From lyorg@scorpio.com Sun Sep 1 00:24:39 1996 From: lyorg@scorpio.com (Lyor Goldstein) Date: Sun Sep 1 00:24:41 PDT 1996 Subject: http daemon source code Does anyone know of source code for the "http" daemon on VxWorks ? ***************************************************************** * Lyor Goldstein. Email : lyorg@scorpio.com * * Scorpio Communications Ltd. Phone : 972-3-5339654 ext 115 * * Software group Fax : 972-3-5339518 * * 1c Yoni Netanyahu st. * * Or Yehuda 60250 * * P.O.Box 564, Israel * ***************************************************************** From daemon@csg.lbl.gov Sun Sep 1 04:01:09 1996 From: daemon@csg.lbl.gov Date: Sun Sep 1 04:01:11 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Sep 1 04:01:07 PDT 1996 Subject: Re: MV162 SRAM ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: MV162 SRAM Date: 31 Aug 1996 02:47:33 GMT From: steved@pacifier.com@199.2.117.163 (Steve Doiel) Organization: Pacifier BBS, Vancouver, Wa. ((360) 693-0325) Message-ID: <508945$2gg@news.pacifier.com> References: <3225A755.59E2B600@bbn.com> Reply-To: steved@pacifier.com (Steve Doiel) Fred Roeber writes: [snip] >The SRAM, on the other hand, isn't used explicitely by VxWorks at all >and is available for any type of application use. To use it you simply >set a pointer to its local address and store whatever you want there. >Hope this helps. Fred > Yes, very much, Thank you. SteveD steved@pacifier.com --------------------------- End of New-News digest ********************** From slirosi@gic.gi.com Sun Sep 1 17:11:36 1996 From: "Lirosi, Salvatore (HT-MS)" Date: Sun Sep 1 17:11:38 PDT 1996 Subject: FW: IP broadcast loopbacks Does anyone know of a way to inhibit the loop-back of broadcasted IP frames on 5.2? I've tried setting the "useloopback" to zero but to no avail. Thanks in advance, Sal LiRosi From daemon@csg.lbl.gov Mon Sep 2 04:00:17 1996 From: daemon@csg.lbl.gov Date: Mon Sep 2 04:00:20 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Sep 2 04:00:15 PDT 1996 Subject: Starting up multiple VxSim sessions? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Starting up multiple VxSim sessions? Date: 30 Aug 1996 17:17:45 GMT From: markham@bnr.ca (Andrew Markham) Organization: Bell Northern Research Message-ID: <5077np$cds@nrtphba6.bnr.ca> We are going to be doing some testing with a standard configuration of X simulators with (X-1) having one software load and 1 have another. Is it possible to write some sort of script that will crank up a bunch of simulators, load them with certain loads, and then run some initialization functions? Any help will be greatly appreciated, but source code would be even better... :^) - -- Andy Markham markham@nortel.ca Nortel, Inc. --------------------------- End of New-News digest ********************** From leonid@rst.co.il Mon Sep 2 05:31:49 1996 From: leonid@rst.co.il (Leonid Rosenboim) Date: Mon Sep 2 05:31:51 PDT 1996 Subject: Re: Starting up multiple VxSim sessions? One easy way to do what you need is to copy the config/sunos directory in to two new directories, and in each add your application objects to the Makefile, and statically link the VxWorks image with the application, and add initialization. SInce you have only two distinct "versions", it should be pretty straight-forward. Another way, is not to change anything in the VxWorks tree, but to prepare a "bootline" for each of the simulators. In sunos/sysLib.c there are sysNvramGet() and sysNvramSet() functions. VxSIm stores the boot paramters in a plain file. If you call bootChange() from the Shell, you will force VxSIm to create one such nvram file. If there is no write permission in the config directory, the file will be set in your home directory. The name of the file is vxWorks.nvram0 for CPU# and so forth. DOnt need to do a bootChange for each CPU, make one with bootChange, then edit it manually and create a bootline for each of the other processors, setting for each a startup script to load and initialize the application. Just remember to change the e=xxx.xxx.xxx.xxx IP number for each CPU#. Them you can fire up all VxSim's from an ordinary UNIX shell script, with config/vxWorks -p setting the CPU# from the command line, and each of them will use its distinct "bootline" and use it's own startup script. Hope this helps. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 P.O.Box 11502, AZUR 58017, Israel Fax: +972-3-559-8244 WWW: http://www.rst.co.il E-Mail: leonid@rst.co.il From froeber@bbn.com Mon Sep 2 10:55:37 1996 From: Fred Roeber Date: Mon Sep 2 10:55:39 PDT 1996 Subject: Re: Subject: Sockets with a timeout Kareem Fawell wrote: > I am using SOCK_STREAM type sockets and I would like to have a connect > and read with a timeout. If I were using UNIX I could do something > with signals, but this is not supported on vxWorks. Any ideas would be > appreciated. I usually just use the "select" system call to implement reads with timeouts since you can give that call a timeout and also ask it to return when the socket you want to read from is readable. Books like "Unix Network Programming" can show examples of how to set this up. A few points to be careful of are to realize that when select says your socket has data to read, it doesn't mean that all the data you want to read is necessarily there yet; you may have to read the data in multiple chunks if you are using a TCP/IP connection. Also, there can actually be cases where it could say that data is readable when there isn't actually any data available so you really should make sure to set the socket up as non-blocking and be prepared to get no data back from your read call (as far as I can tell, this last case can only happen when multiple tasks are sharing the same socket). Fred -- | Fred J Roeber, BBN Systems & Technologies | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | | See http://www.bbn.com/tv for TraceMaker info | From jmarc@wrsec.fr Tue Sep 3 02:39:59 1996 From: jmarc@wrsec.fr (Jmarc Jean-Marc de Maraumont) Date: Tue Sep 3 02:40:01 PDT 1996 Subject: Xterminal emulator for PC Dear VxUsers, Does any one of you use a Xterminal emulator successfully with tornado: PC client using an emulator + Unix station (X11 R5/R6) running Tornado. Tornado should run on the Unix station and being displayed on the PC. So far, I have tried using Xoftware (from Chameleon) but it crashes Windows 95. Thanks for your help, Jean-Marc De Maraumont. ***************************************************************** * Jean-Marc de Maraumont email : jmarc@wrsec.fr * * WIND RIVER SYSTEM EUROPE tel : 33 1 60 92 63 19 * * 19, Av. de la Norvege fax : 33 1 69 92 63 15 * * ZA de Courtaboeuf 1 * * 91962 LES ULIS cedex. France http://www.wrs.com * ***************************************************************** From daemon@csg.lbl.gov Tue Sep 3 04:00:32 1996 From: daemon@csg.lbl.gov Date: Tue Sep 3 04:00:34 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Sep 3 04:00:29 PDT 1996 Subject: Re: How to detect end of socket connection Subject: Re: How to detect end of socket connection Subject: Re: Subject: Sockets with a timeout Subject: Re: Programming Boot ROMS Subject: Hiding symbols with external linkage scope ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to detect end of socket connection Date: 31 Aug 1996 02:59:45 GMT From: steved@pacifier.com@199.2.117.163 (Steve Doiel) Organization: Pacifier BBS, Vancouver, Wa. ((360) 693-0325) Message-ID: <5089r1$2gg@news.pacifier.com> References: <507e7g$j6v@fu-berlin.de> Reply-To: steved@pacifier.com (Steve Doiel) IHeinz Junkes writes: >I have open a socket connection with socket, bind, listen, accept > >then i read endless commands form the connection..... >now the other computer crached ..... >the read request gets now always data ... without end > >How can i check that a connection is still alive ??? > >Heinz > I have found that the "read" routine apparently only returns zero length (on blocking sockets) when the connection is lost. I can't guarantee this but it appears to work this way. Also, check out the setSockOpt( ...KEEPALIVE ) option (sorry I don't have a manual in front of me to check the exact syntax). I hope this helps, Steve Doiel --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to detect end of socket connection Date: Fri, 30 Aug 1996 16:19:33 -0700 From: "H.J. Bae" Organization: Peaceful Star Message-ID: <32277705.78BA@peacefulstar.oakland.ca.us> References: <507e7g$j6v@fu-berlin.de> Reply-To: hjb@peacefulstar.oakland.ca.us Heinz Junkes wrote: > > I have open a socket connection with socket, bind, listen, accept > > then i read endless commands form the connection..... > now the other computer crached ..... > the read request gets now always data ... without end > > How can i check that a connection is still alive ??? > Since the other computer simply crashed, you can't be sure whether the connection has been closed properly by the other computer. If it had been correctly closed you could try sending data to it, and it should come back with error. If it is unknown whether the other computer had a chance to properly close the connection, there is not any quick way to know the other computer simply crashed and connection is no longer valid. You can try using KEEPALIVE timeout, but it is set to a long default value (hours). Pretty much the only reliable and timely way of knowing something like this (the other computer crashing) is by implementing application specific "are-you-there" or "i'm-alive" type of messages on a seperate channel and timeout dead connections. - -- H.J. Bae http://peacefulstar.oakland.ca.us --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Subject: Sockets with a timeout Date: 31 Aug 1996 03:05:39 GMT From: steved@pacifier.com@199.2.117.163 (Steve Doiel) Organization: Pacifier BBS, Vancouver, Wa. ((360) 693-0325) Message-ID: <508a63$2gg@news.pacifier.com> References: <9608310057.AA04625@beavis.isys> Reply-To: steved@pacifier.com (Steve Doiel) Kareem Fawell writes: >Hello: > >I am using SOCK_STREAM type sockets and I would like to have a connect >and read with a timeout. If I were using UNIX I could do something >with signals, but this is not supported on vxWorks. Any ideas would be >appreciated. > Gee.. I asked EXACTLY the same question several months back. WRS sent me a tech note that documents "ConnectWithTimeout". It does just what you're looking for. If you want particulars on the parameters check in the header files, if you don't find them email me at steved@woodland.usnr.com. SteveD steved@pacifier.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Programming Boot ROMS Date: Sat, 24 Aug 1996 19:06:11 GMT From: wb6w@netcom.com (Glenn Thomas) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: References: Sender: wb6w@netcom19.netcom.com My newsreader claims that CHRIS JONES (cjones@digsys1.bwi.wec.com) wrote: : Are there any good sources available with instructions for : programming boot roms? : Our target is an MVME177. : How does one seperate ROM image into the even and odd prom? : Thanks, : Chris Jones : cjones@digsys1.bwi.wec.com : (410) 765-7959 : (410) 993-8822 FAX : Northrop Grumman : Electronic Sensors & Systems Division Chris - The several examples I know of all involve someone writing a program on the host system that reads the absolute file produced by the linker (and possibly translated by something like elmHex) into a memory image buffer and then writeing a file or files in whatever format is convienent. If you lean of another way to do it, I'd appreciate an e-mail about it! Thanks! - -- ********************************************************************* * "Two wrongs don't make a right, but three lefts do." * * * * wb6w@netcom.com - Glenn Thomas * ********************************************************************* --------------------------- Newsgroups: comp.os.vxworks Subject: Hiding symbols with external linkage scope Date: 2 Sep 1996 23:30:52 GMT From: "Arve Sollie" Organization: ABB Network Partner Message-ID: <01bb9925$fd9d3a40$6663d4c1@nenu02h1.nopow.abb.no> We are about to start up a rather large project with vxWorks as platform. The dynamic link facilites is ideal as we will develop a rather large set of optional modules. Each of the modules will have a rather narrow API, and most inter-module communication will be database oriented. Some of the modules will be rather large, and will be made up of several source files. Question: Is it possible to link a set of object files together, but only to expose the API routines in the symbol table left after the module link process.? Symbols that are visible between the object files in a module, should not be visible outside the module. The main problem is to avoid multiple symbol declarations, as each module will be designed and implemented by different teams. - ---------------------------------------------------------------- arve.sollie@serly.mail.abb.se asollie@online.no - ---------------------------------------------------------------- --------------------------- End of New-News digest ********************** From MikeMcQuade@connect.com Tue Sep 3 07:07:50 1996 From: Mike McQuade Date: Tue Sep 3 07:07:52 PDT 1996 Subject: Re: Xterminal emulator for PC At 02:40 AM 9/3/96 PDT, you wrote: >Submitted-by jmarc@wrsec.fr Tue Sep 3 02:39:59 1996 >Submitted-by: jmarc@wrsec.fr (Jmarc Jean-Marc de Maraumont) > > > > Dear VxUsers, > > > Does any one of you use a Xterminal emulator successfully with tornado: > >PC client using an emulator + Unix station (X11 R5/R6) running Tornado. > >Tornado should run on the Unix station and being displayed on the PC. >So far, I have tried using Xoftware (from Chameleon) but it crashes Windows 95. > We have done it with Linux, of course it will cost you about $6.00 to get Linux. Using Linux as a remote X display worked great. MikeMcQuade@connect.com From krausvk@vs-ulm.dasa.de Tue Sep 3 07:17:11 1996 From: "Volker" Date: Tue Sep 3 07:17:13 PDT 1996 Subject: Re: Xterminal emulator for PC > Dear VxUsers, > > > Does any one of you use a Xterminal emulator successfully with tornado: > > PC client using an emulator + Unix station (X11 R5/R6) running Tornado. > > Tornado should run on the Unix station and being displayed on the PC. > So far, I have tried using Xoftware (from Chameleon) but it crashes Windows 95. > > Thanks for your help, > > Jean-Marc De Maraumont. > Hi, we use Exceed for Windows95 /NT from "Hummingbird". We thought as well about Linux. Regards Volker ########################################################### Dr. Volker Kraus debis Systemhaus GEI Magirusstrasse 43 D-89077 Ulm Germany Tel.: ++49-(0)731-9344-2051 FAX : ++49-(0)731-9344-100 Mail: vkraus@gei-ulm.daimler.benz.com currently reachable at: Daimler-Benz Aerospace Telesensomatik / Airportlogistik Woerthstrasse 85 D-89077 Ulm Tel.: ++49-(0)731-392-4610 FAX : ++49-(0)731-392-3663 Mail: krausvk@vs-ulm.dasa.de ########################################################### From tmorris@me.umn.edu Tue Sep 3 07:30:19 1996 From: "Ted Morris" Date: Tue Sep 3 07:30:21 PDT 1996 Subject: Re: http daemon source code > > Submitted-by lyorg@scorpio.com Sun Sep 1 00:24:39 1996 > Submitted-by: lyorg@scorpio.com (Lyor Goldstein) > > Does anyone know of source code for the "http" daemon on VxWorks ? > Here's something I saved a while back. Haven't had a chance to use it yet. In the same line of thought -- is there a server that runs JAVA out there? yikes! Ted > > > A lot of people asked for the toy HTTP server for VxWorks. > It is in: > http://peacefulstar.oakland.ca.us/vxhacks.html > > It's a hack, so don't expect much. It's useful for letting your > embedded target respond to HTTP clients (netscape) > for management functions, etc. > > HJB > _/_/_/ _/_/ _/_/ _/_/ _/_/_/ _/_/ _/_/_/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/ _/_/ _/ _/ _/ _/ _/ _/ _/ _/_/_/_/ _/ _/ _/ _/ _/ _/_/_/_/ _/ _/_/_/ _/ _/ _/ _/ _/_/_/ _/ _/ _/_/_/ ------------------------------------------------------------------------------ Ted Morris tmorris@me.umn.edu 612-625-3520 Center For Advanced Manufacturing Design And Control 612-625-9881 Institute of Technology FAX: 612-625-8884 University of Minnesota 111 Church St. S.E. (Rm 125) Mpls MN 55455 "A lie can travel halfway around the world before the truth can get its shoes on." [Actually, NOT Mark Twain] ------------------------------------------------------------------------------ From c_mcfarr@linkabit.titan.com Tue Sep 3 08:20:37 1996 From: c_mcfarr@linkabit.titan.com (Shawn McFarr) Date: Tue Sep 3 08:20:39 PDT 1996 Subject: Using crosswind in system-mode I am using Tornado 1.0 and would like to set breakpoints in my tasks and ISR routines using crosswind. What are the steps involved to set breakpoints using the system mode? The target agent is configured with a serial port as the communication interface to host based tools. Thanks, Shawn McFarr From clugo@emu.sp.trw.com Tue Sep 3 10:12:36 1996 From: "Carol L. Lugo" Date: Tue Sep 3 10:12:38 PDT 1996 Subject: rpcgen for vxworks Hi ! Looking for a rpcgen for VxWorks...does anyone know of anything ?? Carol From SJohnson@systran.com Tue Sep 3 13:46:46 1996 From: Steve Johnson Date: Tue Sep 3 13:46:48 PDT 1996 Subject: PPC 1604 VME A32 Mapping Summary of problem: Cannot access A32 VME space when using default values in sysLib.c Description of problem: Found the following values defined in mv1600.h: VME_WINDOW_A32_BASE 0xd0000000 VME_WINDOW_A32_SIZE 0x04000000 This will allocate a VME window from CPU addresses 0xd0000000 to 0xd3ffffff which is consistent with the values found in the VME2PCI registers 0x48 = 1000 0x4a = 13ff 0x4c = 0001 0x4e = f000 which causes any access in the range d0000000 to d3ffffff to be translated to the VME addresses 00000000 to 03ffffff But, when I place our card at VME address 0 and try to read the board's memory, I get: 0xd0000000: data access Exception... ... The problem seems to be with the addition of the LOCAL_BUS_MEM_A32 in the sysPhysMemDesc[] array. This seems to allow only VME addresses from 0xd8000000 to 0xdcffffff but the VME2PCI chip does not pass these addresses on. I can get around this by widening the window in VME2PCI and placing our card at 0x08000000 or by placing our card at 0x00000000 and removing the LOCAL_BUS_MEM_A32 and rebuilding the kernel. Am I doing something wrong, or is the VME A32 space disabled by default on purpose? From idg!idg.encomix.com!Mr.Chip@encomix.com Tue Sep 3 14:06:43 1996 From: Mr.Chip@idg.encomix.com (Roberto Artola) Date: Tue Sep 3 14:06:45 PDT 1996 unsubscribe --- OffRoad 1.9n unregistered From bqv@eng102.eng.i-o.com Tue Sep 3 15:19:59 1996 From: "Bang Q. Vu" Date: Tue Sep 3 15:20:02 PDT 1996 Subject: Re: rpcgen for vxworks Carol Lugo penned: > >Hi ! >Looking for a rpcgen for VxWorks...does anyone know of anything ?? > >Carol > Well, i'm using rpcgen which comes with SunOS 5.4 (/usr/bin/rpcgen) to generate codes for VxWorks. You should be able to find rpcgen for other flavor with Unix. -- Bang Vu | I/O, Inc. bvu@eng102.eng.i-o.com | 12300 Parc Crest Dr. Voice 713-879-2087 | Stafford, Texas 77477 Fax 713-879-2096 | From MikeMcQuade@connect.com Tue Sep 3 15:57:21 1996 From: Mike McQuade Date: Tue Sep 3 15:57:23 PDT 1996 Subject: Re: PPC 1604 VME A32 Mapping At 01:46 PM 9/3/96 PDT, you wrote: >Submitted-by SJohnson@systran.com Tue Sep 3 13:46:46 1996 >Submitted-by: Steve Johnson > > >Summary of problem: > >Cannot access A32 VME space when using default values in sysLib.c > snip... I had a similar problem trying to access A24 space, I fixed my problem with the following code in sysLib.c Also, someone a while back had a similar problem, thhey were in A32 space, I sent them this code and it worked for them... /* * sysBatDesc[] is used to initialize the block address translation (BAT) * registers within the PowerPC 603/604 MMU. */ UINT32 sysBatDesc [2 * (_MMU_NUM_IBAT + _MMU_NUM_DBAT)] = { ((VME_WINDOW_A24_BASE & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_16M | _MMU_UBAT_VS | _MMU_UBAT_VP), /* Instruction BAT 0 */ ((VME_WINDOW_A24_BASE & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW | _MMU_LBAT_GUARDED | _MMU_LBAT_CACHE_INHIBIT), ((VME_WINDOW_A32_BASE & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | _MMU_UBAT_VS | _MMU_UBAT_VP), /* Instruction BAT 1 */ ((VME_WINDOW_A32_BASE & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW | _MMU_LBAT_GUARDED | _MMU_LBAT_CACHE_INHIBIT), _MMU_UBAT_NOT_USED, /* Instruction BAT 2 */ _MMU_LBAT_NOT_USED, _MMU_UBAT_NOT_USED, /* Instruction BAT 3 */ _MMU_LBAT_NOT_USED, ((VME_WINDOW_A24_BASE & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_16M | _MMU_UBAT_VS | _MMU_UBAT_VP), /* Data BAT 0 */ ((VME_WINDOW_A24_BASE & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW | _MMU_LBAT_CACHE_INHIBIT), ((VME_WINDOW_A32_BASE & _MMU_UBAT_BEPI_MASK) | _MMU_UBAT_BL_256M | _MMU_UBAT_VS | _MMU_UBAT_VP), /* Data BAT 1 */ ((VME_WINDOW_A32_BASE & _MMU_LBAT_BRPN_MASK) | _MMU_LBAT_PP_RW | _MMU_LBAT_CACHE_INHIBIT), _MMU_UBAT_NOT_USED, /* Data BAT 2 */ _MMU_LBAT_NOT_USED, _MMU_UBAT_NOT_USED, /* Data BAT 3 */ _MMU_LBAT_NOT_USED }; >Yep, your sysBatDesc worked, but I had to make a change in >VME_WINDOW_A32_SIZE defined in mv1600.h. The default was way too small >at 0x04000000. With LOCAL_MEM_BUS_A32 starting at 0x08000000, this >size is not enough to reach the PCI Slave Start Adress at 0x10000000. >I changed VME_WINDOW_A32_SIZE to 0x10000000, and started my addressing >at 0x0a000000. (Also I commented out the entry for A32 space in >sysPhysMemDesc). Now I can read and write to A32 VME memory! Hope it helps. MikeMcQuade@connect.com From bff@dspcon.com Tue Sep 3 18:01:46 1996 From: bff@dspcon.com (Robert F. Franz) Date: Tue Sep 3 18:01:51 PDT 1996 Subject: 100BaseT4 I have 2 100BaseT4 boards and I am wondering if I can hook them together without a hub if I cross the wiring. Are the signal levels compatible? The price of 100BaseT4 hub is dropping every week and I only need two computers networked at the moment. ------------------------------------------------------------------------------- DSPCon, Inc. -- leaders in digital signal processing solutions ------------------------------------------------------------------------------- Bob Franz 380 FootHill Rd. Bridgewater, NJ 08807 (908)722-5656 Fax:(908)722-3259 bff@dspcon.com ------------------------------------------------------------------------------- From ensslin@dreisoft.de Wed Sep 4 00:45:03 1996 From: Friedrich Ensslin Date: Wed Sep 4 00:45:05 PDT 1996 Subject: AW: Xterminal emulator for PC the vxWorks Users Group Exploder[SMTP:vxwexplo@lbl.gov] wrote: > > Dear VxUsers, > > > Does any one of you use a Xterminal emulator successfully with tornado: > >PC client using an emulator + Unix station (X11 R5/R6) running Tornado. > >Tornado should run on the Unix station and being displayed on the PC. >So far, I have tried using Xoftware (from Chameleon) but it crashes Windows 95. > > Thanks for your help, > > Jean-Marc De Maraumont. > I am using Hummingbird's eXceed on Win95 successfully for exactly that purpose, and it works fine. Hope this helps. ------------------------------------------------ Friedrich Ensslin 3SOFT GmbH Wetterkreuz 19a , D-91058 Erlangen , Germany phone: +49 9131 7701-0 , fax: +49 9131 7701-80 email: ensslin@dreisoft.de From rip@wrsec.fr Wed Sep 4 01:30:50 1996 From: rip@wrsec.fr (Rip richard Williamson ) Date: Wed Sep 4 01:30:52 PDT 1996 Subject: Re: PPC 1604 VME A32 Mapping Steve Johnson might have said: > Summary of problem: > > Cannot access A32 VME space when using default values in sysLib.c > > Description of problem: > > Found the following values defined in mv1600.h: > VME_WINDOW_A32_BASE 0xd0000000 > VME_WINDOW_A32_SIZE 0x04000000 > > This will allocate a VME window from CPU addresses > 0xd0000000 to 0xd3ffffff > which is consistent with the values found in the VME2PCI registers > 0x48 = 1000 > 0x4a = 13ff > 0x4c = 0001 > 0x4e = f000 > > which causes any access in the range > d0000000 to d3ffffff > to be translated to the VME addresses > 00000000 to 03ffffff > > But, when I place our card at VME address 0 and try to read the board's > memory, I get: > > 0xd0000000: > data access > Exception... > ... > > The problem seems to be with the addition of the LOCAL_BUS_MEM_A32 in the > sysPhysMemDesc[] array. > This seems to allow only VME addresses from > 0xd8000000 to 0xdcffffff > but the VME2PCI chip does not pass these addresses on. > > I can get around this by widening the window in VME2PCI and placing our card > at 0x08000000 or by placing our card at 0x00000000 and removing the > LOCAL_BUS_MEM_A32 and rebuilding the kernel. > Am I doing something wrong, or is the VME A32 space disabled by default on > purpose? > The mv1604 bsp has in its sysLib.c the following entry in the sysPhysMemDesc regarding A32 space on the VME: { (void *) (VME_WINDOW_A32_BASE + LOCAL_MEM_BUS_A32), (void *) (VME_WINDOW_A32_BASE + LOCAL_MEM_BUS_A32), 0x04000000, /* 64 meg */ VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, This entry has the effect of opening a window on the A32 mapping from 0x0800'0000 to 0x0c00'0000. This is mapped this way, because the LOCAL_MEM_BUS_A32 is set to 0x0800'0000 + (0x0100'0000 * procNum), and if you are trying to boot n ppc cards via shared memory, this is where they will begin in A32 space. LOCAL_MEM_BUS_A32 is set to 0x0800'0000 due to a hardware addressing conflict. If, however, you are only trying to read an i/o card at A32 0x0, you will get this data access error. The three simplest methods then to fix this feature: 1) change the sysPhysMemDesc entry addresses: { (void *) (VME_WINDOW_A32_BASE), (void *) (VME_WINDOW_A32_BASE), 0x04000000, /* 64 meg */ VM_STATE_MASK_VALID | VM_STATE_MASK_WRITABLE | VM_STATE_MASK_CACHEABLE, VM_STATE_VALID | VM_STATE_WRITABLE | VM_STATE_CACHEABLE_NOT }, (Allows A32 access from 0x0 to 0x03ff'ffff) 2) If you need to boot n ppc cards, as well as an i/o card at 0x0, then you will need to change the size of the window to something outrageous like 0x1000'0000. This allows A32 access from 0x0 to whereever you have set, but eats up a ton of memory creating the page tables for the MMU. It might be smarter to rejumper your i/o board. 3) Rejumper your target board for 0x0800'0000. Wed Sep 4 10:36:02 MET DST 1996 ----------------------------------------------------------------------- Merci de corriger mes fautes. Sie koennen mich gern berichtigen. ----------------------------------------------------------------------- Richard Williamson rip@wrsec.fr Technical Support Engineer vox: +33 (1) 60.92.63.13 Wind River Systems - EC fax: +33 (1) 60.92.63.15 ----------------------------------------------------------------------- INCLUDED MESSAGES: 36: Certe, Toto, sentio nos in Kansate non iam adesse... --------------------------------------------------END INCLUDED MESSAGES From daemon@csg.lbl.gov Wed Sep 4 04:01:39 1996 From: daemon@csg.lbl.gov Date: Wed Sep 4 04:01:41 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Sep 4 04:00:47 PDT 1996 Subject: Re: work queue overflow Subject: Re: Two Ethernets Subject: lptDrv, lptIntr never called Subject: Re: HELP!!! ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: work queue overflow Date: 25 Aug 1996 18:19:15 -0400 From: thesharps@aol.com (THE SHARPS) Organization: America Online, Inc. (1-800-827-6364) Message-ID: <4vqjh3$9dr@newsbf02.news.aol.com> References: <321F7F76.D0B@aminet.co.kr> Reply-To: thesharps@aol.com (THE SHARPS) Sender: root@newsbf02.news.aol.com In general, that message indicates that the kernal work queue, which is very small, has overflowed. Since this should not be happening, it is a symptom of another problem, probably dealing with interrupts or interrupt handlers. I have seen that message when I forget to remove a printf from an ISR being developed, or have done other silly things in ISRs. If you are using interrupts in your application, make sure that you are clearing the status bits and the source of the interrupt properly, and that you are not doing anything improper in the ISR. The basic rules of thumb for ISRs are not to do anything that will block, and to get in and out as quickly as possible. Hope this helps. Please let me know what it ultimately turns out to be. Tim Sharp Software Engineer, Cambridge Parallel Processing tsharp@cppus.