: Last block transmitted via tftp from Host SUN Sol. 2.5.1): ... Boot device: /ne File and args: 4c800 Instruction Access Exception Type b(boot)... if i use the original bootrom_uncmp delivered from WRS it looks like this: ... Boot device: /ne File and args: 4c600 Watchdog Reset at 0x4000 Level 14 Interrupt Type b(boot)... If i use the original bootrom from 1995 (VxW 5.1) i get the behaviour some people already know: The TFTP-Boot counts up to... ... The IDPROM contents are invalid Boot device /ne File and args 51800 snoop shows: ... zfssun6 -> zfsrc5 TFTP Data block 619 (512 bytes) zfsrc5 -> zfssun6 TFTP Ack block 619 zfssun6 -> zfsrc5 TFTP Data block 620 (512 bytes) zfsrc5 -> zfssun6 TFTP Ack block 620 zfssun6 -> zfsrc5 TFTP Data block 621 (99 bytes) (last block) zfsrc5 -> zfssun6 TFTP Ack block 621 ? -> (multicast) ETHER Type=0000 (LLC/802.3), size = 52 by Anyway no error occured during the first phase!!! Continuing VxWorks Boot >Attaching network interface fn0... done >Subnet mask 0xffffff00 >Attaching network interface lo0... done >Loading... 563664 + 30968 + 36900 >Starting at 0x20000... That's what i see through snoop on the network: zfssun7 -> zfssun1 NIS C MATCH 4006 in passwd.byuid zfssun1 -> zfssun7 NIS R MATCH OK zfsrc5 -> zfssun6 TFTP Ack block 1164 zfssun6 -> zfsrc5 TFTP Data block 1165 (512 bytes) zfsrc5 -> zfssun6 TFTP Ack block 1165 zfssun6 -> zfsrc5 TFTP Data block 1166 (512 bytes) zfsrc5 -> zfssun6 TFTP Error: not defined ? -> (multicast) ETHER Type=0000 (LLC/802.3), size = 52 bytes Anyway an error occured!!! And there it hangs!! Is it possible, that a memory misconfiguration is the cause? Any help appreciated! Best regards Friedrich Scheurer -- +-------------------------------------------------+ | | | Dipl.-Ing. F. Scheurer | | | | c/o Zentrum Fertigungstechnik Stuttgart | | Arbeitsbereich Steuerungstechnik | | Nobelstrasse 15 | | D-70569 Stuttgart | | | | Telefon +49 711 13162-41(-0) | | Fax +49 711 13162-11 | | E-Mail scheurer@zfs.rus.uni-stuttgart.de | | zfssrr@zfssun6.tz.rus.uni-stuttgart.de | +-------------------------------------------------+ | http://zfs.rus.uni-stuttgart.de | +-------------------------------------------------+ | | | privat: | | Telefon +49 7161 686362 | | Fax +49 7161 686301 | | E-Mail Friedrich.Scheurer@t-online.de | | | +-------------------------------------------------+ From cruz_nojunk@xyplex.com Fri Jul 24 15:01:57 1998 From: Roger Cruz Date: Fri Jul 24 15:02:01 PDT 1998 Subject: Network Performance Info for vxWorks stack? Has anyone using VxWorks conducted a performance analysis of their network stack? I'm interested in finding out what their ethernet to ethernet routing capability is. I've seen reports where FreeBSD can do full bandwidth at 100MBs. I assume that vxWorks being lighter will be able to match this. Any info would be greatly appreciated. Please include CPU type, speed, etc if possible. If any folks at WRS read this newsgroup, could you please put me in contact with an engineer in your network group? Thanks, Roger. PS: remove _nojunk from email address vxworks VxWorks -- Roger Cruz cruz_nojunk@xyplex.com Xyplex Networks w: 978-952-4783 295 Foster Street f: 978-952-4887 Littleton, MA 01460 From daemon@csg.lbl.gov Sat Jul 25 04:02:29 1998 From: daemon@csg.lbl.gov Date: Sat Jul 25 04:06:49 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 25 04:02:26 PDT 1998 Subject: Re: DEC chip ... ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: DEC chip ... Date: Fri, 24 Jul 1998 15:48:34 -0400 From: Victor Sperry Organization: Paradyne Message-ID: <35B8E512.FF5C42D@eng.paradyne.com> References: <6p8b7j$nah$1@nnrp1.dejanews.com> mahershi@my-dejanews.com wrote: > > Hi! > > Back with another question. > > I am using a DEC 21140 based Ethernet controller. There's > a whole bunch of config parameters for it. (Check the dec21x40EndLoad > call in the SENS manual.) From my Win95 PC (which happens to > be my vxWorks Target) I found out the IRQ, Mem and I/O addresses. > > The dec21x40EndLoad() call requires (primarily) these parameters: > o base