From tei@adv.sbc.sony.co.jp Mon Sep 1 11:05:26 1997 From: Tom Isaacson Date: Mon Sep 1 11:05:29 PDT 1997 Subject: Zilog z85230 Driver --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" Does anyone have a working version of a Zilog z85230 ESCC serial device driver? I have an old version (1993) of one from Wind River, but it was given to us untested, and I am having a lot of trouble getting it working. Any help would be greatly appreciated. Tom Isaacson --MimeMultipartBoundary-- From daemon@csg.lbl.gov Tue Sep 2 04:02:37 1997 From: daemon@csg.lbl.gov Date: Tue Sep 2 04:02:40 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Sep 2 04:02:34 PDT 1997 Subject: Re: struct variables memory alignment ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: struct variables memory alignment Date: 1 Sep 97 18:35:24 GMT From: glennt@ibm.net Organization: die.spam.die Message-ID: <340b0aec.0@news1.ibm.net> References: <5u64fp$3pu$1@news02.btx.dtag.de> Reply-To: glenntT@ibm.net In <5u64fp$3pu$1@news02.btx.dtag.de>, schuehle@t-online.de (Martin Sauter) writes: >Hello, > >I have a problem with the alignment of struct variables with the GNU x86 >compiler for VxWorks. Take the following case > >typedef struct mystruct >{ > short test; > float ftest; >} mystruct; > >sizeof(mystruct); returns 8 instead of 6 as I guess there are two bytes >added after the short variable to align the float variable in some way. >No big deal until you write that struct to a file and expect another >program to read the stuff. My question therefore is how it is possible >to disable that memory alignment for this struct or in general? > >Thanks in advance for your help >Martin Martin - You may be SOL. I suspect that the best you can do is order the items in your struct from largest to smallest and hope that the compiler will order them the same way. The struct itself will probably be the same size, but the padding might be at the end rather than in the middle. To see exactly what the compiler DID do, compile the source file that defines the struct to assembly language instead of object and see what the compiler really did give you. See your gnu manual for the switch, I think it's "-s". Good Luck! - Glenn ************************************************************ * * * Two wrongs don't make a right, BUT * * Three lefts do! * * * * (remove upper case characters to de-spamproof my address) * ************************************************************ --------------------------- End of New-News digest ********************** From currie@ghg.net Tue Sep 2 06:41:43 1997 From: currie@ghg.net (Alan Currie) Date: Tue Sep 2 06:41:46 PDT 1997 Subject: Re: Memory alignment --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit the vxWorks Users Group Exploder wrote: > > Submitted-by daemon@csg.lbl.gov Tue Sep 2 04:02:37 1997 > Submitted-by: daemon@csg.lbl.gov > > Comp.Os.Vxworks Daily Digest Tue Sep 2 04:02:34 PDT 1997 > > Subject: Re: struct variables memory alignment > > ------------------------------------------------------- > > Newsgroups: comp.os.vxworks > Subject: Re: struct variables memory alignment > Date: 1 Sep 97 18:35:24 GMT > From: glennt@ibm.net > Organization: die.spam.die > Message-ID: <340b0aec.0@news1.ibm.net> > References: <5u64fp$3pu$1@news02.btx.dtag.de> > Reply-To: glenntT@ibm.net > > In <5u64fp$3pu$1@news02.btx.dtag.de>, schuehle@t-online.de (Martin Sauter) writes: > >Hello, > > > >I have a problem with the alignment of struct variables with the GNU x86 > >compiler for VxWorks. Take the following case > > > >typedef struct mystruct > >{ > > short test; > > float ftest; > >} mystruct; > > > >sizeof(mystruct); returns 8 instead of 6 as I guess there are two bytes > >added after the short variable to align the float variable in some way. > >No big deal until you write that struct to a file and expect another > >program to read the stuff. My question therefore is how it is possible > >to disable that memory alignment for this struct or in general? > > > >Thanks in advance for your help We padded our structures out to 32 bits because we are passing data