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Two Ethernets Date: Tue, 03 Sep 1996 10:10:50 -0700 From: Greg Rasche Organization: General Intrument Message-ID: <322C669A.6424@gi.com> References: <2.2.32.19960829212248.0068cf90@mailhost.mclean.sparta.com> Yes two enet interfaces are supported. There is a 68EN360 manual marked "revision 1", which is newer than the "no revisions" version that clearly states this. Greg Rasche Thomas K. Buchanan wrote: > > At 01:10 PM 8/29/96 PDT, you wrote: > >Submitted-by budh@venus.gvg.tek.com Thu Aug 29 13:10:14 1996 > >Submitted-by: budh@venus.gvg.tek.com (Bud Hensley) > > > > > >Has anyone run the Motorola MC68EN360 processor at 25mHZ > >with two ethernets. There is a controversy on performance, > >if two can be active the same time. Does the CPM risc have > >enough horsepower to run 2 copies of the ethernet protocol > >microcode at the same time. > > It is my understanding that the '360 can only physically and electrically > support one 10Mbit/s interface at a time. At least that is what my > technical manual says... > > adios > ----------------------------------------- > /\\ Keith Buchanan > ///\\ > /////\\ SPARTA, Inc. > ///// \\ 7926 Jones Branch Drive > / \\\\\ \\ Suite 900 > \ \\\\\ // McLean, VA 22102 > \ ////// > \////// Voice: 703 448 0210 > \//// FAX: 703 893 5494 > \// WWW: www.mclean.sparta.com --------------------------- Newsgroups: comp.os.vxworks Subject: lptDrv, lptIntr never called Date: Tue, 03 Sep 1996 14:46:22 -0600 From: Tom Knowlton Organization: Evans and Sutherland Computer Corp. Message-ID: <322C991E.48F5@es.com> I am using the standard (WRS) lptDrv.c to drive a parallel printer on an 80486 target system. Everything appears to be working right, except for the fact that the printer won't print (although I know it isn't hardware that is the problem... it prints when I boot the same hardware up into NT. a typical session might be as follows: - -> lptDrv(0x0f, 07) /* IRQ 7 = INT 0x0f */ 0x00 (return code) /* I hear the printer make its typical "reset" sound */ - -> lptDevCreate("/lpt/0", 0) 0x00 - -> fd = open("/lpt/0", 1) /* 1 = WRITE */ symbol fd added to symbol table... (shortened for brevity) fd = 0x13 = 19 - -> fdprintf(fd, "this is a test") lptWrite: secCnt = 1 intCnt = 0 lptWrite: secCnt = 2 intCnt = 0 lptWrite: secCnt = 3 intCnt = 0 lptWrite: secCnt = 4 intCnt = 0 /* you get the picture? */ This appears to be telling me that the IRQ is not going high, therefore INT 0x0f is not being triggered. The driver code sets the parallel port control register to enable the IRQ, so my big question is, why isn't the doggone ISR being called? (as a result of the ISR not being called, semaphores aren't being given up, and that is what causes the loop as described above). Can anyone explain what I am doing wrong? If you need more information about what I am doing, please email me. Your help is appreciated. - -- *********************************************************************** |Thomas E. Knowlton, Jr. | "Some people without any brains | |tknowlto@es.com | do an _awful_ lot of talking." | |Evans & Sutherland Computer Corp. | --the Scarecrow, from | |http://www.cc.utah.edu/~tek08360 | _The_Wizard_of_OZ_ | |#include | | *********************************************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: HELP!!! Date: 23 Aug 1996 18:16:55 GMT From: scott@tcville.es.hac.com Organization: Hughes Aircraft Message-ID: <4vksin$9o9@hacgate2.hac.com> References: <2.2.32.19960821220149.0069c118@mail.wrs.com> Reply-To: scott@tcville.es.hac.com In article 0069c118@mail.wrs.com, Jon Hillman () writes: >At 11:21 AM 8/21/96 PDT, Tai Cao wrote: >> Does anyone know the address of the GUI news group. >>I need some information on the GUI design tool that interfaces with >>Ada application on the PC platform (486 processor). Any suggestion >>would be greatly appreciated. >I understand that Green Hills ADA-MULTI will be available soon on >PC/Windows95. I have forwarded your request to Pat Rodenbeck of >Green Hills (pat@ghs.com) to get in touch with you. I believe their PC version supports Ada95 !! Quite a versatile builder utility and host side debugger (we have 6months of use with their Ada83+C builder on an AIX host with SELF/RAD6k target). Scott - --- Hughes Aircraft Co. voice: (310) 616-1059 Image and Signal Processing Lab addr: scott@tcville.es.hac.com PO Box 902, EO/E01/A172, El Segundo, Ca. 90245 --------------------------- End of New-News digest ********************** From Peter_D_Smith@ccmail.res.ray.com Wed Sep 4 05:43:31 1996 From: Peter_D_Smith@ccmail.res.ray.com Date: Wed Sep 4 05:43:33 PDT 1996 Subject: VxWorks Exploder Subscription List Please remove my e-mail address from the VxWorks Exploder subscription list. Thank you. Peter D. Smith From clugo@emu.sp.trw.com Wed Sep 4 12:45:38 1996 From: "Carol L. Lugo" Date: Wed Sep 4 12:45:40 PDT 1996 Subject: shell hell ! Here is the problem: We needed to be able to launch a startup script when our target board boots. We modified our config.h file and added quite a few 5.2 options in order to get the startup script to work under 5.3 (Tornado). Everything was fine until I used "m" command inside the shell to modify a register in the A16 address space of a custom board. It looks as though executing "m" under the Tornado shell caused the wrong value to be written into the register while executing the "m" command under the target resident shell caused the correct value to be written to the register. Yukk...please tell me there is a simple explanation .... Thanks in advance, Carol Lugo From daemon@csg.lbl.gov Thu Sep 5 04:02:37 1996 From: daemon@csg.lbl.gov Date: Thu Sep 5 04:02:39 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 5 04:02:34 PDT 1996 Subject: Re: Subject: Sockets with a timeout Subject: Re: UDP packet - max size? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Subject: Sockets with a timeout Date: 3 Sep 1996 20:12:15 GMT From: alexm@vlsi.gsfc.nasa.gov (Alex Measday) Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <50i3ev$oil@post.gsfc.nasa.gov> References: <322B1EFE.59E2B600@bbn.com> In article <322B1EFE.59E2B600@bbn.com>, Fred Roeber wrote: > > I usually just use the "select" system call to implement reads with > timeouts ... > Also, there can actually be cases where [select()] could say that data > is readable when there isn't actually any data available so you really > should make sure to set the socket up as non-blocking ... If you're using select(), there's no reason (that I'm aware of) for the socket to be non-blocking. > ... and be prepared to get no data back from your read call (as far > as I can tell, this last case can only happen when multiple tasks are > sharing the same socket). Say what?! :) Having multiple tasks simultaneously reading from a single socket is not a good idea. Alex ======================================================================== | Alex Measday Code 521 - Microelectronic Systems Branch | | Integral Systems, Inc. NASA Goddard Space Flight Center | | c.a.measday@ieee.org Greenbelt, MD | ======================================================================== ) http://purl.oclc.org/net/alex/ ( ======================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UDP packet - max size? Date: Wed, 04 Sep 1996 18:27:16 -0700 From: "H.J. Bae" Organization: Peaceful Star Message-ID: <322E2C74.2040@peacefulstar.oakland.ca.us> References: <3221CCFD.794B@boris.msfc.nasa.gov> <502g9o$m7k@idiom.com> Reply-To: hjb@peacefulstar.oakland.ca.us Michael Charnoky wrote: > > First of all, these buffers can only be set to a maximum size of roughly > 52224 bytes. Why the limit? Is this limit universal or only inherent to > VxWorks? (I've seen no documentation...) > The limit in VxWorks is actually something like 52387. This is BSD UNIX legacy (BSD 4.3 Tahoe release, a.k.a net-1). The limit is imposed in sbreserve() routine in uipc_socket2.c. This limit is rather arbitrary and should be changed and/or made tunable by users. Later and eariler versions of BSD UNIX network code had different limits. > Second, sending a UDP packet succesfully required extensive tinkering. I > found that once the receive buffer is set to be twice as large as the send > buffer, the transfer succeeds. Why is this so? No such magic is needed > when setting the TCP bufs... UDP is inherently unreliable. If you want a reliable service, use TCP. - -- H.J. Bae http://peacefulstar.oakland.ca.us mailto:hjb@peacefulstar.oakland.ca.us --------------------------- End of New-News digest ********************** From SJohnson@systran.com Thu Sep 5 06:00:30 1996 From: Steve Johnson Date: Thu Sep 5 06:00:33 PDT 1996 Subject: dosFs operation We are working on a vxWorks block device driver for disc drives. We are using the dosFs file system. The driver is designed to support concurrent read/write requests to multiple drives to optimize throughput. When I spawn a number of tasks to do operations to the discs I observe the value "0x3d0001" in the ERRNO field of the i() command. This value appears to be consistent until each of the tasks terminate. According to the vxWorks module numbering and errno reporting logic this value resolves to the error: "S_objLib_OBJ_ID_ERROR". I cannot find any information as to what condition produces this error. Although, I have obeserved this same return value when doing an "ls()" from the shell on a disc that has no files. Is this expected or is there some condition I have completely overlooked in the driver design ? Any insight into this situation will be greatly appreciated. Thanx, S.L. Johnson Systran Corp. Dayton, Ohio sjohnson@systran.com From legoas_marco_a@space.honeywell.com Thu Sep 5 07:37:47 1996 From: "Legoas Marco A" Date: Thu Sep 5 07:37:50 PDT 1996 Subject: dos FS file size dependency? Is there a file size limitation when using the VxWorks DOS file system? We are noticing a speed deterioration as we are accessing the end of large files (> 250 Megs). We are using file oriented IO (fopen,fseek,fread,fclose,...etc). Also we fopen and fclose the file every time. Marco A. Legoas malegoas@space.honeywell.com From sblachma@aoc.nrao.edu Thu Sep 5 08:17:25 1996 From: Steve Blachman Date: Thu Sep 5 08:17:27 PDT 1996 Subject: Re: dos FS file size dependency? > Is there a file size limitation when using the VxWorks DOS file system? > We are noticing a speed deterioration as we are accessing the end of large > files (> 250 Megs). We are using file oriented IO > (fopen,fseek,fread,fclose,...etc). Also we fopen and fclose the file every > time. Yes, or at least we noticed the same effect when writing files. It depends on cluster size, so if your files have a size limit, you can adjust the clusters so the first speed shelf does not occur until after that limit. We also notice a series of speed drops every X mb. I do not know what causes it though, perhaps some multiple FAT table access. Steve Blachman From dmidkiff@trandes.com Thu Sep 5 13:22:08 1996 From: "DENNIS MIDKIFF" Date: Thu Sep 5 13:22:11 PDT 1996 Subject: 8 Port Serial boards for X86 platform Hello, Could anyone recommend a 8 channel serial board for a 486 PC. I have a project I am working on and want to stay compatable with WindRiver and Tornado. So I want to be able to plug in the Card, compile or include the serial driver and be on my way. Any help will be appreciated. Thanks in advance Dennis midkiff dmidkiff@trandes.com From joergb@wrs.com Thu Sep 5 15:29:32 1996 From: Joerg Bertholdt Date: Thu Sep 5 15:29:34 PDT 1996 Subject: Perl on VxWorks Hi, is anybody aware of a Perl interpreter for VxWorks? Any information is highly appreciated. Thanks in advance Joerg __________________________________________________________________ Joerg Bertholdt joergb@wrs.com Product Marketing Manager OS Wind River Systems +1-510-814-2056 voice Alameda, CA 94501 +1-510-814-2010 fascimile http://www.wrs.com __________________________________________________________________ From KURT.STAUBLI@roche.com Fri Sep 6 00:08:19 1996 From: "NAME \"Staubli Kurt\" 2607" Date: Fri Sep 6 00:08:22 PDT 1996 Subject: VxWork on Motorola 68332 We write real time application for medical analyzers and will use VxWorks on a 68360 in the near future. Another group evaluate a real time operating system for a 68332@24MHz. They control DC motor controllers, step motor controllers, digital I/O, AD Converters, serial communication 375kBaud on that board. Has anybody used VxWorks on a 68332 Target ? What about speed and response performance of VxWorks for such applications ? Are there other Real Time OS for such application ? Kurt Staubli Kurt Staubli c/o Tegimenta AG Forrenstrasse 6343 Rotkreuz Switzerland ++41 799 26 07 kurt.staubli@roche.com From kiran@vms.masd.northgrum.com Fri Sep 6 07:15:03 1996 From: kiran@vms.masd.northgrum.com (Kiran Suwal) Date: Fri Sep 6 07:17:00 PDT 1996 Subject: Executing application code residing in SYS68K RR-2 memory board Hi, Does anyone know how to configure vxWorks such that after booting it executes application code stored in a SYS68K RR-2 memory board EPROMs. My standalone system consists of a FRC40 CPU and a SYS68K RR-2 memory board. Any insight into this will be greatly appreciated. Kiran R. Suwal Northrop Grumman Corp. From MikeMcQuade@connect.com Fri Sep 6 07:53:59 1996 From: Mike McQuade Date: Fri Sep 6 07:54:00 PDT 1996 Subject: Linked List DMA xfers Has anyone setup a linked list DMA xfer, Im trying to set one up using the Newbridge Universe chip. Id love to see an example in C. Thanks. Mike MikeMcQuade@connect.com From anthonyl@atlantis.softeng.lamrc.com Fri Sep 6 13:55:26 1996 From: Anthony Le Date: Fri Sep 6 13:55:28 PDT 1996 Subject: Tonado development environment Hi VxWorks users, My company is ready to move to Tonado using Window NT as development environment. We plan to use the cross compiler run on PC to genenrate embeded code. We like to move away from the SUN. Can anybody make comments about this enviroment. Thanks Anthony Le LAM Research Corp. From slirosi@gic.gi.com Fri Sep 6 14:05:04 1996 From: "Lirosi, Salvatore (HT-MS)" Date: Fri Sep 6 14:05:06 PDT 1996 Subject: zbufSockLib I am having a lot of trouble getting improved performance using zbufSockLib. In receiving UDP frames, the supplied zbuf contains multiple segments. These are the kernel mbufs no doubt. However, I cannot see how non-contiguous data received can be made to work efficiently. Either one has to copy the data out or if not, incur unreasonable overhead in using it. Does anyone have any experiences with utilizing UDP data received within zbufs? Sal From YF.Siu@eng.efi.com Fri Sep 6 16:51:44 1996 From: "Yuet Fung Siu" Date: Fri Sep 6 16:51:46 PDT 1996 Subject: Token-ring driver My colleague is seeking the routing code for token-ring drivers under VxWorks 5.2. Any relevant information is greatly appreciated. YF From daemon@csg.lbl.gov Sat Sep 7 04:00:51 1996 From: daemon@csg.lbl.gov Date: Sat Sep 7 04:00:55 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Sep 7 04:00:48 PDT 1996 Subject: PC BIOS extension creation HELP ?? Subject: Offset for VxGDB ------------------------------------------------------- Newsgroups: comp.arch.embedded,comp.os.vxworks,comp.sys.intel Subject: PC BIOS extension creation HELP ?? Date: 6 Sep 1996 11:30:10 GMT From: David Costain <100072.3624@CompuServe.COM> Organization: VG Organic Message-ID: <50p202$7d1$1@mhade.production.compuserve.com> I am currently trying to create a vxWorks bootrom for the 486/586 platform. To this end, I am working on a BIOS extension-like EPROM that contains the BIOS identifier header, and an in-house loader to load vxWorks into RAM, prior to executing the code. My BIOS header format is as follows: byte0: 55 byte1: AA byte2: 01 /* 1 x 512 byte block */ byte3: start of loader code Is this format correct ? I have a couple of technical references which have conflicting formats. The above format looks the same as my video BIOS. I am aware of commercially available bootrom cards, but we would like to put the EPROM on one of our in-house ISA cards. Any information gratefully accepted ! Thanks, Dave Costain Micromass UK Ltd dave.costain@micromass.co.uk --------------------------- Newsgroups: comp.os.vxworks Subject: Offset for VxGDB Date: 6 Sep 1996 19:16:55 GMT From: ashrafis@nt.com (Siamak Razzaghe-Ashrafi) Organization: Northern Telecom Inc., Mission Park Message-ID: <50ptb7$hiq@nrchh52.rich.nt.com> Is there a way to load a dynamicly linked Dot.o without symbols. Then use the sym command to load the symbol table into VxGDB and get VxGDB to know the offset and debug the program? ashrafis@nt.com everThing, everyWhere, anyTime. - -- The thing is it so I believe it is the thing. -SubZero --------------------------- End of New-News digest ********************** From leonid@rst.co.il Sat Sep 7 23:37:32 1996 From: leonid@rst.co.il (Leonid Rosenboim) Date: Sat Sep 7 23:37:38 PDT 1996 Subject: Re: PC BIOS extension creation HELP ?? > Subject: PC BIOS extension creation HELP ?? > Date: 6 Sep 1996 11:30:10 GMT > From: David Costain <100072.3624@CompuServe.COM> > Organization: VG Organic > Message-ID: <50p202$7d1$1@mhade.production.compuserve.com> > I am currently trying to create a vxWorks bootrom for the 486/586 > platform. To this end, I am working on a BIOS extension-like > EPROM that contains the BIOS identifier header, and an in-house > loader to load vxWorks into RAM, prior to executing the code. > ... > Is this format correct ? I have a couple of technical references > which have conflicting formats. The above format looks the same > as my video BIOS. Basically the format you described is ok, except there is also a checksum that should appear at te end of the block(s) according to the number of blocks specified in the 3rd byte. Some BIOS check the checksum, some dont. So if you want to be sure your BOOTROM will work with different BIOS versions, be sure to have the checksum too. Luckily, the lonker supplied with Tornado knows how to do it. Use the "-b" option, and it will add the signature in front and the checksum in the back, and will automatically calculate the 3rd byte and checksum for you. The Tornado assembler will give you trouble though trying to compile 16-bit code. Beware. We have been through this when creating a BOOTABLE Flash File System. Essentially it is an ISA (or PC104) Flash card, we reserve its first 64K block to put a bootstrap code which will enter protected mode and load a specific file from the rest of the Flash card formatted as a Flash File System in to RAM and execute it. The result is pretty cute - you plug in the Flash card, turn power on - and voila! standalone VxWorks greets you nicely. Then of course you can replace the file just booted, or add other files as it is a fully functional file system. From daemon@csg.lbl.gov Sun Sep 8 04:01:06 1996 From: daemon@csg.lbl.gov Date: Sun Sep 8 04:01:08 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Sep 8 04:01:01 PDT 1996 Subject: Re: Problems with select() and recv(). Subject: Re: rpcgen for vxworks Subject: Re: VxWork on Motorola 68332 ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with select() and recv(). Date: 7 Sep 1996 13:14:01 GMT From: tfm@sei.cmu.edu (Theodore Marz) Organization: The Software Engineering Institute Message-ID: <50rsep$89o@news.sei.cmu.edu> References: <3230A5B4.3C6F@htc.honeywell.com> In article <3230A5B4.3C6F@htc.honeywell.com>, Lee Graba writes: |> We are using VxWorks 5.1 on Force 40 VME boards (68040). In part of our application, a |> task running at a priority of 33, goes into a select loop, where it waits on a few |> sockets on which connections already exist. Its basic task is below: |> |> for (;;) |> { |> select_cnt=select(readfds, 0, 0, 0, 10 seconds); |> if (select_cnt > 0) |> { |> find the FD with stuff to read |> read the stuff into a buffer using 2 calls to recv |> spawn a task at lower priority to process the buffer |> } |> else if (select_cnt < 0) |> { |> get errno and print error string |> } |> } |> |> After the select, I have use ioctl to check the bytes to read, and compared with the |> number of bytes read by the recv calls, and these add up. The task that is spawned to |> process the buffer owns the buffer, and the parent task doesn't touch it after the task |> is spawned. |> |> The problem is, we are getting many errors from select after first getting a good |> select, i.e. we successfully read the proper number of bytes from the fd, spawn a task, |> and when we then loop around to call select again, it immediately returns select_cnt = |> -1, and errno = 0xD0003, which strerror_r reports as having to do with an |> INVALID_FILE_DESCRIPTOR. If I put a taskDelay of 100 ms somewhere in the select loop, I |> no longer get these messages. |> |> Anybody have any idea what is causing select to return the error, and why timing has |> anything to do with it? |> |> |> -- |> =========================================================================== |> Lee Graba = |> Honeywell Technology Center = |> 3660 Technology Drive = |> Minneapolis, MN 55418 = |> (612) 951-7503 = |> Graba_Lee@htc.honeywell.com = |> graba@htc.honeywell.com = |> =========================================================================== I not sure what is happening, but here are some thoughts: What priority is the spawned process running at, relative to the pri 33 data detection routine? If the data consumer is not running at higher priority than the data detection routine, you could be re-cycling into the data detection while the data is in the process of being consumed. The file descriptor (socket) may be locked by the consumer task, and so when you try and select(), it fails. The 100mS delay you add pretty much ensures that the consumer task has completed reading the data before another select() gets called on the same fd/socket. Ted Marz --------------------------- Newsgroups: comp.os.vxworks Subject: Re: rpcgen for vxworks Date: Fri, 06 Sep 1996 00:15:15 GMT From: traymond@pcix.com Organization: Crafted Smalltalk Message-ID: References: <199609031712.KAA04659@daisy.sp.trw.com> Reply-To: traymond@pcix.com (Terry Raymond) In <199609031712.KAA04659@daisy.sp.trw.com>, "Carol L. Lugo" writes: > >Hi ! >Looking for a rpcgen for VxWorks...does anyone know of anything ?? It is in the unsupported/rpc4.0 directory. But, I have used the SUN and HP rpcgen to generate code for vxworks with no problem. =========================================================== Terry Raymond Smalltalk Professional Debug Package Crafted Smalltalk *Breakpoints* and *Watchpoints* for 19 Tilley Ave. VW and Envy Newport, RI 02840 (401) 846-6573 traymond@pcix.com Crafted Smalltalk =========================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWork on Motorola 68332 Date: Sat, 07 Sep 1996 05:31:04 GMT From: johnd@reflex-tech.co.uk (John Dalton) Organization: Reflex Technology Message-ID: <50r1g2$2pf@luggage.rednet.co.uk> References: <8619070906091996/A15811/TROD03/11A932471200*@MHS> Reply-To: johnd@reflex-tech.co.uk "NAME \"Staubli Kurt\" 2607" wrote: >We write real time application for medical analyzers and will use VxWorks on a >68360 in the near future. Another group evaluate a real time operating system >for a 68332@24MHz. They control DC motor controllers, step motor controllers, >digital I/O, AD Converters, serial communication 375kBaud on that board. Has >anybody used VxWorks on a 68332 Target ? What about speed and response >performance of VxWorks for such applications ? Are there other Real Time OS for >such application ? >Kurt Staubli You'd get better performance and easier programming with OSE from Enea OSE Systems, Sweden - the kernel is only 7KB in production and interrupt handling is very fast indeed. Look at these figs : Eg. event timings measured on MC68020 at 25MHz :- where n is a system-state dependency, eg. number of messages in queue, number of ready processes, message length etc. (The table is tabbed every 10 chars) Function OSE pSOS VRTX32 OS-9 Alloc memory 9 37+2n 29+3n 9+?*n Free memory 5 32+2n 33+3n 434 Create Process 29 87 73 1190+?*n Send msg, no swap 13 46+2n 40 77+?n Send msg & swap in 20 66+2n 55 134+?n - -> implies process swap 7 20 16 56+?n Receive message 8 31+2n 29 ? Disable Interrupts 4 9 10 ? Initialise INT (2) - (15) 9+6n Quit INT to next INT 3 ? 12 ? Set semaphore from INT 3 38 41 ? Quit INT to background 20 39 37+21n ? (Times in microseconds. ? = couldn't measure value reliably) Note that times for OSE are fully deterministic : there is no system-state dependency, so you can easily calculate worst-case timings. The only system-state-dependent function in OSE is a selective receive. The other thing to note is that the OSE timings include executing run-time checks on system integrity (stack overwrites, message ownership conflicts, memory corruption etc) which the others don't provide at all, and that the application programs you write for OSE have less code than the others to execute because of the OSE error handler system architecture integral to the kernel. Check them out on http://www.enea.se/ose John >Kurt Staubli >c/o Tegimenta AG >Forrenstrasse >6343 Rotkreuz >Switzerland > >++41 799 26 07 >kurt.staubli@roche.com --------------------------- End of New-News digest ********************** From uchenick@tis4000.com Mon Sep 9 05:43:29 1996 From: Gordon Uchenick Date: Mon Sep 9 05:43:31 PDT 1996 Subject: PD TCP/IP Sources Recently someone posted the location of the public domain sources to TCP on gatekeeper.dec.com. Could that person please repeat that post, or could someone tell me where else to find the TCP source? TIA, Gordon Uchenick uchenick@tis4000.com 410-581-0422 From davida@laplace.idec.sdl.usu.edu Mon Sep 9 08:19:55 1996 From: davida@laplace.idec.sdl.usu.edu (David Anderson) Date: Mon Sep 9 08:19:56 PDT 1996 Subject: VxWorks 5.3 and Tornado We are considering upgrading from vxWorks 5.2 to vxWorks 5.3 and Tornado. We have about four to five years experience with vxWorks but to help us up the learning curve with Tornado we are considering taking the Tornado-5.3 class at Wind River. My question is has anyone out there in vxWorks land taken the class and if so do you feel that the information and training you receive is worth the expense of the class? We are particularly interested in the the current status of Tornado, its C++ development tools and environment, information on tools.h++ and other C++ libraries etc. Thanks in advance for any comments/advice, Dave. From jfd@rtp.co.uk Mon Sep 9 10:18:44 1996 From: Julian Day Date: Mon Sep 9 10:18:46 PDT 1996 Subject: Ada Bindings for VxWorks??? Hi, Are you developing for Ada on VxWorks? If so, have you done any work on bindings to VxWorks libraries, e.g Networking, timers etc. If so, can you let me know please. I'll collate results and feed back to the exploder, the information I receive. TVMIA, Julian --------------------------------------------------------------- Julian Day Real Time Products Ltd. Unit 8, Holt Court, e-mail: jfd@rtp.co.uk Aston Science Park, Jennens Road, tel: +44 121 333 6955 Birmingham. B7 4EJ. England. fax: +44 121 333 5433 --------------------------------------------------------------- From 75353.1150@compuserve.com Mon Sep 9 18:57:12 1996 From: "William G. St.Clair" <75353.1150@compuserve.com> Date: Mon Sep 9 18:57:14 PDT 1996 Subject: Seeking VxWorks expert Start-up, developing highly innovative mobile Internet & workgroup product seeks VxWorks expert with experience in C++ Booch Components and/or CORBA ORB design. Does this ring any bells? Let me know. Bill St.Clair, Greystone Peripherals From daemon@csg.lbl.gov Tue Sep 10 04:00:56 1996 From: daemon@csg.lbl.gov Date: Tue Sep 10 04:00:58 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Sep 10 04:00:53 PDT 1996 Subject: Re: PC BIOS extension creation HELP ?? ------------------------------------------------------- Newsgroups: comp.arch.embedded,comp.os.vxworks,comp.sys.intel Subject: Re: PC BIOS extension creation HELP ?? Date: Fri, 6 Sep 1996 19:02:43 GMT From: kam@plex.nl (Gerald van Kampen) Organization: Plex -- a public access Internet provider -- Message-ID: References: <50p202$7d1$1@mhade.production.compuserve.com> <323044F9.31FD@shellx.best.com> Sender: news@plex.nl (Usenet News) In article <323044F9.31FD@shellx.best.com>, Pascal Dornier wrote: >David Costain wrote: >> >> I am currently trying to create a vxWorks bootrom for the 486/586 >> platform. To this end, I am working on a BIOS extension-like >> EPROM that contains the BIOS identifier header, and an in-house >> loader to load vxWorks into RAM, prior to executing the code. >> >> My BIOS header format is as follows: >> >> byte0: 55 >> byte1: AA >> byte2: 01 /* 1 x 512 byte block */ >> byte3: start of loader code >> >> Is this format correct ? I have a couple of technical references >> which have conflicting formats. The above format looks the same >> as my video BIOS. >> >> I am aware of commercially available bootrom cards, but we would >> like to put the EPROM on one of our in-house ISA cards. > >You also need to make sure that the 8 bit checksum (add up all bytes >in the EPROM) is 0. > The first two bytes of a ROM extension are always 55h and aah. The third byte specifies the number of 512-byte blocks to scan when performing a checksum of the ROM BIOS extension. The BIOS multiplies the number by 512 and scans that many bytes starting with the first byte of the header. If all the bytes add up to zero, then the ROM BIOS extension is actually called during initialization. So the last byte of the last block must be calculated in such a way that this condition will be met. The next sample code gives an impression how a BIOS call's a ROM BIOS extension (in this case the VGA BIOS at C0000h-C8000h). I adapted this code from the public domain BIOS source code "bios-asm.zip". My goal was to initialize the VGA BIOS on an embedded PC without BIOS. (BTW. This didn't work because the VGA BIOS execution depends on previous BIOS initializations like equipment list service etc.). - ------------------------start------------- ER_ROM EQU 20h ; Bad ROM in expansion area, bad checksum ASSUME CS:cseg,DS:cseg cseg SEGMENT PARA PUBLIC USE16 'CODE' public main main proc near MOV DX,0C000h ; (VGA-)ROM segment start PUSH DS FNDROM: MOV DS,DX ; Load ROM segment XOR BX,BX ; ...ID offset MOV AX,[BX] ; Read the ROM id CMP AX,0AA55h JNZ NXTROM ; ...not valid ROM MOV AX,40h MOV ES,AX MOV AH,0 MOV AL,[BX+2] ; Get ROM size (bytes * 512) MOV CL,5 SHL AX,CL ; Now ROM size in segments ADD DX,AX ; ...add base segment MOV CL,4 SHL AX,CL ; ROM address in bytes MOV CX,AX ; ...checksum requires CX CHK_01: MOV AL,0 ; ...zero checksum ADDBYT: ADD AL,[BX] ; Add byte to checksum INC BX ; ...BX --> next byte LOOP ADDBYT ; ...loop until done OR AL,AL ; Set condition codes JNZ BADROM ; ...bad ROM PUSH DX MOV Word ptr ES:67h,3 ; Offset for ROM being setup MOV ES:69h,DS ; Segment for ROM being setup CALL Dword ptr ES:67h ; ...call ROM initialization POP DX JMP short FND_01 BADROM: OR Byte ptr ES:15h,ER_ROM ; ROM present, bad checksum NXTROM: ADD DX,80h ; Segment for next ROM FND_01: CMP DX,0C800h ; End of (VGA-)ROM space JL FNDROM ; ...no, continue POP DS ; Exit program to DOS (should be removed, this is for testing the code ; stand-alone as a DOS program. Normally the code is part of real-mode ; initialization of an embedded PC without BIOS. MOV AL,0 ; Exit code MOV AH,4Ch ; Exit to DOS function INT 21h main endp cseg ends end main - --------------------end------------------- Gerald van Kampen kam@plex.nl --------------------------- End of New-News digest ********************** From atl@adaclabs.com Tue Sep 10 06:46:15 1996 From: atl@adaclabs.com (Anthony Le) Date: Tue Sep 10 06:46:17 PDT 1996 Subject: SUBSCRIBE SUBSCRIBE Thank you From malcolm.cifuentes@nautronix.com.au Tue Sep 10 20:16:06 1996 From: malcolm.cifuentes@nautronix.com.au (Malcolm Cifuentes) Date: Tue Sep 10 20:16:08 PDT 1996 Subject: VxWorks 5.2 C++ Hi, I have the following problem :- I have a fresh installation of vxworks 5.2 with gcc. I wish to do development in C++ but I do not have any libg++ libraries, I am not really sure why? I have tried compiling libg++ but I am having trouble tracking down a libg++ version early enough which will compile with the gcc(cygnus-2.2.3.1) compiler. I really would like to compile the latest version of libg++ with the compiler. Any help appreciated, thanks! -- Malcolm Cifuentes ( Nautronix LTD malcom.cifuentes@nautronix.com.au PH : +61-09-430-5900 FAX : +61-09-430-5901 ) From daemon@csg.lbl.gov Wed Sep 11 04:00:27 1996 From: daemon@csg.lbl.gov Date: Wed Sep 11 04:00:29 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Sep 11 04:00:23 PDT 1996 Subject: Re: How to detect end of socket connection Subject: Re: rpcgen for vxworks Subject: Re: Tonado development environment Subject: SMTP (Mail) client Subject: Migration From VAXELN ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to detect end of socket connection Date: 3 Sep 1996 19:58:59 GMT From: alexm@vlsi.gsfc.nasa.gov (Alex Measday) Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <50i2m3$oil@post.gsfc.nasa.gov> References: <507e7g$j6v@fu-berlin.de> In article <507e7g$j6v@fu-berlin.de>, Heinz Junkes wrote: > >How can i check that a connection is still alive ??? > If 1. select() says the socket is ready for reading and 2. read() (or an ioctl(FIONREAD)) returns zero, your connection is broken. Note that you must perform the select() before the read() or ioctl() since data may be received in between the calls. Also see the "Socket Programming Frequently Asked Questions" at: http://www.interlog.com/~vic/sock-faq/ ======================================================================== | Alex Measday Code 521 - Microelectronic Systems Branch | | Integral Systems, Inc. NASA Goddard Space Flight Center | | c.a.measday@ieee.org Greenbelt, MD | ======================================================================== ) http://purl.oclc.org/net/alex/ ( ======================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: rpcgen for vxworks Date: 10 Sep 1996 11:49:18 GMT From: kadionik@aramis (Patrice KADIONIK (IXL - CNRS URA 846 - FRANCE)) Organization: CRIBX1 , Universite de Bordeaux I , France Message-ID: <513kju$8om@news.u-bordeaux.fr> References: <199609031712.KAA04659@daisy.sp.trw.com> Carol L. Lugo (clugo@emu.sp.trw.com) wrote: : Hi ! : Looking for a rpcgen for VxWorks...does anyone know of anything ?? : Carol - -- Use the rpcgen binary of your development environment in order to generate all the C structures and codes (for example the Sun rpcgen command). Rememenber that in the VXWorks side you must call at first and one time the routine rpcTaskInit() in your code. Cheers; Pat. P. Kadionik Laboratory of Microelectronics - IXL University of Bordeaux I Bordeaux - FRANCE \^ ^/ )@ @( +---------oOO--(_)-------------+ | Patrice Kadionik | | kadionik@ixl.u-bordeaux.fr | +-----------------Oooo--oOO----+ oooO ( ) ( ) ) / \ ( (_/ \_) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tonado development environment Date: Tue, 10 Sep 1996 14:16:34 -0700 From: Frank Weissig Organization: STEP Technology, Inc. Message-ID: <3235DAB2.204A@steptech.com> References: <9609062055.AA00864@softeng.lamrc.com> <3235D7A0.40B2@steptech.com> Reply-To: frank@steptech.com Frank Weissig wrote: > > - The architecture of the Tornado environment is generally well thought > out. The "division of labor" between the host and target is sensible > and > allows effective work to be done even on somewhat anemic hosts. > The host-based shell is particularly handy. > Oops! That was supposed to read "anemic TARGETS". P.S. Anthony, I think your email address is incorrect. - -- Frank Weissig Senior Staff Engineer STEP Technology, Inc. frank@steptech.com --------------------------- Newsgroups: comp.os.vxworks Subject: SMTP (Mail) client Date: Tue, 10 Sep 1996 14:33:07 GMT From: lyorg@Scorpio.Com (Lyor Goldstein) Organization: Scorpio Communications Ltd. Message-ID: Reply-To: lyorg@Scorpio.Com Sender: news@Scorpio.Com Does anyone konw of a simple source for a "mail" client (i.e. send some simple text mail to a known server thru the Ethernet adaptor) ? - --- ***************************************************************** * Lyor Goldstein. Email : lyorg@scorpio.com * * Scorpio Communications Ltd. Phone : 972-3-5339654 ext 115 * * Software group Fax : 972-3-5339518 * * 1c Yoni Netanyahu st. * * Or Yehuda 60250 * * P.O.Box 564, Israel * ***************************************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Migration From VAXELN Date: Sun, 08 Sep 1996 14:56:45 -0400 From: Brian & Karen Bell Organization: At Home... Message-ID: <323316ED.736E@mdn.net> Reply-To: bgbell@mdn.net We're obliged (for hardware obsolescence reasons) to port a communication front-end system from an rtVAX/VAXELN platform to something else; and like all projects these days -- or at least so it seems -- we're on a very tight time line to get the job done. In addition to moving the application (~70,000 SLOC of VAX C) we need to replace a custom UNIBUS interface card with something a little more current and lot more broadly supported (like VME, or PMC if it will fit). In the spirit of 'getting the job done' we're considering using DEC's VAXELN emulation library running over VxWorks. While everything I've read about VxWorks (and also about Wind River Systems) is encouraging (quality, service etc) there is one, somewhat disturbing catch; DEC supports the VAXELN emulation library _only_ on Alpha targets. This is of course not surprising -- since DEC is the dominant vendor of Alpha SBCs -- but such a conversion doesn't take us as far as we'd like to go in moving from a proprietary/single-source environment to one in which healthy market competition works more to our advantage. We've not ruled out converting the source to replace all VAXELN dependencies with VxWorks equivalents, but obviously on a source of that size this could be a substantial undertaking. A two-step process in which we first make the emulation-enabled to conversion to an Alpha target and later remove the VAXELN (and Alpha) dependencies is also being considered. I have two questions for this forum: 1) Has anyone undertaken such project before? If so, do you have any hard-won advice you'd be willing to share? 