address of device registers. > o main memory address over PCI bus. (??) > o interrupt vector. > o interrupt level. > > Are the mem and i/o addresses I have in any way the same as > the PCI address and the base address? > > Your answers would be greatly helpful. > > thanks, > Mahesh R S > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- > http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum The way Win95 maps the DEC device into PCI address space has nothing to do with how vxWorks maps it. I'm using vxWorks with a Motorola MTX 603 PowerPC, which also uses the DEC 21140. In order to figure out how all the PCI devices were mapped, I had to dig through the BSP documentation and the .c files that implement the driver for the PCI bridge device. In my case, this device is Mot's raven ASIC, so the files are raven.c and ravenMpic.c. Your board will have some other type of bridge, of course. If you have vxWorks up and running, it's helpful to use pciDeviceShow at the WindSh command line: - -> pciDeviceShow value = 0 = 0x0 Scanning function 0 of each PCI device on bus 0 Using configuration mechanism 1 bus device function vendorID deviceID class 00000000 00000000 00000000 00001057 00004801 00060000 00000000 0000000b 00000000 000010ad 00000565 00060100 00000000 0000000c 00000000 00001000 0000000f 00010000 00000000 0000000e 00000000 00001011 00000009 00020000 00000000 00000010 00000000 0000109e 00008233 00020300 00000000 00000012 00000000 0000110a 00002102 00028000 The DEC 21140 is vendor ID = 0x1011 and device ID = 0x09. This gives you the PCI bus, device and function, in my case, 0, 0x0e, and 0. Now I can use pciHeaderShow. - -> pciHeaderShow 0,0xe,0 value = 0 = 0x0 vendor ID = 0x1011 device ID = 0x0009 command register = 0x0005 status register = 0x0280 revision ID = 0x22 class code = 0x02 sub class code = 0x00 programming interface = 0x00 cache line = 0x20 latency time = 0xff header type = 0x00 BIST = 0x00 base address 0 = 0x01020001 base address 1 = 0x00000000 base address 2 = 0x00000000 base address 3 = 0x00000000 base address 4 = 0x00000000 base address 5 = 0x00000000 cardBus CIS pointer = 0x00000000 sub system vendor ID = 0x0000 sub system ID = 0x0000 expansion ROM base address = 0x7e140000 interrupt line = 0x39 interrupt pin = 0x01 min Grant = 0x14 max Latency = 0x28 This is your PCI config space. The interrupt pin will be 01 for PCI INTA, 02 for PCI INTB, 03 for PCI INTC, and 04 for PCI INTD. This is by definition in the PCI spec. I've never seen a device that uses anything but INTA. How INTA, INTB, INTC, and INTD are routed through your interrupt controller to your CPU is defined by how your motherboard is wired. It's different for each PCI slot. To figure this out, I poured through my ravenMpic.c file and the system hardware docs. Base address 0 is the base PCI address that the device will respond to. The device knows how big its own address space is and it tells the CPU by how many bits in the BARs it will allow the CPU to write to. This is part of plug and play. In a Win95 system, the CPU finds all the devices on the PCI bus (descending across PCI-to-PCI bridges if it has to) and writes to all the BARs in all the devices to see how much address space it has to give to each device. In a vxWorks system, the BSP is written with a particular mapping in mind. Note that the lowest-order bit is an indication of whether the device uses I/O type PCI accesses or memory-type PCI accesses. And bit 0 is used for *mumble* I forget, so the base address is always on a 32-bit boundary. In the case above, when the DEC 21140 sees an address on the PCI bus between 0x01020000 and 0x0102fffc, it will claim the address by handshaking with the PCI bridge device. But you're not home yet! The PCI address is not the address that the CPU knows. They're mapped by the bridge device. So I had to dig into raven.c to figure out how all the devices on the motherboard were mapped. On my motherboard, with my BSP, the DEC 21140 lives at: MPC Address Range Device's BAR Size Definition - ----------------- ------------ ---- ---------- 81020000 8102FFFF 01020000 64K LAN device IO (21140) Ahah! Now we know some of what to tell the END driver. From the man page for dec21x40End: The format of the parameter string is: :::::::: TARGET-SPECIFIC PARAMETERS This is the base address at which the hardware device registers are located. This parameter defines the main memory address over the PCI bus. It is used to translate a physical memory address into a PCI-accessible address. And in configNet.h I have: # define DEC_LOAD_STRING "0x81020000:0x80000000:0x12:0x12:-1:-1:-1:0:0x80000 So: is where the device lives from the CPU's point of view is what you subtract from to get a corresponding PCI address. Hope this helps. Vic Sperry --------------------------- End of New-News digest ********************** From Jim.Rosenow@mts.com Mon Jul 27 04:58:31 1998 From: "Rosenow, Jim" Date: Mon Jul 27 04:58:34 PDT 1998 Subject: vxWorks 5.3.1 pc486 BSP 1.1 EOI question. Is an interrupt handler implemented for this target supposed to issue the PIC EOI or does vxWorks built in code do it (intEnt()/intExit())? I find in src/drv/intrCtl/i8259Pic.c a routine, sysIntEOI(), that unconditionally issues EOI's to both PIC's irrespective of IRQ (which I find odd BTW). This routine is assigned to the variable intEOI in sysLib.c. I can't find any documentation on intEOI(), sysIntEOI(), or recommended interrupt handler structure for this architecture. Can anyone out there help? Thanks. From mwu@alacritycom.com Mon Jul 27 18:39:52 1998 From: Ming Wu Date: Mon Jul 27 18:39:55 PDT 1998 Subject: performance of TCP/IP stack Does anyone know the performance of vxWorks TCP/IP stack on a Motorola860 at 50 or 60 Mhz? Ming From James_M_Flynn@res.raytheon.com Tue Jul 28 13:28:29 1998 From: James_M_Flynn@res.raytheon.com Date: Tue Jul 28 13:28:32 PDT 1998 Subject: MVME Parallel Port Configuration I am using an MVME2604 PPC board, running Tornado and vxWorks 5.3.1 with BSP 1.1/4. I want to use the parallel port to read TTL input signals. I am having trouble setting up the National PC87308 Super IO Controller. I believe I just have to configure the SIO controller initially, and then perform a read sequence on the SIO to read the status of the input signals over the Parallel Port, but I can't seem to get this to work. Can someone let me know how to do this, and/or provide a sample driver? Any help would be greatly appreciated Thanks, Jim From daemon@csg.lbl.gov Wed Jul 29 04:00:33 1998 From: daemon@csg.lbl.gov Date: Wed Jul 29 04:05:18 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 29 04:00:31 PDT 1998 Subject: problem launching Windview ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: problem launching Windview Date: Wed, 29 Jul 1998 06:57:08 GMT From: markus_mitterer@sbu1.storkgroup.com (Markus Mitterer) Organization: STK Message-ID: <35bec826.0@mailserv.stk.co.at> Reply-To: "Markus Mitterer" Hi, I have a problem with WindView. I4m using VxWorks5.3.1 with a mv2303-BSP and a NT-Workstation as host. In my host I have 2 network cards. One is for the LAN and the other one I use for the connection to my target. When I launch WindView I get the message: "waiting for connection 6164" but nothing else happens. When I uninstall the LAN-network card it works properly !!! I hope anybody can help me. Thanks in advance. Markus --------------------------- End of New-News digest ********************** From gwryan@bell-labs.com Wed Jul 29 05:12:04 1998 From: Jerry Ryan Date: Wed Jul 29 05:12:07 PDT 1998 Subject: remote wdb connections VxWorks I am trying to figure out if it is possible to use remote connections to wdb. Here is what I mean: I am running all the development tools on a solaris machine that doesn't have a lot of serial ports handy. The target systems run on little standalone boxes that are themselves connected to PCs running LINUX via a serial port. I am wondering if it is possible for the debugger agent to run remotely; that is, for the handler for the serial ports to run on the LINUX machines, but for the tools to actually run on the Solaris