through sockets to a labview application on a Win '95 PC. Just group your short ints together, bits, etc. and add the appropriate pad to make it a full word boundary. This way the structure definitions on both machines is identical, you don't have to figure out what the compiler is doing. In your example you would have: > >typedef struct mystruct > >{ > > short test; short pad1; > > float ftest; > >} mystruct; --MimeMultipartBoundary-- From R.Kirnum@dialogic.com Tue Sep 2 10:54:04 1997 From: "Kirnum, Robert" Date: Tue Sep 2 10:54:07 PDT 1997 Subject: Re: struct variables memory alignment --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit What about using bit fields? You will still have padding, but it will be at the end. You will have to play games with type conversions though. typedef struct { unsigned shortVal : 16, longVal : 32; } myStruct; myStruct aStruct; short aShort; float aFloat; aStruct.shortVal = aShort; aStruct.longVal = *((unsigned *) &aFloat); This should work. You can then write the necessary number of bytes starting at the beginning of the structure. (I could not get it to work under Visual C++, but I am sure it is due to some MS setting. Check your compiler for specifics on bit fields.) Bob ________________________________________________________________________ D I A L Bob Kirnum _|_|_|O Software Engineer _|_|_|G 973-993-3000 Ext. 6589 _|_|_|I 973-993-5916 Fax _|_|_|C mailto:R.Kirnum@dialogic.com Get the Dialogic Edge at http://www.dialogic.com Dialogic Corp. / 1515 Rt. 10 / Parsippany, NJ 07054 ________________________________________________________________________ Did you know . . . that FireWire was developed by Apple? . . . that Apple is 150 in the Fortune 500 and Microsoft is 172? Believe in Macintosh? Check out http://www.evangelist.macaddict.com/ Join by emailing mailto:evangelist@macway.com --MimeMultipartBoundary-- From cgrames@mdc.com Tue Sep 2 11:10:53 1997 From: Charlie Grames Date: Tue Sep 2 11:10:56 PDT 1997 Subject: comp.os.vxworks newsdigest -Reply --MimeMultipartBoundary Content-Type: text/plain Content-Disposition: inline Newsgroups: comp.os.vxworks Subject: Network Load of Symbol Table using FTP Troy, Are you using Windows NT? We were having similar problems when using the wftpd32 FTP server that comes bundled with Tornado. We switched to the FTP server included with the Microsoft Peer Web Services (bundled with NT Workstation), and the problems went away. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com Date: Fri, 29 Aug 1997 18:12:05 GMT From: "Troy A. Moon" Organization: Hughes Defense Communications Message-ID: <340710F5.3AC9@most.fw.hac.com> Reply-To: tamoon@most.fw.hac.com Sender: usenet@most.fw.hac.com (News Administration) We've been experiencing problems downloading the Symbol Table (vxWorks.sum) from a host using FTP. If several boards try to load the Symbol Table all at once, one of them will get the file successfully, but very slowly. All the other boards will fail to load the file. Has anyone experienced this problem? - -- Troy A. Moon Software Engineer Hughes Defense Communications tamoon@most.fw.hac.com --MimeMultipartBoundary-- From dje@dmc10.com Tue Sep 2 17:14:22 1997 From: dje@dmc10.com Date: Tue Sep 2 17:14:25 PDT 1997 Subject: Real Time Software Engineer Needed Please excuse the intrusion. Your name was linked to a software engineering web site related to real time product development and I was hoping you could help me out. I'm looking to hire a Software Engineer experienced with real time embedded systems to be applied to the full life cycle development of medical devices such as patient monitors and surgical equipment. Experience with C and C++ is required. Position involves developing specs with marketing, defining system requirements, development of the systems, and overseeing the creation of documentation and test procedures. We can provide salary to $110,000.00 and great benefits. If you know anyone that might be interested, I can be reached at: Voice: (609) 586-9000 ext 273 Fax: (609)584-9575 E-mail: dje@dmc10.com Thanks, Dave Eide From daemon@csg.lbl.gov Thu