2) Is there a message implicit in Wind River's failure to list Alpha among the supported targets for its most recent (V5.2) offering? (As near as I can tell, the newest Alpha-supported release is some two major rev behind at V3.2.) As much I admire DEC's technology (including Alpha) I have no desire to be dependent on a _any_ vendor's products if there's a way to avoid it. Any advice will be greatly appreciated. Brian Bell bgbell@dow.com Software Engineering Associate Global Process Control The Dow Chemical Company --------------------------- End of New-News digest ********************** From rip@wrsec.fr Wed Sep 11 07:42:43 1996 From: rip@wrsec.fr (Rip richard Williamson ) Date: Wed Sep 11 07:42:45 PDT 1996 Subject: Re: comp.os.vxworks newsdigest Brian Bell said: ? ? I have two questions for this forum: ? ? 1) Has anyone undertaken such project before? If so, do you have ? any hard-won advice you'd be willing to share? ? ? 2) Is there a message implicit in Wind River's failure to list ? Alpha among the supported targets for its most recent (V5.2) ? offering? (As near as I can tell, the newest Alpha-supported ? release is some two major rev behind at V3.2.) ? ? As much I admire DEC's technology (including Alpha) I have no ? desire to be dependent on a _any_ vendor's products if there's ? a way to avoid it. ? ? Any advice will be greatly appreciated. ? Don't have any hard-won experience with that, sorry. This is about point 2: DEC's version of VxWorks (v3.2) is equivalent to WRS v5.1.1 . When they purchased the rights to use/sell it, they decided to modify the version numbering for one reason or another. For more information along these lines, contact DEC for information about 3.2 for the Alpha, or WRS for Tornado (5.3) for most other architectures (those that we support, that is). Regards, Richard Wed Sep 11 15:57:07 MET DST 1996 ----------------------------------------------------------------------- Richard Williamson rip@wrsec.fr Technical Support Engineer vox: +33 (1) 60.92.63.13 Wind River Systems - EC fax: +33 (1) 60.92.63.15 ----------------------------------------------------------------------- INCLUDED MESSAGES: 57: Change is inevitable, except from a vending machine. --------------------------------------------------END INCLUDED MESSAGES From jimb@rational.com Wed Sep 11 11:28:10 1996 From: Jim Bushway Date: Wed Sep 11 11:28:12 PDT 1996 Subject: MVME1604 board load problem When trying to load the boot file on our MVME1604 board we are getting the following. Any ideas??? Network Boot File load in progress... To abort hit Bytes Received =&622160, Bytes Loaded =&622160 Bytes/Second =&207386, Elapsed Time =3 Second(s) Residual-Data Located at: $00F80000 Exception: Machine Check SRR0 =00013078 SRR1 =000C3040 Vector-Offset =00200 IP =00013078 MSR =00003040 CR =04000000 FPSCR =00000000 R0 =1D203040 R1 =DFE00018 R2 =00000000 R3 =FFFFFFFF R4 =00005000 R5 =00000000 R6 =00000000 R7 =00000000 R8 =00000000 R9 =00000000 R10 =00000000 R11 =00000000 R12 =00000000 R13 =00000000 R14 =00000000 R15 =00000000 R16 =00000000 R17 =00000000 R18 =00000000 R19 =00000000 R20 =00000000 R21 =00000000 R22 =00000000 R23 =00000000 R24 =00000000 R25 =00000000 R26 =00000000 R27 =00000000 R28 =00000000 R29 =00000000 R30 =00000000 R31 =1D203040 SPR0 =00000000 SPR1 =00000000 SPR8 =1D203040 SPR9 =00000000 00013078 4E800020 BCLR 20,0 From csuka@desy.de Thu Sep 12 01:03:09 1996 From: csuka@desy.de (Gabor Csuka) Date: Thu Sep 12 01:03:11 PDT 1996 Subject: Re: MVME1604 board load problem > > Submitted-by jimb@rational.com Wed Sep 11 11:28:10 1996 > Submitted-by: Jim Bushway > > > When trying to load the boot file on our MVME1604 board > we are getting the following. Any ideas??? > > > Network Boot File load in progress... To abort hit From this line - it seems to me you are using the original motrola Bug - not the VxWorks boot prom, does it ? -- Gabor CSUKA DESY, Germany csuka@heras1.desy.de From daemon@csg.lbl.gov Thu Sep 12 04:00:41 1996 From: daemon@csg.lbl.gov Date: Thu Sep 12 04:00:44 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 12 04:00:37 PDT 1996 Subject: compiling kernel .c files with C++? Subject: Re: PC BIOS extension creation HELP ?? Subject: Re: UDP packet - max size? Subject: Needed: VxWorks 5.3 driver for DP83934 lan chip Subject: Torando/a.out to iee695 Subject: Problems with select() and recv(). ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: compiling kernel .c files with C++? Date: Tue, 10 Sep 1996 07:01:05 -0700 From: Todd Hoff Organization: Possibility Outpost Message-ID: <323574A1.1A3C@possibility.com> Has anyone compiled the kernel c files using C++? Any tips or traps? We would like to call C++ code directly without performance reducing c wrappers but don't have the bandwidth to waste if it won't work. Thanx - ------------------------------------------------------------- tmh@possibility.com | The loyalty of small men can be http://www.possibility.com | bought cheaply, for greed has no | pride. - Michael Kube-McDowell --------------------------- Newsgroups: comp.arch.embedded,comp.os.vxworks,comp.sys.intel Subject: Re: PC BIOS extension creation HELP ?? Date: Mon, 9 Sep 1996 06:40:38 GMT From: ken@syd.dit.CSIRO.AU (Ken Yap) Organization: CSIRO Division of Information Technology Message-ID: <510e56INNbvv@degas.syd.dit.csiro.au> References: <50p202$7d1$1@mhade.production.compuserve.com> <323044F9.31FD@shellx.best.com> Reply-To: ken@syd.dit.csiro.au (Ken Yap) Sender: ken@syd.dit.csiro.au (Ken Yap) |>> My BIOS header format is as follows: |>> |>> byte0: 55 |>> byte1: AA |>> byte2: 01 /* 1 x 512 byte block */ |>> byte3: start of loader code |>> |>> Is this format correct ? I have a couple of technical references |>> which have conflicting formats. The above format looks the same |>> as my video BIOS. |>> |>> I am aware of commercially available bootrom cards, but we would |>> like to put the EPROM on one of our in-house ISA cards. |> |>You also need to make sure that the 8 bit checksum (add up all bytes |>in the EPROM) is 0. |> | |The first two bytes of a ROM extension are always 55h and aah. |The third byte specifies the number of 512-byte blocks to scan when performing |a checksum of the ROM BIOS extension. The BIOS multiplies the number by 512 |and scans that many bytes starting with the first byte of the header. If all |the bytes add up to zero, then the ROM BIOS extension is actually called |during initialization. So the last byte of the last block must be calculated |in such a way that this condition will be met. Yes, that's correct, but the checksum doesn't have to be anywhere in particular, you just have to make all the bytes sum to 8 bit zero. It may be that it's convenient to make the last byte in the block the checksum, or rather, the value that makes the checksum zero. Typically this "check" byte is altered by an auxiliary program scanning the binary image just before burning an EPROM. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UDP packet - max size? Date: 11 Sep 1996 14:23:30 GMT From: David Perkinson Organization: TRW, Inc. Message-ID: <3236CBA4.41C6@boris.msfc.nasa.gov> References: <3221CCFD.794B@boris.msfc.nasa.gov> <502g9o$m7k@idiom.com> <322E2C74.2040@peacefulstar.oakland.ca.us> H.J. Bae wrote: > > > The limit in VxWorks is actually something like 52387. This is BSD UNIX > legacy (BSD 4.3 Tahoe release, a.k.a net-1). The limit is imposed > in sbreserve() routine in uipc_socket2.c. This limit is rather > arbitrary > and should be changed and/or made tunable by users. Later and eariler > versions of BSD UNIX network code had different limits. > > > UDP is inherently unreliable. If you want a reliable service, use TCP. > > -- > H.J. Bae > http://peacefulstar.oakland.ca.us > mailto:hjb@peacefulstar.oakland.ca.us I increased udp_sendspace to 5*1024 and set udp_recvspace to twice the send space. This fixed my UDP packet problem. I am sending data to a device that only talks using these large UDP packets, so I had to make UDP work. - -- =============================================================================== David Perkinson, TRW Opinions that may be expressed in this email perkinsond@boris.msfc.nasa.gov are my own and not neccessarily that of TRW. =============================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Needed: VxWorks 5.3 driver for DP83934 lan chip Date: 11 Sep 1996 15:58:55 GMT From: elw@lvld.hp.com (Eddie Williamson) Organization: Hewlett Packard Loveland (Mountain IT) Site Message-ID: <516njv$ov0@hplvejl.lvld.hp.com> Does anyone have anyone have a driver for sale/free :) for the National DP83934 lan chip? We will be using VxWorks 5.3 Tornado. Thanks for any help. ______________________________________________________________________ Hewlett-Packard Eddie L. Williamson Jr. Manufacturing Test Division elw@lvld.hp.com 815 SW 14th Street, MS AU100 (970) 679-3674 Loveland, CO --------------------------- Newsgroups: comp.os.vxworks,comp.sys.m68k Subject: Torando/a.out to iee695 Date: Wed, 11 Sep 1996 14:02:50 -0500 From: Steve Woloschek Organization: GE Medical Systems Message-ID: <32370CDA.271E@med.ge.com> I really need to get my hands on a conversion program for a.out to iee695 for the Windows NT Tornado environment. Or perhaps a sensible gnu ld68k program that can output ieee695. Can anybody help? Steve Woloschek --------------------------- Newsgroups: comp.os.vxworks Subject: Problems with select() and recv(). Date: Fri, 06 Sep 1996 17:29:08 -0500 From: Lee Graba Organization: Honeywell Technology Center Message-ID: <3230A5B4.3C6F@htc.honeywell.com> We are using VxWorks 5.1 on Force 40 VME boards (68040). In part of our application, a task running at a priority of 33, goes into a select loop, where it waits on a few sockets on which connections already exist. Its basic task is below: for (;;) { select_cnt=select(readfds, 0, 0, 0, 10 seconds); if (select_cnt > 0) { find the FD with stuff to read read the stuff into a buffer using 2 calls to recv spawn a task at lower priority to process the buffer } else if (select_cnt < 0) { get errno and print error string } } After the select, I have use ioctl to check the bytes to read, and compared with the number of bytes read by the recv calls, and these add up. The task that is spawned to process the buffer owns the buffer, and the parent task doesn't touch it after the task is spawned. The problem is, we are getting many errors from select after first getting a good select, i.e. we successfully read the proper number of bytes from the fd, spawn a task, and when we then loop around to call select again, it immediately returns select_cnt = - -1, and errno = 0xD0003, which strerror_r reports as having to do with an INVALID_FILE_DESCRIPTOR. If I put a taskDelay of 100 ms somewhere in the select loop, I no longer get these messages. Anybody have any idea what is causing select to return the error, and why timing has anything to do with it? - -- =========================================================================== Lee Graba = Honeywell Technology Center = 3660 Technology Drive = Minneapolis, MN 55418 = (612) 951-7503 = Graba_Lee@htc.honeywell.com = graba@htc.honeywell.com = =========================================================================== --------------------------- End of New-News digest ********************** From 75353.1150@compuserve.com Thu Sep 12 06:30:03 1996 From: "William G. St.Clair" <75353.1150@compuserve.com> Date: Thu Sep 12 06:30:05 PDT 1996 Subject: Rogue Wave/Corba/Active X experts Start-up, developing highly innovative mobile Internet & workgroup product seeks VxWorks expert with experience in C++ Booch Components and/or CORBA ORB design. Active X (DCOM) is a plus. Let me know. Bill St.Clair, Greystone Peripherals 410 761-6073 (phone or fax) From ahw@clalf1.mps.ohio-state.edu Thu Sep 12 07:50:56 1996 From: Andreas Wolf Date: Thu Sep 12 07:50:57 PDT 1996 Subject: iostream library in VxWorks Hi there, we are using Tornado 1.0/VxWorks 5.3 with the GNU C++ compiler and found that the standard iostreams object library is unbundled. One has to buy the (optional?) Wind Foundation Class, to do a simple cout << "Hello World" << endl; Is there any reason why the objects provided in the GNU libg++ source should not just work? Did anybody try this? What's the catch? -- ---------------------------------------------------------------------- ** Andreas H. Wolf ****** ******* * * ** The Ohio State University * * * * * ** Smith Lab * * * * * ** 174 West, 18th Ave * * ****** * * ** Columbus, OH 43210-1106 * * * * * ** Phn: (614) 292-1530 * * * * * ** FAX: (614) 292-8261 ****** ******* ****** ** E-Mail: ahw@mps.ohio-state.edu ** WWW: http://www-physics.mps.ohio-state.edu/~ahw/ From wong@vms.masd.northgrum.com Thu Sep 12 14:42:02 1996 From: wong@vms.masd.northgrum.com (J. Pierre Wong) Date: Thu Sep 12 14:42:54 PDT 1996 Subject: PowerPC VME Boards Greetings, I'm looking for information about who makes PowerPC 603 and 604 VME boards (that also use vxWorks, of course) in the 160~200MHz range. If you know of any vendors that will have them on the market by January 97, could you let me know who they are and a possible contact. Thanks for any information, Pierre Wong Vehicle Management Technology Laboratory Northrop Grumman Corp. Tel: 310.332.6840 Fax: 310.644.3652 Email: wong@vms.masd.northgrum.com From intrepid!alberts@uunet.uu.net Fri Sep 13 02:12:25 1996 From: intrepid!alberts@uunet.uu.net (Ed Alberts x7819) Date: Fri Sep 13 02:12:27 PDT 1996 Subject: Re: PowerPC VME Boards PowerPC Boards: VI computer "http://www.vicomp.com/" is releasing a single slot VME 166mhz or 200mhz VI604B board early next month. I currently have their VI604A(100 mhz) board. Ed Alberts alberts@eci-cec.com From daemon@csg.lbl.gov Fri Sep 13 04:00:40 1996 From: daemon@csg.lbl.gov Date: Fri Sep 13 04:00:43 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Sep 13 04:00:37 PDT 1996 Subject: Re: Tonado development environment Subject: DEC/VxWORKS KNOW HOW IN AUSTRALIA ? Subject: Signal Handler Context Subject: Purify-like tool Subject: Re: Tonado development environment Subject: Re: Xterminal emulator for PC ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tonado development environment Date: Thu, 12 Sep 1996 13:18:36 +0200 From: Casey Crellin Organization: CCII Systems Message-ID: <3237F18C.766E4098@ccii.co.za> References: <9609062055.AA00864@softeng.lamrc.com> <3235D7A0.40B2@steptech.com> Frank Weissig wrote: > - The actual implementation on NT needs a bit of work. It has the feel > of a "1.0 product." I've encountered quite a few problems, particularly > with the debugger GUI. I find myself using the gdb command line > interface > most of the time as a result. The on-line help is mediocre at best, and > the > built-in editor is pathetic. I am using the Windows 95 version, the above holds true for it too. Additional problems with windows stability and zombie processes of gdbi386 and something called runCmd and WinoldApp added. The command line length on dos is 128 bytes so forget long command lines for make or run another version of make... > - I've been pretty much underwhelmed with Wind River tech support. > They've > blown off all of my bug reports because I'm running NT 4.0, for which > Tornado hasn't been "validated." Funny that I should want to run the > current > release of my operating system... The same problems mentioned above were ecxperienced with the "validated" win95. Casey - -- casey@ccii.co.za - ---------------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: DEC/VxWORKS KNOW HOW IN AUSTRALIA ? Date: Fri, 30 Aug 96 11:22:18 PDT From: wb@mkc.com.au Organization: AUSNet Services Message-ID: Is anybody out there in Australia who can report about DEC/VxWorks experience ?? Walter Breunig, Mayer Krieg & Co. wb@mkc.com.au --------------------------- Newsgroups: comp.os.vxworks Subject: Signal Handler Context Date: Wed, 04 Sep 1996 14:58:33 -0800 From: russell.brill@jpl.nasa.gov (Russ Brill) Organization: Jet Propulsion Laboratory Message-ID: Do signal handlers run in task context? Specifically, do the restrictions for interrupt handler code also apply to signal handlers? A manual reference would be appreciated. --------------------------- Newsgroups: comp.os.vxworks Subject: Purify-like tool Date: Thu, 12 Sep 1996 20:31:56 GMT From: Ed Goforth Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: <3238733C.7BF4B952@netcom.com> Sender: egoforth@netcom15.netcom.com We are currently running VxWorks 5.1.1 on Force CPU-5V boards (but we will be upgrading at some point for another project). Are there any tools similar to Purify available? We definitely have a need for VxWorks 5.1.1, but also for Tornado. I have not been receiving messages from the exploder lately, so I would appreciate it if you could copy any replies to egoforth@dsrnet.com. Thanks, Ed Goforth - egoforth@dsrnet.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tonado development environment Date: 13 Sep 1996 03:34:47 GMT From: steved@pacifier.com@199.2.117.163 (Steve Doiel) Organization: Pacifier BBS, Vancouver, Wa. ((360) 693-0325) Message-ID: <51akon$9o1@news.pacifier.com> References: <9609062055.AA00864@softeng.lamrc.com> <3235D7A0.40B2@steptech.com> <3237F18C.766E4098@ccii.co.za> Reply-To: steved@pacifier.com (Steve Doiel) Casey Crellin writes: >Frank Weissig wrote: > >> - The actual implementation on NT needs a bit of work. It has the feel >> of a "1.0 product." I've encountered quite a few problems, particularly >> with the debugger GUI. I find myself using the gdb command line >> interface >> most of the time as a result. The on-line help is mediocre at best, and >> the >> built-in editor is pathetic. Agreed. After speaking with their tech support I learned that they were doing their development (of the environment) with Microsoft Visual C++. I told him they should be required to develop in the development environment they were generating... he said something like "they'd never go for that". > > I am using the Windows 95 version, the above holds true for it too. > Additional problems with windows stability and zombie processes of > gdbi386 and something called runCmd and WinoldApp added. > The command line length on dos is 128 bytes so forget long command > lines for make or run another version of make... > After getting a couple of patches for Tornado, we haven't had may problems with the system locking up. Maybe that's partly because we have learned what to avoid. I prefer to run Tornado on NT... that way when it locks up you can just kill the task without having to reboot. >> - I've been pretty much underwhelmed with Wind River tech support. >> They've >> blown off all of my bug reports because I'm running NT 4.0, for which >> Tornado hasn't been "validated." Funny that I should want to run the >> current >> release of my operating system... I have been using Tornado on NT 4.0 for about a week now. It does not appear to behave any differently on NT 4.0 than it did on NT 3.51. At this point the only siginficant problem I have is with the debugger. It has proven to be very inconsistant in it's reliability. Sometimes I can debug a process, and sometimes I cannot. I've learned that it is a good idea to first "detach" before "stop debugging", and definitely don't just reset the target CPU, unless you want to lock Tornado. On the other hand though, I have found VxWorks to be an EXCELLENT product. The only thing it is missing is a lot of documentation. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Xterminal emulator for PC Date: Thu, 12 Sep 1996 16:26:21 -0700 From: Graham Waters Organization: TRIUMF Message-ID: <32389C1D.7F1B@triumf.ca> References: <9609030937.AA01711@wrsec.fr> <32383B74.5B7@steptech.com> > Jmarc Jean-Marc de Maraumont wrote: > > > > Dear VxUsers, > > > > Does any one of you use a Xterminal emulator successfully with tornado: > > You might like to try Exceed from Hummingbird. I have used versions for Windows 3.2 and Windows 95. I find it faster than X-Win32. You have a choice of contacts: e-mail: support@hummingbird.com www: http://www.hummingbird.com North America: 1 (416) 496 2200 Europe (Geneva): +41 (22) 733 18 58 Graham +-------------------------------------------------------------------+ | Graham Waters | Internet waters@triumf.ca | | Electronics Development | | | TRIUMF | | | 4004 Wesbrook Mall | | | Vancouver BC | Tel: (604)-222-1047 X 6531 | | V6T 2A3 | Fax: (604)-222-7307 | | Canada | TRIUMF Home Page: http://www.triumf.ca | +-------------------------------------------------------------------+ --------------------------- End of New-News digest ********************** From BARONK@hub01.tds-gn.lmco.com Fri Sep 13 05:56:50 1996 From: "Baron, Kenneth @ GRNK (LI-JUPITER DIRSYNC)" Date: Fri Sep 13 05:56:52 PDT 1996 Subject: NFS data rate Hello, I have a sonar application that needs to read about 350 Mbytes of data which I presently have on a fast SCSI hard disk on a Solaris 2.5 SPARC-20 workstation. I am using VxWorks 5.2 on a Motorola PowerPC MVME1604-014 100 MHz board. In my VxWorks application I use fread() to read the data as needed. Since my application is data starved, I would like the reads to be faster. When I mount the Solaris file system and use NFS I get a transfer rate of 5.7 Mbits of my data per second (not counting the packet overhead). When I use netDrv (RSH) I get a transfer rate of 4.1 Mbits per second. Ethernet is supposed to be able to handle 10 Mbits/sec. I have only three nodes on the ethernet: SPARC-20; MVME1604 and an HP printer. 1) Is there any way to tune NFS on Solaris or VxWorks to increase the NFS data rate? 2) What data rates do others experience with PowerPC and 68K systems? 3) How fast a data rate could I expect if I used a client/server TCP socket approach to read the data? Thanks in advance for the info. Ken Baron Lockheed Martin (516) 574-2549 kenneth.baron@lmco.com or BARONK@hub01.tds-gn.lmco.com From garyf@arrowtech.uucp.mcd.mot.com Fri Sep 13 07:07:29 1996 From: garyf@arrowtech.uucp.mcd.mot.com Date: Fri Sep 13 07:07:31 PDT 1996 Subject: Re: PowerPC VME Boards -- Begin original message -- |> From: uumcg!lbl.gov!vxwexplo (the vxWorks Users Group Exploder) |> Date: Fri, 13 Sep 96 02:12:29 PDT |> Subject: Re: PowerPC VME Boards |> To: csg.lbl.gov!vxworks_users |> |> Submitted-by intrepid!alberts@uunet.uu.net Fri Sep 13 02:12:25 1996 |> Submitted-by: intrepid!alberts@uunet.uu.net (Ed Alberts x7819) |> |> PowerPC Boards: |> |> VI computer "http://www.vicomp.com/" is releasing a single slot |> VME 166mhz or 200mhz VI604B board early next month. I currently |> have their VI604A(100 mhz) board. |> |> Ed Alberts |> alberts@eci-cec.com |> -- End original message -- Motorola has already announced their new single slot PowerPC VME boards with either 603 or 604 microprocessors at 200MHz. They expect to start shipping quantity orders with a couple weeks. This new board will be model number series MVME2600. Their new dual slot PowerPC boards, also initially introduced with 200MHz processors will be available in December, and will be model number series MVME3600. Both new board series will support PCI-64 for significant performance increases over the current PowerPC boards. Motorola has been shipping the MVME1600 series PowerPC VME boards for about 18 months with 603-66, 603E-100, 604-100 and 604-133 processors. The prices on the new MVME2600 and MVME3600 boards is also lower than the older MVME1600 boards with equal memory sizes. If you want any info on the new or old boards, just email, call or FAX me and I can FAX you data sheets and pricing. +---------------------------------------------------------------+ | DO NOT USE THE 'REPLY' PATH WHEN REPLYING TO THIS EMAIL. YOU | | MUST USE THE EMAIL PATH BELOW TO REPLY OR INITIATE EMAIL. | |---------------------------------------------------------------| |Gary W. Furr, Manager |email : garyf@mclean.mcd.mot.com | |Technical Sales Support |hotline : 800-470-0704 | |Arrow Electronics, Inc. |phone : 410-309-0686 | |9800J Patuxent Woods Drive|fax : 410-290-7617 | |Columbia, MD 21046-1561 |voicemail: 800-232-2776 #3912710 | |---------------------------------------------------------------| From MikeMcQuade@connect.com Fri Sep 13 07:58:45 1996 From: Mike McQuade Date: Fri Sep 13 07:58:48 PDT 1996 Subject: Re: PowerPC VME Boards At 02:12 AM 9/13/96 PDT, you wrote: >Submitted-by intrepid!alberts@uunet.uu.net Fri Sep 13 02:12:25 1996 >Submitted-by: intrepid!alberts@uunet.uu.net (Ed Alberts x7819) > >PowerPC Boards: > You might want to look at the New Motorola MVME-2600 boards, they are about a 2:1 performance gain over existing PPC boards. I think what really makes the differance is the Falcon and Raven ASICs on these boards. Internal memory to memory xfers on the MVME2600/3600 boards much faster than other mfg boards. The boards have the usual complement of goodies: DEC 21140A 10/100 Mbit ethernet Universe VME-PCI lots of Ram and Flash. MikeMcQuade@connect.com From MikeMcQuade@connect.com Fri Sep 13 10:48:01 1996 From: Mike McQuade Date: Fri Sep 13 10:48:04 PDT 1996 Subject: Re: NFS data rate At 05:56 AM 9/13/96 PDT, you wrote: >Submitted-by BARONK@hub01.tds-gn.lmco.com Fri Sep 13 05:56:50 1996 >Submitted-by: "Baron, Kenneth @ GRNK (LI-JUPITER DIRSYNC)" > >Hello, > >I have a sonar application that needs to read about 350 Mbytes of data >which I presently have on a fast SCSI hard disk on a Solaris 2.5 >SPARC-20 workstation. I am using VxWorks 5.2 on a Motorola PowerPC >MVME1604-014 100 MHz board. In my VxWorks application I use fread() to >read the data as needed. Since my application is data starved, I would >like the reads to be faster. > >When I mount the Solaris file system and use NFS I get a transfer rate >of 5.7 Mbits of my data per second (not counting the packet overhead). >When I use netDrv (RSH) I get a transfer rate of 4.1 Mbits per second. >Ethernet is supposed to be able to handle 10 Mbits/sec. I have only >three nodes on the ethernet: SPARC-20; MVME1604 and an HP printer. > >1) Is there any way to tune NFS on Solaris or VxWorks to increase the >NFS data rate? >2) What data rates do others experience with PowerPC and 68K systems? >3) How fast a data rate could I expect if I used a client/server TCP >socket approach to read the data? > With TCP/IP we have seen about that same data rate (5.7Mbits sustained) on a 1604 133Mhz using VxWorks 5.3 on a moderate trafic network. I think you are at the limitation of 10base-T, if you want to move the data faster, go to 100Base-T. You can get a 100Base-T PMC card for the 1604, from Ramix for about $300.00 I think you will hit the next bottleneck at the SCSI disk xfer rate. MikeMcQuade@connect.com From BARONK@hub01.tds-gn.lmco.com Fri Sep 13 11:13:46 1996 From: "Baron, Kenneth @ GRNK (LI-JUPITER DIRSYNC)" Date: Fri Sep 13 11:13:48 PDT 1996 Subject: NFS data rate Hello, I have a sonar application that needs to read about 350 Mbytes of data which I presently have on a fast SCSI hard disk on a Solaris 2.5 SPARC-20 workstation. I am using VxWorks 5.2 on a Motorola PowerPC MVME1604-014 100 MHz board. In my VxWorks application I use fread() to read the data as needed. Since my application is data starved, I would like the reads to be faster. When I mount the Solaris file system and use NFS I get a transfer rate of 5.7 Mbits of my data per second (not counting the packet overhead). When I use netDrv (RSH) I get a transfer rate of 4.1 Mbits per second. Ethernet is supposed to be able to handle 10 Mbits/sec. I have only three nodes on the ethernet: SPARC-20; MVME1604 and an HP printer. 