machine. Ideally the LINUX agents would just forward messages between the SOlaris box and the target systems transparently. Anybody ever do something like this, and if so how? Or is this just plain nuts? any help appreciated thanks jerry ryan gwryan@bell-labs.com From matthew.pardoe@widney-aish.co.uk Wed Jul 29 06:37:13 1998 From: "Matt Pardoe" Date: Wed Jul 29 06:37:21 PDT 1998 Subject: Sending/ Receiving Ethernet packets Keywords : vxworks, tornado Hello I am looking for some help on sending/ receiving ethernet packets to/ from vxWorks Please reply to me direct as I am not currently subscribing to the exploder, and do not have access to a news reader at present. If requested I will summarise and post any responses I get I am currently starting a new vxworks development using tornado 1.0.1 hosted on Windows NT 4.0 and targeted on an No. of motorola powerPC boards conected over ethernet. Initially I need to do some basic benchmarking/evaluation of the comms, and how much data we can send over the ethernet. We have calculated that we will probably have to send data in raw ethernet packets (avoiding tcp/ip overhead) to meet our throughput requirements. Initially we will not have all of the target boards that will be in the final system, and wish to simulate some of them using PCs connected to the network. The PCs are running NT 4.0 and have 3com 3C905B-TX ethernet cards (100 and 10Mbit ethernet). I've worked out how to send and receive ethernet packets between the vxWorks boards by adding hooks into the drivers. What I want to be able to do is send and reveive raw ethernet packets from the PC end. The content of packet is not too important at this stage, what I need to be able to do is send various lenght packets at various rates, and if possible to monitor no.s sent/received. Thanks Matt Pardoe -------------------------------------------------- Matthew Pardoe Widney Aish Ltd., Fleets House, Fleets Lane, Poole, Dorset, Bh15 3BU Tel: 01202 682600 Fax: 01202 665445 email matthew.pardoe@widney-aish.co.uk -------------------------------------------------- From bpringle@teklogix.com Wed Jul 29 07:20:39 1998 From: Bill Pringlemeir Date: Wed Jul 29 07:20:43 PDT 1998 Subject: SENS drivers, loading and unloading. Can an END driver be dynamically loaded, unloaded ? In other words, can we call muxDevLoad(), muxDevStart() ourselves, without adding the new driver to endDevTbl ? Will the tcp/ip stack talk to a driver that is started later like this? How is the tcp/ip protocol stack associated with a specific end driver? This is neccessary for hot-swappable network cards. Thanks, Bill VxWorks, Tornado, VxWorks, Tornado, VxWorks, Tornado From matthew.pardoe@widney-aish.co.uk Wed Jul 29 08:36:20 1998 From: "Matt Pardoe" Date: Wed Jul 29 08:36:28 PDT 1998 Subject: vxworks tornado Sending/ Receiving Ethernet packets Keywords : vxworks, tornado Hello I am looking for some help on sending/ receiving ethernet packets to/ from vxWorks Please reply to me direct as I am not currently subscribing to the exploder, and do not have access to a news reader at present. If requested I will summarise and post any responses I get I am currently starting a new vxworks development using tornado 1.0.1 hosted on Windows NT 4.0 and targeted on an No. of motorola powerPC boards conected over ethernet. Initially I need to do some basic benchmarking/evaluation of the comms, and how much data we can send over the ethernet. We have calculated that we will probably have to send data in raw ethernet packets (avoiding tcp/ip overhead) to meet our throughput requirements. Initially we will not have all of the target boards that will be in the final system, and wish to simulate some of them using PCs connected to the network. The PCs are running NT 4.0 and have 3com 3C905B-TX ethernet cards (100 and 10Mbit ethernet). I've worked out how to send and receive ethernet packets between the vxWorks boards by adding hooks into the drivers. What I want to be able to do is send and reveive raw ethernet packets from the PC end. The content of packet is not