Sep 4 04:01:46 1997 From: daemon@csg.lbl.gov Date: Thu Sep 4 04:01:50 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 4 04:01:43 PDT 1997 Subject: Re: struct variables memory alignment ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: struct variables memory alignment Date: 3 Sep 97 17:27:24 GMT From: "Keith Kingsley" Message-ID: <01bcb88d$b72f6f50$fd35afce@hopscotch> References: <5u64fp$3pu$1@news02.btx.dtag.de> The GNU compiler has an extension which allows you to specify that a certain structure field should be byte-aligned. Define your structure like this: typedef struct mystruct { short test; float ftest __attribute__ ((packed)); } mystruct; You should find that sizeof(mystruct) returns 6 in this case. Of course, __attribute__ ((packed)) is not ANSI-compliant-- you might want to do #define PACKED __attribute__ ((packed)) in a .h file and define your structure this way: typedef struct mystruct { short test; float ftest PACKED; } mystruct; That way if you change compilers you could redefine PACKED. - --Keith Kingsley Aeroflex Lintek --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Fri Sep 5 04:01:17 1997 From: daemon@csg.lbl.gov Date: Fri Sep 5 04:01:20 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Sep 5 04:01:14 PDT 1997 Subject: Re: X25 comms card to VME Subject: Exceptions in Tornado 1.0.1/gnu 2.7.2? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: X25 comms card to VME Date: Mon, 01 Sep 1997 12:08:25 +0100 From: Thomas Liszka Organization: Siemens Plessey Systems Message-ID: <340AA229.3A3D@coweseng.demon.co.uk> References: <3406C5D3.7643@coweseng.demon.co.uk> Thomas Liszka wrote: > > I am trying to interface an X25 comms card to VME. > Can anyone tell me where I might be able to purchase such a card?. The required card must be X25 and HDLC. --------------------------- Newsgroups: comp.os.vxworks,gnu.g++.help Subject: Exceptions in Tornado 1.0.1/gnu 2.7.2? Date: 1 Sep 1997 11:21:03 GMT From: "Fredrik Jonsson" Organization: Ericsson Austria AG Message-ID: <01bcb6c8$de2df700$8bcbbac1@stpc139.sea.ericsson.se> When Tornado 1.0.1 came with the gnu 2.7.2 compiler I was hoping to be able to do some tests of the exception support. (Unfortunately the exception support was rejected by WindRiver so any user wanting to try it has to do it on his own.) We are using the cpu32 (68k) compiler for VxWorks 5.3. The shipped compiler includes the support for exception compilation using "-fhandle-exceptions" and "-frtti", and, it compiles c++ code with exception handling without any complaints. But, when loading binaries compiled with exceptions I get warnings on the following: warning: object module contains undefined symbols: ___register_exceptions, ___find_first_exception_table_match, ___unwind_function, _terminate__Fv Anyone who knows how to get around this? What should/can I rebuild in order resolve those symbols? /FJ --------------------------- End of New-News digest ********************** From johill@lanl.gov Fri Sep 5 09:13:06 1997 From: Jeff Hill Date: Fri Sep 5 09:13:09 PDT 1997 Subject: RE: Exceptions in Tornado 1.0.1/gnu 2.7.2? --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit > > When Tornado 1.0.1 came with the gnu 2.7.2 compiler I was > hoping to be able to do some tests of the exception support. > > (Unfortunately the exception support was rejected by WindRiver > so any user wanting to try it has to do it on his own.) > > We are using the cpu32 (68k) compiler for VxWorks 5.3. > > The shipped compiler includes the support for exception > compilation using "-fhandle-exceptions" and "-frtti", and, it > compiles c++ code with exception handling without any > complaints. > > But, when loading binaries compiled with exceptions I get > warnings on the following: > > warning: object module contains undefined symbols: > ___register_exceptions, ___find_first_exception_table_match, > ___unwind_function, _terminate__Fv > > Anyone who knows how to get around this? What should/can > I rebuild in order resolve those symbols? > As I understand it use of exceptions under g++ 2.7.2 also precludes use of the -O