1) Is there any way to tune NFS on Solaris or VxWorks to increase the NFS data rate? 2) What data rates do others experience with PowerPC and 68K systems? 3) How fast a data rate could I expect if I used a client/server TCP socket approach to read the data? Thanks in advance for the info. Ken Baron Lockheed Martin (516) 574-2549 kenneth.baron@lmco.com or BARONK@hub01.tds-gn.lmco.com From hesham@baynetworks.com Fri Sep 13 16:26:14 1996 From: Hesham El Bakoury Date: Fri Sep 13 16:26:16 PDT 1996 Subject: debugging exceptions I would like to instrument the exception/handling code to be able to print the values of the PC (program counter registers) and other registers. Is there a way to do this ? my understanding is that I can not do printf from within an exception/interrupt handler. I am not sure also how I can get the value of the PC register. Thanks ======================================================= = Hesham El Bakoury Tel: (408) 486-3410 = = Bay Networks, CA = ======================================================= From dandy@netmision.net Sat Sep 14 17:50:56 1996 From: dandy@netmision.net Date: Sat Sep 14 17:50:58 PDT 1996 Subject: Your work could be faster, easier and even fun ! Didn't you ever got the feeling that working with a clipboard that holds and pastes just one object at a time is just too limited? "Clip Baby Clip" breaks that limitation. It collects the text objects as you copy them,displays them in a menu, and lets you paste each one of them whenever you need it ! Enough leaping back and forth re-typing, re-coping the same things again and again. Avoid those embarrassing human mistakes in names, addresses, difficult terms and numbers. Let "Clip Baby Clip" do the job ! You are invited to download your FREE fully functional shareware (for Windows 3.1x and Windows 95). Visit us at http://www.xmission.com/~wintrnx/regnet/297p.htm ! Thank you in advance, Dandy Quarter From daemon@csg.lbl.gov Mon Sep 16 04:00:40 1996 From: daemon@csg.lbl.gov Date: Mon Sep 16 04:00:43 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Sep 16 04:00:32 PDT 1996 Subject: Re: Subject: Sockets with a timeout Subject: Token Ring Driver Subject: Re: Xterminal emulator for PC Subject: Re: mvme162fx Subject: Re: Tonado development environment Subject: Re: Subject: Sockets with a timeout Subject: Re: Tonado development environment Subject: Re: Problems with select() and recv(). Subject: Re: Problems with select() and recv(). Subject: OUTPOST PATCH ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Subject: Sockets with a timeout Date: 12 Sep 1996 22:25:23 -0700 From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: References: <322B1EFE.59E2B600@bbn.com> <50i3ev$oil@post.gsfc.nasa.gov> <519ie9$gd@post.gsfc.nasa.gov> In article <519ie9$gd@post.gsfc.nasa.gov> alexm@vlsi.gsfc.nasa.gov (Alex Measday) writes: >... >I'm really getting confused. I just found out that one of my programs >that select()s a blocking socket to see if it is writeable before >write()ing to it works as I expected under SunOS but not VxWorks. Ok. It would be nice to get an authoritative answer to this so I have posted to comp.unix.programmer requesting clarification. Martin Shepherd (mcs@astro.caltech.edu) --------------------------- Newsgroups: comp.os.vxworks Subject: Token Ring Driver Date: Tue, 10 Sep 1996 12:04:16 -0400 From: Jeremy Frank Organization: Teklogix Incorporated Message-ID: <32359180.39AF@teklogix.com> Reply-To: jfrank@teklogix.com hi all, I am looking for a token ring driver. The hardware we are using is an IBM mini-card which contains the TROPIC chip. The first application just uses the connectionless mode of the chip (?) and runs TCP/IP over it. However, we will eventually need a driver that can handle both TCP/IP and SNA (not simultaneously; configurable at runtime). Any help would be appreciated. Please email a reply to me, as I do not keep up with this newsgroup constantly. Thank you! Jeremy Frank, Teklogix Inc. jfrank@teklogix.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Xterminal emulator for PC Date: Thu, 12 Sep 1996 09:35:33 -0700 From: Frank Weissig Organization: STEP Technology, Inc. Message-ID: <32383B74.5B7@steptech.com> References: <9609030937.AA01711@wrsec.fr> Reply-To: frank@steptech.com Jmarc Jean-Marc de Maraumont wrote: > > Dear VxUsers, > > Does any one of you use a Xterminal emulator successfully with tornado: > > PC client using an emulator + Unix station (X11 R5/R6) running Tornado. > > Tornado should run on the Unix station and being displayed on the PC. > So far, I have tried using Xoftware (from Chameleon) but it crashes Windows 95. > > Thanks for your help, > > Jean-Marc De Maraumont. > > ***************************************************************** > * Jean-Marc de Maraumont email : jmarc@wrsec.fr * > * WIND RIVER SYSTEM EUROPE tel : 33 1 60 92 63 19 * > * 19, Av. de la Norvege fax : 33 1 69 92 63 15 * > * ZA de Courtaboeuf 1 * > * 91962 LES ULIS cedex. France http://www.wrs.com * > ***************************************************************** For a PC-based XServer, I've been using Micro X-Win32 from StarNet Communications. Seems to work pretty well on NT 3.51, 4.0 and Win95. I haven't used it with Tornado, since I run the NT version, but I do use it quite a bit for administering a Linux system from my NT desktop and my Win95 system at home. The company can be reached at +1.408.739.0881 or sales@starnet.com They have a single-node demo version you can try out for free. - -- Frank Weissig Senior Staff Engineer STEP Technology, Inc. frank@steptech.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: mvme162fx Date: Fri, 13 Sep 1996 17:37:51 -0700 From: Frank Weissig Organization: STEP Technology, Inc. Message-ID: <3239FE5F.CB0@steptech.com> References: <51cqei$56t@news.internetmci.com> Reply-To: frank@steptech.com Chuck Woodring wrote: > > Hi, > > Has anyone successfully put a VxWorks boot image into the Flash RAM of > the Motorola MVME162FX, 33MHz card. I'm having a heck of a time. > Windriver is sending me a boot eprom but I would still like to get the > code into the fllash. > > Any help/comments appreciated. I didn't have any problems burning the flash in an MVME162 per the instructions in the target.txt file, but I don't know what's different about an "FX" board. - -- Frank Weissig Senior Staff Engineer STEP Technology, Inc. frank@steptech.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tonado development environment Date: Fri, 13 Sep 1996 17:48:02 -0700 From: Frank Weissig Organization: STEP Technology, Inc. Message-ID: <323A00C2.1DE2@steptech.com> References: <9609062055.AA00864@softeng.lamrc.com> <3235D7A0.40B2@steptech.com> <3237F18C.766E4098@ccii.co.za> <51akon$9o1@news.pacifier.com> Reply-To: frank@steptech.com Steve Doiel wrote: > > Casey Crellin writes: > >Frank Weissig wrote: > > > >> - The actual implementation on NT needs a bit of work. It has the feel > >> of a "1.0 product." I've encountered quite a few problems, particularly > >> with the debugger GUI. I find myself using the gdb command line > >> interface > >> most of the time as a result. The on-line help is mediocre at best, and > >> the > >> built-in editor is pathetic. > > Agreed. After speaking with their tech support I learned that they were doing > their development (of the environment) with Microsoft Visual C++. I told him > they should be required to develop in the development environment they were > generating... he said something like "they'd never go for that". > > > > > I am using the Windows 95 version, the above holds true for it too. > > Additional problems with windows stability and zombie processes of > > gdbi386 and something called runCmd and WinoldApp added. > > The command line length on dos is 128 bytes so forget long command > > lines for make or run another version of make... > > > > After getting a couple of patches for Tornado, we haven't had may problems > with the system locking up. Maybe that's partly because we have learned what > to avoid. I prefer to run Tornado on NT... that way when it locks up you can > just kill the task without having to reboot. > > >> - I've been pretty much underwhelmed with Wind River tech support. > >> They've > >> blown off all of my bug reports because I'm running NT 4.0, for which > >> Tornado hasn't been "validated." Funny that I should want to run the > >> current > >> release of my operating system... > > I have been using Tornado on NT 4.0 for about a week now. It does not > appear to behave any differently on NT 4.0 than it did on NT 3.51. I suspect that at least some of my problems are unrelated to 4.0, but the fact that I'm using 4.0 has caused WRS tech support to blow most of them off. And I don't have time to install on 3.51 just to prove the point. (None of the problems are show-stoppers.) > > At this point the only siginficant problem I have is with the debugger. It has > proven to be very inconsistant in it's reliability. Sometimes I can debug a > process, and sometimes I cannot. I've learned that it is a good idea to first > "detach" before "stop debugging", and definitely don't just reset the target > CPU, unless you want to lock Tornado. > > On the other hand though, I have found VxWorks to be an EXCELLENT product. > The only thing it is missing is a lot of documentation. As I mentioned in my original post, I've found the quality of the print docs to be quite good. If I have to make a choice, I guess I'll take that over quantity. I do wish that the on-line help was more complete and better organized, so I don't have to drag the print docs with me when I run tests at my client's lab. - -- Frank Weissig Senior Staff Engineer STEP Technology, Inc. frank@steptech.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Subject: Sockets with a timeout Date: 9 Sep 1996 16:35:18 GMT From: alexm@vlsi.gsfc.nasa.gov (Alex Measday) Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <511h06$msi@post.gsfc.nasa.gov> References: <322B1EFE.59E2B600@bbn.com> <50i3ev$oil@post.gsfc.nasa.gov> Martin Shepherd wrote: > Alex Measday wrote: > > > If you're using select(), there's no reason (that I'm aware of) for the > > socket to be non-blocking. > > It isn't necessary but it certainly helps. The fact that select() > indicates that there is data available to be read doesn't mean that > the whole message that you are expecting is available, only that at > least one byte is available. If you then attempt to read from the > indicated socket, the socket may block if you attempt to read more > than a single byte, whereas if the socket were set up for non-blocking > I/O then you could attempt to read the complete remainder of the > expected message without danger of blocking. This is clearly more > efficient. If select() indicates that input is available, then a read(), blocking or non-blocking, will return whatever is available. A blocking read() does NOT wait until the requested amount of data (e.g., a complete message) has been received before returning to the caller. That's why you have to loop on a read() to collect the complete message. (Or you can buffer what's been read and loop on select() to wait for the remainder of the message.) I don't know if there's any difference in efficiency between a blocking and non-blocking read(). Alex ======================================================================== | Alex Measday Code 521 - Microelectronic Systems Branch | | Integral Systems, Inc. NASA Goddard Space Flight Center | | c.a.measday@ieee.org Greenbelt, MD | ======================================================================== ) http://purl.oclc.org/net/alex/ ( ======================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Tonado development environment Date: Thu, 12 Sep 1996 14:22:05 GMT From: sighg@sesig.abb.se (Henry Gessau) Organization: Adtranz Signal, Stockholm Message-ID: <32381afc.633213502@news.seinf.abb.se> References: <9609062055.AA00864@softeng.lamrc.com> <3235D7A0.40B2@steptech.com> <3237F18C.766E4098@ccii.co.za> Reply-To: sighg@sesig.abb.se On Thu, 12 Sep 1996 13:18:36 +0200, Casey Crellin wrote: > Frank Weissig wrote: > > > - The actual implementation on NT needs a bit of work. It has the feel > > of a "1.0 product." I've encountered quite a few problems, particularly > > with the debugger GUI. I find myself using the gdb command line > > interface > > most of the time as a result. The on-line help is mediocre at best, and > > the > > built-in editor is pathetic. > > I am using the Windows 95 version, the above holds true for it too. > Additional problems with windows stability and zombie processes of > gdbi386 and something called runCmd and WinoldApp added. > The command line length on dos is 128 bytes so forget long command > lines for make or run another version of make... I'll add a "me too". We use Tornado under NT and very often in debugging sessions we get zombie runcmd processes and the whole system deteriorates until it is unusable. We have to use a "kill" utility from the NT Resource Kit to blat the offending processes. At least under NT there is no problem with 128 byte command lines. We have very long make command lines presenting no problem at all. In fact the make and compiler stuff is very fast and stable. - -- -------+ Henry Gessau System Designer/Software Engineer/Programmer/Hacker | | ABB Daimler-Benz Transportation Signal, Stockholm, Sweden | +------- All views and opinions expressed in this message are strictly my own --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with select() and recv(). Date: Mon, 09 Sep 1996 08:35:33 -0500 From: Lee Graba Organization: Honeywell Technology Center Message-ID: <32341D25.154F@htc.honeywell.com> References: <3230A5B4.3C6F@htc.honeywell.com> <3231EA39.6B93@netcom.com> Jim Simpson wrote: > > Lee Graba wrote: > > > > We are using VxWorks 5.1 on Force 40 VME boards (68040). In part of our application, a > > task running at a priority of 33, goes into a select loop, where it waits on a few > > sockets on which connections already exist. Its basic task is below: > > > > for (;;) > > { > > select_cnt=select(readfds, 0, 0, 0, 10 seconds); > > if (select_cnt > 0) > > { > > find the FD with stuff to read > > read the stuff into a buffer using 2 calls to recv > > spawn a task at lower priority to process the buffer > > } > > else if (select_cnt < 0) > > { > > get errno and print error string > > } > > } > > > > After the select, I have use ioctl to check the bytes to read, and compared with the > > number of bytes read by the recv calls, and these add up. The task that is spawned to > > process the buffer owns the buffer, and the parent task doesn't touch it after the task > > is spawned. > > > > The problem is, we are getting many errors from select after first getting a good > > select, i.e. we successfully read the proper number of bytes from the fd, spawn a task, > > and when we then loop around to call select again, it immediately returns select_cnt = > > -1, and errno = 0xD0003, which strerror_r reports as having to do with an > > INVALID_FILE_DESCRIPTOR. If I put a taskDelay of 100 ms somewhere in the select loop, I > > no longer get these messages. > > > > Anybody have any idea what is causing select to return the error, and why timing has > > anything to do with it? > > > > -- > > =========================================================================== > > Lee Graba = > > Honeywell Technology Center = > > 3660 Technology Drive = > > Minneapolis, MN 55418 = > > (612) 951-7503 = > > Graba_Lee@htc.honeywell.com = > > graba@htc.honeywell.com = > > =========================================================================== > > How about this: > > > for (;;) > > { > > readfds = set all the fd bits from which you > want to read. Do this everytime through > the loop. > > > select_cnt=select(readfds, 0, 0, 0, 10 seconds); > > if (select_cnt > 0) > > { > > find the FD with stuff to read > > read the stuff into a buffer using 2 calls to recv > > spawn a task at lower priority to process the buffer > > } > > else if (select_cnt < 0) > > { > > get errno and print error string > > } > > } > > Actually, we already to the bit that you added. My pseudo-code was too 'pseudo' to include it in my original post! - -- =========================================================================== Lee Graba = Honeywell Technology Center = 3660 Technology Drive = Minneapolis, MN 55418 = (612) 951-7503 = Graba_Lee@htc.honeywell.com = graba@htc.honeywell.com = =========================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problems with select() and recv(). Date: Mon, 09 Sep 1996 08:37:40 -0500 From: Lee Graba Organization: Honeywell Technology Center Message-ID: <32341DA4.60FF@htc.honeywell.com> References: <3230A5B4.3C6F@htc.honeywell.com> <50rsep$89o@news.sei.cmu.edu> Theodore Marz wrote: > > In article <3230A5B4.3C6F@htc.honeywell.com>, Lee Graba writes: > |> We are using VxWorks 5.1 on Force 40 VME boards (68040). In part of our application, a > |> task running at a priority of 33, goes into a select loop, where it waits on a few > |> sockets on which connections already exist. Its basic task is below: > |> > |> for (;;) > |> { > |> select_cnt=select(readfds, 0, 0, 0, 10 seconds); > |> if (select_cnt > 0) > |> { > |> find the FD with stuff to read > |> read the stuff into a buffer using 2 calls to recv > |> spawn a task at lower priority to process the buffer > |> } > |> else if (select_cnt < 0) > |> { > |> get errno and print error string > |> } > |> } > |> > |> After the select, I have use ioctl to check the bytes to read, and compared with the > |> number of bytes read by the recv calls, and these add up. The task that is spawned to > |> process the buffer owns the buffer, and the parent task doesn't touch it after the task > |> is spawned. > |> > |> The problem is, we are getting many errors from select after first getting a good > |> select, i.e. we successfully read the proper number of bytes from the fd, spawn a task, > |> and when we then loop around to call select again, it immediately returns select_cnt = > |> -1, and errno = 0xD0003, which strerror_r reports as having to do with an > |> INVALID_FILE_DESCRIPTOR. If I put a taskDelay of 100 ms somewhere in the select loop, I > |> no longer get these messages. > |> > |> Anybody have any idea what is causing select to return the error, and why timing has > |> anything to do with it? > |> > |> > |> -- > |> =========================================================================== > |> Lee Graba = > |> Honeywell Technology Center = > |> 3660 Technology Drive = > |> Minneapolis, MN 55418 = > |> (612) 951-7503 = > |> Graba_Lee@htc.honeywell.com = > |> graba@htc.honeywell.com = > |> =========================================================================== > > I not sure what is happening, but here are some thoughts: > > What priority is the spawned process running at, relative to the pri 33 data > detection routine? If the data consumer is not running at higher priority > than the data detection routine, you could be re-cycling into the data > detection while the data is in the process of being consumed. The file > descriptor (socket) may be locked by the consumer task, and so when you try > and select(), it fails. > > The 100mS delay you add pretty much ensures that the consumer task has > completed reading the data before another select() gets called on the same > fd/socket. > > Ted Marz The spawned task is running at a lower priority than the task that calls select and reads from the socket. The spawned task does not read from the socket, this will have been done by the task that called select(). - -- =========================================================================== Lee Graba = Honeywell Technology Center = 3660 Technology Drive = Minneapolis, MN 55418 = (612) 951-7503 = Graba_Lee@htc.honeywell.com = graba@htc.honeywell.com = =========================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: OUTPOST PATCH Date: Sun, 15 Sep 1996 16:07:09 -0500 From: Jeremy Brandt Organization: Northeast Missouri State University Message-ID: <323C6FFD.5DEB@academic.truman.edu> Help! I need a patch for Sierra's game Outpost. It's actually a single file that's been corrupted and I need a copy of it. It is called: tile6.opc If you know about this, could you e-mail me? I can't access this newsgroup on a reg basis. ThankS! Jeremy Brandt v874@academic.truman.edu O- --------------------------- End of New-News digest ********************** From gema@force.de Mon Sep 16 05:35:05 1996 From: gema@force.de (Gerhard Mayr) Date: Mon Sep 16 05:35:07 PDT 1996 Subject: Re: PowerPC VME Boards FORCE Computers has already PowerPC based single-slot CPU boards in their product program. The current portfolio covers 603 and 604 based boards, with one or two PMC slots. Several operating systems has been ported, also VxWorks. FORCE COMPUTERS has sales offices all over the world, please visit our web page to get details about your next sales office. "http://www.forcecomputers.com". Gerd Mayr gema@force.de From cpr@ocegr.fr Mon Sep 16 09:41:09 1996 From: cpr@ocegr.fr (Craig Preston) Date: Mon Sep 16 09:41:11 PDT 1996 Subject: Purify with VxSim? This may seem like an oddball question, but ... does anybody know if using Purify from Pure Atria with VxSim gives any useful results? Purify is a run-time error detection tool using Object Code Insertion in the Unix environment. I have no idea how VxSim maps its emulation of VxWorks memory management onto Unix, and so I wonder if using Purify on an application running with VxSim isn't a waste of time. I am aware that there are equivalent tools for VxWorks, but my organization is hesitant to invest in them. Any insight, experience, advice, etc, pro or con is solicited. If nothing else, they may serve to justify an investment in dedicated tools. :-) Please reply to me directly. If anyone's interested, I'd be happy to summarize and pass on any responses I get. Thanks in advance for your help, Craig // Craig Preston // Software Quality Engineer // Oce Graphics France S.A. // email: preston@ocegr.fr // phone: +33 1 48 98 80 92 // fax : +33 1 48 98 54 50 From c_mcfarr@linkabit.titan.com Mon Sep 16 09:41:37 1996 From: c_mcfarr@linkabit.titan.com (Shawn McFarr) Date: Mon Sep 16 09:41:39 PDT 1996 Subject: Accessing host files using Tornado? My Tornado configuration excludes network facilities due to small amount of available target memory. I'd like to use open()/close()/read()/write() VxWorks calls in my application to access files that are resident on my host file system (i.e. using WDB target agent). My other method is to use the host resident WTX protocol through either the WTX Tcl API or WTX C API. Any other alternate methods? Thanks, Shawn McFarr From usd@mee.hwm.com Mon Sep 16 09:56:13 1996 From: Uwe Scheffold Date: Mon Sep 16 09:56:15 PDT 1996 Subject: Load and run applications with telnet Hi vxWorker's has anybody out there been able to load and run a application, compiled for tornado, only with a telnet (rlogin host)? We do !not! want to use the "host based shell". Uwe From MAILER-DAEMON Mon Sep 16 10:39:22 1996 From: PMDF Mail Server Date: Mon Sep 16 10:39:27 PDT 1996 Subject: Undeliverable mail: Processing failure --Boundary (ID raoQahN+HFNcBmVBehljuQ) Content-type: TEXT/PLAIN; CHARSET=US-ASCII These addresses were rejected: suan j henry: *** Unknown message recipient *** --Boundary (ID raoQahN+HFNcBmVBehljuQ) Content-type: MESSAGE/RFC822 Return-path: vxwexplo@lbl.gov Date: Mon, 16 Sep 1996 06:00:45 -0500 (EST) From: vxwexplo@lbl.gov Subject: comp.os.vxworks newsdigest To: vxworks_users@csg.lbl.gov MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII; NAME=TEXT.TXT Content-transfer-encoding: X-UUENCODE begin 600 TEXT.TXT M4W5B;6ET=&5D+6)Y(&1A96UO;D!C'=O`T*("`@("`@("!3=6)J96-T.B!293H@5&]N861O(&1E M=F5L;W!M96YT(&5N=FER;VYM96YT#0H@("`@("`@(%-U8FIE8W0Z(%)E.B!3 M=6)J96-T.B!3;V-K971S('=I=&@@82!T:6UE;W5T#0H@("`@("`@(%-U8FIE M8W0Z(%)E.B!4;VYA9&\@9&5V96QO<&UE;G0@96YV:7)O;FUE;G0-"B`@("`@ M("`@4W5B:F5C=#H@4F4Z(%!R;V)L96US('=I=&@@6,Y8G5F8F]U:V,N9G-F0&=O8FQI;BYC86QT M96-H+F5D=3X-"E)E9F5R96YC97,Z(#PS,C)",45&12XU.44R0C8P,$!B8FXN M8V]M/B`\-3!I,V5V)&]I;$!P;W-T+F=S9F,N;F%S82YG;W8^#0H)/'EC.6%F M=7AS,W!V+F9S9D!G;V)L:6XN8V%L=&5C:"YE9'4^(#PU,3EI93DD9V1`<&]S M="YG2D@=W)I=&5S.@T*/BXN+@T*/DDG;2!R96%L;'D@9V5T M=&EN9R!C;VYF=7-E9"X@($D@:G5S="!F;W5N9"!O=70@=&AA="!O;F4@;V8@ M;7D@<')O9W)A;7,-"CYT:&%T('-E;&5C="@I'!E8W1E9"!U;F1E"YP'=OF%T:6]N.B!496ML;V=I>"!);F-O2!H96QP('=O=6QD(&)E(&%P M<')E8VEA=&5D+B`@4&QE87-E(&5M86EL(&$@2X@(%1H86YK('EO=2$-"@T*2F5R96UY($9R86YK+"!496ML;V=I>"!);F,N M#0IJ9G)A;FM`=&5K;&]G:7@N8V]M#0H-"BTM+2TM+2TM+2TM+2TM+2TM+2TM M+2TM+2TM+0T*#0I.97=S9W)O=7!S.B!C;VUP+F]S+G9X=V]R:W,-"E-U8FIE M8W0Z(%)E.B!8=&5R;6EN86P@96UU;&%T;W(@9F]R(%!##0I$871E.B!4:'4L M(#$R(%-E<"`Q.3DV(#`Y.C,U.C,S("TP-S`P#0I&2P@26YC+@T*365S2!O;F4@;V8@>6]U('5S92!A M(%AT97)M:6YA;"!E;75L871O"!S M=&%T:6]N("A8,3$@4C4O4C8I(')U;FYI;F<@5&]R;F%D;RX-"CX@#0H^(%1O M"!S=&%T:6]N(&%N9"!B96EN M9R!D:7-P;&%Y960@;VX@=&AE(%!#+@T*/B!3;R!F87(L($D@:&%V92!T"X@1G)A;F-E("`@("`@("!H='1P M.B\O=W=W+G=R2!7:6XY-2!S>7-T96T@870@:&]M92X- M"@T*5&AE(&-O;7!A;GD@8V%N(&)E(')E86-H960@870@*S$N-#`X+C2!O=70@9F]R(&9R964N M#0H-"BT@+2T@#0I&2P@26YC+@T*9G)A;FM`2!P=70@82!6 M>%=O'0@9FEL92P@8G5T($D@ M9&]N)W0@:VYO=R!W:&%T)W,@9&EF9F5R96YT#0IA8F]U="!A;B`B1E@B(&)O M87)D+@T*#0HM("TM(`T*1G)A;FL@5V5I'=O2P@ M26YC+@T*365S2U4;SH@9G)A;FM`0T*/B`^/B!W:71H('1H92!D96)U9V=E2!P7-T96T@;&]C:VEN9R!U<"X@($UA>6)E('1H870G M6]U(&-A;@T*/B!J=7-T(&MI;&P@=&AE M('1A2!B=6<@2!T:&%T($D@2!D:69F97)E;G1L>2!O;B!.5"`T M+C`@=&AA;B!I="!D:60@;VX@3E0@,RXU,2X-"@T*22!S=7-P96-T('1H870@ M870@;&5A2!P%=O2!T:&EN M9R!I="!I2X@($D@9&\@=VES:"!T:&%T M('1H92!O;BUL:6YE(&AE;'`@=V%S(&UOF5D+"!S;R!)(&1O;B=T(&AA=F4@=&\@9')A9R!T:&4@<')I M;G0@9&]C2!C;&EE M;G0G2P@26YC+@T*9G)A;FM`&U`=FQS:2YG6QA;F0@55-! M#0I-97-S86=E+4E$.B`\-3$Q:#`V)&US:4!P;W-T+F=S9F,N;F%S82YG;W8^ M#0I2969E6,Y6]U)W)E('5S M:6YG('-E;&5C="@I+"!T:&5R92=S(&YO(')E87-O;B`H=&AA="!))VT@87=A M2!H M96QP6]U(&-O=6QD M(&%T=&5M<'0@=&\@'!E8W1E9"!M97-S86=E('=I=&AO=70@9&%N9V5R(&]F(&)L;V-K M:6YG+B!4:&ES(&ES(&-L96%R;'D@;6]R90T*/B!E9F9I8VEE;G0N#0H-"DEF M('-E;&5C="@I(&EN9&EC871E6]U(&AA=F4@=&\@;&]O<"!O;B!A(')E860H*2!T;R!C M;VQL96-T('1H92!C;VUP;&5T92!M97-S86=E+@T**$]R('EO=2!C86X@8G5F M9F5R('=H870G`T*#0H]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T-"GP@($%L M97@@365A2`@("`@("`@("`@("!#;V1E(#4R,2`M($UI8W)O96QE8W1R M;VYI8R!3>7-T96US($)R86YC:"`@?`T*?"`@26YT96=R86P@4WES=&5M4!I965E+F]R9R`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@($=R965N8F5L="P@340@('P-"CT]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/0T**2`@("`@("`@("`@("`@("`@("`@(&AT M='`Z+R]P=7)L+F]C;&,N;W)G+VYE="]A;&5X+R`@("`@("`@("`@("`@("`@ M("`H#0H]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T-"@T*#0HM+2TM+2TM M+2TM+2TM+2TM+2TM+2TM+2TM+2T-"@T*3F5W'=O2!'97-S874I#0I/2!#4!C8VEI+F-O+GIA/B!W0T* M/B`^('=I=&@@=&AE(&1E8G5G9V5R($=522X@($D@9FEN9"!M>7-E;&8@=7-I M;F<@=&AE(&=D8B!C;VUM86YD(&QI;F4-"CX@/B!I;G1E2!O9G1E;B!I;B!D96)U9V=I;F<-"G-E71E(&-O;6UA;F0@;&EN97,N(%=E(&AA=F4-"G9E2!'97-S874@("`@("`@("`@("!3>7-T96T@1&5S:6=N M97(O4V]F='=AB!47=E;&PN8V]M/B`\,S(S,45! M,SDN-D(Y,T!N971C;VTN8V]M/@T*#0I*:6T@4VEM<'-O;B!WPT*/B`^("`@("`@("`@("`@("`@("!S96QE8W1?8VYT/7-E;&5C M="AR96%D9F1S+"`P+"`P+"`P+"`Q,"!S96-O;F1S*3L-"CX@/B`@("`@("`@ M("`@("`@("`@:68@*'-E;&5C=%]C;G0@/B`P*0T*/B`^("`@("`@("`@("`@ M("`@("![#0H^(#X@("`@("`@("`@("`@("`@("`@("`@("`@9FEN9"!T:&4@ M1D0@=VET:"!S='5F9B!T;R!R96%D#0H^(#X@("`@("`@("`@("`@("`@("`@ M("`@("`@71E2!T:&4@2!E M7=E;&P@5&5C:&YO;&]G>2!# M96YT97(@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`@(#T-"CX@/B`S-C8P(%1E8VAN;VQO9WD@1')I=F4@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@(#T- M"CX@/B!-:6YN96%P;VQIPT*/@T*/B`@("`@("`@("`@("`@("`@("!R96%D M9F1S(#T@7=E M;&P@5&5C:&YO;&]G>2!#96YT97(@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@(#T-"C,V-C`@5&5C:&YO;&]G>2!$7=E M;&PN8V]M/@T*3W)G86YI>F%T:6]N.B!(;VYE>7=E;&P@5&5C:&YO;&]G>2!# M96YT97(-"DUE7=E M;&PN8V]M/@T*4F5F97)E;F-EPT*/B!\/B`@("`@("`@("`@("`@("`@("`@("!F:6YD('1H92!&1"!W M:71H('-T=69F('1O(')E860-"CX@?#X@("`@("`@("`@("`@("`@("`@("`@ M71E$0P,#`S+"!W:&EC:"!S=')E7=E;&P@5&5C:&YO M;&]G>2!#96YT97(@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@(#T-"CX@?#X@,S8V,"!496-H;F]L;V=Y($1R:79E("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`]#0H^('P^($UI;FYE87!O;&ES+"!-3B`U-30Q."`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@/0T* M/B!\/B`H-C$R*2`Y-3$M-S4P,R`@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@(#T-"CX@?#X@1W)A M8F%?3&5E0&AT8RYH;VYE>7=E;&PN8V]M("`@("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`]#0H^('P^(&=R86)A0&AT8RYH M;VYE>7=E;&PN8V]M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@/0T*/B!\/B`]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T-"CX-"CX@22!N;W0@2!I2!T:&4@8V]N6]U('1R>0T*/B!A;F0@6]U(&%D9"!P7=E;&PN8V]M("`@("`@("`@("`@("`@("`@ M("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@("`@/0T*/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T]/3T] M/3T]/3T]/3T]/3T]/3T]/3T]/3T]#0H-"BTM+2TM+2TM+2TM+2TM+2TM+2TM M+2TM+2TM+0T*#0I.97=S9W)O=7!S.B!C;VUP+F]S+G9X=V]R:W,-"E-U8FIE M8W0Z($]55%!/4U0@4$%40T@-"D1A=&4Z(%-U;BP@,34@4V5P(#$Y.38@,38Z M,#2!"2!O9B!I="X@($ET(&ES(&-A;&QE9#H-"@T*=&EL M938N;W!C#0H-"DEF('EO=2!K;F]W(&%B;W5T('1H:7,L(&-O=6QD('EO=2!E M+6UA:6P@;64_("!)