too important at this stage, what I need to be able to do is send various lenght packets at various rates, and if possible to monitor no.s sent/received. Thanks Matt Pardoe -------------------------------------------------- Matthew Pardoe Widney Aish Ltd., Fleets House, Fleets Lane, Poole, Dorset, Bh15 3BU Tel: 01202 682600 Fax: 01202 665445 email matthew.pardoe@widney-aish.co.uk -------------------------------------------------- From RadaJA@corning.com Wed Jul 29 09:06:20 1998 From: "Rada, Jeff A" Date: Wed Jul 29 09:06:23 PDT 1998 Subject: RE: remote wdb connections vxWorks users, I have tried something similar and never got good results. I tried to use a terminal server that was on the network and connect the vxWorks target through one of the serial ports. If I set up slip on the target I could get the connection to work but could not get the system mode debugging t o work but I could get the launcher tools to work that way. -----Original Message----- From: vxwexplo@lbl.gov [mailto:vxwexplo@lbl.gov] Sent: Wednesday, July 29, 1998 8:12 AM To: vxworks_users@csg.lbl.gov Subject: remote wdb connections Submitted-by gwryan@bell-labs.com Wed Jul 29 05:12:04 1998 Submitted-by: Jerry Ryan VxWorks I am trying to figure out if it is possible to use remote connections to wdb. Here is what I mean: I am running all the development tools on a solaris machine that doesn't have a lot of serial ports handy. The target systems run on little standalone boxes that are themselves connected to PCs running LINUX via a serial port. I am wondering if it is possible for the debugger agent to run remotely; that is, for the handler for the serial ports to run on the LINUX machines, but for the tools to actually run on the Solaris machine. Ideally the LINUX agents would just forward messages between the SOlaris box and the target systems transparently. Anybody ever do something like this, and if so how? Or is this just plain nuts? any help appreciated thanks jerry ryan gwryan@bell-labs.com From jshaw_shade@compuserve.com Wed Jul 29 09:22:03 1998 From: John Shaw Date: Wed Jul 29 09:22:07 PDT 1998 Subject: Motorola IP-488 drivers for VxWorks Hello VxWorkers, This forum has seen some enquiries in the past about possible VxWorks drivers for IEEE-488 hardware. But there have been few responses! We are about to use a Motorola IEEE-488 IP module for use with a MVME-162. This might either use the TMS9914A chip or the NAT9914A chip. Are there any suitable drivers for the chip or for the IP-488, for VxWorks or UNIX, in = C? Are there any recommended web sites for drivers or EVEN DATASHEETS? Strange that Motorola should have produced a board with no software! How did they test their kit? The board ident is the MVIP300, using the TMS chip. Any advice or useful information will be greatly appreciated. John Shaw, Shade (Computer Services) Limited, Calne, Wilthshire, England, UK. Email: john.shaw@shade.co.uk From aram@ozone.spdg.com Wed Jul 29 09:24:25 1998 From: Aram Nahidipour Date: Wed Jul 29 09:24:28 PDT 1998 Subject: H.323 protocol stack for vxWorks vxWorks Is there any company provide H.323 protocol stack for vxWorks? Is there any freeware implementation of H.323 protocl stack for vxWorks or any unix platform? Thanks Aram From Keith_Buchanan@SPARTA.com Wed Jul 29 13:18:51 1998 From: "Keith Buchanan" Date: Wed Jul 29 13:18:54 PDT 1998 Subject: RE: Motorola IP-488 drivers for VxWorks > This forum has seen some enquiries in the past about possible VxWorks > drivers for IEEE-488 hardware. But there have been few responses! We are > about to use a Motorola IEEE-488 IP module for use with a MVME-162. This > might either use the TMS9914A chip or the NAT9914A chip. Are there any > suitable drivers for the chip or for the IP-488, for VxWorks or UNIX, in = > C? > > Are there any recommended web sites for drivers or EVEN DATASHEETS? The WRS web site has some links to companies who offer drivers though the list is unimpressive. > Strange that Motorola should have produced a board with no software! How Usually with their debug software kits