option. Does anyone know if there is any work underway at GNU to resolve this problem? Does anyone know of any good articles on the performance cost of placing C++ code inside a catch block? I am interested only in the impact on the normal(no exception present) path through the C++ source code. Jeff ______________________________________________________ Jeffrey O. Hill Internet johill@lanl.gov LANL MS H820 Voice 505 665 1831 Los Alamos NM 87545 USA FAX 505 665 5107 --MimeMultipartBoundary-- From SJohnson@SYSTRAN.com Fri Sep 5 09:51:59 1997 From: Steve Johnson Date: Fri Sep 5 09:52:02 PDT 1997 Subject: Uninitialized Interrupt 2 --MimeMultipartBoundary Content-Type: text/plain "Does anyone happen to know ....?" We have written a driver for an mv2603 and mv3604 that uses a PMC card generating PMC interrupts . The driver runs fine, but occasionally we get the message "uninitialized interrupt 2" while the driver is running. The driver never hangs and appears to be working correctly otherwise. Is this a similar situation which occurred in the BSP back when the 1603 boards printed out "uninitialized interrupt 7" back in the earlier days? If I remember, this was due to noisy interrupt lines. The fix was to simply add in a handler which simply returned. Or is this a different type of problem? Thanx for any Help. Steve Johnson Systran Corp. Dayton, Ohio sjohnson@systran.com --MimeMultipartBoundary-- From daemon@csg.lbl.gov Sat Sep 6 04:00:48 1997 From: daemon@csg.lbl.gov Date: Sat Sep 6 04:00:54 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Sep 6 04:00:45 PDT 1997 Subject: VxWorks TIMEZONE environment variable ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks TIMEZONE environment variable Date: Mon, 25 Aug 1997 13:23:02 -0700 From: Curtis Bray Message-ID: <3401E9A6.DB08A97@best.com> Reply-To: cbray@best.com I've a bit confused by the documentation in the ansiTime library. >If the environment variable TIMEZONE is set, the information is >retrieved from this variable, otherwise from the locale information. > TIMEZONE is of the form: > > name_of_zone:(unused): time_in_minutes_from_UTC:daylight_start:daylight_end > > To calculate local time, the value of time_in_minutes_from_UTC is subtracted > from UTC; time_in_minutes_from_UTC must be positive. I saw nothing in the locale.h that support timezone, so I used the environment variable method described above. I followed these examples and always set time_in_minutes_from_UTC to be positive. For example: GMT-1 means time_in_minutes_from_UTC = 60 GMT+1 means time_in_minutes_from_UTC = 12*60+1*60 = 780 (this keeps it positive) This didn't respond like I thought it would and once I set time_in_minutes_from_UTC to -60 for the GMT+1 Timezone everything worked correctly. Any ideas what my the manual states that time_in_minutes_from_UTC must be positive? Thank you! Curtis Bray cbray@best.com --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Sun Sep 7 04:00:21 1997 From: daemon@csg.lbl.gov Date: Sun Sep 7 04:00:24 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Sep 7 04:00:17 PDT 1997 Subject: GET THE CASH YOU NEED //////////// ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: GET THE CASH YOU NEED //////////// Date: Sat, 6 Sep 1997 18:14:55 From: CASH4U@dream.com Organization: Bell Atlantic Internet Solutions Message-ID: <5usknp$d3n@world1.bawave.com> CASH GRANTS CASH GRANTS CASH GRANTS Foundations all over the United States GIVE CASH GRANTS. ANYONE can apply for a Grant from 18 years and up... This money HAS to be given away, WHY not to YOU? Grants from $500.00 to $50,000.00 possible, in some instances. Grants don't have to be paid back. Grants can be ideal for people who are or were bankrupt or just have bad credit. Get the money you need to start that business, you have always wanted. To get your list of FOUNDATIONS that give grants, AND instructions on how to apply. Send a check or money order for ONLY $7.00 to: TO RECEIVE BY US MAIL ENCLOSE A STAMPED, SELF ADDRESSED, #10 ENVELOPE. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ B.C.S. 39 GURLEY ROAD #201 EDISON, NJ 08817 Att: GRANT INFO ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Include your E-MAIL ADDRESS for faster service. --------------------------- End of New-News digest ********************** From ENG.Gordon.Manning@megadrive.cchub.com Mon Sep 8 08:24:40 1997 From: ENG.Gordon.Manning@megadrive.cchub.com (ENG Gordon Manning) Date: Mon Sep 8 08:24:43 PDT 1997 Subject: Galileo 4 BSP --MimeMultipartBoundary Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Has anyone tried to get the Galileo 4 BSP to work? I have been unsuccessful and would appreciate any advice. Thank you, J. Gordon Manning MegaDrive Systems jmanning@megadrive.com --MimeMultipartBoundary-- From RadaJA@corning.com Mon Sep 8 12:05:20 1997 From: "Rada, Jeff A" Date: Mon Sep 8 12:05:24 PDT 1997 Subject: Second Ethernet controller for MVME177 --MimeMultipartBoundary Content-Type: text/plain Content-Transfer-Encoding: 7BIT All, Does any one know if there are ethernet controllers for the VME BUS that will work as a second controller for a MVME177 (68060) processor. We wish to create a secondary lan that Communicates to host processes and a primary using the on board controller. I know there are ethernet controllers out there but will they work with my board and vxWorks? Thanks in advance. --MimeMultipartBoundary-- From owner-vxwexplo-process Wed Sep 10 07:29:46 1997 From: "Mabrouk Aboughanaima" Date: Wed Sep 10 07:29:49 PDT 1997 Subject: C++ Makefile for PowerPC860 Hi, Does anyone have any C++ Makefile that comiples a multi file project for PowerPC860. My CPU is PPC860 VxWorks 5.3.1 TOOL is gnu (GNU Make version 3.74+wrs-2) Please if anyone has a file or an idea e_mail me at mabrouka@nortel.ca or phone <613>763-6000 Thanks a lot. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ + Mabrouk Aboughanaima + Tel: (613) 763-6000 + + CDMA Software + ESN: 393-6000 + + Nortel Wireless Networks + FAX: (613) 763-2686 + + P.O. Box 3511, Station 'C' + Internet: mabrouka@nortel.ca+ + Ottawa K1Y 4H7 Canada + + ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From owner-vxwexplo-process Wed Sep 10 11:43:35 1997 From: Mark Muri Date: Wed Sep 10 11:43:38 PDT 1997 Subject: SM_MAX_TRIES set to 100, yet smObjShow reports 5001 --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" All, We're in the popular cycle of evaluating the current VME PowerPC offerings. The current problem is affecting multiple vendor's cards, so I'm wondering if I am running into a more general VxWorks SM_NET and/or vxMP issue. The smObjShow output (below) shows a bizarre number for the Maximum Tries to Take Lock field, even though the various config.h files set this upper limit to 100. Similarly, the smNetShow output is showing a large number of packet errors that may be attributed to this. I've been able to trace this number down to a consecutive number of 'false' return values from each vendor's sysBusTas routines. That is, at some point, the sysBusTas routine said it could not TAS the backplane 5001 times in a row. I'm discussing with the vendors what the BSP issues may or may not be. I thought I'd open it up and see if any general users may have already seen/fixed this issue. I have a hard time believing that my bus is *that* heavily loaded when this condition occurs (although I have 6 to 8 CPUs on the backplane at the time). We're on Tornado 1.0.1 with the Jumbo Patch 3. We have many, many INCLUDE_ options set, but I'm not convinced they relate to this problem (i.e. INCLUDE_SCSI, INCLUDE_POSIX, etc). Comments? -> smObjShow Shared Mem Anchor Local Addr : 0x4100 Shared Mem Hdr Local Addr : 0x2002c Attached CPU : 8 Max Tries to Take Lock : 5001 Shared Object Type Current Maximum Available -------------------- ---------- --------- ---------- Tasks 0 200 200 Binary Semaphores 0 50 50 Counting Semaphores 0 50 50 Messages Queues 0 100 100 Memory Partitions 0 20 20 Names in Database 0 200 200 value = 0 = 0x0 -> -> smNetShow Anchor Local Addr: 0x4100 Sequential addressing enabled. Master address: 129.46.153.101 heartbeat = 67953, header at 0x101ec, free pkts = 19. cpu int type arg1 arg2 arg3 queued pkts --- -------- ---------- ---------- ---------- ----------- 0 mbox-1 0x2d 0xcc00 0 0 1 mbox-1 0x2d 0xcc08 0 0 2 mbox-1 0x2d 0xcc10 0 0 3 mbox-1 0x2d 0xcc18 0 0 4 mbox-1 0x2d 0xcc20 0 0 5 mbox-1 0x2d 0xcc28 0 0 6 mbox-1 0x2d 0xcc30 0 0 7 mbox-1 0x2d 0xcc38 0 0 input packets = 3442 output packets = 6033 input errors = 6 output errors = 15 collisions = 0 value = 0 = 0x0 -> Mark <==========================o============================================> | Mark E. Muri | E-Mail: mailto:mmuri@qualcomm.com | | Sr. Staff Engineer/Mgr | Alpha: mailto:mmuri@pager.qualcomm.com | | R-241H | Work: (619) 658-3303 | | QUALCOMM, Inc | Fax: (619) 651-1987 | | 10185 McKellar Ct. | Pager: (619) 636-0446 | | San Diego, CA 92121-2779 | PCS: (619) 207-3139 | | Go Gulls! | VME/C++/CCP/GSS/680x0/PowerPC/VxWorks/Ada | <==========================o============================================> --MimeMultipartBoundary-- From daemon@csg.lbl.gov Thu Sep 11 04:00:41 1997 From: daemon@csg.lbl.gov Date: Thu Sep 11 04:00:44 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 11 04:00:36 PDT 1997 Subject: problem with transparent FTP file access ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: problem with transparent FTP file access Date: 9 Sep 1997 06:25:22 GMT From: schuehle@t-online.de (Martin Sauter) Organization: Schuehle Measurement and Control Systems Message-ID: <5v2q4i$br6$1@news02.btx.dtag.de> Reply-To: ms@schuehle.com Hi all, I am using transparent file access with FTP to transfer files generated on my x86 BSP embedded system (Tornado 1.0, VxWorks 5.3) to another computer. This works fine most of the time but sooner or later there seems to be a problem during data transmission to the FTP server and the whole embedded system hangs. The task that is using the transparent FTP file access has one of the lowest priorities in the system so something strange is going on. This is an excerpt from the FTP server log: 16:01:02 Connection accepted from 147.11.72.100 16:01:02 Command "USER target" received 16:01:02 PASSword accepted 16:01:02 User target logged in 16:01:02 Command "TYPE I" received 16:01:02 TYPE set to I N 16:01:02 Command "CWD c:/pr/mm5" received 16:01:02 Command "PORT 147,11,72,100,4,69" received 16:01:02 Port set to 147.11.72.100 - 1093 (4,69) 16:01:02 Command "STOR as05_05.dat" received 16:01:02 STORe started on file as05_05.dat - -> so far this is o.k. but here comes the error : 16:02:38 Transfer aborted -> (usually only takes about a second) 16:02:38 Put file c:/pr/mm5/as05_05.dat unsuccessfully 16:02:38 QUIT or close. User target logged out Any idea what the problem could be? thanks in advance Martin P.S.: Just had a couple of other errors during ftp transfer to the embedded systems. The first time I got the following error message: "Socket error 10054" and the second time the ftp server said "Received data signal FD_Write on socket 6 which I can't find". Both times the embedded system hung afterwards... --------------------------- End of New-News digest ********************** From dsunnarb@wavetech.net Thu Sep 11 09:40:58 1997 From: Dan Sunnarborg Date: Thu Sep 11 09:41:02 PDT 1997 Subject: Bootable VxWorks application --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anyone who can help, I am tying to make a bootable application by using your taskSpawn command. The problem is, I don't wish to build it into the vxworks image due to its large size. Is it possible to spawn a task that is outside the vxworks image, even though it is being loaded PRIOR to the taskSpawn call as follows in usrRoot(): usrRoot() { . . . #ifdef INCLUDE_USER_APPL /* Startup the user's application */ USER_APPL_INIT; /* must be a valid C statement or block */ #endif /* Load module containing hostinit() */ env_id = loadModule(fd_env, LOAD_ALL_SYMBOLS); taskSpawn ("HostInit", 100, 0, 20000, (FUNCPTR) hostinit, 0,0,0,0,0,0,0,0,0,0); } I can declare the task as external, so the "C" file builds, but the link fails with: ld386 -X -N -e _sysInit -Ttext 00108000 \ -o vxWorks dataSegPad.o vxWorks.tmp ctdt.o vxWorks.tmp(.text+0x5a87): undefined reference to `hostinit' I also tried to add