(`T*8V%N)W0@86-C97-S('1H:7,@;F5W2!" Date: Mon Sep 16 13:17:41 PDT 1996 Subject: Re: vxWorks I/O system At 11:54 AM 9/16/96 PDT, you wrote: >Submitted-by schen3@ccgate.hac.com Mon Sep 16 11:54:55 1996 >Submitted-by: schen3@ccgate.hac.com > > Hello, > > When the application make call to the VxWorks I/O system such as > open(), does VxWorks spawn the open() as a task or just a function > call? function call from used in application like Unix, function call called from shell UNlike Unix adios ----------------------------------------- /\\ Keith Buchanan ///\\ /////\\ SPARTA, Inc. ///// \\ 7926 Jones Branch Drive / \\\\\ \\ Suite 900 \ \\\\\ // McLean, VA 22102 \ ////// \////// Voice: 703 448 0210 \//// FAX: 703 893 5494 \// WWW: www.mclean.sparta.com From bwenholz@pacesetter.com Mon Sep 16 16:36:24 1996 From: bwenholz@pacesetter.com (Bruce Wenholz) Date: Mon Sep 16 16:36:32 PDT 1996 Subject: FWD>FW- WARNING (fwd) ----- Begin Included Message ----- From curata@Pacesetter.Com Mon Sep 16 14:19 PDT 1996 Return-Path: Received: from tustin.Pacesetter.Com by hydra.Pacesetter.Com (SMI-8.6/SMI-SVR4) id OAA10368; Mon, 16 Sep 1996 14:19:19 -0700 Received: by tustin.Pacesetter.Com (SMI-8.6/SMI-SVR4-PI) id OAA09585; Mon, 16 Sep 1996 14:19:18 -0700 Date: Mon, 16 Sep 1996 14:19:18 -0700 From: curata@Pacesetter.Com (Chet Urata) To: vsorense@Pacesetter.Com, ffernand@Pacesetter.Com, gmusser@Pacesetter.Com, terzis@Pacesetter.Com, gbenenyan@Pacesetter.Com, hzar@Pacesetter.Com, sshah@Pacesetter.Com, elogan@Pacesetter.Com, smccord@Pacesetter.Com, syang@Pacesetter.Com, ted@sd.inri.com, emiller@SIRIUS.COM, Don_Pugh@zd.com, bwenholz@Pacesetter.Com, cheturata@earthlink.net, lhoffman@xxcal.com, rengan@alsys.com, iwamoto@nvidia.com, rcisnero@Pacesetter.Com, YUENR@cpvz.dcs.saic.com Subject: FWD>FW- WARNING (fwd) Yo! I called the number and it looks like they are ready for the deluge. They say that the database does not contain any lethal info. Use your own judgement. Chet ----- Begin Included Message ----- From sophie@worldnet.att.net Mon Sep 16 09:57 PDT 1996 X-Sender: sophie@postoffice.worldnet.att.net Mime-Version: 1.0 To: CariePaTi@aol.com, curata@Pacesetter.Com, billyogl@minn.net, cafoley@aol.com, MDP5902@aol.com, Freeeeko@aol.com, mitch@hdmg.com, sma@unlinfo.unl.edu (steven anderson), jill@ssch.com, steves@pa.dec.com From: Sophie Anderson Subject: FWD>FW- WARNING (fwd) Date: Mon, 16 Sep 1996 16:14:06 +0000 >From: Kelly Bowen >To: barbara , Sophie >Subject: FWD>FW- WARNING (fwd) >Date: Mon, 16 Sep 96 15:24:17 +0000 > >Mail*Link(r) SMTP FWD>FW: WARNING (fwd) > > >I just got this in the mail. FYI. > >Forwarded message: >> > _______________________________ >> > >> > FYI (sent to me by one our attorneys): >> > >> > Your name, social security number, current address, previous addresses, >> > mother's maiden name, birth date, and other personal information are now >> > available to anyone with a credit card through a new Lexis database called > >> > P-Trax. As I am sure you are aware, this information could be used to >> > commit credit card fraud or otherwise allow someone else to use your >> > identity. >>> >>> You can have your name and information removed from this list by making a >>> telephone request. Call (800) 543-6862, select option4 and then option 3 >>> ("all other questions") and tell the representative answering that you wish > >>> to remove your name from this P-trax database. You may also send a fax to >>> (513) 865-7360, or physical mail to LEXIS-NEXIS / P.O. Box 933 / Dayton, >>> Ohio 45401-0933. Sending physical mail to confirm your name has been >>> removed is always a good idea. >>> >>>As word of the existence of this database has spread on the net, Lexis-Nexis > >>>has been inundated with calls, and has set up a special set of operators to >>> handle the volume. In addition, Andrew Bleh (rhymes with "Play") is a >>> manager responsible for this product, and is the person to whom complaints >>> about the service could be directed. He can be reached at the above 800 >>> number. Ask for extension 3385. According to Lexis, the manager >>> responsible is Bill Fister at extension 1364. >>> >>>I called this morning and had my name removed. The representative will need > >>>your name and social security number to remove you from the list. I suggest > >>>that we inundate these people with requests to remove our info from the list > >>> and FORWARD THIS E-MAIL TO EVERYONE WE KNOW!!!! > >------------------ RFC822 Header Follows ------------------ >Received: by it.berklee.edu with SMTP;13 Sep 1996 11:10:30 -0500 >Received: (from egieskes@localhost) by acs4.bu.edu (8.7.5/) id KAA145440; Fri, >13 Sep 1996 10:55:41 -0400 >From: egieskes@bu.edu >Message-Id: <199609131455.KAA145440@acs4.bu.edu> >Subject: FW: WARNING (fwd) >To: kbowen@it.berklee.edu (Kelly Bowen), steves@pa.dec.com (Steve Schneider), > BGieskes@aol.com (Barbara Gieskes), Freeeeko@aol.com (Michiel Gieskes) >Date: Fri, 13 Sep 1996 10:55:40 -0400 (EDT) >Cc: ahartley@acs.bu.edu, lkehoe@acs.bu.edu (Lauren Kehoe), > jsiemon@acs.bu.edu (James Siemon), > wcarroll@acs.bu.edu (William Carroll), > kbazler@acs.bu.edu (Kirk Melnikoff), > paigemct@acs.bu.edu (Paige McThenia), > zleib@acs.bu.edu (Zahava Leibowitz), slyons@acs.bu.edu (Sarah Lyons), > mwalker@acs.bu.edu (Michael Walker), > kpetersn@acs.bu.edu (Kaara Peterson), > kbrumm@acs.bu.edu (Kathryn Brumm), > kmur-and@acs.bu.edu (Kathryn MurphyAnderson), > jillo@acs.bu.edu (Jill Orofino), > imlac@acs.bu.edu (Eric Johnson-DeBaufre), > jmulroon@acs.bu.edu (Jonathan Mulrooney), > colinh@acs.bu.edu (Colin Harris) >X-Mailer: ELM [version 2.4 PL25] >MIME-Version: 1.0 >Content-Type: text/plain; charset=US-ASCII >Content-Transfer-Encoding: 7bit > > > > ----- End Included Message ----- ----- End Included Message ----- From anthonyl@atlantis.softeng.lamrc.com Mon Sep 16 16:43:46 1996 From: Anthony Le Date: Mon Sep 16 16:43:48 PDT 1996 Subject: Re: vxWorks I/O system Hi everyone, Question: How do I make standard input NON-BLOCKING? I want to check if there is a keyboard character input before I use getc() function. Anthony Le LAM RESEARCH CORP. anthony.le@lamrc.com From bwenholz@pacesetter.com Mon Sep 16 17:30:16 1996 From: bwenholz@pacesetter.com (Bruce Wenholz) Date: Mon Sep 16 17:30:47 PDT 1996 Subject: FWD>FW- WARNING (fwd) ----- Begin Included Message ----- From curata@Pacesetter.Com Mon Sep 16 14:19 PDT 1996 Return-Path: Received: from tustin.Pacesetter.Com by hydra.Pacesetter.Com (SMI-8.6/SMI-SVR4) id OAA10368; Mon, 16 Sep 1996 14:19:19 -0700 Received: by tustin.Pacesetter.Com (SMI-8.6/SMI-SVR4-PI) id OAA09585; Mon, 16 Sep 1996 14:19:18 -0700 Date: Mon, 16 Sep 1996 14:19:18 -0700 From: curata@Pacesetter.Com (Chet Urata) To: vsorense@Pacesetter.Com, ffernand@Pacesetter.Com, gmusser@Pacesetter.Com, terzis@Pacesetter.Com, gbenenyan@Pacesetter.Com, hzar@Pacesetter.Com, sshah@Pacesetter.Com, elogan@Pacesetter.Com, smccord@Pacesetter.Com, syang@Pacesetter.Com, ted@sd.inri.com, emiller@SIRIUS.COM, Don_Pugh@zd.com, bwenholz@Pacesetter.Com, cheturata@earthlink.net, lhoffman@xxcal.com, rengan@alsys.com, iwamoto@nvidia.com, rcisnero@Pacesetter.Com, YUENR@cpvz.dcs.saic.com Subject: FWD>FW- WARNING (fwd) Yo! I called the number and it looks like they are ready for the deluge. They say that the database does not contain any lethal info. Use your own judgement. Chet ----- Begin Included Message ----- From sophie@worldnet.att.net Mon Sep 16 09:57 PDT 1996 X-Sender: sophie@postoffice.worldnet.att.net Mime-Version: 1.0 To: CariePaTi@aol.com, curata@Pacesetter.Com, billyogl@minn.net, cafoley@aol.com, MDP5902@aol.com, Freeeeko@aol.com, mitch@hdmg.com, sma@unlinfo.unl.edu (steven anderson), jill@ssch.com, steves@pa.dec.com From: Sophie Anderson Subject: FWD>FW- WARNING (fwd) Date: Mon, 16 Sep 1996 16:14:06 +0000 >From: Kelly Bowen >To: barbara , Sophie >Subject: FWD>FW- WARNING (fwd) >Date: Mon, 16 Sep 96 15:24:17 +0000 > >Mail*Link(r) SMTP FWD>FW: WARNING (fwd) > > >I just got this in the mail. FYI. > >Forwarded message: >> > _______________________________ >> > >> > FYI (sent to me by one our attorneys): >> > >> > Your name, social security number, current address, previous addresses, >> > mother's maiden name, birth date, and other personal information are now >> > available to anyone with a credit card through a new Lexis database called > >> > P-Trax. As I am sure you are aware, this information could be used to >> > commit credit card fraud or otherwise allow someone else to use your >> > identity. >>> >>> You can have your name and information removed from this list by making a >>> telephone request. Call (800) 543-6862, select option4 and then option 3 >>> ("all other questions") and tell the representative answering that you wish > >>> to remove your name from this P-trax database. You may also send a fax to >>> (513) 865-7360, or physical mail to LEXIS-NEXIS / P.O. Box 933 / Dayton, >>> Ohio 45401-0933. Sending physical mail to confirm your name has been >>> removed is always a good idea. >>> >>>As word of the existence of this database has spread on the net, Lexis-Nexis > >>>has been inundated with calls, and has set up a special set of operators to >>> handle the volume. In addition, Andrew Bleh (rhymes with "Play") is a >>> manager responsible for this product, and is the person to whom complaints >>> about the service could be directed. He can be reached at the above 800 >>> number. Ask for extension 3385. According to Lexis, the manager >>> responsible is Bill Fister at extension 1364. >>> >>>I called this morning and had my name removed. The representative will need > >>>your name and social security number to remove you from the list. I suggest > >>>that we inundate these people with requests to remove our info from the list > >>> and FORWARD THIS E-MAIL TO EVERYONE WE KNOW!!!! > >------------------ RFC822 Header Follows ------------------ >Received: by it.berklee.edu with SMTP;13 Sep 1996 11:10:30 -0500 >Received: (from egieskes@localhost) by acs4.bu.edu (8.7.5/) id KAA145440; Fri, >13 Sep 1996 10:55:41 -0400 >From: egieskes@bu.edu >Message-Id: <199609131455.KAA145440@acs4.bu.edu> >Subject: FW: WARNING (fwd) >To: kbowen@it.berklee.edu (Kelly Bowen), steves@pa.dec.com (Steve Schneider), > BGieskes@aol.com (Barbara Gieskes), Freeeeko@aol.com (Michiel Gieskes) >Date: Fri, 13 Sep 1996 10:55:40 -0400 (EDT) >Cc: ahartley@acs.bu.edu, lkehoe@acs.bu.edu (Lauren Kehoe), > jsiemon@acs.bu.edu (James Siemon), > wcarroll@acs.bu.edu (William Carroll), > kbazler@acs.bu.edu (Kirk Melnikoff), > paigemct@acs.bu.edu (Paige McThenia), > zleib@acs.bu.edu (Zahava Leibowitz), slyons@acs.bu.edu (Sarah Lyons), > mwalker@acs.bu.edu (Michael Walker), > kpetersn@acs.bu.edu (Kaara Peterson), > kbrumm@acs.bu.edu (Kathryn Brumm), > kmur-and@acs.bu.edu (Kathryn MurphyAnderson), > jillo@acs.bu.edu (Jill Orofino), > imlac@acs.bu.edu (Eric Johnson-DeBaufre), > jmulroon@acs.bu.edu (Jonathan Mulrooney), > colinh@acs.bu.edu (Colin Harris) >X-Mailer: ELM [version 2.4 PL25] >MIME-Version: 1.0 >Content-Type: text/plain; charset=US-ASCII >Content-Transfer-Encoding: 7bit > > > > ----- End Included Message ----- ----- End Included Message ----- From MikeMcQuade@connect.com Mon Sep 16 22:30:45 1996 From: MikeMcQuade@connect.com Date: Mon Sep 16 22:30:48 PDT 1996 Subject: Motorola MVME-160x Parallel Port Has anyone ever setup the parallel port on a Moto PowerPC 1603/4 board such that it generates interrupts ? Ive got a signal that I want to bring into my system via interrupt, Id like to use the /ACK line on the parallel port to generate the int. Thanks in advance. MikeMcQuade@connect.com From daemon@csg.lbl.gov Tue Sep 17 04:00:29 1996 From: daemon@csg.lbl.gov Date: Tue Sep 17 04:00:32 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Sep 17 04:00:25 PDT 1996 Subject: mvme162fx Subject: Re: How to get CPU time? Subject: Re: How to get CPU time? Subject: Re: Need recommendations on real-time OS and hardware Subject: iRMX port to VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: mvme162fx Date: Wed, 11 Sep 1996 13:19:46 GMT From: woodring.fam@MCI2000.com (Chuck Woodring) Organization: InternetMCI Message-ID: <51cqei$56t@news.internetmci.com> Hi, Has anyone successfully put a VxWorks boot image into the Flash RAM of the Motorola MVME162FX, 33MHz card. I'm having a heck of a time. Windriver is sending me a boot eprom but I would still like to get the code into the fllash. Any help/comments appreciated. Chuck Woodring NUWCDIVNPT woodring_C@fs107.npt.nuwc.navy.mil woodring.fam@mci2000.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get CPU time? Date: 13 Sep 1996 16:42:51 -0700 From: kla@leland.Stanford.EDU (Earl Mitchell) Organization: Stanford University, CA 94305, USA Message-ID: <51crhr$6te@amy14.Stanford.EDU> References: <32467211.249333462@netline-fddi.jpl.nasa.gov> <3238DD66.2FF5@finley.com> In article <3238DD66.2FF5@finley.com>, John Finley wrote: >Kevin D. Quitt wrote: >> >> I need my program to be able to determine the amount of CPU time it's had, as >> well as how much real-time has elapsed. The spy and timex libraries prove it >> can be done, but only print to the console. >.... > >Spy samples who the current task is at a fixed rate (via interrupts), >and calculates the percent time spent in each task by its frequency >in the overall total. So it's a statistical measure of CPU time in >each task, but that doesn't sound like what you want. > >I think timex just run a routine over and over, timing it each time, >until it thinks it has a good average measurement of the time the >routine takes. That is, if you run it 100 times, and 50 times it takes >3 ticks, and 50 times it takes 4 ticks, then you say the routine takes >about 3.5 ticks to execute. > >You could use the task switch hook if you need a "precise" measurement. >Each switch, you would see if your task is being switched to or from, >read the tick, and act accordingly. This would not account for interrupt >time, though. (I put "precise" in quotes, because you will only get >tick-resolution if you use the system ticker. If a faster (non- >interrupting) timebase is available, you can get better resolution.) > >I don't know how the commercial tools work. The spy tool reports cpu utilization per task in ticks. So the smallest unit of time it will detect is 1 tick. On a typical mv162 target clockrate for VxWorks is 60Hz (so 1 tick = 1/60 seconds = approx. 16 milliseconds). So if you run spy and a task never runs for more than 16 milliseconds the spy report will show it ran for 0 ticks in the sampled time period. Changing the clock rate will make tick granularity better but this will cause "thrashing" from too many interrupts. So if you need more granularity than a tick you must use something else. One option is "WindView" which has microsecond granularity. Another option is using a logic analyzer. If you just want to see if a task ran at all then use a hook. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get CPU time? Date: Mon, 16 Sep 1996 18:23:32 GMT From: kdq@emoryi.jpl.nasa.gov (Kevin D. Quitt) Organization: Speaking only for Myself Message-ID: <3242990c.586993702@netline-fddi.jpl.nasa.gov> References: <32467211.249333462@netline-fddi.jpl.nasa.gov> Reply-To: kdq@emoryi.jpl.nasa.gov As an experiment, I ran spy, then ran my program. To my amazement, the taskTicks field was now being maintained, and my program timed itself correctly! So I tried again using spyClkStart(), which didn't work, then I tried calling spyClkStart() from one task, then having it spawn the real program. This also didn't work. Why should this be so hard? - -- #include _ Kevin D Quitt USA 91351-4454 96.37% of all statistics are made up Per the FCA, this email address may not be added to any commercial mail list --------------------------- Newsgroups: comp.os.os9,comp.os.vxworks Subject: Re: Need recommendations on real-time OS and hardware Date: Thu, 05 Sep 1996 00:32:20 GMT From: afraser@hookup.net (Andrew Fraser) Organization: HookUp Communication Corporation, Ottawa, Ontario, CANADA Message-ID: <322e1936.82193449@nnrp.ott.hookup.net> References: <1996Sep4.192253.25115@nosc.mil> mitch@manta.nosc.mil (Ray Mitchell) wrote: >Greetings All, > This is a long-winded posting since my experience in this area >is limited. We are in the process of putting together a new system and >I need as much information as I can get regarding recommendations based >on everyone's prior experience. The system is intended to implement a >direction finding algorithm and support routines requiring some DSP. >Our current prototype is implemented on a VME system using a Force CPU30 >card, a DMA card, an extended memory card, and a custom hardware card. >We are currently using OS9 and doing all development on the same system that >runs the finished product. I know that this is vague but if possible I would >like recommendations on both software and hardware that are not only good >alone, but are know to work together compatibly. The following are some of >the requirements (and questions) as I see them: > >1. REAL TIME OPERATING SYSTEM > The two operating systems I have heard of in this regard are OS9 > (obviously) and VxWorks (which I've never used). What I need out > of whatever operating system I use is: Not many comments here since I'm hardware, but VxWorks is very commonly used in realtime applications. > A. Ability to compile and develop programs locally on the > same system that will run the finished product. Requiring > reasonable development only on a remote platform is > unacceptable (although the ability to do so is a plus). Vx is usually developed on a seperate Sun or equivalent, but I don't see why this would be a disadvantage. Tornado (ie VxWorks 5.3) seems to have some quite powerful graphical debugging tools. (And if your application crashes the system, you don't have to reboot your development environment?) > B. Real time response > C. Easy low level hardware access Via pointers in C. > D. Ability to read and write PC formatted SCSI disks (OS9 > does this acceptably but not perfectly) VxWorks can host a DOS file system. > E. Ethernet access for I/O of the finished product Available under Vx and I would assume every commercial RTOS; the actual interface may depend on the Ethernet card vendor's firmware? > F. A local GUI interface for the finished product You would probably get this via an X-Windows package for whichever video card you are using. Usually available from the video card manufacturer. You didn't mention including a video card, so I don't know where the GUI would be defined and run? > G. User friendly development environment and good vendor > support >2. HARDWARE > If possible I would like recommendations of products that are known > to work well together as well as with the real time OS chosen. > A. Bus architecture > We are currently using an old VME system. I assume > VME (with possibly VSB) is the way to go on this > system but am open to other suggestions. Whatever > is used must to be stable. Clearly VME, but then this is also the industry I work in! You can find just about any type of card you need, in several form factors and ruggedization levels. You may not need the VSB bus, since VME's bandwidth is quite high, but many CPU cards offer it. > B. CPU card > Currently we are using a Force CPU30(25MHz) with 4MB > on-card memory which is woefully inadequate. I would > like to use whatever the current consensus is for a > good, fast, reliable CPU card with at least 32MB of > on-card memory. My experience with Force sales has > been all bad and I assume is their tech support will > be even worse (although this is just my opinion). I'm > not committed to a particular processor and software > support and reliability are most important. Three > RS422 serial ports not dedicated to anything else are > required (either on the CPU card or another card). You have a pretty wide choice of processors; 680x0, PowerPC, Pentium, etc. 32 MB is easily available. Your serial channels may be entirely on the base card, or some may be from a daughter card. Either way it's a one slot solution. Browse through the on-line catalog at VITA (VMEbus Industry Trade Assoc), at http://www.vita.com to see what is available. > C. DSP card > This is the point where I know the least. I would > like a card with at least 32MB of DMAable RAM and > reasonable (whatever that means) compute power. We > are doing a few FFTs and matrix operations which the > current CPU can't even begin to keep up with. I don't > know whether multiple DSPs on the card are necessary > but if they are available they will probably be used. > C language support is required (and a DSP function > library is desired). Again, visit the VITA catalog. Several DSP processors and architectures are available. > D. DMA card > This card must have at least one 32bit DMA port (single > ended TTL is fine) and will be used to DMA into the DSP > memory. I understand that there is something called > the VSB bus which may be used for such DMA transfers > without tieing up the VME bus. This may be desirable > but I'm not sure it's required in our application. You don't need this. . The DMA card presently (I assume) reads from your custom card, and writes the data to the DSP card. That will be two cycles, since VME does not support "fly-by" transfers. All current VMEbus cards have a bus interface with a decent DMA built into them, so the DSP card would use it's own DMA to read from the data source. > >Thanks to everyone, >Ray Mitchell >mitch@nosc.mil - --------------------------------------------------------------------- ANDREW FRASER DY 4 Systems Inc email afraser@dy4.com Senior Hardware Designer 333 Palladium Dr. voice (613)599-9199x251 Product Support Group Kanata, Ontario fax (613)599-7777 Canada K2V 1A6 support support@dy4.com http://www.dy4.com sales sales@dy4.com - ------------------- CUSTOMER FIRST, QUALITY ALWAYS ------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: iRMX port to VxWorks Date: Wed, 11 Sep 1996 09:49:36 +1100 From: Steve Elliott Organization: Mayer Krieg & Co Message-ID: <3235F080.344@mkc.com.au> Reply-To: selliott@mkc.com.au Can anyone give guidance on porting iRMX applications to VxWorks. Thanks in advance. Steve Elliott --------------------------- End of New-News digest ********************** From uchenick@tis4000.com Tue Sep 17 04:58:49 1996 From: Gordon Uchenick Date: Tue Sep 17 04:58:50 PDT 1996 Subject: RE: vxWorks I/O system > Question: How do I make standard input NON-BLOCKING? I want to check > if there is a keyboard character input > before I use getc() function. Use fileno(stdin) to get the fd. Then you can either use select() or ioctl() with FIONREAD to find out if there is at least one byte present in the stdin buffer. You should not have to bother actually setting ioctl FIONBIO on fileno(stdin) if you use either one of these techniques. If you do set ioctl FIONBIO and then use read() on the fd, I think that you would get a -1 return and errno set to EWOULDBLOCK if there was nothing in the console buffer. However, be cautious because I haven't actually tested this. You should also be careful to test that your software is not adversely affected when someone telnets or rlogins into your shell. I think you'd be OK, but you never really know until you try it. HTH, Gordon Uchenick uchenick@tis4000.com "Life's a reach. Then you jibe." From jhillman@wrs.com Tue Sep 17 06:42:52 1996 From: Jon Hillman Date: Tue Sep 17 06:42:54 PDT 1996 Subject: Re: vxWorks I/O system At 11:54 AM 9/16/96 PDT, Steven wrote: > Hello, > > When the application make call to the VxWorks I/O system such as > open(), does VxWorks spawn the open() as a task or just a function > call? VxWorks makes a function call for the open, close, read, write, ioctl calls. These calls are part of the I/O system library, which looks up the device driver routines in a table look-up related to the name of the device. The appropriate routine is then called to do the operation. No tasks are spawned, although they my communicate with other tasks (e.g., if asynchronous I/O is used) which may manage some portion of the process. It is not efficient to spawn tasks which do a single function and dies. Rather tasks should be spawned once and (for example) accept input via message queues, etc. I tried to be brief but informative. Feel free to contact me directly for more details. _________________________ | Jon Hillman, FAE | | Wind River Systems | /) (407) 273-4646 (\ / ) Email:jhillman@wrs.com ( \ _( (|_________________________) ) /> (((\ \) /,) / ) / //))/ (\\\\ \_/ / \ \_/ ///// \ / \ / \ _/ \_ / ----/ /----------------------\ \---- / / \ \ From c_mcfarr@linkabit.titan.com Tue Sep 17 16:39:54 1996 From: c_mcfarr@linkabit.titan.com (Shawn McFarr) Date: Tue Sep 17 16:39:56 PDT 1996 Subject: Intel hex converter? Anyone knows where I can get a hold of a binary to convert an 'a.out' format to Intel hex? Thanks, Shawn McFarr From pardo@rti.com Wed Sep 18 00:37:26 1996 From: Gerardo Pardo-Castellote Date: Wed Sep 18 00:37:28 PDT 1996 Subject: Re: Load and run applications with telnet >> Hi vxWorker's >> >> has anybody out there been able to load and run a application, >> compiled for tornado, only with a telnet (rlogin host)? >> We do !not! want to use the "host based shell". >> >> Uwe What do you mean by "compiled for tornado". The application doesn't really change, however you need a kernel that has included all the necessary facilities (loader, shell, rlogin, telnet). You can use WindConfig or change configAll.h directly. In any case, you may want to look in target/config/all/configAll.h and search for the #ifdef INCLUDE_CONFIGURATION_5_2 to see what you need to define to have all the facilities on the target. The only "gotcha" is that the CrossWind debugger will not work (and similarly any other tools that use the symbol table in the target-server will not be aware of the code that you loaded directly to the target). You can still degug using vxgdb. -Gerardo =========================================================================== = = = = Gerardo Pardo-Castellote = email: pardo@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 155A Moffett Park Drive, Suite 111 = Fax: (408) 734-5009 = = Sunnyvale, CA 94089 = http://www.rti.com = = = = =========================================================================== From daemon@csg.lbl.gov Wed Sep 18 04:00:20 1996 From: daemon@csg.lbl.gov Date: Wed Sep 18 04:00:22 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Sep 18 04:00:17 PDT 1996 Subject: HELP - tgtsrvr bombs out on me! ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: HELP - tgtsrvr bombs out on me! Date: Tue, 17 Sep 1996 14:37:15 -0400 From: Yamudda Wezamibootz Organization: Data General Corporation, Westboro, MA Message-ID: <323EEFDB.6BE5@inigo.us.dg.com> I'm using tornado 1.0, VxWorks 5.2, trying to debug over TCP/IP ethernet connection. Here's what I get: Attaching backend... succeeded. Connecting to target agent... succeeded. Attached a.out OMF reader. Error: rpccore backend client Procedure unavailable Error: Can't read ctor/dtor strategy in target memory Error: rpccore backend client Procedure unavailable Error: rpccore backend client Remote system error Meanwhile, I can talk to my target via a shell task on a serial port. It pings my host just fine. In response to "i", it gives: NAME ENTRY TID PRI STATUS PC SP ERRNO DELAY - ---------- ------------ -------- --- ---------- -------- -------- - ------- ----- tExcTask _excTask 1fc9cc 0 PEND 3e4b69d 1fc93c 0 0 tLogTask _logTask 1fa0b4 0 PEND 3e4b69d 1fa020 0 0 tShell _shell 1b5964 1 READY 3e1e60c 1b564c 0 0 tWdbTask 3e432dc 1b6b0c 3 PEND 3e43a0d 1b69c0 0 0 dacsmssvc _dacsmssvc 1af480 20 PEND 3e43a0d 1af368 d0003 0 mulvio _mulvioserv 1acbc8 20 PEND 3e43a0d 1acab0 d0003 0 ampctodacse_ampctodacse 1aa310 20 PEND 3e43a0d 1aa1f8 d0003 0 tNetTask _netTask 1f5d7c 50 PEND 3e43a0d 1f5d24 0 0 tPortmapd _portmapd 1d8d78 100 PEND 3e43a0d 1d8c48 16 0 value = 0 = 0x0 ...What does it all mean? What can I do about it? Many thanks in advance. --------------------------- End of New-News digest ********************** From MAILER-DAEMON Wed Sep 18 17:04:39 1996 From: PMDF Mail Server Date: Wed Sep 18 17:04:43 PDT 1996 Subject: Undeliverable mail: Processing failure --Boundary (ID MsPJyuN7x3wVP7dxPwaWOA) Content-type: TEXT/PLAIN; CHARSET=US-ASCII These addresses were rejected: suan j henry: *** Unknown message recipient *** --Boundary (ID MsPJyuN7x3wVP7dxPwaWOA) Content-type: MESSAGE/RFC822 Return-path: vxwexplo@lbl.gov Date: Tue, 17 Sep 1996 06:00:34 -0500 (EST) From: vxwexplo@lbl.gov Subject: comp.os.vxworks newsdigest To: vxworks_users@csg.lbl.gov MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII; NAME=TEXT.TXT Content-transfer-encoding: X-UUENCODE begin 600 TEXT.TXT M4W5B;6ET=&5D+6)Y(&1A96UO;D!C'=O`T* M1&%T93H@5V5D+"`Q,2!397`@,3DY-B`Q,SHQ.3HT-B!'350-"D9R;VTZ('=O M;V1R:6YG+F9A;4!-0TDR,#`P+F-O;2`H0VAU8VL@5V]O9')I;F2!P=70@82!6>%=OB!C87)D+B`@22=M(&AA=FEN9R!A(&AE8VL@;V8@82!T:6UE+@T*5VEN9')I M=F5R(&ES('-E;F1I;F<@;64@82!B;V]T(&5P2P@0T$@.30S,#4L(%5300T*365S M2YC;VT^+"!*;VAN($9I;FQE M>2`@/&IO:&Y`9FEN;&5Y+F-O;3X@=W)O=&4Z#0H^2V5V:6X@1"X@475I='0@ M=W)O=&4Z#0H^/B`-"CX^($D@;F5E9"!M>2!P6]U('=O=6QD('-E92!I9B!Y;W5R('1A2!A;F0@82!T87-K(&YE=F5R(')U;G,@9F]R(&UO M2!B971T97(@8G5T('1H:7,@=VEL;`T*8V%U2!I;G1E6]U(&YE960@;6]R92!G2!T:&%N(&$@=&EC:R!Y;W4@;75S M="!U7IEF%T:6]N.B!3<&5A:VEN9R!O;FQY M(&9O7-E;&8-"DUE'!E M2!A;6%Z96UE;G0L('1H90T*=&%S:U1I8VMS(&9I96QD('=A4-L:U-T87)T M*"DL('=H:6-H(&1I9&XG="!W;W)K+"!T:&5N($D-"G1R:65D(&-A;&QI;F<@ M7-T96T@:7,@:6YT96YD960@=&\@:6UP;&5M96YT(&$-"CYD:7)E8W1I;VX@ M9FEN9&EN9R!A;&=O7-T96T@=7-I;F<@82!&;W)C92!#4%4S,`T* M/F-A2!C87)D+"!A M;F0@82!C=7-T;VT@:&%R9'=A2!G;V]D#0H^86QO;F4L M(&)U="!A%=O"!I2!T:&ES('=O=6QD(&)E(&$@9&ES861V86YT86=E+B`@5&]R;F%D M;R`H:64@5GA7;W)K6]U<@T*87!P;&EC871I;VX@8W)A6]U<@T*9&5V96QO<&UE;G0@96YV:7)O;FUE M;G0_/&<^*0T*/@D)0BX@(%)E86P@=&EM92!R97-P;VYS90T*/@D)0RX@($5A M2!T;R!R96%D(&%N9"!W%=O2!D97!E;F0@;VX@=&AE($5T:&5R M;F5T(&-A7-T M96TN("!)(&%S2!64T(I(&ES M('1H92!W87D@=&\@9V\@;VX@=&AI2!N;W0@;F5E9"!T:&4@5E-"(&)U2!#4%4@8V%R M9',@;V9F97(@:70N#0H^"0E"+B`@0U!5(&-A2X@($UY(&5X<&5R:65N8V4@=VET:"!&;W)C92!S86QE2!O<&EN:6]N*2X@($DG;0T*/@D)"6YO="!C M;VUM:71T960@=&\@82!P87)T:6-U;&%R('!R;V-E2!A71H:6YG(&5L2!B92!E;G1I0T*;VX@=&AE(&)A0T*/@D)"6)U="!I9B!T:&5Y(&%R92!A=F%I;&%B;&4@=&AE>2!W:6QL('!R M;V)A8FQY(&)E('5S960N#0H^"0D)0R!L86YG=6%G92!S=7!P;W)T(&ES(')E M<75I2UB>2(@=')A;G-F97)S+B`@06QL M#0IC=7)R96YT(%9-16)U2!-:71C:&5L;`T*/FUI=&-H0&YO#(U,0T*(%!R;V1U8W0@4W5P<&]R="!'"