or an OS-9 driver... > did they test their kit? The board ident is the MVIP300, using the TMS > chip. > SPARTA can sell you the source for a VxWorks driver for the National Instruments GPIB/PC-104 card. It uses the TNT4882C chip set which can look like the 9914, the 4882 or the 7210. It would be a good piece of code to start out with if the modes I exercised are compatible with the 9914. Contact me off-line... adios ---------------------------------- /\\ Thomas Keith Buchanan ///\\ /////\\ SPARTA, Inc. ///// \\ 519 Interstate 30 / \\\\\ \\ Suite 140 \ \\\\\ // Rockwall, TX 75087 \ ////// \////// Voice: 972 672 9657 \//// FAX: 972 722 1216 \// WWW: www.sparta.com From gnn@wrs.com Wed Jul 29 22:12:17 1998 From: "George V. Neville-Neil" Date: Wed Jul 29 22:12:21 PDT 1998 Subject: Re: SENS drivers, loading and unloading. At 07:20 AM 7/29/98 PDT, the vxWorks Users Group Exploder wrote: >Can an END driver be dynamically loaded, unloaded ? >In other words, can we call muxDevLoad(), muxDevStart() >ourselves, without adding the new driver to endDevTbl ? Yes, that's possible. >Will the tcp/ip stack talk to a driver that is started >later like this? How is the tcp/ip protocol stack associated >with a specific end driver? > Yes, you need to use the ipAttach function. Later, George ---------------------------------------------------------------------------- George V. Neville-Neil gnn@wrs.com "Completely round is the steering wheel that leads to compound fractures..." --- The Kingston Trio From mmenge@dspt.com Thu Jul 30 06:33:10 1998 From: mmenge@dspt.com Date: Thu Jul 30 06:33:13 PDT 1998 Subject: MV2604 200MHz runs too Sloooow. Mark Menge@DSPT 07/30/98 09:32 AM I recently installed the 1.1/4 BSP for the 2604 because the original BSP no longer works with the boards Motorola ships now. Now I find the larger the code size the slower the code runs! I tried this test: With just vxWorks.st : m1 = malloc(100000) m2 = malloc(100000) timexN(memcpy, m1, m2, 100000) 992 microseconds With all our code loaded it takes 188 milliseconds! Does anyone know what possibly could cause it to run almost 200 times slower? I checked interrupts with spy, but that didn't seem to be the problem. vxWorks Tornado From KARTHUR@prodigy.net Thu Jul 30 08:02:54 1998 From: "Kurt A Lambrecht" Date: Thu Jul 30 08:02:57 PDT 1998 Subject: VxWorks-Adding Mem Card to Gespac cPCI System This is a multi-part message in MIME format. ------=_NextPart_000_0022_01BDBBA9.2C461720 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hello, I am working with a Gespac, Inc. PPC603e board on a compact PCI = backplane. My system includes an ethernet card, which communicates to = the Sun host. =20 I cannot seem to get a memory card configured into the system. The BSP = locates it, and writes values to the base address regs 1 and 2, as well = as the command reg in pci config. space. My question is this: What else do I have to do to the BSP to allow this = card to be addressed? It looks like it was placed at an address = overlapping that of the local memory on the cpu card. It turns out that = was the case - I pulled the card from the box, and my memory write/read = routines still worked. (not good). Is anyone out there using GESPAC equipment? =20 Thanx, Kurt Lambrecht ------=_NextPart_000_0022_01BDBBA9.2C461720 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello,
 
I am working with a Gespac, Inc. = PPC603e board=20 on a compact PCI backplane.  My system includes an ethernet card, = which=20 communicates to the Sun host. 
 
I cannot seem to get a memory card = configured=20 into the system.  The BSP locates it, and writes values to the base = address=20 regs 1 and 2, as well as the command reg in pci config. = space.
 
My question is this: What else do I = have to do=20 to the BSP to allow this card to be addressed?  It looks like it = was placed=20 at an address overlapping that of the local memory on the cpu = card.  It=20 turns out that was the case - I pulled the card from the box, and my = memory=20 write/read routines still worked. (not good).
