Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sat Jul 1 04:00:25 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 1 04:00:18 PDT 1995 Subject: Determining transmit host's IP address in UDP Subject: Re: Determining transmit host's IP address in UDP ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Determining transmit host's IP address in UDP Date: 30 Jun 1995 15:03:08 GMT From: Shash Chatterjee Organization: Lockheed Martin Tactical Aircraft Systems Message-ID: <3t13nc$ao9@cliffy.lfwc.lockheed.com> Once I receive a mssage using the read() call on a UDP socket, is there a way to find out what IP address sent the message? Thanks, Shash - -- +-----------------------------------------------------------------+ + ChatterjeeS@lfwc.lockheed.com + + Sasvata Chatterjee | Electronic Systems Design & Integration + + "SHASH-WATA" | Lockheed Martin Tactical Aircraft Systems+ + (817) 763-1495 | Lockheed Boulevard, MZ2273 + + (817) 763-1495 (FAX) | Fort Worth, TX 76108 + +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Determining transmit host's IP address in UDP Date: Fri, 30 Jun 95 16:41:32 GMT From: siposj@pcnet1.ascs.aro.allied.com (Jon C. Sipos) Organization: AlliedSignal Aerospace Communications Systems Message-ID: <3t19gs$1cu@asbe05.phx1.aro.allied.com> References: <3t13nc$ao9@cliffy.lfwc.lockheed.com> In article <3t13nc$ao9@cliffy.lfwc.lockheed.com>, Shash Chatterjee wrote: >Once I receive a mssage using the read() call on a UDP socket, is there >a way to find out what IP address sent the message? > >Thanks, >Shash Try recvfrom() instead of read(). - ---- Jon C. Sipos siposj@pcnet1.ascs.aro.allied.com AlliedSignal Aerospace Government Electronics Systems 1300 East Joppa Rd. Baltimore, MD, USA 21286-5999 VOICE (410) 583-4112 FAX (410) 583-5901 - ---- Jon C. Sipos siposj@pcnet1.ascs.aro.allied.com AlliedSignal Aerospace Government Electronics Systems 1300 East Joppa Rd. Baltimore, MD, USA 21286-5999 VOICE (410) 583-4112 FAX (410) 583-5901 --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sat Jul 1 04:00:25 1995 From: daemon@csg.lbl.gov Date: Sat Jul 1 04:00:28 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 1 04:00:18 PDT 1995 Subject: Determining transmit host's IP address in UDP Subject: Re: Determining transmit host's IP address in UDP ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Determining transmit host's IP address in UDP Date: 30 Jun 1995 15:03:08 GMT From: Shash Chatterjee Organization: Lockheed Martin Tactical Aircraft Systems Message-ID: <3t13nc$ao9@cliffy.lfwc.lockheed.com> Once I receive a mssage using the read() call on a UDP socket, is there a way to find out what IP address sent the message? Thanks, Shash - -- +-----------------------------------------------------------------+ + ChatterjeeS@lfwc.lockheed.com + + Sasvata Chatterjee | Electronic Systems Design & Integration + + "SHASH-WATA" | Lockheed Martin Tactical Aircraft Systems+ + (817) 763-1495 | Lockheed Boulevard, MZ2273 + + (817) 763-1495 (FAX) | Fort Worth, TX 76108 + +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Determining transmit host's IP address in UDP Date: Fri, 30 Jun 95 16:41:32 GMT From: siposj@pcnet1.ascs.aro.allied.com (Jon C. Sipos) Organization: AlliedSignal Aerospace Communications Systems Message-ID: <3t19gs$1cu@asbe05.phx1.aro.allied.com> References: <3t13nc$ao9@cliffy.lfwc.lockheed.com> In article <3t13nc$ao9@cliffy.lfwc.lockheed.com>, Shash Chatterjee wrote: >Once I receive a mssage using the read() call on a UDP socket, is there >a way to find out what IP address sent the message? > >Thanks, >Shash Try recvfrom() instead of read(). - ---- Jon C. Sipos siposj@pcnet1.ascs.aro.allied.com AlliedSignal Aerospace Government Electronics Systems 1300 East Joppa Rd. Baltimore, MD, USA 21286-5999 VOICE (410) 583-4112 FAX (410) 583-5901 - ---- Jon C. Sipos siposj@pcnet1.ascs.aro.allied.com AlliedSignal Aerospace Government Electronics Systems 1300 East Joppa Rd. Baltimore, MD, USA 21286-5999 VOICE (410) 583-4112 FAX (410) 583-5901 --------------------------- End of New-News digest ********************** Subject: Re: Copying large (net) files Submitted-by leonid@rst.co.il Sun Jul 2 00:09:59 1995 Submitted-by: leonid@rst.co.il (Leonid Rosenboim) > Date: 28 Jun 1995 02:04:52 GMT > From: friedl@cher.heurikon.com (Ted Friedl) > I would like to copy a 1 gigabyte file from my Unix host (via Ethernet) > to the target's file system ("/sd0/" in this case). The netDrv does do everything in RAM and copies over the net only when open() and close() are called. This is well known and documented. You could of course enable NFS and then it will work, but much easier is to enable the FTP server on VxWorks, which is rather useful when you have local disks on the target, and do the entire transfer using an FTP client on your host. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il From leonid@rst.co.il Sun Jul 2 00:09:59 1995 From: leonid@rst.co.il (Leonid Rosenboim) Date: Sun Jul 2 00:10:02 PDT 1995 Subject: Re: Copying large (net) files > Date: 28 Jun 1995 02:04:52 GMT > From: friedl@cher.heurikon.com (Ted Friedl) > I would like to copy a 1 gigabyte file from my Unix host (via Ethernet) > to the target's file system ("/sd0/" in this case). The netDrv does do everything in RAM and copies over the net only when open() and close() are called. This is well known and documented. You could of course enable NFS and then it will work, but much easier is to enable the FTP server on VxWorks, which is rather useful when you have local disks on the target, and do the entire transfer using an FTP client on your host. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sun Jul 2 04:00:22 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sun Jul 2 04:00:17 PDT 1995 Subject: tape drive s/w Subject: Re: Copying large (net) files Subject: Flash file system vendors Subject: Using Exabyte Model 8505 Subject: Re: Need driver for Intel 28F020 Flash Memory ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: tape drive s/w Date: 27 Jun 95 20:25:17 GMT From: judd@melair.lockheed.on.ca (Rob Judd) Organization: Lockheed Canada Inc., Kanata, Ontario, Canada Message-ID: <7632@melair.lockheed.on.ca> Hello, we have an application that requires that we write to a tape drive, and I was wondering if someone might have some idea where I might find this software, if there is any. :-). the particulars; a HP35480A 4mm, 2GB tape, Motorola MV162 controller, vxWorks 5.1.1 thanks rob - -- Robert Judd voice : (613) 599 3270 x513 Lockheed Martin Canada Inc. fax : (613) 599 3282 3001 Solandt Road Kanata, Ont, Can K2K 2M8 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Copying large (net) files Date: 28 Jun 1995 06:26:18 GMT From: friedl@cher.heurikon.com (Ted Friedl) Organization: Heurikon Corporation Message-ID: <3sqsma$qdo@zebu.heurikon.com> References: <3sqdc4$om4@zebu.heurikon.com> In article <3sqdc4$om4@zebu.heurikon.com>, Ted Friedl (ME!!!) wrote: >Any simple solutions out there?? There is at least one! My original posting was up for only an hour and I got a bunch of email responses. I thought I would follow up for anyone who is curious. This is what worked best for me (from Fritz Stauffer): : How about running the ftp daemon on vxWorks, : run ftp on the unix host to vxworks, and put the : file? This worked great! Thanks to all who responded. Ted Friedl Senior Software Engineer Heurikon Corporation friedl@heurikon.com --------------------------- Newsgroups: comp.os.vxworks Subject: Flash file system vendors Date: Wed, 28 Jun 1995 12:19:51 GMT From: skv@pcdoh.pgh.wec.com (Suresh K. Vishnubhatla) Organization: Westinghouse Electric Co. Message-ID: <1995Jun28.121951.29224@tron.bwi.wec.com> Reply-To: skv@pcdoh.pgh.wec.com Sender: usenet@tron.bwi.wec.com (Usenet_news poster) I am looking for Flash file system vendors. Any info. will be greatly appreciated. Thanks Suresh --------------------------- Newsgroups: comp.os.vxworks Subject: Using Exabyte Model 8505 Date: Fri, 30 Jun 1995 17:27:31 GMT From: William Oszust Organization: Westinghouse Electric Corporation Message-ID: Reply-To: William Oszust Sender: usenet@tron.bwi.wec.com (Usenet_news poster) We are using vxWorks 5.1.1 and are trying to connect an Exabyte Tape Drive Model 8505 through a scsi adapter. Vxworks does not support the Exabyte Tape drive. Has anyone ever used this drive with vxWorks or know how I would go about doing this? Including writting/reading and programming for real time applications. Any help would be greatly appreciated! Thanks, Bill ====================================================================== William Oszust e-mail: oszust@norden.com Westinghouse Norden Systems Telephone: 1-516-845-2525 75 Maxess Road Fax: 1-516-845-2907 Melville, New York 11747 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Need driver for Intel 28F020 Flash Memory Date: 1 Jul 1995 13:50:53 GMT From: Paul Anderson Organization: Compware Corporation Message-ID: <3t3jrt$1j5@infoman.net99.net> References: <2FF4337B@gicpo.gic.gi.com> "Moore, Rich (HT-MS)" wrote: > >Does anyone know where I can get a driver for the Intel 28F020 >2048K (256 x 8) Flash Memory. I have a MVME162 with 4 of these >chips on-board (1Mbyte). I want to run a Block Device Driver over >this flash so that I can use the dosFs (DOS File System) that comes >with VxWorks. > >Thanks in advance! >Rich Moore >General Instrument Corp. >Hatboro, PA Rich Compware offers a source package for the GreenSpring IP-Flash[2M/8M] Industry Pack under VxWorks. That device is the 28F008. The FlashLib package handles programming & erase, which supports copying a file system to the Flash. I'll look into 28F020 compatibility. Compware is a VAR/SysIntegrator, we sell SUN, MOT, & various hardware, device driver integration is our proprietary product with about a dozen VxWorks drivers, mostly IP's. Please email your fax no. and I'll send a user's guide asap. Paul Anderson (paul@compware.com) - -------------------------------------------- Compware Corporation 9735 N. 90th Place, Suite 200 Tel.: 602-998-8650, 800-788-5855 Scottsdale, AZ 85258 Facs.: 602-661-0019 - -------------------------------------------- --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sun Jul 2 04:00:22 1995 From: daemon@csg.lbl.gov Date: Sun Jul 2 04:00:25 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 2 04:00:17 PDT 1995 Subject: tape drive s/w Subject: Re: Copying large (net) files Subject: Flash file system vendors Subject: Using Exabyte Model 8505 Subject: Re: Need driver for Intel 28F020 Flash Memory ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: tape drive s/w Date: 27 Jun 95 20:25:17 GMT From: judd@melair.lockheed.on.ca (Rob Judd) Organization: Lockheed Canada Inc., Kanata, Ontario, Canada Message-ID: <7632@melair.lockheed.on.ca> Hello, we have an application that requires that we write to a tape drive, and I was wondering if someone might have some idea where I might find this software, if there is any. :-). the particulars; a HP35480A 4mm, 2GB tape, Motorola MV162 controller, vxWorks 5.1.1 thanks rob - -- Robert Judd voice : (613) 599 3270 x513 Lockheed Martin Canada Inc. fax : (613) 599 3282 3001 Solandt Road Kanata, Ont, Can K2K 2M8 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Copying large (net) files Date: 28 Jun 1995 06:26:18 GMT From: friedl@cher.heurikon.com (Ted Friedl) Organization: Heurikon Corporation Message-ID: <3sqsma$qdo@zebu.heurikon.com> References: <3sqdc4$om4@zebu.heurikon.com> In article <3sqdc4$om4@zebu.heurikon.com>, Ted Friedl (ME!!!) wrote: >Any simple solutions out there?? There is at least one! My original posting was up for only an hour and I got a bunch of email responses. I thought I would follow up for anyone who is curious. This is what worked best for me (from Fritz Stauffer): : How about running the ftp daemon on vxWorks, : run ftp on the unix host to vxworks, and put the : file? This worked great! Thanks to all who responded. Ted Friedl Senior Software Engineer Heurikon Corporation friedl@heurikon.com --------------------------- Newsgroups: comp.os.vxworks Subject: Flash file system vendors Date: Wed, 28 Jun 1995 12:19:51 GMT From: skv@pcdoh.pgh.wec.com (Suresh K. Vishnubhatla) Organization: Westinghouse Electric Co. Message-ID: <1995Jun28.121951.29224@tron.bwi.wec.com> Reply-To: skv@pcdoh.pgh.wec.com Sender: usenet@tron.bwi.wec.com (Usenet_news poster) I am looking for Flash file system vendors. Any info. will be greatly appreciated. Thanks Suresh --------------------------- Newsgroups: comp.os.vxworks Subject: Using Exabyte Model 8505 Date: Fri, 30 Jun 1995 17:27:31 GMT From: William Oszust Organization: Westinghouse Electric Corporation Message-ID: Reply-To: William Oszust Sender: usenet@tron.bwi.wec.com (Usenet_news poster) We are using vxWorks 5.1.1 and are trying to connect an Exabyte Tape Drive Model 8505 through a scsi adapter. Vxworks does not support the Exabyte Tape drive. Has anyone ever used this drive with vxWorks or know how I would go about doing this? Including writting/reading and programming for real time applications. Any help would be greatly appreciated! Thanks, Bill ====================================================================== William Oszust e-mail: oszust@norden.com Westinghouse Norden Systems Telephone: 1-516-845-2525 75 Maxess Road Fax: 1-516-845-2907 Melville, New York 11747 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Need driver for Intel 28F020 Flash Memory Date: 1 Jul 1995 13:50:53 GMT From: Paul Anderson Organization: Compware Corporation Message-ID: <3t3jrt$1j5@infoman.net99.net> References: <2FF4337B@gicpo.gic.gi.com> "Moore, Rich (HT-MS)" wrote: > >Does anyone know where I can get a driver for the Intel 28F020 >2048K (256 x 8) Flash Memory. I have a MVME162 with 4 of these >chips on-board (1Mbyte). I want to run a Block Device Driver over >this flash so that I can use the dosFs (DOS File System) that comes >with VxWorks. > >Thanks in advance! >Rich Moore >General Instrument Corp. >Hatboro, PA Rich Compware offers a source package for the GreenSpring IP-Flash[2M/8M] Industry Pack under VxWorks. That device is the 28F008. The FlashLib package handles programming & erase, which supports copying a file system to the Flash. I'll look into 28F020 compatibility. Compware is a VAR/SysIntegrator, we sell SUN, MOT, & various hardware, device driver integration is our proprietary product with about a dozen VxWorks drivers, mostly IP's. Please email your fax no. and I'll send a user's guide asap. Paul Anderson (paul@compware.com) - -------------------------------------------- Compware Corporation 9735 N. 90th Place, Suite 200 Tel.: 602-998-8650, 800-788-5855 Scottsdale, AZ 85258 Facs.: 602-661-0019 - -------------------------------------------- --------------------------- End of New-News digest ********************** Subject: CDROM FIlesystem for VxWorks - Beta Testers needed Submitted-by leonid@rst.co.il Sun Jul 2 06:04:37 1995 Submitted-by: leonid@rst.co.il (Leonid Rosenboim) We are about to release a fully functional CD-ROM File System package for VxWorks, and would like to have a couple of additional users to test it with real-life applications. The module is a fully VxWorks-compatible file system that supports read-only drives. It has been tested with ISO-9660 disks with and without RockRidge extentions. POSIX readdir() and stat() functions are supported. Selected testers will receive a binary (object) file, and manual pages at no charge, and will be expected to report their results (good or bad) to us reasonably soon. The source code will be available subsequently for sale. If you are interested to participate please fill in the following form and return it to us: ------------------------------------------------------------------------ Organization / Company / Project: Mail Address: Technical Contact: Phone: Fax: E-Mail: Administrative Contact: Phone: Fax: E-Mail: VxWorks Version: Target CPU Architecture: Model: Peripherals Configuration: (including CDROm drive vendor and model) Application Description: Special Considerations: ----------------------------------------------------------------------- Feel free to contact us for further information. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il From leonid@rst.co.il Sun Jul 2 06:04:37 1995 From: leonid@rst.co.il (Leonid Rosenboim) Date: Sun Jul 2 06:04:41 PDT 1995 Subject: CDROM FIlesystem for VxWorks - Beta Testers needed We are about to release a fully functional CD-ROM File System package for VxWorks, and would like to have a couple of additional users to test it with real-life applications. The module is a fully VxWorks-compatible file system that supports read-only drives. It has been tested with ISO-9660 disks with and without RockRidge extentions. POSIX readdir() and stat() functions are supported. Selected testers will receive a binary (object) file, and manual pages at no charge, and will be expected to report their results (good or bad) to us reasonably soon. The source code will be available subsequently for sale. If you are interested to participate please fill in the following form and return it to us: ------------------------------------------------------------------------ Organization / Company / Project: Mail Address: Technical Contact: Phone: Fax: E-Mail: Administrative Contact: Phone: Fax: E-Mail: VxWorks Version: Target CPU Architecture: Model: Peripherals Configuration: (including CDROm drive vendor and model) Application Description: Special Considerations: ----------------------------------------------------------------------- Feel free to contact us for further information. ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Mon Jul 3 04:00:52 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Mon Jul 3 04:00:46 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: Mon, 3 Jul 1995 05:21:14 GMT From: Ron Kellam Organization: Hewlett-Packard Co. boundary="-------------------------------66818931989" Message-ID: References: Sender: news@hparc12.aus.hp.com (News Daemon ID) This is a multi-part message in MIME format. - ---------------------------------66818931989 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii guy@boi.hp.com (Guy Burnham) wrote: >I'm questioning whether I'm loosing my marbles! When I ask the shell in >VxWorks 5.1.1 ...... > > -> .7 <= .6 > >It comes back with ...... > >value = 1 > >So I wrote a little C hack .... > >int LTorEQ (float a, float b) > { > return (a <= b); > } > >.... and called it with the following result > >-> LTorEQ 0.7,0.6 >value = 1 = 0x1 > >Does anyone elses system think 0.7 is less than or equal to 0.6?? > Yes Guy, my system behaves the same... Explanations anyone?? Ron. - ---------------------------------66818931989 Content-Transfer-Encoding: 7bit Content-Type: text/plain - ---------------------------------------------------------------------------- Ron Kellam | Phone: +613 272 4016 Hewlett Packard | Fax: +613 898 9257 .-_!\ Australian Telecom Operation | Email: ronk@hpato.aus.hp.com / ATO \ 31-41 Joseph St; Blackburn | HPDesk: KELLAM_RON/HP9061_RG@hpausa1 \_.-._/ Victoria, Australia, 3130 | "My opinions are my own, not HP's" v - ---------------------------------------------------------------------------- "Even if you're on the right track, you still get run over if you don't keep moving" - ---------------------------------------------------------------------------- - ---------------------------------66818931989-- --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Mon Jul 3 04:00:52 1995 From: daemon@csg.lbl.gov Date: Mon Jul 3 04:00:55 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 3 04:00:46 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: Mon, 3 Jul 1995 05:21:14 GMT From: Ron Kellam Organization: Hewlett-Packard Co. boundary="-------------------------------66818931989" Message-ID: References: Sender: news@hparc12.aus.hp.com (News Daemon ID) This is a multi-part message in MIME format. - ---------------------------------66818931989 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii guy@boi.hp.com (Guy Burnham) wrote: >I'm questioning whether I'm loosing my marbles! When I ask the shell in >VxWorks 5.1.1 ...... > > -> .7 <= .6 > >It comes back with ...... > >value = 1 > >So I wrote a little C hack .... > >int LTorEQ (float a, float b) > { > return (a <= b); > } > >.... and called it with the following result > >-> LTorEQ 0.7,0.6 >value = 1 = 0x1 > >Does anyone elses system think 0.7 is less than or equal to 0.6?? > Yes Guy, my system behaves the same... Explanations anyone?? Ron. - ---------------------------------66818931989 Content-Transfer-Encoding: 7bit Content-Type: text/plain - ---------------------------------------------------------------------------- Ron Kellam | Phone: +613 272 4016 Hewlett Packard | Fax: +613 898 9257 .-_!\ Australian Telecom Operation | Email: ronk@hpato.aus.hp.com / ATO \ 31-41 Joseph St; Blackburn | HPDesk: KELLAM_RON/HP9061_RG@hpausa1 \_.-._/ Victoria, Australia, 3130 | "My opinions are my own, not HP's" v - ---------------------------------------------------------------------------- "Even if you're on the right track, you still get run over if you don't keep moving" - ---------------------------------------------------------------------------- - ---------------------------------66818931989-- --------------------------- End of New-News digest ********************** Subject: Re: Why is 0.5 <= 0.4 ???? Submitted-by phil@naic.edu Mon Jul 3 06:30:13 1995 Submitted-by: phil@naic.edu (Phil Perillat) > > guy@boi.hp.com (Guy Burnham) wrote: > >I'm questioning whether I'm loosing my marbles! When I ask the shell in > >VxWorks 5.1.1 ...... > > > > -> .7 <= .6 > > > >It comes back with ...... > > > >value = 1 > > > >So I wrote a little C hack .... > > > >int LTorEQ (float a, float b) > > { > > return (a <= b); > > } > > > >.... and called it with the following result > > > >-> LTorEQ 0.7,0.6 > >value = 1 = 0x1 > > > >Does anyone elses system think 0.7 is less than or equal to 0.6?? > > > > Yes Guy, my system behaves the same... > > Explanations anyone?? > > Ron. For your test program, the shell is passing the variables as doubles. Changing your program to > >int LTorEQ (double a, double b) > > { > > return (a <= b); > > } will work correctly. For the shell case .7 <= .6 i'll bet that the numbers are stored as doubles and then the shell is doing an integer compare on the first word of each double... phil perillat arecibo observatory phil@naic.edu From phil@naic.edu Mon Jul 3 06:30:13 1995 From: phil@naic.edu (Phil Perillat) Date: Mon Jul 3 06:30:16 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? > > guy@boi.hp.com (Guy Burnham) wrote: > >I'm questioning whether I'm loosing my marbles! When I ask the shell in > >VxWorks 5.1.1 ...... > > > > -> .7 <= .6 > > > >It comes back with ...... > > > >value = 1 > > > >So I wrote a little C hack .... > > > >int LTorEQ (float a, float b) > > { > > return (a <= b); > > } > > > >.... and called it with the following result > > > >-> LTorEQ 0.7,0.6 > >value = 1 = 0x1 > > > >Does anyone elses system think 0.7 is less than or equal to 0.6?? > > > > Yes Guy, my system behaves the same... > > Explanations anyone?? > > Ron. For your test program, the shell is passing the variables as doubles. Changing your program to > >int LTorEQ (double a, double b) > > { > > return (a <= b); > > } will work correctly. For the shell case .7 <= .6 i'll bet that the numbers are stored as doubles and then the shell is doing an integer compare on the first word of each double... phil perillat arecibo observatory phil@naic.edu Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Tue Jul 4 04:00:35 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Tue Jul 4 04:00:31 PDT 1995 Subject: Re: Ruggedized SBCs Subject: Re: C++-Compiler for VxWorks Subject: Need VME adapter for PCI card ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Ruggedized SBCs Date: Fri, 30 Jun 1995 14:55:51 GMT From: afraser@hookup.net (Andrew Fraser) Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <3t135g$nin@noc.tor.hookup.net> References: <9506291826.AA07300@lbl.gov> Roland Mandler wrote: >Does anyone have a list of ruggedized SBCs (680X0 or SPARC) for which there >are VxWorks BSPs available? (Even a partial list would be appreciated) >Thanks. Hi Roland, How rugged do you want? :) Our cards are available in grades from plain commercial up to full mil, conduction cooled, conformally coated. We have the following 680x0 SBCs, with VxWorks BSPs. No SPARC cards, but the SVME-170 is a PowerPC 603. SVME-153: A basic 68020 SBC, 16 MHz, RAM/Flash/EPROM etc options SVME-155: Similar to '153 but with MIL-STD-1553 interface SVME-160: A complete 68040 SBC, 25 MHz, various memory options SVME-162: 25 or 33 MHz 68040, up to 8 MB SRAM, MAXpack daughtercards SVME-163: 25 or 33 MHz 68040, up to 64 MB DRAM, MAXpack daughtercards Please contact our office below for more info. VITA also has a Web page and is in the process of putting the industry catalog on-line; they're worth visiting at http://www.vita.com/ - -Andy - ------------------------------------------------------------------------ ANDREW FRASER DY 4 Systems Inc email: afraser@dy4.com Senior H/W Designer 21 Fitzgerald Rd phone: (613)596-9922 x251 Product Support Group Nepean, Ontario fax: (613)596-0574 (veni, vidi, fixi) Canada K2H 9J4 marketing: sales@dy4.com - -------------------- CUSTOMER FIRST, QUALITY ALWAYS -------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++-Compiler for VxWorks Date: 22 Jun 1995 19:28:45 -0700 From: kla@leland.Stanford.EDU (Earl Mitchell) Organization: Stanford University, CA 94305, USA Message-ID: <3sd8st$sct@elaine30.Stanford.EDU> References: <3ri7gp$o0d@news-s01.ca.us.ibm.net> <3rojak$7oj@f111.iassf.easams.com.au> <3rq41e$is8@elaine30.stanford.edu> <3s8egq$7fn@f111.iassf.easams.com.au> In article <3s8egq$7fn@f111.iassf.easams.com.au>, Rohan LENARD wrote: >Well the GNU compiler implements ansi and non-ansi features and any non-ansi >feature can be turned off with -ansi -pedantic. If you're compiler doesn't >allow this then I suggest getting a new compiler. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Try compiling your GNU code with ObjectCenter or SparcWorks compiler. Last time I tried neither mode worked. I had to rewrite code to get around this. Strike 1. >The GNU libg++ has the ANSI definition of iostream. If Microsloth doesn't >get yourself a copy of libg++. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Which won't work with Macrohard C++ compiler without MAJOR hacking. For example, both have defined some manipulators which the other doesn't, and some manipulators are implemented differently (i.e. different args). That's just the tip of the iceberg with Macrohard C++ compiler problems. So forget using Macrohard's Visual C++ environment with GNU code without changes. Strike 2. >Templates are actually quite easy to get around once you understand the >models the different compilers use. For most real time applications templates are practically a joke, most people I know avoid using them due to code size explosion problems (i.e. most embedded applications have limited memory). Your excessive use of templates says a lot. ;-) But ignoring this issue here's a simple test. Try compiling the standard template library with the GNU compiler. The standard template library is now part of the ANSI standard. If you can get it to compile without changing code your a better man than me. Strike 3. >Who codes to the ARM ? There's been a draft working paper for a *long* time. >If you're developing in C++ you should be involved in the process of >developing the final standard otherwise you will be surprised when it >is released. Who codes to a "draft"? Especially one that most compilers don't support yet. Final comment: If you stick to the simple stuff you probably won't have to change much code but most C++ programmers don't like to stick to the simple stuff (e.g. the heavy use of templates). If you don't care about ANSI compatibility your ok also. However, once you venture out to use third party tools and class libraries all bets are of and you'll become interested in ANSI compatibility very quickly. And don't forget CORBA. --------------------------- Newsgroups: comp.arch.bus.vmebus,comp.os.vxworks Subject: Need VME adapter for PCI card ? Date: Mon, 26 Jun 1995 17:44:33 GMT From: riendeau@ireq-sat.hydro.qc.ca (Sylvain Riendeau) Organization: Institut de Recherche d'Hydro-Quebec,Varennes,Canada Message-ID: Reply-To: riendeau@ireq-sat.hydro.qc.ca Sender: news@ireq.hydro.qc.ca (Netnews Admin) As strange as it could be... Is there an adapter of some sort available that will adapt a PCI board to a VME B-size chassis? An adapter without CPU would be ideal. That adapter should hopefully: - - map PCI's 32bit memory/registers to VME A32/D32, - - map PCI's 4 interrupts to VME interrupts, - - no PCI or VME bus masterring is needed. Thanks. P.S.: Please reply by e-mail. I will post a summary to the newsgroup if there is interest. - ------------------------------------------------------------------ Sylvain Riendeau Engineer/Ingenieur Tel: (514) 652-8573 Hydro-Quebec (IREQ/LAB-M4) Fax: (514) 652-8570 1800, Montee Ste-Julie Internet: riendeau@ireq-sat.hydro.qc.ca Varennes, Quebec CANADA J3X 1S1 - ------------------------------------------------------------------ --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Tue Jul 4 04:00:35 1995 From: daemon@csg.lbl.gov Date: Tue Jul 4 04:00:38 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 4 04:00:31 PDT 1995 Subject: Re: Ruggedized SBCs Subject: Re: C++-Compiler for VxWorks Subject: Need VME adapter for PCI card ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Ruggedized SBCs Date: Fri, 30 Jun 1995 14:55:51 GMT From: afraser@hookup.net (Andrew Fraser) Organization: HookUp Communication Corporation, Oakville, Ontario, CANADA Message-ID: <3t135g$nin@noc.tor.hookup.net> References: <9506291826.AA07300@lbl.gov> Roland Mandler wrote: >Does anyone have a list of ruggedized SBCs (680X0 or SPARC) for which there >are VxWorks BSPs available? (Even a partial list would be appreciated) >Thanks. Hi Roland, How rugged do you want? :) Our cards are available in grades from plain commercial up to full mil, conduction cooled, conformally coated. We have the following 680x0 SBCs, with VxWorks BSPs. No SPARC cards, but the SVME-170 is a PowerPC 603. SVME-153: A basic 68020 SBC, 16 MHz, RAM/Flash/EPROM etc options SVME-155: Similar to '153 but with MIL-STD-1553 interface SVME-160: A complete 68040 SBC, 25 MHz, various memory options SVME-162: 25 or 33 MHz 68040, up to 8 MB SRAM, MAXpack daughtercards SVME-163: 25 or 33 MHz 68040, up to 64 MB DRAM, MAXpack daughtercards Please contact our office below for more info. VITA also has a Web page and is in the process of putting the industry catalog on-line; they're worth visiting at http://www.vita.com/ - -Andy - ------------------------------------------------------------------------ ANDREW FRASER DY 4 Systems Inc email: afraser@dy4.com Senior H/W Designer 21 Fitzgerald Rd phone: (613)596-9922 x251 Product Support Group Nepean, Ontario fax: (613)596-0574 (veni, vidi, fixi) Canada K2H 9J4 marketing: sales@dy4.com - -------------------- CUSTOMER FIRST, QUALITY ALWAYS -------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: C++-Compiler for VxWorks Date: 22 Jun 1995 19:28:45 -0700 From: kla@leland.Stanford.EDU (Earl Mitchell) Organization: Stanford University, CA 94305, USA Message-ID: <3sd8st$sct@elaine30.Stanford.EDU> References: <3ri7gp$o0d@news-s01.ca.us.ibm.net> <3rojak$7oj@f111.iassf.easams.com.au> <3rq41e$is8@elaine30.stanford.edu> <3s8egq$7fn@f111.iassf.easams.com.au> In article <3s8egq$7fn@f111.iassf.easams.com.au>, Rohan LENARD wrote: >Well the GNU compiler implements ansi and non-ansi features and any non-ansi >feature can be turned off with -ansi -pedantic. If you're compiler doesn't >allow this then I suggest getting a new compiler. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Try compiling your GNU code with ObjectCenter or SparcWorks compiler. Last time I tried neither mode worked. I had to rewrite code to get around this. Strike 1. >The GNU libg++ has the ANSI definition of iostream. If Microsloth doesn't >get yourself a copy of libg++. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Which won't work with Macrohard C++ compiler without MAJOR hacking. For example, both have defined some manipulators which the other doesn't, and some manipulators are implemented differently (i.e. different args). That's just the tip of the iceberg with Macrohard C++ compiler problems. So forget using Macrohard's Visual C++ environment with GNU code without changes. Strike 2. >Templates are actually quite easy to get around once you understand the >models the different compilers use. For most real time applications templates are practically a joke, most people I know avoid using them due to code size explosion problems (i.e. most embedded applications have limited memory). Your excessive use of templates says a lot. ;-) But ignoring this issue here's a simple test. Try compiling the standard template library with the GNU compiler. The standard template library is now part of the ANSI standard. If you can get it to compile without changing code your a better man than me. Strike 3. >Who codes to the ARM ? There's been a draft working paper for a *long* time. >If you're developing in C++ you should be involved in the process of >developing the final standard otherwise you will be surprised when it >is released. Who codes to a "draft"? Especially one that most compilers don't support yet. Final comment: If you stick to the simple stuff you probably won't have to change much code but most C++ programmers don't like to stick to the simple stuff (e.g. the heavy use of templates). If you don't care about ANSI compatibility your ok also. However, once you venture out to use third party tools and class libraries all bets are of and you'll become interested in ANSI compatibility very quickly. And don't forget CORBA. --------------------------- Newsgroups: comp.arch.bus.vmebus,comp.os.vxworks Subject: Need VME adapter for PCI card ? Date: Mon, 26 Jun 1995 17:44:33 GMT From: riendeau@ireq-sat.hydro.qc.ca (Sylvain Riendeau) Organization: Institut de Recherche d'Hydro-Quebec,Varennes,Canada Message-ID: Reply-To: riendeau@ireq-sat.hydro.qc.ca Sender: news@ireq.hydro.qc.ca (Netnews Admin) As strange as it could be... Is there an adapter of some sort available that will adapt a PCI board to a VME B-size chassis? An adapter without CPU would be ideal. That adapter should hopefully: - - map PCI's 32bit memory/registers to VME A32/D32, - - map PCI's 4 interrupts to VME interrupts, - - no PCI or VME bus masterring is needed. Thanks. P.S.: Please reply by e-mail. I will post a summary to the newsgroup if there is interest. - ------------------------------------------------------------------ Sylvain Riendeau Engineer/Ingenieur Tel: (514) 652-8573 Hydro-Quebec (IREQ/LAB-M4) Fax: (514) 652-8570 1800, Montee Ste-Julie Internet: riendeau@ireq-sat.hydro.qc.ca Varennes, Quebec CANADA J3X 1S1 - ------------------------------------------------------------------ --------------------------- End of New-News digest ********************** Subject: Need driver for Intel 28F020 Flash Memory Submitted-by rmoore@gic.gi.com Wed Jul 5 05:53:28 1995 Submitted-by: "Moore, Rich (HT-MS)" Does anyone know where I can get a driver for the Intel 28F020 2048K (256 x 8) Flash Memory. I have a MVME162 with 4 of these chips on-board (1Mbyte). I want to run a Block Device Driver over this flash so that I can use the dosFs (DOS File System) that comes with VxWorks. Thanks in advance! Rich Moore General Instrument Corp. Hatboro, PA From rmoore@gic.gi.com Wed Jul 5 05:53:28 1995 From: "Moore, Rich (HT-MS)" Date: Wed Jul 5 05:53:30 PDT 1995 Subject: Need driver for Intel 28F020 Flash Memory Does anyone know where I can get a driver for the Intel 28F020 2048K (256 x 8) Flash Memory. I have a MVME162 with 4 of these chips on-board (1Mbyte). I want to run a Block Device Driver over this flash so that I can use the dosFs (DOS File System) that comes with VxWorks. Thanks in advance! Rich Moore General Instrument Corp. Hatboro, PA Subject: Open Position Submitted-by brodrigu@polaris.cv.nrao.edu Wed Jul 5 08:18:34 1995 Submitted-by: "Billie Rodriguez, 804 296-0312" Electronics Engineer (Software and Digital Systems) Will assist in the design, construction, and maintenance of software and hardware systems for the Green Bank Telescope. Responsibilities will include coding of real-time control software, hardware/software integration, and maintenance of hardware/software systems. Some hardware design and maintenance of existing systems will be required. At least a Bachelors degree in Electrical or Computer Engineering, with experience in programming real-time systems, digital and analog electronics, writing driver software, and integration of hardware/software systems. Familiarity with modern engineering and programming practices and state-of-the-art digital devices is necessary. The applicant should be an experienced user of oscilloscopes, logic analyzers, and various other hardware/software integration tools. Familiarity with some or all of the following is required: Object Oriented design, C/C++, UNIX, VxWorks, and sockets/RPCs. E-mail resumes to brodrigu@nrao.edu or mail them to: Mr. Tim Weadon NRAO P.O. Box 2 Green Bank, WV 24944-0002 PLEASE NOTE: THIS POSITION HAS RECENTLY BEEN LISTED ON THE VXWORKS EXPLODER. IT IS NOT NECESSARY FOR PREVIOUS APPLICANTS TO RE-APPLY. From brodrigu@polaris.cv.nrao.edu Wed Jul 5 08:18:34 1995 From: "Billie Rodriguez, 804 296-0312" Date: Wed Jul 5 08:18:36 PDT 1995 Subject: Open Position Electronics Engineer (Software and Digital Systems) Will assist in the design, construction, and maintenance of software and hardware systems for the Green Bank Telescope. Responsibilities will include coding of real-time control software, hardware/software integration, and maintenance of hardware/software systems. Some hardware design and maintenance of existing systems will be required. At least a Bachelors degree in Electrical or Computer Engineering, with experience in programming real-time systems, digital and analog electronics, writing driver software, and integration of hardware/software systems. Familiarity with modern engineering and programming practices and state-of-the-art digital devices is necessary. The applicant should be an experienced user of oscilloscopes, logic analyzers, and various other hardware/software integration tools. Familiarity with some or all of the following is required: Object Oriented design, C/C++, UNIX, VxWorks, and sockets/RPCs. E-mail resumes to brodrigu@nrao.edu or mail them to: Mr. Tim Weadon NRAO P.O. Box 2 Green Bank, WV 24944-0002 PLEASE NOTE: THIS POSITION HAS RECENTLY BEEN LISTED ON THE VXWORKS EXPLODER. IT IS NOT NECESSARY FOR PREVIOUS APPLICANTS TO RE-APPLY. Subject: vxWorks in Satellites Submitted-by prb@aplexus.jhuapl.edu Wed Jul 5 11:58:36 1995 Submitted-by: prb@aplexus.jhuapl.edu (Paul R. Bade) Hi, Has anybody been using vxWorks in a satellite? If so, what hardware are you using? Thanks, +====================================================================+ | __ ____ __ __ | | /\ \ / \ \ /\_\__ /\ \ Johns Hopkins University | | / \_\ | /\ \ \ / / /\_\\ \ \ Applied Physics Lab. | | / /\ | |\ \/ \ \ / / / / / \ \ \ | | / \/ |_| \ /\ \_\ / / / / / \ \ \ Paul R. Bade | | / /\__/_/ \ \ \/_// / / / / / / / (301)-953-6000 x8681 | | / / / \ \_\ \ \/ / / / / / prb@aplexus.jhuapl.edu | | \/_/ \/_/ \__/_/ \/_/ | | __ ____ __ __ | | /\ \ / \ \ /\ \ /\ \ | | / \_\ | /\ \ \ / \_\ / \ \ | | / /\ | |\ \/ \ \ / /\ |_| / /\ \ \ | | / \/ |_| \ /\ \_\ / / / | | \/ / \ \ | | / /\ / / \ \ \/_// / / / / / /\ \_\ | | \ \/ / / \ \_\ \ \/ / / \/ / / / | | \__/_/ \/_/ \__/_/ /_/_/ | | | +====================================================================+ From prb@aplexus.jhuapl.edu Wed Jul 5 11:58:36 1995 From: prb@aplexus.jhuapl.edu (Paul R. Bade) Date: Wed Jul 5 11:58:39 PDT 1995 Subject: vxWorks in Satellites Hi, Has anybody been using vxWorks in a satellite? If so, what hardware are you using? Thanks, +====================================================================+ | __ ____ __ __ | | /\ \ / \ \ /\_\__ /\ \ Johns Hopkins University | | / \_\ | /\ \ \ / / /\_\\ \ \ Applied Physics Lab. | | / /\ | |\ \/ \ \ / / / / / \ \ \ | | / \/ |_| \ /\ \_\ / / / / / \ \ \ Paul R. Bade | | / /\__/_/ \ \ \/_// / / / / / / / (301)-953-6000 x8681 | | / / / \ \_\ \ \/ / / / / / prb@aplexus.jhuapl.edu | | \/_/ \/_/ \__/_/ \/_/ | | __ ____ __ __ | | /\ \ / \ \ /\ \ /\ \ | | / \_\ | /\ \ \ / \_\ / \ \ | | / /\ | |\ \/ \ \ / /\ |_| / /\ \ \ | | / \/ |_| \ /\ \_\ / / / | | \/ / \ \ | | / /\ / / \ \ \/_// / / / / / /\ \_\ | | \ \/ / / \ \_\ \ \/ / / \/ / / / | | \__/_/ \/_/ \__/_/ /_/_/ | | | +====================================================================+ Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Thu Jul 6 04:00:21 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Thu Jul 6 04:00:13 PDT 1995 Subject: Green Hills embedded development experience using pSOS+ or vxWorks? Subject: COMPUTER/SOFTWARE ENGINEERS Subject: Why is 0.5 <= 0.4 ???? Subject: Using GNU tools under MSDOS for cross development for VxWorks Subject: Wanted: Real-Time OS Bibliographies (2 Specific Ones) ------------------------------------------------------- Newsgroups: comp.realtime,comp.sys.m68k,comp.os.vxworks Subject: Green Hills embedded development experience using pSOS+ or vxWorks? Date: 5 Jul 1995 08:14:41 -0400 From: bennett@cs.unc.edu (Brad Bennett) Organization: The University of North Carolina Message-ID: <3tdvnh$df7@medusa.cs.unc.edu> Followup-To: poster Does anyone have any development experience using the Green Hills development environment for embedded systems cross-development? In particular, using remote debugging for real-time based O/S's, such as pSOS+ and vxWorks? The development host would be an IBM RS6000. Any feedback would be greatly appreciated. Thanks. Bradley Bennett --------------------------- Newsgroups: comp.os.vxworks,comp.realtime,sdnet.jobs Subject: COMPUTER/SOFTWARE ENGINEERS Date: 26 Jun 1995 20:30:54 GMT From: recruit@sd.aplabs.com Organization: Advanced Processing Laboratories Keywords: COMPUTER SOFTWARE ENGINEERS Message-ID: <3sn5du$59r@sulu.sd.aplabs.com> Reply-To: recruit@sd.aplabs.com Rapidly growing real-time systems integrator needs motivated engineers for software development and systems integration. Working knowledge of UNIX & C is required. Familiarity with VxWorks & VMEbus hardware is a plus. We are looking for individuals with 0-5 years of experience in software development and systems integration. Recent college grads are encouraged to apply. The successful candidate will have the following qualifications: - BS or MS in Computer Science, Computer Engineering or related field - A GPA of 3.25 or above if recent college grad - Strong background in UNIX & C - Good software engineering practices - Experience in real-time software development Submit resume and salary requirements to: AP Labs 5871 Oberlin Dr San Diego, CA 92121 Attn: Recruiting or fax to (619)546-0278 addressed to Recruiting or via email to recruit@sd.aplabs.com --------------------------- Newsgroups: comp.os.vxworks Subject: Why is 0.5 <= 0.4 ???? Date: Fri, 30 Jun 1995 22:17:25 GMT From: guy@boi.hp.com (Guy Burnham) Organization: Hewlett-Packard / Boise, Idaho Message-ID: Sender: news@boi.hp.com (Boise Site News Server) I'm questioning whether I'm loosing my marbles! When I ask the shell in VxWorks 5.1.1 ...... -> .7 <= .6 It comes back with ...... value = 1 So I wrote a little C hack .... int LTorEQ (float a, float b) { return (a <= b); } .... and called it with the following result - -> LTorEQ 0.7,0.6 value = 1 = 0x1 Does anyone elses system think 0.7 is less than or equal to 0.6?? - -- - --|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|-- Guy Burnham, Advanced Laserjet Operations, Hewlett Packard Co. Inc. guy@boi.hp.com 208(396-3355) Disclaimer: Don't you be the only one to take me seriously! - --|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|-- --------------------------- Newsgroups: comp.os.vxworks Subject: Using GNU tools under MSDOS for cross development for VxWorks Date: 4 Jul 1995 05:35:40 GMT From: "Mark R. Milligan" Organization: Coquitlam Custom Computing Message-ID: <3tajvc$jm0@scipio.cyberstore.ca> Hello Everyone: I thought I heared once that there is a version of the GNU compiler that runs under MSDOS and can be used for cross development for VxWorks. I know that there is a version of the GNU compiler and toolset that runs under MSDOS but i think it is only a native intel version and cannot be used for cross compiling. Has anybody tried this?? or knows of someone who has a setup like this? A place I worked for a while back used the Siera High C compiler under MSDOS for Vxworks cross development, this seemed to work ok for Version 5.0.x of vxworks. I have just tried the new VxWorks for Windows with the Green Hills compiler and cross development toolset. I dont like it very much! You cannot control the compiler with a makefile you have to use there "Builder" Tool which I find incomprehensible. It is also increadably slow and memory intensive. I long for the good old days of a compiler and a makefile. God I hate Windows! Somebody please put me out of my misery. --------------------------- Newsgroups: comp.sys.m68k,comp.arch.embedded,comp.realtime,comp.os.chorus,comp.os.vxworks,comp.os.qnx,comp.os.lynx Subject: Wanted: Real-Time OS Bibliographies (2 Specific Ones) Date: 28 Jun 1995 21:12:01 GMT From: c2xkjb@mail.delcoelect.com (Kevin J Brewer) Organization: Delco Electronics Corp. Keywords: RTOS, realtime, bibliography Message-ID: <3ssgj1$gi4@kocrsv08.delcoelect.com> I am looking for books on real-time OSes which describe the most efficient inter-process communication methods, by way of specific case studies, for various process situations. Such books would discuss, in case studies, which synchronization or communication method is the best for interaction between two (or more) particular processes in a certain situation or type of processing. For the particular situation, which is better: a mailbox or a message queue, a counting semaphore or a binary semaphore, a binary semaphore or a mutex, etc.? A couple weeks ago, of the library books I happened to pick up, _one_ had such an example (case study) in which the scan codes of the keys depressed were communicated from the keyboard ISR to an ASCII code building process. I believe that the scan codes were passed as messages and I think that they were in a message queue, rather than in single message mailboxes. It gave the "whys and wherefores" of why this particular method was "best" (most efficient, but transmitted all of the required information) for the situation. I don't remember any other details about the interactions of the processes. Now that I wish to study this information in detail, I can't remember in which book I saw this example. This example was the kind of case study form of describing the communication methods that I was interested in. So, if any of you have a bibliography on such books, please forward it to me. I'm also looking for books on real-time OSes which contain defining algorithms, in pseudo-code, of the various process synchronization and communication operations (mailboxes, message queues, counting semaphores, binary semaphores, mutexes, rendezvouses, etc. - any which can be found), including those particularly useful algorithmic forms in which, say, binary semaphores are manifested as merely restricted counting semaphores or a mutex is merely a run-time modified binary semaphore, etc. If anyone has a bibliography of this kind of books, please forward it to me also. Thanks, much in advance for any help!! - -- - - Kevin J. Brewer Phone: (317) 451-0648 GM Line: 8-322-0648 Just Say No To C(ryptic) FAX: (317) 451-0174 8-322-0174 Microcomputer Architecture Delco Electronics Corporation --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Thu Jul 6 04:00:21 1995 From: daemon@csg.lbl.gov Date: Thu Jul 6 04:00:24 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 6 04:00:13 PDT 1995 Subject: Green Hills embedded development experience using pSOS+ or vxWorks? Subject: COMPUTER/SOFTWARE ENGINEERS Subject: Why is 0.5 <= 0.4 ???? Subject: Using GNU tools under MSDOS for cross development for VxWorks Subject: Wanted: Real-Time OS Bibliographies (2 Specific Ones) ------------------------------------------------------- Newsgroups: comp.realtime,comp.sys.m68k,comp.os.vxworks Subject: Green Hills embedded development experience using pSOS+ or vxWorks? Date: 5 Jul 1995 08:14:41 -0400 From: bennett@cs.unc.edu (Brad Bennett) Organization: The University of North Carolina Message-ID: <3tdvnh$df7@medusa.cs.unc.edu> Followup-To: poster Does anyone have any development experience using the Green Hills development environment for embedded systems cross-development? In particular, using remote debugging for real-time based O/S's, such as pSOS+ and vxWorks? The development host would be an IBM RS6000. Any feedback would be greatly appreciated. Thanks. Bradley Bennett --------------------------- Newsgroups: comp.os.vxworks,comp.realtime,sdnet.jobs Subject: COMPUTER/SOFTWARE ENGINEERS Date: 26 Jun 1995 20:30:54 GMT From: recruit@sd.aplabs.com Organization: Advanced Processing Laboratories Keywords: COMPUTER SOFTWARE ENGINEERS Message-ID: <3sn5du$59r@sulu.sd.aplabs.com> Reply-To: recruit@sd.aplabs.com Rapidly growing real-time systems integrator needs motivated engineers for software development and systems integration. Working knowledge of UNIX & C is required. Familiarity with VxWorks & VMEbus hardware is a plus. We are looking for individuals with 0-5 years of experience in software development and systems integration. Recent college grads are encouraged to apply. The successful candidate will have the following qualifications: - BS or MS in Computer Science, Computer Engineering or related field - A GPA of 3.25 or above if recent college grad - Strong background in UNIX & C - Good software engineering practices - Experience in real-time software development Submit resume and salary requirements to: AP Labs 5871 Oberlin Dr San Diego, CA 92121 Attn: Recruiting or fax to (619)546-0278 addressed to Recruiting or via email to recruit@sd.aplabs.com --------------------------- Newsgroups: comp.os.vxworks Subject: Why is 0.5 <= 0.4 ???? Date: Fri, 30 Jun 1995 22:17:25 GMT From: guy@boi.hp.com (Guy Burnham) Organization: Hewlett-Packard / Boise, Idaho Message-ID: Sender: news@boi.hp.com (Boise Site News Server) I'm questioning whether I'm loosing my marbles! When I ask the shell in VxWorks 5.1.1 ...... -> .7 <= .6 It comes back with ...... value = 1 So I wrote a little C hack .... int LTorEQ (float a, float b) { return (a <= b); } .... and called it with the following result - -> LTorEQ 0.7,0.6 value = 1 = 0x1 Does anyone elses system think 0.7 is less than or equal to 0.6?? - -- - --|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|-- Guy Burnham, Advanced Laserjet Operations, Hewlett Packard Co. Inc. guy@boi.hp.com 208(396-3355) Disclaimer: Don't you be the only one to take me seriously! - --|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|--|-- --------------------------- Newsgroups: comp.os.vxworks Subject: Using GNU tools under MSDOS for cross development for VxWorks Date: 4 Jul 1995 05:35:40 GMT From: "Mark R. Milligan" Organization: Coquitlam Custom Computing Message-ID: <3tajvc$jm0@scipio.cyberstore.ca> Hello Everyone: I thought I heared once that there is a version of the GNU compiler that runs under MSDOS and can be used for cross development for VxWorks. I know that there is a version of the GNU compiler and toolset that runs under MSDOS but i think it is only a native intel version and cannot be used for cross compiling. Has anybody tried this?? or knows of someone who has a setup like this? A place I worked for a while back used the Siera High C compiler under MSDOS for Vxworks cross development, this seemed to work ok for Version 5.0.x of vxworks. I have just tried the new VxWorks for Windows with the Green Hills compiler and cross development toolset. I dont like it very much! You cannot control the compiler with a makefile you have to use there "Builder" Tool which I find incomprehensible. It is also increadably slow and memory intensive. I long for the good old days of a compiler and a makefile. God I hate Windows! Somebody please put me out of my misery. --------------------------- Newsgroups: comp.sys.m68k,comp.arch.embedded,comp.realtime,comp.os.chorus,comp.os.vxworks,comp.os.qnx,comp.os.lynx Subject: Wanted: Real-Time OS Bibliographies (2 Specific Ones) Date: 28 Jun 1995 21:12:01 GMT From: c2xkjb@mail.delcoelect.com (Kevin J Brewer) Organization: Delco Electronics Corp. Keywords: RTOS, realtime, bibliography Message-ID: <3ssgj1$gi4@kocrsv08.delcoelect.com> I am looking for books on real-time OSes which describe the most efficient inter-process communication methods, by way of specific case studies, for various process situations. Such books would discuss, in case studies, which synchronization or communication method is the best for interaction between two (or more) particular processes in a certain situation or type of processing. For the particular situation, which is better: a mailbox or a message queue, a counting semaphore or a binary semaphore, a binary semaphore or a mutex, etc.? A couple weeks ago, of the library books I happened to pick up, _one_ had such an example (case study) in which the scan codes of the keys depressed were communicated from the keyboard ISR to an ASCII code building process. I believe that the scan codes were passed as messages and I think that they were in a message queue, rather than in single message mailboxes. It gave the "whys and wherefores" of why this particular method was "best" (most efficient, but transmitted all of the required information) for the situation. I don't remember any other details about the interactions of the processes. Now that I wish to study this information in detail, I can't remember in which book I saw this example. This example was the kind of case study form of describing the communication methods that I was interested in. So, if any of you have a bibliography on such books, please forward it to me. I'm also looking for books on real-time OSes which contain defining algorithms, in pseudo-code, of the various process synchronization and communication operations (mailboxes, message queues, counting semaphores, binary semaphores, mutexes, rendezvouses, etc. - any which can be found), including those particularly useful algorithmic forms in which, say, binary semaphores are manifested as merely restricted counting semaphores or a mutex is merely a run-time modified binary semaphore, etc. If anyone has a bibliography of this kind of books, please forward it to me also. Thanks, much in advance for any help!! - -- - - Kevin J. Brewer Phone: (317) 451-0648 GM Line: 8-322-0648 Just Say No To C(ryptic) FAX: (317) 451-0174 8-322-0174 Microcomputer Architecture Delco Electronics Corporation --------------------------- End of New-News digest ********************** Subject: Open Position, New Description Submitted-by brodrigu@polaris.cv.nrao.edu Thu Jul 6 06:46:27 1995 Submitted-by: "Billie Rodriguez, 804 296-0312" The Job Opening I sent yesterday, July 5, was an outdated version of the description below. Please disregard yesterday's e-mail. DESCRIPTION: The National Radio Astronomy Observatory (NRAO) has an opening for an Electronics Engineer (Software and Digital Systems) to work in Green Bank, West Virginia on the Green Bank Telescope Project (GBT). The GBT is a fully offset, 100-meter diameter radio telescope which, when completed, will be the largest and most advanced antenna of its type in the world. The successful applicant will assist in the design, construction, and support of software and hardware systems for the GBT. Responsibilities will include coding of real-time control software, hardware/software integration, and support of hardware/software systems. Some hardware design and support of existing systems will be required. This position is based at our Green Bank, West Virginia facility. Green Bank is a small community nestled in the Allegheny Mountains, near Snowshoe Ski Resort. Here you will have a stimulating professional challenge in an academic-type environment accompanied by a carefree country lifestyle. QUALIFICATIONS: Bachelors degree in Electrical or Computer Engineering, with 5 years experience in programming real-time systems, digital and analog electronics, writing driver software, and integration of hardware/software systems. Familiarity with modern engineering and programming practices, state-of-the-art digital devices, and with C or C++, UNIX, and VxWorks, or a similar real-time operating system is required. The applicant should be experienced in hardware/software integration and trouble-shooting real-time systems. Experience with Object Oriented design, network programming, and the Motif toolkit is also preferred. For immediate consideration, please send resume to: Mr. Tim Weadon National Radio Astronomy Observatory P.O. Box 2 Green Bank, WV 24944-0002 or, via e-mail to: brodrigu@nrao.edu From brodrigu@polaris.cv.nrao.edu Thu Jul 6 06:46:27 1995 From: "Billie Rodriguez, 804 296-0312" Date: Thu Jul 6 06:46:30 PDT 1995 Subject: Open Position, New Description The Job Opening I sent yesterday, July 5, was an outdated version of the description below. Please disregard yesterday's e-mail. DESCRIPTION: The National Radio Astronomy Observatory (NRAO) has an opening for an Electronics Engineer (Software and Digital Systems) to work in Green Bank, West Virginia on the Green Bank Telescope Project (GBT). The GBT is a fully offset, 100-meter diameter radio telescope which, when completed, will be the largest and most advanced antenna of its type in the world. The successful applicant will assist in the design, construction, and support of software and hardware systems for the GBT. Responsibilities will include coding of real-time control software, hardware/software integration, and support of hardware/software systems. Some hardware design and support of existing systems will be required. This position is based at our Green Bank, West Virginia facility. Green Bank is a small community nestled in the Allegheny Mountains, near Snowshoe Ski Resort. Here you will have a stimulating professional challenge in an academic-type environment accompanied by a carefree country lifestyle. QUALIFICATIONS: Bachelors degree in Electrical or Computer Engineering, with 5 years experience in programming real-time systems, digital and analog electronics, writing driver software, and integration of hardware/software systems. Familiarity with modern engineering and programming practices, state-of-the-art digital devices, and with C or C++, UNIX, and VxWorks, or a similar real-time operating system is required. The applicant should be experienced in hardware/software integration and trouble-shooting real-time systems. Experience with Object Oriented design, network programming, and the Motif toolkit is also preferred. For immediate consideration, please send resume to: Mr. Tim Weadon National Radio Astronomy Observatory P.O. Box 2 Green Bank, WV 24944-0002 or, via e-mail to: brodrigu@nrao.edu Subject: unsubscribe Submitted-by john@paul.visicom.com Thu Jul 6 10:43:42 1995 Submitted-by: "John Bloomquist" unsubscribe Please unsubscribe. ---------------------------------------------------------------------------------------------------------------------------- John Bloomquist e-mail: john@visicom.com 16703 SE McGillivary Blvd Suite 210 voice: (360) 604-9771 Camas, WA 98607 fax: (360) 604-9773 From john@paul.visicom.com Thu Jul 6 10:43:42 1995 From: "John Bloomquist" Date: Thu Jul 6 10:43:45 PDT 1995 Subject: unsubscribe unsubscribe Please unsubscribe. ---------------------------------------------------------------------------------------------------------------------------- John Bloomquist e-mail: john@visicom.com 16703 SE McGillivary Blvd Suite 210 voice: (360) 604-9771 Camas, WA 98607 fax: (360) 604-9773 Subject: unsubscribe Submitted-by john@paul.visicom.com Thu Jul 6 10:43:48 1995 Submitted-by: "John Bloomquist" unsubscribe Please unsubscribe. ---------------------------------------------------------------------------------------------------------------------------- John Bloomquist e-mail: john@visicom.com 16703 SE McGillivary Blvd Suite 210 voice: (360) 604-9771 Camas, WA 98607 fax: (360) 604-9773 From john@paul.visicom.com Thu Jul 6 10:43:48 1995 From: "John Bloomquist" Date: Thu Jul 6 10:43:50 PDT 1995 Subject: unsubscribe unsubscribe Please unsubscribe. ---------------------------------------------------------------------------------------------------------------------------- John Bloomquist e-mail: john@visicom.com 16703 SE McGillivary Blvd Suite 210 voice: (360) 604-9771 Camas, WA 98607 fax: (360) 604-9773 Subject: Re: Why is 0.5 <= 0.4 ???? Submitted-by steele@telerobotics.jpl.nasa.gov Thu Jul 6 19:27:18 1995 Submitted-by: Rob Steele Relative to the recent e-mail on this curiousity, I did the following user 5.1.1 -> 1.5 <= 1.1 value = 1 -> 1.1 <= 1.5 value = 0 -> 0.6 <= 0.7 value = 0 -> 0.7 <= 0.6 value = 1 Does anyone know why yet? Rob Steele From steele@telerobotics.jpl.nasa.gov Thu Jul 6 19:27:18 1995 From: Rob Steele Date: Thu Jul 6 19:27:21 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? Relative to the recent e-mail on this curiousity, I did the following user 5.1.1 -> 1.5 <= 1.1 value = 1 -> 1.1 <= 1.5 value = 0 -> 0.6 <= 0.7 value = 0 -> 0.7 <= 0.6 value = 1 Does anyone know why yet? Rob Steele Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Fri Jul 7 04:00:21 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Fri Jul 7 04:00:17 PDT 1995 Subject: Re: vxWorks in Satellites Subject: Re: RIP for VxWorks Subject: Re: 162FX and Quad RS422 IP module. Subject: Write a file with NFS Subject: cplusCtorsLink() segmentation error Subject: Re: Can not Make vxWorks_rom Subject: Analysis software for time-stamped event data? Subject: Re: sizeof() in vxWorks Subject: stripping objects for vxworks Subject: GCC w/C++ Subject: Re: Why is 0.5 <= 0.4 ???? Subject: Using the 167-bug ROMS with custom ROMS ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks in Satellites Date: 6 Jul 1995 11:13:33 GMT From: dyer@alx.sticomet.com (Doug Dyer) Organization: Naval Research Laboratory Message-ID: <3tgggt$i6l@ra.nrl.navy.mil> References: <9507051858.AA05061@aplexus.jhuapl.edu> prb@aplexus.jhuapl.edu (Paul R. Bade) writes: >Hi, >Has anybody been using vxWorks in a satellite? >If so, what hardware are you using? > Thanks, Paul, we have used VxWorks on Clementine but it ran on a mil-standard MIPS chip, not a rad-hard part (sort of an experiment in itself). There are a few other projects that will be using a mil-spec MIPs chip in orbit to save $$$. VxWorks will also fly on the Air Force "Argos" satellite in the Harris RH3000 (an experimental rad-hard MIPS chip). There are a few other experimental RH-MIPS boards but none will be tested earlier than Argos. I don't recommend this particular chip, but the Harris RHC-3000 BSP is 99% complete. The concern would be the price over MIL-STD parts. Also, I have lost track of the progress of Honeywell's RH-32 - a MIPS variant which comes with its own compiler and vxWorks BSP. I know there are rad-hard 386s and VxWorks supports Intel now. I don't know what the state of the RH386 or the Intel BSP is. The cool thing about vxWorks in satellites is where you can significantly improve the development cycle by chucking those 1970s-era development tools as well as a more sophisticated embedded system design (heh, let alone the 1750 cpu). - -- Doug Dyer - dyer@alx.sticomet.com | Blatant Ad: ECL - embedded "C-like" Software Technology, Inc. (STI) | scripting language. DC office: (703) 329-9707 | Fun for the whole family! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RIP for VxWorks Date: 23 Jun 1995 21:59:06 GMT From: switt@triceratops.hac.com (Steve Witt) Organization: Hughes Aircraft Company Message-ID: <3sfdfa$55k@hacgate2.hac.com> References: <9506231048.AA06396@wrsec.fr> jerome@wrsec.fr wrote: : Hello, : Does anyone have ported RIP (Routing IP) to VxWorks targets and ready to : share this software? If not how much effort would be necssary to port it? : Thanks for your help. : Jerome BRING : ------------------------------------------------------------------- : Jerome BRING | Phone: +33 1 69-07-78-78 : Southern Europe Region Manager | FAX: +33 1 69-07-08-26 : Wind River Systems | Mobile: +33 07-72-19-48 : 27 Avenue de la Baltique | email: jerome@wrsec.fr : 91962 Les Ulis, FRANCE | WWW: http://www.wrs.com : ------------------------------------------------------------------- Some kind soul has already done it. Look at the VxWorks archive site and I believe it is called 'vxrouted'. I was able to use it as the core of a RIP implementation I did. I needed to modify it to a certain extent to integrate it into our product. Steve Witt --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 162FX and Quad RS422 IP module. Date: 30 Jun 1995 04:15:39 GMT From: Paul Anderson Organization: Compware Corporation Message-ID: <3svtpb$hdc@infoman.net99.net> References: <9506281010.AA05065@redifon.demon.co.uk> "M. S. Armstrong" wrote: > >Hello people, > > I am just about to put together a system in which I hope to use the >MVME162FX. > It's been said that the standard 162 BSP handles >the 162FX, but I recall someone a few months ago claiming problems when using >it. If anyone has used the FX and has opinions on it's performance could you >give me a hint about any nasties before I trial and buy the board, please. > >It is my intension to populate this board with quad serial RS422/RS232 IP >modules from TEWS in Germany. > Tel: +49 (0)4101 42637 or ask your local distributor for a TIP865. > > >Cheers, > > Mark... > ><> Mark Armstrong <> * * <> <> ><> Redifon Technology <> * <> Tel (0372) 376677 <> ><> 9 Mole Business Park, <> * o <> Fax (0372) 379480 <> ><> Randalls Road, <> * <\> <> msa@redifon.demon.co.uk <> ><> Leatherhead, Surrey <> __ /_\__, <> <> ><> KT22 7BA, England <> <> A Thompson CSF Company <> > Hi Mark Some advice on the 162FX and an advert for our serial drivers for IP's. The 162 FX requires the updated 162FX from WRS,Inc. The most evident change is to accommodate the "command register" implementation of the 85C30(85C230?) versus the memory-mapped register set of past 162's. Basically, the previous BSP fails to operate the serial port. Compware offers serial drivers for the IP-Serial(85C30), the IP-UniversalSerial(16C30, pretty high performance), and the IP-OctalSerial(8-ch async up to 38.4). All offer 422 compatibility. I'll fax our price list asap. If you have interest, please email/fax/call, and I will forward docs. Best regards, Paul Anderson (paul@compware.com) Compware Corporation Tel.: 602-998-8650, 800-788-5855 9735 N. 90th Place, Suite 200 Facs.: 602-661-0019 Scottsdale, AZ 85258 --------------------------- Newsgroups: comp.os.vxworks Subject: Write a file with NFS Date: 6 Jul 1995 12:14:53 GMT From: gournay@sigx0.cea.fr (J.F. Gournay) Organization: CEA Message-ID: <3tgk3t$s1c@anemone.saclay.cea.fr> Reply-To: gournay@sigx0.cea.fr We are developing an acquisition system on vxWorks. A task has to write a buffer to a file periodically (buffer length : a few thousand bytes, period : a few seconds). We open the file once at the beginning of the acquisition to avoid wasting time in open - lseek - write - close sequences. We are using NFS protocol to do so. Problem : Usually, the write function takes less than 100 msec, but from time to time it takes 5 or 10 seconds (but never 1 second or 2 or ...) !!! We changed the CPU (68030 or 68040), the NFS server, the length of the file, but the problem persists. Any advice would be greatly appreciated. Thanks. JF Gournay --------------------------- Newsgroups: comp.os.vxworks Subject: cplusCtorsLink() segmentation error Date: Fri, 23 Jun 1995 14:02:18 GMT From: idawson@mlsma.mlm.att.com (nsi3520000-Dawson) Organization: AT&T Message-ID: Sender: news@ssbunews.ih.att.com (Netnews Administration) Does anyone have a patch for vxsim to get around the runtime segmentation error, when adding the call to cplusCtorsLink() in the function usrRoot(), in usrConfig.c. #ifdef INCLUDE_CPLUS /* unbundled C++ product */ cplusLibInit (); cplusCtorsLink(); #endif Thanks Ian idawson@att.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can not Make vxWorks_rom Date: 28 Jun 1995 23:33:12 GMT From: Graham Waters Organization: The University of British Columbia Message-ID: <3ssoro$adi@nntp.ucs.ubc.ca> References: rls@ltis.loral.com (Randy Silagi) wrote: > > > I am trying to make a bootable VxWorks image without the Shell or Symbol > table. According to the makefile this should be vxWorks_rom. But, this > fails during the linking/loading. I get the following errors: > > if_ether.o(.text+1f4): undefined reference to `_loif' > if_ether.o(.text+1fa): undefined reference to `_looutput' > uipc_mbuf.o(.text+de): undefined reference to `_mbufConfig' > uipc_mbuf.o(.text+13e): undefined reference to `_clusterConfig' > > I had the exact same messages when building a rommable standalone application with "make vxWorks.st_rom.hex. In my case I undefined "INCLUDE_NETWORK" because my application didn't need it. However I had left a piece of application code that created a socket. I wrapped it in a conditional and the undefined went away. I suspect you may have done something similar. All it takes is one loose s=socket(AF_INET, SOCK_DGRAM, 0) or any other network function and you have a problem. +------------------------------------------------------------------+ | Graham Waters E-Mail: waters@triumf.ca | TRIUMF, Meson Research Facility | 4004 Wesbrooke Mall Phone: 604-222-1047 x6531 | Vancouver, BC, Canada Fax: 604-222-7307 +----------------------------------------------------------------+ Unix reacts to the user as though it is being annoyed --------------------------- Newsgroups: comp.realtime,comp.os.vxworks,comp.sys.m68k Subject: Analysis software for time-stamped event data? Date: 5 Jul 1995 14:53:36 -0400 From: bennett@cs.unc.edu (Brad Bennett) Organization: The University of North Carolina Message-ID: <3ten3g$eog@medusa.cs.unc.edu> Followup-To: poster I'm in the processing of designing an time-stamped event recording machanism for use in analyzing application performance based on a Real-Time O/S. Doesn anyone know of any tools for analyzing this kind of information? The kind of information I'd like to extract is % task utilization, amount of times in ISR's, etc... Any pointers would be greatly appreciated,,,, - --bennett@cs.unc.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sizeof() in vxWorks Date: 27 Jun 1995 16:14:58 GMT From: garyjohn@spk.hp.com (Gary Johnson) Organization: Hewlett Packard Message-ID: <3spaq2$lv7@hpscit.sc.hp.com> References: <3snivq$2f7@ixnews6.ix.netcom.com> Marc Tillinghast (tilly@ix.netcom.com) wrote: > I have a struct that consists of 49 chars. Under vxWorks > sizeof(myStruct) returns 50. Under UNIX sizeof returns 49. > Does vxWorks consistently return the size + 1 of structs. > Any info would be appreciated. > > Thanks, > Marc Tillinghast (tilly@storm.com) Sizeof() returns the amount of memory allocated for the object rather than the amount of data that the object "contains". The result returned by sizeof() depends on the alignment used by the compiler. This can sometimes be affected by command-line options and/or #pragmas. It appears that your VxWorks compiler aligns structs on word (2-char) boundaries while your UNIX compiler aligns structs on char boundaries. A compiler that aligned structs on long-word (4-char) boundaries would return a sizeof() value of 52. Gary --------------------------- Newsgroups: comp.os.vxworks Subject: stripping objects for vxworks Date: 27 Jun 1995 17:50:10 GMT From: ahrens@informatik.hu-berlin.de (Klaus Ahrens) Organization: Humboldt-University Berlin Message-ID: <3spgci$92f@hahn.informatik.hu-berlin.de> Reply-To: ahrens@informatik.hu-berlin.de Dear vxworkers, we have to reduce the size of a 68k object file which is compiled with the WindRiver C++ gateway from huge generated C++ sources to fit into the RAM of our force board. The size could be reduced by stripping all symbols off (~4Mb less), but unfortunately this also strips the (in the running version) only needed entry point to start up the system. How to strip to an object that contains only some symbols or how to locate the entry point in a fully stripped loaded object ???????? Any suggestions ? Please reply by email ! Thanx in advance Klaus - -- _______________________________________________________________________ /| | | || Dr. K. Ahrens | Phone +49-30-20181 238 | || Institut fuer Informatik | Fax +49-30-20181 234 | || Humboldt-Universitaet | email ahrens@informatik.hu-berlin.de | || | http://www.informatik.hu-berlin.de/~ahrens| || +-------------------------------------------+ || Lindenstrasse 54a | +---------------------------------------+ | || 10117 Berlin | |___ Hindsight is an exact science. ___/| | || Germany | +--------------------------------------++ | ||___________________________|___________________________________________| |/___________________________/__________________________________________/ --------------------------- Newsgroups: comp.os.vxworks Subject: GCC w/C++ Date: 6 Jul 1995 17:46:35 GMT From: halpin@codem.com (Stephen E. Halpin) Organization: CODEM Systems, Inc. Message-ID: <3th7hr$1n6@regina.seh.codem.com> We are using VxWorks 5.1.1 (hopefully soon to be 5.2 when the BSP is validated) with the GCC compiler. We are interested in building GCC 2.7.0 with C++ support enabled. Before I waste a lot of time on this, I was wondering if the object modules that are ultimately produced from C++ sources are compatable with VxWorks, or does the extra linkage support of C++ require support that VxWorks doesnt have out of the box? Thanks, -Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: 02 Jul 1995 05:34:17 GMT From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: References: In article guy@boi.hp.com (Guy Burnham) writes: :I'm questioning whether I'm loosing my marbles! When I ask the shell in :VxWorks 5.1.1 ...... : -> .7 <= .6 :It comes back with ...... :value = 1 While I don't know the reason for the above, which also happens on our VxWorks 5.1.1 setup, I can shed light on your second question. :So I wrote a little C hack .... : :int LTorEQ (float a, float b) : { : return (a <= b); : } : :.... and called it with the following result : :-> LTorEQ 0.7,0.6 :value = 1 = 0x1 The VxWorks shell passes floating point arguments as doubles, whereas your function expects floats. The result is system dependent, but on our MC68040 based system the values of a and b become 1.8 and 2.72008e+23 respectively. This makes the quoted return value understandable. To get your function to work as intended, simply change its declaration to: int LTorEQ (double a, double b) Martin Shepherd (mcs@astro.caltech.edu) --------------------------- Newsgroups: comp.os.vxworks Subject: Using the 167-bug ROMS with custom ROMS Date: Mon, 3 Jul 1995 05:13:49 GMT From: aa719@FreeNet.Carleton.CA (Dave Phillips) Organization: The National Capital FreeNet Message-ID: Reply-To: aa719@FreeNet.Carleton.CA (Dave Phillips) Sender: aa719@freenet3.carleton.ca (Dave Phillips) A short while back, somebody posted info on how to install and use the 167-bug (came with the MVME-167) ROMS and still use custom (boot) ROMS. Could someone please give me some info or a repost? Thanx. --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Fri Jul 7 04:00:21 1995 From: daemon@csg.lbl.gov Date: Fri Jul 7 04:00:25 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 7 04:00:17 PDT 1995 Subject: Re: vxWorks in Satellites Subject: Re: RIP for VxWorks Subject: Re: 162FX and Quad RS422 IP module. Subject: Write a file with NFS Subject: cplusCtorsLink() segmentation error Subject: Re: Can not Make vxWorks_rom Subject: Analysis software for time-stamped event data? Subject: Re: sizeof() in vxWorks Subject: stripping objects for vxworks Subject: GCC w/C++ Subject: Re: Why is 0.5 <= 0.4 ???? Subject: Using the 167-bug ROMS with custom ROMS ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: vxWorks in Satellites Date: 6 Jul 1995 11:13:33 GMT From: dyer@alx.sticomet.com (Doug Dyer) Organization: Naval Research Laboratory Message-ID: <3tgggt$i6l@ra.nrl.navy.mil> References: <9507051858.AA05061@aplexus.jhuapl.edu> prb@aplexus.jhuapl.edu (Paul R. Bade) writes: >Hi, >Has anybody been using vxWorks in a satellite? >If so, what hardware are you using? > Thanks, Paul, we have used VxWorks on Clementine but it ran on a mil-standard MIPS chip, not a rad-hard part (sort of an experiment in itself). There are a few other projects that will be using a mil-spec MIPs chip in orbit to save $$$. VxWorks will also fly on the Air Force "Argos" satellite in the Harris RH3000 (an experimental rad-hard MIPS chip). There are a few other experimental RH-MIPS boards but none will be tested earlier than Argos. I don't recommend this particular chip, but the Harris RHC-3000 BSP is 99% complete. The concern would be the price over MIL-STD parts. Also, I have lost track of the progress of Honeywell's RH-32 - a MIPS variant which comes with its own compiler and vxWorks BSP. I know there are rad-hard 386s and VxWorks supports Intel now. I don't know what the state of the RH386 or the Intel BSP is. The cool thing about vxWorks in satellites is where you can significantly improve the development cycle by chucking those 1970s-era development tools as well as a more sophisticated embedded system design (heh, let alone the 1750 cpu). - -- Doug Dyer - dyer@alx.sticomet.com | Blatant Ad: ECL - embedded "C-like" Software Technology, Inc. (STI) | scripting language. DC office: (703) 329-9707 | Fun for the whole family! --------------------------- Newsgroups: comp.os.vxworks Subject: Re: RIP for VxWorks Date: 23 Jun 1995 21:59:06 GMT From: switt@triceratops.hac.com (Steve Witt) Organization: Hughes Aircraft Company Message-ID: <3sfdfa$55k@hacgate2.hac.com> References: <9506231048.AA06396@wrsec.fr> jerome@wrsec.fr wrote: : Hello, : Does anyone have ported RIP (Routing IP) to VxWorks targets and ready to : share this software? If not how much effort would be necssary to port it? : Thanks for your help. : Jerome BRING : ------------------------------------------------------------------- : Jerome BRING | Phone: +33 1 69-07-78-78 : Southern Europe Region Manager | FAX: +33 1 69-07-08-26 : Wind River Systems | Mobile: +33 07-72-19-48 : 27 Avenue de la Baltique | email: jerome@wrsec.fr : 91962 Les Ulis, FRANCE | WWW: http://www.wrs.com : ------------------------------------------------------------------- Some kind soul has already done it. Look at the VxWorks archive site and I believe it is called 'vxrouted'. I was able to use it as the core of a RIP implementation I did. I needed to modify it to a certain extent to integrate it into our product. Steve Witt --------------------------- Newsgroups: comp.os.vxworks Subject: Re: 162FX and Quad RS422 IP module. Date: 30 Jun 1995 04:15:39 GMT From: Paul Anderson Organization: Compware Corporation Message-ID: <3svtpb$hdc@infoman.net99.net> References: <9506281010.AA05065@redifon.demon.co.uk> "M. S. Armstrong" wrote: > >Hello people, > > I am just about to put together a system in which I hope to use the >MVME162FX. > It's been said that the standard 162 BSP handles >the 162FX, but I recall someone a few months ago claiming problems when using >it. If anyone has used the FX and has opinions on it's performance could you >give me a hint about any nasties before I trial and buy the board, please. > >It is my intension to populate this board with quad serial RS422/RS232 IP >modules from TEWS in Germany. > Tel: +49 (0)4101 42637 or ask your local distributor for a TIP865. > > >Cheers, > > Mark... > ><> Mark Armstrong <> * * <> <> ><> Redifon Technology <> * <> Tel (0372) 376677 <> ><> 9 Mole Business Park, <> * o <> Fax (0372) 379480 <> ><> Randalls Road, <> * <\> <> msa@redifon.demon.co.uk <> ><> Leatherhead, Surrey <> __ /_\__, <> <> ><> KT22 7BA, England <> <> A Thompson CSF Company <> > Hi Mark Some advice on the 162FX and an advert for our serial drivers for IP's. The 162 FX requires the updated 162FX from WRS,Inc. The most evident change is to accommodate the "command register" implementation of the 85C30(85C230?) versus the memory-mapped register set of past 162's. Basically, the previous BSP fails to operate the serial port. Compware offers serial drivers for the IP-Serial(85C30), the IP-UniversalSerial(16C30, pretty high performance), and the IP-OctalSerial(8-ch async up to 38.4). All offer 422 compatibility. I'll fax our price list asap. If you have interest, please email/fax/call, and I will forward docs. Best regards, Paul Anderson (paul@compware.com) Compware Corporation Tel.: 602-998-8650, 800-788-5855 9735 N. 90th Place, Suite 200 Facs.: 602-661-0019 Scottsdale, AZ 85258 --------------------------- Newsgroups: comp.os.vxworks Subject: Write a file with NFS Date: 6 Jul 1995 12:14:53 GMT From: gournay@sigx0.cea.fr (J.F. Gournay) Organization: CEA Message-ID: <3tgk3t$s1c@anemone.saclay.cea.fr> Reply-To: gournay@sigx0.cea.fr We are developing an acquisition system on vxWorks. A task has to write a buffer to a file periodically (buffer length : a few thousand bytes, period : a few seconds). We open the file once at the beginning of the acquisition to avoid wasting time in open - lseek - write - close sequences. We are using NFS protocol to do so. Problem : Usually, the write function takes less than 100 msec, but from time to time it takes 5 or 10 seconds (but never 1 second or 2 or ...) !!! We changed the CPU (68030 or 68040), the NFS server, the length of the file, but the problem persists. Any advice would be greatly appreciated. Thanks. JF Gournay --------------------------- Newsgroups: comp.os.vxworks Subject: cplusCtorsLink() segmentation error Date: Fri, 23 Jun 1995 14:02:18 GMT From: idawson@mlsma.mlm.att.com (nsi3520000-Dawson) Organization: AT&T Message-ID: Sender: news@ssbunews.ih.att.com (Netnews Administration) Does anyone have a patch for vxsim to get around the runtime segmentation error, when adding the call to cplusCtorsLink() in the function usrRoot(), in usrConfig.c. #ifdef INCLUDE_CPLUS /* unbundled C++ product */ cplusLibInit (); cplusCtorsLink(); #endif Thanks Ian idawson@att.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Can not Make vxWorks_rom Date: 28 Jun 1995 23:33:12 GMT From: Graham Waters Organization: The University of British Columbia Message-ID: <3ssoro$adi@nntp.ucs.ubc.ca> References: rls@ltis.loral.com (Randy Silagi) wrote: > > > I am trying to make a bootable VxWorks image without the Shell or Symbol > table. According to the makefile this should be vxWorks_rom. But, this > fails during the linking/loading. I get the following errors: > > if_ether.o(.text+1f4): undefined reference to `_loif' > if_ether.o(.text+1fa): undefined reference to `_looutput' > uipc_mbuf.o(.text+de): undefined reference to `_mbufConfig' > uipc_mbuf.o(.text+13e): undefined reference to `_clusterConfig' > > I had the exact same messages when building a rommable standalone application with "make vxWorks.st_rom.hex. In my case I undefined "INCLUDE_NETWORK" because my application didn't need it. However I had left a piece of application code that created a socket. I wrapped it in a conditional and the undefined went away. I suspect you may have done something similar. All it takes is one loose s=socket(AF_INET, SOCK_DGRAM, 0) or any other network function and you have a problem. +------------------------------------------------------------------+ | Graham Waters E-Mail: waters@triumf.ca | TRIUMF, Meson Research Facility | 4004 Wesbrooke Mall Phone: 604-222-1047 x6531 | Vancouver, BC, Canada Fax: 604-222-7307 +----------------------------------------------------------------+ Unix reacts to the user as though it is being annoyed --------------------------- Newsgroups: comp.realtime,comp.os.vxworks,comp.sys.m68k Subject: Analysis software for time-stamped event data? Date: 5 Jul 1995 14:53:36 -0400 From: bennett@cs.unc.edu (Brad Bennett) Organization: The University of North Carolina Message-ID: <3ten3g$eog@medusa.cs.unc.edu> Followup-To: poster I'm in the processing of designing an time-stamped event recording machanism for use in analyzing application performance based on a Real-Time O/S. Doesn anyone know of any tools for analyzing this kind of information? The kind of information I'd like to extract is % task utilization, amount of times in ISR's, etc... Any pointers would be greatly appreciated,,,, - --bennett@cs.unc.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: sizeof() in vxWorks Date: 27 Jun 1995 16:14:58 GMT From: garyjohn@spk.hp.com (Gary Johnson) Organization: Hewlett Packard Message-ID: <3spaq2$lv7@hpscit.sc.hp.com> References: <3snivq$2f7@ixnews6.ix.netcom.com> Marc Tillinghast (tilly@ix.netcom.com) wrote: > I have a struct that consists of 49 chars. Under vxWorks > sizeof(myStruct) returns 50. Under UNIX sizeof returns 49. > Does vxWorks consistently return the size + 1 of structs. > Any info would be appreciated. > > Thanks, > Marc Tillinghast (tilly@storm.com) Sizeof() returns the amount of memory allocated for the object rather than the amount of data that the object "contains". The result returned by sizeof() depends on the alignment used by the compiler. This can sometimes be affected by command-line options and/or #pragmas. It appears that your VxWorks compiler aligns structs on word (2-char) boundaries while your UNIX compiler aligns structs on char boundaries. A compiler that aligned structs on long-word (4-char) boundaries would return a sizeof() value of 52. Gary --------------------------- Newsgroups: comp.os.vxworks Subject: stripping objects for vxworks Date: 27 Jun 1995 17:50:10 GMT From: ahrens@informatik.hu-berlin.de (Klaus Ahrens) Organization: Humboldt-University Berlin Message-ID: <3spgci$92f@hahn.informatik.hu-berlin.de> Reply-To: ahrens@informatik.hu-berlin.de Dear vxworkers, we have to reduce the size of a 68k object file which is compiled with the WindRiver C++ gateway from huge generated C++ sources to fit into the RAM of our force board. The size could be reduced by stripping all symbols off (~4Mb less), but unfortunately this also strips the (in the running version) only needed entry point to start up the system. How to strip to an object that contains only some symbols or how to locate the entry point in a fully stripped loaded object ???????? Any suggestions ? Please reply by email ! Thanx in advance Klaus - -- _______________________________________________________________________ /| | | || Dr. K. Ahrens | Phone +49-30-20181 238 | || Institut fuer Informatik | Fax +49-30-20181 234 | || Humboldt-Universitaet | email ahrens@informatik.hu-berlin.de | || | http://www.informatik.hu-berlin.de/~ahrens| || +-------------------------------------------+ || Lindenstrasse 54a | +---------------------------------------+ | || 10117 Berlin | |___ Hindsight is an exact science. ___/| | || Germany | +--------------------------------------++ | ||___________________________|___________________________________________| |/___________________________/__________________________________________/ --------------------------- Newsgroups: comp.os.vxworks Subject: GCC w/C++ Date: 6 Jul 1995 17:46:35 GMT From: halpin@codem.com (Stephen E. Halpin) Organization: CODEM Systems, Inc. Message-ID: <3th7hr$1n6@regina.seh.codem.com> We are using VxWorks 5.1.1 (hopefully soon to be 5.2 when the BSP is validated) with the GCC compiler. We are interested in building GCC 2.7.0 with C++ support enabled. Before I waste a lot of time on this, I was wondering if the object modules that are ultimately produced from C++ sources are compatable with VxWorks, or does the extra linkage support of C++ require support that VxWorks doesnt have out of the box? Thanks, -Steve --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: 02 Jul 1995 05:34:17 GMT From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: References: In article guy@boi.hp.com (Guy Burnham) writes: :I'm questioning whether I'm loosing my marbles! When I ask the shell in :VxWorks 5.1.1 ...... : -> .7 <= .6 :It comes back with ...... :value = 1 While I don't know the reason for the above, which also happens on our VxWorks 5.1.1 setup, I can shed light on your second question. :So I wrote a little C hack .... : :int LTorEQ (float a, float b) : { : return (a <= b); : } : :.... and called it with the following result : :-> LTorEQ 0.7,0.6 :value = 1 = 0x1 The VxWorks shell passes floating point arguments as doubles, whereas your function expects floats. The result is system dependent, but on our MC68040 based system the values of a and b become 1.8 and 2.72008e+23 respectively. This makes the quoted return value understandable. To get your function to work as intended, simply change its declaration to: int LTorEQ (double a, double b) Martin Shepherd (mcs@astro.caltech.edu) --------------------------- Newsgroups: comp.os.vxworks Subject: Using the 167-bug ROMS with custom ROMS Date: Mon, 3 Jul 1995 05:13:49 GMT From: aa719@FreeNet.Carleton.CA (Dave Phillips) Organization: The National Capital FreeNet Message-ID: Reply-To: aa719@FreeNet.Carleton.CA (Dave Phillips) Sender: aa719@freenet3.carleton.ca (Dave Phillips) A short while back, somebody posted info on how to install and use the 167-bug (came with the MVME-167) ROMS and still use custom (boot) ROMS. Could someone please give me some info or a repost? Thanx. --------------------------- End of New-News digest ********************** Subject: Shared memory network problems Submitted-by manser@radstone.co.uk Fri Jul 7 04:01:02 1995 Submitted-by: Clive Manser I have a problem using VxWorks 5.1.1 and a shared memory network. This problem may have originally have been reported as SPR 2459 and was = reputedly fixed in 5.1.1 (maybe). If the shared memory backplane anchor region is filled with zeroes then = things seem to be OK. However if the shared memory anchor region has the 5th Longword= set to a non zero value (example 0x1000000) *and* the ready value longword is zero= (i.e. not ready / 0x87654321) then a bus error results. I have tracked this down to the card adding the fifth longword pointer = (?) to the anchor address to find the keepalive value. But the card appears to ignore the = ready value. I suspect that this is a library problem. Anybody else seen this or any suggestions ? Clive Manser manser@radstone.co.uk From manser@radstone.co.uk Fri Jul 7 04:01:02 1995 From: Clive Manser Date: Fri Jul 7 04:01:04 PDT 1995 Subject: Shared memory network problems I have a problem using VxWorks 5.1.1 and a shared memory network. This problem may have originally have been reported as SPR 2459 and was = reputedly fixed in 5.1.1 (maybe). If the shared memory backplane anchor region is filled with zeroes then = things seem to be OK. However if the shared memory anchor region has the 5th Longword= set to a non zero value (example 0x1000000) *and* the ready value longword is zero= (i.e. not ready / 0x87654321) then a bus error results. I have tracked this down to the card adding the fifth longword pointer = (?) to the anchor address to find the keepalive value. But the card appears to ignore the = ready value. I suspect that this is a library problem. Anybody else seen this or any suggestions ? Clive Manser manser@radstone.co.uk Subject: Re: Why is 0.5 <= 0.4 ???? Submitted-by Phillip.Shaffer@ae.ge.com Fri Jul 7 07:06:18 1995 Submitted-by: Phillip.Shaffer@ae.ge.com (Phil Shaffer) Based on experimentation with VxWorks 2.0.3b, the problem is with the "<=" function in the shell, which returns the inverted (i.e. wrong) value. Limited tests of <, >, ==, and >= all show these work, while <= always (in my tests) returns the opposite of what it should. The problem seems to have nothing to do with floating point representation (except in the case others have addressed of a function with float arguments that receives doubles). Phillip L. Shaffer phillip.shaffer@ae.ge.com GE Aircraft Engines, MS G57 1 Neumann Way Cincinnati, OH 45215 From Phillip.Shaffer@ae.ge.com Fri Jul 7 07:06:18 1995 From: Phillip.Shaffer@ae.ge.com (Phil Shaffer) Date: Fri Jul 7 07:06:21 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? Based on experimentation with VxWorks 2.0.3b, the problem is with the "<=" function in the shell, which returns the inverted (i.e. wrong) value. Limited tests of <, >, ==, and >= all show these work, while <= always (in my tests) returns the opposite of what it should. The problem seems to have nothing to do with floating point representation (except in the case others have addressed of a function with float arguments that receives doubles). Phillip L. Shaffer phillip.shaffer@ae.ge.com GE Aircraft Engines, MS G57 1 Neumann Way Cincinnati, OH 45215 Subject: Apology for repeated messages. Submitted-by ssellers@gic.gi.com Fri Jul 7 07:55:32 1995 Submitted-by: "Sellers, Scott (HT-MS)" I'm sorry for the repeated transmissions of my request for flash info. It wasn't intentional. We are having problems with our mail server (MS-Mail, grrrrrr!!!). I apologize for the inconveniences I caused. Rich Moore (borrowing Scott Sellers' account because my mail's broke!!!) From ssellers@gic.gi.com Fri Jul 7 07:55:32 1995 From: "Sellers, Scott (HT-MS)" Date: Fri Jul 7 07:55:34 PDT 1995 Subject: Apology for repeated messages. I'm sorry for the repeated transmissions of my request for flash info. It wasn't intentional. We are having problems with our mail server (MS-Mail, grrrrrr!!!). I apologize for the inconveniences I caused. Rich Moore (borrowing Scott Sellers' account because my mail's broke!!!) Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sat Jul 8 04:00:11 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 8 04:00:09 PDT 1995 Subject: Macintosh development Workstations for vxWorks? Subject: vxworks usage survey ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Macintosh development Workstations for vxWorks? Date: 7 Jul 1995 14:19:20 GMT From: johncoby@blkbox.com (John Cobarruvias) Organization: The Black Box, Houston, Tx (713) 480-2686 Message-ID: <3tjfp8$rlt@news.blkbox.com> Is there an enviorment on the Mac for developing applications for vxworks? - -- | _ | _ _ | | |\ |\ | |_ | |/ //| | | |/| |/ _| //| // | John R. Cobarruvias johncobyblkbox.com --------------------------- Newsgroups: comp.os.vxworks Subject: vxworks usage survey Date: 7 Jul 1995 19:20:38 GMT From: squeeby@tuna.hooked.net (Hwajin Bae) Organization: Peaceful Star Software, Menlo Park, CA Message-ID: <3tk1e6$6u9@its.hooked.net> i am gathering data for a private report on vxworks. if you are a vxworks user, please take a moment and drop me a note on what you are developing. please do NOT include any proprietary information. a general description of the project will suffice (e.g. purpose, intent, usage, overall architecture?). thanks for your cooperation. please use EMAIL: squeeby@hooked.net hwajin bae an ex-vxworks developer --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sat Jul 8 04:00:11 1995 From: daemon@csg.lbl.gov Date: Sat Jul 8 04:00:14 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 8 04:00:09 PDT 1995 Subject: Macintosh development Workstations for vxWorks? Subject: vxworks usage survey ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Macintosh development Workstations for vxWorks? Date: 7 Jul 1995 14:19:20 GMT From: johncoby@blkbox.com (John Cobarruvias) Organization: The Black Box, Houston, Tx (713) 480-2686 Message-ID: <3tjfp8$rlt@news.blkbox.com> Is there an enviorment on the Mac for developing applications for vxworks? - -- | _ | _ _ | | |\ |\ | |_ | |/ //| | | |/| |/ _| //| // | John R. Cobarruvias johncobyblkbox.com --------------------------- Newsgroups: comp.os.vxworks Subject: vxworks usage survey Date: 7 Jul 1995 19:20:38 GMT From: squeeby@tuna.hooked.net (Hwajin Bae) Organization: Peaceful Star Software, Menlo Park, CA Message-ID: <3tk1e6$6u9@its.hooked.net> i am gathering data for a private report on vxworks. if you are a vxworks user, please take a moment and drop me a note on what you are developing. please do NOT include any proprietary information. a general description of the project will suffice (e.g. purpose, intent, usage, overall architecture?). thanks for your cooperation. please use EMAIL: squeeby@hooked.net hwajin bae an ex-vxworks developer --------------------------- End of New-News digest ********************** Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sun Jul 9 04:00:12 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sun Jul 9 04:00:09 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? Subject: test Subject: Open House Subject: Re: GCC w/C++ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: 7 Jul 95 15:42:52 GMT From: support@wrs.com (WRS Support) Organization: Wind River Systems, Inc. Message-ID: References: Rob Steele (steele@telerobotics.jpl.nasa.gov) writes: > Relative to the recent e-mail on this curiousity, > I did the following user 5.1.1 > > > -> 1.5 <= 1.1 > value = 1 > -> 1.1 <= 1.5 > value = 0 > -> 0.6 <= 0.7 > value = 0 > -> 0.7 <= 0.6 > value = 1 > > Does anyone know why yet? The bug you are demonstrating is described by SPR #1517, "Shell expressions involving <= operator return the wrong value." It was "fixed" in 1992, but as it turns out, it is only fixed for integer arguments. Here's a work-around: -> (0.7 < 0.6) || (0.7 == 0.6) Thanks for the feedback! - ------------- WindRiver Technical Support --------------------------- Newsgroups: comp.os.vxworks Subject: test Date: 7 Jul 1995 23:15:50 GMT From: Kim Whitney Organization: WindRiver Message-ID: <3tkf76$el3@darya.wrs.com> aaaaaaa --------------------------- Newsgroups: comp.os.vxworks Subject: Open House Date: 7 Jul 1995 23:30:02 GMT From: Kim Whitney Organization: WindRiver Message-ID: <3tkg1q$el3@darya.wrs.com> SOFTWARE ENGINEERS--- PRODUCT MARKETING---CUSTOMER SUPPORT---SYSTEMS ADMINISTRATION--TECHNICAL INSTRUCTOR ITS YOUR CAREER----WHAT'S IMPORTANT TO YOU? * GROWTH----- one of CALIFORNIA'S FASTEST GROWING COMPANIES. * THE MARKET LEADER---- THE WORLD'S LEADER IN EMBEDDED DEVELOPMENT TOOLS AND OPERATING SYSTEMS. *GREAT ENVIRONMENT---STOCK OPTIONS, SABBATICAL, 401K,FULL BENEFITS. *GREAT LOCATION---PARK LIKE ALAMEDA SETTING,ONE BLOCK FROM MARINA, 5 MIN. TO THE BEACH. *PERSONAL GROWTH AND RESPONSIBILITY--- 35% GROWTH FUELING MAJOR INVESTMENT IN ENGINEERING -MARKETING. * TEAM---PROFESSIONAL, SMART, DYNAMIC OPEN HOUSE------OPEN HOUSE-------OPEN HOUSE SNACKS AND REFRESHMENTS TUESDAY JULY 25 AND THURSDAY JULY 27 5:30 PM TO 8:00 PM IDEAL BACKGROUNDS; CS/EE degrees; UNIX--Windows--Networking--GUIs--- embedded-real time.Strong track record of proven performance. Recent graduates to heavily experienced. GET A JUMP START--SEND YOUR RESUME SOON OR BRING IT WITH YOU. OUR ENGINEERS AND MARKETING STAFF WILL BE READY TO TALK WITH YOU. WIND RIVER SYSTEMS, 1010 Atlantic Ave, Alameda, CA 94507 Attn.; Kim Whitney HOW TO GET THERE--From the South Bay: Take 880 North. Exit at Broadway-Alameda exit. Turn right on Broadway. Turn right on 7th street. Turn right on Webster. Exit at Constitution Ave. Go to Atlantic Ave, make left. From SF: Take Bay Bridge to 980. Exit at 11th/12th street. Turn left on 5th. Follow signs to Alameda. Exit at Constitution. Go to Atlantic Ave, make left. Access our home page on the WEB http://www.wrs.com for more complete job descriptions and general WRS information. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GCC w/C++ Date: 8 Jul 1995 13:38:23 -0600 From: egott@unm.edu (Eric Gottlieb) Organization: University of New Mexico, Albuquerque Message-ID: <3tmmrf$t2m@callisto.unm.edu> References: <3th7hr$1n6@regina.seh.codem.com> In article <3th7hr$1n6@regina.seh.codem.com>, Stephen E. Halpin wrote: >We are using VxWorks 5.1.1 (hopefully soon to be 5.2 when >the BSP is validated) with the GCC compiler. We are interested >in building GCC 2.7.0 with C++ support enabled. Before I waste >a lot of time on this, I was wondering if the object modules >that are ultimately produced from C++ sources are compatable >with VxWorks, or does the extra linkage support of C++ require >support that VxWorks doesnt have out of the box? > >Thanks, -Steve > I'm far from knowledgeable about VxWorks, c++, or g++; that said, I've played a bit with compiling a g++ 2.7.0 cross compiler for VxWorks 5.1.1 (m68k) that runs under Irix 5.3. It seems to work for simple c++ programs, but is lacking the sort of linker support I think you are asking about. For example, static objects do not seem to get constructed. I'd be very interested in hearing about other people's experiences, especially if they have been able to build more functional versions of g++ for VxWorks. -- Eric Gottlieb (egott@unm.edu) - -- - -- Eric Gottlieb (egott@unm.edu) (http://www.unm.edu/~egott/finger.html) -- --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sun Jul 9 04:00:12 1995 From: daemon@csg.lbl.gov Date: Sun Jul 9 04:00:15 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 9 04:00:09 PDT 1995 Subject: Re: Why is 0.5 <= 0.4 ???? Subject: test Subject: Open House Subject: Re: GCC w/C++ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Why is 0.5 <= 0.4 ???? Date: 7 Jul 95 15:42:52 GMT From: support@wrs.com (WRS Support) Organization: Wind River Systems, Inc. Message-ID: References: Rob Steele (steele@telerobotics.jpl.nasa.gov) writes: > Relative to the recent e-mail on this curiousity, > I did the following user 5.1.1 > > > -> 1.5 <= 1.1 > value = 1 > -> 1.1 <= 1.5 > value = 0 > -> 0.6 <= 0.7 > value = 0 > -> 0.7 <= 0.6 > value = 1 > > Does anyone know why yet? The bug you are demonstrating is described by SPR #1517, "Shell expressions involving <= operator return the wrong value." It was "fixed" in 1992, but as it turns out, it is only fixed for integer arguments. Here's a work-around: -> (0.7 < 0.6) || (0.7 == 0.6) Thanks for the feedback! - ------------- WindRiver Technical Support --------------------------- Newsgroups: comp.os.vxworks Subject: test Date: 7 Jul 1995 23:15:50 GMT From: Kim Whitney Organization: WindRiver Message-ID: <3tkf76$el3@darya.wrs.com> aaaaaaa --------------------------- Newsgroups: comp.os.vxworks Subject: Open House Date: 7 Jul 1995 23:30:02 GMT From: Kim Whitney Organization: WindRiver Message-ID: <3tkg1q$el3@darya.wrs.com> SOFTWARE ENGINEERS--- PRODUCT MARKETING---CUSTOMER SUPPORT---SYSTEMS ADMINISTRATION--TECHNICAL INSTRUCTOR ITS YOUR CAREER----WHAT'S IMPORTANT TO YOU? * GROWTH----- one of CALIFORNIA'S FASTEST GROWING COMPANIES. * THE MARKET LEADER---- THE WORLD'S LEADER IN EMBEDDED DEVELOPMENT TOOLS AND OPERATING SYSTEMS. *GREAT ENVIRONMENT---STOCK OPTIONS, SABBATICAL, 401K,FULL BENEFITS. *GREAT LOCATION---PARK LIKE ALAMEDA SETTING,ONE BLOCK FROM MARINA, 5 MIN. TO THE BEACH. *PERSONAL GROWTH AND RESPONSIBILITY--- 35% GROWTH FUELING MAJOR INVESTMENT IN ENGINEERING -MARKETING. * TEAM---PROFESSIONAL, SMART, DYNAMIC OPEN HOUSE------OPEN HOUSE-------OPEN HOUSE SNACKS AND REFRESHMENTS TUESDAY JULY 25 AND THURSDAY JULY 27 5:30 PM TO 8:00 PM IDEAL BACKGROUNDS; CS/EE degrees; UNIX--Windows--Networking--GUIs--- embedded-real time.Strong track record of proven performance. Recent graduates to heavily experienced. GET A JUMP START--SEND YOUR RESUME SOON OR BRING IT WITH YOU. OUR ENGINEERS AND MARKETING STAFF WILL BE READY TO TALK WITH YOU. WIND RIVER SYSTEMS, 1010 Atlantic Ave, Alameda, CA 94507 Attn.; Kim Whitney HOW TO GET THERE--From the South Bay: Take 880 North. Exit at Broadway-Alameda exit. Turn right on Broadway. Turn right on 7th street. Turn right on Webster. Exit at Constitution Ave. Go to Atlantic Ave, make left. From SF: Take Bay Bridge to 980. Exit at 11th/12th street. Turn left on 5th. Follow signs to Alameda. Exit at Constitution. Go to Atlantic Ave, make left. Access our home page on the WEB http://www.wrs.com for more complete job descriptions and general WRS information. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GCC w/C++ Date: 8 Jul 1995 13:38:23 -0600 From: egott@unm.edu (Eric Gottlieb) Organization: University of New Mexico, Albuquerque Message-ID: <3tmmrf$t2m@callisto.unm.edu> References: <3th7hr$1n6@regina.seh.codem.com> In article <3th7hr$1n6@regina.seh.codem.com>, Stephen E. Halpin wrote: >We are using VxWorks 5.1.1 (hopefully soon to be 5.2 when >the BSP is validated) with the GCC compiler. We are interested >in building GCC 2.7.0 with C++ support enabled. Before I waste >a lot of time on this, I was wondering if the object modules >that are ultimately produced from C++ sources are compatable >with VxWorks, or does the extra linkage support of C++ require >support that VxWorks doesnt have out of the box? > >Thanks, -Steve > I'm far from knowledgeable about VxWorks, c++, or g++; that said, I've played a bit with compiling a g++ 2.7.0 cross compiler for VxWorks 5.1.1 (m68k) that runs under Irix 5.3. It seems to work for simple c++ programs, but is lacking the sort of linker support I think you are asking about. For example, static objects do not seem to get constructed. I'd be very interested in hearing about other people's experiences, especially if they have been able to build more functional versions of g++ for VxWorks. -- Eric Gottlieb (egott@unm.edu) - -- - -- Eric Gottlieb (egott@unm.edu) (http://www.unm.edu/~egott/finger.html) -- --------------------------- End of New-News digest ********************** Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Mon Jul 10 04:00:21 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Mon Jul 10 04:00:18 PDT 1995 Subject: pty and shell ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: pty and shell Date: 10 Jul 1995 08:25:54 GMT From: mcw@aus.hp.com (M C Wong) Organization: HP Australian Telecom Operation Message-ID: <3tqo6i$egb@hpscit.sc.hp.com> Hi, I want to experiment with pty under vxWorks, and will like to write a simple `script' program to exercise the pty stuff. The only thing that bothers me is that, will I be able to taskSpawn another shell ? Also, does taskSpawn replace the UNIX fork( ), exec( ) behaviour ? If not, how can I achieve that ? Regards. --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Mon Jul 10 04:00:21 1995 From: daemon@csg.lbl.gov Date: Mon Jul 10 04:00:24 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 10 04:00:18 PDT 1995 Subject: pty and shell ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: pty and shell Date: 10 Jul 1995 08:25:54 GMT From: mcw@aus.hp.com (M C Wong) Organization: HP Australian Telecom Operation Message-ID: <3tqo6i$egb@hpscit.sc.hp.com> Hi, I want to experiment with pty under vxWorks, and will like to write a simple `script' program to exercise the pty stuff. The only thing that bothers me is that, will I be able to taskSpawn another shell ? Also, does taskSpawn replace the UNIX fork( ), exec( ) behaviour ? If not, how can I achieve that ? Regards. --------------------------- End of New-News digest ********************** Subject: use of taskRestart() Submitted-by rich@immrc.eng.mcmaster.ca Mon Jul 10 08:56:18 1995 Submitted-by: rich@immrc.eng.mcmaster.ca Dear Vxworkers, I am trying to safely restart a task that has already initialised device and TCP/IP socket interfaces. I would like to avoid reinitialising everything for each time I disable/enable the task. Does anyone out there have any info on using the lib routine taskRestart()? The documentation says that it will first `terminate` my task, and will then restart. Does terminate here refer to a forced termination? What if I am blocked on a socket? What state is the socket left in (ie: are buffers flushed?)? Any tips here would be appreciated... Thanks in advance, Richard Teltz, Intelligent Machines and Manufacturing Research Centre McMaster University, Hamilton, Ontario, Canada. From rich@immrc.eng.mcmaster.ca Mon Jul 10 08:56:18 1995 From: rich@immrc.eng.mcmaster.ca Date: Mon Jul 10 08:56:21 PDT 1995 Subject: use of taskRestart() Dear Vxworkers, I am trying to safely restart a task that has already initialised device and TCP/IP socket interfaces. I would like to avoid reinitialising everything for each time I disable/enable the task. Does anyone out there have any info on using the lib routine taskRestart()? The documentation says that it will first `terminate` my task, and will then restart. Does terminate here refer to a forced termination? What if I am blocked on a socket? What state is the socket left in (ie: are buffers flushed?)? Any tips here would be appreciated... Thanks in advance, Richard Teltz, Intelligent Machines and Manufacturing Research Centre McMaster University, Hamilton, Ontario, Canada. Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Tue Jul 11 04:00:23 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Tue Jul 11 04:00:19 PDT 1995 Subject: Re: Will WRS support Linux? Subject: Re: GCC w/C++ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Will WRS support Linux? Date: Mon, 10 Jul 1995 23:04:55 GMT From: gerlach@netcom.com (Matthew H. Gerlach) Organization: Netcom Online Communications Services (408-241-9760 login: guest) Message-ID: References: <3ts1l3$7ch@news.greatbasin.net> Sender: gerlach@netcom18.netcom.com In article <3ts1l3$7ch@news.greatbasin.net> mike@ringo.reno.nv.us writes: > > >Since my Pentium running linux kicks the pants off my Sparc 5 >in performance at a MUCH lower price, it sure would be nice if WRS >distributed a Linux version of vxworks. Will WRS ever support Linux??? > > If Lynx were supported that would be my first choice of Host Platforms as well. That being said, one does not necessarily need WRS support to get Linux to be VxWorks Host. All the host needs is the basic IP protocols, which Lynx most certainly has, and it needs the ability to build gnu cros-compilers, which is certainly does. In theory one would just need to configure the gnu tools for HOST=Lynx and TARGET=m68k-vxworks or something like that. Rumor has it that this variant of the gnu tools needs to hacks to get working. A gentlemen sent my the hacks, but I have not used them, and they may not be necessary anymore. Matthew H. Gerlach - -- ******************************************************* * Gerlach Computer Consulting * * * * Real-Time Embedded Systems * * Networking Protocols * * Software Development Environments * * Network and Sun System Administration * * * * gerlach@netcom.com * * * ******************************************************* --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GCC w/C++ Date: 11 Jul 1995 11:09:59 +1000 From: rjl@f111.iassf.easams.com.au (Rohan LENARD) Organization: EASAMS (Australia) Pty Ltd Message-ID: <3tsj17$8ji@f111.iassf.easams.com.au> References: <3th7hr$1n6@regina.seh.codem.com> <3tmmrf$t2m@callisto.unm.edu> In article <3tmmrf$t2m@callisto.unm.edu>, Eric Gottlieb wrote: : :I'm far from knowledgeable about VxWorks, c++, or g++; that :said, I've played a bit with compiling a g++ 2.7.0 cross :compiler for VxWorks 5.1.1 (m68k) that runs under Irix 5.3. :It seems to work for simple c++ programs, but is lacking the :sort of linker support I think you are asking about. For :example, static objects do not seem to get constructed. I'd :be very interested in hearing about other people's experiences, :especially if they have been able to build more functional :versions of g++ for VxWorks. You need to use the GNU linker (not the crappy one you get with VxWorks). Last time I checked the linker from binutils 2.5.2 worked. It's part of the binutils package available at the following FTP sites - Here is a list of anonymous FTP archive sites for GNU software. ASIA: ftp.cs.titech.ac.jp, utsun.s.u-tokyo.ac.jp:/ftpsync/prep, cair.kaist.ac.kr:/pub/gnu, ftp.nectec.or.th:/pub/mirrors/gnu AUSTRALIA: archie.oz.au:/gnu (archie.oz or archie.oz.au for ACSnet) AFRICA: ftp.sun.ac.za:/pub/gnu MIDDLE-EAST: ftp.technion.ac.il:/pub/unsupported/gnu EUROPE: irisa.irisa.fr:/pub/gnu, ftp.univ-lyon1.fr:pub/gnu, ftp.mcc.ac.uk, unix.hensa.ac.uk:/pub/uunet/systems/gnu, ftp.denet.dk, src.doc.ic.ac.uk:/gnu, ftp.eunet.ch, nic.switch.ch:/mirror/gnu, ftp.informatik.rwth-aachen.de:/pub/gnu, ftp.informatik.tu-muenchen.de, ftp.win.tue.nl:/pub/gnu, ftp.funet.fi:/pub/gnu, ftp.stacken.kth.se, isy.liu.se, ftp.luth.se:/pub/unix/gnu, ftp.sunet.se:/pub/gnu, archive.eu.net SOUTH AMERICA: ftp.unicamp.br:/pub/gnu WESTERN CANADA: ftp.cs.ubc.ca:/mirror2/gnu USA: wuarchive.wustl.edu:/systems/gnu, labrea.stanford.edu, ftp.digex.net:/pub/gnu, ftp.kpc.com:/pub/mirror/gnu, f.ms.uky.edu:/pub3/gnu, jaguar.utah.edu:/gnustuff, ftp.hawaii.edu:/mirrors/gnu, vixen.cso.uiuc.edu:/gnu, mrcnext.cso.uiuc.edu:/pub/gnu, ftp.cs.columbia.edu:/archives/gnu/prep, col.hp.com:/mirrors/gnu, gatekeeper.dec.com:/pub/GNU, ftp.uu.net:/systems/gnu The "official site" is prep.ai.mit.edu, but your transfer will probably go faster if you use one of the above machines. Regards, Rohan - -- - ---------------------------------------------------------------------------- rjl@iassf.easams.com.au | All quotes can be attributed to my automated quote Rohan Lenard | writing tool. Yours for just $19.95; and if you +61-2-367-4555 | call now you'll get a free set of steak knives ... --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Tue Jul 11 04:00:23 1995 From: daemon@csg.lbl.gov Date: Tue Jul 11 04:00:26 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 11 04:00:19 PDT 1995 Subject: Re: Will WRS support Linux? Subject: Re: GCC w/C++ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Will WRS support Linux? Date: Mon, 10 Jul 1995 23:04:55 GMT From: gerlach@netcom.com (Matthew H. Gerlach) Organization: Netcom Online Communications Services (408-241-9760 login: guest) Message-ID: References: <3ts1l3$7ch@news.greatbasin.net> Sender: gerlach@netcom18.netcom.com In article <3ts1l3$7ch@news.greatbasin.net> mike@ringo.reno.nv.us writes: > > >Since my Pentium running linux kicks the pants off my Sparc 5 >in performance at a MUCH lower price, it sure would be nice if WRS >distributed a Linux version of vxworks. Will WRS ever support Linux??? > > If Lynx were supported that would be my first choice of Host Platforms as well. That being said, one does not necessarily need WRS support to get Linux to be VxWorks Host. All the host needs is the basic IP protocols, which Lynx most certainly has, and it needs the ability to build gnu cros-compilers, which is certainly does. In theory one would just need to configure the gnu tools for HOST=Lynx and TARGET=m68k-vxworks or something like that. Rumor has it that this variant of the gnu tools needs to hacks to get working. A gentlemen sent my the hacks, but I have not used them, and they may not be necessary anymore. Matthew H. Gerlach - -- ******************************************************* * Gerlach Computer Consulting * * * * Real-Time Embedded Systems * * Networking Protocols * * Software Development Environments * * Network and Sun System Administration * * * * gerlach@netcom.com * * * ******************************************************* --------------------------- Newsgroups: comp.os.vxworks Subject: Re: GCC w/C++ Date: 11 Jul 1995 11:09:59 +1000 From: rjl@f111.iassf.easams.com.au (Rohan LENARD) Organization: EASAMS (Australia) Pty Ltd Message-ID: <3tsj17$8ji@f111.iassf.easams.com.au> References: <3th7hr$1n6@regina.seh.codem.com> <3tmmrf$t2m@callisto.unm.edu> In article <3tmmrf$t2m@callisto.unm.edu>, Eric Gottlieb wrote: : :I'm far from knowledgeable about VxWorks, c++, or g++; that :said, I've played a bit with compiling a g++ 2.7.0 cross :compiler for VxWorks 5.1.1 (m68k) that runs under Irix 5.3. :It seems to work for simple c++ programs, but is lacking the :sort of linker support I think you are asking about. For :example, static objects do not seem to get constructed. I'd :be very interested in hearing about other people's experiences, :especially if they have been able to build more functional :versions of g++ for VxWorks. You need to use the GNU linker (not the crappy one you get with VxWorks). Last time I checked the linker from binutils 2.5.2 worked. It's part of the binutils package available at the following FTP sites - Here is a list of anonymous FTP archive sites for GNU software. ASIA: ftp.cs.titech.ac.jp, utsun.s.u-tokyo.ac.jp:/ftpsync/prep, cair.kaist.ac.kr:/pub/gnu, ftp.nectec.or.th:/pub/mirrors/gnu AUSTRALIA: archie.oz.au:/gnu (archie.oz or archie.oz.au for ACSnet) AFRICA: ftp.sun.ac.za:/pub/gnu MIDDLE-EAST: ftp.technion.ac.il:/pub/unsupported/gnu EUROPE: irisa.irisa.fr:/pub/gnu, ftp.univ-lyon1.fr:pub/gnu, ftp.mcc.ac.uk, unix.hensa.ac.uk:/pub/uunet/systems/gnu, ftp.denet.dk, src.doc.ic.ac.uk:/gnu, ftp.eunet.ch, nic.switch.ch:/mirror/gnu, ftp.informatik.rwth-aachen.de:/pub/gnu, ftp.informatik.tu-muenchen.de, ftp.win.tue.nl:/pub/gnu, ftp.funet.fi:/pub/gnu, ftp.stacken.kth.se, isy.liu.se, ftp.luth.se:/pub/unix/gnu, ftp.sunet.se:/pub/gnu, archive.eu.net SOUTH AMERICA: ftp.unicamp.br:/pub/gnu WESTERN CANADA: ftp.cs.ubc.ca:/mirror2/gnu USA: wuarchive.wustl.edu:/systems/gnu, labrea.stanford.edu, ftp.digex.net:/pub/gnu, ftp.kpc.com:/pub/mirror/gnu, f.ms.uky.edu:/pub3/gnu, jaguar.utah.edu:/gnustuff, ftp.hawaii.edu:/mirrors/gnu, vixen.cso.uiuc.edu:/gnu, mrcnext.cso.uiuc.edu:/pub/gnu, ftp.cs.columbia.edu:/archives/gnu/prep, col.hp.com:/mirrors/gnu, gatekeeper.dec.com:/pub/GNU, ftp.uu.net:/systems/gnu The "official site" is prep.ai.mit.edu, but your transfer will probably go faster if you use one of the above machines. Regards, Rohan - -- - ---------------------------------------------------------------------------- rjl@iassf.easams.com.au | All quotes can be attributed to my automated quote Rohan Lenard | writing tool. Yours for just $19.95; and if you +61-2-367-4555 | call now you'll get a free set of steak knives ... --------------------------- End of New-News digest ********************** Subject: Motorola MVME167 wait states Submitted-by Phillip.Shaffer@ae.ge.com Tue Jul 11 13:30:28 1995 Submitted-by: Phillip.Shaffer@ae.ge.com (Phil Shaffer) Does anyone know how many wait states are used by the 68040 accessing RAM and EPROM on a 25-MHz MVME167? Thanks. Phillip L. Shaffer phillip.shaffer@ae.ge.com GE Aircraft Engines, MS G57 1 Neumann Way Cincinnati, OH 45215 From Phillip.Shaffer@ae.ge.com Tue Jul 11 13:30:28 1995 From: Phillip.Shaffer@ae.ge.com (Phil Shaffer) Date: Tue Jul 11 13:30:31 PDT 1995 Subject: Motorola MVME167 wait states Does anyone know how many wait states are used by the 68040 accessing RAM and EPROM on a 25-MHz MVME167? Thanks. Phillip L. Shaffer phillip.shaffer@ae.ge.com GE Aircraft Engines, MS G57 1 Neumann Way Cincinnati, OH 45215 Subject: Software Engineer Position Submitted-by steele@telerobotics.jpl.nasa.gov Tue Jul 11 15:57:19 1995 Submitted-by: Rob Steele The Rover Technology Research Program at the Jet Propulsion Laboratory, California Instite of Technlogy, is looking for a real-time systems programmer. This is a research program to build prototype Mars rovers and develop technology and capabilities for upcoming missions to Mars. For instance, our current efforts are in building Rocky 7, a 10kg fully autonomous, rough terrain mobile robot. Rocky 7 has a 3U VME chasis with 68040, A/D, DIO, framegrabber, peripheral LM629 motor control, 13 actuators, 6 cameras, etc. Please see http://robotics.jpl.nasa.gov/tasks/rovertech/homepage.html for more. Required experience/knowledge: VxWorks, device driver code development, real-time C language programming Desired experience/knowledge: RTI Control Shell development environment, C++, X Programming, Control Theory, Robotics Please send resumes to: Dr. Homayoun Seraji Mail Stop 198-219 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena CA 91109 seraji@jpl.nasa.gov From steele@telerobotics.jpl.nasa.gov Tue Jul 11 15:57:19 1995 From: Rob Steele Date: Tue Jul 11 15:57:22 PDT 1995 Subject: Software Engineer Position The Rover Technology Research Program at the Jet Propulsion Laboratory, California Instite of Technlogy, is looking for a real-time systems programmer. This is a research program to build prototype Mars rovers and develop technology and capabilities for upcoming missions to Mars. For instance, our current efforts are in building Rocky 7, a 10kg fully autonomous, rough terrain mobile robot. Rocky 7 has a 3U VME chasis with 68040, A/D, DIO, framegrabber, peripheral LM629 motor control, 13 actuators, 6 cameras, etc. Please see http://robotics.jpl.nasa.gov/tasks/rovertech/homepage.html for more. Required experience/knowledge: VxWorks, device driver code development, real-time C language programming Desired experience/knowledge: RTI Control Shell development environment, C++, X Programming, Control Theory, Robotics Please send resumes to: Dr. Homayoun Seraji Mail Stop 198-219 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena CA 91109 seraji@jpl.nasa.gov Subject: Parallel port driver for mv147? Submitted-by key@diasonx.com Tue Jul 11 16:54:04 1995 Submitted-by: key@diasonx.com (Charles Key) We want to drive an HP 540C Deskjet printer through the parallel port of an mv147. Any code, suggestions, or other information would be much appreciated. ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com From key@diasonx.com Tue Jul 11 16:54:04 1995 From: key@diasonx.com (Charles Key) Date: Tue Jul 11 16:54:07 PDT 1995 Subject: Parallel port driver for mv147? We want to drive an HP 540C Deskjet printer through the parallel port of an mv147. Any code, suggestions, or other information would be much appreciated. ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Wed Jul 12 04:00:21 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Wed Jul 12 04:00:18 PDT 1995 Subject: gzip/compress for vxworks ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: gzip/compress for vxworks ? Date: 12 Jul 1995 06:09:31 GMT From: mcw@aus.hp.com (M C Wong) Organization: HP Australian Telecom Operation Message-ID: <3tvour$2ut@hpscit.sc.hp.com> Hi, Has anyone ported any compression/descompression utitilies to vxworks, such as gzip and compress ? --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Wed Jul 12 04:00:21 1995 From: daemon@csg.lbl.gov Date: Wed Jul 12 04:00:25 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 12 04:00:18 PDT 1995 Subject: gzip/compress for vxworks ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: gzip/compress for vxworks ? Date: 12 Jul 1995 06:09:31 GMT From: mcw@aus.hp.com (M C Wong) Organization: HP Australian Telecom Operation Message-ID: <3tvour$2ut@hpscit.sc.hp.com> Hi, Has anyone ported any compression/descompression utitilies to vxworks, such as gzip and compress ? --------------------------- End of New-News digest ********************** Subject: vi running on VxWorks? Submitted-by revbob@eight-ball.hv.boeing.com Wed Jul 12 08:18:34 1995 Submitted-by: revbob@eight-ball.hv.boeing.com (Bob Crispen) I thought I'd surely seen this somewhere, but can't find it in my archives -- has anyone ported a simple text editor, similar to vi, so that it will run on a VxWorks system? The challenge is to edit a file on the SCSI drive, and while it's always possible to copy it to an NFS mounted partition, edit it on the workstation, and copy it back, it would save some money if we could edit the file in situ on the VxWorks terminal. Has anyone done this? Bob Crispen revbob@eight-ball.hv.boeing.com From revbob@eight-ball.hv.boeing.com Wed Jul 12 08:18:34 1995 From: revbob@eight-ball.hv.boeing.com (Bob Crispen) Date: Wed Jul 12 08:18:37 PDT 1995 Subject: vi running on VxWorks? I thought I'd surely seen this somewhere, but can't find it in my archives -- has anyone ported a simple text editor, similar to vi, so that it will run on a VxWorks system? The challenge is to edit a file on the SCSI drive, and while it's always possible to copy it to an NFS mounted partition, edit it on the workstation, and copy it back, it would save some money if we could edit the file in situ on the VxWorks terminal. Has anyone done this? Bob Crispen revbob@eight-ball.hv.boeing.com Subject: Connection between VxWorks and PLC Submitted-by rhin!wrsec.fr!joergb@lbl.gov Wed Jul 12 08:59:47 1995 Submitted-by: rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) Greetings. Has anybody realized a connection between a VxWorks target and Siemens S5 PLC? Thanks for any info you are willing to share. Best Regards Joerg ************************************************************************* * Joerg Bertholdt Wind River Systems GmbH * * Field Application Engineer Freisinger Strasse 34 * * Email: joergb@wrsec.fr D-85737 Ismaning * * Tel: +49-89-96094942 Fax: +49-89-96094940 * ************************************************************************* From rhin!wrsec.fr!joergb@lbl.gov Wed Jul 12 08:59:47 1995 From: rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) Date: Wed Jul 12 08:59:49 PDT 1995 Subject: Connection between VxWorks and PLC Greetings. Has anybody realized a connection between a VxWorks target and Siemens S5 PLC? Thanks for any info you are willing to share. Best Regards Joerg ************************************************************************* * Joerg Bertholdt Wind River Systems GmbH * * Field Application Engineer Freisinger Strasse 34 * * Email: joergb@wrsec.fr D-85737 Ismaning * * Tel: +49-89-96094942 Fax: +49-89-96094940 * ************************************************************************* Subject: Re: unsubscribe Submitted-by schultzr@source.asset.com Wed Jul 12 13:08:39 1995 Submitted-by: schultzr@source.asset.com (Richard Schultz) Please unsubscribe me to the VxWorks exploder. Rhanks, Richard From schultzr@source.asset.com Wed Jul 12 13:08:39 1995 From: schultzr@source.asset.com (Richard Schultz) Date: Wed Jul 12 13:08:41 PDT 1995 Subject: Re: unsubscribe Please unsubscribe me to the VxWorks exploder. Rhanks, Richard Subject: Re: comp.os.vxworks newsdigest Submitted-by burt@xylan.com Wed Jul 12 15:43:37 1995 Submitted-by: burt@xylan.com (Burt Cyr) M C Wong writes: > Hi, > Has anyone ported any compression/descompression utitilies to > vxworks, such as gzip and compress ? I ported gzip-1.2.3 to our environment. Forget about the PD compress - it only will net you a little over 40% compression. gzip/gunzip will net you approximately 60-65%. The port was accomplished as follows: 1) I downloaded the gzip src from our local GNU archive site and copied it into two directories (meaning I now have two subdirs that have identical source code subtrees in them.) [I did this for several reasons: The gzip package produces one executable that becomes 3 different programs (gzip, gunzip, zcat); I really only want gzip to run on our host (a Sun) and gunzip to run out of ROM (our datacomm product - a LAN switch); therefore, I used two subdirectories - one for the host and one for the product since I was going to do a lot of stripping to conserve space in our product. (No need to have the compression code if we're onnly going to decompress.)] 2) Gzip is programmer friendly and the only thing I had to do to get it running on our Sun pretty much was minor configuration to the Makefile and tailor.h. 3) To get gunzip to run in our product was a little bit more tricky since gunzip wants to malloc() space and the vxWorks malloc package wants to include the vx world (which would defeat the gzip gains.) We'd compress our executable from ~2.5 MB down to ~950K, add the decompression code and vxworks mallocator, and this would cause it to reinflate above 1.3MB - which was the compression ratio we got with the vxWorks public-domain compress routine - a noop. So I wrote my own simple mallocator. I experimented with how the host side gunzip malloc() and free() routines worked and found that they produced and accordian effect (malloc a lot - 6K - then free a lot). Each set of mallocs and frees had to do with internal compression tables inserted by gzip. Over the course of our 2.5 MB file, it would malloc() approximately a MB, and free() the data. At any given time however, it only has approximately 8KB outstanding. 4) I then stripped out the compression code from the product subtree and ended up with a nice tight package (17K text 2.5K data + pool for mallocating from which is discarded after decompression). 5) Finally, I integrated it into the vxWorks.st_rom in our BSP - replacing the compress routines. I also wrote a simple gtest program to simulate what our product will do with the compressed image. This I also inserted into our vxWorks.st_rom recipe. It produces a report of the final image, mallocator performance and will cause the make of the .st_rom executable to abort if certain conditions are exceeded (such as overrunning the memory pool we have allocated.) I've heard others mention a desire for this but never responded until I reached a conviction that it's time to give back a little to the community what it's abundantly provided to our development effort. If anyone is interested in this, I'll unbundle it from our product and archive it at UCAR. I wasn't sure if this would be a wasted effort on my part or not. In any case - you can let me know directly or post to the news group. Burt Cyr ================================================================== v _ ____(_)___ Burt Cyr Xylan, Inc. _-_-__-_-_-_ burt@xylan.com 26679 Agoura Rd. Suite 100 --_-_--_-_- (818) 880-3500 x3533 Calabasas, CA 91302 ================================================================== From burt@xylan.com Wed Jul 12 15:43:37 1995 From: burt@xylan.com (Burt Cyr) Date: Wed Jul 12 15:43:40 PDT 1995 Subject: Re: comp.os.vxworks newsdigest M C Wong writes: > Hi, > Has anyone ported any compression/descompression utitilies to > vxworks, such as gzip and compress ? I ported gzip-1.2.3 to our environment. Forget about the PD compress - it only will net you a little over 40% compression. gzip/gunzip will net you approximately 60-65%. The port was accomplished as follows: 1) I downloaded the gzip src from our local GNU archive site and copied it into two directories (meaning I now have two subdirs that have identical source code subtrees in them.) [I did this for several reasons: The gzip package produces one executable that becomes 3 different programs (gzip, gunzip, zcat); I really only want gzip to run on our host (a Sun) and gunzip to run out of ROM (our datacomm product - a LAN switch); therefore, I used two subdirectories - one for the host and one for the product since I was going to do a lot of stripping to conserve space in our product. (No need to have the compression code if we're onnly going to decompress.)] 2) Gzip is programmer friendly and the only thing I had to do to get it running on our Sun pretty much was minor configuration to the Makefile and tailor.h. 3) To get gunzip to run in our product was a little bit more tricky since gunzip wants to malloc() space and the vxWorks malloc package wants to include the vx world (which would defeat the gzip gains.) We'd compress our executable from ~2.5 MB down to ~950K, add the decompression code and vxworks mallocator, and this would cause it to reinflate above 1.3MB - which was the compression ratio we got with the vxWorks public-domain compress routine - a noop. So I wrote my own simple mallocator. I experimented with how the host side gunzip malloc() and free() routines worked and found that they produced and accordian effect (malloc a lot - 6K - then free a lot). Each set of mallocs and frees had to do with internal compression tables inserted by gzip. Over the course of our 2.5 MB file, it would malloc() approximately a MB, and free() the data. At any given time however, it only has approximately 8KB outstanding. 4) I then stripped out the compression code from the product subtree and ended up with a nice tight package (17K text 2.5K data + pool for mallocating from which is discarded after decompression). 5) Finally, I integrated it into the vxWorks.st_rom in our BSP - replacing the compress routines. I also wrote a simple gtest program to simulate what our product will do with the compressed image. This I also inserted into our vxWorks.st_rom recipe. It produces a report of the final image, mallocator performance and will cause the make of the .st_rom executable to abort if certain conditions are exceeded (such as overrunning the memory pool we have allocated.) I've heard others mention a desire for this but never responded until I reached a conviction that it's time to give back a little to the community what it's abundantly provided to our development effort. If anyone is interested in this, I'll unbundle it from our product and archive it at UCAR. I wasn't sure if this would be a wasted effort on my part or not. In any case - you can let me know directly or post to the news group. Burt Cyr ================================================================== v _ ____(_)___ Burt Cyr Xylan, Inc. _-_-__-_-_-_ burt@xylan.com 26679 Agoura Rd. Suite 100 --_-_--_-_- (818) 880-3500 x3533 Calabasas, CA 91302 ================================================================== Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Thu Jul 13 04:00:25 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Thu Jul 13 04:00:20 PDT 1995 Subject: Need: UNIX heap management code similar to memPartAlloc/memPartFree Subject: Re: Connection between VxWorks and PLC ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Need: UNIX heap management code similar to memPartAlloc/memPartFree Date: 12 Jul 1995 17:45:33 GMT From: Shash Chatterjee Organization: Lockheed Martin Tactical Aircraft Systems Message-ID: <3u11nt$f3p@cliffy.lfwc.lockheed.com> I'd like to manage some unix (SYSV-IPC) shared memory segments with memory management routines like memPartAlloc and memPartFree. Are there any free sources available on the net anywhere? I've already looked at GNU malloc...but it doesn't seem to fit the bill: managing multiple user-defined regions. Thanks in advance, Shash - -- +-----------------------------------------------------------------+ + ChatterjeeS@lfwc.lockheed.com + + Sasvata Chatterjee | Electronic Systems Design & Integration + + "SHASH-WATA" | Lockheed Martin Tactical Aircraft Systems+ + (817) 763-1495 | Lockheed Boulevard, MZ2273 + + (817) 763-1495 (FAX) | Fort Worth, TX 76108 + +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Connection between VxWorks and PLC Date: 13 Jul 1995 10:21:26 +0100 From: cpg@jet.uk (Chris Gatcombe) Organization: JET Joint Undertaking Message-ID: References: <9507121454.AA04701@rhin> In <9507121454.AA04701@rhin> rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) writes: >Greetings. >Has anybody realized a connection between a VxWorks target and Siemens S5 PLC? >Thanks for any info you are willing to share. >Best Regards >Joerg I am using a Siemens S5-115U-942 with a CP524 comms module. On my VME system, I have an MVME167 (with inbuilt comms ports) and also an ASIO16, a dedicated 16 channel serial port board. As you will no doubt know, the PLC requires 9600 baud, even parity, 8 databits and 1 stop bit. The drivers for the MVME167 on-board comms do not allow for changing of parity, but the drivers for the ASIO16 do. One of our people here has hacked the driver supplied with the ASIO16 and also the MVME167 serial driver to suppport parity changes. I am currently trying to get the VME rack to talk to the PLC via a VxWorks program called Sprite/VxWorks. According to the manual, Sprite "...will allow an application program to communicate over a serial line with a Siemens 3964R communications processor or other communications device". Sprite has been ported to several operating systems, eg MSDOS, VAX, VxWorks and others by a company in the UK called Mentec. Their address is: Mentec International Limited, Industrial Systems Division, Mentec House, York Way, High Wycombe, Buckinghamshire HP12 3PY. Tel: +44 1494 472800 Fax: +44 1494 449256 The setup here is that the serial port(s) have to be setup to be 9600,e,8,1, then sprite has to be loaded and told which serial port to use. Your application prtogram to read/write the PLC datablocks then has to read/write message queues to access the datablock data. Hope this is useful. Chris. - -- Dr. Chris Gatcombe, contractor at JET Joint European Undertaking *********************** http://www.jet.uk ********************** ******** "Don't let the Sun go down on me" - Elton John ******** *** "Its the wrong trousers Gromit, and they've gone wrong!" *** =============================================================================== The above article is the personal view of the poster and should not be considered as an official comment from the JET Joint Undertaking =============================================================================== --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Thu Jul 13 04:00:25 1995 From: daemon@csg.lbl.gov Date: Thu Jul 13 04:00:28 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 13 04:00:20 PDT 1995 Subject: Need: UNIX heap management code similar to memPartAlloc/memPartFree Subject: Re: Connection between VxWorks and PLC ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Need: UNIX heap management code similar to memPartAlloc/memPartFree Date: 12 Jul 1995 17:45:33 GMT From: Shash Chatterjee Organization: Lockheed Martin Tactical Aircraft Systems Message-ID: <3u11nt$f3p@cliffy.lfwc.lockheed.com> I'd like to manage some unix (SYSV-IPC) shared memory segments with memory management routines like memPartAlloc and memPartFree. Are there any free sources available on the net anywhere? I've already looked at GNU malloc...but it doesn't seem to fit the bill: managing multiple user-defined regions. Thanks in advance, Shash - -- +-----------------------------------------------------------------+ + ChatterjeeS@lfwc.lockheed.com + + Sasvata Chatterjee | Electronic Systems Design & Integration + + "SHASH-WATA" | Lockheed Martin Tactical Aircraft Systems+ + (817) 763-1495 | Lockheed Boulevard, MZ2273 + + (817) 763-1495 (FAX) | Fort Worth, TX 76108 + +-----------------------------------------------------------------+ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Connection between VxWorks and PLC Date: 13 Jul 1995 10:21:26 +0100 From: cpg@jet.uk (Chris Gatcombe) Organization: JET Joint Undertaking Message-ID: References: <9507121454.AA04701@rhin> In <9507121454.AA04701@rhin> rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) writes: >Greetings. >Has anybody realized a connection between a VxWorks target and Siemens S5 PLC? >Thanks for any info you are willing to share. >Best Regards >Joerg I am using a Siemens S5-115U-942 with a CP524 comms module. On my VME system, I have an MVME167 (with inbuilt comms ports) and also an ASIO16, a dedicated 16 channel serial port board. As you will no doubt know, the PLC requires 9600 baud, even parity, 8 databits and 1 stop bit. The drivers for the MVME167 on-board comms do not allow for changing of parity, but the drivers for the ASIO16 do. One of our people here has hacked the driver supplied with the ASIO16 and also the MVME167 serial driver to suppport parity changes. I am currently trying to get the VME rack to talk to the PLC via a VxWorks program called Sprite/VxWorks. According to the manual, Sprite "...will allow an application program to communicate over a serial line with a Siemens 3964R communications processor or other communications device". Sprite has been ported to several operating systems, eg MSDOS, VAX, VxWorks and others by a company in the UK called Mentec. Their address is: Mentec International Limited, Industrial Systems Division, Mentec House, York Way, High Wycombe, Buckinghamshire HP12 3PY. Tel: +44 1494 472800 Fax: +44 1494 449256 The setup here is that the serial port(s) have to be setup to be 9600,e,8,1, then sprite has to be loaded and told which serial port to use. Your application prtogram to read/write the PLC datablocks then has to read/write message queues to access the datablock data. Hope this is useful. Chris. - -- Dr. Chris Gatcombe, contractor at JET Joint European Undertaking *********************** http://www.jet.uk ********************** ******** "Don't let the Sun go down on me" - Elton John ******** *** "Its the wrong trousers Gromit, and they've gone wrong!" *** =============================================================================== The above article is the personal view of the poster and should not be considered as an official comment from the JET Joint Undertaking =============================================================================== --------------------------- End of New-News digest ********************** Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Fri Jul 14 04:00:23 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Fri Jul 14 04:00:19 PDT 1995 Subject: Task memory protection required in VxWorks development environment.. ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Task memory protection required in VxWorks development environment.. Date: Thu, 13 Jul 1995 18:21:39 GMT From: joshisx@ntmtv.com (Suhas Joshi) Organization: Northern Telecom Inc., Mountain View Message-ID: Sender: news@ntmtv.com Hi, VxWorks does provide very little memory management and control. i.e. any task can write to any other task's memory and screw up the entire working system. This lack of control can be justified obviously as it is always by the argument of real time impact. However any new task can screw up the entire running system. In a development environment where lots of people are working to develop multiple features on the same product it becomes essential to find out what task is currupting the memory. IS THERE ANY WAY OF INTRODUCING MEMORY CONTROL AT THE OS LEVEL IN THE DEVELOPMENT ENVIRONMENT. This will greately reduce the debugging time in the development environment. The run time system or the customer supported system need not have this control and hence no real time impact in a true real life system. Is any one aware of such a product from Wind River Systems. Any one who has been through this can you explain. Regards - -Suhas --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Fri Jul 14 04:00:23 1995 From: daemon@csg.lbl.gov Date: Fri Jul 14 04:00:26 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 14 04:00:19 PDT 1995 Subject: Task memory protection required in VxWorks development environment.. ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Task memory protection required in VxWorks development environment.. Date: Thu, 13 Jul 1995 18:21:39 GMT From: joshisx@ntmtv.com (Suhas Joshi) Organization: Northern Telecom Inc., Mountain View Message-ID: Sender: news@ntmtv.com Hi, VxWorks does provide very little memory management and control. i.e. any task can write to any other task's memory and screw up the entire working system. This lack of control can be justified obviously as it is always by the argument of real time impact. However any new task can screw up the entire running system. In a development environment where lots of people are working to develop multiple features on the same product it becomes essential to find out what task is currupting the memory. IS THERE ANY WAY OF INTRODUCING MEMORY CONTROL AT THE OS LEVEL IN THE DEVELOPMENT ENVIRONMENT. This will greately reduce the debugging time in the development environment. The run time system or the customer supported system need not have this control and hence no real time impact in a true real life system. Is any one aware of such a product from Wind River Systems. Any one who has been through this can you explain. Regards - -Suhas --------------------------- End of New-News digest ********************** Subject: Re: Task memory protection required in VxWorks development environment Submitted-by prb@aplexus.jhuapl.edu Fri Jul 14 05:19:20 1995 Submitted-by: prb@aplexus.jhuapl.edu (Paul R. Bade) (Suhas Joshi) Wrote: > > VxWorks does provide very little memory management > and control. i.e. any task can write to any other > task's memory and screw up the entire working system. > > This lack of control can be justified obviously as > it is always by the argument of real time impact. > > However any new task can screw up the entire running > system. > > In a development environment where lots of people are > working to develop multiple features on the same product > it becomes essential to find out what task is currupting > the memory. > > IS THERE ANY WAY OF INTRODUCING MEMORY CONTROL AT THE OS > LEVEL IN THE DEVELOPMENT ENVIRONMENT. > Have you looked at the vmLib option from Wind River? Paul From prb@aplexus.jhuapl.edu Fri Jul 14 05:19:20 1995 From: prb@aplexus.jhuapl.edu (Paul R. Bade) Date: Fri Jul 14 05:19:22 PDT 1995 Subject: Re: Task memory protection required in VxWorks development environment (Suhas Joshi) Wrote: > > VxWorks does provide very little memory management > and control. i.e. any task can write to any other > task's memory and screw up the entire working system. > > This lack of control can be justified obviously as > it is always by the argument of real time impact. > > However any new task can screw up the entire running > system. > > In a development environment where lots of people are > working to develop multiple features on the same product > it becomes essential to find out what task is currupting > the memory. > > IS THERE ANY WAY OF INTRODUCING MEMORY CONTROL AT THE OS > LEVEL IN THE DEVELOPMENT ENVIRONMENT. > Have you looked at the vmLib option from Wind River? Paul Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sat Jul 15 04:00:22 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 15 04:00:19 PDT 1995 Subject: Re: Task memory protection required in VxWorks development environment.. Subject: Re: pty and shell ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Task memory protection required in VxWorks development environment.. Date: 14 Jul 1995 15:48:56 GMT From: ericr@vcd.hp.com (Eric Ross) Organization: Hewlett-Packard Message-ID: <3u63l8$kct@news.vcd.hp.com> References: Wind River does have a "Vm" product that gets you part way to where you want to be. It allows private memory contexts but these contexts are created manually by each task that needs them. If you need full protection, Wind River does not have anything(at least that they have announced) that gives you that capability. - -- Eric Ross | Hewlett-Packard Company, VPR ericr@vcd.hp.com | Vancouver, WA, USA (360)212-2372 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: pty and shell Date: 14 Jul 1995 22:33:14 GMT From: vanandel@stout.atd.ucar.edu (Joe Van Andel) Organization: National Center for Atmospheric Research Message-ID: <3u6rba$1uc@ncar.ucar.edu> Sender: vanandel@stay.atd.ucar.edu (Joe Van Andel) The vxworks shell is not re-entrant (sigh . ), so you can only have one copy running. You can't really emulate Unix's fork and exec, since there isn't a mechanism for duplicating a process, including the current program counter and a matching stack pointer. You can run multiple copies of a task, provided the code you write IS reentrant. There isn't any "exec" function either. The only way to start a new process is taskSpawn, or one of its variants, like the shell "sp". - -- Joe VanAndel Internet: vanandel@ncar.ucar.edu National Center for Web http://www.atd.ucar.edu/jva/home.html Atmospheric Research --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sat Jul 15 04:00:22 1995 From: daemon@csg.lbl.gov Date: Sat Jul 15 04:00:26 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 15 04:00:19 PDT 1995 Subject: Re: Task memory protection required in VxWorks development environment.. Subject: Re: pty and shell ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Task memory protection required in VxWorks development environment.. Date: 14 Jul 1995 15:48:56 GMT From: ericr@vcd.hp.com (Eric Ross) Organization: Hewlett-Packard Message-ID: <3u63l8$kct@news.vcd.hp.com> References: Wind River does have a "Vm" product that gets you part way to where you want to be. It allows private memory contexts but these contexts are created manually by each task that needs them. If you need full protection, Wind River does not have anything(at least that they have announced) that gives you that capability. - -- Eric Ross | Hewlett-Packard Company, VPR ericr@vcd.hp.com | Vancouver, WA, USA (360)212-2372 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: pty and shell Date: 14 Jul 1995 22:33:14 GMT From: vanandel@stout.atd.ucar.edu (Joe Van Andel) Organization: National Center for Atmospheric Research Message-ID: <3u6rba$1uc@ncar.ucar.edu> Sender: vanandel@stay.atd.ucar.edu (Joe Van Andel) The vxworks shell is not re-entrant (sigh . ), so you can only have one copy running. You can't really emulate Unix's fork and exec, since there isn't a mechanism for duplicating a process, including the current program counter and a matching stack pointer. You can run multiple copies of a task, provided the code you write IS reentrant. There isn't any "exec" function either. The only way to start a new process is taskSpawn, or one of its variants, like the shell "sp". - -- Joe VanAndel Internet: vanandel@ncar.ucar.edu National Center for Web http://www.atd.ucar.edu/jva/home.html Atmospheric Research --------------------------- End of New-News digest ********************** Subject: Makefile/compile question Submitted-by cgonter@inept.stanford.edu Mon Jul 17 10:28:33 1995 Submitted-by: cgonter@inept.stanford.edu I use to compile code using the simple makefile and when I modified it for vxWorks 5.2 it gives me the following error: inept:cgonter->mbi:make gcc-sde -g -DCPU=R3000 -I/devl/vw/h -I/devl/chex/star/drv/mbi -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 -c mbistar.c cpp: Invalid option `-lang-c' *** Error code 1 make: Fatal error: Command failed for target `mbistar.o' All the makefile does is the compile. I can execute the compile line from the shell and it works fine.. Any ideas? craig cgonter@inept.stanford.edu ######################################################################### CPU = R3000 TOOL = gnu TARGET_DIR = /devl/vw/bin BOARD = R3000 # VxWorks root directory. This is often "/usr/vw". VWHOME = /devl/vw # Where the cross compiler is installed GCCHOME = ${VWHOME}/gnu/sun4.mips MBIHOME = /devl/chex/star/drv/mbi USEROBJDIR = $(MBIHOME)/lib MBIINCLUDEDIR = $(MBIHOME) GCC_EXEC_PREFIX = ${GCCHOME}/lib/gcc-lib/sdebig/2.5.7 ######################################################################### # To change the compiler switches, for example to change from -O # to -g, change the following line: CFLAGS = -g CC_DEFINES = -DCPU=$(CPU) ALL_INCLUDES = -I$(VX_VW_BASE)/h -I$(MBIINCLUDEDIR) MIPSFLAGS = -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 CC_SWITCHES = ${CFLAGS} $(CC_DEFINES) $(ALL_INCLUDES) $(MIPSFLAGS) LD = $(GCCHOME)/bin/ld-sde LDRFLAGS = -r CC = gcc-sde ######################################################################### SRCS = mbistar.c OBJS = mbistar.o all: $(OBJS) $(OBJS): $(SRCS) $(CC) $(CC_SWITCHES) -c $(SRCS) depend: makedepend -- $(ALL_INCLUDES) -- $(SRCS) clean: - rm -f *.lo *.o .c.o: $(CC) -c $(CC_SWITCHES) $< # DO NOT DELETE THIS LINE -- make depend depends on it. mbistar.o: /devl/vw/h/vxWorks.h /devl/vw/h/types/vxCpu.h mbistar.o: /devl/vw/h/types/vxArch.h /devl/vw/h/types/vxParams.h mbistar.o: /devl/vw/h/types/vxTypesBase.h /devl/vw/h/types/vxTypes.h mbistar.o: /devl/vw/h/types/vxANSI.h /devl/vw/h/types/vxTypesOld.h mbistar.o: /devl/vw/h/sys/types.h /devl/vw/h/ioLib.h /devl/vw/h/vwModNum.h mbistar.o: /devl/vw/h/limits.h /devl/vw/h/net/uio.h /devl/vw/h/fcntl.h mbistar.o: /devl/vw/h/sys/fcntlcom.h /devl/vw/h/unistd.h /devl/vw/h/iv.h mbistar.o: /devl/vw/h/vme.h /devl/vw/h/logLib.h /devl/vw/h/taskLib.h mbistar.o: /devl/vw/h/classLib.h /devl/vw/h/objLib.h /devl/vw/h/errno.h mbistar.o: /devl/vw/h/memLib.h /devl/vw/h/qLib.h /devl/vw/h/qClass.h mbistar.o: /devl/vw/h/regs.h /devl/vw/h/excLib.h /devl/vw/h/private/eventP.h mbistar.o: /devl/vw/h/private/semLibP.h /devl/vw/h/semLib.h mbistar.o: /devl/vw/h/private/classLibP.h /devl/vw/h/private/objLibP.h mbistar.o: /devl/vw/h/stdlib.h /devl/vw/h/msgQLib.h mbi.h cmdmsg.h From cgonter@inept.stanford.edu Mon Jul 17 10:28:33 1995 From: cgonter@inept.stanford.edu Date: Mon Jul 17 10:28:36 PDT 1995 Subject: Makefile/compile question I use to compile code using the simple makefile and when I modified it for vxWorks 5.2 it gives me the following error: inept:cgonter->mbi:make gcc-sde -g -DCPU=R3000 -I/devl/vw/h -I/devl/chex/star/drv/mbi -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 -c mbistar.c cpp: Invalid option `-lang-c' *** Error code 1 make: Fatal error: Command failed for target `mbistar.o' All the makefile does is the compile. I can execute the compile line from the shell and it works fine.. Any ideas? craig cgonter@inept.stanford.edu ######################################################################### CPU = R3000 TOOL = gnu TARGET_DIR = /devl/vw/bin BOARD = R3000 # VxWorks root directory. This is often "/usr/vw". VWHOME = /devl/vw # Where the cross compiler is installed GCCHOME = ${VWHOME}/gnu/sun4.mips MBIHOME = /devl/chex/star/drv/mbi USEROBJDIR = $(MBIHOME)/lib MBIINCLUDEDIR = $(MBIHOME) GCC_EXEC_PREFIX = ${GCCHOME}/lib/gcc-lib/sdebig/2.5.7 ######################################################################### # To change the compiler switches, for example to change from -O # to -g, change the following line: CFLAGS = -g CC_DEFINES = -DCPU=$(CPU) ALL_INCLUDES = -I$(VX_VW_BASE)/h -I$(MBIINCLUDEDIR) MIPSFLAGS = -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 CC_SWITCHES = ${CFLAGS} $(CC_DEFINES) $(ALL_INCLUDES) $(MIPSFLAGS) LD = $(GCCHOME)/bin/ld-sde LDRFLAGS = -r CC = gcc-sde ######################################################################### SRCS = mbistar.c OBJS = mbistar.o all: $(OBJS) $(OBJS): $(SRCS) $(CC) $(CC_SWITCHES) -c $(SRCS) depend: makedepend -- $(ALL_INCLUDES) -- $(SRCS) clean: - rm -f *.lo *.o .c.o: $(CC) -c $(CC_SWITCHES) $< # DO NOT DELETE THIS LINE -- make depend depends on it. mbistar.o: /devl/vw/h/vxWorks.h /devl/vw/h/types/vxCpu.h mbistar.o: /devl/vw/h/types/vxArch.h /devl/vw/h/types/vxParams.h mbistar.o: /devl/vw/h/types/vxTypesBase.h /devl/vw/h/types/vxTypes.h mbistar.o: /devl/vw/h/types/vxANSI.h /devl/vw/h/types/vxTypesOld.h mbistar.o: /devl/vw/h/sys/types.h /devl/vw/h/ioLib.h /devl/vw/h/vwModNum.h mbistar.o: /devl/vw/h/limits.h /devl/vw/h/net/uio.h /devl/vw/h/fcntl.h mbistar.o: /devl/vw/h/sys/fcntlcom.h /devl/vw/h/unistd.h /devl/vw/h/iv.h mbistar.o: /devl/vw/h/vme.h /devl/vw/h/logLib.h /devl/vw/h/taskLib.h mbistar.o: /devl/vw/h/classLib.h /devl/vw/h/objLib.h /devl/vw/h/errno.h mbistar.o: /devl/vw/h/memLib.h /devl/vw/h/qLib.h /devl/vw/h/qClass.h mbistar.o: /devl/vw/h/regs.h /devl/vw/h/excLib.h /devl/vw/h/private/eventP.h mbistar.o: /devl/vw/h/private/semLibP.h /devl/vw/h/semLib.h mbistar.o: /devl/vw/h/private/classLibP.h /devl/vw/h/private/objLibP.h mbistar.o: /devl/vw/h/stdlib.h /devl/vw/h/msgQLib.h mbi.h cmdmsg.h Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Tue Jul 18 04:00:18 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Tue Jul 18 04:00:11 PDT 1995 Subject: boot error 0x1a9 Subject: Message Logging Subject: VXworks FAQ Subject: Re: VXworks FAQ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: boot error 0x1a9 Date: 17 Jul 1995 09:18:14 -0700 From: jhudson@kaiwan.kaiwan.com (John G Hudson) Organization: KAIWAN Internet (310-527-4279,818-756-0180,909-785-9712,714-638-4133,805-294-9338) Keywords: boot Message-ID: <3ue2g6$na2@kaiwan.kaiwan.com> Hi, I'm having trouble loading the VxWorks boot file into a Radstone RS41-68 processor. Sometimes it boots immediately after a system reset and other times it takes 5+ attempts. In all cases when it fails errorno=0x1a9 is displayed. Can anyone help? Thanks, John Hudson Rockwell jhudson@power.amasd.anatcp.rockwell.com --------------------------- Newsgroups: comp.os.vxworks Subject: Message Logging Date: 17 Jul 1995 19:20:55 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: I couldn't find this in the FAQ: I have a question about using logMsg() from a procedure which uses a string to repeatedly write messages. Since this is volatile memory, I may overwrite the string with a new message before the old one was dispatched by the logTask. Is there some trick to get around this problem? Thanks in advance, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: VXworks FAQ Date: 17 Jul 1995 20:54:17 -0400 From: arrghh@aol.com (Arrghh) Organization: America Online, Inc. (1-800-827-6364) Message-ID: <3uf0np$3p8@newsbf02.news.aol.com> Reply-To: arrghh@aol.com (Arrghh) Sender: root@newsbf02.news.aol.com anyone know where I can get the vxworks faq? Thanks, Monica --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VXworks FAQ Date: 18 Jul 1995 08:56:25 GMT From: Chonticha Organization: NECTEC - TTL Message-ID: <3ufsvp$7pp@senior.nectec.or.th> References: <3uf0np$3p8@newsbf02.news.aol.com> arrghh@aol.com (Arrghh) wrote: >anyone know where I can get the vxworks faq? Thanks, Monica Try anonymous ftp at the following ftp site: rtfm.mit.edu:/pub/usenet/news.answers/vxworks-faq.z ftp.uu.net:/usenet/news.answers/vxworks-faq.z cs.toronto.edu/pub/usenet/comp.answers/vxworks-faq.z ftp.netcom.com:/pub/hjb/vxfaq.gz good luck Chontp. --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Tue Jul 18 04:00:18 1995 From: daemon@csg.lbl.gov Date: Tue Jul 18 04:00:20 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 18 04:00:11 PDT 1995 Subject: boot error 0x1a9 Subject: Message Logging Subject: VXworks FAQ Subject: Re: VXworks FAQ ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: boot error 0x1a9 Date: 17 Jul 1995 09:18:14 -0700 From: jhudson@kaiwan.kaiwan.com (John G Hudson) Organization: KAIWAN Internet (310-527-4279,818-756-0180,909-785-9712,714-638-4133,805-294-9338) Keywords: boot Message-ID: <3ue2g6$na2@kaiwan.kaiwan.com> Hi, I'm having trouble loading the VxWorks boot file into a Radstone RS41-68 processor. Sometimes it boots immediately after a system reset and other times it takes 5+ attempts. In all cases when it fails errorno=0x1a9 is displayed. Can anyone help? Thanks, John Hudson Rockwell jhudson@power.amasd.anatcp.rockwell.com --------------------------- Newsgroups: comp.os.vxworks Subject: Message Logging Date: 17 Jul 1995 19:20:55 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: I couldn't find this in the FAQ: I have a question about using logMsg() from a procedure which uses a string to repeatedly write messages. Since this is volatile memory, I may overwrite the string with a new message before the old one was dispatched by the logTask. Is there some trick to get around this problem? Thanks in advance, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: VXworks FAQ Date: 17 Jul 1995 20:54:17 -0400 From: arrghh@aol.com (Arrghh) Organization: America Online, Inc. (1-800-827-6364) Message-ID: <3uf0np$3p8@newsbf02.news.aol.com> Reply-To: arrghh@aol.com (Arrghh) Sender: root@newsbf02.news.aol.com anyone know where I can get the vxworks faq? Thanks, Monica --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VXworks FAQ Date: 18 Jul 1995 08:56:25 GMT From: Chonticha Organization: NECTEC - TTL Message-ID: <3ufsvp$7pp@senior.nectec.or.th> References: <3uf0np$3p8@newsbf02.news.aol.com> arrghh@aol.com (Arrghh) wrote: >anyone know where I can get the vxworks faq? Thanks, Monica Try anonymous ftp at the following ftp site: rtfm.mit.edu:/pub/usenet/news.answers/vxworks-faq.z ftp.uu.net:/usenet/news.answers/vxworks-faq.z cs.toronto.edu/pub/usenet/comp.answers/vxworks-faq.z ftp.netcom.com:/pub/hjb/vxfaq.gz good luck Chontp. --------------------------- End of New-News digest ********************** Subject: Calculating remaining stack space Submitted-by uchenick@tate.com Tue Jul 18 06:42:29 1995 Submitted-by: uchenick@tate.com (Gordon Uchenick) ----- Begin Included Message ----- >From MAILER-DAEMON Tue Jul 18 08:39:27 1995 Date: Tue, 18 Jul 95 08:39:22 EDT >From: MAILER-DAEMON (Mail Delivery Subsystem) Subject: Returned mail: User unknown How do you go about calculating the amount of stack space available to your task? I'm trying to do this without all the overhead of calling taskInfoGet(). I tried subtracting some of the fields in the WIND_TCB (such as pStackLimit or pStackEnd) from the current stack pointer, but I never get the same results as if I had called taskInfoGet(). I'm working on a recursive algorithm which I'd like to stop "recursing" when the stack space gets too low for another interation without overflowing my own stack which can crash the CPU. Thanks in advance. I'll summarize findings back to vxwexplo. Gordon Uchenick Tate Integrated Systems, L.P. OwingsMills, MD uchenick@tate.com 410-581-0422 Fax: 410-581-5738 From uchenick@tate.com Tue Jul 18 06:42:29 1995 From: uchenick@tate.com (Gordon Uchenick) Date: Tue Jul 18 06:42:32 PDT 1995 Subject: Calculating remaining stack space ----- Begin Included Message ----- >From MAILER-DAEMON Tue Jul 18 08:39:27 1995 Date: Tue, 18 Jul 95 08:39:22 EDT >From: MAILER-DAEMON (Mail Delivery Subsystem) Subject: Returned mail: User unknown How do you go about calculating the amount of stack space available to your task? I'm trying to do this without all the overhead of calling taskInfoGet(). I tried subtracting some of the fields in the WIND_TCB (such as pStackLimit or pStackEnd) from the current stack pointer, but I never get the same results as if I had called taskInfoGet(). I'm working on a recursive algorithm which I'd like to stop "recursing" when the stack space gets too low for another interation without overflowing my own stack which can crash the CPU. Thanks in advance. I'll summarize findings back to vxwexplo. Gordon Uchenick Tate Integrated Systems, L.P. OwingsMills, MD uchenick@tate.com 410-581-0422 Fax: 410-581-5738 Subject: non blocking read on serial port Submitted-by punamia@newkla.kla.com Tue Jul 18 12:57:41 1995 Submitted-by: punamia@newkla.kla.com (Manoj Punamia) I would appreciate if anyone has any comments or help on the following problem: DESCRIPTION: I am trying to do non blocking "read" from serial port on 68030 Force board. Before calling read, I start watchdog timer routine to timeout the read. The watchdog routine calls ioctl(my_fd, FIOCANCEL,0). PROBLEM: The VxWorks hangsup and I think read is still blocked. The ioctl call does not seem to cancel read. WHY ? Is "select" call a better way to do this ? Version: VxWorks (for Force SYS68K/CPU-30) version 5.0.2b. Kernel: WIND version 2.0. From punamia@newkla.kla.com Tue Jul 18 12:57:41 1995 From: punamia@newkla.kla.com (Manoj Punamia) Date: Tue Jul 18 12:57:43 PDT 1995 Subject: non blocking read on serial port I would appreciate if anyone has any comments or help on the following problem: DESCRIPTION: I am trying to do non blocking "read" from serial port on 68030 Force board. Before calling read, I start watchdog timer routine to timeout the read. The watchdog routine calls ioctl(my_fd, FIOCANCEL,0). PROBLEM: The VxWorks hangsup and I think read is still blocked. The ioctl call does not seem to cancel read. WHY ? Is "select" call a better way to do this ? Version: VxWorks (for Force SYS68K/CPU-30) version 5.0.2b. Kernel: WIND version 2.0. Subject: Re: unsubscribe Submitted-by bordua@csg.lbl.gov Tue Jul 18 17:06:32 1995 Submitted-by: bordua@csg.lbl.gov (Michael Bordua) Unsubscribed from the vxWorks Exploder. Mike Bordua MGBordua@lbl.gov VxWorks Exploder Administrator Computer Systems Group Lawrence Berkeley Laboratory From bordua@csg.lbl.gov Tue Jul 18 17:06:32 1995 From: bordua@csg.lbl.gov (Michael Bordua) Date: Tue Jul 18 17:06:34 PDT 1995 Subject: Re: unsubscribe Unsubscribed from the vxWorks Exploder. Mike Bordua MGBordua@lbl.gov VxWorks Exploder Administrator Computer Systems Group Lawrence Berkeley Laboratory Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Wed Jul 19 04:00:16 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Wed Jul 19 04:00:12 PDT 1995 Subject: Motorola MVME147SA available Subject: Problem with ld on AXPvme160 VxWorks Subject: Re: Calculating remaining stack space Subject: New archive entries for gcc-2.7.0 Subject: Interphase FDDI works on VxWorks 5.02 but not on 5.2 Subject: Re: VXworks FAQ Subject: Converting drivers for on-board chips to same chips on IP bus Subject: Re: Makefile/compile question Subject: SPARCstation/2 motherboards for SALE Subject: rt11 file systems can't be bigger than 2 gig? Subject: Parallel port driver for mv167? Subject: timer_t conflict ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Motorola MVME147SA available Date: 18 Jul 1995 13:43:31 GMT From: innov8@ix.netcom.com (Marty ) Organization: Netcom Message-ID: <3ugdq3$8jb@ixnews7.ix.netcom.com> I have the following Motorola VME card available. Configuration: MVME147SA 68030 CPU, 20Mhz, 8MB DRAM MC68882 floating point coprocessor 4 asynchronous serial port 1 ethernet interface 1 SCSI with DMA and A32/D32 VMEbus 1 parallel port Please e-mail if interested. - -- Marty Norton E-mail: innov8@ix.netcom.com --------------------------- Newsgroups: comp.os.vxworks Subject: Problem with ld on AXPvme160 VxWorks Date: 18 Jul 1995 15:37:53 GMT From: jschamba@physics.utexas.edu (Jo Schambach) Organization: Physics Department, University of Texas at Austin Message-ID: <3ugkgh$ftu@geraldo.cc.utexas.edu> Reply-To: JSchamba@utpapa.ph.utexas.edu After compiling a function on a DECstation 2000/Model 300 with either cc or gcc, I don't seem to be able to load the resulting object file to a VxWorks target: (the target is a DEC AXPvme160 running DEC's version of VxWorks) - -> ld < rft.o ld error: error reading file (errno = 0x3d0001). value = 0 = 0x0 When I look up the error message, I find: #define S_objLib_OBJ_ID_ERROR (M_objLib | 1) where #define M_objLib (61 << 16). I have to admit, that I don't understand what that means. This error seems to be specific to the source file in question, since other source files load just fine. Could someone please enlighten me, what is going on here? Any help would be greatly appreciated. Thank you, Jo ***************************************************************** Joachim Schambach Research Associate Physics Department, RLM 5.208 The University of Texas at Austin Austin, TX 78712 USA email: JSchamba@physics.utexas.edu Tel: x1 (512) 471-1303 FAX: x1 (512) 471-9637 ***************************************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Calculating remaining stack space Date: 18 Jul 1995 16:01:03 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: References: <9507181246.AA13109@tate.com> In article <9507181246.AA13109@tate.com> uchenick@tate.com (Gordon Uchenick) writes: > How do you go about calculating the amount of stack space available to > your task? While we're at it - how do you calculate how much stack space to allocate to a task when you are spawning it? For example, I want to spawn a process to run taskLog(char* format string, int arg1, int arg2, int arg3, int arg4, int arg5, int arg6) > Thanks in advance. I'll summarize findings back to vxwexplo. So what's vxwexplo? Thanks, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: New archive entries for gcc-2.7.0 Date: 18 Jul 1995 17:14:43 GMT From: thor@ymir.atd.ucar.edu (Rich Neitzel) Organization: National Center for Atmospheric Research Message-ID: <3ugq63$31r@ncar.ucar.edu> I've just added an updated vx_cplusplus file to reflect the changes to build gcc-2.7.0 for VxWorks and added a new file, libg++-2.7.0.patch, that has the patches needed for libg++-2.7.0. Premade binaries for a complete cross tools set for 68k targets on a Solaris 2 host are also available via FTP in the vx_cross directory. FTP address: ftp://ftp.atd.ucar.edu/pub/vxworks email server address: vxworks_archive@ncar.ucar.edu - -- Richard Neitzel thor@atd.ucar.edu Torren med sitt skjegg National Center For Atmospheric Research lokkar borni under sole-vegg Box 3000 Boulder, CO 80307-3000 Gjx'i med sitt shinn 303-497-2057 jagar borni inn. --------------------------- Newsgroups: comp.os.vxworks Subject: Interphase FDDI works on VxWorks 5.02 but not on 5.2 Date: 18 Jul 1995 17:15:54 GMT From: getz@getz.manassas.ibm.com Organization: Loral Federal Systems, Manassas Message-ID: <3ugq8a$9hl@lfsserv1.fsc.ibm.com> Reply-To: getz@getz.manassas.ibm.com We have an Interphase 5211 FDDI adapter which is operational under VxWorks 5.02 but we are upgrading to VxWorks 5.2. Using the 5.2 beta (we haven't received the 5.2 release yet), the FDDI driver loads (after compensating for a couple of system calls that the driver uses which are not available in 5.2) and all of the messages look good. However, we cannot communicate with this adapter while running VxWorks 5.2. If we switch back to 5.02, it works as expected. To compensate for the deleted library calls (sysDualPortMap(), and sysCachSet()), we put our SPARC 5CE into enhanced mode to map all memory to the VME bus. Since we don't yet have all of the VxWorks documentation, we turned off caching of all memory on the SPARC card. I know this isn't a good idea, but the FDDI driver complained without doing this. Any help would be greatly appreciated. Fred Getz --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VXworks FAQ Date: 18 Jul 1995 18:25:33 GMT From: hjb@nova.netapp.com (hwajin bae-Consultant) Organization: Network Appliance Corporation Message-ID: References: <3uf0np$3p8@newsbf02.news.aol.com> <3ufsvp$7pp@senior.nectec.or.th> Reply-To: squeeby@hooked.net >>>>> "Chonticha" == Chonticha writes: In article <3ufsvp$7pp@senior.nectec.or.th> Chonticha writes: Chonticha> Try anonymous ftp at the following ftp site: Chonticha> rtfm.mit.edu:/pub/usenet/news.answers/vxworks-faq.z Chonticha> ftp.uu.net:/usenet/news.answers/vxworks-faq.z Chonticha> cs.toronto.edu/pub/usenet/comp.answers/vxworks-faq.z Chonticha> ftp.netcom.com:/pub/hjb/vxfaq.gz ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ this last one is no longer valid. there is no hjb on netcom.com anymore. ftp.mclean.sparta.com should be the place for the most up-to-date FAQ, which is now being maintained by Mike A. hwajin - -- Hwa-Jin Bae *IX/Realtime/Embbeded/Network/HW/SW Consultant hjb@netcom.com - Peaceful Star Project Vox: 510-536-7607 Page: 510-466-9166 hjb@netapp.com - Network Appliance hjb@hybrid.com - Hybrid Networks --------------------------- Newsgroups: comp.os.vxworks Subject: Converting drivers for on-board chips to same chips on IP bus Date: Tue, 18 Jul 95 13:18:00 MDT From: jeffh@lfs.loral.com Message-ID: <9507181923.AA56674@inetgw.fsc.ibm.com> Our hardware platform is a Motorola MVME162LX SBC. We have the drivers for the on board SCC, a Z85230 chip. Does anyone have an idea/experience in converting the drivers to work with the same chip on an IP board? Purchasing forgot to order the technical documentation set, so until that comes in, I won't be able to answer this for myself. If you would answer directly back to my e-mail address, I would appreciate the advice. Thanks, /\ /\ ____o_o o Jeff Hartt /\ / *\/\ / *\ /\ /_*_||*/\ Loral Federal Systems /* \ * * \/ * \/ *\| __ | | jeffh@lfs.loral.com / \_______*__*_\___|__||__|__| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile/compile question Date: 18 Jul 1995 21:48:16 GMT From: Don Small Organization: Sandia National Laboratories Message-ID: <3uha70$kru@news.sandia.gov> References: <199507171028.DAA22693@inept.stanford.edu> I get the same error any time I have incorrectly set the GCC_EXEC_PREFIX variable. It is executing the host cpp instead of cpp-sde as you see on the line cpp: Invalid option. Our GCC_EXEC_PREFIX seems to work properly and the only difference I see is a final `/` at the end: GCC_EXEC_PREFIX = /usr/local/gnu/gnu2.5.7/lib/gcc-lib/sdebig/2.5.7/ Hope that helps. Don --------------------------- Newsgroups: comp.os.vxworks Subject: SPARCstation/2 motherboards for SALE Date: 18 Jul 1995 21:49:47 GMT From: jtalbot@3do.com (Jim Talbot) Organization: The 3DO Company, Redwood City, CA Keywords: unix, sparc, vme, vxworks Message-ID: <3uha9r$1ce@badger.3do.com> Two (2) SPARCstation/2 