`@("@V,3,I-3DY+330N8V]M('-A;&5S("`@65R M($MR:65G("8@0V\-"DUE Date: Thu Sep 19 06:43:21 PDT 1996 Subject: Search for PC platform Hi VxWorkers, is there anyone who can recommend us a PC platform for use in industrial environment. We have the following requirements: Power supply: 24VDC or 115VAC / 230VAC Environment temperature up to 70 degrees C CPU: 486 33MHz RAM: 4 MB IDE interface: for HD min 10 MB or SANDISK Serial interface: RS485, 115kBaud, galv. sep., (16550) RS232, 19.2kBaud, (16550) Ethernet: RJ45, HW Jumper for NE2000/3COM Keyboard interface VGA LCD interface Parallel interface ============================================================ Paul Geiter Zellweger Uster phone +41 1 943 22 51 Wilstr 11 fax +41 1 943 38 11 8610 Uster email uzgeit@zellweger.ch ============================================================ From usd@mee.hwm.com Thu Sep 19 07:35:30 1996 From: Uwe Scheffold Date: Thu Sep 19 07:35:32 PDT 1996 Subject: Use of Tools.h++ iostreams WFC on SPARC host Hi vxWorker's has anybody out there installed (and in use) the WFC, Booch, Tools.h++ and iostreams for a SPARC target and host? What installation procedure did you take? I get some undefined (WFC) references on compiling the Kernel: vxWorks.tmp(.data+0x814): undefined reference to `__iosout_o' vxWorks.tmp(.data+0x818): undefined reference to `__manip_o' etc. So what can be done to clear this? Uwe From smith.ian@traf.com Thu Sep 19 08:13:25 1996 From: smith.ian@traf.com Date: Thu Sep 19 08:13:27 PDT 1996 Subject: Serial lines on 167 Hi We are trying to configure the serial lines on a MVME-167 . Does anyone have a note of what needs to be done to cd2400Sio.c, to permit the use of the hardware control functions SIO_HW_OPTS_SET and SIO_HW_OPTS_GET for character size,stop bits etc, that they would be willing to share. Many Thanks Ian =================================== Ian D. Smith (smith.ian@traf.com) Phone: +44- (114) -244 9971 Davy International =================================== From pardo@rti.com Thu Sep 19 11:47:28 1996 From: Gerardo Pardo-Castellote Date: Thu Sep 19 11:47:29 PDT 1996 Subject: RE: Load and run applications with telnet >> >> >> Snip >> When using tornado with the host shell configured in etc. >> >> The only "gotcha" is that the CrossWind debugger will not work. >> >> Firstly, you can use Crosswind. just type: >> target vxworks >> in the command window. It then uses RDB/RPC like vxgdb, but >> the interface is nicer. :) >> I hadn't tried that. Still, you can use either strategy: windshell-loading & CrossWind, or target-shell loading & CorssWind(VxGDB mode). But, assuming there are any dependencies) you cannot mix where you load the code because objects loaded though windsh will fail to link to objects loaded directly to the target and vice-versa. >> What would be really nice (WRS are you listening?) is some >> commmunication between the target tools and the host tools. >> e.g. if I load something using the tgtsvr it should check to >> see if I have a symbol table on the target and if so updates it >> and the module list etc. How radical! It would also be nice if >> when I ld on the target it causes the tgtsvr to update it's >> records. I completely agree, in fact we already developped such facility internally. >> This is a fundemental flaw in the way the system behaves at the >> moment. For years people have used startup scripts to load the >> relevant objects to the target when it boots, so they don't have >> to keep relinking BSPs. Now with tornado, as soon as you want to >> run a startup script, without host intervention, you need the >> old shell. You also need the old loader. So you can't see the >> module or any of the symbols from the tgtsvr, so the new flashy >> tools become almost unusable. This is definitely a problem, I wouldn't go as far as calling it a fundamental flaw in the design because WRS can probably solve it by adding the communicatuons you mentioned above and keeping the target and host sides consistent. >> Don't get me wrong, things are moving in the right direction. >> It's just I see a lot of people using this software and this is >> frequently a problem. I agree. -Gerardo =========================================================================== = = = = Gerardo Pardo-Castellote = email: pardo@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 155A Moffett Park Drive, Suite 111 = Fax: (408) 734-5009 = = Sunnyvale, CA 94089 = http://www.rti.com = = = = =========================================================================== From jhillman@wrs.com Thu Sep 19 21:26:58 1996 From: Jon Hillman Date: Thu Sep 19 21:27:00 PDT 1996 Subject: Re: Use of Tools.h++ iostreams WFC on SPARC host At 07:35 AM 9/19/96 PDT, Uwe wrote: > >Hi vxWorker's > >has anybody out there installed (and in use) the WFC, Booch, Tools.h++ >and iostreams for a SPARC target and host? > >What installation procedure did you take? > >I get some undefined (WFC) references on compiling the Kernel: >vxWorks.tmp(.data+0x814): undefined reference to `__iosout_o' >vxWorks.tmp(.data+0x818): undefined reference to `__manip_o' >etc. > >So what can be done to clear this? Uwe, At the risk of telling you something you may already know: Make sure you include INCLUDE_IOSTREAMS in your configuration before you build vxWorks. _________________________ | Jon Hillman, FAE | | Wind River Systems | /) (407) 273-4646 (\ / ) Email:jhillman@wrs.com ( \ _( (|_________________________) ) /> (((\ \) /,) / ) / //))/ (\\\\ \_/ / \ \_/ ///// \ / \ / \ _/ \_ / ----/ /----------------------\ \---- / / \ \ From daemon@csg.lbl.gov Fri Sep 20 04:00:16 1996 From: daemon@csg.lbl.gov Date: Fri Sep 20 04:00:19 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Sep 20 04:00:13 PDT 1996 Subject: Re: How to detect end of socket connection Subject: Programming PB-TIM2, VADI3, VADI4, PB-DIN2 Subject: Re: Subject: Sockets with a timeout Subject: Re: multi-protocol implementation Subject: Performace Data of VxWorks Subject: Re: debugging exceptions Subject: Re: multi-protocol implementation ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to detect end of socket connection Date: 11 Sep 1996 15:28:13 GMT From: David Perkinson Organization: TRW, Inc. Message-ID: <3236DACF.167E@boris.msfc.nasa.gov> References: <507e7g$j6v@fu-berlin.de> <50i2m3$oil@post.gsfc.nasa.gov> Alex Measday wrote: > > In article <507e7g$j6v@fu-berlin.de>, Heinz Junkes wrote: > > > >How can i check that a connection is still alive ??? > > > > If > > 1. select() says the socket is ready for reading and > 2. read() (or an ioctl(FIONREAD)) returns zero, > > your connection is broken. Note that you must perform the select() > before the read() or ioctl() since data may be received in between > the calls. > > Also see the "Socket Programming Frequently Asked Questions" at: > > http://www.interlog.com/~vic/sock-faq/ > > ======================================================================== > | Alex Measday Code 521 - Microelectronic Systems Branch | > | Integral Systems, Inc. NASA Goddard Space Flight Center | > | c.a.measday@ieee.org Greenbelt, MD | > ======================================================================== > ) http://purl.oclc.org/net/alex/ ( > ======================================================================== I have sockets that are connected to processes that don't always notify me when they are leaving. I have found the zero return of the read call to be a reliable check to determine that the connection is no longer valid. - -- =============================================================================== David Perkinson, TRW Opinions that may be expressed in this email perkinsond@boris.msfc.nasa.gov are my own and not neccessarily that of TRW. =============================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Programming PB-TIM2, VADI3, VADI4, PB-DIN2 Date: Thu, 19 Sep 1996 12:31:16 +0200 From: Bartlomiej Piwek Organization: University of Mining & Metallurgy Message-ID: I'm starting with VxWorks & VME programming, therefore I'm looking for examples of code for PB-TIM2, VADI3, VADI4, PB-DIN2. Is there any archive for VxWorks & devices source code ? Thanks, Bartek Piwek Dept. of Automatic Control, UMM, Krakow, Poland. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Subject: Sockets with a timeout Date: 11 Sep 1996 04:14:52 -0700 From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: References: <322B1EFE.59E2B600@bbn.com> <50i3ev$oil@post.gsfc.nasa.gov> <511h06$msi@post.gsfc.nasa.gov> In article <511h06$msi@post.gsfc.nasa.gov> alexm@vlsi.gsfc.nasa.gov (Alex Measday) writes: >Martin Shepherd wrote: >> Alex Measday wrote: >> >> > If you're using select(), there's no reason (that I'm aware of) for the >> > socket to be non-blocking. >>... >> least one byte is available. If you then attempt to read from the >> indicated socket, the socket may block if you attempt to read more >> than a single byte, whereas if the socket were set up for non-blocking >>... >If select() indicates that input is available, then a read(), blocking >or non-blocking, will return whatever is available. I stand corrected. My confusion is related to the following excerpt from Richard Steven's "Advanced Programming in the UNIX Environment". "When reading from a network, buffering within the network may cause less than the requested amount to be returned." This says nothing about how long read() will wait before giving up and returning what is available. Furthermore a cautious interpretation of this sentence would have the word "may" refering to what read() will do in the situation where there is insufficient data to satisfy the request. Can you name and/or provide a quote from any documentation that is more explicit on this point? >I don't know if there's any >difference in efficiency between a blocking and non-blocking read(). Well, since I was comparing reading one byte at a time to reading larger chunks, this is irrelevant in light of the above. Martin Shepherd (mcs@astro.caltech.edu) --------------------------- Newsgroups: comp.os.vxworks Subject: Re: multi-protocol implementation Date: 19 Sep 1996 11:26:10 -0700 From: Venu Organization: Zippo Message-ID: <51s382$372@lex.zippo.com> References: <51q4m5$rq@lex.zippo.com> In article , gerlach@netcom.com says... > > >It's been some time that I messed with VxWorks protocol stacks, but >at one time I worked on a product with IP and Link Layer SNA. > >In short the protocol stacks in VxWorks are not implemented as individual >tasks. There is no IP task per say. There is the netjob task which actually >executes the IP code. In our case the SNA Link Layer protocol stack >was implemented so that the code executes in the context of the net job >tasks as well. > >Usually in multi-protocol envirnments the different protocol don't usually >talk to each other directly. Usually the device driver hands the given >packet to the appropriate protocol stack and that's about it. One might >have an application that interacts with multiple protocols, like a >protocol translater, but in that case data is received from on protocol >and then transmittedd out another protocol. > Now I understand my question was confusing. Let me make it clear. I didn't intend to ask about inter-protocol communication. I wanted to ask how does multiple protocols talk to same MAC/physical interface. For example, there are two protocol stacks -- A and B -- and a MAC/phy interface C. One way to implement is to run all the three entities as single task. Other way is to implement all there as seperate tasks where A and B talk to C using message queues or any other inter-task communication mechanism. One another way is that A and C run as a task, and B run as a different task. Are there any more ways of implementing them? What are the advantages and disadvantages of these various scenarios? Venu >I hope this helps rather than confuses. > >Matthew >In article <51q4m5$rq@lex.zippo.com> Venu writes: >>Hi all, >> >>I would like to know how multi-protocol stacks (for e.g., TCP/IP and SPX/IPX), >>using same MAC interface (say 802.3), are implemented in VxWorks. I would >>appreciate if anybody can provide me information as to how these three entities >>(in the example) -- TCP/IP, SPX/IPX, and 802.3 -- are organized as tasks and >>how they communicate. >> >>Thanks, >>Venu > > --------------------------- Newsgroups: comp.os.vxworks Subject: Performace Data of VxWorks Date: Thu, 19 Sep 1996 12:31:28 -0400 From: Yafan An Organization: Northern Telecom Message-ID: <32417560.1115@nortel.ca> Reply-To: yafan@nortel.ca Hi there, I am trying to collect some performance data of VxWorks in order to compare it with OS9. The following are what I am interested: 1. context switching overhead; 2. memory management performance and details; 3. interupt latency; 4. networking library performance. Test data on 68xxx and MIPS are most interested. I think there must be some data somethere, I am trying to find. I appreciate any help. I am also interested in writing some performance testing routines for some specific functions, but I need to compile enough data in order to assist RTOS selection for our wireless phone product. Thanks, Yafan An, yafan@nortel.ca Computing Research Lab. Nortel, tel: 613-763-2536 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: debugging exceptions Date: 19 Sep 1996 17:50:39 -0700 From: kla@leland.Stanford.EDU (Earl Mitchell) Organization: Stanford University, CA 94305, USA Message-ID: <51spov$dtn@amy15.Stanford.EDU> References: <2.2c.32.19960913184516.0069af00@sc-mail1.corpwest.baynetworks.com> In article <2.2c.32.19960913184516.0069af00@sc-mail1.corpwest.baynetworks.com>, Hesham El Bakoury wrote: > > I would like to instrument the exception/handling code to be able to >print the values > of the PC (program counter registers) and other registers. > > Is there a way to do this ? my understanding is that I can not do printf >from within > an exception/interrupt handler. > > I am not sure also how I can get the value of the PC register. The following code fragments should answer your questions: 1. install signal handlers for for all possible exceptions SIGCONTEXT localContext; void sigInstall (void) { SIGVEC pVec; SIGVEC oVec; if (sigvec (SIGSEGV, NULL, &oVec) == ERROR) logMsg ("Cannot get info from signal vector\n"); pVec = oVec; pVec.sv_handler = sigHandler; if (sigvec (SIGSEGV, &pVec, &oVec) == ERROR) logMsg ("Cannot initialize signal vector\n"); /* .. others deleted .. */ } 2. Create a signal handler which dumps info like PC reg by accessing the VxWorks context info. This struct is CPU specific. Code below is for a 68020 cpu. void sigHandler (int sig, int code, SIGCONTEXT *sigContext) { unsigned short op_code, i; infoBuf.taskID = taskIdSelf(); infoBuf.signal = sig; infoBuf.code = code; localContext = *sigContext; logMsg ("*****************************************\n"); printSignal(); op_code = *(unsigned short *)localContext.sc_regs.pc; logMsg ("PC: %#08x SR: %#04x OP_CODE: %#04x SIG: %d\n", localContext.sc_regs.pc, localContext.sc_regs.sr, op_code, sig); /* * print out information to user and re-start task. */ printf ("A0-6:"); for (i = 0; i < 7; i++) printf(" %08lx", localContext.sc_regs.addrReg[i]); printf("\n"); printf ("D0-7:"); for (i = 0; i < 8; i++) printf(" %08lx", localContext.sc_regs.dataReg[i]); printf("\n"); logMsg("Restarting: %s (%#x)\n", taskName(infoBuf.taskID), infoBuf.taskID); taskRestart(infoBuf.taskID); } --------------------------- Newsgroups: comp.os.vxworks Subject: Re: multi-protocol implementation Date: Thu, 19 Sep 1996 21:21:11 -0700 From: "H.J. Bae" Organization: Peaceful Star Message-ID: <32421BB7.291B@peacefulstar.oakland.ca.us> References: <51q4m5$rq@lex.zippo.com> <51s382$372@lex.zippo.com> Reply-To: hjb@peacefulstar.oakland.ca.us Venu wrote: > > Now I understand my question was confusing. Let me make it clear. I didn't > intend to ask about inter-protocol communication. I wanted to ask how does > multiple protocols talk to same MAC/physical interface. For example, there > are two protocol stacks -- A and B -- and a MAC/phy interface C. One way to > implement is to run all the three entities as single task. Other way is to > implement all there as seperate tasks where A and B talk to C using message > queues or any other inter-task communication mechanism. One another way is > that A and C run as a task, and B run as a different task. Are there any more > ways of implementing them? In VxWorks, as well as other BSD-derived network software, interfaces to MAC/phy interface is done through various if_***() routines. For example, to send a packet from IP or any other network protocol, you will use if_output() indicating which interface to send the packet out, and pointer to the data to be sent. In practice, it does not make sense to partition out interfaces between MAC/phy interface and protocols into seperate tasks. There is no need, and it will slow things down. - -- H.J. Bae http://peacefulstar.oakland.ca.us mailto:hjb@peacefulstar.oakland.ca.us --------------------------- End of New-News digest ********************** From MikeMcQuade@connect.com Fri Sep 20 13:03:45 1996 From: Mike McQuade Date: Fri Sep 20 13:03:47 PDT 1996 Subject: RPC and VxWorks Im trying to use RPC to run functions on my VxWorks target, from a Unix application, and XDR to send data structs to the target. does anyone know of, or have any sample code for RPC ? Thank you. MikeMcQuade@connect.com From daemon@csg.lbl.gov Sat Sep 21 04:00:18 1996 From: daemon@csg.lbl.gov Date: Sat Sep 21 04:00:21 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Sep 21 04:00:16 PDT 1996 Subject: Boot failures Subject: Exit handler/task ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Boot failures Date: Fri, 20 Sep 1996 15:11:56 -0500 From: Alan Currie Organization: NASA/Johnson Space Center Message-ID: <3242FA71.6A85@ghgcorp.com> Reply-To: currie@ghgcorp.com I am having a problem that when I include my application to the vxWorks.st load, it auto reboots after I get the loading... xxxxxx xxxxxx xxxxxx. When I remove my application, and do a separate load from the shell after vxWorks has booted, it works fine. Any suggestions? I am using a DY4 SVME 163 board which has a 68040 processor. Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Exit handler/task Date: 20 Sep 1996 21:39:09 GMT From: wiltzius@coral.llnl.gov (David P Wiltzius,LC,21551,) Organization: Lawrence Livermore Nat'l Lab Message-ID: <51v2tt$6ar@lll-winken.llnl.gov> I'm using VxWorks 5.2. I'm finding the exit handler mechanism (e.g., taskDeleteHookAdd etc) rather clumsy. This exit task is invoked when any task is deleted, not just speicific tasks. And it runs in the context of the shell or exception task, depending on how the task is terminated, rather than in the context of the task that is terminating. I also tried using task variables to get some context of the exiting task, but I was unable to successfully (i.e., w/o error) invoke taskVarGet, even tho the task variables did successfully follow the context of the tasks - that is the taskVarAdd worked as expected. All this makes such "exit handlers" more challenging (not impossible tho!) to utilize. Or am I missing something (I consider myself a vxWorks novice)? Any suggestions or help welcomed. Thanks. Dave Wiltzius LLNL wiltzius@llnl.gov --------------------------- End of New-News digest ********************** From tkb@mclean.sparta.com Sat Sep 21 12:01:24 1996 From: "Thomas K. Buchanan" Date: Sat Sep 21 12:01:30 PDT 1996 Subject: Re: Boot failures >I am having a problem that when I include my application to the >vxWorks.st load, it auto reboots after I get the loading... xxxxxx >xxxxxx xxxxxx. When I remove my application, and do a separate load >from the shell after vxWorks has booted, it works fine. Any >suggestions? > >I am using a DY4 SVME 163 board which has a 68040 processor. Thanks. The big kernel and your application must fit under the boot code in RAM. You BSP comes setup so that the kernel itself can be peoperly loaded underneath the boot code. If you link in application code of any size, it may not fit any longer. The fix is to burn a new bootrom which loads itself to a higher location in RAM. ----------------------------------------- /\\ Keith Buchanan ///\\ /////\\ SPARTA, Inc. ///// \\ 7926 Jones Branch Drive / \\\\\ \\ Suite 900 \ \\\\\ // McLean, VA 22102 \ ////// \////// Voice: 703 448 0210 \//// FAX: 703 893 5494 \// WWW: www.mclean.sparta.com From Teerasak=Rengpraphun%EOL%MSDTWK@vines.msd.ray.com Sun Sep 22 10:05:51 1996 From: Teerasak=Rengpraphun%EOL%MSDTWK@vines.msd.ray.com Date: Sun Sep 22 10:05:53 PDT 1996 Subject: Console Tasks in VxWorks not Functioning I have an intermittent problem that may be related to VxWorks. Does anyone have a suggestion on what the problem is or how I may be able to obtain more information to get more clues on what the problem is? The system has a SUN GUI interface which talks to an MVME162-262 68040 processor card via ethernet. The boot and the kernel load to the processor card has no problem. Under normal running conditions the following tasks are running on the processor board: The asterisk next to the task name is an application generated tasks. NAME ENTRY TID PRI STATUS PC SP ERRNO DELAY ---------- ------------ -------- --- ---------- -------- -------- ------- ----- tExcTask _excTask fd8244 0 PEND 80140 fd81ac d0003 0 tLogTask _logTask fd5944 0 PEND 80140 fd58a8 0 0 tShell _shell fa3fa0 1 READY 4ead4 fa3c68 d0003 0 tRlogind _rlogind fb6118 2 PEND 29664 fb5f1c 0 0 tTelnetd _telnetd fb4274 2 PEND 29664 fb41a4 0 0 tRlogOutTas_rlogOutTask fff218 2 READY 29664 fff0b8 0 0 tRlogInTask_rlogInTask fa1700 2 READY 2963c fa1578 0 0 tRdbTask _rdbTask fb1844 20 PEND 29664 fb1724 d0003 0 tNetTask _netTask fd15b8 50 READY 4d248 fd1550 0 0 tFtpdTask _ftpdTask faf38c 55 PEND 29664 faf2d0 0 0 *EndOfFrame _eof dfb148 80 READY 29664 dfb100 0 0 *GOUCommsWS _GOUCommsWS e34c1c 82 PEND 29664 e346c0 0 0 *GOUCommsVME_GOUCommsVME e2fc6c 83 PEND 80140 e2f7d4 0 0 *DualPortMon_DualPortMon e2acbc 84 PEND+T 80140 e2a82c 3d0004 12 *MV200 _MV200Main e20d5c 85 PEND 80140 e208c8 0 0 *Ampex _AmpexMessag e25d0c 87 PEND 80140 e25874 0 0 tPortmapd _portmapd fb2d5c 100 PEND 29664 fb2c28 16 0 *DAC _DataAcquisi e111cc 102 PEND 80140 e10d24 0 0 *Display _Display d340f4 130 READY f1f966 d33e0c 0 0 value = 0 = 0x0 -> The mechanism used for inter-task communication is through VxWorks's msQLib.h which contains calls for creating a message queue and for sending and receiving these queues. The system would sometimes run all day, and sometimes fail several times a day. The time and occurrence of failure seems random, but the symptoms are exactly the same. Normally, the RUN light on the processor board is very active. At the time of the failure, the RUN light is still on but is very dim. The logic analyzer monitoring the VME backplane IRQs indicates that the processor board never came out of the interrupt service routine for IRQ3 because higher priority interrupts are still being processed, but IRQ3 remains activated. Then when lower priority IRQs are activated, they too remain unserviced. At the time of the failure, the tShell task for the console display is frozen. The console is directly connected from the console port of the processor board to the SUN Sparc station. Any comments or suggestions would be greatly appreciated. _______________________________________________________________________ Luke Rengpraphun Raytheon Company, ESD/EOL 50 Apple Hill Drive MS/T3LC21 Tewksbury, MA 01876 Tel : 508-858-1259 Fax : 508-858-5522 E-Mail : Teerasak=Rengpraphun@vines.msd.ray.com ________________________________________________________________________ From kelvin.martin@def.bae.co.uk Mon Sep 23 05:58:19 1996 From: kelvin.martin@def.bae.co.uk (MARTIN) Date: Mon Sep 23 05:58:21 PDT 1996 Subject: Unsubscribe unsubscribe From froeber@bbn.com Mon Sep 23 06:30:38 1996 From: Fred Roeber Date: Mon Sep 23 06:30:40 PDT 1996 Subject: Re: Console Tasks in VxWorks not Functioning Luke Rengpraphun wrote: > I have an intermittent problem that may be related to VxWorks. Does anyone > have a suggestion on what the problem is or how I may be able to obtain more > information to get more clues on what the problem is? > > The system has a SUN GUI interface which talks to an MVME162-262 68040 > processor card via ethernet. The boot and the kernel load to the processor > card has no problem. <> > The system would sometimes run all day, and sometimes fail several times a > day. The time and occurrence of failure seems random, but the symptoms are > exactly the same. > > Normally, the RUN light on the processor board is very active. At the time > of the failure, the RUN light is still on but is very dim. The logic > analyzer monitoring the VME backplane IRQs indicates that the processor board > never came out of the interrupt service routine for IRQ3 because higher > priority interrupts are still being processed, but IRQ3 remains activated. > Then when lower priority IRQs are activated, they too remain unserviced. > > At the time of the failure, the tShell task for the console display is > frozen. The console is directly connected from the console port of the > processor board to the SUN Sparc station. The most straight-forward way to track down a problem like this would be to use the WindView tool in "post mortem" mode. This way you would see exactly what was happening before the problem showed up. Lacking WindView, you need to resort to guess-work on what hung the system. From the task list you provided, it would appear that this processor is interfacing with a lot of other cards across the backplane. From what you say about IRQ3 it would appear that your processor gets hung up processing a higher level interrupt. Since the problem is random, it is probably due to some hardware glitch in one of the cards where under certain conditions it keeps the interrupt line high when it shouldn't causing an infinite interrupt condition. I would suspect any home-grown boards first. As far as finding out what interrupt is hanging things you can try a trick like toggling the FAIL LED control bit on and off in the system clock interrupt handler (use the TIMERCR_BDFLO bit in the VMECHIP2_TIMERCR register since you are using an MVME162LX). Then when the system locks up if the fail LED is still blinking then it means the interrupt that is locking the system is at a lower level than the system clock interrupt (level 6 by default). You can then change the system clock level by changing the SYS_CLK_LEVEL value in the mv162.h BSP file until you find which interrupt level is locking the system. Once you see that, figure out what devices are on that level and track things from there. Like I said, WindView would make it a lot simpler since you would be able to see directly what interrupt level hung things. Good luck. Fred -- | Fred J Roeber, BBN Systems & Technologies | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | | See http://www.bbn.com/tv for TraceMaker info | From leif.thuresson@tsl.uu.se Mon Sep 23 06:47:02 1996 From: leif.thuresson@tsl.uu.se (Leif Thuresson) Date: Mon Sep 23 06:47:03 PDT 1996 Subject: Driver for Kinetics 2917 CAMAC interface Hi, Does anyone know of a vxWorks driver for Kinetics 2917 VME bus interface for 3922 CAMAC crate controller ? Regards, /Leif ----------------------------------------------------------------------------- Leif Thuresson Tel: +46-18-183860 Uppsala University Fax: +46-18-183833 The Svedberg Laboratory E-mail: leif.thuresson@tsl.uu.se P.O.Box 533, S-751 21 Uppsala, Sweden From bqv@eng102.eng.i-o.com Mon Sep 23 07:06:49 1996 From: "Bang Q. Vu" Date: Mon Sep 23 07:06:51 PDT 1996 Subject: Re: RPC and VxWorks Mike McQuad penned: >Submitted-by MikeMcQuade@connect.com Fri Sep 20 13:03:45 1996 >Submitted-by: Mike McQuade > >Im trying to use RPC to run functions on my VxWorks target, from a Unix >application, and XDR to send data structs to the target. > >does anyone know of, or have any sample code for RPC ? > >Thank you. > > >MikeMcQuade@connect.com > There are many books on RPC programming. A good one is "Power Programming with RPC" by John Bloomer from O'Reilly & Associates. Tons of good RPC stuff in there, including codes that should get you going. Bang Vu IO Inc. From mea@mclean.sparta.com Mon Sep 23 07:13:56 1996 From: Mike Anderson Date: Mon Sep 23 07:13:58 PDT 1996 Subject: Re: Console Tasks in VxWorks not Functioning At 10:05 AM 9/22/96 PDT, you wrote: >Submitted-by Teerasak=Rengpraphun%EOL%MSDTWK@vines.msd.ray.com Sun Sep 22 10:05:51 1996 >Submitted-by: Teerasak=Rengpraphun%EOL%MSDTWK@vines.msd.ray.com > >I have an intermittent problem that may be related to VxWorks. Does anyone >have a suggestion on what the problem is or how I may be able to obtain more >information to get more clues on what the problem is? > [snip...] > >Normally, the RUN light on the processor board is very active. At the time >of the failure, the RUN light is still on but is very dim. The logic >analyzer monitoring the VME backplane IRQs indicates that the processor board >never came out of the interrupt service routine for IRQ3 because higher >priority interrupts are still being processed, but IRQ3 remains activated. >Then when lower priority IRQs are activated, they too remain unserviced. > >At the time of the failure, the tShell task for the console display is >frozen. The console is directly connected from the console port of the >processor board to the SUN Sparc station. > >Any comments or suggestions would be greatly appreciated. Fred Roeber's comments are certainly a good approach. However, just for yucks, why not try boosting the ISR_STACK_SIZE in configAll.h to something greater than the 1K default. It sounds like your ISR nesting level may be blowing away the hardware ISR stack. Do something wreckless like setting it to 2K ;-). HTH, =============================================================================== __ Real-Time System Development, Integration, Training and Services //\\ // \\ Mike Anderson // /\ \\ Chief Engineer Voice : (703) 448-0210 ext. 235 // / \ \\ SPARTA, Inc. FAX : (703) 734-3323 // \ \\ 7926 Jones Branch Drive EMAIL : mea@mclean.sparta.com \\ \ // Suite 900 Web : http://www.mclean.sparta.com \\ \ / // McLean, VA 22102 \\ \/ // "Software development is like making \\ // a baby... You can't make a baby in one \\// month by impregnating nine women. -- "Pride in Performance" Some things just take time." =============================================================================== From feng@bnlls1.nsls.bnl.gov Mon Sep 23 11:51:10 1996 From: Kate Feng Date: Mon Sep 23 11:51:12 PDT 1996 Subject: Re: Driver for Kinetics 2917 CAMAC interface > Hi, > Does anyone know of a vxWorks driver for Kinetics 2917 VME bus interface > for 3922 CAMAC crate controller ? > I worte it for VxWorks 5.1. You can get it from anonymous ftp ftp.nsls.bnl.gov:/nsls/pub/beamline.driver. ---------------------------------------------------------------------- S. Kate Feng E-mail: feng@bnlls1.nsls.bnl.gov Brookhaven National Laboratory, feng1@bnl.gov 725B, Upton, NY 11973 Phone: (516)344-2647 U.S.A. FAX: (516)344-4745 From aac!sol!applanix.com!babak@seraph.uunet.ca Tue Sep 24 07:44:35 1996 From: babak@applanix.com Date: Tue Sep 24 07:44:37 PDT 1996 Subject: BSP for AMPRO 486i and P5i ----- From: Bob Derak Does anyone know of a BSP for the following AMPRO boards: - AMPRO Little Board/486i - AMPRO Little Board/P5i If you do, I appreciate a note. Thanks. Email: babak@applanix.com ----- Bob Derak Applanix Corp. Markham, Ontario, Canada From Gerald.Pelissier@france.sun.com Tue Sep 24 09:25:49 1996 From: Gerald.Pelissier@france.sun.com (Gerald Pelissier - SME - SUN Microelectronics) Date: Tue Sep 24 09:25:51 PDT 1996 Subject: VxWORKS SBUS card drivers Dear Sir(s), I'm looking for a SBUS cards VxWORKS driver as SCSIS, Ethenet etc.. and more especially now, for an SBUS card VxWORKS driver for a SUN card called SBUS Quad Ethernet Controller, Ref. #X1058A. An application engineer at Wind River Systems advised me to post a request on this alias, because Wind River Systems doesn't provide such driver. Best regards, thank in advance. Gerald Pelissier ----------------------------------------------------------------------- ______ /_____/\ Gerald Pelissier, Southern Europe /_____\\ \ Field Application Engineer /_____\ \\ / Sun Microelectronics /_____/ \/ / / A Division of Sun Microsystems Inc. /_____/ / \//\ 13 Avenue Morane Saulnier - BP 53 \_____\//\ / / 78142 Velizy Cedex - France \_____/ / /\ / Email : gerald.pelissier@france.sun.com \_____/ \\ \ Phone : +33 1 30 67 52 78 \_____\ \\ Fax : +33 1 30 67 53 05 \_____\/ Mobile : +33 07 39 44 79 ------------------------------------------------------------------------ From cwcssk@leonis.nus.sg Tue Sep 24 18:00:11 1996 From: Suresh Santhana Krishnan Date: Tue Sep 24 18:00:14 PDT 1996 Subject: Anyone working on PCI with VxWorks! Hi, I am interested to know whether anyone working with PCI bus drivers for VxWorks. I am involved in programming S5933 (AMCC). If any help on PCI-BIOS access and method of data xfer (Mailbox, Fifo, Burst) will mostly be appreciated. Thanks, Suresh.S.K -- __^^^^^__ / @ @ \ =====================ooO==(_)=Ooo=========================== Suresh Santhana Krishnan Tel: 065-7715151 (O) Senior Research Engineer Fax: 065-7795441 (O) Centre For Wireless Communication Email: cwcssk@leonis.nus.sg (O) National University Of Singapore sureener@pacific.net.sg (H) Engineering Blk 2,#05-02, 10, Kent Ridge Crescent Singapore - 119260 ooO =======================( )-Ooo============================ \ ( ( ) \_) ) / (_/ From leonid@rst.co.il Wed Sep 25 00:13:26 1996 From: leonid@rst.co.il (Leonid Rosenboim) Date: Wed Sep 25 00:13:28 PDT 1996 Subject: Re: Console Tasks in VxWorks not Functioning Luke Rengpraphun wrote: > I have an intermittent problem that may be related to VxWorks. Does anyone > have a suggestion on what the problem is or how I may be able to obtain more > information to get more clues on what the problem is? Fred Roeber wrote: >From the task list you provided, it would appear that this processor >is interfacing with a lot of other cards across the backplane. From >what you say about IRQ3 it would appear that your processor gets hung >up processing a higher level interrupt. Since the problem is random, >... I agree with Fred, but would like to add some thoughts: From the description it seems that the interrupt that is looping infinitely is not a "spurious" one, it is one that has an interrupt handler installed, and It seems this interrupt handler either: a. has an infinite loop inside the ISR or b. does not do any system calls. Because if it did system calls for each iteration of the ISR, or if it was a spurious (unconnected) vector, then "panic: qorkQ overflow" would appear and system would restart. That should narrow down the options. Now in order to "debug" this situation, I can recommend another "primitive" technique - to add some debug messages to the suspected ISR, which are written to some memory instead of peing printed to logMsg(). This memory could be on another VME board, or simply in the EXC_MSG_ADRS area, writing to it as a circular buffer. So when the card is stock, you can access this area and see a couple of last messages, which should give you a hint. Note that EXC_MSG_ADRS is zeroed when you hit the RESET button (aka cold start). Activating the ABORT switch with an NMI is a better option, or else use NVRAM or something for the messages. Hope this helps. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 P.O.Box 11502, AZUR 58017, Israel Fax: +972-3-559-8244 WWW: http://www.rst.co.il E-Mail: leonid@rst.co.il From info@triada.si Wed Sep 25 05:32:46 1996 From: info@triada.si Date: Wed Sep 25 05:32:49 PDT 1996 Subject: subscribe subscribe info@triada.si From rich@immrc1.eng.mcmaster.ca Wed Sep 25 06:41:32 1996 From: rich@immrc1.eng.mcmaster.ca (Richard Teltz) Date: Wed Sep 25 06:41:34 PDT 1996 Subject: VxWorks TCP/IP backplane bandwidth Does anyone have any benchmark data for data transfer rates (maximum) when using the VxWorks implementation of TCP/IP over the backplane? I am particularly interested in the following configuration (or similar): -VMEbus, -Motorola 147 or 167 cards, -both stream and datagram comm. Also, has VxWorks implemented this over the PCI bus? Can anyone guide me to any good reference material for the PCI bus? Thanks in advance, Rich Teltz *------------------------------------------------------------------* | INTELLIGENT MACHINES AND MANUFACTURING RESEARCH CENTRE (IMMRC) | *------------------------------------------------------------------* | Richard Teltz, | Dept. of Mechanical Engineering,| | principal research engineer, | McMaster University, | | tel: (905) 525-9140, x27351 | Hamilton, Ontario, | | fax: (905) 572-7944 | Canada, | | email: rich@immrc.mcmaster.ca | L8S 4L7 | *------------------------------------------------------------------* From anthonyl@atlantis.softeng.lamrc.com Wed Sep 25 07:47:49 1996 From: Anthony Le Date: Wed Sep 25 07:47:51 PDT 1996 Subject: Re: Console Tasks in VxWorks not Functioning Hi Vxperts, I have a problem related to console window that needs to be solved. I use a "tip" program (run in a Xwindow UNIX) to interact with the console connected over a RS-232 line (9600 baud) of the SUN . Sometime it badly hung up specially when I display a lot of messages on the UNIX screen. If I use an external VT100 terminal, it is OK or if I rlogin to the target It's also OK too! Does anyone know why? Thanks for the answer. Anthony Le LAM RESEARCH CORP. anthony.le@lamrc.com From byron@wrs.com Wed Sep 25 14:38:16 1996 From: Byron Goulding Date: Wed Sep 25 14:38:20 PDT 1996 Subject: Adaptec Drivers Hello, I am a Field Applications Engineer for Wind River Systems. I have several customers looking for Adaptec SCSI drivers for the x86 target platform. Has anyone written a vxWorks driver for any Adaptec SCSI chips? Thanks for your attention. Regards, Byron Goulding Wind River Systems Field Applications Engineer (408)-562-5769 byron@wrs.com From shg@adv.sbc.sony.co.jp Thu Sep 26 02:09:37 1996 From: shg@adv.sbc.sony.co.jp (Stuart Gray) Date: Thu Sep 26 02:09:39 PDT 1996 Subject: Tornado Virtual I/O Hello - I am a bit confised about the virtual i/o mechanism that the target server implements under Tornado. When I boot my target and type "devs", /vf0 appears as a device. And sure enough, I can open a path to this redirect all target standard i/o to that channel (as the Users manual says). However, unless I specify that I want a virtual console window (when starting the tgtsvr) then all my host shells are connected to the virtual i/o channel. The manual page on tgtsvr says the following:: Virtual I/O Driver must be included in the target system. When this driver is included, any task on the target may open a virtual channel to read from, or write to, that chan- nel. On the host, any tool may open the same virtual chan- nel to write to, or read from, that channel. Thus the tar- get server acts as an I/O dispatcher, multiplexing whatever physical communications layer is available to allow run-time tasks and host tools to communicate easily. Is it possible to open multiple virtual io channels that connect to different host shells? And is it possible for different target tasks to redirect standard i/o to different virtual io channels? All the best... Stuart. ############################################################################ Stuart Gray New Technologies (D&D) Tel: +44 (0) 1256 455011 Sony Broadcast & Professional Europe DDI: +44 (0) 1256 483105 Jays Close Fax: +44 (0) 1256 811430 Viables Basingstoke EMail: shg@adv.sbc.sony.co.jp RG22 4SB United Kingdom ############################################################################# From shg@adv.sbc.sony.co.jp Thu Sep 26 02:36:29 1996 From: shg@adv.sbc.sony.co.jp (Stuart Gray) Date: Thu Sep 26 02:36:31 PDT 1996 Subject: Re: VxWorks TCP/IP backplane bandwidth >Submitted-by rich@immrc1.eng.mcmaster.ca Wed Sep 25 06:41:32 1996 >Submitted-by: rich@immrc1.eng.mcmaster.ca (Richard Teltz) > > >Does anyone have any benchmark data for data transfer >rates (maximum) when using the VxWorks implementation >of TCP/IP over the backplane? > >I am particularly interested in the following configuration >(or similar): > > -VMEbus, > -Motorola 147 or 167 cards, > -both stream and datagram comm. > >Also, >has VxWorks implemented this over the PCI bus? Can anyone >guide me to any good reference material for the PCI bus? > > >Thanks in advance, > >Rich Teltz Rich - Have you seen the PCI Special Intereset Group Web home page? Everything you ever wanted to know....... http://www.pcisig.com All the best... Stuart. ############################################################################ Stuart Gray New Technologies (D&D) Tel: +44 (0) 1256 455011 Sony Broadcast & Professional Europe DDI: +44 (0) 1256 483105 Jays Close Fax: +44 (0) 1256 811430 Viables Basingstoke EMail: shg@adv.sbc.sony.co.jp RG22 4SB United Kingdom ############################################################################# From daemon@csg.lbl.gov Thu Sep 26 04:00:20 1996 From: daemon@csg.lbl.gov Date: Thu Sep 26 04:00:22 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 26 04:00:16 PDT 1996 Subject: Boot failures Subject: Re: MVME167, FDDI and vxworks Subject: Linux and vxWorks Subject: Re: How to get CPU time? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Boot failures Date: Fri, 20 Sep 1996 15:11:56 -0500 From: Alan Currie Organization: NASA/Johnson Space Center Message-ID: Reply-To: currie@ghgcorp.com Reposting article removed by rogue canceller. I am having a problem that when I include my application to the vxWorks.st load, it auto reboots after I get the loading... xxxxxx xxxxxx xxxxxx. When I remove my application, and do a separate load from the shell after vxWorks has booted, it works fine. Any suggestions? I am using a DY4 SVME 163 board which has a 68040 processor. Thanks. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: MVME167, FDDI and vxworks Date: Wed, 25 Sep 1996 11:29:14 -0700 From: David McCullough Organization: Advanced Processing Laboratories, Inc. Message-ID: <324979FA.63DECDAD@sd.aplabs.com> References: <3247DB87.56CD@morgan.ucs.mun.ca> <32494351.18F2@lsoc114x.ksc.nasa.gov> Richard Price wrote: > > Bill Hunt wrote: > > > > I'm wondering about peoples experiences using the combination of a > > MVME167 CPU board, an Interphase 5211 FDDI board, and Vxworks. > > Interphase supplies a "sample" Vxworks driver for the board. What kind > > of performance might I expect from it ? How about third party drivers > > (the two I know of are from AP Labs and Sparta Inc). > > > > Any information you can share would be much appreciated. > > > > BillFirst understand that the 5211 is older,slower etc. than the 5411. but if that is what you've got > it is a fine product. We used the APL Driver (costly) and then researched inhouse drivers trying > different modes (UDP TCP IP. Depending upon word size and ring topology you may be able to write > your own driver. Certainly our driver is sold to make a profit. But one would be hard pressed to fund the necessary engineering effort to develop even a scaled down driver, for what we sell our driver for. The AP Labs 5211 VxWorks driver is a mature product that is currently used in a wide variety of real time systems. It is available for both SPARC and 68k targets. If is a full VxWorks network driver, supporting all protocols available under VxWorks. Performance will vary depending on whether you running an a 25mhz or 33mhz MVME167, your load level, and the packet size of your transmissions. Our sales organization can provide benchmarks. David McCullough Software Engineer AP Labs dave@sd.aplabs.com www.sd.aplabs.com (619) 546-8626 --------------------------- Newsgroups: comp.os.vxworks Subject: Linux and vxWorks Date: Tue, 17 Sep 1996 16:41:28 +0200 From: Casey Crellin Organization: CCII Systems Message-ID: <323EB898.15C1@ccii.co.za> Hi, A little while back there was a thread wrt Linux and vxworks Has anyone ported the 5.3 Tornado system to Linux? With advice from Lyor Goldstein I got the gnu tools going OK. Although I am having hassles with gdb. It is compiled for my target seems to be in order. WDBREADY comes through on the right device. But upon initialising the gdb for the target, it get malformed packets sent from the target and gdb shuts down... I am guessing but could this be because of the target server not being run? I tried to run tgtsvr undr dosemu but it cannot run in DOS mode. Could anyone out there spare me aclue? Casey - -- Casey Crellin casey@ccii.co.za CCII Systems (Pty) Ltd - ------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: How to get CPU time? Date: Fri, 13 Sep 1996 20:56:37 -0700 From: John Finley Organization: Netcom Message-ID: <323A2CF5.4AE5@finley.com> References: <32467211.249333462@netline-fddi.jpl.nasa.gov> <3238DD66.2FF5@finley.com> <51crhr$6te@amy14.Stanford.EDU> Reply-To: john@finley.com Earl Mitchell wrote: > .... > > The spy tool reports cpu utilization per task in ticks. So the smallest > unit of time it will detect is 1 tick. On a typical mv162 target clockrate > for VxWorks is 60Hz (so 1 tick = 1/60 seconds = approx. 16 milliseconds). > So if you run spy and a task never runs for more than 16 milliseconds the > spy report will show it ran for 0 ticks in the sampled time period. > Changing the clock rate will make tick granularity better but this will > cause "thrashing" from too many interrupts. > > So if you need more granularity than a tick you must use something else. > One option is "WindView" which has microsecond granularity. Another option > is using a logic analyzer. > > If you just want to see if a task ran at all then use a hook. Not really. Spy turns on an ISR driven by the aux clock at the rate you specify. On each interrupt, the spy ISR looks to see what task was interrupted, and credits one "tick" to the task that was interrupted. If the CPU was at interrupt level, then "interrupt" gets credited one tick; same for idle. The "ticks" column in the spy report is a little misleading; it should really be labelled "hits". So no matter how briefly a task executes, it might get caught by spy. In general, to get good spy results, you should: 1) Set the spy clock rate to something reasonably high, to get more samples. 2) Set the spy clock rate to something unrelated to (relatively prime to?) the main clock. This prevents tasks that do lots of taskDelay() calls from getting in or out of sync with the spy clock, and getting more or less hits than they deserve. 3) Let it run "a long time". Whatever that means.... A statistician could figure out how long to let it run at what spy clock rate to give whatever assurance you'd catch a task waking up every X microseconds and running for Y microseconds. With a long enough run, you *will* see hits on a task or ISR that runs for periods much less than the spy clock period. A habit I had once was "spy 10,777" on a 167/33 MHz. Spy's one of those neat tools that's simple, yet still does a very good job. But you really need to know what it's doing to use it correctly. I can only guess on how a windview works. And that guess is that it hooks the task switch, and reads a clock with very good resolution (usec). Does windview count interrupt time? If so, then it has to hook ISR entry and return, I think. John - -------------------------------------------------- John Finley Engineering Consultant john@finley.com Real Time Systems (619) 689-0032 Networking Protocols - -------------------------------------------------- --------------------------- End of New-News digest ********************** From froeber@bbn.com Thu Sep 26 05:15:12 1996 From: Fred Roeber Date: Thu Sep 26 05:15:14 PDT 1996 Subject: Re: Console Tasks in VxWorks not Functioning Anthony Le wrote: > I have a problem related to console window that needs to be solved. > I use a "tip" program (run in a Xwindow UNIX) to interact with the > console connected over a RS-232 line (9600 baud) of the SUN . > Sometime it badly hung up specially when I display a lot of messages > on the UNIX screen. Just an idea. If you run tip inside a cmdtool window on the SUN you can have problems. In fact, after some amount of time the window usually goes away because too much past data has been saved for scrolling back to. Things usually work better if you run tip from an xterm window or even a shelltool window. We do this with xterm windows on SUNs all the time with no problems. HTH. Fred -- | Fred J Roeber, BBN Systems & Technologies | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | | See http://www.bbn.com/tv for TraceMaker info | From leonid@rst.co.il Thu Sep 26 06:11:24 1996 From: leonid@rst.co.il (Leonid Rosenboim) Date: Thu Sep 26 06:11:27 PDT 1996 Subject: Re: How to get CPU time? > Date: Fri, 13 Sep 1996 20:56:37 -0700 > From: John Finley > ... > I can only guess on how a windview works. And that guess is > that it hooks the task switch, and reads a clock with very > good resolution (usec). Does windview count interrupt time? > If so, then it has to hook ISR entry and return, I think. You are quite right. WindView adds "instrumentation" i.e. hoks to all system events, including task switch, semaphore operations, and interrupt entry and exit. You can enable instrumentation of specific event classes, and when you do, a WindView function is added, and it records the event with a very precise time stamp. It is a very powerful tool, it has nothing to do with statistics when it comes down to measuring an event which occurs once in a very long period of time. Leaving our statistics means a better precentage of truth in the results. Still, it takes some experience to guess how the system would perform without the hooks installed, and without the overhead they add. I heard there is a hardware product by BBN that takes the measurement part and assotiated overhead out of the equation. I think it's in the TradeWinds directory on Wind River's Web site. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 P.O.Box 11502, AZUR 58017, Israel Fax: +972-3-559-8244 WWW: http://www.rst.co.il E-Mail: leonid@rst.co.il From froeber@bbn.com Thu Sep 26 06:20:46 1996 From: Fred Roeber Date: Thu Sep 26 06:20:49 PDT 1996 Subject: Re: spy use john finley wrote: > Spy's one of those neat tools that's simple, yet still does > a very good job. But you really need to know what it's doing > to use it correctly. > > A statistician could figure out how long to let it run at what > spy clock rate to give whatever assurance you'd catch a task > waking up every X microseconds and running for Y microseconds. > With a long enough run, you *will* see hits on a task or ISR > that runs for periods much less than the spy clock period. I totally agree with John. If you have periodic applications then spy will give you good information if you let it run long enough. I had a paper from a while ago where someone actually did the analysis to show how long you had to run things to get whatever accuracy you wanted on the task timing. One thing to note about measuring interrupt timings with spy, is that you will only get the timing for interrupt handlers and code that don't mask the aux clock interrupt that is used to generate the data needed for the spy report. I have run spy off of a non maskable interrupt before to get accurate timings including all interrupt levels and code that locked out interrupts but that was back when I had source access and could tweak the spy code a little to get it to work in this mode. > I can only guess on how a windview works. And that guess is > that it hooks the task switch, and reads a clock with very > good resolution (usec). Does windview count interrupt time? > If so, then it has to hook ISR entry and return, I think. WindView works entirely differently from spy (which is a profiler). With WindView, the object code for vxWorks has extra code added to many key points in the operating system to create a record of the execution of that code in a large circular buffer. Each of these event records is time stamped using a local high resolution clock. This mode of operation means that WindView is much more useful to look at the explicit timing of sequences of code including code that doesn't run periodically. The WindView manual has a complete description of all the points in the operating system that are instrumented (plus you can add your own instrumentation points). We actually sell a product called TraceMaker that makes WindView work even better. Fred -- | Fred J Roeber, BBN Systems & Technologies | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | | See http://www.bbn.com/tv for TraceMaker info | From MikeMcQuade@connect.com Thu Sep 26 08:03:28 1996 From: Mike McQuade Date: Thu Sep 26 08:03:30 PDT 1996 Subject: 100 Base-Tx PMC cards Is anyone using a 100Base-Tx PMC card ? if so who makes the card, and what are your thoughts on the card ? thank you. MikeMcQuade@connect.com From compston@radstone.co.uk Thu Sep 26 09:52:33 1996 From: David Compston Date: Thu Sep 26 09:52:35 PDT 1996 Subject: Re: 100 Base-Tx PMC cards Replying to: Mike McQuade,Thu, 26 Sep 96 17:16:41 BST > >Is anyone using a 100Base-Tx PMC card ? > >if so who makes the card, and what are your thoughts on the card ? This question seems to crop up from time to time. Radstone has a 100Base-TX PMC adapter that comes complete with VxWorks driver for PowerPC -- checkout our website for further details... http://www.radstone.com/pmc.html Regards -- David Compston compston@radstone.co.uk Radstone Technology PLC Tel: +44 (0)1327 359444 Water Lane, Towcester Fax: +44 (0)1327 358113 NN12 6JN, UK Web: http://www.radstone.com/ From mea@mclean.sparta.com Thu Sep 26 10:47:57 1996 From: Mike Anderson Date: Thu Sep 26 10:47:59 PDT 1996 Subject: Re: MVME167, FDDI and vxworks At 04:00 AM 9/26/96 PDT, you wrote: > >Richard Price wrote: >> >> Bill Hunt wrote: >> > >> > I'm wondering about peoples experiences using the combination of a >> > MVME167 CPU board, an Interphase 5211 FDDI board, and Vxworks. >> > Interphase supplies a "sample" Vxworks driver for the board. What kind >> > of performance might I expect from it ? How about third party drivers >> > (the two I know of are from AP Labs and Sparta Inc). >> > >> > Any information you can share would be much appreciated. >> > >> > BillFirst understand that the 5211 is older,slower etc. than the 5411. but if that is what you've got >> it is a fine product. We used the APL Driver (costly) and then researched inhouse drivers trying >> different modes (UDP TCP IP. Depending upon word size and ring topology you may be able to write >> your own driver. > >Certainly our driver is sold to make a profit. But one would >be hard pressed to fund the necessary engineering effort to >develop even a scaled down driver, for what we sell our driver >for. The AP Labs 5211 VxWorks driver is a mature product that >is currently used in a wide variety of real time systems. It is >available for both SPARC and 68k targets. If is a full VxWorks >network driver, supporting all protocols available under VxWorks. > >Performance will vary depending on whether you running an a 25mhz >or 33mhz MVME167, your load level, and the packet size of your >transmissions. Our sales organization can provide benchmarks. > > Speaking for SPARTA, I can certainly agree with Dave's sentiments. Clearly, a company who writes a driver and then offers it for sale is at least trying to recoup their development costs and hopefully make a little extra ;-). There are generally two pricing models that companies will follow. One, price it high and sell only a few units. This was certainly WRS's pricing model for VxWorks B.T. (Before Tornado). At $23K+, this price was too high for the 1-2 man company to be able to purchase the product, especially given that they'd have to buy another ODL if they ever got a second project. But, when your target marketplace is the GMs, TRWs and HPs of the world, they can afford it. The other model is price it cheap and try to make it up in volume. I refer to this as the Turbo Pascal model. This is where you find the $5K (and less) real-time kernels for the X86 platforms. But, think about how much money a senior engineer will bill out for on a daily basis. It's not at all uncommon to pay $1K/day for a guy who really knows his stuff as a consultant. When you think of it in those terms, even the old VxWorks pricing was a bargain. Because it would certainly take more than 23 days to duplicate all of the facilities of VxWorks. So, as for my view point on drivers, here we tend to look at those markets were we believe there's sufficient volume and then use the Turbo Pascal model. For instance, our 5211 driver sells for $800 per target and you get the source code. We support 68K, SPARC, MIPS and have versions running on PPC. Sure, folks could try to cheat us, but for $800? Why bother to risk the litigation? We also tend to assume that people are basically honest and giving them the source code gives us a fighting chance at customer support (just change line 572 and recompile...). Of course, even with the source, we run into a fair share of problems with all of the possible combinations of hardware that exists out there. Especially since our FDDI driver supports VME32, VME32 BLT and VME64 MBLT. Setting these options up on the arbitrary VxWorks system can be non-trivial. But that's why you *pay* for the driver so you can get some tech support when you need it. Therefore, before rejecting that driver as being "too expensive". Think about how much time it would *realistically* take to duplicate the effort and multiply that out through the loading factors of salary and benefits. You'll find that in many cases, it's cheaper to buy it than built it yourself. Regards, =============================================================================== __ Real-Time System Development, Integration, Training and Services //\\ // \\ Mike Anderson // /\ \\ Chief Engineer Voice : (703) 448-0210 ext. 235 // / \ \\ SPARTA, Inc. FAX : (703) 734-3323 // \ \\ 7926 Jones Branch Drive EMAIL : mea@mclean.sparta.com \\ \ // Suite 900 Web : http://www.mclean.sparta.com \\ \ / // McLean, VA 22102 \\ \/ // "Software development is like making \\ // a baby... You can't make a baby in one \\// month by impregnating nine women. -- "Pride in Performance" Some things just take time." =============================================================================== From mea@mclean.sparta.com Thu Sep 26 10:56:00 1996 From: Mike Anderson Date: Thu Sep 26 10:56:02 PDT 1996 Subject: Re: Boot failures At 04:00 AM 9/26/96 PDT, you wrote: > >I am having a problem that when I include my application to the >vxWorks.st load, it auto reboots after I get the loading... xxxxxx >xxxxxx xxxxxx. When I remove my application, and do a separate load >from the shell after vxWorks has booted, it works fine. Any >suggestions? > >I am using a DY4 SVME 163 board which has a 68040 processor. Thanks. > I think that Keith Buchanan from SPARTA answered this one already, but I believe that the problem is that your standalone load image is too large to fit under the RAM_HIGH_ADRS defined location for the ROM in memory. That location is set for 0x90000 by default. This means that if your load image is > ~576K bytes, you'll overwrite the boot ROM image in RAM and the load process will die a horrible death. What you need to do to fix this problem is to redefine the RAM_HIGH_ADRS value to something > 0x90000 and remake your boot ROMs. Remember to make these changes in both the config.h and Makefile.XXX.gnu files before rebuilding the ROMs. Regards, =============================================================================== __ Real-Time System Development, Integration, Training and Services //\\ // \\ Mike Anderson // /\ \\ Chief Engineer Voice : (703) 448-0210 ext. 235 // / \ \\ SPARTA, Inc. FAX : (703) 734-3323 // \ \\ 7926 Jones Branch Drive EMAIL : mea@mclean.sparta.com \\ \ // Suite 900 Web : http://www.mclean.sparta.com \\ \ / // McLean, VA 22102 \\ \/ // "Software development is like making \\ // a baby... You can't make a baby in one \\// month by impregnating nine women. -- "Pride in Performance" Some things just take time." =============================================================================== From schen3@ccgate.hac.com Thu Sep 26 20:33:31 1996 From: schen3@ccgate.hac.com Date: Thu Sep 26 20:33:33 PDT 1996 Subject: Bare-Bone configuration? We try to cut vxWorks OS to bare-bone, I need to know what option need to be included or excluded. Can someone tell me where I can find all the descriptions for all the facilities (INCLUDE_*) ? Thanks in advance. Steven schen3@ccgate.hac.com From daemon@csg.lbl.gov Fri Sep 27 04:01:00 1996 From: daemon@csg.lbl.gov Date: Fri Sep 27 04:01:02 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Sep 27 04:00:58 PDT 1996 Subject: NTP for VxWorks PPC603e ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: NTP for VxWorks PPC603e Date: Tue, 17 Sep 1996 16:25:09 +0200 From: Thomas Rosen Organization: ABB Sweden Message-ID: <323EB4C5.25B0@rly.abb.se> Hi there! I need to know if there is a NTP (version3) implementation for PPC603 available somewhere.... I'll be very glad to see an example in C :-) - Thomas [rlythro@rly.abb.se] --------------------------- End of New-News digest ********************** From gema@force.de Fri Sep 27 05:11:02 1996 From: gema@force.de (Gerhard Mayr) Date: Fri Sep 27 05:11:04 PDT 1996 Subject: Re: 100 Base-Tx PMC cards > >Is anyone using a 100Base-Tx PMC card ? > >if so who makes the card, and what are your thoughts on the card ? FORCE COMPUTERS has a 10Base/100BaseT PMC module in their product portfolio. The drivers for VxWorks on the PowerPC are also available. Please contact one of our sales offices. More infos on our web-page: http://www.forcecomputers.com Best regards, Gerhard Mayr ------------------------------------------------------------------- Gerhard Mayr FORCE COMPUTERS GmbH Tel: +49-89-60814-300 Prof.