 
Is anyone out there using GESPAC=20 equipment? 
 
Thanx,
Kurt Lambrecht
------=_NextPart_000_0022_01BDBBA9.2C461720-- From cgrames@mdc.com Thu Jul 30 12:20:32 1998 From: Charlie Grames Date: Thu Jul 30 12:20:36 PDT 1998 Subject: MV2604 200MHz runs too Sloooow. -Reply Mark, You say you are using the 200 MHz MVME2604? Is this the latest release with the Universe II chip? If so, I believe you need a patch. I got this patch some time back from Motorola for support of the MVME2700. You would need to check whether it is available from Wind River now. Even if you have the Universe II boards, I don't know that your problem is related to that. We have not seen any issues like this. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com >>> the vxWorks Users Group Exploder 07/30/98 08:33am >>> Submitted-by mmenge@dspt.com Thu Jul 30 06:33:10 1998 Submitted-by: mmenge@dspt.com Mark Menge@DSPT 07/30/98 09:32 AM I recently installed the 1.1/4 BSP for the 2604 because the original BSP no longer works with the boards Motorola ships now. Now I find the larger the code size the slower the code runs! I tried this test: With just vxWorks.st : m1 = malloc(100000) m2 = malloc(100000) timexN(memcpy, m1, m2, 100000) 992 microseconds With all our code loaded it takes 188 milliseconds! Does anyone know what possibly could cause it to run almost 200 times slower? I checked interrupts with spy, but that didn't seem to be the problem. vxWorks Tornado From kareem@cognex.com Thu Jul 30 12:45:56 1998 From: kareem@cognex.com (Kareem Fawell) Date: Thu Jul 30 12:45:59 PDT 1998 Subject: Template inline member functions I cannot get inline member functions of template classes to actually inline properly, and I am seeking help. For example, assume I have a header file with this declaration: template class Register { public: Reg(T *a) : addr(a) { }; T read() { return *addr; }; void write(T t) { *addr = t; }; private: volatile T *addr; }; and I include this header in a couple of .cpp files. When I check the assembly, the read and write functions are not inlined. I can get them inlined by explicitly instantiating the classes using template class Reg; for example, but then I get errors about multiple definitions. I would greatly appreciate any help. Thank you very much. Kareem Fawell, SWE Cognex, Inc Alameda, CA 94501 (510) 865-9882, ext 116 kareem@cognex.com VxWorks From daemon@csg.lbl.gov Fri Jul 31 04:05:04 1998 From: daemon@csg.lbl.gov Date: Fri Jul 31 04:05:19 PDT 1998 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 31 04:05:01 PDT 1998 Subject: Re: Host file access Subject: Re: 5.3 VxWorks BSP for 486 - Setting System Clock Rate ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Host file access Date: 30 Jul 1998 08:13:02 -0400 From: Michael Carr Organization: UltraNet Communications , an RCN Company http://www.ultranet.com/ Message-ID: References: <35BF8048.604002B@lmco.com> Are you opening the file by it's full path name? I.e., "://"? Steve O'Neill writes: > > We are seeing some rather bizarre behavior wrt accessing host files from > target boards... > > Our configuration - Tornado 1.0.1 hosted on NT workstations, multiple > MVME2307 boards, a slightly modified version of WRS's 2304 BSP. > > The odd behavior is that we can access host files from WinShell when > doing redirection of scripts or to load object files. However, when we > attempt to access files via fopen, for example, we get "Connection > Refused". > > Has anyone seen similar behavior? Any ideas? > > TIA, > Steve O'Neill > - -- - ----------------------------------------------------------- Michael Carr Redstone Communications *** Remove 'NOSPAM' from my email address when replying *** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 5.3 VxWorks BSP for 486 - Setting System Clock Rate Date: Thu, 30 Jul 1998 09:31:32 -0400 From: Kevin Bradley Organization: Doctoral student, Electrical and Computer Engineering, Carnegie Mellon, Pittsburgh, PA Message-ID: <0pk7Ko600iWP01nVE0@andrew.cmu.edu> References: <6pm3md$nh4@drn.newsguy.com> <6pnr8b$5gp@niven.ksc.nasa.gov> <6po6oi$83d@drn.newsguy.com> Excerpts