Motherboards for Sale ============================================= I have two (2) SPARCstation/2 motherboards. These boards can be used to build your own SPARCstation (add keyboard, disk, monitor) and/or VxWorks target. They are 6U VMEbus boards manufactured by a SPARC licensee that I used to develop a GREAT VxWorks BSP (Board Support Package). I also did the SPARC kernel when I was working at Wind River Systems. Technical info: - -------------- 40MHz clock rate 16MB or 32MB of DRAM 64K or 128K of cache 2 Sbus slots High-speed VMEbus interface chips Connectors for: keyboard, SCSI, ethernet, Sbus, VME A32/D32 For more details contact me at jim.talbot@3do.com or (415) 261-2738. --------------------------- Newsgroups: comp.os.vxworks Subject: rt11 file systems can't be bigger than 2 gig? Date: 18 Jul 1995 22:10:02 GMT From: vanandel@stout.atd.ucar.edu (Joe Van Andel) Organization: National Center for Atmospheric Research Message-ID: <3uhbfq$or6@ncar.ucar.edu> Sender: vanandel@clipper.atd.ucar.edu (Joe Van Andel) I recently upgraded a disk in one of our VxWorks systems from 2gig to 4 gig. Naturally, I was a bit upset to find that with VxWorks 5.1.2, I couldn't build an RT11 file system bigger than 2 gig. Has this been fixed in 5.2, or does WRS have a patch? Thanks much. - -- Joe VanAndel Internet: vanandel@ncar.ucar.edu National Center for Web http://www.atd.ucar.edu/jva/home.html Atmospheric Research --------------------------- Newsgroups: comp.os.vxworks Subject: Parallel port driver for mv167? Date: 18 Jul 95 12:49:57 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: driver mv167 Message-ID: <1995Jul18.124958.23460@hrbicf> - -- Hi, We are upgrading a system to use mv167 cards replacing some mv121 cards. The mv167 card contains a parallel printer port on it which we have accessed via the original 167bug. Has anyone written a parallel port printer driver for VxWorks 5.1 that they are willing to share or could someone enlighten us to what we need to do to setup the printer port? Thanks, - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: timer_t conflict Date: 19 Jul 1995 01:24:56 GMT From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: The following regards VxWorks (for Motorola MVME167) version 5.1.1. The VxWorks man page for timer_create() says that timer_create() returns -1 on error. However the return type of timer_create() is timer_t, which under VxWorks is typedef'd as: typedef struct __timer *timer_t in /usr/vw/h/types/vxTypesBase.h. Given that timer_t is a pointer, a return value of -1 makes little sense, so what is the correct error value under VxWorks, and is this likely to be changed in the future? Martin Shepherd (mcs@astro.caltech.edu) --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Wed Jul 19 04:00:16 1995 From: daemon@csg.lbl.gov Date: Wed Jul 19 04:00:21 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 19 04:00:12 PDT 1995 Subject: Motorola MVME147SA available Subject: Problem with ld on AXPvme160 VxWorks Subject: Re: Calculating remaining stack space Subject: New archive entries for gcc-2.7.0 Subject: Interphase FDDI works on VxWorks 5.02 but not on 5.2 Subject: Re: VXworks FAQ Subject: Converting drivers for on-board chips to same chips on IP bus Subject: Re: Makefile/compile question Subject: SPARCstation/2 motherboards for SALE Subject: rt11 file systems can't be bigger than 2 gig? Subject: Parallel port driver for mv167? Subject: timer_t conflict ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Motorola MVME147SA available Date: 18 Jul 1995 13:43:31 GMT From: innov8@ix.netcom.com (Marty ) Organization: Netcom Message-ID: <3ugdq3$8jb@ixnews7.ix.netcom.com> I have the following Motorola VME card available. Configuration: MVME147SA 68030 CPU, 20Mhz, 8MB DRAM MC68882 floating point coprocessor 4 asynchronous serial port 1 ethernet interface 1 SCSI with DMA and A32/D32 VMEbus 1 parallel port Please e-mail if interested. - -- Marty Norton E-mail: innov8@ix.netcom.com --------------------------- Newsgroups: comp.os.vxworks Subject: Problem with ld on AXPvme160 VxWorks Date: 18 Jul 1995 15:37:53 GMT From: jschamba@physics.utexas.edu (Jo Schambach) Organization: Physics Department, University of Texas at Austin Message-ID: <3ugkgh$ftu@geraldo.cc.utexas.edu> Reply-To: JSchamba@utpapa.ph.utexas.edu After compiling a function on a DECstation 2000/Model 300 with either cc or gcc, I don't seem to be able to load the resulting object file to a VxWorks target: (the target is a DEC AXPvme160 running DEC's version of VxWorks) - -> ld < rft.o ld error: error reading file (errno = 0x3d0001). value = 0 = 0x0 When I look up the error message, I find: #define S_objLib_OBJ_ID_ERROR (M_objLib | 1) where #define M_objLib (61 << 16). I have to admit, that I don't understand what that means. This error seems to be specific to the source file in question, since other source files load just fine. Could someone please enlighten me, what is going on here? Any help would be greatly appreciated. Thank you, Jo ***************************************************************** Joachim Schambach Research Associate Physics Department, RLM 5.208 The University of Texas at Austin Austin, TX 78712 USA email: JSchamba@physics.utexas.edu Tel: x1 (512) 471-1303 FAX: x1 (512) 471-9637 ***************************************************************** --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Calculating remaining stack space Date: 18 Jul 1995 16:01:03 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: References: <9507181246.AA13109@tate.com> In article <9507181246.AA13109@tate.com> uchenick@tate.com (Gordon Uchenick) writes: > How do you go about calculating the amount of stack space available to > your task? While we're at it - how do you calculate how much stack space to allocate to a task when you are spawning it? For example, I want to spawn a process to run taskLog(char* format string, int arg1, int arg2, int arg3, int arg4, int arg5, int arg6) > Thanks in advance. I'll summarize findings back to vxwexplo. So what's vxwexplo? Thanks, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: New archive entries for gcc-2.7.0 Date: 18 Jul 1995 17:14:43 GMT From: thor@ymir.atd.ucar.edu (Rich Neitzel) Organization: National Center for Atmospheric Research Message-ID: <3ugq63$31r@ncar.ucar.edu> I've just added an updated vx_cplusplus file to reflect the changes to build gcc-2.7.0 for VxWorks and added a new file, libg++-2.7.0.patch, that has the patches needed for libg++-2.7.0. Premade binaries for a complete cross tools set for 68k targets on a Solaris 2 host are also available via FTP in the vx_cross directory. FTP address: ftp://ftp.atd.ucar.edu/pub/vxworks email server address: vxworks_archive@ncar.ucar.edu - -- Richard Neitzel thor@atd.ucar.edu Torren med sitt skjegg National Center For Atmospheric Research lokkar borni under sole-vegg Box 3000 Boulder, CO 80307-3000 Gjx'i med sitt shinn 303-497-2057 jagar borni inn. --------------------------- Newsgroups: comp.os.vxworks Subject: Interphase FDDI works on VxWorks 5.02 but not on 5.2 Date: 18 Jul 1995 17:15:54 GMT From: getz@getz.manassas.ibm.com Organization: Loral Federal Systems, Manassas Message-ID: <3ugq8a$9hl@lfsserv1.fsc.ibm.com> Reply-To: getz@getz.manassas.ibm.com We have an Interphase 5211 FDDI adapter which is operational under VxWorks 5.02 but we are upgrading to VxWorks 5.2. Using the 5.2 beta (we haven't received the 5.2 release yet), the FDDI driver loads (after compensating for a couple of system calls that the driver uses which are not available in 5.2) and all of the messages look good. However, we cannot communicate with this adapter while running VxWorks 5.2. If we switch back to 5.02, it works as expected. To compensate for the deleted library calls (sysDualPortMap(), and sysCachSet()), we put our SPARC 5CE into enhanced mode to map all memory to the VME bus. Since we don't yet have all of the VxWorks documentation, we turned off caching of all memory on the SPARC card. I know this isn't a good idea, but the FDDI driver complained without doing this. Any help would be greatly appreciated. Fred Getz --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VXworks FAQ Date: 18 Jul 1995 18:25:33 GMT From: hjb@nova.netapp.com (hwajin bae-Consultant) Organization: Network Appliance Corporation Message-ID: References: <3uf0np$3p8@newsbf02.news.aol.com> <3ufsvp$7pp@senior.nectec.or.th> Reply-To: squeeby@hooked.net >>>>> "Chonticha" == Chonticha writes: In article <3ufsvp$7pp@senior.nectec.or.th> Chonticha writes: Chonticha> Try anonymous ftp at the following ftp site: Chonticha> rtfm.mit.edu:/pub/usenet/news.answers/vxworks-faq.z Chonticha> ftp.uu.net:/usenet/news.answers/vxworks-faq.z Chonticha> cs.toronto.edu/pub/usenet/comp.answers/vxworks-faq.z Chonticha> ftp.netcom.com:/pub/hjb/vxfaq.gz ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ this last one is no longer valid. there is no hjb on netcom.com anymore. ftp.mclean.sparta.com should be the place for the most up-to-date FAQ, which is now being maintained by Mike A. hwajin - -- Hwa-Jin Bae *IX/Realtime/Embbeded/Network/HW/SW Consultant hjb@netcom.com - Peaceful Star Project Vox: 510-536-7607 Page: 510-466-9166 hjb@netapp.com - Network Appliance hjb@hybrid.com - Hybrid Networks --------------------------- Newsgroups: comp.os.vxworks Subject: Converting drivers for on-board chips to same chips on IP bus Date: Tue, 18 Jul 95 13:18:00 MDT From: jeffh@lfs.loral.com Message-ID: <9507181923.AA56674@inetgw.fsc.ibm.com> Our hardware platform is a Motorola MVME162LX SBC. We have the drivers for the on board SCC, a Z85230 chip. Does anyone have an idea/experience in converting the drivers to work with the same chip on an IP board? Purchasing forgot to order the technical documentation set, so until that comes in, I won't be able to answer this for myself. If you would answer directly back to my e-mail address, I would appreciate the advice. Thanks, /\ /\ ____o_o o Jeff Hartt /\ / *\/\ / *\ /\ /_*_||*/\ Loral Federal Systems /* \ * * \/ * \/ *\| __ | | jeffh@lfs.loral.com / \_______*__*_\___|__||__|__| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile/compile question Date: 18 Jul 1995 21:48:16 GMT From: Don Small Organization: Sandia National Laboratories Message-ID: <3uha70$kru@news.sandia.gov> References: <199507171028.DAA22693@inept.stanford.edu> I get the same error any time I have incorrectly set the GCC_EXEC_PREFIX variable. It is executing the host cpp instead of cpp-sde as you see on the line cpp: Invalid option. Our GCC_EXEC_PREFIX seems to work properly and the only difference I see is a final `/` at the end: GCC_EXEC_PREFIX = /usr/local/gnu/gnu2.5.7/lib/gcc-lib/sdebig/2.5.7/ Hope that helps. Don --------------------------- Newsgroups: comp.os.vxworks Subject: SPARCstation/2 motherboards for SALE Date: 18 Jul 1995 21:49:47 GMT From: jtalbot@3do.com (Jim Talbot) Organization: The 3DO Company, Redwood City, CA Keywords: unix, sparc, vme, vxworks Message-ID: <3uha9r$1ce@badger.3do.com> Two (2) SPARCstation/2 Motherboards for Sale ============================================= I have two (2) SPARCstation/2 motherboards. These boards can be used to build your own SPARCstation (add keyboard, disk, monitor) and/or VxWorks target. They are 6U VMEbus boards manufactured by a SPARC licensee that I used to develop a GREAT VxWorks BSP (Board Support Package). I also did the SPARC kernel when I was working at Wind River Systems. Technical info: - -------------- 40MHz clock rate 16MB or 32MB of DRAM 64K or 128K of cache 2 Sbus slots High-speed VMEbus interface chips Connectors for: keyboard, SCSI, ethernet, Sbus, VME A32/D32 For more details contact me at jim.talbot@3do.com or (415) 261-2738. --------------------------- Newsgroups: comp.os.vxworks Subject: rt11 file systems can't be bigger than 2 gig? Date: 18 Jul 1995 22:10:02 GMT From: vanandel@stout.atd.ucar.edu (Joe Van Andel) Organization: National Center for Atmospheric Research Message-ID: <3uhbfq$or6@ncar.ucar.edu> Sender: vanandel@clipper.atd.ucar.edu (Joe Van Andel) I recently upgraded a disk in one of our VxWorks systems from 2gig to 4 gig. Naturally, I was a bit upset to find that with VxWorks 5.1.2, I couldn't build an RT11 file system bigger than 2 gig. Has this been fixed in 5.2, or does WRS have a patch? Thanks much. - -- Joe VanAndel Internet: vanandel@ncar.ucar.edu National Center for Web http://www.atd.ucar.edu/jva/home.html Atmospheric Research --------------------------- Newsgroups: comp.os.vxworks Subject: Parallel port driver for mv167? Date: 18 Jul 95 12:49:57 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: driver mv167 Message-ID: <1995Jul18.124958.23460@hrbicf> - -- Hi, We are upgrading a system to use mv167 cards replacing some mv121 cards. The mv167 card contains a parallel printer port on it which we have accessed via the original 167bug. Has anyone written a parallel port printer driver for VxWorks 5.1 that they are willing to share or could someone enlighten us to what we need to do to setup the printer port? Thanks, - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: timer_t conflict Date: 19 Jul 1995 01:24:56 GMT From: mcs@goblin.caltech.edu (Martin Shepherd) Organization: California Institute of Technology. Message-ID: The following regards VxWorks (for Motorola MVME167) version 5.1.1. The VxWorks man page for timer_create() says that timer_create() returns -1 on error. However the return type of timer_create() is timer_t, which under VxWorks is typedef'd as: typedef struct __timer *timer_t in /usr/vw/h/types/vxTypesBase.h. Given that timer_t is a pointer, a return value of -1 makes little sense, so what is the correct error value under VxWorks, and is this likely to be changed in the future? Martin Shepherd (mcs@astro.caltech.edu) --------------------------- End of New-News digest ********************** Subject: cplusCtorsLink() causes bus error Submitted-by mmilde@scires.com Wed Jul 19 10:20:35 1995 Submitted-by: mmilde@scires.com I've taken object code, munched it and linked it in with a bootable image of vxWorks running in flash memory. To execute the global constructors in my code I've include cplusCtorsLink() as follows in usrConfig.c (WindC++ Gateway User's Guide 1.0, page 25). #define INCLUDE_CPLUS cplusLibInit(); cplusCtorsLink(); <---- New line of code #endif The system crashes with a bus error when it executes cplusCtorsLink(). There is not a problem with any of the constructors in the global instances of objects because the same object code runs just fine when downloaded from the host using "ld" and executing the constructors with cplusCtors(). The bus error also indicates that the program counter is the same as the address that caused the bus error. This tells me that maybe cplusCtorsLink() is attempting to jump to a constructor but is jumping somewhere else instead. Any help would be greatly appreciated! Thanks, Mike Milde Scientific Research Corporation (404) 859-9161 mmilde@scires.com From mmilde@scires.com Wed Jul 19 10:20:35 1995 From: mmilde@scires.com Date: Wed Jul 19 10:20:37 PDT 1995 Subject: cplusCtorsLink() causes bus error I've taken object code, munched it and linked it in with a bootable image of vxWorks running in flash memory. To execute the global constructors in my code I've include cplusCtorsLink() as follows in usrConfig.c (WindC++ Gateway User's Guide 1.0, page 25). #define INCLUDE_CPLUS cplusLibInit(); cplusCtorsLink(); <---- New line of code #endif The system crashes with a bus error when it executes cplusCtorsLink(). There is not a problem with any of the constructors in the global instances of objects because the same object code runs just fine when downloaded from the host using "ld" and executing the constructors with cplusCtors(). The bus error also indicates that the program counter is the same as the address that caused the bus error. This tells me that maybe cplusCtorsLink() is attempting to jump to a constructor but is jumping somewhere else instead. Any help would be greatly appreciated! Thanks, Mike Milde Scientific Research Corporation (404) 859-9161 mmilde@scires.com Subject: fopen() problem Submitted-by perryd@boris.lbl.gov Wed Jul 19 13:45:02 1995 Submitted-by: "David B. Perry" Earlier I wrote: > I am experiencing a problem with the fopen routine. I am accessing files > resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D (68040/VxWorks > 5.1.1) accross Ethernet. The fopen routine will open and close the files, but > it does not return a NULL file pointer when the file does not exist (as > documented). I have performed the fopen function from code and from the shell > and in both cases the fopen routine does not recognize that the file does not > exist. RSH is being used for the remote file access method. > > Has anyone seen this problem before? Wind River has not come up with an answer... Sergi Casas responded with the suggestion to look at errno. It was a good idea (thanks!) (why didn't WE think of it...), but unfortunately, did not help as errno was not set. I think the fault is in the driver (both the original problem & lack of errno). Sure could use some help folks... -- ===================================================================== David B. Perry e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== From perryd@boris.lbl.gov Wed Jul 19 13:45:02 1995 From: "David B. Perry" Date: Wed Jul 19 13:45:11 PDT 1995 Subject: fopen() problem Earlier I wrote: > I am experiencing a problem with the fopen routine. I am accessing files > resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D (68040/VxWorks > 5.1.1) accross Ethernet. The fopen routine will open and close the files, but > it does not return a NULL file pointer when the file does not exist (as > documented). I have performed the fopen function from code and from the shell > and in both cases the fopen routine does not recognize that the file does not > exist. RSH is being used for the remote file access method. > > Has anyone seen this problem before? Wind River has not come up with an answer... Sergi Casas responded with the suggestion to look at errno. It was a good idea (thanks!) (why didn't WE think of it...), but unfortunately, did not help as errno was not set. I think the fault is in the driver (both the original problem & lack of errno). Sure could use some help folks... -- ===================================================================== David B. Perry e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== Subject: Re: fopen() problem Submitted-by stan@rti.com Wed Jul 19 15:16:10 1995 Submitted-by: Stan Schneider Hi, >> > I am experiencing a problem with the fopen routine. I am accessing >> > files resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D >> > (68040/VxWorks 5.1.1) accross Ethernet. The fopen routine will open >> > and close the files, but it does not return a NULL file pointer when >> > the file does not exist (as documented). I have performed the fopen >> > function from code and from the shell and in both cases the fopen >> > routine does not recognize that the file does not exist. RSH is being >> > used for the remote file access method. >> > >> > Has anyone seen this problem before? Wind River has not come up with an >> > answer... >> I don't know if this is related, but we had strange problems with files accessed through symbolic links set by our Challenge (IRIX 5.3). For some reason, VxWorks had trouble with symlinks set by the SGI. We solved the problem by destroying the links, & resetting them from a Sun. Of course, that doesn't help if you don't have a Sun. I think it was a permission problem; symlinks set by the SGI don't have 777 permission mode like the Sun does. You might try changing your umask to 000, deleting any symbolic links, and recreating them. You also might want to try the nfs file system... HTH, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Wed Jul 19 15:16:10 1995 From: Stan Schneider Date: Wed Jul 19 15:16:12 PDT 1995 Subject: Re: fopen() problem Hi, >> > I am experiencing a problem with the fopen routine. I am accessing >> > files resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D >> > (68040/VxWorks 5.1.1) accross Ethernet. The fopen routine will open >> > and close the files, but it does not return a NULL file pointer when >> > the file does not exist (as documented). I have performed the fopen >> > function from code and from the shell and in both cases the fopen >> > routine does not recognize that the file does not exist. RSH is being >> > used for the remote file access method. >> > >> > Has anyone seen this problem before? Wind River has not come up with an >> > answer... >> I don't know if this is related, but we had strange problems with files accessed through symbolic links set by our Challenge (IRIX 5.3). For some reason, VxWorks had trouble with symlinks set by the SGI. We solved the problem by destroying the links, & resetting them from a Sun. Of course, that doesn't help if you don't have a Sun. I think it was a permission problem; symlinks set by the SGI don't have 777 permission mode like the Sun does. You might try changing your umask to 000, deleting any symbolic links, and recreating them. You also might want to try the nfs file system... HTH, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: booting vxworks on Solaris system Submitted-by cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 Submitted-by: cgonter@inept.stanford.edu Can anyone clue me in getting vxworks to boot on a Solaris 2.4 craig cgonter@inept.stanford.edu What has been done: host -> modified the /etc/dfs/dfstab with share -F nfs -o rw -d "vxworks stuff" /opt1 then executed shareall /usr/lib/nfs/nfsd -a 16 /usr/lib/nfs/mountd then check the nfs system by mounting the partition onto another workstation. The results are: target -> [VxWorks Boot]: @ boot device : egl processor number : 0 host name : inept file name : /opt1/devl/vw/config/star/vxWorks inet on ethernet (e) : 36.64.0.200:ff000000 host inet (h) : 36.64.0.195 user (u) : cgonter flags (f) : 0x0 target name (tn) : whoosh Attaching network interface egl0... done. Attaching network interface lo0... done. Loading... permission denied Error loading file: errno = 0x250002. From cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 From: cgonter@inept.stanford.edu Date: Wed Jul 19 17:16:39 PDT 1995 Subject: booting vxworks on Solaris system Can anyone clue me in getting vxworks to boot on a Solaris 2.4 craig cgonter@inept.stanford.edu What has been done: host -> modified the /etc/dfs/dfstab with share -F nfs -o rw -d "vxworks stuff" /opt1 then executed shareall /usr/lib/nfs/nfsd -a 16 /usr/lib/nfs/mountd then check the nfs system by mounting the partition onto another workstation. The results are: target -> [VxWorks Boot]: @ boot device : egl processor number : 0 host name : inept file name : /opt1/devl/vw/config/star/vxWorks inet on ethernet (e) : 36.64.0.200:ff000000 host inet (h) : 36.64.0.195 user (u) : cgonter flags (f) : 0x0 target name (tn) : whoosh Attaching network interface egl0... done. Attaching network interface lo0... done. Loading... permission denied Error loading file: errno = 0x250002. Subject: Re: booting vxworks on Solaris system Submitted-by mmuri@qualcomm.com Wed Jul 19 18:22:42 1995 Submitted-by: mmuri@qualcomm.com (Mark Muri) >Submitted-by cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 >Submitted-by: cgonter@inept.stanford.edu > > >Can anyone clue me in getting vxworks to boot on a Solaris 2.4 >[VxWorks Boot]: @ > >boot device : egl >processor number : 0 >host name : inept >file name : /opt1/devl/vw/config/star/vxWorks >inet on ethernet (e) : 36.64.0.200:ff000000 >host inet (h) : 36.64.0.195 >user (u) : cgonter >flags (f) : 0x0 >target name (tn) : whoosh > >Attaching network interface egl0... done. >Attaching network interface lo0... done. >Loading... permission denied > >Error loading file: errno = 0x250002. Don't know about Solaris booting, but this error message (when I'm booting a VME card from a Solaris machine) usually means that 'whoosh' has not been added to the .rhost file in 'cgonter's home directory accessable via 'inept'. And, of course, the vxWorks kernel (and dir paths) needs to be readable by the universe. You probably gnu this. Mark <=================o=====================================================> | Mark Muri | E-Mail: mmuri@qualcomm.com (619) Work: 658-3303 | | Senior Engineer | Office: Q-244A Fax: 658-2108 | | QUALCOMM, Inc <+> Car: xxx-xxxx | | 6455 Lusk Blvd., San Diego, CA 92121 Pager: 636-0446 | <=======================================================================> From mmuri@qualcomm.com Wed Jul 19 18:22:42 1995 From: mmuri@qualcomm.com (Mark Muri) Date: Wed Jul 19 18:22:45 PDT 1995 Subject: Re: booting vxworks on Solaris system >Submitted-by cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 >Submitted-by: cgonter@inept.stanford.edu > > >Can anyone clue me in getting vxworks to boot on a Solaris 2.4 >[VxWorks Boot]: @ > >boot device : egl >processor number : 0 >host name : inept >file name : /opt1/devl/vw/config/star/vxWorks >inet on ethernet (e) : 36.64.0.200:ff000000 >host inet (h) : 36.64.0.195 >user (u) : cgonter >flags (f) : 0x0 >target name (tn) : whoosh > >Attaching network interface egl0... done. >Attaching network interface lo0... done. >Loading... permission denied > >Error loading file: errno = 0x250002. Don't know about Solaris booting, but this error message (when I'm booting a VME card from a Solaris machine) usually means that 'whoosh' has not been added to the .rhost file in 'cgonter's home directory accessable via 'inept'. And, of course, the vxWorks kernel (and dir paths) needs to be readable by the universe. You probably gnu this. Mark <=================o=====================================================> | Mark Muri | E-Mail: mmuri@qualcomm.com (619) Work: 658-3303 | | Senior Engineer | Office: Q-244A Fax: 658-2108 | | QUALCOMM, Inc <+> Car: xxx-xxxx | | 6455 Lusk Blvd., San Diego, CA 92121 Pager: 636-0446 | <=======================================================================> Subject: Re: booting vxworks on Solaris system Submitted-by stan@rti.com Wed Jul 19 18:25:02 1995 Submitted-by: Stan Schneider >> >> Can anyone clue me in getting vxworks to boot on a Solaris 2.4 >> >> craig >> cgonter@inept.stanford.edu >> >> What has been done: >> >> The results are: >> >> target -> >> >> [VxWorks Boot]: @ >> >> boot device : egl >> processor number : 0 >> host name : inept >> file name : /opt1/devl/vw/config/star/vxWorks >> inet on ethernet (e) : 36.64.0.200:ff000000 >> host inet (h) : 36.64.0.195 >> user (u) : cgonter >> flags (f) : 0x0 >> target name (tn) : whoosh >> >> Attaching network interface egl0... done. >> Attaching network interface lo0... done. >> Loading... permission denied >> >> Error loading file: errno = 0x250002. NFS has nothing to do with loading. The problem is lack of permissions to rsh into the host. For instance, it's possible that the user "cgonter" doesn't have the target "whoosh" in your .rhosts file, or in the /etc/hosts.equiv file on inept. Also, I think Stanford is running its class A net (36.) with a netmask of 0xffff0000. But that wouldn't give you the error above. -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Wed Jul 19 18:25:02 1995 From: Stan Schneider Date: Wed Jul 19 18:25:05 PDT 1995 Subject: Re: booting vxworks on Solaris system >> >> Can anyone clue me in getting vxworks to boot on a Solaris 2.4 >> >> craig >> cgonter@inept.stanford.edu >> >> What has been done: >> >> The results are: >> >> target -> >> >> [VxWorks Boot]: @ >> >> boot device : egl >> processor number : 0 >> host name : inept >> file name : /opt1/devl/vw/config/star/vxWorks >> inet on ethernet (e) : 36.64.0.200:ff000000 >> host inet (h) : 36.64.0.195 >> user (u) : cgonter >> flags (f) : 0x0 >> target name (tn) : whoosh >> >> Attaching network interface egl0... done. >> Attaching network interface lo0... done. >> Loading... permission denied >> >> Error loading file: errno = 0x250002. NFS has nothing to do with loading. The problem is lack of permissions to rsh into the host. For instance, it's possible that the user "cgonter" doesn't have the target "whoosh" in your .rhosts file, or in the /etc/hosts.equiv file on inept. Also, I think Stanford is running its class A net (36.) with a netmask of 0xffff0000. But that wouldn't give you the error above. -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: Re: booting vxworks on Solaris system Submitted-by pat@wrs.com Wed Jul 19 19:13:52 1995 Submitted-by: Patrick Boylan the vxWorks Users Group Exploder writes: > > Submitted-by cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 > Submitted-by: cgonter@inept.stanford.edu > > > Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > ... > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > Try using FTP by specifying a password. It that works then you probably only need to add the host to your ~/.rhosts or /etc/hosts.equiv file. If it doesn't work, then check the read and execute permissions on the file and it's parent directories. It you're still having trouble, check your .cshrc file. Occasionally this causes problems when booting with "rsh" as you are. For example "stty" commands in the .cshrc file can really cause a real head-scratching boot failure. Good luck. -Patrick From pat@wrs.com Wed Jul 19 19:13:52 1995 From: Patrick Boylan Date: Wed Jul 19 19:13:54 PDT 1995 Subject: Re: booting vxworks on Solaris system the vxWorks Users Group Exploder writes: > > Submitted-by cgonter@inept.stanford.edu Wed Jul 19 17:16:37 1995 > Submitted-by: cgonter@inept.stanford.edu > > > Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > ... > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > Try using FTP by specifying a password. It that works then you probably only need to add the host to your ~/.rhosts or /etc/hosts.equiv file. If it doesn't work, then check the read and execute permissions on the file and it's parent directories. It you're still having trouble, check your .cshrc file. Occasionally this causes problems when booting with "rsh" as you are. For example "stty" commands in the .cshrc file can really cause a real head-scratching boot failure. Good luck. -Patrick Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Thu Jul 20 04:00:22 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Thu Jul 20 04:00:17 PDT 1995 Subject: Driver for the SCC2692 DUART Subject: Re: Makefile/compile question Subject: Re: Problem with ld on AXPvme160 VxWorks Subject: Re: Makefile/compile question Subject: Transfer from VAXELN to VXworks Subject: System clock synchronzation Subject: C programmer needed Subject: VxWorks SNMP agent questions? Subject: Re: non blocking read on serial port Subject: VxWorks / VxWindows Problem ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Driver for the SCC2692 DUART Date: 19 Jul 1995 10:42:49 GMT From: Adi Regev Organization: RADVision Ltd. Message-ID: <3uinj9$fa1@radmail.rad.co.il> I want to write a driver for the PHILIPS SCC2692 DUART. I will enjoy some writen code, other info will be accepted in joy! - ------------------------------------------------------------------- Adi Regev RADVision Ltd. adi@radvision.rad.co.il --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile/compile question Date: 19 Jul 1995 15:56:07 GMT From: michaelv@qualcomm.com (Michael Vakulenko) Organization: Qualcomm Israel Message-ID: References: <199507171028.DAA22693@inept.stanford.edu> In article <199507171028.DAA22693@inept.stanford.edu>, cgonter@inept.stanford.edu wrote: > I use to compile code using the simple makefile and when I modified it for > vxWorks 5.2 it gives me the following error: > > > inept:cgonter->mbi:make > gcc-sde -g -DCPU=R3000 -I/devl/vw/h -I/devl/chex/star/drv/mbi - -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 -c mbistar.c > cpp: Invalid option `-lang-c' > *** Error code 1 > make: Fatal error: Command failed for target `mbistar.o' > > I had similar problem with VxWorks v5.2 installed from CD. We have 68K target with Solaris host. The problem was that INSTALL script has incorrectly defined links in "$VX_GNU_BASE/gnu/solaris.68k/lib/gcc-lib/m68k-wrs-vxworks/cygnus-2.2.3.1/" directory. This directory contains links to execs which will be launched during the compilation (e.g. cc68k starts assembler from here). After the installation, the links pointed nowhere in our case and the compiler failed to start the correct exec. In our configuration, the links should point to "$VX_GNU_BASE/gnu/solaris.68k/lib/m68k-wrs-vxworks/bin/" directory. Of course, you have different directory names, but I hope you've got an idea. ______________________________________________________________ Michael Vakulenko, Voice: +972-4-577999 Software Engineer, Fax: +972-4-577998 Qualcomm Israel, Ltd. Email: michaelv@qualcomm.com ______________________________________________________________ __________________________________________________________ Michael Vakulenko, Voice: +972-4-577999 Software Engineer, Fax: +972-4-577998 Qualcomm Israel, Ltd Email: michaelv@qualcomm.com __________________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem with ld on AXPvme160 VxWorks Date: 19 Jul 1995 18:48:41 GMT From: Ron Reid Organization: Honeywell Message-ID: <3ujk29$as@killerbee.jsc.nasa.gov> References: <3ugkgh$ftu@geraldo.cc.utexas.edu> This is what works for me on a similar setup (replace "ws04", "193.0.0.1", "myobj.o" and "/mnt/usr/mydir" with your info) : vxworks> hostAdd "ws04","193.0.0.1" vxworks> netDevCreate "ws04:", "ws04", 1 vxworks> cd ws04:/mnt/usr/mydir vxworks> ld References: <199507171028.DAA22693@inept.stanford.edu> In article michaelv@qualcomm.com (Michael Vakulenko) writes: I had similar problem with VxWorks v5.2 installed from CD. We have 68K target with Solaris host. The problem was that INSTALL script has incorrectly defined links in "$VX_GNU_BASE/gnu/solaris.68k/lib/gcc-lib/m68k-wrs-vxworks/cygnus-2.2.3.1/" directory. This directory contains links to execs which will be launched during the compilation (e.g. cc68k starts assembler from here). Michael is right; cc68k (aka gcc) was probably picking up the native cpp which didn't understand the command-line parameters being passed to it. If you're ever debugging a problem like this, try adding the ``-v'' parameter to your cc68k (aka gcc) command-line. cc68k is really just an executor; it first executes the C pre-processor, then the C front-end (cc1), then the assembler and finally the linker. By using -v, it should show you (a) exactly which programs it is calling and (b) what command line parameters it is passing to its sub-programs. It can be very informative. In the case of this particular problem, you might have noticed that just ``cpp'' was being called; normally cpp would have a long path in front of it, like $VX_GNU_BASE/gnu/solaris.68k/lib/m68k-wrs-vxworks/bin/ or whatever. Jason Molenda Cygnus Support --------------------------- Newsgroups: comp.os.vxworks Subject: Transfer from VAXELN to VXworks Date: 19 Jul 1995 23:08:51 GMT From: "Phil S. Wilshire" Organization: Internet On-Ramp, Inc. Message-ID: <3uk3a3$g29@express.ior.com> HI, I am new to The VxWorks world but I am excited by its potential. I am working on high speed high performance rolling mill control systems using Alpha AXPvme CPU's in an Aluminum Rolling Mill. We are looking at replacement systems for our current mill controllers. I am evaluating the feasability of tranferring a project written under VAXELN to VxWorks under DEC Unix ( OSF/1 ) I was wondering if any one else has done this ??? ( They must of otherwise why would DEC API product ) I have some questions if so. 1/ How complete is the API ? Will all the VAXELN system service calls etc be satisfied ?? 2/ How efficient is the resulting code ?? 3/ Our VAXELN stuff was not written to be Posix compliant. Does this matter ? Hope to hear from someone. Email or post. I'll post back if I get any sort of response. Thanks Phil Wilshire --------------------------- Newsgroups: comp.os.vxworks Subject: System clock synchronzation Date: 19 Jul 1995 17:23 PST From: cadfael@reg.triumf.ca (MORRIS, DAVID) Organization: TRIUMF: Tri-University Meson Facility Message-ID: <19JUL199517234271@reg.triumf.ca> Hello What is the recommended method of synchronizing the Real Time Clock used in VxWorks with another host? Is there a network request protocol for a time reference that you then load into VxWorks with clock_settime or do you have to initiate from outside VxWorks with an RPC call or socket? I need to ensure the time stamp inside my data files agrees with the file time and date on an NFS mounted drive. I am mounting and writing files from VxWorks. Thanks in advance for any help. David Morris cadfael@decu16.triumf.ca --------------------------- Newsgroups: comp.os.vxworks,comp.lang.c,comp.realtime Subject: C programmer needed Date: Thu, 20 Jul 1995 03:27:16 GMT From: uhlik@otis.arraycomm.com (Chris Uhlik) Organization: ArrayComm, Inc. Message-ID: Followup-To: comp.os.vxworks Sender: uhlik@arraycomm.com (Chris Uhlik) Are you a super-productive C programmer interested in complex embedded systems? Do you enjoy working with a small group of highly competent engineers on something new where you learn lots of new things every day? Do you want to get into wireless communications? ArrayComm in, San Jose California, is developing base station technology for wireless communications systems based on state-of-the-art signal processing techniques. In particular we build boards packed with DSPs and high performance RISC processors that process signals from arrays of antennas to improve the performance and capacity of cellular telephone systems. These base-stations interface to various voice and data networks (SS7, ISDN, Abis, A, raw T1/E1, etc.) We are looking for a few strong C programmers who enjoy working with a superb team of engineers. If you learn quickly, don't mind reading protocol specifications, work well in a loosely organized setting, and write really good C code, then we have a job for you. Strong pluses: master's degree or higher good understanding of basic mathematics real-time programming experience DSP experience work on Unix workstations (not PC based development) experience bringing up new hardware knowledge of cellular communications protocols (e.g. GSM, AMPS, PACS, etc) While these attributes would be nice, what is really important to us is enthusiasm, hunger to learn new things, and excellent C design and coding skills. Please send your resume (ascii text or postscript) via e-mail to uhlik@arraycomm.com or by fax to (408) 428-9083 Chris --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks SNMP agent questions? Date: 19 Jul 95 18:13:53 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: snmp WindNet Message-ID: <1995Jul19.181354.23471@hrbicf> - -- Hello, I have some questions concerning VxWork's SNMP agent available via the WindNet SNMP option. (The agent supports SNMPv1 and SNMPv2.) (1) What SNMP traps are generated by this agent for SNMPv1? (2) What additional SNMP traps are generated by this agent for SNMPv2? (3) If I have both SNMPv1 and SNMPv2 managers on the network how do I modify the configuration files to give the SNMPv1 manager an expanded view of the mib? (4) How do I get a SNMPv1 manager on a different subnet to see traps from this agent? Thanks in advance. - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: non blocking read on serial port Date: 19 Jul 1995 15:46:59 GMT From: miked@wrs.com (Mike Deliman) Organization: Wind River Systems, Inc. Message-ID: <3uj9dj$1f2@darya.wrs.com> References: <9507181955.AA00218@newkla.kla.com> Dear VxWorld, punamia@newkla.kla.com (Manoj Punamia) writes: >I would appreciate if anyone has any comments or help on the following >problem: > >DESCRIPTION: >I am trying to do non blocking "read" from serial port on 68030 Force board. >Before calling read, I start watchdog timer routine to timeout the read. >The watchdog routine calls ioctl(my_fd, FIOCANCEL,0). This is a fairly common problem. The crux of the issue is that the watchdog routine runs as an ISR, not at task level. From the man page on wdLib: Note that the timeout routine is invoked at interrupt level, rather than in the context of the task. Thus, there are restrictions on what the routine may do. Watchdog routines are constrained to the same rules as interrupt service rou- tines. For example, they may not take semaphores or issue other calls that may block. Attached is a "technical summary" on this problem, pretty much explains it all. Best Regards, -mike - ------------------- begin included file (cut here) ------------------- Summary #8868 AUTHOR: miked - ------------------------------------------------------- DATE: Oct 30 1992 10:13AM TITLE: ioctl(FD,FIOCANCEL,NULL) on tyLib device from watchdog hangs system PROBLEM: tyIoctl takes the device protection semaphore to be able to access pTyDev->rdState.canceled and set it TRUE. SYMPTOM: ioctl(FD, FIOCANCEL,NULL) on a FD opened on a tyLib based device (serial device) called from a watchdog hangs the system. SOLUTION: Write a task, to run at a high priority, which either pends on a semTake, or is taskSuspended. Have the wdog routine, or ISR, do a semGive or a taskResume. See SPR 966. read will return 0, so one may check errno to see if it's S_ioLib_CANCELLED. - -- - ------------------------------------------------------------------------------ Mike Deliman <> miked@wrs.com <> phone: 510-748-4100 <> facs: 510-814-2164 Wind River Systems, 1010 Atlantic Ave, Alameda CA 94501 USA;http://www.wrs.com - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks / VxWindows Problem Date: 20 Jul 1995 09:20:12 GMT From: Colin Ford Organization: Northern Telecom Message-ID: <3ul74c$bl4@bcrkh13.bnr.ca> Hi, I'm new to this game of using a real time system. I'm from a UNIX workstation background and have been thrown in at the deep end of VxWorks/VxWindows and porting our code. Let me fill you in....... We are developing a Motif frontend for one of our products on a HP712. The target platform is a PC75Mhz running VxWorks and VxWindows. We are quite a long way down the development life cycle and have almost finished the GUI. We could not port over to the target platform before as the VxWindows did not support a driver for the screen we have on the laptop PC. So we port it over and run up our lovely interface and then bang ! MWM falls over. Now as I'm new to this debugging in VxWorks seems a nightmare. It appears to either be on a SELECT or a XtDestroyWidget call I can't find out for shure. All I know is that if I take XtDestroyWidget out then it does not go BANG ! But when I do a tt on the mwm task it looks like its in some reply routine, and everything else is in the WaitForSomethingToHappen routine. I don't know if its my code of the mwm or X-Server ?????? Now the guys at vxWorks are looking into this and they are great but I was just wondering if anyone else out there has used VxWindows with their applications and come accross anything like this ????? Also has anyone tried work processes and been sucssessful ? any suggestions would be recived with many thanks.... Col. --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Thu Jul 20 04:00:22 1995 From: daemon@csg.lbl.gov Date: Thu Jul 20 04:00:25 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 20 04:00:17 PDT 1995 Subject: Driver for the SCC2692 DUART Subject: Re: Makefile/compile question Subject: Re: Problem with ld on AXPvme160 VxWorks Subject: Re: Makefile/compile question Subject: Transfer from VAXELN to VXworks Subject: System clock synchronzation Subject: C programmer needed Subject: VxWorks SNMP agent questions? Subject: Re: non blocking read on serial port Subject: VxWorks / VxWindows Problem ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Driver for the SCC2692 DUART Date: 19 Jul 1995 10:42:49 GMT From: Adi Regev Organization: RADVision Ltd. Message-ID: <3uinj9$fa1@radmail.rad.co.il> I want to write a driver for the PHILIPS SCC2692 DUART. I will enjoy some writen code, other info will be accepted in joy! - ------------------------------------------------------------------- Adi Regev RADVision Ltd. adi@radvision.rad.co.il --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Makefile/compile question Date: 19 Jul 1995 15:56:07 GMT From: michaelv@qualcomm.com (Michael Vakulenko) Organization: Qualcomm Israel Message-ID: References: <199507171028.DAA22693@inept.stanford.edu> In article <199507171028.DAA22693@inept.stanford.edu>, cgonter@inept.stanford.edu wrote: > I use to compile code using the simple makefile and when I modified it for > vxWorks 5.2 it gives me the following error: > > > inept:cgonter->mbi:make > gcc-sde -g -DCPU=R3000 -I/devl/vw/h -I/devl/chex/star/drv/mbi - -mcpu=r3000 -O0 -funroll-loops -nostdinc -G 0 -c mbistar.c > cpp: Invalid option `-lang-c' > *** Error code 1 > make: Fatal error: Command failed for target `mbistar.o' > > I had similar problem with VxWorks v5.2 installed from CD. We have 68K target with Solaris host. The problem was that INSTALL script has incorrectly defined links in "$VX_GNU_BASE/gnu/solaris.68k/lib/gcc-lib/m68k-wrs-vxworks/cygnus-2.2.3.1/" directory. This directory contains links to execs which will be launched during the compilation (e.g. cc68k starts assembler from here). After the installation, the links pointed nowhere in our case and the compiler failed to start the correct exec. In our configuration, the links should point to "$VX_GNU_BASE/gnu/solaris.68k/lib/m68k-wrs-vxworks/bin/" directory. Of course, you have different directory names, but I hope you've got an idea. ______________________________________________________________ Michael Vakulenko, Voice: +972-4-577999 Software Engineer, Fax: +972-4-577998 Qualcomm Israel, Ltd. Email: michaelv@qualcomm.com ______________________________________________________________ __________________________________________________________ Michael Vakulenko, Voice: +972-4-577999 Software Engineer, Fax: +972-4-577998 Qualcomm Israel, Ltd Email: michaelv@qualcomm.com __________________________________________________________ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Problem with ld on AXPvme160 VxWorks Date: 19 Jul 1995 18:48:41 GMT From: Ron Reid Organization: Honeywell Message-ID: <3ujk29$as@killerbee.jsc.nasa.gov> References: <3ugkgh$ftu@geraldo.cc.utexas.edu> This is what works for me on a similar setup (replace "ws04", "193.0.0.1", "myobj.o" and "/mnt/usr/mydir" with your info) : vxworks> hostAdd "ws04","193.0.0.1" vxworks> netDevCreate "ws04:", "ws04", 1 vxworks> cd ws04:/mnt/usr/mydir vxworks> ld References: <199507171028.DAA22693@inept.stanford.edu> In article michaelv@qualcomm.com (Michael Vakulenko) writes: I had similar problem with VxWorks v5.2 installed from CD. We have 68K target with Solaris host. The problem was that INSTALL script has incorrectly defined links in "$VX_GNU_BASE/gnu/solaris.68k/lib/gcc-lib/m68k-wrs-vxworks/cygnus-2.2.3.1/" directory. This directory contains links to execs which will be launched during the compilation (e.g. cc68k starts assembler from here). Michael is right; cc68k (aka gcc) was probably picking up the native cpp which didn't understand the command-line parameters being passed to it. If you're ever debugging a problem like this, try adding the ``-v'' parameter to your cc68k (aka gcc) command-line. cc68k is really just an executor; it first executes the C pre-processor, then the C front-end (cc1), then the assembler and finally the linker. By using -v, it should show you (a) exactly which programs it is calling and (b) what command line parameters it is passing to its sub-programs. It can be very informative. In the case of this particular problem, you might have noticed that just ``cpp'' was being called; normally cpp would have a long path in front of it, like $VX_GNU_BASE/gnu/solaris.68k/lib/m68k-wrs-vxworks/bin/ or whatever. Jason Molenda Cygnus Support --------------------------- Newsgroups: comp.os.vxworks Subject: Transfer from VAXELN to VXworks Date: 19 Jul 1995 23:08:51 GMT From: "Phil S. Wilshire" Organization: Internet On-Ramp, Inc. Message-ID: <3uk3a3$g29@express.ior.com> HI, I am new to The VxWorks world but I am excited by its potential. I am working on high speed high performance rolling mill control systems using Alpha AXPvme CPU's in an Aluminum Rolling Mill. We are looking at replacement systems for our current mill controllers. I am evaluating the feasability of tranferring a project written under VAXELN to VxWorks under DEC Unix ( OSF/1 ) I was wondering if any one else has done this ??? ( They must of otherwise why would DEC API product ) I have some questions if so. 1/ How complete is the API ? Will all the VAXELN system service calls etc be satisfied ?? 2/ How efficient is the resulting code ?? 3/ Our VAXELN stuff was not written to be Posix compliant. Does this matter ? Hope to hear from someone. Email or post. I'll post back if I get any sort of response. Thanks Phil Wilshire --------------------------- Newsgroups: comp.os.vxworks Subject: System clock synchronzation Date: 19 Jul 1995 17:23 PST From: cadfael@reg.triumf.ca (MORRIS, DAVID) Organization: TRIUMF: Tri-University Meson Facility Message-ID: <19JUL199517234271@reg.triumf.ca> Hello What is the recommended method of synchronizing the Real Time Clock used in VxWorks with another host? Is there a network request protocol for a time reference that you then load into VxWorks with clock_settime or do you have to initiate from outside VxWorks with an RPC call or socket? I need to ensure the time stamp inside my data files agrees with the file time and date on an NFS mounted drive. I am mounting and writing files from VxWorks. Thanks in advance for any help. David Morris cadfael@decu16.triumf.ca --------------------------- Newsgroups: comp.os.vxworks,comp.lang.c,comp.realtime Subject: C programmer needed Date: Thu, 20 Jul 1995 03:27:16 GMT From: uhlik@otis.arraycomm.com (Chris Uhlik) Organization: ArrayComm, Inc. Message-ID: Followup-To: comp.os.vxworks Sender: uhlik@arraycomm.com (Chris Uhlik) Are you a super-productive C programmer interested in complex embedded systems? Do you enjoy working with a small group of highly competent engineers on something new where you learn lots of new things every day? Do you want to get into wireless communications? ArrayComm in, San Jose California, is developing base station technology for wireless communications systems based on state-of-the-art signal processing techniques. In particular we build boards packed with DSPs and high performance RISC processors that process signals from arrays of antennas to improve the performance and capacity of cellular telephone systems. These base-stations interface to various voice and data networks (SS7, ISDN, Abis, A, raw T1/E1, etc.) We are looking for a few strong C programmers who enjoy working with a superb team of engineers. If you learn quickly, don't mind reading protocol specifications, work well in a loosely organized setting, and write really good C code, then we have a job for you. Strong pluses: master's degree or higher good understanding of basic mathematics real-time programming experience DSP experience work on Unix workstations (not PC based development) experience bringing up new hardware knowledge of cellular communications protocols (e.g. GSM, AMPS, PACS, etc) While these attributes would be nice, what is really important to us is enthusiasm, hunger to learn new things, and excellent C design and coding skills. Please send your resume (ascii text or postscript) via e-mail to uhlik@arraycomm.com or by fax to (408) 428-9083 Chris --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks SNMP agent questions? Date: 19 Jul 95 18:13:53 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: snmp WindNet Message-ID: <1995Jul19.181354.23471@hrbicf> - -- Hello, I have some questions concerning VxWork's SNMP agent available via the WindNet SNMP option. (The agent supports SNMPv1 and SNMPv2.) (1) What SNMP traps are generated by this agent for SNMPv1? (2) What additional SNMP traps are generated by this agent for SNMPv2? (3) If I have both SNMPv1 and SNMPv2 managers on the network how do I modify the configuration files to give the SNMPv1 manager an expanded view of the mib? (4) How do I get a SNMPv1 manager on a different subnet to see traps from this agent? Thanks in advance. - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: non blocking read on serial port Date: 19 Jul 1995 15:46:59 GMT From: miked@wrs.com (Mike Deliman) Organization: Wind River Systems, Inc. Message-ID: <3uj9dj$1f2@darya.wrs.com> References: <9507181955.AA00218@newkla.kla.com> Dear VxWorld, punamia@newkla.kla.com (Manoj Punamia) writes: >I would appreciate if anyone has any comments or help on the following >problem: > >DESCRIPTION: >I am trying to do non blocking "read" from serial port on 68030 Force board. >Before calling read, I start watchdog timer routine to timeout the read. >The watchdog routine calls ioctl(my_fd, FIOCANCEL,0). This is a fairly common problem. The crux of the issue is that the watchdog routine runs as an ISR, not at task level. From the man page on wdLib: Note that the timeout routine is invoked at interrupt level, rather than in the context of the task. Thus, there are restrictions on what the routine may do. Watchdog routines are constrained to the same rules as interrupt service rou- tines. For example, they may not take semaphores or issue other calls that may block. Attached is a "technical summary" on this problem, pretty much explains it all. Best Regards, -mike - ------------------- begin included file (cut here) ------------------- Summary #8868 AUTHOR: miked - ------------------------------------------------------- DATE: Oct 30 1992 10:13AM TITLE: ioctl(FD,FIOCANCEL,NULL) on tyLib device from watchdog hangs system PROBLEM: tyIoctl takes the device protection semaphore to be able to access pTyDev->rdState.canceled and set it TRUE. SYMPTOM: ioctl(FD, FIOCANCEL,NULL) on a FD opened on a tyLib based device (serial device) called from a watchdog hangs the system. SOLUTION: Write a task, to run at a high priority, which either pends on a semTake, or is taskSuspended. Have the wdog routine, or ISR, do a semGive or a taskResume. See SPR 966. read will return 0, so one may check errno to see if it's S_ioLib_CANCELLED. - -- - ------------------------------------------------------------------------------ Mike Deliman <> miked@wrs.com <> phone: 510-748-4100 <> facs: 510-814-2164 Wind River Systems, 1010 Atlantic Ave, Alameda CA 94501 USA;http://www.wrs.com - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks / VxWindows Problem Date: 20 Jul 1995 09:20:12 GMT From: Colin Ford Organization: Northern Telecom Message-ID: <3ul74c$bl4@bcrkh13.bnr.ca> Hi, I'm new to this game of using a real time system. I'm from a UNIX workstation background and have been thrown in at the deep end of VxWorks/VxWindows and porting our code. Let me fill you in....... We are developing a Motif frontend for one of our products on a HP712. The target platform is a PC75Mhz running VxWorks and VxWindows. We are quite a long way down the development life cycle and have almost finished the GUI. We could not port over to the target platform before as the VxWindows did not support a driver for the screen we have on the laptop PC. So we port it over and run up our lovely interface and then bang ! MWM falls over. Now as I'm new to this debugging in VxWorks seems a nightmare. It appears to either be on a SELECT or a XtDestroyWidget call I can't find out for shure. All I know is that if I take XtDestroyWidget out then it does not go BANG ! But when I do a tt on the mwm task it looks like its in some reply routine, and everything else is in the WaitForSomethingToHappen routine. I don't know if its my code of the mwm or X-Server ?????? Now the guys at vxWorks are looking into this and they are great but I was just wondering if anyone else out there has used VxWindows with their applications and come accross anything like this ????? Also has anyone tried work processes and been sucssessful ? any suggestions would be recived with many thanks.... Col. --------------------------- End of New-News digest ********************** Subject: VxWorks on MIPS controllers Submitted-by rhin!wrsec.fr!joergb@lbl.gov Thu Jul 20 08:29:55 1995 Submitted-by: rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) Greetings, is anybody willing to share his experiance with VxWorks on a MIPS controller (e.g. R3051, R3081). Drivers for on-chip devices are of special interest. Thanks for any info. Joerg ************************************************************************* * Joerg Bertholdt Wind River Systems GmbH * * Field Application Engineer Freisinger Strasse 34 * * Email: joergb@wrsec.fr D-85737 Ismaning * * Tel: +49-89-962445-42 Fax: +49-89-962445-55 * * http://www.wrs.com * ************************************************************************* From rhin!wrsec.fr!joergb@lbl.gov Thu Jul 20 08:29:55 1995 From: rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) Date: Thu Jul 20 08:29:57 PDT 1995 Subject: VxWorks on MIPS controllers Greetings, is anybody willing to share his experiance with VxWorks on a MIPS controller (e.g. R3051, R3081). Drivers for on-chip devices are of special interest. Thanks for any info. Joerg ************************************************************************* * Joerg Bertholdt