-Messerschmitt-Str. 1 Fax: +49-89-60814-276 D-85579 Neubiberg/Munich Email: gema@force.de Germany http://www.forcecomputers.com ------------------------------------------------------------------- From Lennart.Bie@emw.ericsson.se Fri Sep 27 07:10:27 1996 From: Lennart.Bie@emw.ericsson.se (FY/DX Lennart Bie) Date: Fri Sep 27 07:10:29 PDT 1996 Subject: Driver to MVME1604 serial 3&4 Can anyone give a hint where to find a synchronous driver using dma transfers operating the Z85230 (serial 3&4) on the Moto MVME1604 PPC board. Thanks in advance. Lennart Bie, Ericsson Microwave Systems AB, Molndal, Sweden From Richard_W_Mattox@ccmail.orl.mmc.com Fri Sep 27 10:53:32 1996 From: Richard_W_Mattox@ccmail.orl.mmc.com Date: Fri Sep 27 10:53:35 PDT 1996 Subject: SCSI Driver Hello out there, I am trying to use the scsiIoctl function with the FIOSCSICOMMAND parameter to send a SCSI command to a direct access device. I followed the code example in the Online Manuals of the scsiLib. I can use the library routine scsiTestUnitRdy invoked from windsh, but if I construct the TEST_UNIT_READY command and send it using the scsiIoctl function it fails. I need to send a PREVENT ALLOW MEDIUM REMOVAL SCSI command and VxWorks does not provide a library routine for this command. Therefore, I need the scsiIoctl function to work. If anybody can help I would greatly appreciate it. The architecture is a MVME-1603 PowerPC board running version 5.3 of VxWorks. I know the SCSI2 driver is BETA so maybe this is a bug. Thanks in advance, Rick Mattox ( richard_w_mattox@ccmail.orl.lmco.com ) From stan@rti.com Fri Sep 27 16:01:35 1996 From: Stan Schneider Date: Fri Sep 27 16:01:37 PDT 1996 Subject: Re: spy use >> > A statistician could figure out how long to let it run at what >> > spy clock rate to give whatever assurance you'd catch a task >> > waking up every X microseconds and running for Y microseconds. >> > With a long enough run, you *will* see hits on a task or ISR >> > that runs for periods much less than the spy clock period. >> >> I totally agree with John. If you have periodic applications then >> spy will give you good information if you let it run long enough. >> I had a paper from a while ago where someone actually did the analysis >> to show how long you had to run things to get whatever accuracy you >> wanted on the task timing. >> Hey! I wrote that paper! :-) I'll post it to our web page next week & announce when it's up. It analyzes the statistics of sampling profilers. The accuracy available with even modest sampling is really pretty surprising. The analysis applies to all statistical profilers, including spy and also our dynamic execution profiler called "ScopeProfile". Since there seems to be some confusion in this thread, I'll take the liberty of pointing out some of the differences between Spy, WindView, and ScopeProfile. All these tools are useful for tuning your application code for maximum performance. They are very different and mostly complementary. Spy is a task-level profiler; it shows you how much CPU each task is using. ScopeProfile is a *function-level* profiler; it analyzes how much CPU each routine in your system is using (much finer resolution information than reported by spy). WindView shows when events occur, which tasks are running, and when and why changes occur. It's not really a profiler. Spy reports snapshots of task CPU usage. It's useful to see at large granularity which tasks are hogs. One caveat: spy should be run with a non-integer multiple of the system clock (preferably a relative prime), and should collect at least 1500 samples or so per report. The default parameters (5,100) are horrible. spy(15,107) will work for most apps. These tools give you different information. They are most useful in different situations. For example, if your application is thrashing (spending most of its time context switching), ScopeProfile (and spy) will simply report that most of the CPU is being spent in the kernel. WindView will show you which tasks are switching, and display the events (e.g. interrupts) that are causing the switches. It will show how long it took to respond to the interrupts, what was delayed, etc. Windview is clearly the tool to use in this situation. On the other hand, if you have one task hogging the CPU, WindView (and spy) will only tell you that one task is running most of the time. ScopeProfile will provide a detailed function-by-function analysis, breaking down the individual routines within the task that are burning the CPU. It will give you a map of what that task is doing, what routines are being called, what routines they call, and point out the inefficiencies. You could sprinkle user events around your code with WindView, but that's not really practical. This is the domain of a profiler. So, all three tools help you tune the performance of your code. However, they provide different perspectives: WindView offers precise measurements of event timing, ScopeProfile provides a detailed functional analysis of where your CPU time is being used. Spy simply tells you which tasks are active. HTH, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 x104 = = 155A Moffett Park Drive, Suite 111 = Fax: (408) 734-5009 = = Sunnyvale, CA 94089 = http://www.rti.com = = = = ============================================================================= From richard@carapace.demon.co.uk Sat Sep 28 00:57:11 1996 From: rms Date: Sat Sep 28 00:57:13 PDT 1996 Subject: ethernet I am currently using vxWorks on a PC arechetecture but I am going to move to a platform where everyting is made by ourselves - we would still use the intel processors but we will be glueing down most of the peripherals. The target is a 'process controller' with little or no man- machine interface and certainly no VDU or keybaord. It will therfore not be much like a PC any more! I need to put a network chipset for one ( or posibly two ) network interfaces on the board. The reason for multiple interfaces is twofold firstly for relyability and secondly so that time critical data will not get held up by heavy traffic. What chipsets do people recomend - of course vxWorks drivers are a requirement. We woudld like to support 100mbps/10mbps or at least have a way of migrating to faster LANs. Thanking you in anticipation.... -- rms From AS_Bridge@msmail3.hac.com Sat Sep 28 03:32:44 1996 From: "AS Bridge" Date: Sat Sep 28 03:32:46 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: ethernet Sent: Sat, Sep 28, 1996 2:03 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 3:31 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 06:06:16 1996 From: "AS Bridge" Date: Sat Sep 28 06:06:25 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 4:38 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 5:39 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 08:28:23 1996 From: "AS Bridge" Date: Sat Sep 28 08:28:25 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 7:03 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 8:27 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 10:26:25 1996 From: "AS Bridge" Date: Sat Sep 28 10:26:27 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 9:21 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 10:25 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 12:27:30 1996 From: "AS Bridge" Date: Sat Sep 28 12:27:32 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 11:24 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 12:26 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 14:49:51 1996 From: "AS Bridge" Date: Sat Sep 28 14:49:53 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 1:21 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 2:48 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 16:55:59 1996 From: "AS Bridge" Date: Sat Sep 28 16:56:01 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 3:49 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 4:52 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 19:40:05 1996 From: "AS Bridge" Date: Sat Sep 28 19:40:07 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 6:06 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 7:38 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sat Sep 28 21:54:12 1996 From: "AS Bridge" Date: Sat Sep 28 21:54:14 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sat, Sep 28, 1996 8:50 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sat, Sep 28, 1996 9:53 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 01:05:15 1996 From: "AS Bridge" Date: Sun Sep 29 01:05:17 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 12:01 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 1:03 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From lyorg@scorpio.com Sun Sep 29 02:02:43 1996 From: lyorg@scorpio.com (Lyor Goldstein) Date: Sun Sep 29 02:02:45 PDT 1996 Subject: SMTP client Hello, I am posting this simple SMTP client which enables sending e-mail from a VxWorks target to any SMTP server. I make no guarantees or warranties as to its performance (since I have not tested it thoroughly). However, it seems simple enough and may well serve as a basis for further development. By doing this I hope to promote useful utilities code exchange in the spirit of the FSF. I ask only that my name be left as the author for anyone using this code. ***************************************************************** * Lyor Goldstein. Email : lyorg@scorpio.com * * Scorpio Communications Ltd. Phone : 972-3-5339654 ext 115 * * Software group Fax : 972-3-5339518 * * 1c Yoni Netanyahu st. * * Or Yehuda 60250 * * P.O.Box 564, Israel * ***************************************************************** #include #include #include #include #include #include #include #include #include #include #include #include #include /***************************************************************** * Lyor Goldstein. Email : lyorg@scorpio.com * Scorpio Communications Ltd. * Software group *****************************************************************/ /*---------------------------------------------------------------------------*/ #define SMTP_E_SYS_STAT 211 /* System status, or system help reply */ #define SMTP_E_HELP_MSG 214 /* Help message */ #define SMTP_E_DOMAIN_RDY 220 /* Service ready */ #define SMTP_E_DOMAIN_CLS 221 /* Service closing transmit channel */ #define SMTP_E_ACTION_OK 250 /* Requested mail action okay, completed */ #define SMTP_E_USR_NLOCAL 251 /* User not local; will forward to */ #define is_bad_smtp_rcode(c) (((c) < 200) || ((c) >= 300)) #define SMTP_E_START_INP 354 /* Start mail input; end with . */ #define SMTP_E_SRVC_NA 421 /* Service not available */ #define SMTP_E_ACTION_NA 450 /* Requested mail action not taken */ #define SMTP_E_ACTION_ABRT 451 /* Requested action aborted */ #define SMTP_E_ACTION_MEM 452 /* insufficient system storage */ #define SMTP_E_SYNTAX 500 /* Syntax error, command unrecognized */ #define SMTP_E_PARAMS 501 /* Syntax error in parameters or arguments */ #define SMTP_E_CMD_NA 502 /* Command not implemented */ #define SMTP_E_BAD_CMD_SEQ 503 /* Bad sequence of commands */ #define SMTP_E_BAD_PARAM 504 /* Command parameter not implemented */ #define SMTP_E_MBOX_NA 550 /* mailbox unavailable */ #define SMTP_E_USR_TRY 551 /* User not local; please try */ #define SMTP_E_MEM_OVRFLW 552 /* exceeded storage allocation */ #define SMTP_E_MBOX_NAME 553 /* mailbox name not allowed */ #define SMTP_E_TRAN_FAILED 554 /* Transaction failed */ /*---------------------------------------------------------------------------*/ static int smtpOpen (const char hname[], int pnum) { int s=(-1), t=1; struct sockaddr_in sin; struct sockaddr *saddrp=(struct sockaddr *) &sin; bzero((char *) &sin, (sizeof sin)); sin.sin_family = AF_INET; sin.sin_port = htons(pnum); if (isdigit((int) hname[0])) { if ((sin.sin_addr.s_addr=inet_addr(hname)) == ((u_long) ERROR)) return (-1); } else { if ((sin.sin_addr.s_addr=hostGetByName(hname)) == ERROR) return (-1); } s = socket(AF_INET, SOCK_STREAM, 0); if (s < 0) return (-1); if((setsockopt(s,SOL_SOCKET,SO_KEEPALIVE,(void *)&t,(sizeof t))) != OK) { close(s); return (-1); } if((setsockopt(s,IPPROTO_TCP,TCP_NODELAY,(void *)&t,(sizeof t))) == -1) { close(s); return (-1); } if (connect(s, saddrp, (sizeof sin)) != OK) { close(s); return (-1); } return s; } /*---------------------------------------------------------------------------*/ static STATUS smtpClose (FILE *fp, int s, const char reason[], STATUS retval) { if ((fp != NULL) && (fp != stdin)) fclose(fp); if (s > 0) close(s); return retval; } /*---------------------------------------------------------------------------*/ static STATUS smtpGetResponse (int s, char l[], int *rcode) { char *lp=l; do { if (read(s, lp, sizeof(char)) != sizeof(char)) return ERROR; if (*lp == '\n') break; lp++; } while (*(lp-1) != '\n'); *lp = '\0'; if (!isdigit((int) l[0])) return ERROR; sscanf(l, "%d", rcode); #ifdef VERBOSE fprintf(stderr, "RSP: %s\n", l); #endif return OK; } /*---------------------------------------------------------------------------*/ static STATUS smtpSendCmdSync (int s, FILE *fp, const char cmd[], char l[], int expcode) { int rcode; if (write(s, cmd, strlen(cmd)) <= 0) return smtpClose(fp, s, cmd, ERROR); /* skip all OK codes til ACTION-OK code found */ do { if (smtpGetResponse(s, l, &rcode) != OK) return smtpClose(fp, s, l, ERROR); if (rcode == expcode) return OK; if (is_bad_smtp_rcode(rcode)) return smtpClose(fp, s, l, ERROR); } while (rcode != SMTP_E_ACTION_OK); return OK; } /*---------------------------------------------------------------------------*/ static STATUS smtpAddRecepient (int s, const char rcpt[], char l[], int *rcode) { sprintf(l, "RCPT TO: %s\n", rcpt); if (write(s, l, strlen(l)) <= 0) return smtpClose(NULL, s, "'RCPT TO:' failed", ERROR); if (smtpGetResponse(s, l, rcode) != OK) return smtpClose(NULL, s, "bad 'RCPT TO:' response", ERROR); return OK; } /*---------------------------------------------------------------------------*/ #define MAX_LL 127 /* max input line fro text file */ #define MAX_CMD 79 /* max command/response */ int smtpDataSockOpen (const char host[], const char fromUsr[], const char *tusrs[], int port) { static const char data_cmd[]="DATA\n"; int pnum=(port != 0) ? port : IPPORT_SMTP; static const char defhost[]="bootHost"; const char *hname=(host != NULL) ? host : defhost; char usr[32], *stp, *enp; const char *fusr=fromUsr; int s=(-1), rcode, i, tnum; char l[MAX_CMD+1]; if (fusr == NULL) { remCurIdGet(usr, NULL); fusr = usr; } sprintf(l, "MAIL FROM: %s\n", fusr); if ((s=smtpOpen(hname, pnum)) < 0) return smtpClose(NULL, s, "'smtpOpen' failed", ERROR); if (smtpSendCmdSync(s, NULL, l, l, (-1)) != OK) return ERROR; for (i=0, tnum=0; tusrs[i] != NULL; i++) { if (smtpAddRecepient(s,tusrs[i], l, &rcode) != OK) return smtpClose(NULL, s, "add recepient", ERROR); /* this is a special failure code - it suggests someone else to try - its * suggested path is enclosed in '<...>' */ if ((rcode == SMTP_E_USR_TRY) && ((stp=strchr(l, '<')) != NULL)) { stp++; if ((enp=strchr(stp, '>')) != NULL) { *enp = '\0'; strncpy(usr, stp, (sizeof usr)); usr[(sizeof usr)-1] = '\0'; /* try the suggested user path */ if (smtpAddRecepient(s, usr, l, &rcode) != OK) return smtpClose(NULL, s, "add recepient", ERROR); } } /* if we get a bad response we go to next recepient and don't count it */ if ((rcode == SMTP_E_ACTION_OK) || (rcode == SMTP_E_USR_NLOCAL)) tnum++; } /* make sure at least ONE recepient */ if (tnum == 0) return smtpClose(NULL, s, "no recepients", ERROR); if (smtpSendCmdSync(s, NULL, data_cmd, l, SMTP_E_START_INP) != OK) return ERROR; return s; } /*---------------------------------------------------------------------------*/ STATUS smtpDataSockClose (int s, int add_crlf) { char l[MAX_CMD+1]; static const char quit_cmd[]="QUIT\n"; if (add_crlf) sprintf(l, "\r\n.\r\n"); else sprintf(l, ".\r\n"); if (smtpSendCmdSync(s, NULL, l, l, (-1)) != OK) return smtpClose(NULL, s, "send '.'", ERROR); if (smtpSendCmdSync(s, NULL, quit_cmd, l, SMTP_E_DOMAIN_CLS) != OK) return smtpClose(NULL, s, "send 'QUIT' cmd", ERROR); return smtpClose(NULL, s, "OK", OK); } /*---------------------------------------------------------------------------*/ static STATUS smtpFcopy (int s, const char fname[], int *add_crlf) { FILE *fp = (fname != NULL) ? fopen(fname, "r") : stdin; int is_stdin = (fname != NULL) ? 0 : 1; char l[MAX_LL+1]; if (fp == NULL) return smtpClose(fp, s, "open file", ERROR); while (fgets(l, MAX_LL, fp) == l) { if (write(s, l, strlen(l)) <= 0) return smtpClose(fp, s, "Line send", ERROR); if (l[strlen(l)-1] == '\n') *add_crlf = 0; else *add_crlf = 1; } if (!is_stdin) fclose(fp); return OK; } /*---------------------------------------------------------------------------*/ static STATUS smtpTcopy (int s, const char txt[], int *add_crlf) { if (write(s, txt, strlen(txt)) <= 0) return smtpClose(NULL, s, "Text send", ERROR); if (txt[strlen(txt)-1] == '\n') *add_crlf = 0; else *add_crlf = 1; return OK; } /*---------------------------------------------------------------------------*/ typedef enum enum_smtp_arg_case { SMTP_FILE_CASE, SMTP_TEXT_CASE, SMTP_BAD_CASE } SMTP_ARG_CASE; static STATUS smtpSend (const char host[], const char fromUsr[], const char *tusrs[], int port, SMTP_ARG_CASE acase, void *arg) { int s=(-1), add_crlf=1; if ((s=smtpDataSockOpen(host, fromUsr, tusrs, port)) == ERROR) return smtpClose(NULL, s, "open data sock", ERROR); switch(acase) { case SMTP_FILE_CASE : if (smtpFcopy(s, (const char *) arg, &add_crlf) != OK) return ERROR; break; case SMTP_TEXT_CASE : if (smtpTcopy(s, (const char *) arg, &add_crlf) != OK) return ERROR; break; default : return smtpClose(NULL, s, "Unknown SMTP case", ERROR); } return smtpDataSockClose(s, add_crlf); } /*---------------------------------------------------------------------------*/ S STATUS smtpFsend (const char host[], const char fromUsr[], const char *tusrs[], int port, const char fname[]) { return smtpSend(host, fromUsr, tusrs, port, SMTP_FILE_CASE, (void *) fname); } /*---------------------------------------------------------------------------*/ STATUS smtpSendText (const char host[], /* name/address - default=bootHost*/ const char fromUsr[], /* user - default=remCurIdGet() */ const char toUsr[], int port, /* port - default=IPPORT_SMTP */ const char txt[]) { const char *tusrs[2]; tusrs[0] = toUsr; tusrs[1] = NULL; return smtpSend(host, fromUsr, tusrs, port, SMTP_TEXT_CASE, (void *) txt); } /*---------------------------------------------------------------------------*/ STATUS smtpSendFile (const char host[], /* name/address - default=bootHost*/ const char fromUsr[], /* user - default=remCurIdGet() */ const char toUsr[], int port, /* port - default=IPPORT_SMTP */ const char fname[]) /* default=stdin */ { const char *tusrs[2]; tusrs[0] = toUsr; tusrs[1] = NULL; return smtpSend(host, fromUsr, tusrs, port, SMTP_FILE_CASE, (void *) fname); } /*---------------------------------------------------------------------------*/ /* uses the defaults for "smtpSendFile" */ STATUS smtpSendFQuick (const char toUsr[], const char fname[]) { return smtpSendFile(NULL, NULL, toUsr, 0, fname); } /* uses the defaults for "smtpSendText" */ STATUS smtpSendTQuick (const char toUsr[], const char txt[]) { return smtpSendText(NULL, NULL, toUsr, 0, txt); } From AS_Bridge@msmail3.hac.com Sun Sep 29 03:35:22 1996 From: "AS Bridge" Date: Sun Sep 29 03:35:24 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 2:14 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 3:34 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 04:06:28 1996 From: "AS Bridge" Date: Sun Sep 29 04:06:30 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: SMTP client Sent: Sun, Sep 29, 1996 3:02 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 4:05 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 05:38:30 1996 From: "AS Bridge" Date: Sun Sep 29 05:38:32 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 4:28 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 5:37 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 06:09:31 1996 From: "AS Bridge" Date: Sun Sep 29 06:09:33 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 4:58 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 6:08 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 08:08:36 1996 From: "AS Bridge" Date: Sun Sep 29 08:08:39 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 7:05 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 8:07 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 08:08:36 1996 From: "AS Bridge" Date: Sun Sep 29 08:08:39 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 6:36 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 8:07 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 10:11:41 1996 From: "AS Bridge" Date: Sun Sep 29 10:11:43 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 9:05 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 10:10 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 10:42:40 1996 From: "AS Bridge" Date: Sun Sep 29 10:42:42 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 9:22 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 10:41 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 12:29:45 1996 From: "AS Bridge" Date: Sun Sep 29 12:29:47 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 11:21 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 12:28 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 13:02:44 1996 From: "AS Bridge" Date: Sun Sep 29 13:02:47 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 11:36 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 1:01 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 15:06:49 1996 From: "AS Bridge" Date: Sun Sep 29 15:06:52 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 1:34 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 3:06 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 15:06:49 1996 From: "AS Bridge" Date: Sun Sep 29 15:06:52 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 2:03 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 3:06 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 17:10:55 1996 From: "AS Bridge" Date: Sun Sep 29 17:10:57 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 4:06 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 5:09 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 17:10:55 1996 From: "AS Bridge" Date: Sun Sep 29 17:10:58 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 4:06 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 5:09 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 19:31:01 1996 From: "AS Bridge" Date: Sun Sep 29 19:31:04 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 6:15 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 7:29 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 21:38:05 1996 From: "AS Bridge" Date: Sun Sep 29 21:38:08 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 8:23 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 9:37 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 21:38:05 1996 From: "AS Bridge" Date: Sun Sep 29 21:38:08 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 8:23 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 9:37 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From cwcssk@leonis.nus.sg Sun Sep 29 21:50:25 1996 From: Suresh Santhana Krishnan Date: Sun Sep 29 21:50:27 PDT 1996 Subject: unsubscribe unsubscribe -- __^^^^^__ / @ @ \ =====================ooO==(_)=Ooo=========================== Suresh Santhana Krishnan Tel: 065-7715151 (O) Senior Research Engineer Fax: 065-7795441 (O) Centre For Wireless Communication Email: cwcssk@leonis.nus.sg (O) National University Of Singapore sureener@pacific.net.sg (H) Engineering Blk 2,#05-02, 10, Kent Ridge Crescent Singapore - 119260 ooO =======================( )-Ooo============================ \ ( ( ) \_) ) / (_/ From AS_Bridge@msmail3.hac.com Sun Sep 29 23:58:12 1996 From: "AS Bridge" Date: Sun Sep 29 23:58:16 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 10:32 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 11:57 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 23:58:12 1996 From: "AS Bridge" Date: Sun Sep 29 23:58:16 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: unsubscribe Sent: Sun, Sep 29, 1996 10:47 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 11:57 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Sun Sep 29 23:58:12 1996 From: "AS Bridge" Date: Sun Sep 29 23:58:17 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Sun, Sep 29, 1996 10:34 PM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Sun, Sep 29, 1996 11:57 PM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 02:07:16 1996 From: "AS Bridge" Date: Mon Sep 30 02:07:20 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 1:02 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 2:06 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 02:07:17 1996 From: "AS Bridge" Date: Mon Sep 30 02:07:23 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 1:01 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 2:06 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 02:07:18 1996 From: "AS Bridge" Date: Mon Sep 30 02:07:24 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 1:03 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 2:06 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From daemon@csg.lbl.gov Mon Sep 30 04:01:32 1996 From: daemon@csg.lbl.gov Date: Mon Sep 30 04:01:34 PDT 1996 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Sep 30 04:01:29 PDT 1996 Subject: Password protected patches from WindRiver Subject: Help wanted on FLASH file systems ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Password protected patches from WindRiver Date: Thu, 19 Sep 1996 14:20:37 GMT From: rlykrha@rly.abb.se (Krister Hagman) Organization: ABB Network Partner AB Message-ID: <51rlk6$14uu@sdaw04.seinf.abb.se> Reply-To: rlykrha@rly.abb.se Hi, I've received a patch (SPR6522) from http://www.wrs.com (WindRiver Systems). My problem now is that the patch is a password-protected-self-extracting-archive! And I have no idea of what the password is. Is there anyone out here who knows how the passwords are constructed for their patches? Send me an e-mail with the answer. Best regards Krister Hagman --------------------------- Newsgroups: comp.os.vxworks Subject: Help wanted on FLASH file systems Date: Wed, 18 Sep 1996 12:09:01 -0700 From: Lasse Kinnunen Organization: LOA AB Message-ID: <324048CD.7250@rly.abb.se> Is there someone that knows where to find FLASH file systems, be they true FFS or not, that are also fault-tolerant? The FLASH types in question would be both small and large sector (but not at the same time). --------------------------- End of New-News digest ********************** From AS_Bridge@msmail3.hac.com Mon Sep 30 04:31:04 1996 From: "AS Bridge" Date: Mon Sep 30 04:31:06 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 3:26 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 4:30 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 04:38:55 1996 From: "AS Bridge" Date: Mon Sep 30 04:38:58 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 3:36 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 4:37 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 04:38:55 1996 From: "AS Bridge" Date: Mon Sep 30 04:38:58 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 3:34 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 4:37 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From cpires@br.homeshopping.com.br Mon Sep 30 05:43:01 1996 From: Carlos Pires Date: Mon Sep 30 05:43:03 PDT 1996 Subject: Unsubscribe Please, unsubscribe me. Thanks. From usd@mee.hwm.com Mon Sep 30 06:09:06 1996 From: Uwe Scheffold Date: Mon Sep 30 06:09:09 PDT 1996 Subject: munch process for C++ Hi vxWorker's, does anybody out there know what exactly the munching process does? Uwe From ahw@clalf1.mps.ohio-state.edu Mon Sep 30 06:26:32 1996 From: Andreas Wolf Date: Mon Sep 30 06:26:35 PDT 1996 Subject: Re: Undeliverable Mail HELP!!!! I got over 20 mails of this kind: >Submitted-by AS_Bridge@msmail3.hac.com Mon Sep 30 04:38:55 1996 >Submitted-by: "AS Bridge" > >Unknown Microsoft mail form. Approximate representation follows. > >Message: Undeliverable Mail >Sent: Mon, Sep 30, 1996 3:34 AM >To: Carroll, John C; Claudio, Michael A >On Server: 631 Segment5 >Date: Mon, Sep 30, 1996 4:37 AM >Reason: Could not be delivered because the destination Microsoft Mail server >could not be found. Can the the exploder manager end this soon??? Andreas ==:-( From AS_Bridge@msmail3.hac.com Mon Sep 30 06:28:55 1996 From: "AS Bridge" Date: Mon Sep 30 06:28:57 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: comp.os.vxworks newsdigest Sent: Mon, Sep 30, 1996 5:23 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 6:24 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 07:01:23 1996 From: "AS Bridge" Date: Mon Sep 30 07:01:26 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 5:49 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 6:58 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 07:01:23 1996 From: "AS Bridge" Date: Mon Sep 30 07:01:27 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 5:53 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 6:58 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From AS_Bridge@msmail3.hac.com Mon Sep 30 07:10:48 1996 From: "AS Bridge" Date: Mon Sep 30 07:10:50 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Undeliverable Mail Sent: Mon, Sep 30, 1996 6:06 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 7:09 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From jhillman@wrs.com Mon Sep 30 07:13:55 1996 From: Jon Hillman Date: Mon Sep 30 07:13:57 PDT 1996 Subject: Re: munch process for C++ At 06:09 AM 9/30/96 PDT, Uwe wrote: >does anybody out there know what exactly the munching process does? Munching is the process of finding global/static classes and adding the constructor and destructor functions to a "to-do" list (my terminology). The munching process is normally handling covertly in the compilation process for OS's such as Unix and DOS, since Unix and DOS have explicitly starting/ending points for their processes. For Unix, calling constructors for global objects is typically done immediately as a result of the exec system call in Unix and destructors for global objects are called immediately as a result of the exit() system call. With vxWorks (and other embedded RTOS's), this "starting/ending point" of the application is rather fuzzy, so calling of global/static constructors must initiated explicitly by the application code (with vxWorks call is made to cplusCtors/cplusDtors), which executes those functions in the "to-do" list. The constructors, by the way, are called in the order which the are instantiated, and destructors the opposite. Non-global/static objects are otherwise within a SCOPE, which upon entrance to the scope of the object (function, statement block, new operator) the constructors are called, and upon exiting the scope of the object, the destructor is called. I hope this helps. Others may more succinctly explain this process, and after reading this, I hope they do. :-\ _________________________ | Jon Hillman, FAE | | Wind River Systems | /) (407) 273-4646 (\ / ) Email:jhillman@wrs.com ( \ _( (|_________________________) ) /> (((\ \) /,) / ) / //))/ (\\\\ \_/ / \ \_/ ///// \ / \ / \ _/ \_ / ----/ /----------------------\ \---- / / \ \ From AS_Bridge@msmail3.hac.com Mon Sep 30 08:03:00 1996 From: "AS Bridge" Date: Mon Sep 30 08:03:02 PDT 1996 Subject: Undeliverable Mail Unknown Microsoft mail form. Approximate representation follows. Message: Unsubscribe Sent: Mon, Sep 30, 1996 6:57 AM To: Carroll, John C; Claudio, Michael A On Server: 631 Segment5 Date: Mon, Sep 30, 1996 8:01 AM Reason: Could not be delivered because the destination Microsoft Mail server could not be found. From jjoly@glenayre.com Mon Sep 30 09:53:51 1996 From: jjoly@glenayre.com (J.P. Joly [4629]) Date: Mon Sep 30 09:53:53 PDT 1996 Subject: unsubscribe unsubscribe