from netnews.comp.os.vxworks: 29-Jul-98 Re: 5.3 VxWorks BSP for 486.. by Tom Haley@??? > When I read the resolution of the clock and saw that it was exactly > what I expected it to be I started to suspect WindView. I'm pretty sure > but not 100% sure. Has anybody had any experience changing the clock > rate from the default one for the BSP and seeing how WindView displays > data? I was thinking that Wind River probably did this in testing out > WindView. In its default configuration, WindView uses the auxiliary clock driver ($WIND_BASE/target/src/drv/timer/i8253Timer.c), which should technically be unaffected by the system clock rate. If you want to use an independent clock source for WindView, I suggest using the Pentium cycle counter. A WindView timestamp driver for this is available at ftp://usa.ece.cmu.edu/pub/VxWorks/pentium.c, which can either be "#include"'d into usrConfig.c or linked later. You only need to change config.h in the following way: /* Set up the PENTIUM timer drivers */ #define INCLUDE_TIMESTAMP #define INCLUDE_USERTIMESTAMP #define USER_TIMESTAMP pentium_TimeStamp #define USER_TIMESTAMPLOCK pentium_TimeStampLock #define USER_TIMEENABLE pentium_TimerEnable #define USER_TIMEDISABLE pentium_TimerDisable #define USER_TIMECONNECT pentium_TimerIntConnect #define USER_TIMEPERIOD pentium_Period #define USER_TIMEFREQ pentium_Freq This should tell usrConfig.c to not use the default and use this instead. Look for wvTmrRegister(...) in usrConfig.c to see how it's initialized. I have used this driver with many different clock rates, and found it to be very accurate. Can't speak for the auxiliary clock. -- Kevin --------------------------- End of New-News digest ********************** From mcuddy@fensende.com Fri Jul 31 16:12:13 1998 From: Mike Cuddy Date: Fri Jul 31 16:12:16 PDT 1998 Subject: BSP questions ... I'm a new VxWorks user (however, I'm not new to embedded programming ;-), and I'm trying to whip the DEC SA-110/EBSA-285 BSP into a shape that more closely matches our development target system (which should be arriving from FAB any day now ;-). Most notable is the exclusion of the "dc" ethernet SENS driver and replacing it with the AMD AM79C97x driver (included in the BSP developer's kit) so, my two questions are thus: 1) what are the ramifications of replacing the SENS driver with the old-style driver? how do I kick out the DEC driver (used by the rom vxworks for booting) and replace it with the AMD ethernet driver... 2) what's the earliest that I can get printf()'s out the console port in the BSP bootup cycle? Optimally, I'd really like to be able to get some debugging information out of the BSP in sysHwInit(), unfortunately, as I understand it, that's before the I/O system is initialized, so I can't just print out information there. Thanks for your help. -- Mike Cuddy (mcuddy@FensEnde.com, MC312), Programmer, Daddy, Human. Fen's Ende Software, Redwood City, CA, USA, Earth, Sol System, Milky Way. I remember asking why ... Let it rain, and protect us from this Cruel Sun. Join CAUCE: The Coalition Against Unsolicited Commercial E-mail. From bob@seaweed.com Fri Jul 31 16:48:07 1998 From: bob schulman Date: Fri Jul 31 16:48:10 PDT 1998 Subject: Re: [Q:] Memory management. Joachim Fabini (Joachim.Fabini@sea.ericsson.se) asked: Is there anyone around who implemented (or tried to implement) a proprietary memory-management on VxWorks to avoid fragmentation and/or to enhance performance? The malloc on VxWorks is so slow that proprietary mechanisms (e.g. memory pools for reusing fixed-size buffers) could imho significantly reduce fragmentation and enhance performance. Seaweed Systems has a written-from-scratch memLib replacement. It has all the functionality which Wind River's memLib has...and more! It has memory leak detection, sanity checks at malloc/realloc/free time, user-callable memory arena sanity checker, constant-time allocation, better fragmentation behavior, callbacks for various low-memory conditions, and more! A real ginzu knife of an app. Please call or email for information: info@seaweed.com 425.895-1721 Bob Schulman