Wind River Systems GmbH * * Field Application Engineer Freisinger Strasse 34 * * Email: joergb@wrsec.fr D-85737 Ismaning * * Tel: +49-89-962445-42 Fax: +49-89-962445-55 * * http://www.wrs.com * ************************************************************************* Subject: NTDS-B Support under VxWorks Submitted-by mstull@columbia1.fci.com Thu Jul 20 09:11:52 1995 Submitted-by: mstull@columbia1.fci.com Hello VxWorkers, Does anyone know of any NTDS-B boards which are supported under vxWorks? I think someone sent a message on this recently. Any info is greatly appreciated. Thanks, ********************************************************************** Mark Stull phone:(410) 674-0003 Field Applications Engineer fax: (410) 674-0008 FORCE COMPUTERS, Inc. email: mstull@fci.com Northeast U.S. Sales/MD ********************************************************************** From mstull@columbia1.fci.com Thu Jul 20 09:11:52 1995 From: mstull@columbia1.fci.com Date: Thu Jul 20 09:11:55 PDT 1995 Subject: NTDS-B Support under VxWorks Hello VxWorkers, Does anyone know of any NTDS-B boards which are supported under vxWorks? I think someone sent a message on this recently. Any info is greatly appreciated. Thanks, ********************************************************************** Mark Stull phone:(410) 674-0003 Field Applications Engineer fax: (410) 674-0008 FORCE COMPUTERS, Inc. email: mstull@fci.com Northeast U.S. Sales/MD ********************************************************************** Subject: UPDATE info: booting vxworks on Solaris system Submitted-by cgonter@inept.stanford.edu Thu Jul 20 10:29:25 1995 Submitted-by: cgonter@inept.stanford.edu IN THE PREVIOUS POSTING: > What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > > then executed > > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > > then check the nfs system by mounting the partition onto another > workstation. > > > The results are: > > target -> > > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > New info .... The .rhosts file has been in the hosts users area(cgonter) with the following permissions: -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* .rhosts contains --> inept:cgonter->cgonter:more .rhosts emprise cgonter dyslexic + whoosh + Error return is the same as above: > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > ADDED /etc/hosts.equiv and the results are: > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... The target system whoosh hangs at this point. .... thanks for all the previous responses craig cgonter@inept.stanford.edu From cgonter@inept.stanford.edu Thu Jul 20 10:29:25 1995 From: cgonter@inept.stanford.edu Date: Thu Jul 20 10:29:28 PDT 1995 Subject: UPDATE info: booting vxworks on Solaris system IN THE PREVIOUS POSTING: > What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > > then executed > > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > > then check the nfs system by mounting the partition onto another > workstation. > > > The results are: > > target -> > > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > New info .... The .rhosts file has been in the hosts users area(cgonter) with the following permissions: -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* .rhosts contains --> inept:cgonter->cgonter:more .rhosts emprise cgonter dyslexic + whoosh + Error return is the same as above: > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. > ADDED /etc/hosts.equiv and the results are: > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... The target system whoosh hangs at this point. .... thanks for all the previous responses craig cgonter@inept.stanford.edu Subject: Re: UPDATE info: booting vxworks on Solaris system Submitted-by stan@rti.com Thu Jul 20 13:47:23 1995 Submitted-by: Stan Schneider >> inept:cgonter->cgonter:more .rhosts >> emprise cgonter >> dyslexic + >> whoosh + >> >> >> Error return is the same as above: >> >> > Attaching network interface egl0... done. >> > Attaching network interface lo0... done. >> > Loading... permission denied >> > >> > Error loading file: errno = 0x250002. >> > >> >> ADDED /etc/hosts.equiv and the results are: >> >> >> > Attaching network interface egl0... done. >> > Attaching network interface lo0... done. >> > Loading... You don't want the "+" sign in your .rhosts file, just the target hostname. The optional username in a .rhosts file allows a differently-named user to access your account. For example, you've allowed cgonter on emprise to rlogin as cgonter on inept (which will work). If you had "emprise fish", then fish@emprise could rlogin as cgonter to inept. I don't think the "+" syntax is allowed in .rhosts; it doesn't really make sense. The username means something different in /etc/hosts.equiv; there the username specifies specific users with rlogin capability as like-named users on your host. In other words, "whoosh +" is identical to just "whoosh" only in /etc/hosts.equiv. Anyhow, that explains your permissions problem. >> >> The target system whoosh hangs at this point. >> Now you're loaded, and the system is trying to get the symbol table, via nfs if you've compiled in that option (which you apparently have, since otherwise you get a notice that nfs isn't installed). The fact that you get nothing may mean you've got a bad network configuration; my bet is on the netmask. It's also possible that you loaded garbage. Did you get the "starting at" message? Are you monitoring the serial line? Does it resond to ^C? Can you still ping the target? Are the run lights still on? Is this a "vanilla" kernel? Is your BSP correctly configured? HTH, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Thu Jul 20 13:47:23 1995 From: Stan Schneider Date: Thu Jul 20 13:47:25 PDT 1995 Subject: Re: UPDATE info: booting vxworks on Solaris system >> inept:cgonter->cgonter:more .rhosts >> emprise cgonter >> dyslexic + >> whoosh + >> >> >> Error return is the same as above: >> >> > Attaching network interface egl0... done. >> > Attaching network interface lo0... done. >> > Loading... permission denied >> > >> > Error loading file: errno = 0x250002. >> > >> >> ADDED /etc/hosts.equiv and the results are: >> >> >> > Attaching network interface egl0... done. >> > Attaching network interface lo0... done. >> > Loading... You don't want the "+" sign in your .rhosts file, just the target hostname. The optional username in a .rhosts file allows a differently-named user to access your account. For example, you've allowed cgonter on emprise to rlogin as cgonter on inept (which will work). If you had "emprise fish", then fish@emprise could rlogin as cgonter to inept. I don't think the "+" syntax is allowed in .rhosts; it doesn't really make sense. The username means something different in /etc/hosts.equiv; there the username specifies specific users with rlogin capability as like-named users on your host. In other words, "whoosh +" is identical to just "whoosh" only in /etc/hosts.equiv. Anyhow, that explains your permissions problem. >> >> The target system whoosh hangs at this point. >> Now you're loaded, and the system is trying to get the symbol table, via nfs if you've compiled in that option (which you apparently have, since otherwise you get a notice that nfs isn't installed). The fact that you get nothing may mean you've got a bad network configuration; my bet is on the netmask. It's also possible that you loaded garbage. Did you get the "starting at" message? Are you monitoring the serial line? Does it resond to ^C? Can you still ping the target? Are the run lights still on? Is this a "vanilla" kernel? Is your BSP correctly configured? HTH, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: Re: NTDS-B Support under VxWorks Submitted-by swtillman@ccgate.hac.com Thu Jul 20 15:17:00 1995 Submitted-by: swtillman@ccgate.hac.com mstull@columbia1.fci.com wrote: > Does anyone know of any NTDS-B boards which are supported under vxWorks? > I think someone sent a message on this recently. Any info is greatly > appreciated. Mark, You might try GET Engineering Corp. Last I talked to them, they had drivers for SUN OS, AT&T, VRTX. Driver for VxWorks was soon to be released. I have not used their VME boards, but I'm about to use their PC-based product in an upcoming project. Here is their address. Hope this helps... GET Engineering Corp. 9350 Bond Ave. El Cajon, CA 92021 (619)443-8295 -Scott W. Tillman -Hughes Missile Systems Co. -RAM GMLS Software Design From swtillman@ccgate.hac.com Thu Jul 20 15:17:00 1995 From: swtillman@ccgate.hac.com Date: Thu Jul 20 15:17:03 PDT 1995 Subject: Re: NTDS-B Support under VxWorks mstull@columbia1.fci.com wrote: > Does anyone know of any NTDS-B boards which are supported under vxWorks? > I think someone sent a message on this recently. Any info is greatly > appreciated. Mark, You might try GET Engineering Corp. Last I talked to them, they had drivers for SUN OS, AT&T, VRTX. Driver for VxWorks was soon to be released. I have not used their VME boards, but I'm about to use their PC-based product in an upcoming project. Here is their address. Hope this helps... GET Engineering Corp. 9350 Bond Ave. El Cajon, CA 92021 (619)443-8295 -Scott W. Tillman -Hughes Missile Systems Co. -RAM GMLS Software Design Subject: Re: boot error 0x1a9 Submitted-by uri.porat@indigo.co.il Fri Jul 21 00:47:38 1995 Submitted-by: uri.porat@indigo.co.il (Uri Porat) Hello John Follows a full description of the problem you met. I spent a lot of energy trying to solve it. I used the assistance of Mr. Leonid Rosenboim (RST Software Industries Israel) and Mr. Kelly Caudill from the Sun's support center at SunLtd. ============================ Problem analysis ====================================== 1. The problem arises whenever a reboot is perform, close enough (time domain) to previous boot. 2. VxWorks is using FTP to fetch both OS and startup script file. When failure is reported, it is the first file (OS) that fails. 3. When using "snoop", the FTP-server reports error 425: "425 Can't build data connection: Cannot assign requested add" 4. VxWorks whenever perform an FTP boot, establishes the FTP-data-link through port 1025 _ALWAYS_ i.e.: command-link: port 1024 (VxWorks) - port 21 (FTP-server) 1st-data-link: port 1025 (VxWorks) - port 20 (FTP-server) # OS file 2nd-data-link: port 1027 (VxWorks) - port 20 (FTP-server) # startup file 3nd-data-link: port 1028 (VxWorks) - port 20 (FTP-server) # application follows netstat info: Local Address Remote Address Swind Send-Q Rwind Recv-Q State -------------------- -------------------- ----- ------ ----- ------ ------- 192.168.100.30.20 192.168.100.32.1025 4096 0 25600 0 TIME_WAIT 192.168.100.30.20 192.168.100.32.1027 4096 0 25600 0 TIME_WAIT 192.168.100.30.20 192.168.100.32.1029 4096 0 25600 0 TIME_WAIT ..... ============================== SunSolve "bug report" =============================== Bug Id: 1176794 Category: network Subcategory: internet-utility Release summary: s1093 Synopsis: ftp fails after several iterations Integrated in releases: Patch id: Description: The FTP client machine (a "vxworks" system) boots using FTP. The snoop trace shows a series of successful "boot operations". Each time, it makes the control connection, and then the data connection, and then procedes to transfer the data and all is well. When the failure occurs, the snoop trace shows the control connection, and then, in response to the PORT command from the client, the server sends back: 425 Can't build data connection: Cannot assign requested address The addresses are the same. In each PORT command, the client specifies 1025 because, each time, the client just booting up and so 1025 is the next free port. In ftpd, I only found one place where "425 Can't build data connection" is returned. There are three "425..." returns, but the associated text makes them unique. This one expects success or EADDRINUSE in response to the connect() call. If it gets an error and it is EADDRINUSE, it waits and retries up to a limit. If the limit is reached or any other error returned, the error is appended to the end of the "425 Can't build data connection" message and returned to the client. So it is getting EADDRNOTAVAIL instead of EADDRINUSE. Bug 1141225 is similar but not the same. Why would it suddenly get EADDRNOTAVAIL after having done exactly the same thing several times in a row? History: Submitter: kelly.caudill@east Date: 09/08/94 Dispatch Operator: bugtraq Date: 09/08/94 ============================== Kelly Caudill reply ============================== > I just found your bug report (Bug Id: 1176794) regarding VxWorks, on the > sunSolve. > > It is already several days that I spend dealing with this problem. > > It seems to me that two major limitations make this bug so hard to solve: > > 1. Solaris (is it just a SOLARIS problem?) does not close the FTP-DATA > connection although the FTP protocol is fulfilled Does it really not close the connection? Or does it leave it in TIME_WAIT or some other state? What is shown by netstat -f inet ? > 2. On the other side, VxWorks always use port 1025 to establish FTP_DATA > connection. Since this connection is not closed by SOLARIS FTP fails If the connection on the Solaris side is in one of the normal "shutdown but wait a short time" states, then the VxWorks system should get a reset back. You can reduce the time these connections stay in these post-shutdown states with ndd. ndd /dev/tcp will list the variables you can set. ndd /dev/tcp name will list the current value of that name. The following, despite what its name implies, controls how long an endpoint stays in the TIME_WAIT state. ndd -set /dev/tcp tcp_close_wait_interval > Do you agree with my assumptions? Mostly. But we should see what state the endpoints are in - netstat -af inet > Could you please tell me how did (if at all) you solve this problem? Did > you give-up FTP and used another protocol? Did SUN support you with any > kind of patch? I work in the support center here at Sun. Here is some info from the comments part of the bug report, which you probably do not get in the publicly available version. I will update the summary in the bug report with some of this: I think we can close this bug. The EADDRNOTAVAIL is a feature: If the FTP client machine reboots, comes back up, and starts an FTP again while the server thinks an FTP is still in progress - as the trace shows - the server will probably still be trying to send data on the data connection and one of these packets will cause the client to send back a reset which will completely free up the previous connection. The transition would be from ESTABLISHED to CLOSED by way of tcpabort() and then, very soon thereafter, to LISTEN and on toward ESTABLISHED again just like normal. However, if the transfer completes before the client is rebooted, the previous connection is in TIME_WAIT and cannot be restarted until it is cleaned up after tcp_close_wait_interval which defaults to 240,000 milliseconds (4 minutes).. So, reducing tcp_close_wait_interval (to maybe 10 seconds - or less) might be a fine sollution to his problem. The customer was trying to reproduce his problem by rebooting the vxworks box constantly. When he finally happened to reboot it just *after* it had successfully booted, it displayed the same symptoms as the real problem. However, it is not the same as the real problem which will probably turn out to be a configuration error or bug in the other system. I say this is a feature because it is correct for us to keep the end-point in the TIME_WAIT state for some time after a normal close. For more information on this, refer to the sections about the TCP finite state machine in Douglas E. Comer's and Daved L. Stevens' book, Internetworking with TCP/IP Volume II. Normally, this would not be an issue. A connection is identified by the IP address, protocol, and port number of both end points. Since a subsequent connection from a client would normally involve a new port number, there is no problem with holding onto a the end-point on the server for a short time. However, when, as in this case, the client reboots and starts over with the same port number, you end up with this problem. A workaround is to reduce the time spent in TIME_WAIT state with the tcp_close_wait_interval ndd parameter. Kelly /\ -------------------------------------------------------- \\ \ Kelly Caudill \ \\ / North American Solution Center / \/ / / Sun Microsystems / / \//\ 5 Omni Way \//\ / / Chelmsford, MA 01824-4141 / / /\ / / \\ \ Phone: (508) 442-1343 Fax: (508) 442-1418 \ \\ EMail: Kelly.Caudill@East.Sun.Com \/ ============================== ---------- ============================== ============================== My summary ============================== ============================== ---------- ============================== These mail-exchanges are very "fresh" and therefore I had not yet enough time to test this solution. But I hope this information will help you. BTW, our configuration is Motorola MV162 and a Sun SPARCclassic (Solaris). .--------------------------------------------------------------------. | Uri Porat, Software engineer Tel: (972-8) 381-996/818 | | Indigo LTD, POB 150, FAX: (972-8) 381-341/338 | | Kiryat Weizmann, E-mail: uri.porat@Indigo.co.il | | Rehovot 76101, Israel "My views are my own" | '--------------------------------------------------------------------' From uri.porat@indigo.co.il Fri Jul 21 00:47:38 1995 From: uri.porat@indigo.co.il (Uri Porat) Date: Fri Jul 21 00:47:41 PDT 1995 Subject: Re: boot error 0x1a9 Hello John Follows a full description of the problem you met. I spent a lot of energy trying to solve it. I used the assistance of Mr. Leonid Rosenboim (RST Software Industries Israel) and Mr. Kelly Caudill from the Sun's support center at SunLtd. ============================ Problem analysis ====================================== 1. The problem arises whenever a reboot is perform, close enough (time domain) to previous boot. 2. VxWorks is using FTP to fetch both OS and startup script file. When failure is reported, it is the first file (OS) that fails. 3. When using "snoop", the FTP-server reports error 425: "425 Can't build data connection: Cannot assign requested add" 4. VxWorks whenever perform an FTP boot, establishes the FTP-data-link through port 1025 _ALWAYS_ i.e.: command-link: port 1024 (VxWorks) - port 21 (FTP-server) 1st-data-link: port 1025 (VxWorks) - port 20 (FTP-server) # OS file 2nd-data-link: port 1027 (VxWorks) - port 20 (FTP-server) # startup file 3nd-data-link: port 1028 (VxWorks) - port 20 (FTP-server) # application follows netstat info: Local Address Remote Address Swind Send-Q Rwind Recv-Q State -------------------- -------------------- ----- ------ ----- ------ ------- 192.168.100.30.20 192.168.100.32.1025 4096 0 25600 0 TIME_WAIT 192.168.100.30.20 192.168.100.32.1027 4096 0 25600 0 TIME_WAIT 192.168.100.30.20 192.168.100.32.1029 4096 0 25600 0 TIME_WAIT ..... ============================== SunSolve "bug report" =============================== Bug Id: 1176794 Category: network Subcategory: internet-utility Release summary: s1093 Synopsis: ftp fails after several iterations Integrated in releases: Patch id: Description: The FTP client machine (a "vxworks" system) boots using FTP. The snoop trace shows a series of successful "boot operations". Each time, it makes the control connection, and then the data connection, and then procedes to transfer the data and all is well. When the failure occurs, the snoop trace shows the control connection, and then, in response to the PORT command from the client, the server sends back: 425 Can't build data connection: Cannot assign requested address The addresses are the same. In each PORT command, the client specifies 1025 because, each time, the client just booting up and so 1025 is the next free port. In ftpd, I only found one place where "425 Can't build data connection" is returned. There are three "425..." returns, but the associated text makes them unique. This one expects success or EADDRINUSE in response to the connect() call. If it gets an error and it is EADDRINUSE, it waits and retries up to a limit. If the limit is reached or any other error returned, the error is appended to the end of the "425 Can't build data connection" message and returned to the client. So it is getting EADDRNOTAVAIL instead of EADDRINUSE. Bug 1141225 is similar but not the same. Why would it suddenly get EADDRNOTAVAIL after having done exactly the same thing several times in a row? History: Submitter: kelly.caudill@east Date: 09/08/94 Dispatch Operator: bugtraq Date: 09/08/94 ============================== Kelly Caudill reply ============================== > I just found your bug report (Bug Id: 1176794) regarding VxWorks, on the > sunSolve. > > It is already several days that I spend dealing with this problem. > > It seems to me that two major limitations make this bug so hard to solve: > > 1. Solaris (is it just a SOLARIS problem?) does not close the FTP-DATA > connection although the FTP protocol is fulfilled Does it really not close the connection? Or does it leave it in TIME_WAIT or some other state? What is shown by netstat -f inet ? > 2. On the other side, VxWorks always use port 1025 to establish FTP_DATA > connection. Since this connection is not closed by SOLARIS FTP fails If the connection on the Solaris side is in one of the normal "shutdown but wait a short time" states, then the VxWorks system should get a reset back. You can reduce the time these connections stay in these post-shutdown states with ndd. ndd /dev/tcp will list the variables you can set. ndd /dev/tcp name will list the current value of that name. The following, despite what its name implies, controls how long an endpoint stays in the TIME_WAIT state. ndd -set /dev/tcp tcp_close_wait_interval > Do you agree with my assumptions? Mostly. But we should see what state the endpoints are in - netstat -af inet > Could you please tell me how did (if at all) you solve this problem? Did > you give-up FTP and used another protocol? Did SUN support you with any > kind of patch? I work in the support center here at Sun. Here is some info from the comments part of the bug report, which you probably do not get in the publicly available version. I will update the summary in the bug report with some of this: I think we can close this bug. The EADDRNOTAVAIL is a feature: If the FTP client machine reboots, comes back up, and starts an FTP again while the server thinks an FTP is still in progress - as the trace shows - the server will probably still be trying to send data on the data connection and one of these packets will cause the client to send back a reset which will completely free up the previous connection. The transition would be from ESTABLISHED to CLOSED by way of tcpabort() and then, very soon thereafter, to LISTEN and on toward ESTABLISHED again just like normal. However, if the transfer completes before the client is rebooted, the previous connection is in TIME_WAIT and cannot be restarted until it is cleaned up after tcp_close_wait_interval which defaults to 240,000 milliseconds (4 minutes).. So, reducing tcp_close_wait_interval (to maybe 10 seconds - or less) might be a fine sollution to his problem. The customer was trying to reproduce his problem by rebooting the vxworks box constantly. When he finally happened to reboot it just *after* it had successfully booted, it displayed the same symptoms as the real problem. However, it is not the same as the real problem which will probably turn out to be a configuration error or bug in the other system. I say this is a feature because it is correct for us to keep the end-point in the TIME_WAIT state for some time after a normal close. For more information on this, refer to the sections about the TCP finite state machine in Douglas E. Comer's and Daved L. Stevens' book, Internetworking with TCP/IP Volume II. Normally, this would not be an issue. A connection is identified by the IP address, protocol, and port number of both end points. Since a subsequent connection from a client would normally involve a new port number, there is no problem with holding onto a the end-point on the server for a short time. However, when, as in this case, the client reboots and starts over with the same port number, you end up with this problem. A workaround is to reduce the time spent in TIME_WAIT state with the tcp_close_wait_interval ndd parameter. Kelly /\ -------------------------------------------------------- \\ \ Kelly Caudill \ \\ / North American Solution Center / \/ / / Sun Microsystems / / \//\ 5 Omni Way \//\ / / Chelmsford, MA 01824-4141 / / /\ / / \\ \ Phone: (508) 442-1343 Fax: (508) 442-1418 \ \\ EMail: Kelly.Caudill@East.Sun.Com \/ ============================== ---------- ============================== ============================== My summary ============================== ============================== ---------- ============================== These mail-exchanges are very "fresh" and therefore I had not yet enough time to test this solution. But I hope this information will help you. BTW, our configuration is Motorola MV162 and a Sun SPARCclassic (Solaris). .--------------------------------------------------------------------. | Uri Porat, Software engineer Tel: (972-8) 381-996/818 | | Indigo LTD, POB 150, FAX: (972-8) 381-341/338 | | Kiryat Weizmann, E-mail: uri.porat@Indigo.co.il | | Rehovot 76101, Israel "My views are my own" | '--------------------------------------------------------------------' Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Fri Jul 21 04:00:18 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Fri Jul 21 04:00:14 PDT 1995 Subject: ECC DRAM vs Parity DRAM ??? Subject: malloc more than 1 MB, how ? Subject: booting (errno = 0x3d) Subject: PCMCIA for 68xxx ? Subject: Wind C++ Subject: open files take up memory Subject: JPL s/w eng. position Subject: Re: booting vxworks on Solaris system Subject: C/C++ choice ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: ECC DRAM vs Parity DRAM ??? Date: 20 Jul 1995 10:57:26 GMT From: kccheng@hycpsg01 (Kuang-chun Cheng) Organization: FERMILAB, Batavia, IL Message-ID: <3ulcqm$dpd@fnnews.fnal.gov> Hi folks, Let me describle my problem first. We have one RMF3573 board connect one Exabyte 8505 8mm tape device. We use MV167 for our VME master CPU. Couple month ago, we finished the driver of RMF3573 and tested the tape writing rate. The result is about 0.49 MB/sec the same as the Exabyte 8505 User's Manual said. Recently, we got a new MV167-32B board. The first thing to do for us is test the RMF3573 driver and tape writing rate again on the new board. But this time the max. rate we got is 0.38 MB/sec. We spend a lot of time try to figure out what happen, but ... The only possibility we got at this moment is come from DRAM. The MV167 is a 25 MHz with 4MB DRAM (parity) cpu board while MV167-32B is a 33 MHz with 8 MB ECC DRAM cpu board. I guess ECC DRAM do more stuff than parity DRAM so it slow down the tape writing rate (since Exabyte 8505 must access ECC DRAM to get data). But I am not sure ... I did enable the FSTRD bit at MCECC Defaults Register 1 and set MCECC BCLK Frequency register to 33 MHz (as MV167 programmer's ref. guide said, this will push DRAM into full speed operating). But I can't get any improvement. Any suggestion is appreciated. Thanks. ________________ Kuang-chun Cheng kccheng@phys.sinica.edu.tw http://hycpsg01.fnal.gov/kccheng/kccheng.html --------------------------- Newsgroups: comp.os.vxworks Subject: malloc more than 1 MB, how ? Date: 20 Jul 1995 11:09:57 GMT From: kccheng@hycpsg01 (Kuang-chun Cheng) Organization: FERMILAB, Batavia, IL Message-ID: <3uldi5$dpd@fnnews.fnal.gov> Hi, I try to alloc more than 1 MB memory by using malloc() command on MV167 running vxWorks 5.11. But it complains memory size too big. Can I allocate a memory area more than 1 MB ? How to do that ? Thanks a lot. ________________ Kuang-chun Cheng kccheng@phys.sinica.edu.tw --------------------------- Newsgroups: comp.os.vxworks Subject: booting (errno = 0x3d) Date: 20 Jul 1995 16:55:08 GMT From: "Matthew W. K. Brown" Organization: Sandia National Laboratories Message-ID: <3um1pc$o6b@news.sandia.gov> References: <199507191717.KAA10688@inept.stanford.edu> Since the group has been on the topic of booting lately, I thought that I'd ask about this problem that we've been having ... Whenever we boot up 4 or more (sometimes it only takes 3) of our targets at once we get (at various points in the booting process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). I understand that there are limited resources and all of the targets are competing for them; however, I would rather that they wait a bit longer (instead of just dying). Surely, this is a common problem whenever a big system is powering up ... so does anyone know the easy solution (like do I just bump up some timeout value)? Thanks, Matthew Brown mwbrown@sandia.gov --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: PCMCIA for 68xxx ? Date: 20 Jul 1995 16:41:13 GMT From: fitz@aim.cse.tek.com (Bob Fitzsimmons) Organization: Tektronix, Inc., Beaverton, OR. Message-ID: <3um0v9$2h5@tekadm1.cse.tek.com> Reply-To: fitz@aim.cse.tek.com I'm working on a product with an embedded 68xxx system. We're interested in supporting several peripheral devices including: extended memory (ram disks), and serial ports. Because of our small form factor, we're most interested in providing this support via PCMCIA cards. Questions: 1. Which of the available bus adapters is most compatible with 68xxx? 2. Do any vendors provide 68xxx driver support? Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx embedded environment would be greatly appreciated. Thanks in advance, Bob Fitzsimmons Tektonix, Inc. Measurement Business Division (503) 627-5777 --------------------------- Newsgroups: comp.os.vxworks Subject: Wind C++ Date: 20 Jul 1995 18:47:57 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: I'm considering purchasing the Wind C++ Gateway between VxWorks and ObjectCenter and was wondering anybody has some experiences, positive or negative, to share. Thanks, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: open files take up memory Date: Thu, 20 Jul 1995 18:04:34 GMT From: doliver@nswc.navy.mil (Donna Oliver) Organization: NAVAL SURFACE WARFARE CENTER DAHLGREN DIVISION Message-ID: <1995Jul20.180434.9403@relay.nswc.navy.mil> Sender: news@relay.nswc.navy.mil There had been a discussion on VxWorks using memory to hold opened files. If someone has that discussion could you please repost it or forward it on to me. Thank you... --------------------------- Newsgroups: comp.os.vxworks Subject: JPL s/w eng. position Date: 20 Jul 1995 21:23:16 GMT From: Rob Steele Organization: JPL Message-ID: <3umhg4$a12@netline-fddi.jpl.nasa.gov> The Rover Technology Research Program at the Jet Propulsion Laboratory, California Instite of Technlogy, is looking for a real-time systems programmer. This is a research program to build prototype Mars rovers and develop technology and capabilities for upcoming missions to Mars. For instance, our current efforts are in building Rocky 7, a 10kg fully autonomous, rough terrain mobile robot. Rocky 7 has a 3U VME chasis with 68040, A/D, DIO, framegrabber, peripheral LM629 motor control, 13 actuators, 6 cameras, etc. Please see http://robotics.jpl.nasa.gov/tasks/rovertech/homepage.html for more. Required experience/knowledge: VxWorks, device driver code development, real-time C language programming Desired experience/knowledge: RTI Control Shell development environment, C++, X Programming, Control Theory, Robotics Please send resumes to: Dr. Homayoun Seraji Mail Stop 198-219 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena CA 91109 seraji@jpl.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting vxworks on Solaris system Date: 21 Jul 1995 00:25:05 GMT From: puppalag Organization: dialogic.com Message-ID: <3ums51$iv7@pravda.dialogic.com> References: <199507191717.KAA10688@inept.stanford.edu> cgonter@inept.stanford.edu wrote: > >Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > >craig >cgonter@inept.stanford.edu > >What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > > then executed > > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > > then check the nfs system by mounting the partition onto another > workstation. > > >The results are: > > target -> > >[VxWorks Boot]: @ > >boot device : egl >processor number : 0 >host name : inept >file name : /opt1/devl/vw/config/star/vxWorks >inet on ethernet (e) : 36.64.0.200:ff000000 >host inet (h) : 36.64.0.195 >user (u) : cgonter >flags (f) : 0x0 >target name (tn) : whoosh > >Attaching network interface egl0... done. >Attaching network interface lo0... done. >Loading... permission denied > >Error loading file: errno = 0x250002. Hi, When VxWorks starts booting depending on the flags It will use either RSH or FTP to get your bootfile and not the NFS mount partition. After you boot it will mount all the NFS exported partitions. If RSH: In the home directory for user cgonter on your system 36.64.0.195 create a .rhosts file which has the following line in it 36.64.0.200 Or if Whoosh is in the /etc/hosts or the DNS you can have the following line in the .rhosts whoosh Try tha above solution. It may solve your problem Goutham Puppala puppalag@dialogic.com --------------------------- Newsgroups: comp.os.vxworks Subject: C/C++ choice Date: 21 Jul 1995 07:45:39 GMT From: vasileia@polhp5.in2p3.fr () Organization: CERN European Laboratory for Particle Physics Message-ID: <3unlv3$n8k@ccpnws.in2p3.fr> Reply-To: vasileia@polhp5.in2p3.fr () Hi, We are a group here just starting/considering VxWorks for our aplications. So we would very much like to get your opinion regarding the programming language you think has the more potential with the VxWorks enviroment. Thats to say do we stick with GNU-C or it would be wiser to try C++. In principle we would like C++ as a choise but I have no idea if the facilities (compilers (gnu-g++ etc) are on a level where we will have no problems from this point of view. Any comments so will be very wellcome thank you - -- *------------------------------------------------------------------------* | George Vasileiadis Email : vacs@hpna49-2.cern.ch | | LPNHE,Ecole Polytechnique vasileia@polhp5.in2p3.fr | | 91128,Palaiseau vassild@cernvm.cern.ch | | France vasileia@slac.stanford.edu | | __________________________ | | Tel: 33-1-6933-3151 (Off.) | | 33-1-43294548 (Hom.) | *------------------------------------------------------------------------* --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Fri Jul 21 04:00:18 1995 From: daemon@csg.lbl.gov Date: Fri Jul 21 04:00:22 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 21 04:00:14 PDT 1995 Subject: ECC DRAM vs Parity DRAM ??? Subject: malloc more than 1 MB, how ? Subject: booting (errno = 0x3d) Subject: PCMCIA for 68xxx ? Subject: Wind C++ Subject: open files take up memory Subject: JPL s/w eng. position Subject: Re: booting vxworks on Solaris system Subject: C/C++ choice ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: ECC DRAM vs Parity DRAM ??? Date: 20 Jul 1995 10:57:26 GMT From: kccheng@hycpsg01 (Kuang-chun Cheng) Organization: FERMILAB, Batavia, IL Message-ID: <3ulcqm$dpd@fnnews.fnal.gov> Hi folks, Let me describle my problem first. We have one RMF3573 board connect one Exabyte 8505 8mm tape device. We use MV167 for our VME master CPU. Couple month ago, we finished the driver of RMF3573 and tested the tape writing rate. The result is about 0.49 MB/sec the same as the Exabyte 8505 User's Manual said. Recently, we got a new MV167-32B board. The first thing to do for us is test the RMF3573 driver and tape writing rate again on the new board. But this time the max. rate we got is 0.38 MB/sec. We spend a lot of time try to figure out what happen, but ... The only possibility we got at this moment is come from DRAM. The MV167 is a 25 MHz with 4MB DRAM (parity) cpu board while MV167-32B is a 33 MHz with 8 MB ECC DRAM cpu board. I guess ECC DRAM do more stuff than parity DRAM so it slow down the tape writing rate (since Exabyte 8505 must access ECC DRAM to get data). But I am not sure ... I did enable the FSTRD bit at MCECC Defaults Register 1 and set MCECC BCLK Frequency register to 33 MHz (as MV167 programmer's ref. guide said, this will push DRAM into full speed operating). But I can't get any improvement. Any suggestion is appreciated. Thanks. ________________ Kuang-chun Cheng kccheng@phys.sinica.edu.tw http://hycpsg01.fnal.gov/kccheng/kccheng.html --------------------------- Newsgroups: comp.os.vxworks Subject: malloc more than 1 MB, how ? Date: 20 Jul 1995 11:09:57 GMT From: kccheng@hycpsg01 (Kuang-chun Cheng) Organization: FERMILAB, Batavia, IL Message-ID: <3uldi5$dpd@fnnews.fnal.gov> Hi, I try to alloc more than 1 MB memory by using malloc() command on MV167 running vxWorks 5.11. But it complains memory size too big. Can I allocate a memory area more than 1 MB ? How to do that ? Thanks a lot. ________________ Kuang-chun Cheng kccheng@phys.sinica.edu.tw --------------------------- Newsgroups: comp.os.vxworks Subject: booting (errno = 0x3d) Date: 20 Jul 1995 16:55:08 GMT From: "Matthew W. K. Brown" Organization: Sandia National Laboratories Message-ID: <3um1pc$o6b@news.sandia.gov> References: <199507191717.KAA10688@inept.stanford.edu> Since the group has been on the topic of booting lately, I thought that I'd ask about this problem that we've been having ... Whenever we boot up 4 or more (sometimes it only takes 3) of our targets at once we get (at various points in the booting process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). I understand that there are limited resources and all of the targets are competing for them; however, I would rather that they wait a bit longer (instead of just dying). Surely, this is a common problem whenever a big system is powering up ... so does anyone know the easy solution (like do I just bump up some timeout value)? Thanks, Matthew Brown mwbrown@sandia.gov --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: PCMCIA for 68xxx ? Date: 20 Jul 1995 16:41:13 GMT From: fitz@aim.cse.tek.com (Bob Fitzsimmons) Organization: Tektronix, Inc., Beaverton, OR. Message-ID: <3um0v9$2h5@tekadm1.cse.tek.com> Reply-To: fitz@aim.cse.tek.com I'm working on a product with an embedded 68xxx system. We're interested in supporting several peripheral devices including: extended memory (ram disks), and serial ports. Because of our small form factor, we're most interested in providing this support via PCMCIA cards. Questions: 1. Which of the available bus adapters is most compatible with 68xxx? 2. Do any vendors provide 68xxx driver support? Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx embedded environment would be greatly appreciated. Thanks in advance, Bob Fitzsimmons Tektonix, Inc. Measurement Business Division (503) 627-5777 --------------------------- Newsgroups: comp.os.vxworks Subject: Wind C++ Date: 20 Jul 1995 18:47:57 GMT From: jkr@cs.cmu.edu (Julio Ken Rosenblatt) Organization: School of Computer Science, CMU,Pittsburgh, PA Message-ID: I'm considering purchasing the Wind C++ Gateway between VxWorks and ObjectCenter and was wondering anybody has some experiences, positive or negative, to share. Thanks, Julio __________________________________________________________________________ Julio K. Rosenblatt Carnegie Mellon University jkr+@cmu.edu Robotics Institute Phone: (412) 268-6880 5000 Forbes Ave. Fax: (412) 268-5571 Pittsburgh, PA 15213-3891 Practice random kindness and senseless acts of beauty. --------------------------- Newsgroups: comp.os.vxworks Subject: open files take up memory Date: Thu, 20 Jul 1995 18:04:34 GMT From: doliver@nswc.navy.mil (Donna Oliver) Organization: NAVAL SURFACE WARFARE CENTER DAHLGREN DIVISION Message-ID: <1995Jul20.180434.9403@relay.nswc.navy.mil> Sender: news@relay.nswc.navy.mil There had been a discussion on VxWorks using memory to hold opened files. If someone has that discussion could you please repost it or forward it on to me. Thank you... --------------------------- Newsgroups: comp.os.vxworks Subject: JPL s/w eng. position Date: 20 Jul 1995 21:23:16 GMT From: Rob Steele Organization: JPL Message-ID: <3umhg4$a12@netline-fddi.jpl.nasa.gov> The Rover Technology Research Program at the Jet Propulsion Laboratory, California Instite of Technlogy, is looking for a real-time systems programmer. This is a research program to build prototype Mars rovers and develop technology and capabilities for upcoming missions to Mars. For instance, our current efforts are in building Rocky 7, a 10kg fully autonomous, rough terrain mobile robot. Rocky 7 has a 3U VME chasis with 68040, A/D, DIO, framegrabber, peripheral LM629 motor control, 13 actuators, 6 cameras, etc. Please see http://robotics.jpl.nasa.gov/tasks/rovertech/homepage.html for more. Required experience/knowledge: VxWorks, device driver code development, real-time C language programming Desired experience/knowledge: RTI Control Shell development environment, C++, X Programming, Control Theory, Robotics Please send resumes to: Dr. Homayoun Seraji Mail Stop 198-219 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena CA 91109 seraji@jpl.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting vxworks on Solaris system Date: 21 Jul 1995 00:25:05 GMT From: puppalag Organization: dialogic.com Message-ID: <3ums51$iv7@pravda.dialogic.com> References: <199507191717.KAA10688@inept.stanford.edu> cgonter@inept.stanford.edu wrote: > >Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > >craig >cgonter@inept.stanford.edu > >What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > > then executed > > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > > then check the nfs system by mounting the partition onto another > workstation. > > >The results are: > > target -> > >[VxWorks Boot]: @ > >boot device : egl >processor number : 0 >host name : inept >file name : /opt1/devl/vw/config/star/vxWorks >inet on ethernet (e) : 36.64.0.200:ff000000 >host inet (h) : 36.64.0.195 >user (u) : cgonter >flags (f) : 0x0 >target name (tn) : whoosh > >Attaching network interface egl0... done. >Attaching network interface lo0... done. >Loading... permission denied > >Error loading file: errno = 0x250002. Hi, When VxWorks starts booting depending on the flags It will use either RSH or FTP to get your bootfile and not the NFS mount partition. After you boot it will mount all the NFS exported partitions. If RSH: In the home directory for user cgonter on your system 36.64.0.195 create a .rhosts file which has the following line in it 36.64.0.200 Or if Whoosh is in the /etc/hosts or the DNS you can have the following line in the .rhosts whoosh Try tha above solution. It may solve your problem Goutham Puppala puppalag@dialogic.com --------------------------- Newsgroups: comp.os.vxworks Subject: C/C++ choice Date: 21 Jul 1995 07:45:39 GMT From: vasileia@polhp5.in2p3.fr () Organization: CERN European Laboratory for Particle Physics Message-ID: <3unlv3$n8k@ccpnws.in2p3.fr> Reply-To: vasileia@polhp5.in2p3.fr () Hi, We are a group here just starting/considering VxWorks for our aplications. So we would very much like to get your opinion regarding the programming language you think has the more potential with the VxWorks enviroment. Thats to say do we stick with GNU-C or it would be wiser to try C++. In principle we would like C++ as a choise but I have no idea if the facilities (compilers (gnu-g++ etc) are on a level where we will have no problems from this point of view. Any comments so will be very wellcome thank you - -- *------------------------------------------------------------------------* | George Vasileiadis Email : vacs@hpna49-2.cern.ch | | LPNHE,Ecole Polytechnique vasileia@polhp5.in2p3.fr | | 91128,Palaiseau vassild@cernvm.cern.ch | | France vasileia@slac.stanford.edu | | __________________________ | | Tel: 33-1-6933-3151 (Off.) | | 33-1-43294548 (Hom.) | *------------------------------------------------------------------------* --------------------------- End of New-News digest ********************** Subject: Max File descriptors Submitted-by leonid@rst.co.il Fri Jul 21 05:46:29 1995 Submitted-by: leonid@rst.co.il (Leonid Rosenboim) Dear VxWorrkers, have you ever reached the VxWorks limit on file descriptors / sockets ? Have you ever tried to define NUM_FILES in configAll.h to a higher value than the OS could support ? If so, what was the upper absolute limit ? Thanks, ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il From leonid@rst.co.il Fri Jul 21 05:46:29 1995 From: leonid@rst.co.il (Leonid Rosenboim) Date: Fri Jul 21 05:46:32 PDT 1995 Subject: Max File descriptors Dear VxWorrkers, have you ever reached the VxWorks limit on file descriptors / sockets ? Have you ever tried to define NUM_FILES in configAll.h to a higher value than the OS could support ? If so, what was the upper absolute limit ? Thanks, ----------------------------------------------------------------------- Leonid Rosenboim Phone: +972-3-559-8144 R S T Software Industries Ltd. Mobile: +972-50-307-142 3 Hashikma st. Industrial Zone, Fax: +972-3-559-8244 P.O.Box 11502, AZUR 58017, Israel E-Mail: leonid@rst.co.il Subject: Re: boot error 0x1a9 & UPDATE info:booting vxworks on Solaris system Submitted-by fbh@vlsi9.gsfc.nasa.gov Fri Jul 21 07:56:03 1995 Submitted-by: fbh@vlsi9.gsfc.nasa.gov (Francis B. Hallahan) When ".rhosts" is good, it is very good, but when ".rhosts" is bad, it is horrid. We switched from ftp to .rhosts just a few weeks ago because of "boot error 0x1a9". We use a .rhosts file in a special vxworks account since we want to limit access to our Workstations from VxWorks. The "secret" for .rhosts is hidden at the bottom of the man page for "rhosts" -- only OFFICIAL hostnames may be used; No aliases, nicknames, or IP addresses. First you must determine the official name for your vxWorks system. For your system, this may be "whoosh". If you have NIS, it is more likely that your system name is "whoosh.stanford.edu" or something similar. This would make your .rhosts look like: #. inept:cgonter->cgonter:more .rhosts emprise cgonter dyslexic + whoosh.stanford.edu + - allow all users from whoosh better still whoosh.stanford.edu cgonter - allow only cgonter best whoosh.stanford.edu special_vx_account - allow only from limited user We created, although we did this for FTP, a special vxWorks account that was on in group "nobody" for security reasons. This way anyone using the account only has access to world readable files! When using RSH to there are a few other things to watch out for. The most important is that your ".cshrc" or ".profile" files MUST produce NO (and I do mean zero) output to an rsh command. The vxWorks Programmers Guide has more in section 2, especially the Troubleshooting section. When all is setup correctly Attaching network interface egl0... done. Attaching network interface lo0... done. Loading... 371428 + 22308 + 22584 - your numbers will vary Starting at 0x1000... The 3 numbers after "Loading" show that the transfer is working. The "Starting at" shows the handoff from ROM code to the vxWorks image. #. From cgonter@inept.stanford.edu Thu Jul 20 10:29:25 1995 #. #. IN THE PREVIOUS POSTING: #. > . . . #. > then check the nfs system by mounting the partition onto another #. > workstation. #. > #. > As others have already said, NFS is not used until after your vxWorks image is loaded and started #. > . . . #. #. #. New info .... #. #. The .rhosts file has been in the hosts users area(cgonter) with the following #. permissions: #. #. -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* #. #. .rhosts contains --> #. #. #. inept:cgonter->cgonter:more .rhosts #. emprise cgonter #. dyslexic + #. whoosh + #. #. #. #. Error return is the same as above: #. #. > Attaching network interface egl0... done. #. > Attaching network interface lo0... done. #. > Loading... permission denied #. > #. > Error loading file: errno = 0x250002. #. > #. This is a host configuration problem; most likely "whoosh" is not your official name. #. ADDED /etc/hosts.equiv and the results are: #. #. #. #. The target system whoosh hangs at this point. #. #. Don't know why your system hangs, but since no numbers showed up, the file transfer didn't work. This might be because of something in your .cshrc file. #. #. #. .... thanks for all the previous responses #. craig #. cgonter@inept.stanford.edu #. #. Hope this helps, Frank ----------------------------------------------------------------------- | Frank Hallahan | Hallahan@gsfc.nasa.gov | | Computer Sciences Corporation | fbh@vlsi.gsfc.nasa.gov | | at Nasa Goddard Space Flight Center | Phone (301) 286 7064 | ----------------------------------------------------------------------- From fbh@vlsi9.gsfc.nasa.gov Fri Jul 21 07:56:03 1995 From: fbh@vlsi9.gsfc.nasa.gov (Francis B. Hallahan) Date: Fri Jul 21 07:56:06 PDT 1995 Subject: Re: boot error 0x1a9 & UPDATE info:booting vxworks on Solaris system When ".rhosts" is good, it is very good, but when ".rhosts" is bad, it is horrid. We switched from ftp to .rhosts just a few weeks ago because of "boot error 0x1a9". We use a .rhosts file in a special vxworks account since we want to limit access to our Workstations from VxWorks. The "secret" for .rhosts is hidden at the bottom of the man page for "rhosts" -- only OFFICIAL hostnames may be used; No aliases, nicknames, or IP addresses. First you must determine the official name for your vxWorks system. For your system, this may be "whoosh". If you have NIS, it is more likely that your system name is "whoosh.stanford.edu" or something similar. This would make your .rhosts look like: #. inept:cgonter->cgonter:more .rhosts emprise cgonter dyslexic + whoosh.stanford.edu + - allow all users from whoosh better still whoosh.stanford.edu cgonter - allow only cgonter best whoosh.stanford.edu special_vx_account - allow only from limited user We created, although we did this for FTP, a special vxWorks account that was on in group "nobody" for security reasons. This way anyone using the account only has access to world readable files! When using RSH to there are a few other things to watch out for. The most important is that your ".cshrc" or ".profile" files MUST produce NO (and I do mean zero) output to an rsh command. The vxWorks Programmers Guide has more in section 2, especially the Troubleshooting section. When all is setup correctly Attaching network interface egl0... done. Attaching network interface lo0... done. Loading... 371428 + 22308 + 22584 - your numbers will vary Starting at 0x1000... The 3 numbers after "Loading" show that the transfer is working. The "Starting at" shows the handoff from ROM code to the vxWorks image. #. From cgonter@inept.stanford.edu Thu Jul 20 10:29:25 1995 #. #. IN THE PREVIOUS POSTING: #. > . . . #. > then check the nfs system by mounting the partition onto another #. > workstation. #. > #. > As others have already said, NFS is not used until after your vxWorks image is loaded and started #. > . . . #. #. #. New info .... #. #. The .rhosts file has been in the hosts users area(cgonter) with the following #. permissions: #. #. -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* #. #. .rhosts contains --> #. #. #. inept:cgonter->cgonter:more .rhosts #. emprise cgonter #. dyslexic + #. whoosh + #. #. #. #. Error return is the same as above: #. #. > Attaching network interface egl0... done. #. > Attaching network interface lo0... done. #. > Loading... permission denied #. > #. > Error loading file: errno = 0x250002. #. > #. This is a host configuration problem; most likely "whoosh" is not your official name. #. ADDED /etc/hosts.equiv and the results are: #. #. #. #. The target system whoosh hangs at this point. #. #. Don't know why your system hangs, but since no numbers showed up, the file transfer didn't work. This might be because of something in your .cshrc file. #. #. #. .... thanks for all the previous responses #. craig #. cgonter@inept.stanford.edu #. #. Hope this helps, Frank ----------------------------------------------------------------------- | Frank Hallahan | Hallahan@gsfc.nasa.gov | | Computer Sciences Corporation | fbh@vlsi.gsfc.nasa.gov | | at Nasa Goddard Space Flight Center | Phone (301) 286 7064 | ----------------------------------------------------------------------- Subject: Re: Wind C++ Submitted-by key@diasonx.com Fri Jul 21 09:21:39 1995 Submitted-by: key@diasonx.com (Charles Key) > I'm considering purchasing the Wind C++ Gateway between VxWorks > and ObjectCenter and was wondering anybody has some experiences, > positive or negative, to share. We purchased Gateway/ObjectCenter a couple of years ago. ObjectCenter is fairly pricey, and really only serves as a frontend to vxgdb. It's somewhat easier to use than vxgdb, IMO, including a nice class/data browser. You are, however, using a minimal subset of ObjectCenter's capabilities and paying the full price. If you do some development and testing on the workstation, it might be cost effective to have the ObjectCenter development environment. Our code is so hardware dependent that we do no workstation testing, and ObjectCenter has not been cost effective. Also realize that Gateway is cfront-based, and there are tradeoffs there, as well (e.g. cfront doesn't recognize the volatile keyword, which tends to be kind of important for embedded systems). Others use gcc, which now supports vxworks out of the box. ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com From key@diasonx.com Fri Jul 21 09:21:39 1995 From: key@diasonx.com (Charles Key) Date: Fri Jul 21 09:21:41 PDT 1995 Subject: Re: Wind C++ > I'm considering purchasing the Wind C++ Gateway between VxWorks > and ObjectCenter and was wondering anybody has some experiences, > positive or negative, to share. We purchased Gateway/ObjectCenter a couple of years ago. ObjectCenter is fairly pricey, and really only serves as a frontend to vxgdb. It's somewhat easier to use than vxgdb, IMO, including a nice class/data browser. You are, however, using a minimal subset of ObjectCenter's capabilities and paying the full price. If you do some development and testing on the workstation, it might be cost effective to have the ObjectCenter development environment. Our code is so hardware dependent that we do no workstation testing, and ObjectCenter has not been cost effective. Also realize that Gateway is cfront-based, and there are tradeoffs there, as well (e.g. cfront doesn't recognize the volatile keyword, which tends to be kind of important for embedded systems). Others use gcc, which now supports vxworks out of the box. ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com Subject: Re: booting (errno = 0x3d) Submitted-by lharad@atl.com Fri Jul 21 10:05:38 1995 Submitted-by: lharad@atl.com (Les Harada) > > Subject: booting (errno = 0x3d) > Date: 20 Jul 1995 16:55:08 GMT > >From: "Matthew W. K. Brown" > Organization: Sandia National Laboratories > > Since the group has been on the topic of booting lately, I > thought that I'd ask about this problem that we've been having ... > > Whenever we boot up 4 or more (sometimes it only takes 3) of our > targets at once we get (at various points in the booting > process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). > I understand that there are limited resources and all of the > targets are competing for them; however, I would rather that > they wait a bit longer (instead of just dying). Surely, this > is a common problem whenever a big system is powering up ... so > does anyone know the easy solution (like do I just bump up some > timeout value)? > > Thanks, > > Matthew Brown > mwbrown@sandia.gov > > At a previous job, we had a similar problem, but I don't remember the error message. Inetd on the boot host had a feature that it would only allow 40 socket connection in a minute. Since each target uses several socket connection to boot, it didn't take us long to reach 40. I don't think we had problems with only 4 targets, but your S_errno_ECONNREFUSED seems to indicate that the limited resource is on the host. I think this feature was implimented to protect the host from runaway processes on the client. We had to get a patched inetd from our vendor, Harris Computers, to allow us to change the limit. BTW, the targets also saturated the subnet, so we put a delay in their boot scripts based on cpu number (they were in the same chassis) to spread out some of their network traffic. Les Harada lharad@atl.com From lharad@atl.com Fri Jul 21 10:05:38 1995 From: lharad@atl.com (Les Harada) Date: Fri Jul 21 10:05:40 PDT 1995 Subject: Re: booting (errno = 0x3d) > > Subject: booting (errno = 0x3d) > Date: 20 Jul 1995 16:55:08 GMT > >From: "Matthew W. K. Brown" > Organization: Sandia National Laboratories > > Since the group has been on the topic of booting lately, I > thought that I'd ask about this problem that we've been having ... > > Whenever we boot up 4 or more (sometimes it only takes 3) of our > targets at once we get (at various points in the booting > process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). > I understand that there are limited resources and all of the > targets are competing for them; however, I would rather that > they wait a bit longer (instead of just dying). Surely, this > is a common problem whenever a big system is powering up ... so > does anyone know the easy solution (like do I just bump up some > timeout value)? > > Thanks, > > Matthew Brown > mwbrown@sandia.gov > > At a previous job, we had a similar problem, but I don't remember the error message. Inetd on the boot host had a feature that it would only allow 40 socket connection in a minute. Since each target uses several socket connection to boot, it didn't take us long to reach 40. I don't think we had problems with only 4 targets, but your S_errno_ECONNREFUSED seems to indicate that the limited resource is on the host. I think this feature was implimented to protect the host from runaway processes on the client. We had to get a patched inetd from our vendor, Harris Computers, to allow us to change the limit. BTW, the targets also saturated the subnet, so we put a delay in their boot scripts based on cpu number (they were in the same chassis) to spread out some of their network traffic. Les Harada lharad@atl.com Subject: Re: UPDATE info: booting vxworks on Solaris system Submitted-by punamia@newkla.kla.com Fri Jul 21 11:15:26 1995 Submitted-by: punamia@newkla.kla.com (Manoj Punamia) > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... The target system whoosh hangs at this point. I had the same problem when tryiing to boot my target which is Force-68030 board with VxWorks-5.0.2b from SunOS. I Changed the VME Chassis and problem went away. From punamia@newkla.kla.com Fri Jul 21 11:15:26 1995 From: punamia@newkla.kla.com (Manoj Punamia) Date: Fri Jul 21 11:15:29 PDT 1995 Subject: Re: UPDATE info: booting vxworks on Solaris system > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... The target system whoosh hangs at this point. I had the same problem when tryiing to boot my target which is Force-68030 board with VxWorks-5.0.2b from SunOS. I Changed the VME Chassis and problem went away. Subject: Re: booting (errno = 0x3d) Submitted-by stan@rti.com Fri Jul 21 11:30:48 1995 Submitted-by: Stan Schneider >> Whenever we boot up 4 or more (sometimes it only takes 3) of our >> targets at once we get (at various points in the booting >> process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). This is usually symptomatic of bus-grant starvation, especially if the targets that don't boot are in high-numbered slots of a single VME cage. Are all your CPUs on different Bus Grant lines? Is round-robin arbitration enabled? -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Fri Jul 21 11:30:48 1995 From: Stan Schneider Date: Fri Jul 21 11:30:50 PDT 1995 Subject: Re: booting (errno = 0x3d) >> Whenever we boot up 4 or more (sometimes it only takes 3) of our >> targets at once we get (at various points in the booting >> process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). This is usually symptomatic of bus-grant starvation, especially if the targets that don't boot are in high-numbered slots of a single VME cage. Are all your CPUs on different Bus Grant lines? Is round-robin arbitration enabled? -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: Multi-processing on PCI bus Submitted-by stan@rti.com Fri Jul 21 12:40:42 1995 Submitted-by: Stan Schneider Hi all, Does anyone have any experience with fast multi-processing on a PCI bus? TIA, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Fri Jul 21 12:40:42 1995 From: Stan Schneider Date: Fri Jul 21 12:41:44 PDT 1995 Subject: Multi-processing on PCI bus Hi all, Does anyone have any experience with fast multi-processing on a PCI bus? TIA, -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sat Jul 22 04:00:25 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 22 04:00:15 PDT 1995 Subject: Does anyone have a synchronous SDLC/HDLC driver for the 85x30 SCC? Subject: Re: booting (errno = 0x3d) Subject: Re: booting vxworks on Solaris system Subject: Smallest VxWorks Image ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Does anyone have a synchronous SDLC/HDLC driver for the 85x30 SCC? Date: Fri, 21 Jul 95 08:52:26 MDT From: jeffh@lfs.loral.com Message-ID: <9507211456.AA38958@inetgw.fsc.ibm.com> We have an asynchronous driver for the SCC, but not the SDLC driver. Does anyone out there happen to have one? Thanks, /\ /\ ____o_o o Jeff Hartt /\ / *\/\ / *\ /\ /_*_||*/\ Loral Federal Systems /* \ * * \/ * \/ *\| __ | | jeffh@lfs.loral.com / \_______*__*_\___|__||__|__| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting (errno = 0x3d) Date: 21 Jul 1995 16:11:01 GMT From: hartl@Informatik.TU-Muenchen.DE (Anton Hartl) Organization: Technische Universitaet Muenchen, Germany Message-ID: <3uojil$lph@sunsystem5.informatik.tu-muenchen.de> References: <199507191717.KAA10688@inept.stanford.edu> <3um1pc$o6b@news.sandia.gov> "Matthew W. K. Brown" writes: >Whenever we boot up 4 or more (sometimes it only takes 3) of our >targets at once we get (at various points in the booting >process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). A common source for these kind of problems is a safe-guard in the inetd (we've had this on SUN OS 4.1.x) that tries to handle the case that a number of daemons that it forks terminate in a very short time interval. Inetd then believes the daemon is broken somehow and with the goal to protect the system as a whole from cray daemons, it disables the service for some minutes. >I understand that there are limited resources and all of the >targets are competing for them; however, I would rather that >they wait a bit longer (instead of just dying). Surely, this >is a common problem whenever a big system is powering up ... so >does anyone know the easy solution (like do I just bump up some >timeout value)? Given that you cannot arrange the systems that are booting not to snarf everything at the same time, the are two ways to handle this: (1) make a wrapper around ftpd that just forks the real ftpd, waits for it to exit and then adds a sleep of some seconds; (2) tell inetd to wait for the ftpd to terminate and not to fork others in the meanwhile, that's done by changed the column for ftp in inetd.conf that says "nowait" to "wait"; but you have to be aware that you can have only ONE ftp connection at a time on that machine, which is probably something you cannot do for other reasons. In addition I'm not sure wether inetds security mechanism isn't still active then and thus might result in the very same behaviour. The other thing is that you could switch to NFS as a method to fetch files for your tasks (but that's not an option for the actual boot process, you have only ftp and rcp here). Hope that helps, -Toni - -- ________________________________________________________________________________ Anton Hartl | WWW means What a Waste of WWW resources. Anton.Hartl@informatik.tu-muenchen.de | It is a way to use more to achieve less. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting vxworks on Solaris system Date: 21 Jul 1995 20:15:28 GMT From: witts@PROBLEM_WITH_INEWS_GATEWAY_FILE (Steve Witt) Organization: TRW Data Technologies Division, Carson CA Message-ID: <3up1t1$bo5@venice.sedd.trw.com> References: <199507191717.KAA10688@inept.stanford.edu> cgonter@inept.stanford.edu wrote: > Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > craig > cgonter@inept.stanford.edu > What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > then executed > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > then check the nfs system by mounting the partition onto another > workstation. > The results are: > target -> > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. Just had the same problem when we upgraded from SunOS 4.1.3 to Solaris 2.4. I have solved the problem in a way that isn't too great. The problem is your hosts.equiv or .rhosts file probably doesn't have an entry for your target in it. At least that is how it should work. I was only able to get our boards to load off our development systems by putting a .rhosts file in each developer's account. This file has a single + in it. Read the man page on hosts.equiv. This allows ANY other machine to execute rlogin, rsh, rcp, and rcmd in 'trusted' mode, i.e. trusted users don't need to supply a password when accessing this machine. This is a big security hole for your system. What should work is to put a + [+whoosh (in your case)] line in the .rhosts file to allow trusted access only to specific hosts. This did not work for me and I'm not sure why at this point. When vxWorks boots the bootrom code tries to execute 'rsh' commands on the host to load the vxWorks image from where ever you tell it load from. The 'permission denied' that you saw is returned by the rsh daemon because it doesn't think your target is authorized. (at least that's what I think is going on) I'm not sure why you're 'sharing' a portion of a filesystem for an NFS mount. Perhaps you're mounting this filesystem for some other reason, but I don't think this is involved with loading the vxWorks image onto your target so that it can boot. If there is anyone else out there who has solved this better than I, please let us know. I'm not pleased with the universal access to our machines, but it does work and I've got work to do. I'll try to trouble shoot it when I can. - --------------------------------------------------------------------- Steve Witt Hughes Aircraft Company sawitt@hac2arpa.hac.com --------------------------- Newsgroups: comp.os.vxworks Subject: Smallest VxWorks Image Date: 21 Jul 1995 21:30:27 GMT From: t.wrather@genie.geis.com (Ted Wrather) Organization: Bell-Northern Research, RTP, NC Message-ID: <3up69j$ai0@brtph500.bnr.ca> What is the size of the smallest usable VxWorks image for the MC68302? I seem to remember 24Kbytes, but I can't find that number written down in any of the literature. Thanks, Ted --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sat Jul 22 04:00:25 1995 From: daemon@csg.lbl.gov Date: Sat Jul 22 04:00:29 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 22 04:00:15 PDT 1995 Subject: Does anyone have a synchronous SDLC/HDLC driver for the 85x30 SCC? Subject: Re: booting (errno = 0x3d) Subject: Re: booting vxworks on Solaris system Subject: Smallest VxWorks Image ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Does anyone have a synchronous SDLC/HDLC driver for the 85x30 SCC? Date: Fri, 21 Jul 95 08:52:26 MDT From: jeffh@lfs.loral.com Message-ID: <9507211456.AA38958@inetgw.fsc.ibm.com> We have an asynchronous driver for the SCC, but not the SDLC driver. Does anyone out there happen to have one? Thanks, /\ /\ ____o_o o Jeff Hartt /\ / *\/\ / *\ /\ /_*_||*/\ Loral Federal Systems /* \ * * \/ * \/ *\| __ | | jeffh@lfs.loral.com / \_______*__*_\___|__||__|__| --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting (errno = 0x3d) Date: 21 Jul 1995 16:11:01 GMT From: hartl@Informatik.TU-Muenchen.DE (Anton Hartl) Organization: Technische Universitaet Muenchen, Germany Message-ID: <3uojil$lph@sunsystem5.informatik.tu-muenchen.de> References: <199507191717.KAA10688@inept.stanford.edu> <3um1pc$o6b@news.sandia.gov> "Matthew W. K. Brown" writes: >Whenever we boot up 4 or more (sometimes it only takes 3) of our >targets at once we get (at various points in the booting >process) an ethernet error--errno = 0x3d (S_errno_ECONNREFUSED). A common source for these kind of problems is a safe-guard in the inetd (we've had this on SUN OS 4.1.x) that tries to handle the case that a number of daemons that it forks terminate in a very short time interval. Inetd then believes the daemon is broken somehow and with the goal to protect the system as a whole from cray daemons, it disables the service for some minutes. >I understand that there are limited resources and all of the >targets are competing for them; however, I would rather that >they wait a bit longer (instead of just dying). Surely, this >is a common problem whenever a big system is powering up ... so >does anyone know the easy solution (like do I just bump up some >timeout value)? Given that you cannot arrange the systems that are booting not to snarf everything at the same time, the are two ways to handle this: (1) make a wrapper around ftpd that just forks the real ftpd, waits for it to exit and then adds a sleep of some seconds; (2) tell inetd to wait for the ftpd to terminate and not to fork others in the meanwhile, that's done by changed the column for ftp in inetd.conf that says "nowait" to "wait"; but you have to be aware that you can have only ONE ftp connection at a time on that machine, which is probably something you cannot do for other reasons. In addition I'm not sure wether inetds security mechanism isn't still active then and thus might result in the very same behaviour. The other thing is that you could switch to NFS as a method to fetch files for your tasks (but that's not an option for the actual boot process, you have only ftp and rcp here). Hope that helps, -Toni - -- ________________________________________________________________________________ Anton Hartl | WWW means What a Waste of WWW resources. Anton.Hartl@informatik.tu-muenchen.de | It is a way to use more to achieve less. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: booting vxworks on Solaris system Date: 21 Jul 1995 20:15:28 GMT From: witts@PROBLEM_WITH_INEWS_GATEWAY_FILE (Steve Witt) Organization: TRW Data Technologies Division, Carson CA Message-ID: <3up1t1$bo5@venice.sedd.trw.com> References: <199507191717.KAA10688@inept.stanford.edu> cgonter@inept.stanford.edu wrote: > Can anyone clue me in getting vxworks to boot on a Solaris 2.4 > craig > cgonter@inept.stanford.edu > What has been done: > host -> modified the /etc/dfs/dfstab with > > share -F nfs -o rw -d "vxworks stuff" /opt1 > then executed > shareall > /usr/lib/nfs/nfsd -a 16 > /usr/lib/nfs/mountd > then check the nfs system by mounting the partition onto another > workstation. > The results are: > target -> > [VxWorks Boot]: @ > > boot device : egl > processor number : 0 > host name : inept > file name : /opt1/devl/vw/config/star/vxWorks > inet on ethernet (e) : 36.64.0.200:ff000000 > host inet (h) : 36.64.0.195 > user (u) : cgonter > flags (f) : 0x0 > target name (tn) : whoosh > > Attaching network interface egl0... done. > Attaching network interface lo0... done. > Loading... permission denied > > Error loading file: errno = 0x250002. Just had the same problem when we upgraded from SunOS 4.1.3 to Solaris 2.4. I have solved the problem in a way that isn't too great. The problem is your hosts.equiv or .rhosts file probably doesn't have an entry for your target in it. At least that is how it should work. I was only able to get our boards to load off our development systems by putting a .rhosts file in each developer's account. This file has a single + in it. Read the man page on hosts.equiv. This allows ANY other machine to execute rlogin, rsh, rcp, and rcmd in 'trusted' mode, i.e. trusted users don't need to supply a password when accessing this machine. This is a big security hole for your system. What should work is to put a + [+whoosh (in your case)] line in the .rhosts file to allow trusted access only to specific hosts. This did not work for me and I'm not sure why at this point. When vxWorks boots the bootrom code tries to execute 'rsh' commands on the host to load the vxWorks image from where ever you tell it load from. The 'permission denied' that you saw is returned by the rsh daemon because it doesn't think your target is authorized. (at least that's what I think is going on) I'm not sure why you're 'sharing' a portion of a filesystem for an NFS mount. Perhaps you're mounting this filesystem for some other reason, but I don't think this is involved with loading the vxWorks image onto your target so that it can boot. If there is anyone else out there who has solved this better than I, please let us know. I'm not pleased with the universal access to our machines, but it does work and I've got work to do. I'll try to trouble shoot it when I can. - --------------------------------------------------------------------- Steve Witt Hughes Aircraft Company sawitt@hac2arpa.hac.com --------------------------- Newsgroups: comp.os.vxworks Subject: Smallest VxWorks Image Date: 21 Jul 1995 21:30:27 GMT From: t.wrather@genie.geis.com (Ted Wrather) Organization: Bell-Northern Research, RTP, NC Message-ID: <3up69j$ai0@brtph500.bnr.ca> What is the size of the smallest usable VxWorks image for the MC68302? I seem to remember 24Kbytes, but I can't find that number written down in any of the literature. Thanks, Ted --------------------------- End of New-News digest ********************** Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sun Jul 23 04:00:19 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sun Jul 23 04:00:16 PDT 1995 Subject: Free Real-time O/S magazine subscription Subject: survey says... ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Free Real-time O/S magazine subscription Date: 22 Jul 1995 12:18:59 GMT From: Wayne Kristoff Organization: Publications Office Message-ID: <3uqqbj$cvd@nnrp3.primenet.com> Real-time Engineering is published quarterly for software and system engineers involved in the development of real-time software. Our web page provides subscriber services to current and potential readers (http://www.primenet.com/~magpub). Subscriptions can also be obtained by requesting a subscription form from magpub@aol.com. Provide us a fax number or address and we will send you the form. --------------------------- Newsgroups: comp.os.vxworks Subject: survey says... Date: Sun, 23 Jul 1995 00:27:13 -0700 From: squeeby Organization: Hooked Online Services Message-ID: Here are results of my recent survey on VxWorks usage. VxWorks Usage Survey - ---------------------- 1. Realtime data acquisition, analysis and storage 2. Aircraft tracking system 3. Hand-held MTDR, OTDR testers 4. Testbeds for various hardware 5. Time synchronization systems 6. Distributed Health monitoring for flight/vehicle systems 7. Automated ground checkout 8. Fault tolerant networks 9. On-board distributed flight control systems 10. Autonomous and Semi-autonomous ground vehicles 11. Check-out system for unusual hardware: Russian rocket engines 12. I/O engine for simulation of the other systems 13. Autoglider - Image processing 14. Spectrograph control 15. Adaptive optics 16. Robotic manipulators 17. Spacecraft to Mars (Pathfinder) 18. X-terminals (i960-based) 19. Remotely operated vehicle at Monterey Bay Aquarium Research 20. Television station automation 21. Telecommunications 22. Electric Power generation monitoring 23. Navy communications system 24. Machine vision 25. Radar Image generatory, symbology image generator 26. Low resolution imaging spectrograph for Keck 10-meter telescope in Hawaii 27. Guide camera for Palomar Observatory 5-meter telescope 28. Telescope control system for Palomar 29. Telerobots for extreme environments (subsea oil platform, nuclear equipment operation) 30. Northern Telecom PBX call processing 31. SUN Network Coprocessor: NFS accelerator 32. Silicon wafer prober 33. Interactive TV 34. Automotive control systems 35. NYC traffic control http://www.hooked.net/users/squeeby/ --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sun Jul 23 04:00:19 1995 From: daemon@csg.lbl.gov Date: Sun Jul 23 04:00:22 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 23 04:00:16 PDT 1995 Subject: Free Real-time O/S magazine subscription Subject: survey says... ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Free Real-time O/S magazine subscription Date: 22 Jul 1995 12:18:59 GMT From: Wayne Kristoff Organization: Publications Office Message-ID: <3uqqbj$cvd@nnrp3.primenet.com> Real-time Engineering is published quarterly for software and system engineers involved in the development of real-time software. Our web page provides subscriber services to current and potential readers (http://www.primenet.com/~magpub). Subscriptions can also be obtained by requesting a subscription form from magpub@aol.com. Provide us a fax number or address and we will send you the form. --------------------------- Newsgroups: comp.os.vxworks Subject: survey says... Date: Sun, 23 Jul 1995 00:27:13 -0700 From: squeeby Organization: Hooked Online Services Message-ID: Here are results of my recent survey on VxWorks usage. VxWorks Usage Survey - ---------------------- 1. Realtime data acquisition, analysis and storage 2. Aircraft tracking system 3. Hand-held MTDR, OTDR testers 4. Testbeds for various hardware 5. Time synchronization systems 6. Distributed Health monitoring for flight/vehicle systems 7. Automated ground checkout 8. Fault tolerant networks 9. On-board distributed flight control systems 10. Autonomous and Semi-autonomous ground vehicles 11. Check-out system for unusual hardware: Russian rocket engines 12. I/O engine for simulation of the other systems 13. Autoglider - Image processing 14. Spectrograph control 15. Adaptive optics 16. Robotic manipulators 17. Spacecraft to Mars (Pathfinder) 18. X-terminals (i960-based) 19. Remotely operated vehicle at Monterey Bay Aquarium Research 20. Television station automation 21. Telecommunications 22. Electric Power generation monitoring 23. Navy communications system 24. Machine vision 25. Radar Image generatory, symbology image generator 26. Low resolution imaging spectrograph for Keck 10-meter telescope in Hawaii 27. Guide camera for Palomar Observatory 5-meter telescope 28. Telescope control system for Palomar 29. Telerobots for extreme environments (subsea oil platform, nuclear equipment operation) 30. Northern Telecom PBX call processing 31. SUN Network Coprocessor: NFS accelerator 32. Silicon wafer prober 33. Interactive TV 34. Automotive control systems 35. NYC traffic control http://www.hooked.net/users/squeeby/ --------------------------- End of New-News digest ********************** Subject: Re: boot error 0x1a9 & UPDATE info:booting vxworks on Solaris system Submitted-by froeber@bbn.com Sun Jul 23 18:41:00 1995 Submitted-by: Fred Roeber Frank Hallahan gave a pretty good rundown on rsh issues when booting from VxWorks. One thing he didn't mention is that as Craig reported below, his .rhosts file had permission 755. At least on some systems, a .rhosts file has to have no group or world access to be considered valid. This is a security measure that some OS versions seem to enforce. Thus, you should use mode 600 on .rhosts files (i.e. -rw-------). >#. New info .... >#. >#. The .rhosts file has been in the hosts users area(cgonter) with the following >#. permissions: >#. >#. -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* >#. It would be interesting to hear back from Craig in the end what ended up fixing his problem. Fred | Fred J Roeber, Bolt Beranek and Newman Inc | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | From froeber@bbn.com Sun Jul 23 18:41:00 1995 From: Fred Roeber Date: Sun Jul 23 18:41:02 PDT 1995 Subject: Re: boot error 0x1a9 & UPDATE info:booting vxworks on Solaris system Frank Hallahan gave a pretty good rundown on rsh issues when booting from VxWorks. One thing he didn't mention is that as Craig reported below, his .rhosts file had permission 755. At least on some systems, a .rhosts file has to have no group or world access to be considered valid. This is a security measure that some OS versions seem to enforce. Thus, you should use mode 600 on .rhosts files (i.e. -rw-------). >#. New info .... >#. >#. The .rhosts file has been in the hosts users area(cgonter) with the following >#. permissions: >#. >#. -rwxr-xr-x 1 cgonter staff 56 Jul 07 02:57 .rhosts* >#. It would be interesting to hear back from Craig in the end what ended up fixing his problem. Fred | Fred J Roeber, Bolt Beranek and Newman Inc | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Mon Jul 24 04:00:16 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Mon Jul 24 04:00:12 PDT 1995 Subject: Translator: Intel ASM -> GNU ASM ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Translator: Intel ASM -> GNU ASM ? Date: Sun, 23 Jul 1995 20:23:06 GMT From: compulit@actcom.co.il (Amit Resh) Organization: ACTCOM - Internet Services in Israel Message-ID: Sender: news@wang.com Does anyone out there have, or can point me to, a translator from Intel 80x86 Assembler syntax to GNU Assembler (AT&T) syntax ? - ------------------------------------------------------------------- Amit Resh Tel. (972)-3-5401268 V.P. R&D, COMPULITE R&D Inc. FAX. (972)-3-5401276 3 Haroshet St. E-Mail: compulit@actcom.co.il Ramat Hasharon, Israel --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Mon Jul 24 04:00:16 1995 From: daemon@csg.lbl.gov Date: Mon Jul 24 04:00:19 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Jul 24 04:00:12 PDT 1995 Subject: Translator: Intel ASM -> GNU ASM ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Translator: Intel ASM -> GNU ASM ? Date: Sun, 23 Jul 1995 20:23:06 GMT From: compulit@actcom.co.il (Amit Resh) Organization: ACTCOM - Internet Services in Israel Message-ID: Sender: news@wang.com Does anyone out there have, or can point me to, a translator from Intel 80x86 Assembler syntax to GNU Assembler (AT&T) syntax ? - ------------------------------------------------------------------- Amit Resh Tel. (972)-3-5401268 V.P. R&D, COMPULITE R&D Inc. FAX. (972)-3-5401276 3 Haroshet St. E-Mail: compulit@actcom.co.il Ramat Hasharon, Israel --------------------------- End of New-News digest ********************** Subject: Re: booting (errno = 0x3d) Submitted-by mwbrown@somnet.sandia.gov Mon Jul 24 08:16:14 1995 Submitted-by: "Matthew W. Brown" > ... that don't boot are in high-numbered slots of a single VME cage. Actually, we only have one target/cage. It's just that we have several cages booting at one time. A few people have suggested that I change my bootroms so that they boot up at different times; however, not only do I want all of my bootroms to be the same, but I am not convinced that would solve all of my problems as sometimes a target will fail after it has nearly finished running my (somewhat lengthy) startup script. Les Harada (lharad@atl.com) has (probably correctly) suggested that it might actually be a function of my host--which is only allowing so many socket connections at a time. I intend to explore that a bit. He also suggested that I could taskDelay() a bit in my startup scripts with longer delays for certain targets--I like that ideal too. Anyway, thanks for the response. Matthew Brown mwbrown@sandia.gov From mwbrown@somnet.sandia.gov Mon Jul 24 08:16:14 1995 From: "Matthew W. Brown" Date: Mon Jul 24 08:16:16 PDT 1995 Subject: Re: booting (errno = 0x3d) > ... that don't boot are in high-numbered slots of a single VME cage. Actually, we only have one target/cage. It's just that we have several cages booting at one time. A few people have suggested that I change my bootroms so that they boot up at different times; however, not only do I want all of my bootroms to be the same, but I am not convinced that would solve all of my problems as sometimes a target will fail after it has nearly finished running my (somewhat lengthy) startup script. Les Harada (lharad@atl.com) has (probably correctly) suggested that it might actually be a function of my host--which is only allowing so many socket connections at a time. I intend to explore that a bit. He also suggested that I could taskDelay() a bit in my startup scripts with longer delays for certain targets--I like that ideal too. Anyway, thanks for the response. Matthew Brown mwbrown@sandia.gov Subject: Re: booting (errno = 0x3d) Submitted-by mwbrown@somnet.sandia.gov Mon Jul 24 08:20:39 1995 Submitted-by: "Matthew W. Brown" Whoops, I sent that last response to the whole group (I had meant to send it to Stan Schneider). Anyway, thanks for all of the responses. My last post (the one that I had meant to send to Stan) pretty much summarizes my status ... and it looks like I can work around this one. Matthew Brown mwbrown@sandia.gov From mwbrown@somnet.sandia.gov Mon Jul 24 08:20:39 1995 From: "Matthew W. Brown" Date: Mon Jul 24 08:20:42 PDT 1995 Subject: Re: booting (errno = 0x3d) Whoops, I sent that last response to the whole group (I had meant to send it to Stan Schneider). Anyway, thanks for all of the responses. My last post (the one that I had meant to send to Stan) pretty much summarizes my status ... and it looks like I can work around this one. Matthew Brown mwbrown@sandia.gov Subject: VGA VME cards Submitted-by km@spacenet.lanl.gov Mon Jul 24 08:51:56 1995 Submitted-by: km@spacenet.lanl.gov (Kim Mcculla) We're considering two boards to handle our VGA output requirements for a MVME167 system. 1. Raster Graphics 741 VxWorks drivers and C-library 2. VIGRA GDM-9000 VxWorks driver At first glance it seems that Raster Graphics has better software support and VIGRA has better resolution. Has anyone had any experience with either or both of these boards? Thanks! -Kim ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Kimberly McCulla E-mail: km@spacenet.lanl.gov Space Data Systems, NIS-3 PO Box 1663, MS-D440 Phone: (505) 667-3295 Los Alamos National Laboratory FAX: (505) 665-4197 Los Alamos, NM 87545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From km@spacenet.lanl.gov Mon Jul 24 08:51:56 1995 From: km@spacenet.lanl.gov (Kim Mcculla) Date: Mon Jul 24 08:51:58 PDT 1995 Subject: VGA VME cards We're considering two boards to handle our VGA output requirements for a MVME167 system. 1. Raster Graphics 741 VxWorks drivers and C-library 2. VIGRA GDM-9000 VxWorks driver At first glance it seems that Raster Graphics has better software support and VIGRA has better resolution. Has anyone had any experience with either or both of these boards? Thanks! -Kim ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Kimberly McCulla E-mail: km@spacenet.lanl.gov Space Data Systems, NIS-3 PO Box 1663, MS-D440 Phone: (505) 667-3295 Los Alamos National Laboratory FAX: (505) 665-4197 Los Alamos, NM 87545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Subject: 386EX. Submitted-by james.larsen@pixie.co.za Mon Jul 24 12:21:46 1995 Submitted-by: james.larsen@pixie.co.za I would like to hear feedback from users of Vxworks with the Intel 386EX. From james.larsen@pixie.co.za Mon Jul 24 12:21:46 1995 From: james.larsen@pixie.co.za Date: Mon Jul 24 12:21:49 PDT 1995 Subject: 386EX. I would like to hear feedback from users of Vxworks with the Intel 386EX. Subject: Re: booting (errno = 0x3d) Submitted-by stan@rti.com Mon Jul 24 13:30:20 1995 Submitted-by: Stan Schneider >> Submitted-by mwbrown@somnet.sandia.gov Mon Jul 24 08:16:14 1995 >> Les Harada (lharad@atl.com) has (probably correctly) suggested that it >> might actually be a function of my host--which is only allowing so many >> socket connections at a time. I intend to explore that a bit. He also >> suggested that I could taskDelay() a bit in my startup scripts with >> longer delays for certain targets--I like that ideal too. You shouldn't have to stagger your boots. Check out the inetd patch; it's a very likely cause. I forget the patch number, but I think there's a discussion in the VxWorks FAQ. -- Stan stan@rti.com From stan@rti.com Mon Jul 24 13:30:20 1995 From: Stan Schneider Date: Mon Jul 24 13:30:52 PDT 1995 Subject: Re: booting (errno = 0x3d) >> Submitted-by mwbrown@somnet.sandia.gov Mon Jul 24 08:16:14 1995 >> Les Harada (lharad@atl.com) has (probably correctly) suggested that it >> might actually be a function of my host--which is only allowing so many >> socket connections at a time. I intend to explore that a bit. He also >> suggested that I could taskDelay() a bit in my startup scripts with >> longer delays for certain targets--I like that ideal too. You shouldn't have to stagger your boots. Check out the inetd patch; it's a very likely cause. I forget the patch number, but I think there's a discussion in the VxWorks FAQ. -- Stan stan@rti.com Subject: No VME address space in 5.2 Submitted-by key@diasonx.com Mon Jul 24 18:22:24 1995 Submitted-by: key@diasonx.com (Charles Key) I have just installed and built 5.2 for mv147, including the customizations that we use in 5.1. The new kernel crashes if I try to access any memory address above the limit of physical memory on the board. We have boards with both 8 and 16 Mb of memory, and each can access just that much address space. Following this, I've built a kernel with no customizations. Same result. Note that I have not changed the sysPhysMemDesc array in sysLib.c. It indicates that there should be 32 Mb of VME address space. Has anyone seen this problem? ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com From key@diasonx.com Mon Jul 24 18:22:24 1995 From: key@diasonx.com (Charles Key) Date: Mon Jul 24 18:22:26 PDT 1995 Subject: No VME address space in 5.2 I have just installed and built 5.2 for mv147, including the customizations that we use in 5.1. The new kernel crashes if I try to access any memory address above the limit of physical memory on the board. We have boards with both 8 and 16 Mb of memory, and each can access just that much address space. Following this, I've built a kernel with no customizations. Same result. Note that I have not changed the sysPhysMemDesc array in sysLib.c. It indicates that there should be 32 Mb of VME address space. Has anyone seen this problem? ___________________________________________________________________________ Charles Key Diasonics Ultrasound key@diasonx.com Subject: Key pressed routine Submitted-by norbert.schillemans@asml.nl Tue Jul 25 01:04:45 1995 Submitted-by: norbert.schillemans@asml.nl (Norbert Schillemans) Hi, I'am trying to develop a routine which will look whether a key is pressed on the keyboard. This routine should not wait for a character to be entered (like getc). I tried using ioctl for this purpose as follows: ioctl(fd, FIONREAD, (int)&no_chr); This should return the number of characters present in the input buffer. For some reason this does not work. I tried different values for "fd", ranging from 0 to 2. We use a Sun which is connected via ethernet with the vxworks cpu (so I do a rlogin to the vxworks cpu from the Sun, load my routine and run it). Does anyone know what I am doing wrong here or have any suggestions? Thanks, Norbert Schillemans From norbert.schillemans@asml.nl Tue Jul 25 01:04:45 1995 From: norbert.schillemans@asml.nl (Norbert Schillemans) Date: Tue Jul 25 01:04:48 PDT 1995 Subject: Key pressed routine Hi, I'am trying to develop a routine which will look whether a key is pressed on the keyboard. This routine should not wait for a character to be entered (like getc). I tried using ioctl for this purpose as follows: ioctl(fd, FIONREAD, (int)&no_chr); This should return the number of characters present in the input buffer. For some reason this does not work. I tried different values for "fd", ranging from 0 to 2. We use a Sun which is connected via ethernet with the vxworks cpu (so I do a rlogin to the vxworks cpu from the Sun, load my routine and run it). Does anyone know what I am doing wrong here or have any suggestions? Thanks, Norbert Schillemans Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Tue Jul 25 04:00:14 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Tue Jul 25 04:00:10 PDT 1995 Subject: Running 68562 DUSCC at 250kbps Subject: HELP! How to force a structure alignment? Subject: Re: HELP! How to force a structure alignment? Subject: VxWorks port to Motorola MC68360 Quads Subject: Re: UPDATE info: booting vxworks on Solaris system ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Running 68562 DUSCC at 250kbps Date: 24 Jul 1995 20:05:54 GMT From: rmm@frc.ri.cmu.edu (Richard Moore) Organization: Field Robotics Center, Carnegie Mellon University Message-ID: Reply-To: rmm@frc2.frc.ri.cmu.edu (Richard Moore) Hello VxWorks Gurus: I have a Matrix MX-CPU330 VME single board computer which uses the 68562 DUSCC chip for serial i/o. The hardware is apparently set up to run as fast as 250kbps, asynch. Unfortunately, the board support package included the generic m68562Serial.c driver, which only supports up to 38.4kbps. I understand that 38.4kbps is the fastest baud rate available from the DUSCC's internal clock and that the CPU330 board is capable of generating the required 4MHz external clock required for the 250kbps. Has anyone modified the 68562 serial device driver to go at faster rates? I don't mind hacking on this, but if somebody has already done it and wouldn't mind sharing at least some pointers, I would greatly appreciate it. Thanks, Rich - -- +------------------------------+------------------------ |Richard Moore | Office: FRC 219 |Senior Research Engineer | Phone: (412) 268-8208 |Field Robotics Center | Fax: (412) 268-5895 |Carnegie Mellon University | rmm@frc.ri.cmu.edu +------------------------------+------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: HELP! How to force a structure alignment? Date: Mon, 24 Jul 1995 16:22:38 -0700 From: cobarruvias@asd1.jsc.nasa.gov (John R. Cobarruvias) Organization: NASA/JSC Message-ID: I inherited code which assumes a fullword boundary in structures. So a struct: struct DATE { unsigned short day ; char hour ; char minute ; char second ; }; /* End DATE TOTAL OF 5 BYTES! */ and one using the above struct: struct EVENT_LOG { struct DATE date ; int eventNum ; }; /* End structure of events TOTAL OF 9 BYTES! */ The above EVENT_LOG struct will align eventNum on a fullword boundary for a total of 12 bytes. MY problem is the board we are using now, PEP PVM30, aligns on a halfword boundary. This is ok with in the board, but the data of the structs are being provided to another box via the serial port. This box assumes a full word boundary! The easiest fix for all this is to force the gnu c compiler to align on fullwords. Is there a switch to do this with the compiler? _ _ _ / _ /_ _ /_/ / ` _ /_ _ _ _ . _ _ (_/ /_// // / / \ /_, /_//_//_|/ / /_/|// /_|_\ John R. Cobarruvias cobarruvias@asd1.jsc.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: HELP! How to force a structure alignment? Date: 24 Jul 1995 22:11:00 GMT From: ericr@vcd.hp.com (Eric Ross) Organization: Hewlett-Packard Message-ID: <3v15pk$eom@news.vcd.hp.com> References: Look at the #pragma align and #pragma pack commands to the C preprocessor. - -- Eric Ross | Hewlett-Packard Company, VPR ericr@vcd.hp.com | Vancouver, WA, USA (360)212-2372 --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks port to Motorola MC68360 Quads Date: 25 Jul 1995 01:20:25 GMT From: Basil Ma Organization: InterNex ISDN Internet Access is our Business Message-ID: <3v1gsp$cpi@voyager.internex.net> I'm doing a VxWorks port to the Motorola MC68360 Quads board. Is the MC68360QUADS bsp, QUICC Ethernet, and VxWorks 683xx object appropriate places to start? bma@iwv.com (415) 261-6200 x129 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UPDATE info: booting vxworks on Solaris system Date: 25 Jul 1995 04:30:47 GMT From: jjwang@clark.net (Jason J Wang) Organization: Clark Internet Services, Inc., Ellicott City, MD USA Message-ID: <3v1s1n$d3e@clarknet.clark.net> References: <9507211813.AA18154@newkla.kla.com> I had a similar problem that caused the the vxWorks board to hang while booting from Solaris (the problem did not seem to occur when booting from a SunOS4.1.3 Machine). Someone here in c.o.vxworks suggested that I edit my boot image (the bootrom_uncmp, not the actual kernel) to be an even number of bytes if it happens to compile out to an odd number of bytes. This addresses a known bug in either Solaris or vxWorks (I can't recall which). Usually, I just append a space onto the end of my boot image. Hope this helps! - -jason --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Tue Jul 25 04:00:14 1995 From: daemon@csg.lbl.gov Date: Tue Jul 25 04:00:17 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Jul 25 04:00:10 PDT 1995 Subject: Running 68562 DUSCC at 250kbps Subject: HELP! How to force a structure alignment? Subject: Re: HELP! How to force a structure alignment? Subject: VxWorks port to Motorola MC68360 Quads Subject: Re: UPDATE info: booting vxworks on Solaris system ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Running 68562 DUSCC at 250kbps Date: 24 Jul 1995 20:05:54 GMT From: rmm@frc.ri.cmu.edu (Richard Moore) Organization: Field Robotics Center, Carnegie Mellon University Message-ID: Reply-To: rmm@frc2.frc.ri.cmu.edu (Richard Moore) Hello VxWorks Gurus: I have a Matrix MX-CPU330 VME single board computer which uses the 68562 DUSCC chip for serial i/o. The hardware is apparently set up to run as fast as 250kbps, asynch. Unfortunately, the board support package included the generic m68562Serial.c driver, which only supports up to 38.4kbps. I understand that 38.4kbps is the fastest baud rate available from the DUSCC's internal clock and that the CPU330 board is capable of generating the required 4MHz external clock required for the 250kbps. Has anyone modified the 68562 serial device driver to go at faster rates? I don't mind hacking on this, but if somebody has already done it and wouldn't mind sharing at least some pointers, I would greatly appreciate it. Thanks, Rich - -- +------------------------------+------------------------ |Richard Moore | Office: FRC 219 |Senior Research Engineer | Phone: (412) 268-8208 |Field Robotics Center | Fax: (412) 268-5895 |Carnegie Mellon University | rmm@frc.ri.cmu.edu +------------------------------+------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: HELP! How to force a structure alignment? Date: Mon, 24 Jul 1995 16:22:38 -0700 From: cobarruvias@asd1.jsc.nasa.gov (John R. Cobarruvias) Organization: NASA/JSC Message-ID: I inherited code which assumes a fullword boundary in structures. So a struct: struct DATE { unsigned short day ; char hour ; char minute ; char second ; }; /* End DATE TOTAL OF 5 BYTES! */ and one using the above struct: struct EVENT_LOG { struct DATE date ; int eventNum ; }; /* End structure of events TOTAL OF 9 BYTES! */ The above EVENT_LOG struct will align eventNum on a fullword boundary for a total of 12 bytes. MY problem is the board we are using now, PEP PVM30, aligns on a halfword boundary. This is ok with in the board, but the data of the structs are being provided to another box via the serial port. This box assumes a full word boundary! The easiest fix for all this is to force the gnu c compiler to align on fullwords. Is there a switch to do this with the compiler? _ _ _ / _ /_ _ /_/ / ` _ /_ _ _ _ . _ _ (_/ /_// // / / \ /_, /_//_//_|/ / /_/|// /_|_\ John R. Cobarruvias cobarruvias@asd1.jsc.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: Re: HELP! How to force a structure alignment? Date: 24 Jul 1995 22:11:00 GMT From: ericr@vcd.hp.com (Eric Ross) Organization: Hewlett-Packard Message-ID: <3v15pk$eom@news.vcd.hp.com> References: Look at the #pragma align and #pragma pack commands to the C preprocessor. - -- Eric Ross | Hewlett-Packard Company, VPR ericr@vcd.hp.com | Vancouver, WA, USA (360)212-2372 --------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks port to Motorola MC68360 Quads Date: 25 Jul 1995 01:20:25 GMT From: Basil Ma Organization: InterNex ISDN Internet Access is our Business Message-ID: <3v1gsp$cpi@voyager.internex.net> I'm doing a VxWorks port to the Motorola MC68360 Quads board. Is the MC68360QUADS bsp, QUICC Ethernet, and VxWorks 683xx object appropriate places to start? bma@iwv.com (415) 261-6200 x129 --------------------------- Newsgroups: comp.os.vxworks Subject: Re: UPDATE info: booting vxworks on Solaris system Date: 25 Jul 1995 04:30:47 GMT From: jjwang@clark.net (Jason J Wang) Organization: Clark Internet Services, Inc., Ellicott City, MD USA Message-ID: <3v1s1n$d3e@clarknet.clark.net> References: <9507211813.AA18154@newkla.kla.com> I had a similar problem that caused the the vxWorks board to hang while booting from Solaris (the problem did not seem to occur when booting from a SunOS4.1.3 Machine). Someone here in c.o.vxworks suggested that I edit my boot image (the bootrom_uncmp, not the actual kernel) to be an even number of bytes if it happens to compile out to an odd number of bytes. This addresses a known bug in either Solaris or vxWorks (I can't recall which). Usually, I just append a space onto the end of my boot image. Hope this helps! - -jason --------------------------- End of New-News digest ********************** Subject: Re: Key pressed routine Submitted-by froeber@bbn.com Tue Jul 25 05:17:59 1995 Submitted-by: Fred Roeber Norbert asked: > >I'am trying to develop a routine which will look whether >a key is pressed on the keyboard. This routine should >not wait for a character to be entered (like getc). > >I tried using ioctl for this purpose as follows: >ioctl(fd, FIONREAD, (int)&no_chr); The main thing I think that is missing is that the terminal line has to be set to "raw" mode instead of the default "line buffered" mode in order to get a keystroke right away. We use the code below to set the terminal into raw mode before starting operation: /* * set terminal to raw mode to get key right away (disable ^C too so that * user must exit in controlled fashion) */ rc = ioctl(0, FIOSETOPTIONS, OPT_ECHO | OPT_TANDEM | OPT_7_BIT | OPT_CRMOD); assert(rc != ERROR); When we are done and want to restore normal terminal operation we use the following code: /* restore terminal to normal mode */ rc = ioctl(0, FIOSETOPTIONS, OPT_TERMINAL); and finally, we use the following routine to read characters. Hope this helps. Fred | Fred J Roeber, Bolt Beranek and Newman Inc | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | /*************************************************************************** * keyhit - check if operator entered a key signifying end of run * * Flush the key and return TRUE if operator has typed a key */ STATIC int keyhit(void) { int nread; int rc; /* probe STDIN to see if operator entered any characters */ rc = ioctl(0, FIONREAD, (int) &nread); assert(rc != ERROR); if (nread) { /* got a char */ /* flush all the characters entered */ for (; nread; nread--) getchar(); printf("got keyhit\n"); return (TRUE); } return (FALSE); } From froeber@bbn.com Tue Jul 25 05:17:59 1995 From: Fred Roeber Date: Tue Jul 25 05:18:01 PDT 1995 Subject: Re: Key pressed routine Norbert asked: > >I'am trying to develop a routine which will look whether >a key is pressed on the keyboard. This routine should >not wait for a character to be entered (like getc). > >I tried using ioctl for this purpose as follows: >ioctl(fd, FIONREAD, (int)&no_chr); The main thing I think that is missing is that the terminal line has to be set to "raw" mode instead of the default "line buffered" mode in order to get a keystroke right away. We use the code below to set the terminal into raw mode before starting operation: /* * set terminal to raw mode to get key right away (disable ^C too so that * user must exit in controlled fashion) */ rc = ioctl(0, FIOSETOPTIONS, OPT_ECHO | OPT_TANDEM | OPT_7_BIT | OPT_CRMOD); assert(rc != ERROR); When we are done and want to restore normal terminal operation we use the following code: /* restore terminal to normal mode */ rc = ioctl(0, FIOSETOPTIONS, OPT_TERMINAL); and finally, we use the following routine to read characters. Hope this helps. Fred | Fred J Roeber, Bolt Beranek and Newman Inc | | 50 Enterprise Place Middletown, RI 02842-5202 | | froeber@bbn.com 401-849-2543 (X48) | /*************************************************************************** * keyhit - check if operator entered a key signifying end of run * * Flush the key and return TRUE if operator has typed a key */ STATIC int keyhit(void) { int nread; int rc; /* probe STDIN to see if operator entered any characters */ rc = ioctl(0, FIONREAD, (int) &nread); assert(rc != ERROR); if (nread) { /* got a char */ /* flush all the characters entered */ for (; nread; nread--) getchar(); printf("got keyhit\n"); return (TRUE); } return (FALSE); } Subject: X window type converter Submitted-by pdw@eel-mail.mc.duke.edu Tue Jul 25 10:30:48 1995 Submitted-by: "Patrick D. Wolf" I just switched from windx to vxWindows and I am seeing some of the following warnings. Anybody know the cause and solution? Warning: No type converter registered for 'TranslationTable' to 'Btn1Down' conversion. Warning: No type converter registered for 'TranslationTable' to 'InitialResourcesPersistent' conversion. Warning: No type converter registered for 'ButtonPress' to 'KeyRelease' conversion. Warning: No type converter registered for 'TranslationTable' to 'Color' conversion. Warning: No type converter registered for 'TranslationTable' to 'Translations' conversion. Warning: No type converter registered for 'TranslationTable' to 'InitialResourcesPersistent' conversion. Warning: No type converter registered for 'ButtonPress' to 'Btn2Up' conversion. Thanks in advance, Patrick Wolf Department of Biomedical Engineering Duke University Durham, NC 27708 (919) 660-5114 pdw@eel-mail.mc.duke.edu From pdw@eel-mail.mc.duke.edu Tue Jul 25 10:30:48 1995 From: "Patrick D. Wolf" Date: Tue Jul 25 10:30:51 PDT 1995 Subject: X window type converter I just switched from windx to vxWindows and I am seeing some of the following warnings. Anybody know the cause and solution? Warning: No type converter registered for 'TranslationTable' to 'Btn1Down' conversion. Warning: No type converter registered for 'TranslationTable' to 'InitialResourcesPersistent' conversion. Warning: No type converter registered for 'ButtonPress' to 'KeyRelease' conversion. Warning: No type converter registered for 'TranslationTable' to 'Color' conversion. Warning: No type converter registered for 'TranslationTable' to 'Translations' conversion. Warning: No type converter registered for 'TranslationTable' to 'InitialResourcesPersistent' conversion. Warning: No type converter registered for 'ButtonPress' to 'Btn2Up' conversion. Thanks in advance, Patrick Wolf Department of Biomedical Engineering Duke University Durham, NC 27708 (919) 660-5114 pdw@eel-mail.mc.duke.edu Subject: Re: Key pressed routine Submitted-by stan@rti.com Tue Jul 25 11:25:51 1995 Submitted-by: Stan Schneider >> Submitted-by norbert.schillemans@asml.nl Tue Jul 25 01:04:45 1995 >> Submitted-by: norbert.schillemans@asml.nl (Norbert Schillemans) >> >> Hi, >> >> I'am trying to develop a routine which will look whether >> a key is pressed on the keyboard. This routine should >> not wait for a character to be entered (like getc). >> >> I tried using ioctl for this purpose as follows: >> ioctl(fd, FIONREAD, (int)&no_chr); >> >> This should return the number of characters present in the >> input buffer. For some reason this does not work. >> I tried different values for "fd", ranging from 0 to 2. >> >> Does anyone know what I am doing wrong here or have any suggestions? You're missing data that may be in the stream buffer. This always returns the number of characters ready to read (taken from RTILib's QueryKeyReady routine): long n; ioctl(0, FIONREAD, &n); /* Anything in the channel? */ n += (stdin)->_r; /* or in buffer? */ -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= From stan@rti.com Tue Jul 25 11:25:51 1995 From: Stan Schneider Date: Tue Jul 25 11:25:54 PDT 1995 Subject: Re: Key pressed routine >> Submitted-by norbert.schillemans@asml.nl Tue Jul 25 01:04:45 1995 >> Submitted-by: norbert.schillemans@asml.nl (Norbert Schillemans) >> >> Hi, >> >> I'am trying to develop a routine which will look whether >> a key is pressed on the keyboard. This routine should >> not wait for a character to be entered (like getc). >> >> I tried using ioctl for this purpose as follows: >> ioctl(fd, FIONREAD, (int)&no_chr); >> >> This should return the number of characters present in the >> input buffer. For some reason this does not work. >> I tried different values for "fd", ranging from 0 to 2. >> >> Does anyone know what I am doing wrong here or have any suggestions? You're missing data that may be in the stream buffer. This always returns the number of characters ready to read (taken from RTILib's QueryKeyReady routine): long n; ioctl(0, FIONREAD, &n); /* Anything in the channel? */ n += (stdin)->_r; /* or in buffer? */ -- Stan ============================================================================= = = = = Stan Schneider = email: stan@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 954 Aster, Sunnyvale, CA 94086 = Fax: (408) 720-8419 = = = = ============================================================================= Subject: subscribe Submitted-by wilhite_david@ccmail.ncsc.navy.mil Tue Jul 25 13:45:10 1995 Submitted-by: wilhite_david@ccmail.ncsc.navy.mil subscribe From wilhite_david@ccmail.ncsc.navy.mil Tue Jul 25 13:45:10 1995 From: wilhite_david@ccmail.ncsc.navy.mil Date: Tue Jul 25 13:45:13 PDT 1995 Subject: subscribe subscribe Subject: gcc Cross for i86 Submitted-by vince@lassen.rti.com Tue Jul 25 15:13:57 1995 Submitted-by: vince@rti.com (Vince Chen) Does anyone have experience with building a gcc cross-compiler for VxWorks with 80x86 as the target and SunOS4 (or Solaris) as the host? --target=i386-wrs-vxworks --host=sparc-sun-sunos4 or sparc-sun-solaris2 I'm not sure what to use for the configuration files for the target side, but have created some based on other vxworks config files. My first attempt seems to build a working compiler, but I'm having trouble with libgcc.a. It contains unresolved symbols for things like __exit and _cleanup. It also includes things like _muldi3, _divdi3, etc. I don't remember these being included in the library for the m68k cross tools. Any help would be appreciated! -vince P.S. I'm trying to build for 2.6.x. I don't think this configuration is supported "out-of-the-box" even for 2.7.0. From vince@lassen.rti.com Tue Jul 25 15:13:57 1995 From: vince@rti.com (Vince Chen) Date: Tue Jul 25 15:13:59 PDT 1995 Subject: gcc Cross for i86 Does anyone have experience with building a gcc cross-compiler for VxWorks with 80x86 as the target and SunOS4 (or Solaris) as the host? --target=i386-wrs-vxworks --host=sparc-sun-sunos4 or sparc-sun-solaris2 I'm not sure what to use for the configuration files for the target side, but have created some based on other vxworks config files. My first attempt seems to build a working compiler, but I'm having trouble with libgcc.a. It contains unresolved symbols for things like __exit and _cleanup. It also includes things like _muldi3, _divdi3, etc. I don't remember these being included in the library for the m68k cross tools. Any help would be appreciated! -vince P.S. I'm trying to build for 2.6.x. I don't think this configuration is supported "out-of-the-box" even for 2.7.0. Subject: re: Key pressed Thanks! Submitted-by norbert.schillemans@asml.nl Wed Jul 26 00:51:20 1995 Submitted-by: norbert.schillemans@asml.nl (Norbert Schillemans) Hi, I indeed forgot to put the terminal in raw mode. It works fine now. Thanks, Norbert Schillemans From norbert.schillemans@asml.nl Wed Jul 26 00:51:20 1995 From: norbert.schillemans@asml.nl (Norbert Schillemans) Date: Wed Jul 26 00:51:23 PDT 1995 Subject: re: Key pressed Thanks! Hi, I indeed forgot to put the terminal in raw mode. It works fine now. Thanks, Norbert Schillemans Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Wed Jul 26 04:00:48 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Wed Jul 26 04:00:27 PDT 1995 Subject: Re: re: unsubscribe Subject: Problems linking a 68040 VxWorks kernel with gnu-ld 2.5.2 Subject: MC68302 Performance Data Subject: LAPD (ITU Q.921) Subject: Re: Key pressed routine Subject: US-MI-Detroit Job Opening: Realtime Support Engineer Subject: Re: PCMCIA for 68xxx ? Subject: Re: Key pressed routine ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: re: unsubscribe Date: 25 Jul 1995 14:45:20 GMT From: Monica Abitol Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <3v3020$pff@post.gsfc.nasa.gov> References: <9507122010.AA55616@source.asset.com> How do you subscribe to the vxworks exploder? what's the email address? Thanks, Monica --------------------------- Newsgroups: comp.os.vxworks Subject: Problems linking a 68040 VxWorks kernel with gnu-ld 2.5.2 Date: 25 Jul 1995 13:44:40 GMT From: Ninane@fynu.ucl.ac.be Organization: University of Louvain (LLN) - Nuclear Physics Dept. Message-ID: <3v2sg8$b3o@sci3.sri.ucl.ac.be> Folks, I have a problem when linking the VxWorks kernel for a 68040 (MVME167) with gnu-ld68k (v2.5.2 with BFD 2.5 - configured for vxworks) : > ... stuff deleted ... > ld68k -X -N -e _sysInit -Ttext 00001000 \ > sysALib.o sysLib.o sysLocalLib.o tyCoDrv.o usrConfig.o version.o > ../../lib/68040/rdb.a ../../lib/68040/drv.a ../../lib/68040/all.a \ > ../../lib/68040/config.a ../../lib/68040/netif.a ../../lib/68040/net.a \ > ../../lib/68040/rpc.a ../../lib/68040/wind.a ../../lib/68040/all.a \ > ../../lib/68040/netif.a ../../lib/68040/68k.a ../../lib/68040/config.a \ > ../../lib/68040/all.a ../../lib/68040/libm.a * excFppALib.o(.text+0x112): relocation truncated to fit: DISP16 _x_fpsp_unfl * excFppALib.o(.text+0xfe): relocation truncated to fit: DISP16 _x_fpsp_ovfl * excFppALib.o(.text+0xea): relocation truncated to fit: DISP16 _x_fpsp_snan * do_func.o(.data+0x178): relocation truncated to fit: DISP16 _x_t_dz2 * do_func.o(.data+0x174): relocation truncated to fit: DISP16 _x_slognp1 * srem_mod.o(.data+0x2d8): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xc2): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xaa): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xa4): relocation truncated to fit: 16 .data * ... stuff deleted ... * l_srem_mod.o(.data+0x2d8): relocation truncated to fit: 16 .data * *** Error code 1 * make: Fatal error: Command failed for target `vxWorks' sysALib.o, sysLib.o, ... have