the -r option to the vxworks makefile, but upon loading the image, I received an error and nothing loaded. Is there a solution to this problem, or do I have to build an incredibly large vxworks image with everything linked ahead of time? Thanks in advance! Dan Sunnarborg Applied Biometrics Inc. Burnsville, MN (612) 890-1123 dsunnarb@wavetech.net --MimeMultipartBoundary-- From pbanta@adc.com Thu Sep 11 13:20:14 1997 From: pbanta@adc.com (Paul Banta) Date: Thu Sep 11 13:20:17 PDT 1997 Subject: gdbm port Before I reinvent the wheel, has anyone ported gdbm to VxWorks? If so, would you mind sharing your work? -Paul -- Paul Banta voice 972-680-7666 ADC Telecommunications fax 972-680-7672 Network Services Division email pbanta@adc.com From daemon@csg.lbl.gov Fri Sep 12 04:01:53 1997 From: daemon@csg.lbl.gov Date: Fri Sep 12 04:01:56 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Fri Sep 12 04:01:48 PDT 1997 Subject: Re: VxWorks won't close old socket Subject: How do I read the keyboard? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks won't close old socket Date: Thu, 11 Sep 1997 06:47:16 -0700 From: Chet Kackman Organization: GDE Systems Inc. Message-ID: <3417F664.7F3@GDEsystems.com> References: <873835177.263688@wagasa.cts.com> Mathue Moyer wrote: > > Here's the story. I'm evaluating VxWorks (5.3.1) networking performance on > an i486 platform. I have an application put together that accepts telnet > connections on a well-known port, copies text amongst the various > connections, and when instructed to do so, shuts itself down. On various > other platforms, all connections are cleanly terminated when the application > shuts down, and I can immediately reuse the port that was being used. > For those familiar with using sockets, I use setsockopt to set SO_REUSEADDR > to ON and SO_LINGER to OFF (I've also tried ON with a linger time of 0). > Problem is, it _doesn't_ work that way on the VxWorks platform. The > existing connections are not closed cleanly (remote client is just left > sitting instead of getting the "Connection closed by foreign host" message) > and no matter how long I wait (up to several hours) I can't reuse the same > port number (it complains about being in use when I try to bind() to it). > I figured maybe it was a problem with counting on the implicit "close()" > performed when this app runs as a process rather than a VxWorks task, but > explicitly closing all sockets (all connections and the host socket) didn't > fix the problem. I tried using shutdown(fd, 2) on all of the sockets as > well, but that didn't do it either. > > So the question is... has anybody else experienced problems with VxWorks > failing to properly terminate network connections? > > -- > Mathue Moyer | Anyone who says they have only > Email: mathue@cts.com | one life to live must not know > URL: http://www.users.cts.com/king/m/mathue | how to read a book. Matt Try using the select function to test for read/write or exception for each FD in use, It may help you detect when your socket is closed. Regards Chet --------------------------- Newsgroups: comp.os.vxworks Subject: How do I read the keyboard? Date: 11 Sep 1997 15:25:09 GMT From: jim@geigerpc.gsfc.nasa.gov (Jim Geiger) Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Message-ID: <5v92gl$f7r@post.gsfc.nasa.gov> Hello, I am running Tornado 1.0.1 on a Win95 host with a PPC 603e target (a MVME 2600-2 board in a VME chassie). I would like to be able to read/capture keyboard input. In a DOS environment, I would use the functions in , and, in a unix environment, I would use functions in to set up the console so I could read the keyboard. Does anyone know how I do this in the VxWorks environment? Thanks, Jim --------------------------- End of New-News digest ********************** From nmehrotra@yurie.com Fri Sep 12 08:21:54 1997 From: nmehrotra@yurie.com Date: Fri Sep 12 09:04:30 PDT 1997 Subject: Bootable VxWorks application --MimeMultipartBoundary Content-type: text/plain; charset=US-ASCII From: Nitin Mehrotra on 09/12/97 11:19 AM Hi Dan, What you're trying to do is interesting but