been compiled without any problems with gnu-cc68k 2.6.3. My development platform is a Solaris 2.4 system. I'm still using VxWorks 5.0.2b. Any hints ? Alain - -- Dr. Alain H. Ninane | Tel : +32-10-47.32.32 - Fax: +32-10-45.21.83 University of Louvain | Internet: Ninane@fynu.ucl.ac.be Nuclear Physics Dept. | Ch. du Cyclotron, 2 B-1348 Louvain-la-Neuve | BELGIUM --------------------------- Newsgroups: comp.os.vxworks Subject: MC68302 Performance Data Date: 25 Jul 1995 15:02:02 GMT From: Bill Gutknecht Organization: BNR/Nortel Message-ID: <3v311a$9ic@nrtphba6.bnr.ca> Hello ... Does anyone have any performance data (in particular ISR/Task switch lacency) for a 68302? Data for a 68000/68008 would suffice! Thanks! Bill - ------------------------------------------------------------------------------ Bill Gutknecht "If I die, I will go before Crom and he will BNR/Northern Telecom ask me 'What is the Riddle of Steel?' If I Research Triangle Park, NC do not know it, he will cast me out of billag@bnr.ca Valhalla and laugh at me ... " -Conan - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: LAPD (ITU Q.921) Date: 25 Jul 1995 15:24:18 GMT From: Bill Gutknecht Organization: BNR/Nortel Message-ID: <3v32b2$a7f@nrtphba6.bnr.ca> Has anyone had any experience/recommendations for vendors that sell ISDN protocol stacks? In particular, LAPD layer 2 protocol (ITU Q.921). - ------------------------------------------------------------------------------ Bill Gutknecht "If I die, I will go before Crom and he will BNR/Nortel ask me 'What is the Riddle of Steel?' If I Research Triangle Park, NC do not know it, he will cast me out of billag@bnr.ca Valhalla and laugh at me ... " -Conan - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: 25 Jul 1995 20:06:30 GMT From: rypma@waterloo.hp.com (Ted Rypma) Organization: H-P Panacom Div, Waterloo, ON Canada Message-ID: <3v3is6$657@hppadbk.waterloo.hp.com> References: <9507250804.AA01917@mihr.asml.nl> Norbert Schillemans (norbert.schillemans@asml.nl) wrote: : I tried using ioctl for this purpose as follows: : ioctl(fd, FIONREAD, (int)&no_chr); : This should return the number of characters present in the : input buffer. For some reason this does not work. Do you have the serial port in LINE or RAW mode? FIONREAD returns the number of bytes available to read. In LINE mode, nothing is available to be read until a newline is read. Ted Rypma Waterloo, Ontario --------------------------- Newsgroups: mi.jobs,comp.os.vxworks,comp.realtime,comp.arch.embedded,misc.jobs.offered Subject: US-MI-Detroit Job Opening: Realtime Support Engineer Date: 25 Jul 1995 22:47:54 GMT From: dwybo@ix.netcom.com (David R. Wybo ) Organization: Netcom Message-ID: <3v3saq$laq@ixnews6.ix.netcom.com> Wind River Systems, the leading supplier of real-time operating systems, has openings at their Detroit MI facilities. Technical Support Engineer, Automotive Applications Location: Detroit Michigan area. Responsibilities: Respond to a variety of technical support calls for Wind River's deeply embedded operating system for in-vehicle applications, WindStream, and associated development tools. Requirements: BSCS or EE or equivalent experience. Excellent verbal and written skills. RTOS programming and development, C programming. Understanding of UNIX and/or MS Windows OS internals. Team oriented workskills. Automotive electronics application experience, VxWorks experience a plus. Contact: Dave Wybo dwybo@wrs.com or Dave Leahy dleahy@wrs.com Wind River Systems 39111 W. Six Mile Rd. Livonia, MI 48152 313-591-7220 313-591-7283 FAX --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: 24 Jul 1995 11:05:09 +0200 From: jj@ekf.werries.de ("J. Jansen") Organization: EKF Elektronik GmbH, Hamm FRG Message-ID: <3uvno5$5f8@ekf.werries.de> References: <3um0v9$2h5@tekadm1.cse.tek.com> fitz@aim.cse.tek.com (Bob Fitzsimmons) writes: >I'm working on a product with an embedded 68xxx system. We're interested in >supporting several peripheral devices including: extended memory (ram disks), >and serial ports. Because of our small form factor, we're most interested >in providing this support via PCMCIA cards. >Questions: >1. Which of the available bus adapters is most compatible with 68xxx? >2. Do any vendors provide 68xxx driver support? >Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx >embedded environment would be greatly appreciated. >Thanks in advance, >Bob Fitzsimmons >Tektonix, Inc. >Measurement Business Division >(503) 627-5777 EKF offers 3/6HU VMEbus PC Card (PCMCIA/Jeida) adapter boards. OS-9 support includes ATA drivers (harddisk, sundisk flash). Anybody interested in a free data sheet feel free to write to info@ekf.werrries.de Please do not forget to include your full postal address, thank you! - --------------------- EKF Elektronik Messtechnik GmbH --------------------- Joachim Jansen e-mail: jj@ekf.werries.de Philipp Reis Str. 4 phone: ++49 (0)2381-6890-0 D-59065 Hamm (Germany) fax: ++49 (0)2381-6890-90 - ------------------- Advanced VMEbus Hardware + Software ------------------- - -- - --------------------------------------------------------------------------- jj@ekf.werries.de - --------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: 26 Jul 1995 04:23:52 GMT From: "Mark R. Milligan" Organization: Coquitlam Custom Computing Message-ID: <3v4g0o$fpi@sulla.cyberstore.ca> References: <9507250804.AA01917@mihr.asml.nl> norbert.schillemans@asml.nl (Norbert Schillemans) wrote: >Hi, > >I'am trying to develop a routine which will look whether >a key is pressed on the keyboard. This routine should >not wait for a character to be entered (like getc). > >I tried using ioctl for this purpose as follows: >ioctl(fd, FIONREAD, (int)&no_chr); > >This should return the number of characters present in the >input buffer. For some reason this does not work. >I tried different values for "fd", ranging from 0 to 2. > >Does anyone know what I am doing wrong here or have any suggestions? > This is a classic problem with VxWorks. The shell is running at a higher priority than your task and is stealing all the characters pressed on the keyboard. I got around this by using another serial port and attaching a terminal to it. This allowed me to send data to the terminal and get responses and still use the shell for debugging and control etc. You can try to kill the shell or lower it's priority. Anyone else found a work around for this quandry ??? Mark R. Milligan --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Wed Jul 26 04:00:48 1995 From: daemon@csg.lbl.gov Date: Wed Jul 26 04:00:52 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Wed Jul 26 04:00:27 PDT 1995 Subject: Re: re: unsubscribe Subject: Problems linking a 68040 VxWorks kernel with gnu-ld 2.5.2 Subject: MC68302 Performance Data Subject: LAPD (ITU Q.921) Subject: Re: Key pressed routine Subject: US-MI-Detroit Job Opening: Realtime Support Engineer Subject: Re: PCMCIA for 68xxx ? Subject: Re: Key pressed routine ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: re: unsubscribe Date: 25 Jul 1995 14:45:20 GMT From: Monica Abitol Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <3v3020$pff@post.gsfc.nasa.gov> References: <9507122010.AA55616@source.asset.com> How do you subscribe to the vxworks exploder? what's the email address? Thanks, Monica --------------------------- Newsgroups: comp.os.vxworks Subject: Problems linking a 68040 VxWorks kernel with gnu-ld 2.5.2 Date: 25 Jul 1995 13:44:40 GMT From: Ninane@fynu.ucl.ac.be Organization: University of Louvain (LLN) - Nuclear Physics Dept. Message-ID: <3v2sg8$b3o@sci3.sri.ucl.ac.be> Folks, I have a problem when linking the VxWorks kernel for a 68040 (MVME167) with gnu-ld68k (v2.5.2 with BFD 2.5 - configured for vxworks) : > ... stuff deleted ... > ld68k -X -N -e _sysInit -Ttext 00001000 \ > sysALib.o sysLib.o sysLocalLib.o tyCoDrv.o usrConfig.o version.o > ../../lib/68040/rdb.a ../../lib/68040/drv.a ../../lib/68040/all.a \ > ../../lib/68040/config.a ../../lib/68040/netif.a ../../lib/68040/net.a \ > ../../lib/68040/rpc.a ../../lib/68040/wind.a ../../lib/68040/all.a \ > ../../lib/68040/netif.a ../../lib/68040/68k.a ../../lib/68040/config.a \ > ../../lib/68040/all.a ../../lib/68040/libm.a * excFppALib.o(.text+0x112): relocation truncated to fit: DISP16 _x_fpsp_unfl * excFppALib.o(.text+0xfe): relocation truncated to fit: DISP16 _x_fpsp_ovfl * excFppALib.o(.text+0xea): relocation truncated to fit: DISP16 _x_fpsp_snan * do_func.o(.data+0x178): relocation truncated to fit: DISP16 _x_t_dz2 * do_func.o(.data+0x174): relocation truncated to fit: DISP16 _x_slognp1 * srem_mod.o(.data+0x2d8): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xc2): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xaa): relocation truncated to fit: 16 .data * l_scosh.o(.data+0xa4): relocation truncated to fit: 16 .data * ... stuff deleted ... * l_srem_mod.o(.data+0x2d8): relocation truncated to fit: 16 .data * *** Error code 1 * make: Fatal error: Command failed for target `vxWorks' sysALib.o, sysLib.o, ... have been compiled without any problems with gnu-cc68k 2.6.3. My development platform is a Solaris 2.4 system. I'm still using VxWorks 5.0.2b. Any hints ? Alain - -- Dr. Alain H. Ninane | Tel : +32-10-47.32.32 - Fax: +32-10-45.21.83 University of Louvain | Internet: Ninane@fynu.ucl.ac.be Nuclear Physics Dept. | Ch. du Cyclotron, 2 B-1348 Louvain-la-Neuve | BELGIUM --------------------------- Newsgroups: comp.os.vxworks Subject: MC68302 Performance Data Date: 25 Jul 1995 15:02:02 GMT From: Bill Gutknecht Organization: BNR/Nortel Message-ID: <3v311a$9ic@nrtphba6.bnr.ca> Hello ... Does anyone have any performance data (in particular ISR/Task switch lacency) for a 68302? Data for a 68000/68008 would suffice! Thanks! Bill - ------------------------------------------------------------------------------ Bill Gutknecht "If I die, I will go before Crom and he will BNR/Northern Telecom ask me 'What is the Riddle of Steel?' If I Research Triangle Park, NC do not know it, he will cast me out of billag@bnr.ca Valhalla and laugh at me ... " -Conan - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: LAPD (ITU Q.921) Date: 25 Jul 1995 15:24:18 GMT From: Bill Gutknecht Organization: BNR/Nortel Message-ID: <3v32b2$a7f@nrtphba6.bnr.ca> Has anyone had any experience/recommendations for vendors that sell ISDN protocol stacks? In particular, LAPD layer 2 protocol (ITU Q.921). - ------------------------------------------------------------------------------ Bill Gutknecht "If I die, I will go before Crom and he will BNR/Nortel ask me 'What is the Riddle of Steel?' If I Research Triangle Park, NC do not know it, he will cast me out of billag@bnr.ca Valhalla and laugh at me ... " -Conan - ------------------------------------------------------------------------------ --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: 25 Jul 1995 20:06:30 GMT From: rypma@waterloo.hp.com (Ted Rypma) Organization: H-P Panacom Div, Waterloo, ON Canada Message-ID: <3v3is6$657@hppadbk.waterloo.hp.com> References: <9507250804.AA01917@mihr.asml.nl> Norbert Schillemans (norbert.schillemans@asml.nl) wrote: : I tried using ioctl for this purpose as follows: : ioctl(fd, FIONREAD, (int)&no_chr); : This should return the number of characters present in the : input buffer. For some reason this does not work. Do you have the serial port in LINE or RAW mode? FIONREAD returns the number of bytes available to read. In LINE mode, nothing is available to be read until a newline is read. Ted Rypma Waterloo, Ontario --------------------------- Newsgroups: mi.jobs,comp.os.vxworks,comp.realtime,comp.arch.embedded,misc.jobs.offered Subject: US-MI-Detroit Job Opening: Realtime Support Engineer Date: 25 Jul 1995 22:47:54 GMT From: dwybo@ix.netcom.com (David R. Wybo ) Organization: Netcom Message-ID: <3v3saq$laq@ixnews6.ix.netcom.com> Wind River Systems, the leading supplier of real-time operating systems, has openings at their Detroit MI facilities. Technical Support Engineer, Automotive Applications Location: Detroit Michigan area. Responsibilities: Respond to a variety of technical support calls for Wind River's deeply embedded operating system for in-vehicle applications, WindStream, and associated development tools. Requirements: BSCS or EE or equivalent experience. Excellent verbal and written skills. RTOS programming and development, C programming. Understanding of UNIX and/or MS Windows OS internals. Team oriented workskills. Automotive electronics application experience, VxWorks experience a plus. Contact: Dave Wybo dwybo@wrs.com or Dave Leahy dleahy@wrs.com Wind River Systems 39111 W. Six Mile Rd. Livonia, MI 48152 313-591-7220 313-591-7283 FAX --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: 24 Jul 1995 11:05:09 +0200 From: jj@ekf.werries.de ("J. Jansen") Organization: EKF Elektronik GmbH, Hamm FRG Message-ID: <3uvno5$5f8@ekf.werries.de> References: <3um0v9$2h5@tekadm1.cse.tek.com> fitz@aim.cse.tek.com (Bob Fitzsimmons) writes: >I'm working on a product with an embedded 68xxx system. We're interested in >supporting several peripheral devices including: extended memory (ram disks), >and serial ports. Because of our small form factor, we're most interested >in providing this support via PCMCIA cards. >Questions: >1. Which of the available bus adapters is most compatible with 68xxx? >2. Do any vendors provide 68xxx driver support? >Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx >embedded environment would be greatly appreciated. >Thanks in advance, >Bob Fitzsimmons >Tektonix, Inc. >Measurement Business Division >(503) 627-5777 EKF offers 3/6HU VMEbus PC Card (PCMCIA/Jeida) adapter boards. OS-9 support includes ATA drivers (harddisk, sundisk flash). Anybody interested in a free data sheet feel free to write to info@ekf.werrries.de Please do not forget to include your full postal address, thank you! - --------------------- EKF Elektronik Messtechnik GmbH --------------------- Joachim Jansen e-mail: jj@ekf.werries.de Philipp Reis Str. 4 phone: ++49 (0)2381-6890-0 D-59065 Hamm (Germany) fax: ++49 (0)2381-6890-90 - ------------------- Advanced VMEbus Hardware + Software ------------------- - -- - --------------------------------------------------------------------------- jj@ekf.werries.de - --------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: 26 Jul 1995 04:23:52 GMT From: "Mark R. Milligan" Organization: Coquitlam Custom Computing Message-ID: <3v4g0o$fpi@sulla.cyberstore.ca> References: <9507250804.AA01917@mihr.asml.nl> norbert.schillemans@asml.nl (Norbert Schillemans) wrote: >Hi, > >I'am trying to develop a routine which will look whether >a key is pressed on the keyboard. This routine should >not wait for a character to be entered (like getc). > >I tried using ioctl for this purpose as follows: >ioctl(fd, FIONREAD, (int)&no_chr); > >This should return the number of characters present in the >input buffer. For some reason this does not work. >I tried different values for "fd", ranging from 0 to 2. > >Does anyone know what I am doing wrong here or have any suggestions? > This is a classic problem with VxWorks. The shell is running at a higher priority than your task and is stealing all the characters pressed on the keyboard. I got around this by using another serial port and attaching a terminal to it. This allowed me to send data to the terminal and get responses and still use the shell for debugging and control etc. You can try to kill the shell or lower it's priority. Anyone else found a work around for this quandry ??? Mark R. Milligan --------------------------- End of New-News digest ********************** Subject: fopen() problem (again) Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 Submitted-by: "David B. Perry" Some how I got temporarily un-subscribed from vxwexplo and may have missed any suggestions that came in for my problem. Once more, here it is... MUCH earlier I wrote: > I am experiencing a problem with the fopen routine. I am accessing files > resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D (68040/VxWorks > 5.1.1) accross Ethernet. The fopen routine will open and close the files, but > it does not return a NULL file pointer when the file does not exist (as > documented). I have performed the fopen function from code and from the shell > and in both cases the fopen routine does not recognize that the file does not > exist. RSH is being used for the remote file access method. > > Has anyone seen this problem before? Wind River has not come up with an answer... Sergi Casas responded with the suggestion to look at errno. It was a good idea (thanks!) (why didn't WE think of it...), but unfortunately, did not help as errno was not set. I think the fault is in the driver (both the original problem & lack of errno). Leonid Rosenboim (leonid@rst.co.il) responded with: Try to use FTP instead of RSH. in RSH you can easily damage the netDrv functionality be including commands that emit output in .cshrc file of the user to which the target is set. Your help is appreciated! David -- ===================================================================== David B. Perry e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== From perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 From: "David B. Perry" Date: Wed Jul 26 06:00:10 PDT 1995 Subject: fopen() problem (again) Some how I got temporarily un-subscribed from vxwexplo and may have missed any suggestions that came in for my problem. Once more, here it is... MUCH earlier I wrote: > I am experiencing a problem with the fopen routine. I am accessing files > resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D (68040/VxWorks > 5.1.1) accross Ethernet. The fopen routine will open and close the files, but > it does not return a NULL file pointer when the file does not exist (as > documented). I have performed the fopen function from code and from the shell > and in both cases the fopen routine does not recognize that the file does not > exist. RSH is being used for the remote file access method. > > Has anyone seen this problem before? Wind River has not come up with an answer... Sergi Casas responded with the suggestion to look at errno. It was a good idea (thanks!) (why didn't WE think of it...), but unfortunately, did not help as errno was not set. I think the fault is in the driver (both the original problem & lack of errno). Leonid Rosenboim (leonid@rst.co.il) responded with: Try to use FTP instead of RSH. in RSH you can easily damage the netDrv functionality be including commands that emit output in .cshrc file of the user to which the target is set. Your help is appreciated! David -- ===================================================================== David B. Perry e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== Subject: Radstone SCSI problems Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:19:11 1995 Submitted-by: "David B. Perry" Folks, We have seen several problems with Radstone boards, in particular the CPU40D-213E and CPU40D-212E (68040 cpus). The current show-stopper is this: when reading from or writing a SCSI drive using the dosFS, the task performing the action hangs after some number of operations. I did a tt on the non-running task and got the following as the last few entries: ... _dosFsReadyChange _scsiWrtSecs _scsiTransact _scsiReqSense _scsiBusReset +64: _semQPut([...]) In addition, most tasks with a priority number greater than 1 get stuck at the same place (!). We have been talking to Radstone since the end of June, but they have not provided any solutions - or even any suggestions. I can supply more information, like a test file that reliably hangs, to anyone interested. Thanks, David -- ===================================================================== David B. Perry TRW e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== From perryd@boris.lbl.gov Wed Jul 26 06:19:11 1995 From: "David B. Perry" Date: Wed Jul 26 06:19:13 PDT 1995 Subject: Radstone SCSI problems Folks, We have seen several problems with Radstone boards, in particular the CPU40D-213E and CPU40D-212E (68040 cpus). The current show-stopper is this: when reading from or writing a SCSI drive using the dosFS, the task performing the action hangs after some number of operations. I did a tt on the non-running task and got the following as the last few entries: ... _dosFsReadyChange _scsiWrtSecs _scsiTransact _scsiReqSense _scsiBusReset +64: _semQPut([...]) In addition, most tasks with a priority number greater than 1 get stuck at the same place (!). We have been talking to Radstone since the end of June, but they have not provided any solutions - or even any suggestions. I can supply more information, like a test file that reliably hangs, to anyone interested. Thanks, David -- ===================================================================== David B. Perry TRW e-mail: perryd@boris.msfc.nasa.gov 213 Wynn Dr. phone: (205) 961-1033 Huntsville, AL 35805 fax: (205) 544-2238 ===================================================================== Subject: Re: fopen() problem (again) Submitted-by bwenholz@pacesetter.com Wed Jul 26 09:38:07 1995 Submitted-by: bwenholz@pacesetter.com (Bruce Wenholz) }Subject: fopen() problem (again) } }Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 }Submitted-by: "David B. Perry" } }Some how I got temporarily un-subscribed from vxwexplo and may have missed any } suggestions that came in for my problem. Once more, here it is... } }MUCH earlier I wrote: } }> I am experiencing a problem with the fopen routine. I am accessing files }> resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D }(68040/VxWorks }> 5.1.1) accross Ethernet. The fopen routine will open and close the files, }but }> it does not return a NULL file pointer when the file does not exist (as }> documented). I have performed the fopen function from code and from the }shell }> and in both cases the fopen routine does not recognize that the file does not }> exist. RSH is being used for the remote file access method. }> }> Has anyone seen this problem before? Wind River has not come up with an }answer... } }Sergi Casas responded with the suggestion to look at errno. I have a problem that seem like the source would be fopen. I am doing an open for wright. The file discriptor is assigned and the file seems to be open. I can write the file. The problem comes when I close the file, and the directory does not exist. That is when VxWorks actually opens the file but since the directory does not exist, my file is lost. From bwenholz@pacesetter.com Wed Jul 26 09:38:07 1995 From: bwenholz@pacesetter.com (Bruce Wenholz) Date: Wed Jul 26 09:38:15 PDT 1995 Subject: Re: fopen() problem (again) }Subject: fopen() problem (again) } }Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 }Submitted-by: "David B. Perry" } }Some how I got temporarily un-subscribed from vxwexplo and may have missed any } suggestions that came in for my problem. Once more, here it is... } }MUCH earlier I wrote: } }> I am experiencing a problem with the fopen routine. I am accessing files }> resident on an SGI Challenge L (IRIX 5.2) from a Radstone CPU40D }(68040/VxWorks }> 5.1.1) accross Ethernet. The fopen routine will open and close the files, }but }> it does not return a NULL file pointer when the file does not exist (as }> documented). I have performed the fopen function from code and from the }shell }> and in both cases the fopen routine does not recognize that the file does not }> exist. RSH is being used for the remote file access method. }> }> Has anyone seen this problem before? Wind River has not come up with an }answer... } }Sergi Casas responded with the suggestion to look at errno. I have a problem that seem like the source would be fopen. I am doing an open for wright. The file discriptor is assigned and the file seems to be open. I can write the file. The problem comes when I close the file, and the directory does not exist. That is when VxWorks actually opens the file but since the directory does not exist, my file is lost. Subject: ARINC575 Submitted-by eugenel@condorsys.com Wed Jul 26 10:45:24 1995 Submitted-by: eugenel@condorsys.com (Eugene Leung X714) Anyone knows if there is a vendor suppling VME ARINC575 interface card or a ARINC 575 IP w vxWorks driver? Thanks in advance. Eugene ================================================================================ - - - Eugene Leung + email: eugenel@condorsys.com - - Sr. S/W Engineer + phone: (408)371-9580 ext 714 - - Condor Systems Inc. + fax: (408)371-9589 - - 2133 Samaritan Dr. - - San Jose, CA 95124 - - - ================================================================================ From eugenel@condorsys.com Wed Jul 26 10:45:24 1995 From: eugenel@condorsys.com (Eugene Leung X714) Date: Wed Jul 26 10:45:26 PDT 1995 Subject: ARINC575 Anyone knows if there is a vendor suppling VME ARINC575 interface card or a ARINC 575 IP w vxWorks driver? Thanks in advance. Eugene ================================================================================ - - - Eugene Leung + email: eugenel@condorsys.com - - Sr. S/W Engineer + phone: (408)371-9580 ext 714 - - Condor Systems Inc. + fax: (408)371-9589 - - 2133 Samaritan Dr. - - San Jose, CA 95124 - - - ================================================================================ Subject: UK: Job Opportunity Submitted-by klf@redifon.demon.co.uk Wed Jul 26 10:57:39 1995 Submitted-by: Kieren Feakes Permanent position, salary depending on experience to 28K Redifon MEL Technology Division ------------------------------- This is a well established company specialising in development and consultancy for complex real-time embedded systems. Location : Surrey/Sussex, UK Position (permanent) -------------------- We are at the start of a large project to develop a leading-edge sonar receiver processing system for a government agency. The system entails the use of a number of DSP devices to perform various signal processing functions, a number of off-the-shelf sonar interface boards, control processors running VxWorks and an X11 based Motif GUI. Extremely high data throughput rates are involved. Essential Requirements ---------------------- British Citizen Programming in C (4 years min) Real-time O/S skills Systems Development and Integration Self motivated team player Work under own steam to deadlines Good degree in computing/electronics (2.1 Hons or above ideally) Development experience using Unix/Suns. Desirable --------- VxWorks experience including :- Writing device drivers Use of TCP and UDP sockets Intertask/processor communications High speed VME transfers. ATM knowledge. Solaris 2.4 experience X11/Motif Hardware interfacing SPOX O/S DSP techniques Working to QA standards such as ISO 9001 & TickIT Please send CV/resumes to: -------------------------- <> Kieren Feakes <> * <> <> <> Redifon Technology <> * <> Tel (01372) 376677 <> <> 9 Mole Business Park, <> * o * <> Fax (01372) 379480 <> <> Randalls Road, <> * <\> <> klf@redifon.demon.co.uk <> <> Leatherhead, Surrey <> __ /_\__, <> <> <> KT22 7BA, England <> <> <> From klf@redifon.demon.co.uk Wed Jul 26 10:57:39 1995 From: Kieren Feakes Date: Wed Jul 26 10:57:41 PDT 1995 Subject: UK: Job Opportunity Permanent position, salary depending on experience to 28K Redifon MEL Technology Division ------------------------------- This is a well established company specialising in development and consultancy for complex real-time embedded systems. Location : Surrey/Sussex, UK Position (permanent) -------------------- We are at the start of a large project to develop a leading-edge sonar receiver processing system for a government agency. The system entails the use of a number of DSP devices to perform various signal processing functions, a number of off-the-shelf sonar interface boards, control processors running VxWorks and an X11 based Motif GUI. Extremely high data throughput rates are involved. Essential Requirements ---------------------- British Citizen Programming in C (4 years min) Real-time O/S skills Systems Development and Integration Self motivated team player Work under own steam to deadlines Good degree in computing/electronics (2.1 Hons or above ideally) Development experience using Unix/Suns. Desirable --------- VxWorks experience including :- Writing device drivers Use of TCP and UDP sockets Intertask/processor communications High speed VME transfers. ATM knowledge. Solaris 2.4 experience X11/Motif Hardware interfacing SPOX O/S DSP techniques Working to QA standards such as ISO 9001 & TickIT Please send CV/resumes to: -------------------------- <> Kieren Feakes <> * <> <> <> Redifon Technology <> * <> Tel (01372) 376677 <> <> 9 Mole Business Park, <> * o * <> Fax (01372) 379480 <> <> Randalls Road, <> * <\> <> klf@redifon.demon.co.uk <> <> Leatherhead, Surrey <> __ /_\__, <> <> <> KT22 7BA, England <> <> <> Subject: Forcing Structure Alignment Submitted-by nichols@lx34.nl.nuwc.navy.mil Wed Jul 26 14:58:05 1995 Submitted-by: George Nichols Data alignment is a common problem on code dealing with fixed data types and ported across processors and operation systems. Constant porting problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove me to the following solutions: Define fixed data types in header file keyed to CPU and OS, (ie): typedef char INT8; typedef short INT16; typedef long INT32; Always manually group/align elements in a data structure to 32bits, (ie): struct DEMO { INT8 this[5]; INT8 pad1[3]; INT16 that[3]; INT16 pad2[1]; INT32 theother[4]; }; Grouping elements of similar types together minimizes the overall number of pads require(and wasted space). This technique produces trackable data structures when data is communicated from one CPU/OS to a different CPU/OS. ~ ~ ~ ~ ~ ~ ~ | ~ ~ ~ ~ | ~ ~ | ~ ____ | | _O _|_ | | | ____|_|__ | | | \ | / ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ George Nichols Naval Undersea Warfare Center (NUWC) Submarine Electromagnetics Department (34) Communication Systems Division (341) Systems Analysis Branch (3411) Detachment New London Phone: 203-440-5503 Fax: 203-440-4007 Email: nichols@lx34.nl.nuwc.navy.mil From nichols@lx34.nl.nuwc.navy.mil Wed Jul 26 14:58:05 1995 From: George Nichols Date: Wed Jul 26 14:58:08 PDT 1995 Subject: Forcing Structure Alignment Data alignment is a common problem on code dealing with fixed data types and ported across processors and operation systems. Constant porting problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove me to the following solutions: Define fixed data types in header file keyed to CPU and OS, (ie): typedef char INT8; typedef short INT16; typedef long INT32; Always manually group/align elements in a data structure to 32bits, (ie): struct DEMO { INT8 this[5]; INT8 pad1[3]; INT16 that[3]; INT16 pad2[1]; INT32 theother[4]; }; Grouping elements of similar types together minimizes the overall number of pads require(and wasted space). This technique produces trackable data structures when data is communicated from one CPU/OS to a different CPU/OS. ~ ~ ~ ~ ~ ~ ~ | ~ ~ ~ ~ | ~ ~ | ~ ____ | | _O _|_ | | | ____|_|__ | | | \ | / ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ George Nichols Naval Undersea Warfare Center (NUWC) Submarine Electromagnetics Department (34) Communication Systems Division (341) Systems Analysis Branch (3411) Detachment New London Phone: 203-440-5503 Fax: 203-440-4007 Email: nichols@lx34.nl.nuwc.navy.mil Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Thu Jul 27 04:00:16 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Thu Jul 27 04:00:13 PDT 1995 Subject: Re: Max File descriptors Subject: Re: PCMCIA for 68xxx ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max File descriptors Date: 26 Jul 1995 21:16:54 GMT From: randyc@soda.CSUA.Berkeley.EDU (Randy Yen-pang Chou) Organization: Computer Science Undergrad Assoc., Univ. of Calif. Berkeley Message-ID: <3v6bc6$p4n@agate.berkeley.edu> References: <9507211146.AA07939@rst.co.il> In article <9507211146.AA07939@rst.co.il>, Leonid Rosenboim wrote: >Dear VxWorrkers, > >have you ever reached the VxWorks limit on file descriptors / sockets ? >Have you ever tried to define NUM_FILES in configAll.h to a higher >value than the OS could support ? If so, what was the upper >absolute limit ? > Yes. We've increased NUM_FILES to 255. Anything higher and you won't be able to use RPCs unless WRS recompiles the RPC libraries for you. - ---Randy --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: Wed, 26 Jul 1995 17:32:44 GMT From: olin@netcom.com (B.W. 'Ollie' Olin) Organization: NOT Message-ID: References: <3um0v9$2h5@tekadm1.cse.tek.com> <3uvno5$5f8@ekf.werries.de> Sender: olin@netcom9.netcom.com jj@ekf.werries.de ("J. Jansen") wrote: >Anybody interested in a free data sheet feel free to write to > info@ekf.werrries.de ^^^^^^^^ For those interested in sending them e-mail, note the type-O. It should read info@ekf.werries.de (only two Rs). This has been incorrectly repeated several times, please correct your master text file. Thanks. Ollie - -- B.W. 'Ollie' Olin "Reality plays no significant role in this mess." olin@netcom.com Unknown --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Thu Jul 27 04:00:16 1995 From: daemon@csg.lbl.gov Date: Thu Jul 27 04:00:19 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Jul 27 04:00:13 PDT 1995 Subject: Re: Max File descriptors Subject: Re: PCMCIA for 68xxx ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Max File descriptors Date: 26 Jul 1995 21:16:54 GMT From: randyc@soda.CSUA.Berkeley.EDU (Randy Yen-pang Chou) Organization: Computer Science Undergrad Assoc., Univ. of Calif. Berkeley Message-ID: <3v6bc6$p4n@agate.berkeley.edu> References: <9507211146.AA07939@rst.co.il> In article <9507211146.AA07939@rst.co.il>, Leonid Rosenboim wrote: >Dear VxWorrkers, > >have you ever reached the VxWorks limit on file descriptors / sockets ? >Have you ever tried to define NUM_FILES in configAll.h to a higher >value than the OS could support ? If so, what was the upper >absolute limit ? > Yes. We've increased NUM_FILES to 255. Anything higher and you won't be able to use RPCs unless WRS recompiles the RPC libraries for you. - ---Randy --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: Wed, 26 Jul 1995 17:32:44 GMT From: olin@netcom.com (B.W. 'Ollie' Olin) Organization: NOT Message-ID: References: <3um0v9$2h5@tekadm1.cse.tek.com> <3uvno5$5f8@ekf.werries.de> Sender: olin@netcom9.netcom.com jj@ekf.werries.de ("J. Jansen") wrote: >Anybody interested in a free data sheet feel free to write to > info@ekf.werrries.de ^^^^^^^^ For those interested in sending them e-mail, note the type-O. It should read info@ekf.werries.de (only two Rs). This has been incorrectly repeated several times, please correct your master text file. Thanks. Ollie - -- B.W. 'Ollie' Olin "Reality plays no significant role in this mess." olin@netcom.com Unknown --------------------------- End of New-News digest ********************** Subject: Re: Max File Descriptors Submitted-by uchenick@tate.com Thu Jul 27 06:23:45 1995 Submitted-by: uchenick@tate.com (Gordon Uchenick) > Subject: Re: Max File descriptors > Date: 26 Jul 1995 21:16:54 GMT > From: randyc@soda.CSUA.Berkeley.EDU (Randy Yen-pang Chou) > Organization: Computer Science Undergrad Assoc., Univ. of Calif. Berkeley > Leonid Rosenboim wrote: > >Dear VxWorrkers, > > > >have you ever reached the VxWorks limit on file descriptors / sockets ? > >Have you ever tried to define NUM_FILES in configAll.h to a higher > >value than the OS could support ? If so, what was the upper > >absolute limit ? > > > > Yes. We've increased NUM_FILES to 255. Anything higher and you won't > be able to use RPCs unless WRS recompiles the RPC libraries for you. > > We have also increased NUM_FILES. If you plan to go past 255, also be aware that you have to change the symbol FD_SETSIZE in h/types/vxTypesOld.h. or the FD_SET macro that is a part of the select() facility will hammer memory. 255 files was enough for us and we stopped there, so there may be others files that need a "tweak". Did anybody else go past 255? Regards, Gordon From uchenick@tate.com Thu Jul 27 06:23:45 1995 From: uchenick@tate.com (Gordon Uchenick) Date: Thu Jul 27 06:23:48 PDT 1995 Subject: Re: Max File Descriptors > Subject: Re: Max File descriptors > Date: 26 Jul 1995 21:16:54 GMT > From: randyc@soda.CSUA.Berkeley.EDU (Randy Yen-pang Chou) > Organization: Computer Science Undergrad Assoc., Univ. of Calif. Berkeley > Leonid Rosenboim wrote: > >Dear VxWorrkers, > > > >have you ever reached the VxWorks limit on file descriptors / sockets ? > >Have you ever tried to define NUM_FILES in configAll.h to a higher > >value than the OS could support ? If so, what was the upper > >absolute limit ? > > > > Yes. We've increased NUM_FILES to 255. Anything higher and you won't > be able to use RPCs unless WRS recompiles the RPC libraries for you. > > We have also increased NUM_FILES. If you plan to go past 255, also be aware that you have to change the symbol FD_SETSIZE in h/types/vxTypesOld.h. or the FD_SET macro that is a part of the select() facility will hammer memory. 255 files was enough for us and we stopped there, so there may be others files that need a "tweak". Did anybody else go past 255? Regards, Gordon Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Fri Jul 28 04:00:21 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Fri Jul 28 04:00:17 PDT 1995 Subject: Booting VxWorks and Unix boot server Subject: Using both VxWorks and native proms? Subject: Localtalk/Appletalk support for VxWorks Subject: Re: Parallel port driver for mv167? Subject: Re: Key pressed routine Subject: HELP! Bus Error with MVME162LX Subject: Re: Forcing Structure Alignment Subject: Re: re: fopen() problem (again) Subject: Re: PCMCIA for 68xxx ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Booting VxWorks and Unix boot server Date: 27 Jul 1995 11:17:05 GMT From: tronds@nftdf.fof.nft.no (Trond Skaar) Organization: NFT, Norsk Forsvarsteknologi A/S, 3600 Kongsberg Keywords: FTP, boot Message-ID: <3v7sjh$s2p@nftdf.nft.no> Sender: tronds@nftdf (Trond Skaar) I have a MicroForce VMEbus rack containing two Force 2CE-80 CPUs. One CPU is running Unix, and the other CPU is running VxWorks.The Unix CPU is a boot server (tftpboot) for the VxWorks CPU. PROBLEM: When I boot the complete rack (either at power-UP, or by executing the reboot command on the Unix CPU) the VxWorks CPU will not boot. If I hit the reset button on the VxWorks CPU after the Unix CPU is up and running, then the VxWorks CPU boots without any problem. Also if I press CTRL-x after the VxWorks CPU has been booted, the VxWorks CPU will reboot without any problem. If anyone has experienced the same problem and found a solution to it (so that the VxWorks CPU always boots), then I would appreciate a description of the solution! DIAGNOSTICS: I logged the printouts from the VxWorks CPU when the boot fails: ---------------------------------------------------------------- WARNING: Unable to determine keyboard type Device not found: screen Can't open input device. Keyboard not present. Using tty for input and output. FORCE CPU-2CE, No Keyboard ROM 2.3.10 P/N 910-12437-106 CPU2CE_00012 32 MB memory installed, Serial #737805. Ethernet address 0:80:42:8:2:d, Host ID: 558b420d Copyright (c) 1990, Sun Microsystems, Inc. Copyright (c) 1992, Force Computers Inc. Testing 1 megs of memory. Still to go 0 SBus slot 0 le esp dma SBus slot 1 SBus slot 2 SBus slot 3 Boot device: /sbus/le@0,c00000 File and args: Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet receive failed The file just loaded does not appear to be executable. Type b (boot), c (continue), or n (new command mode) > - -- Trond Skaar (email: tronds@fof.nft.no) --------------------------- Newsgroups: comp.os.vxworks Subject: Using both VxWorks and native proms? Date: 27 Jul 95 08:43:30 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: proms mv167 Message-ID: <1995Jul27.084331.23497@hrbicf> - -- Hello, Does anyone have experience using both the VxWorks and Motorola proms together? We would like to normally boot our Motorola VME167 cards using VxWorks (5.1.1) and have the option to use the diagnostics embedded in the Motorola proms. Could someone please enlighten us to what is required to allow us to accomplish this? Thanks in advance, - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Localtalk/Appletalk support for VxWorks Date: 27 Jul 1995 00:36:14 GMT From: Eric McLaughlin Organization: Grass Valley Group, Grass Valley, CA Message-ID: <3v6n1u$g6l@gv-gate.gvg.tek.com> Has anyone out there used the Localtalk datalink layer to deliver TCP/IP messages? We were thinking of using it in a similar fashion to that used with the shared memory network (i.e. LocalTalk for message delivery with the TCP/IP protocols). We're running VxWorks on a 68302 processor. By adding a 68195 Local Talk adapter and the appropriate drivers, we have a cheap networking solution. We might also consider using Apple talk on top of LocalTalk assuming it wasn't a huge memory pig and a port to Vxworks exists. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Parallel port driver for mv167? Date: 27 Jul 1995 20:55:26 GMT From: srejto Organization: MIT Lincoln Laboratory Message-ID: <3v8ufu$ab6@llnews.ll.mit.edu> References: <1995Jul18.124958.23460@hrbicf> Call it what you will, this will toggle all 1s then all 0s to the output port. May be usefull to you. srejto@ll.mit.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: Thu, 27 Jul 1995 09:11:21 GMT From: kam@ceb-ws03.oce.nl (Kampen G. van) Organization: OCE Nederland B.V. Message-ID: References: <9507250804.AA01917@mihr.asml.nl> <3v4g0o$fpi@sulla.cyberstore.ca> Sender: news@oce.nl (The Daily News @ nntp01.oce.nl) In article <3v4g0o$fpi@sulla.cyberstore.ca> "Mark R. Milligan" writes: >norbert.schillemans@asml.nl (Norbert Schillemans) wrote: >>Hi, >> >>I'am trying to develop a routine which will look whether >>a key is pressed on the keyboard. This routine should >>not wait for a character to be entered (like getc). >> >>I tried using ioctl for this purpose as follows: >>ioctl(fd, FIONREAD, (int)&no_chr); >> >>This should return the number of characters present in the >>input buffer. For some reason this does not work. >>I tried different values for "fd", ranging from 0 to 2. >> >>Does anyone know what I am doing wrong here or have any suggestions? >> >This is a classic problem with VxWorks. The shell is running at a higher >priority than your task and is stealing all the characters pressed on the >keyboard. > >I got around this by using another serial port and attaching a terminal to >it. This allowed me to send data to the terminal and get responses and still >use the shell for debugging and control etc. > >You can try to kill the shell or lower it's priority. > >Anyone else found a work around for this quandry ??? > >Mark R. Milligan > > I think this problem is caused by the default mode a terminal port is configured at boot time. In my case this is done with the call: ioctl (consoleFd, FIOSETOPTIONS, OPT_TERMINAL); This mode passes characters only after a new-line is received (or after the buffer is full which I'm not sure about). If you want characters to be passed directly you have to call: ioctl (consoleFd, FIOSETOPTIONS, OPT_RAW); Only this may screw up your terminal parsing if you enter commands at the monitor program prompt. In this case try toggling between the two modes. Hope this will do the job for you. #include Gerald van Kampen --------------------------- Newsgroups: comp.os.vxworks Subject: HELP! Bus Error with MVME162LX Date: Thu, 27 Jul 1995 17:53:26 From: lentz@gate.net (Raymond P. Lentz, III) Keywords: Bus Error Message-ID: We have an MVME162LX Motorola card and a high speed parallel interface (HiPPI) card in a 6U VME chassis. We wrote a couple of small routines to get the Motorola card and the HiPPI talking to each other. We created a standalone image with VxWorks 5.1 and put it on the Motorola onboard flash for testing and everything worked well. We then created a standalone PROM version and put it into the Motorola card. When the HiPPI tries to write to the mailbox on the Motorola card we get a VME Bus write error. We think the problem is because the Motorola DRAM is not properly mapped to VME space, but we can't seem to figure out what we have done wrong. Any and all help will be greatly appreciated. Ray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: Thu, 27 Jul 1995 16:52:12 GMT From: geger@phantom.den.mmc.com (George Eger) Organization: Martin Marietta Astronautics Group Message-ID: <1995Jul27.165212.8144@den.mmc.com> Sender: news@den.mmc.com (News Admin) From George Nichols, nichols@lx34.nl.nuwc.navy.mil: >> Data alignment is a common problem on code dealing with fixed data types >> and ported across processors and operation systems. Constant porting >> problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove >> me to the following solutions: >> >> Define fixed data types in header file keyed to CPU and OS, (ie): >> >> typedef char INT8; >> typedef short INT16; >> typedef long INT32; >> >> Always manually group/align elements in a data structure to 32bits, (ie): >> >> struct DEMO { >> INT8 this[5]; >> INT8 pad1[3]; >> INT16 that[3]; >> INT16 pad2[1]; >> INT32 theother[4]; >> }; >> >> Grouping elements of similar types together minimizes the overall number of >> pads require(and wasted space). >> >> This technique produces trackable data structures when data is >> communicated from one CPU/OS to a different CPU/OS. >> >> I have found that forced alignment is very useful, also. I have developed a habit that keeps things aligned with the largest elements being located towards the beginning of the structure: struct dummy { IEEEDouble_t x; /* 0x00 */ IEEEDouble_t y /* 0x08 */ struct timespec now; /* 0x10 */ uint32_t cntr; /* 0x18 */ char name[16]; /* 0x1C */ int32_t pad[1]; /* 0x2C */ }; /* 0x30 total */ I try to keep the byte offset of the structure in a comment - that way, when having to use an assembly level debugger or looking at memory dumps, you have a quick way of telling where the various members are. I try to round to an even 16 bytes (or an even power of 2) for the structure size. Various compilers and OS's calculate offsets or allocate blocks that are powers of 2 more easily/quickly, so you might be able to gain a little speed by keeping the blocks a convenient size. There are exceptions, such as when you have pointers for linked lists. On the 680x0 and other processors, doing register indirect with offset is not as fast an addressing mode as just doing register indirect. If you're having to walk like lists to the end, you can save a few clock cycles every element by having the address to the next element in the first member. You can also have a generic set of functions (hmmm, sound's like methods) that operate on just the link pointers for various types of lists (with potentially different list elements but common members for the list things). struct dummy { struct dummy next; /* 0x00 */ struct dummy prev; /* 0x04 */ IEEEDouble_t x; /* 0x08 */ IEEEDouble_t y /* 0x10 */ struct timespec now; /* 0x18 */ uint32_t cntr; /* 0x20 */ char name[16]; /* 0x24 */ int32_t pad[3]; /* 0x30 */ }; /* 0x40 total */ I've used these 'rules' to port these kinds of structures between various flavors of Unix, VxWorks, Mac OS, Windows, Apollo Domain, bare-bones R3000's and other HW & SW and had very few problems. I have seen problems with some RISC processors and interactions between the virtual memory system and the compiler that caused even FORTRAN programs with poorly aligned common blocks to die, so I know how ugly this stuff can get. HTH, GWE ||========================================================================== ||George Eger / geger@den.mmc.com || Voice - (303) - 971 - 6974 || ||Avionics and Ground Systems || Fax - (303) - 977 - 1145 || ||Advanced Launch Systems || MS T320 || ||Lockheed Martin Astronautics || P.O. Box 179, Denver CO 80201 || ||========================================================================== ||We are at a cusp - between the past when humans were more reliable than || ||computers and the future when computers are more reliable than humans. || ||========================================================================== ||All opinions (however truthful or misinformed) are my own. || ||========================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: re: fopen() problem (again) Date: 28 Jul 1995 07:40:55 GMT From: Bruce Wenholz Organization: Mini-Micronets Co. Message-ID: <3va4a7$jql@mars.earthlink.net> References: <9507261636.AA02405@austin.Pacesetter.Com> bwenholz@pacesetter.com (Bruce Wenholz) wrote: >}Subject: fopen() problem (again) >}Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 >}Submitted-by: "David B. Perry" [...] >}MUCH earlier I wrote: >} >}> Has anyone seen this problem before? Wind River has not come up with an >}answer... >} >}Sergi Casas responded with the suggestion to look at errno. > >I have a problem that seems like the source would be fopen. I am doing an >open for write. The file discriptor is assigned and the file seems to be >open. I can write the file. The problem comes when I close the file, and the >directory does not exist. That is when VxWorks actually opens the file but >since the directory does not exist, my file is lost. David: in reply to my messge in reply to your message, my solution follows. I am using a general open that acquires the file name from 1 of several sources all of which feed a common message queue with character input. The solution that I cam up with today at work is as follows. errno = 0; *p_fd = open(...); if ( errno ) { printErrno(errno); if ( *p_fd > ) close ( *p_fd ); *p_fd = ERROR; } if ( *p_fd == ERROR ) { Do normal failure to open processing. } else { Do normal sucessful open processing. } Yes, replying to my own message from my home account. I hope this helps who ever it was that was having the problem with the fopen(); - -- Mini-Micronets: FidoNet: TBD Internet whoforest@eartlink.net --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: Sun, 23 Jul 1995 17:01:55 -0800 From: darnell@primenet.com (Darnell Gadberry) Organization: Phonica Message-ID: References: <3um0v9$2h5@tekadm1.cse.tek.com> In article <3um0v9$2h5@tekadm1.cse.tek.com>, fitz@aim.cse.tek.com wrote: > I'm working on a product with an embedded 68xxx system. We're interested in > supporting several peripheral devices including: extended memory (ram disks), > and serial ports. Because of our small form factor, we're most interested > in providing this support via PCMCIA cards. > > Questions: > > 1. Which of the available bus adapters is most compatible with 68xxx? > 2. Do any vendors provide 68xxx driver support? > > Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx > embedded environment would be greatly appreciated. > > Thanks in advance, > Bob Fitzsimmons > > Tektonix, Inc. > Measurement Business Division > (503) 627-5777 Bob, We went down this path several months ago. Here is what I discovered after a week of research: 1. I was not able to find a single vendor that produced a PCMCIA adapter IC that was directly bus compatible with the 68K family. Unfortunately, the 80x86 family dominates the PCMCIA world. Therefore, you will need to create a pseudo ISA bus (or PCI buss) for talking to one of the Cirrus, Omega, or Zilog controller chips. 2. The latest release of VXWorks OS from WindRiver Systems has fairly complete PCMCIA support. This could dramatically reduce your development time. However, VXWorks is a bit pricey. (13.5K dev system + royalties) 3. If you do not use an OS like VXWorks, you will need to write Card and Socket Services drivers for every device you want to support. Good Luck! Darnell Gadberry Phonica Digital Audio Products darnell@primenet.com --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Fri Jul 28 04:00:21 1995 From: daemon@csg.lbl.gov Date: Fri Jul 28 04:00:25 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Jul 28 04:00:17 PDT 1995 Subject: Booting VxWorks and Unix boot server Subject: Using both VxWorks and native proms? Subject: Localtalk/Appletalk support for VxWorks Subject: Re: Parallel port driver for mv167? Subject: Re: Key pressed routine Subject: HELP! Bus Error with MVME162LX Subject: Re: Forcing Structure Alignment Subject: Re: re: fopen() problem (again) Subject: Re: PCMCIA for 68xxx ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Booting VxWorks and Unix boot server Date: 27 Jul 1995 11:17:05 GMT From: tronds@nftdf.fof.nft.no (Trond Skaar) Organization: NFT, Norsk Forsvarsteknologi A/S, 3600 Kongsberg Keywords: FTP, boot Message-ID: <3v7sjh$s2p@nftdf.nft.no> Sender: tronds@nftdf (Trond Skaar) I have a MicroForce VMEbus rack containing two Force 2CE-80 CPUs. One CPU is running Unix, and the other CPU is running VxWorks.The Unix CPU is a boot server (tftpboot) for the VxWorks CPU. PROBLEM: When I boot the complete rack (either at power-UP, or by executing the reboot command on the Unix CPU) the VxWorks CPU will not boot. If I hit the reset button on the VxWorks CPU after the Unix CPU is up and running, then the VxWorks CPU boots without any problem. Also if I press CTRL-x after the VxWorks CPU has been booted, the VxWorks CPU will reboot without any problem. If anyone has experienced the same problem and found a solution to it (so that the VxWorks CPU always boots), then I would appreciate a description of the solution! DIAGNOSTICS: I logged the printouts from the VxWorks CPU when the boot fails: ---------------------------------------------------------------- WARNING: Unable to determine keyboard type Device not found: screen Can't open input device. Keyboard not present. Using tty for input and output. FORCE CPU-2CE, No Keyboard ROM 2.3.10 P/N 910-12437-106 CPU2CE_00012 32 MB memory installed, Serial #737805. Ethernet address 0:80:42:8:2:d, Host ID: 558b420d Copyright (c) 1990, Sun Microsystems, Inc. Copyright (c) 1992, Force Computers Inc. Testing 1 megs of memory. Still to go 0 SBus slot 0 le esp dma SBus slot 1 SBus slot 2 SBus slot 3 Boot device: /sbus/le@0,c00000 File and args: Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet receive failed The file just loaded does not appear to be executable. Type b (boot), c (continue), or n (new command mode) > - -- Trond Skaar (email: tronds@fof.nft.no) --------------------------- Newsgroups: comp.os.vxworks Subject: Using both VxWorks and native proms? Date: 27 Jul 95 08:43:30 EST From: jmw@hrb.com (Joel M. Whitesel) Organization: HRB Systems, Inc. Keywords: proms mv167 Message-ID: <1995Jul27.084331.23497@hrbicf> - -- Hello, Does anyone have experience using both the VxWorks and Motorola proms together? We would like to normally boot our Motorola VME167 cards using VxWorks (5.1.1) and have the option to use the diagnostics embedded in the Motorola proms. Could someone please enlighten us to what is required to allow us to accomplish this? Thanks in advance, - ------------------------------------------------------------------------- Joel M. Whitesel HRB Systems jmw@hrb.com P.O. Box 60 State College, PA 16804 x2780 (814) 238-4311 - ------------------------------------------------------------------------- --------------------------- Newsgroups: comp.os.vxworks Subject: Localtalk/Appletalk support for VxWorks Date: 27 Jul 1995 00:36:14 GMT From: Eric McLaughlin Organization: Grass Valley Group, Grass Valley, CA Message-ID: <3v6n1u$g6l@gv-gate.gvg.tek.com> Has anyone out there used the Localtalk datalink layer to deliver TCP/IP messages? We were thinking of using it in a similar fashion to that used with the shared memory network (i.e. LocalTalk for message delivery with the TCP/IP protocols). We're running VxWorks on a 68302 processor. By adding a 68195 Local Talk adapter and the appropriate drivers, we have a cheap networking solution. We might also consider using Apple talk on top of LocalTalk assuming it wasn't a huge memory pig and a port to Vxworks exists. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Parallel port driver for mv167? Date: 27 Jul 1995 20:55:26 GMT From: srejto Organization: MIT Lincoln Laboratory Message-ID: <3v8ufu$ab6@llnews.ll.mit.edu> References: <1995Jul18.124958.23460@hrbicf> Call it what you will, this will toggle all 1s then all 0s to the output port. May be usefull to you. srejto@ll.mit.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Key pressed routine Date: Thu, 27 Jul 1995 09:11:21 GMT From: kam@ceb-ws03.oce.nl (Kampen G. van) Organization: OCE Nederland B.V. Message-ID: References: <9507250804.AA01917@mihr.asml.nl> <3v4g0o$fpi@sulla.cyberstore.ca> Sender: news@oce.nl (The Daily News @ nntp01.oce.nl) In article <3v4g0o$fpi@sulla.cyberstore.ca> "Mark R. Milligan" writes: >norbert.schillemans@asml.nl (Norbert Schillemans) wrote: >>Hi, >> >>I'am trying to develop a routine which will look whether >>a key is pressed on the keyboard. This routine should >>not wait for a character to be entered (like getc). >> >>I tried using ioctl for this purpose as follows: >>ioctl(fd, FIONREAD, (int)&no_chr); >> >>This should return the number of characters present in the >>input buffer. For some reason this does not work. >>I tried different values for "fd", ranging from 0 to 2. >> >>Does anyone know what I am doing wrong here or have any suggestions? >> >This is a classic problem with VxWorks. The shell is running at a higher >priority than your task and is stealing all the characters pressed on the >keyboard. > >I got around this by using another serial port and attaching a terminal to >it. This allowed me to send data to the terminal and get responses and still >use the shell for debugging and control etc. > >You can try to kill the shell or lower it's priority. > >Anyone else found a work around for this quandry ??? > >Mark R. Milligan > > I think this problem is caused by the default mode a terminal port is configured at boot time. In my case this is done with the call: ioctl (consoleFd, FIOSETOPTIONS, OPT_TERMINAL); This mode passes characters only after a new-line is received (or after the buffer is full which I'm not sure about). If you want characters to be passed directly you have to call: ioctl (consoleFd, FIOSETOPTIONS, OPT_RAW); Only this may screw up your terminal parsing if you enter commands at the monitor program prompt. In this case try toggling between the two modes. Hope this will do the job for you. #include Gerald van Kampen --------------------------- Newsgroups: comp.os.vxworks Subject: HELP! Bus Error with MVME162LX Date: Thu, 27 Jul 1995 17:53:26 From: lentz@gate.net (Raymond P. Lentz, III) Keywords: Bus Error Message-ID: We have an MVME162LX Motorola card and a high speed parallel interface (HiPPI) card in a 6U VME chassis. We wrote a couple of small routines to get the Motorola card and the HiPPI talking to each other. We created a standalone image with VxWorks 5.1 and put it on the Motorola onboard flash for testing and everything worked well. We then created a standalone PROM version and put it into the Motorola card. When the HiPPI tries to write to the mailbox on the Motorola card we get a VME Bus write error. We think the problem is because the Motorola DRAM is not properly mapped to VME space, but we can't seem to figure out what we have done wrong. Any and all help will be greatly appreciated. Ray --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: Thu, 27 Jul 1995 16:52:12 GMT From: geger@phantom.den.mmc.com (George Eger) Organization: Martin Marietta Astronautics Group Message-ID: <1995Jul27.165212.8144@den.mmc.com> Sender: news@den.mmc.com (News Admin) From George Nichols, nichols@lx34.nl.nuwc.navy.mil: >> Data alignment is a common problem on code dealing with fixed data types >> and ported across processors and operation systems. Constant porting >> problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove >> me to the following solutions: >> >> Define fixed data types in header file keyed to CPU and OS, (ie): >> >> typedef char INT8; >> typedef short INT16; >> typedef long INT32; >> >> Always manually group/align elements in a data structure to 32bits, (ie): >> >> struct DEMO { >> INT8 this[5]; >> INT8 pad1[3]; >> INT16 that[3]; >> INT16 pad2[1]; >> INT32 theother[4]; >> }; >> >> Grouping elements of similar types together minimizes the overall number of >> pads require(and wasted space). >> >> This technique produces trackable data structures when data is >> communicated from one CPU/OS to a different CPU/OS. >> >> I have found that forced alignment is very useful, also. I have developed a habit that keeps things aligned with the largest elements being located towards the beginning of the structure: struct dummy { IEEEDouble_t x; /* 0x00 */ IEEEDouble_t y /* 0x08 */ struct timespec now; /* 0x10 */ uint32_t cntr; /* 0x18 */ char name[16]; /* 0x1C */ int32_t pad[1]; /* 0x2C */ }; /* 0x30 total */ I try to keep the byte offset of the structure in a comment - that way, when having to use an assembly level debugger or looking at memory dumps, you have a quick way of telling where the various members are. I try to round to an even 16 bytes (or an even power of 2) for the structure size. Various compilers and OS's calculate offsets or allocate blocks that are powers of 2 more easily/quickly, so you might be able to gain a little speed by keeping the blocks a convenient size. There are exceptions, such as when you have pointers for linked lists. On the 680x0 and other processors, doing register indirect with offset is not as fast an addressing mode as just doing register indirect. If you're having to walk like lists to the end, you can save a few clock cycles every element by having the address to the next element in the first member. You can also have a generic set of functions (hmmm, sound's like methods) that operate on just the link pointers for various types of lists (with potentially different list elements but common members for the list things). struct dummy { struct dummy next; /* 0x00 */ struct dummy prev; /* 0x04 */ IEEEDouble_t x; /* 0x08 */ IEEEDouble_t y /* 0x10 */ struct timespec now; /* 0x18 */ uint32_t cntr; /* 0x20 */ char name[16]; /* 0x24 */ int32_t pad[3]; /* 0x30 */ }; /* 0x40 total */ I've used these 'rules' to port these kinds of structures between various flavors of Unix, VxWorks, Mac OS, Windows, Apollo Domain, bare-bones R3000's and other HW & SW and had very few problems. I have seen problems with some RISC processors and interactions between the virtual memory system and the compiler that caused even FORTRAN programs with poorly aligned common blocks to die, so I know how ugly this stuff can get. HTH, GWE ||========================================================================== ||George Eger / geger@den.mmc.com || Voice - (303) - 971 - 6974 || ||Avionics and Ground Systems || Fax - (303) - 977 - 1145 || ||Advanced Launch Systems || MS T320 || ||Lockheed Martin Astronautics || P.O. Box 179, Denver CO 80201 || ||========================================================================== ||We are at a cusp - between the past when humans were more reliable than || ||computers and the future when computers are more reliable than humans. || ||========================================================================== ||All opinions (however truthful or misinformed) are my own. || ||========================================================================== --------------------------- Newsgroups: comp.os.vxworks Subject: Re: re: fopen() problem (again) Date: 28 Jul 1995 07:40:55 GMT From: Bruce Wenholz Organization: Mini-Micronets Co. Message-ID: <3va4a7$jql@mars.earthlink.net> References: <9507261636.AA02405@austin.Pacesetter.Com> bwenholz@pacesetter.com (Bruce Wenholz) wrote: >}Subject: fopen() problem (again) >}Submitted-by perryd@boris.lbl.gov Wed Jul 26 06:00:06 1995 >}Submitted-by: "David B. Perry" [...] >}MUCH earlier I wrote: >} >}> Has anyone seen this problem before? Wind River has not come up with an >}answer... >} >}Sergi Casas responded with the suggestion to look at errno. > >I have a problem that seems like the source would be fopen. I am doing an >open for write. The file discriptor is assigned and the file seems to be >open. I can write the file. The problem comes when I close the file, and the >directory does not exist. That is when VxWorks actually opens the file but >since the directory does not exist, my file is lost. David: in reply to my messge in reply to your message, my solution follows. I am using a general open that acquires the file name from 1 of several sources all of which feed a common message queue with character input. The solution that I cam up with today at work is as follows. errno = 0; *p_fd = open(...); if ( errno ) { printErrno(errno); if ( *p_fd > ) close ( *p_fd ); *p_fd = ERROR; } if ( *p_fd == ERROR ) { Do normal failure to open processing. } else { Do normal sucessful open processing. } Yes, replying to my own message from my home account. I hope this helps who ever it was that was having the problem with the fopen(); - -- Mini-Micronets: FidoNet: TBD Internet whoforest@eartlink.net --------------------------- Newsgroups: comp.sys.m68k,comp.os.vxworks Subject: Re: PCMCIA for 68xxx ? Date: Sun, 23 Jul 1995 17:01:55 -0800 From: darnell@primenet.com (Darnell Gadberry) Organization: Phonica Message-ID: References: <3um0v9$2h5@tekadm1.cse.tek.com> In article <3um0v9$2h5@tekadm1.cse.tek.com>, fitz@aim.cse.tek.com wrote: > I'm working on a product with an embedded 68xxx system. We're interested in > supporting several peripheral devices including: extended memory (ram disks), > and serial ports. Because of our small form factor, we're most interested > in providing this support via PCMCIA cards. > > Questions: > > 1. Which of the available bus adapters is most compatible with 68xxx? > 2. Do any vendors provide 68xxx driver support? > > Any other advice on feasibillity of supporting PCMCIA cards in a 68xxx > embedded environment would be greatly appreciated. > > Thanks in advance, > Bob Fitzsimmons > > Tektonix, Inc. > Measurement Business Division > (503) 627-5777 Bob, We went down this path several months ago. Here is what I discovered after a week of research: 1. I was not able to find a single vendor that produced a PCMCIA adapter IC that was directly bus compatible with the 68K family. Unfortunately, the 80x86 family dominates the PCMCIA world. Therefore, you will need to create a pseudo ISA bus (or PCI buss) for talking to one of the Cirrus, Omega, or Zilog controller chips. 2. The latest release of VXWorks OS from WindRiver Systems has fairly complete PCMCIA support. This could dramatically reduce your development time. However, VXWorks is a bit pricey. (13.5K dev system + royalties) 3. If you do not use an OS like VXWorks, you will need to write Card and Socket Services drivers for every device you want to support. Good Luck! Darnell Gadberry Phonica Digital Audio Products darnell@primenet.com --------------------------- End of New-News digest ********************** Subject: SUBSCRIBE Submitted-by tony.holben@def.bae.co.uk Fri Jul 28 13:02:31 1995 Submitted-by: tony.holben@def.bae.co.uk (HOLBEN) From tony.holben@def.bae.co.uk Fri Jul 28 13:02:31 1995 From: tony.holben@def.bae.co.uk (HOLBEN) Date: Fri Jul 28 13:03:23 PDT 1995 Subject: SUBSCRIBE Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sat Jul 29 04:00:29 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sat Jul 29 04:00:18 PDT 1995 Subject: Re: Forcing Structure Alignment Subject: VxWORKS SCSI TAPE DRIVER ? Subject: Booting VxWorks and Unix boot server Subject: Network Time Protocol software Subject: [Q] How to find the memory taken up by application Subject: WWW/FTP Proceedings for the ACM SIGPLAN Real-Time Workshop Subject: SNMP and VxWorks Subject: SNMP and VxWorks Subject: Re: Forcing Structure Alignment Subject: vxworks and snmp ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: Fri, 28 Jul 1995 08:39:40 -0700 From: cobarruvias@asd1.jsc.nasa.gov (John R. Cobarruvias) Organization: NASA/JSC Message-ID: References: In article , George Nichols wrote: Thanks for this post! I ended up "padding" my structures as a fix also. I was looking for a switch to the GNU compiler to align structures on a full word. But GNU doesnt support the switch. Again thanks! > Data alignment is a common problem on code dealing with fixed data types > and ported across processors and operation systems. Constant porting > problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove > me to the following solutions: > > Define fixed data types in header file keyed to CPU and OS, (ie): > > typedef char INT8; > typedef short INT16; > typedef long INT32; > > Always manually group/align elements in a data structure to 32bits, (ie): > > struct DEMO { > INT8 this[5]; > INT8 pad1[3]; > INT16 that[3]; > INT16 pad2[1]; > INT32 theother[4]; > }; > > Grouping elements of similar types together minimizes the overall number of > pads require(and wasted space). > > This technique produces trackable data structures when data is > communicated from one CPU/OS to a different CPU/OS. > > > ~ ~ > ~ ~ > ~ ~ ~ | ~ ~ ~ > ~ | ~ > ~ | ~ ____ > | | _O _|_ > | | | ____|_|__ > | | | \ | / > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > George Nichols > Naval Undersea Warfare Center (NUWC) > Submarine Electromagnetics Department (34) > Communication Systems Division (341) > Systems Analysis Branch (3411) > Detachment New London > Phone: 203-440-5503 > Fax: 203-440-4007 > Email: nichols@lx34.nl.nuwc.navy.mil _ _ _ / _ /_ _ /_/ / ` _ /_ _ _ _ . _ _ (_/ /_// // / / \ /_, /_//_//_|/ / /_/|// /_|_\ John R. Cobarruvias cobarruvias@asd1.jsc.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: VxWORKS SCSI TAPE DRIVER ? Date: 27 Jul 1995 20:52:15 GMT From: srejto Organization: MIT Lincoln Laboratory Message-ID: <3v8u9v$ab6@llnews.ll.mit.edu> ANyone have a SCSI tape driver for a 1/2 inch tape drive, or an exabyte tape drive. A supported product would be preffered but anything would help. Thank you srejto@ll.mit.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Booting VxWorks and Unix boot server Date: 28 Jul 1995 11:06:04 GMT From: tronds@nftdf.fof.nft.no (Trond Skaar) Organization: NFT, Norsk Forsvarsteknologi A/S, 3600 Kongsberg Keywords: FTP, boot Message-ID: <3vagas$cnh@nftdf.nft.no> References: <3v7sjh$s2p@nftdf.nft.no> Sender: tronds@nftdf (Trond Skaar) I tried posting this article yesterday, but something seems to have gone wrong, so I am trying again. === I have a MicroForce VMEbus rack containing two Force 2CE-80 CPUs. One CPU is running Unix, and the other CPU is running VxWorks.The Unix CPU is a boot server (tftpboot) for the VxWorks CPU. PROBLEM: When I boot the complete rack (either at power-UP, or by executing the reboot command on the Unix CPU) the VxWorks CPU will not boot. If I hit the reset button on the VxWorks CPU after the Unix CPU is up and running, then the VxWorks CPU boots without any problem. Also if I press CTRL-x after the VxWorks CPU has been booted, the VxWorks CPU will reboot without any problem. If anyone has experienced the same problem and found a solution to it (so that the VxWorks CPU always boots), then I would appreciate a description of the solution! DIAGNOSTICS: I logged the printouts from the VxWorks CPU when the boot fails: ---------------------------------------------------------------- WARNING: Unable to determine keyboard type Device not found: screen Can't open input device. Keyboard not present. Using tty for input and output. FORCE CPU-2CE, No Keyboard ROM 2.3.10 P/N 910-12437-106 CPU2CE_00012 32 MB memory installed, Serial #737805. Ethernet address 0:80:42:8:2:d, Host ID: 558b420d Copyright (c) 1990, Sun Microsystems, Inc. Copyright (c) 1992, Force Computers Inc. Testing 1 megs of memory. Still to go 0 SBus slot 0 le esp dma SBus slot 1 SBus slot 2 SBus slot 3 Boot device: /sbus/le@0,c00000 File and args: Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet receive failed The file just loaded does not appear to be executable. Type b (boot), c (continue), or n (new command mode) > - -- Trond Skaar (email: tronds@fof.nft.no) --------------------------- Newsgroups: comp.os.vxworks Subject: Network Time Protocol software Date: 28 Jul 1995 15:40 PST From: cadfael@reg.triumf.ca (MORRIS, DAVID) Organization: TRIUMF: Tri-University Meson Facility Message-ID: <28JUL199515405760@reg.triumf.ca> Hello, I have recently been trying to get the NTP V3.1 code from the ftp site running on a Motorola MV162 with vxWorks 5.1 This code was for 5.0 on an SBE VCOM100 computer board. Has anyone already done this, or have any experience with this code? It seems to me that it would be an important aspect of networked computers to have the RTC sync'd to the world. I am just a bit swamped with this and need a few pointers. Thanks in advance! David Morris cadfael@decu16.triumf.ca 604-222-7450 --------------------------- Newsgroups: comp.os.vxworks Subject: [Q] How to find the memory taken up by application Date: Fri, 28 Jul 1995 16:40:32 GMT From: banurup@netcom.com (Prashant K. Banuru) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: Sender: banurup@netcom.netcom.com Hi, I am working on the following platform VxWorks (for DY 4 Systems Inc. SVME-180) version 5.2. Kernel: WIND version 2.4. How would I know how much memory my application is occupying? the text segment, the data segment and the stack. thankyou - -babru babru_thatikunta@fmc.com --------------------------- Newsgroups: comp.os.lynx,comp.os.qnx,comp.os.vxworks,comp.os.os9,comp.os.chorus,comp.os.mach Subject: WWW/FTP Proceedings for the ACM SIGPLAN Real-Time Workshop Date: 28 Jul 1995 23:47:08 -0400 From: rich@cs.umd.edu (Richard Gerber) Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742 Message-ID: <3vcavs$2dn@coltrane.cs.umd.edu> The ACM SIGPLAN Workshop on Languages, Compilers and Tools for Real-Time Systems was held on June 21-22, 1995 in La Jolla CA. All papers from the workshop are now available electronically, and are linked to the LCT-RTS '95 home page: http://www.cs.umd.edu/projects/TimeWare/sigplan95 Papers may also be retrieved via ftp, from ftp://ftp.cs.umd.edu/pub/realtime/sigplan95 - ------------------------------------------------ Richard Gerber Department of Computer Science University of Maryland College Park, MD 20742 USA --------------------------- Newsgroups: comp.os.vxworks Subject: SNMP and VxWorks Date: 29 Jul 1995 07:06:04 GMT From: Joel Estes Organization: uma.com Message-ID: <3vcmks$975@news.unicomp.net> I have done this. If you are going to develop a custom mib, you are better off in the long run using a development library for the SNMP agent that has a mib compiler available. The two I considered were Epilogue and Peer Networks. I chose Peer (I have no connection with them, so there is no monetary gain for me in this) because it supported SMUX and was quicker for the client or whom I was working to get to market because of distributed  processor support through SMUX. Epilogue at the time required a little more work up front in developing the distributed processor support mechanism. Be aware that either will require a fair amount of memory, but Peer is probably a bigger memory hog. Also schedule enough time up front in the project to develop the mib. It makes life a lot easier on your s/w team in the long run. Both of these packages are rather expensive, though--$25k-30k. If all you need is a standard mib-2 support, try some of the PD packages from MIT, CMU, and the like. I am available for consulting, on a contract basis, also. If you wish to ask questions, you can either email me or post to the comp.protocols.snmp news group. Regardz(tm), Joel Estes ------------------------------------------------ | My opinions are mine and only mine...if ya | joele@uma.com | don't like em, don't read em. | ------------------------------------------------ ******************** Long live X *************************** Begin slam: If windows is the question, Microslop (aka Microsoft) is not the answer. With Microslop, the bugs go in before the name goes on. In biblical times, there were several plagues, amongst them one of insects; today we have a plague of Microslop Windoz bugs... When Microslop updates its software, it is considered a bug release: e.g. Microslop Windoz Bug Releases 3.0, 3.0a, 3.1, and 3.11 and Woid Bug Releases 6.0, 6.0a, and 6.0c. Hence one may refer to Microslop Bugware. Microslop Windoz and Herpes Type I have a lot in common: They are both caused by viruses and a large portion of the human population has been exposed to both. --------------------------- Newsgroups: comp.os.vxworks Subject: SNMP and VxWorks Date: 29 Jul 1995 08:00:12 GMT From: Joel Estes Organization: uma.com Message-ID: <3vcpqc$9p8@news.unicomp.net> I have done this. If you are going to develop a custom mib, you are better off in the long run using a development library for the SNMP agent that has a mib compiler available. The two I considered were Epilogue and Peer Networks. I chose Peer (I have no connection with them, so there is no monetary gain for me in this) because it supported SMUX and was quicker for the client or whom I was working to get to market because of distributed  processor support through SMUX. Epilogue at the time required a little more work up front in developing the distributed processor support mechanism. Be aware that either will require a fair amount of memory, but Peer is probably a bigger memory hog. Also schedule enough time up front in the project to develop the mib. It makes life a lot easier on your s/w team in the long run. Both of these packages are rather expensive, though--$25k-30k. If all you need is a standard mib-2 support, try some of the PD packages from MIT, CMU, and the like. I am available for consulting, on a contract basis, also. If you wish to ask questions, you can either email me or post to the comp.protocols.snmp news group. Regardz(tm), Joel Estes ------------------------------------------------ | My opinions are mine and only mine...if ya | joele@uma.com | don't like em, don't read em. | ------------------------------------------------ ******************** Long live X *************************** Begin slam: If windows is the question, Microslop (aka Microsoft) is not the answer. With Microslop, the bugs go in before the name goes on. In biblical times, there were several plagues, amongst them one of insects; today we have a plague of Microslop Windoz bugs... When Microslop updates its software, it is considered a bug release: e.g. Microslop Windoz Bug Releases 3.0, 3.0a, 3.1, and 3.11 and Woid Bug Releases 6.0, 6.0a, and 6.0c. Hence one may refer to Microslop Bugware. Microslop Windoz and Herpes Type I have a lot in common: They are both caused by viruses and a large portion of the human population has been exposed to both. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: 28 Jul 1995 20:44:51 GMT From: marc@wrs.com (Marc Shepard) Organization: Wind River Systems, Inc. Message-ID: <3vbi83$ilu@darya.wrs.com> References: In article George Nichols writes: ... > >Always manually group/align elements in a data structure to 32bits, (ie): > >struct DEMO { > INT8 this[5]; > INT8 pad1[3]; > INT16 that[3]; > INT16 pad2[1]; > INT32 theother[4]; >}; You should also consider using the XDR libraries, which were specifically built to handle passing data structures between machines with (possibly) different endianness and different structure alignments. There is a little more overhead with this approach because the data is copied/twiddled an extra time at each end, but if you can afford the overhead its a nice way to go. The XDR libraries are normally used with RPCs, but they can easily be used by themselves. Any book on RPCs will give details on the XDR facilities. - -- ``` (o o) - -----------------oOO--(_)--OOo-------------------- marc@wrs.com (510)814-2142 --------------------------- Newsgroups: comp.os.vxworks Subject: vxworks and snmp Date: 28 Jul 1995 17:39:02 GMT From: Tim Clotworthy Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <3vb7bm$39c@post.gsfc.nasa.gov> Hello, I am looking for any goods sources of information on how to develop snmp MIBs and agents under vxworks. I have the WindRiver "WindNet SNMP" document (version 1.0 Beta), but find it lacking in good info on agent development. Any leads would be greatly appreciated. Thanks. Tim Clotworthy. --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sat Jul 29 04:00:29 1995 From: daemon@csg.lbl.gov Date: Sat Jul 29 04:00:33 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Jul 29 04:00:18 PDT 1995 Subject: Re: Forcing Structure Alignment Subject: VxWORKS SCSI TAPE DRIVER ? Subject: Booting VxWorks and Unix boot server Subject: Network Time Protocol software Subject: [Q] How to find the memory taken up by application Subject: WWW/FTP Proceedings for the ACM SIGPLAN Real-Time Workshop Subject: SNMP and VxWorks Subject: SNMP and VxWorks Subject: Re: Forcing Structure Alignment Subject: vxworks and snmp ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: Fri, 28 Jul 1995 08:39:40 -0700 From: cobarruvias@asd1.jsc.nasa.gov (John R. Cobarruvias) Organization: NASA/JSC Message-ID: References: In article , George Nichols wrote: Thanks for this post! I ended up "padding" my structures as a fix also. I was looking for a switch to the GNU compiler to align structures on a full word. But GNU doesnt support the switch. Again thanks! > Data alignment is a common problem on code dealing with fixed data types > and ported across processors and operation systems. Constant porting > problems between I80x86(MSDOS/Linux) and 680x0/SPARC(VxWorks/UNIX) drove > me to the following solutions: > > Define fixed data types in header file keyed to CPU and OS, (ie): > > typedef char INT8; > typedef short INT16; > typedef long INT32; > > Always manually group/align elements in a data structure to 32bits, (ie): > > struct DEMO { > INT8 this[5]; > INT8 pad1[3]; > INT16 that[3]; > INT16 pad2[1]; > INT32 theother[4]; > }; > > Grouping elements of similar types together minimizes the overall number of > pads require(and wasted space). > > This technique produces trackable data structures when data is > communicated from one CPU/OS to a different CPU/OS. > > > ~ ~ > ~ ~ > ~ ~ ~ | ~ ~ ~ > ~ | ~ > ~ | ~ ____ > | | _O _|_ > | | | ____|_|__ > | | | \ | / > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > George Nichols > Naval Undersea Warfare Center (NUWC) > Submarine Electromagnetics Department (34) > Communication Systems Division (341) > Systems Analysis Branch (3411) > Detachment New London > Phone: 203-440-5503 > Fax: 203-440-4007 > Email: nichols@lx34.nl.nuwc.navy.mil _ _ _ / _ /_ _ /_/ / ` _ /_ _ _ _ . _ _ (_/ /_// // / / \ /_, /_//_//_|/ / /_/|// /_|_\ John R. Cobarruvias cobarruvias@asd1.jsc.nasa.gov --------------------------- Newsgroups: comp.os.vxworks Subject: VxWORKS SCSI TAPE DRIVER ? Date: 27 Jul 1995 20:52:15 GMT From: srejto Organization: MIT Lincoln Laboratory Message-ID: <3v8u9v$ab6@llnews.ll.mit.edu> ANyone have a SCSI tape driver for a 1/2 inch tape drive, or an exabyte tape drive. A supported product would be preffered but anything would help. Thank you srejto@ll.mit.edu --------------------------- Newsgroups: comp.os.vxworks Subject: Booting VxWorks and Unix boot server Date: 28 Jul 1995 11:06:04 GMT From: tronds@nftdf.fof.nft.no (Trond Skaar) Organization: NFT, Norsk Forsvarsteknologi A/S, 3600 Kongsberg Keywords: FTP, boot Message-ID: <3vagas$cnh@nftdf.nft.no> References: <3v7sjh$s2p@nftdf.nft.no> Sender: tronds@nftdf (Trond Skaar) I tried posting this article yesterday, but something seems to have gone wrong, so I am trying again. === I have a MicroForce VMEbus rack containing two Force 2CE-80 CPUs. One CPU is running Unix, and the other CPU is running VxWorks.The Unix CPU is a boot server (tftpboot) for the VxWorks CPU. PROBLEM: When I boot the complete rack (either at power-UP, or by executing the reboot command on the Unix CPU) the VxWorks CPU will not boot. If I hit the reset button on the VxWorks CPU after the Unix CPU is up and running, then the VxWorks CPU boots without any problem. Also if I press CTRL-x after the VxWorks CPU has been booted, the VxWorks CPU will reboot without any problem. If anyone has experienced the same problem and found a solution to it (so that the VxWorks CPU always boots), then I would appreciate a description of the solution! DIAGNOSTICS: I logged the printouts from the VxWorks CPU when the boot fails: ---------------------------------------------------------------- WARNING: Unable to determine keyboard type Device not found: screen Can't open input device. Keyboard not present. Using tty for input and output. FORCE CPU-2CE, No Keyboard ROM 2.3.10 P/N 910-12437-106 CPU2CE_00012 32 MB memory installed, Serial #737805. Ethernet address 0:80:42:8:2:d, Host ID: 558b420d Copyright (c) 1990, Sun Microsystems, Inc. Copyright (c) 1992, Force Computers Inc. Testing 1 megs of memory. Still to go 0 SBus slot 0 le esp dma SBus slot 1 SBus slot 2 SBus slot 3 Boot device: /sbus/le@0,c00000 File and args: Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet Timeout waiting for ARP/RARP packet receive failed The file just loaded does not appear to be executable. Type b (boot), c (continue), or n (new command mode) > - -- Trond Skaar (email: tronds@fof.nft.no) --------------------------- Newsgroups: comp.os.vxworks Subject: Network Time Protocol software Date: 28 Jul 1995 15:40 PST From: cadfael@reg.triumf.ca (MORRIS, DAVID) Organization: TRIUMF: Tri-University Meson Facility Message-ID: <28JUL199515405760@reg.triumf.ca> Hello, I have recently been trying to get the NTP V3.1 code from the ftp site running on a Motorola MV162 with vxWorks 5.1 This code was for 5.0 on an SBE VCOM100 computer board. Has anyone already done this, or have any experience with this code? It seems to me that it would be an important aspect of networked computers to have the RTC sync'd to the world. I am just a bit swamped with this and need a few pointers. Thanks in advance! David Morris cadfael@decu16.triumf.ca 604-222-7450 --------------------------- Newsgroups: comp.os.vxworks Subject: [Q] How to find the memory taken up by application Date: Fri, 28 Jul 1995 16:40:32 GMT From: banurup@netcom.com (Prashant K. Banuru) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: Sender: banurup@netcom.netcom.com Hi, I am working on the following platform VxWorks (for DY 4 Systems Inc. SVME-180) version 5.2. Kernel: WIND version 2.4. How would I know how much memory my application is occupying? the text segment, the data segment and the stack. thankyou - -babru babru_thatikunta@fmc.com --------------------------- Newsgroups: comp.os.lynx,comp.os.qnx,comp.os.vxworks,comp.os.os9,comp.os.chorus,comp.os.mach Subject: WWW/FTP Proceedings for the ACM SIGPLAN Real-Time Workshop Date: 28 Jul 1995 23:47:08 -0400 From: rich@cs.umd.edu (Richard Gerber) Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742 Message-ID: <3vcavs$2dn@coltrane.cs.umd.edu> The ACM SIGPLAN Workshop on Languages, Compilers and Tools for Real-Time Systems was held on June 21-22, 1995 in La Jolla CA. All papers from the workshop are now available electronically, and are linked to the LCT-RTS '95 home page: http://www.cs.umd.edu/projects/TimeWare/sigplan95 Papers may also be retrieved via ftp, from ftp://ftp.cs.umd.edu/pub/realtime/sigplan95 - ------------------------------------------------ Richard Gerber Department of Computer Science University of Maryland College Park, MD 20742 USA --------------------------- Newsgroups: comp.os.vxworks Subject: SNMP and VxWorks Date: 29 Jul 1995 07:06:04 GMT From: Joel Estes Organization: uma.com Message-ID: <3vcmks$975@news.unicomp.net> I have done this. If you are going to develop a custom mib, you are better off in the long run using a development library for the SNMP agent that has a mib compiler available. The two I considered were Epilogue and Peer Networks. I chose Peer (I have no connection with them, so there is no monetary gain for me in this) because it supported SMUX and was quicker for the client or whom I was working to get to market because of distributed  processor support through SMUX. Epilogue at the time required a little more work up front in developing the distributed processor support mechanism. Be aware that either will require a fair amount of memory, but Peer is probably a bigger memory hog. Also schedule enough time up front in the project to develop the mib. It makes life a lot easier on your s/w team in the long run. Both of these packages are rather expensive, though--$25k-30k. If all you need is a standard mib-2 support, try some of the PD packages from MIT, CMU, and the like. I am available for consulting, on a contract basis, also. If you wish to ask questions, you can either email me or post to the comp.protocols.snmp news group. Regardz(tm), Joel Estes ------------------------------------------------ | My opinions are mine and only mine...if ya | joele@uma.com | don't like em, don't read em. | ------------------------------------------------ ******************** Long live X *************************** Begin slam: If windows is the question, Microslop (aka Microsoft) is not the answer. With Microslop, the bugs go in before the name goes on. In biblical times, there were several plagues, amongst them one of insects; today we have a plague of Microslop Windoz bugs... When Microslop updates its software, it is considered a bug release: e.g. Microslop Windoz Bug Releases 3.0, 3.0a, 3.1, and 3.11 and Woid Bug Releases 6.0, 6.0a, and 6.0c. Hence one may refer to Microslop Bugware. Microslop Windoz and Herpes Type I have a lot in common: They are both caused by viruses and a large portion of the human population has been exposed to both. --------------------------- Newsgroups: comp.os.vxworks Subject: SNMP and VxWorks Date: 29 Jul 1995 08:00:12 GMT From: Joel Estes Organization: uma.com Message-ID: <3vcpqc$9p8@news.unicomp.net> I have done this. If you are going to develop a custom mib, you are better off in the long run using a development library for the SNMP agent that has a mib compiler available. The two I considered were Epilogue and Peer Networks. I chose Peer (I have no connection with them, so there is no monetary gain for me in this) because it supported SMUX and was quicker for the client or whom I was working to get to market because of distributed  processor support through SMUX. Epilogue at the time required a little more work up front in developing the distributed processor support mechanism. Be aware that either will require a fair amount of memory, but Peer is probably a bigger memory hog. Also schedule enough time up front in the project to develop the mib. It makes life a lot easier on your s/w team in the long run. Both of these packages are rather expensive, though--$25k-30k. If all you need is a standard mib-2 support, try some of the PD packages from MIT, CMU, and the like. I am available for consulting, on a contract basis, also. If you wish to ask questions, you can either email me or post to the comp.protocols.snmp news group. Regardz(tm), Joel Estes ------------------------------------------------ | My opinions are mine and only mine...if ya | joele@uma.com | don't like em, don't read em. | ------------------------------------------------ ******************** Long live X *************************** Begin slam: If windows is the question, Microslop (aka Microsoft) is not the answer. With Microslop, the bugs go in before the name goes on. In biblical times, there were several plagues, amongst them one of insects; today we have a plague of Microslop Windoz bugs... When Microslop updates its software, it is considered a bug release: e.g. Microslop Windoz Bug Releases 3.0, 3.0a, 3.1, and 3.11 and Woid Bug Releases 6.0, 6.0a, and 6.0c. Hence one may refer to Microslop Bugware. Microslop Windoz and Herpes Type I have a lot in common: They are both caused by viruses and a large portion of the human population has been exposed to both. --------------------------- Newsgroups: comp.os.vxworks Subject: Re: Forcing Structure Alignment Date: 28 Jul 1995 20:44:51 GMT From: marc@wrs.com (Marc Shepard) Organization: Wind River Systems, Inc. Message-ID: <3vbi83$ilu@darya.wrs.com> References: In article George Nichols writes: ... > >Always manually group/align elements in a data structure to 32bits, (ie): > >struct DEMO { > INT8 this[5]; > INT8 pad1[3]; > INT16 that[3]; > INT16 pad2[1]; > INT32 theother[4]; >}; You should also consider using the XDR libraries, which were specifically built to handle passing data structures between machines with (possibly) different endianness and different structure alignments. There is a little more overhead with this approach because the data is copied/twiddled an extra time at each end, but if you can afford the overhead its a nice way to go. The XDR libraries are normally used with RPCs, but they can easily be used by themselves. Any book on RPCs will give details on the XDR facilities. - -- ``` (o o) - -----------------oOO--(_)--OOo-------------------- marc@wrs.com (510)814-2142 --------------------------- Newsgroups: comp.os.vxworks Subject: vxworks and snmp Date: 28 Jul 1995 17:39:02 GMT From: Tim Clotworthy Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <3vb7bm$39c@post.gsfc.nasa.gov> Hello, I am looking for any goods sources of information on how to develop snmp MIBs and agents under vxworks. I have the WindRiver "WindNet SNMP" document (version 1.0 Beta), but find it lacking in good info on agent development. Any leads would be greatly appreciated. Thanks. Tim Clotworthy. --------------------------- End of New-News digest ********************** Subject: comp.os.vxworks newsdigest Submitted-by daemon@csg.lbl.gov Sun Jul 30 04:00:17 1995 Submitted-by: daemon@csg.lbl.gov Comp.Os.Vxworks Daily Digest Sun Jul 30 04:00:13 PDT 1995 Subject: Re: Driver for the SCC2692 DUART Subject: Re: VxWorks on MIPS controllers ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Driver for the SCC2692 DUART Date: Sat, 29 Jul 1995 05:35:00 GMT From: wellsk@netcom.com (Kevin Wells) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: <3vap97$7bc_001@netcom.com> References: <3uinj9$fa1@radmail.rad.co.il> Sender: wellsk@netcom6.netcom.com In article <3uinj9$fa1@radmail.rad.co.il>, Adi Regev wrote: >I want to write a driver for the PHILIPS SCC2692 DUART. > >I will enjoy some writen code, other info will be accepted in joy! > > >------------------------------------------------------------------- > >Adi Regev >RADVision Ltd. adi@radvision.rad.co.il > Hi. I've just recently written a driver for the 26c92 DUART (in 2692 emulation mode). It supports most of the DUARTs capabilities including the watchdog for the RxFIFO and both channels (through int. polling and modified ioCtl() calls). Unfortunately, I haven't done any testing of the driver as of yet, but am fairly confident that it will work (or may need very small adjustments). If you are willing, maybe we can 'share' the testing of the driver. Since it will be a while until we have target hardware, your experience with the present driver, with any mods or hints, would be helpful, and can save us some time in the future. A 'win-win' situation!?! Please write me to let me know if you are interested. Kevin Wells, s/w engineering, Lear Astronics Corporation --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks on MIPS controllers Date: Sat, 29 Jul 1995 05:49:10 GMT From: wellsk@netcom.com (Kevin Wells) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: <3vaq3q$7bc_002@netcom.com> References: <9507201358.AB09867@rhin> Sender: wellsk@netcom6.netcom.com In article <9507201358.AB09867@rhin>, rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) wrote: >Greetings, > >is anybody willing to share his experiance with VxWorks on a MIPS >controller (e.g. R3051, R3081). Drivers for on-chip devices are of special >interest. > >Thanks for any info. > >Joerg > >************************************************************************* >* Joerg Bertholdt Wind River Systems GmbH * >* Field Application Engineer Freisinger Strasse 34 * >* Email: joergb@wrsec.fr D-85737 Ismaning * >* Tel: +49-89-962445-42 Fax: +49-89-962445-55 * >* http://www.wrs.com * >************************************************************************* > > Hi Joerg. I'll share what experience I can. At present, we are porting VxWorks to a R3081 processor. I have a good familarity with the R3000 BSP, and device driver integration for some common (and uncommon) devices, mostly off-chip. Anything specific I can share? Kevin Wells, s/w engineer, Lear Astronics --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sun Jul 30 04:00:17 1995 From: daemon@csg.lbl.gov Date: Sun Jul 30 04:00:20 PDT 1995 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Jul 30 04:00:13 PDT 1995 Subject: Re: Driver for the SCC2692 DUART Subject: Re: VxWorks on MIPS controllers ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Driver for the SCC2692 DUART Date: Sat, 29 Jul 1995 05:35:00 GMT From: wellsk@netcom.com (Kevin Wells) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: <3vap97$7bc_001@netcom.com> References: <3uinj9$fa1@radmail.rad.co.il> Sender: wellsk@netcom6.netcom.com In article <3uinj9$fa1@radmail.rad.co.il>, Adi Regev wrote: >I want to write a driver for the PHILIPS SCC2692 DUART. > >I will enjoy some writen code, other info will be accepted in joy! > > >------------------------------------------------------------------- > >Adi Regev >RADVision Ltd. adi@radvision.rad.co.il > Hi. I've just recently written a driver for the 26c92 DUART (in 2692 emulation mode). It supports most of the DUARTs capabilities including the watchdog for the RxFIFO and both channels (through int. polling and modified ioCtl() calls). Unfortunately, I haven't done any testing of the driver as of yet, but am fairly confident that it will work (or may need very small adjustments). If you are willing, maybe we can 'share' the testing of the driver. Since it will be a while until we have target hardware, your experience with the present driver, with any mods or hints, would be helpful, and can save us some time in the future. A 'win-win' situation!?! Please write me to let me know if you are interested. Kevin Wells, s/w engineering, Lear Astronics Corporation --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks on MIPS controllers Date: Sat, 29 Jul 1995 05:49:10 GMT From: wellsk@netcom.com (Kevin Wells) Organization: NETCOM On-line Communication Services (408 261-4700 guest) Message-ID: <3vaq3q$7bc_002@netcom.com> References: <9507201358.AB09867@rhin> Sender: wellsk@netcom6.netcom.com In article <9507201358.AB09867@rhin>, rhin!wrsec.fr!joergb@lbl.gov ( =?iso-8859-1?Q?J=F6rg?= Bertholdt) wrote: >Greetings, > >is anybody willing to share his experiance with VxWorks on a MIPS >controller (e.g. R3051, R3081). Drivers for on-chip devices are of special >interest. > >Thanks for any info. > >Joerg > >************************************************************************* >* Joerg Bertholdt Wind River Systems GmbH * >* Field Application Engineer Freisinger Strasse 34 * >* Email: joergb@wrsec.fr D-85737 Ismaning * >* Tel: +49-89-962445-42 Fax: +49-89-962445-55 * >* http://www.wrs.com * >************************************************************************* > > Hi Joerg. I'll share what experience I can. At present, we are porting VxWorks to a R3081 processor. I have a good familarity with the R3000 BSP, and device driver integration for some common (and uncommon) devices, mostly off-chip. Anything specific I can share? Kevin Wells, s/w engineer, Lear Astronics --------------------------- End of New-News digest ********************** Subject: X.25 driver Submitted-by has@info.kobelco.co.jp Mon Jul 31 04:23:11 1995 Submitted-by: has@info.kobelco.co.jp (=?ISO-2022-JP?B?GyRCRDlDK0BuGyhC?= =?ISO-2022-JP?B?GyRCPCJPOhsoQg==?=) Hi, Does anyone know the X.25 driver for VxWorks exist? I'd like to know if it is possible or not to get the driver. Our some customer need it. Best Regards, ---------------------------------------- 長谷川　滋郎＠株式会社　神戸製鋼所 VxWorksグループ　情報システム部 情報エレクトロニクス本部　電子・情報事業本部 〒651 神戸市中央区御幸通6-1-12 三宮ビル東館9階 TEL: 078-261-6461 FAX: 078-261-6499 E-mail: has@info.kobelco.co.jp ---------------------------------------- From has@info.kobelco.co.jp Mon Jul 31 04:23:11 1995 From: has@info.kobelco.co.jp (=?ISO-2022-JP?B?GyRCRDlDK0BuGyhC?= =?ISO-2022-JP?B?GyRCPCJPOhsoQg==?=) Date: Mon Jul 31 04:23:14 PDT 1995 Subject: X.25 driver Hi, Does anyone know the X.25 driver for VxWorks exist? I'd like to know if it is possible or not to get the driver. Our some customer need it. Best Regards, ---------------------------------------- 長谷川　滋郎＠株式会社　神戸製鋼所 VxWorksグループ　情報システム部 情報エレクトロニクス本部　電子・情報事業本部 〒651 神戸市中央区御幸通6-1-12 三宮ビル東館9階 TEL: 078-261-6461 FAX: 078-261-6499 E-mail: has@info.kobelco.co.jp ---------------------------------------- Subject: VME Address Mapping Submitted-by kozubal@jamal.lanl.gov Mon Jul 31 08:02:56 1995 Submitted-by: kozubal@jamal.lanl.gov (Andy Kozubal) I have an I/O board that supports only 24-bit addressing on DMA transfers. This works fine with my 4 MByte MV167, but I can't get it to work with a 16 MByte MV167. The following call-- sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) returns -1 with the 16 MByte processor. I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. Any ideas what's wrong here? ----------------------------------------------------------------------- Andy Kozubal | Mail Stop P942 | Accelerator Operations and Technology Phone: (505) 667-6508 | Los Alamos National Laboratory E-mail: AKozubal@lanl.gov | Los Alamos, New Mexico 87545 ----------------------------------------------------------------------- From kozubal@jamal.lanl.gov Mon Jul 31 08:02:56 1995 From: kozubal@jamal.lanl.gov (Andy Kozubal) Date: Mon Jul 31 08:02:59 PDT 1995 Subject: VME Address Mapping I have an I/O board that supports only 24-bit addressing on DMA transfers. This works fine with my 4 MByte MV167, but I can't get it to work with a 16 MByte MV167. The following call-- sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) returns -1 with the 16 MByte processor. I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. Any ideas what's wrong here? ----------------------------------------------------------------------- Andy Kozubal | Mail Stop P942 | Accelerator Operations and Technology Phone: (505) 667-6508 | Los Alamos National Laboratory E-mail: AKozubal@lanl.gov | Los Alamos, New Mexico 87545 ----------------------------------------------------------------------- Subject: Re: VME Address Mapping Submitted-by pardo@rti.com Mon Jul 31 09:58:09 1995 Submitted-by: Gerardo Pardo-Castellote >> I have an I/O board that supports only 24-bit addressing on DMA >> transfers. This works fine with my 4 MByte MV167, but I can't get >> it to work with a 16 MByte MV167. The following call-- >> >> sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) >> >> returns -1 with the 16 MByte processor. >> I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. >> >> Any ideas what's wrong here? Look at the source for sysLocalToBusAdrs() in config/sysLib.c: For A24 mode, the function only succeeds if you have 4 MB of local memory ( *MEMC_MCR contains the size of the local memory in units of 4 MB). You will also need to modify sysProcNumSet() since the default kernels only map the local memeory onto the bus on A24 mode if the processor has 4 MB of memory. -Gerardo =========================================================================== = = = = Gerardo Pardo-Castellote = email: pardo@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 155A Moffett Park Drive, Suite 111 = Fax: (408) 720-8419 = = = = =========================================================================== From pardo@rti.com Mon Jul 31 09:58:09 1995 From: Gerardo Pardo-Castellote Date: Mon Jul 31 09:58:12 PDT 1995 Subject: Re: VME Address Mapping >> I have an I/O board that supports only 24-bit addressing on DMA >> transfers. This works fine with my 4 MByte MV167, but I can't get >> it to work with a 16 MByte MV167. The following call-- >> >> sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) >> >> returns -1 with the 16 MByte processor. >> I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. >> >> Any ideas what's wrong here? Look at the source for sysLocalToBusAdrs() in config/sysLib.c: For A24 mode, the function only succeeds if you have 4 MB of local memory ( *MEMC_MCR contains the size of the local memory in units of 4 MB). You will also need to modify sysProcNumSet() since the default kernels only map the local memeory onto the bus on A24 mode if the processor has 4 MB of memory. -Gerardo =========================================================================== = = = = Gerardo Pardo-Castellote = email: pardo@rti.com = = Real-Time Innovations, Inc. = Phone: (408) 720-8312 = = 155A Moffett Park Drive, Suite 111 = Fax: (408) 720-8419 = = = = =========================================================================== Subject: Re: VME Address Mapping Submitted-by phil@naic.edu Mon Jul 31 10:53:43 1995 Submitted-by: phil@naic.edu (Phil Perillat) > From root@csg.lbl.gov Mon Jul 31 13:26:47 1995 > Date: Mon, 31 Jul 95 09:58:15 PDT > To: vxworks_users@csg.lbl.gov > From: vxwexplo@lbl.gov (the vxWorks Users Group Exploder) > Subject: Re: VME Address Mapping > Content-Length: 1650 > X-Lines: 46 > > Submitted-by pardo@rti.com Mon Jul 31 09:58:09 1995 > Submitted-by: Gerardo Pardo-Castellote > > > >> I have an I/O board that supports only 24-bit addressing on DMA > >> transfers. This works fine with my 4 MByte MV167, but I can't get > >> it to work with a 16 MByte MV167. The following call-- > >> > >> sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) > >> > >> returns -1 with the 16 MByte processor. > > >> I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. > >> > >> Any ideas what's wrong here? > > Look at the source for sysLocalToBusAdrs() in config/sysLib.c: For A24 mode, > the function only succeeds if you have 4 MB of local memory ( *MEMC_MCR > contains the size of the local memory in units of 4 MB). > > > You will also need to modify sysProcNumSet() since the default kernels > only map the local memeory onto the bus on A24 mode if the processor has > 4 MB of memory. > > > > -Gerardo You need to be a little careful how you map your 16mb board onto A24 space. 24 bits gives you 16mb. If you map the entire memory of the cpu into a24, then no other board in the crate can access a24 space. You might want to create a memory partition on your cpu that will hold the dma buffers. You could then map this portion of your local memory into A24 space. phil perillat arecibo observatory phil@naic.edu From phil@naic.edu Mon Jul 31 10:53:43 1995 From: phil@naic.edu (Phil Perillat) Date: Mon Jul 31 10:53:45 PDT 1995 Subject: Re: VME Address Mapping > From root@csg.lbl.gov Mon Jul 31 13:26:47 1995 > Date: Mon, 31 Jul 95 09:58:15 PDT > To: vxworks_users@csg.lbl.gov > From: vxwexplo@lbl.gov (the vxWorks Users Group Exploder) > Subject: Re: VME Address Mapping > Content-Length: 1650 > X-Lines: 46 > > Submitted-by pardo@rti.com Mon Jul 31 09:58:09 1995 > Submitted-by: Gerardo Pardo-Castellote > > > >> I have an I/O board that supports only 24-bit addressing on DMA > >> transfers. This works fine with my 4 MByte MV167, but I can't get > >> it to work with a 16 MByte MV167. The following call-- > >> > >> sysLocalToBusAdrs (VME_AM_STD_SUP_DATA, pBfr, &busAdrs) > >> > >> returns -1 with the 16 MByte processor. > > >> I am using the mv167 version 1.0 BSP and both VxWorks 5.1 and 5.2. > >> > >> Any ideas what's wrong here? > > Look at the source for sysLocalToBusAdrs() in config/sysLib.c: For A24 mode, > the function only succeeds if you have 4 MB of local memory ( *MEMC_MCR > contains the size of the local memory in units of 4 MB). > > > You will also need to modify sysProcNumSet() since the default kernels > only map the local memeory onto the bus on A24 mode if the processor has > 4 MB of memory. > > > > -Gerardo You need to be a little careful how you map your 16mb board onto A24 space. 24 bits gives you 16mb. If you map the entire memory of the cpu into a24, then no other board in the crate can access a24 space. You might want to create a memory partition on your cpu that will hold the dma buffers. You could then map this portion of your local memory into A24 space. phil perillat arecibo observatory phil@naic.edu Subject: Monthly VxWorks archive posting Submitted-by thor@thor.atd.ucar.edu Mon Jul 31 23:00:15 1995 Submitted-by: thor@thor.atd.ucar.edu (Rich Neitzel) This is the monthly posting showing the current holdings in the VxWorks Software Archive. To get more detailed infomation send email to: vxworks_archive@ncar.ucar.edu The message body must read: send index send index from vx ------------------------------------------------ VxWorks sources: total 14746 lrwxrwxrwx 1 root rsf 5 Feb 10 19:32 README -> index -rw-r--r-- 1 thor rsf 22132 Sep 20 1994 ansi.p1 -rw-r--r-- 1 thor rsf 22717 Sep 20 1994 ansi.p2 -rw-r--r-- 1 thor rsf 24174 Sep 20 1994 ansi.p3 -rw-r--r-- 1 thor rsf 8108 Sep 20 1994 ansi.patch1 -rw-r--r-- 1 thor rsf 37126 Sep 20 1994 ansilib01 -rw-r--r-- 1 thor rsf 18913 Sep 20 1994 ansilib02 -rw-r--r-- 1 thor rsf 2671 Sep 20 1994 benchmarks -rw-r--r-- 1 thor rsf 7168 Sep 20 1994 bitcnt -rw-r--r-- 1 thor rsf 11437 Sep 20 1994 c++builtin.shar -rw-r--r-- 1 thor rsf 22330 Sep 20 1994 c++headers.p1 -rw-r--r-- 1 thor rsf 22775 Sep 20 1994 c++headers.p2 -rw-r--r-- 1 thor rsf 29052 Sep 20 1994 camaclib1 -rw-r--r-- 1 thor rsf 25095 Sep 20 1994 camaclib2 -rw-r--r-- 1 thor rsf 31005 Sep 20 1994 camaclib3 -rw-r--r-- 1 thor rsf 37770 Sep 20 1994 cbench.shar -rw-r--r-- 1 thor rsf 7371 Sep 20 1994 cntsem_class.shar -rw-r--r-- 1 thor rsf 5853 Sep 20 1994 crc.shar -rw-r--r-- 1 thor rsf 8917 Sep 20 1994 deadman.shar -rw-r--r-- 1 thor rsf 41669 Sep 20 1994 dhrystones01 -rw-r--r-- 1 thor rsf 19170 Sep 20 1994 dirlib01 -rw-r--r-- 1 thor rsf 25681 Sep 20 1994 dt1451 -rw-r--r-- 1 thor rsf 4905 Sep 20 1994 dup.shar -rw-r--r-- 1 thor rsf 5944 Sep 20 1994 fcompress.shar -rw-r--r-- 1 thor rsf 11561 Sep 20 1994 flags_class.shar -rw-r--r-- 1 thor rsf 88109 Sep 20 1994 flash.zip -rw-r--r-- 1 thor rsf 44762 Sep 20 1994 force.p1 -rw-r--r-- 1 thor rsf 40154 Sep 20 1994 force.p2 -rw-r--r-- 1 thor rsf 80491 Sep 20 1994 force.shar -rw-r--r-- 1 thor rsf 2453 Sep 20 1994 gcc+68040 -rw-r--r-- 1 thor rsf 6106 Sep 20 1994 getdate -rw-r--r-- 1 thor rsf 9774 Sep 20 1994 hkv30extintutil.shar -rw-r--r-- 1 thor rsf 16177 Jul 18 10:52 index -rw-r--r-- 1 thor rsf 2694 Sep 20 1994 ivecalloc.shar -rw-r--r-- 1 thor rsf 35245 Sep 20 1994 joblib2.p1 -rw-r--r-- 1 thor rsf 18110 Sep 20 1994 joblib2.p2 -rw-r--r-- 1 thor rsf 9079 Sep 20 1994 lclflag.shar -rw-r--r-- 1 thor rsf 174829 Sep 20 1994 lexyacc.tar.gz drwxr-xr-x 2 thor rsf 512 Sep 20 1994 libX11 -rw-r--r-- 1 thor rsf 14059 Nov 4 1994 libg++-2.6.1.patch -rw-r--r-- 1 thor rsf 6612 Jul 18 10:28 libg++-2.7.0.patch -rw-r--r-- 1 thor rsf 1302 Nov 23 1994 libgcc2-2.6.2.patch -rw-r--r-- 1 thor rsf 1862 Dec 29 1994 libgcc2-2.6.3.patch -rw-r--r-- 1 thor rsf 1222 Nov 4 1994 libio-2.6.1.patch -rw-r--r-- 1 thor rsf 277 Nov 4 1994 libobjc.patch drwxr-xr-x 2 thor rsf 512 Sep 20 1994 libx11 -rw-rw-r-- 1 ftp rsf 3515 Nov 2 1994 loadmeter.shar -rw-r--r-- 1 thor rsf 10399 Sep 20 1994 math.shar -rw-r--r-- 1 thor rsf 11950 Sep 20 1994 math2 -rw-r--r-- 1 thor rsf 26655 Sep 20 1994 monitor.shar -rw-r--r-- 1 thor rsf 18733 Sep 20 1994 msgque_class.shar -rw-r--r-- 1 thor rsf 829713 Sep 20 1994 ntpv3.1.tar.gz -rw-r--r-- 1 thor rsf 1010176 Sep 20 1994 ntpv3.tar.gz -rw-r--r-- 1 thor rsf 1082 Sep 20 1994 objc.patch -rw-r--r-- 1 thor rsf 19422 Sep 20 1994 ping01 -rw-r--r-- 1 thor rsf 20494 Sep 20 1994 pipe.shar -rw-r--r-- 1 thor rsf 15418 Sep 20 1994 poolLib -rw-r--r-- 1 thor rsf 15418 Sep 20 1994 poollib -rw-r--r-- 1 thor rsf 13204 Sep 20 1994 ring.shar -rw-r--r-- 1 thor rsf 6614 Sep 20 1994 semCnt -rw-r--r-- 1 thor rsf 6614 Sep 20 1994 semcnt -rw-r--r-- 1 thor rsf 48659 Dec 2 1994 snmp2.01 -rw-r--r-- 1 thor rsf 48658 Dec 2 1994 snmp2.02 -rw-r--r-- 1 thor rsf 48666 Dec 2 1994 snmp2.03 -rw-r--r-- 1 thor rsf 48723 Dec 2 1994 snmp2.04 -rw-r--r-- 1 thor rsf 48869 Dec 2 1994 snmp2.05 -rw-r--r-- 1 thor rsf 48632 Dec 2 1994 snmp2.06 -rw-r--r-- 1 thor rsf 48658 Dec 2 1994 snmp2.07 -rw-r--r-- 1 thor rsf 48652 Dec 2 1994 snmp2.08 -rw-r--r-- 1 thor rsf 48714 Dec 2 1994 snmp2.09 -rw-r--r-- 1 thor rsf 48615 Dec 2 1994 snmp2.10 -rw-r--r-- 1 thor rsf 49185 Dec 2 1994 snmp2.11 -rw-r--r-- 1 thor rsf 49214 Dec 2 1994 snmp2.12 -rw-r--r-- 1 thor rsf 49203 Dec 2 1994 snmp2.13 -rw-r--r-- 1 thor rsf 49239 Dec 2 1994 snmp2.14 -rw-r--r-- 1 thor rsf 49176 Dec 2 1994 snmp2.15 -rw-r--r-- 1 thor rsf 49216 Dec 2 1994 snmp2.16 -rw-r--r-- 1 thor rsf 49182 Dec 2 1994 snmp2.17 -rw-r--r-- 1 thor rsf 49188 Dec 2 1994 snmp2.18 -rw-r--r-- 1 thor rsf 49189 Dec 2 1994 snmp2.19 -rw-r--r-- 1 thor rsf 49200 Dec 2 1994 snmp2.20 -rw-r--r-- 1 thor rsf 48646 Dec 2 1994 snmp2.21 -rw-r--r-- 1 thor rsf 48671 Dec 2 1994 snmp2.22 -rw-r--r-- 1 thor rsf 49229 Dec 2 1994 snmp2.23 -rw-r--r-- 1 thor rsf 49235 Dec 2 1994 snmp2.24 -rw-r--r-- 1 thor rsf 49268 Dec 2 1994 snmp2.25 -rw-r--r-- 1 thor rsf 48682 Dec 2 1994 snmp2.26 -rw-r--r-- 1 thor rsf 49223 Dec 2 1994 snmp2.27 -rw-r--r-- 1 thor rsf 48676 Dec 2 1994 snmp2.28 -rw-r--r-- 1 thor rsf 48818 Dec 2 1994 snmp2.29 -rw-r--r-- 1 thor rsf 49308 Dec 2 1994 snmp2.30 -rw-r--r-- 1 thor rsf 49294 Dec 2 1994 snmp2.31 -rw-r--r-- 1 thor rsf 48634 Dec 2 1994 snmp2.32 -rw-r--r-- 1 thor rsf 48708 Dec 2 1994 snmp2.33 -rw-r--r-- 1 thor rsf 49055 Dec 2 1994 snmp2.34 -rw-r--r-- 1 thor rsf 48724 Dec 2 1994 snmp2.35 -rw-r--r-- 1 thor rsf 48736 Dec 2 1994 snmp2.36 -rw-r--r-- 1 thor rsf 48667 Dec 2 1994 snmp2.37 -rw-r--r-- 1 thor rsf 48626 Dec 2 1994 snmp2.38 -rw-r--r-- 1 thor rsf 48653 Dec 2 1994 snmp2.39 -rw-r--r-- 1 thor rsf 14122 Dec 2 1994 snmp2.40 -rw-r--r-- 1 thor rsf 2308 Sep 20 1994 ss1.bnch -rw-r--r-- 1 thor rsf 41196 Sep 20 1994 stevie01 -rw-r--r-- 1 thor rsf 35279 Sep 20 1994 stevie02 -rw-r--r-- 1 thor rsf 35278 Sep 20 1994 stevie03 -rw-r--r-- 1 thor rsf 35012 Sep 20 1994 stevie04 -rw-r--r-- 1 thor rsf 34502 Sep 20 1994 stevie05 -rw-r--r-- 1 thor rsf 37476 Sep 20 1994 stevie06 -rw-r--r-- 1 thor rsf 30073 Sep 20 1994 stevie07 -rw-r--r-- 1 thor rsf 31562 Sep 20 1994 stevie08 -rw-r--r-- 1 thor rsf 37360 Sep 20 1994 stevie09 -rw-r--r-- 1 thor rsf 20662 Sep 20 1994 stevie10 -rw-r--r-- 1 thor rsf 25717 Sep 20 1994 stevie11 -rw-r--r-- 1 thor rsf 28075 Sep 20 1994 stevie12 -rw-r--r-- 1 thor rsf 31852 Sep 20 1994 stevie13 -rw-r--r-- 1 thor rsf 16225 Sep 20 1994 string.shar -rw-r--r-- 1 thor rsf 8424 Sep 20 1994 syslog.shar -rw-r--r-- 1 thor rsf 15096 Sep 20 1994 task_class.shar -rw-r--r-- 1 thor rsf 16171 Sep 20 1994 taskmon.shar -rw-rw-r-- 1 thor rsf 417608 Jan 31 1995 tcl