tricky; you have to do a little more than what you're doing now. Starting a task involves optionally loading the code in memory and then creating a task with one function from that code. You can keep the code external and load it into memory as you're doing but how do you know the function to start the task with? Specifically, in your taskSpawn call: taskSpawn ("HostInit", 100, 0, 20000, (FUNCPTR)hostinit, 0,0,0,0,0,0,0,0,0,0); Where is the function hostinit()? It is in the code you load, so vxWorks will know it's address at load time but does not know its address at compile/link time. So the compiler flags it as an external function and the linker barfs because it doesn't find it. What you have to do is to find the address of the hostinit function at runtime and then call taskSpawn with that address. The only way I can think of is using the symLib to find the address. Try this piece of code: { FUNCTPTR pFunc; env_id = loadModule(fd_env, LOAD_ALL_SYMBOLS); if (ERROR == (symFindByName(sysSymTbl, "hostinit", &pFunc, &symbolType))) { /* handle function not found error */ } taskSpawn ("HostInit", 100, 0, 20000, pFunc, 0,0,0,0,0,0,0,0,0,0); } That should do the work for you. I haven't tested it out so I'd be interested in knowing your results. Oh, btw, check all errors. I will save you a lot of grief one day. HTH nitin From dsunnarb@wavetech.net Thu Sep 11 09:40:58 1997 From: Dan Sunnarborg Date: Thu Sep 11 09:41:02 PDT 1997 Subject: Bootable VxWorks application --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Anyone who can help, I am tying to make a bootable application by using your taskSpawn command. The problem is, I don't wish to build it into the vxworks image due to its large size. Is it possible to spawn a task that is outside the vxworks image, even though it is being loaded PRIOR to the taskSpawn call as follows in usrRoot(): usrRoot() { . . . #ifdef INCLUDE_USER_APPL /* Startup the user's application */ USER_APPL_INIT; /* must be a valid C statement or block */ #endif /* Load module containing hostinit() */ env_id = loadModule(fd_env, LOAD_ALL_SYMBOLS); taskSpawn ("HostInit", 100, 0, 20000, (FUNCPTR) hostinit, 0,0,0,0,0,0,0,0,0,0); } I can declare the task as external, so the "C" file builds, but the link fails with: ld386 -X -N -e _sysInit -Ttext 00108000 \ -o vxWorks dataSegPad.o vxWorks.tmp ctdt.o vxWorks.tmp(.text+0x5a87): undefined reference to `hostinit' I also tried to add the -r option to the vxworks makefile, but upon loading the image, I received an error and nothing loaded. Is there a solution to this problem, or do I have to build an incredibly large vxworks image with everything linked ahead of time? Thanks in advance! Dan Sunnarborg Applied Biometrics Inc. Burnsville, MN (612) 890-1123 dsunnarb@wavetech.net --MimeMultipartBoundary-- --MimeMultipartBoundary-- From rodmelby@csys.win-uk.net Sat Sep 13 04:23:54 1997 From: rodmelby@csys.win-uk.net (Rodney W Melby) Date: Sat Sep 13 04:24:01 PDT 1997 Subject: No mail since Sept 8th. I haven't had any mail since Sept 8th. To my knowledge I have not cancelled my subscription. Any idea? Regards Rod From daemon@csg.lbl.gov Sun Sep 14 04:00:42 1997 From: daemon@csg.lbl.gov Date: Sun Sep 14 04:00:46 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sun Sep 14 04:00:39 PDT 1997 Subject: executing MMX code ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: executing MMX code Date: 12 Sep 1997 01:55:33 GMT From: dixon@ra.crd.ge.com (walt dixon) Organization: GE Corp. Research & Development, Schenectady, NY Keywords: x86 mmx coff a.out Message-ID: <5va7el$7pv$1@rdsunx.crd.ge.com> I have a need to execute a fairly sizeable chunk of mmx code on a vxworks hosted system. My target is x86 and my host is windows nt. I currently have compiled and tested this code using visual c++. Is there a relatively straight-forward way to convert this code to run on vxWorks. The basic problem seems to be that there is no coff support in the x86 tool set. I have considered the following alternatives: (1) create a custom version of objcopy that supports both a.out and coff. Any thoughts on how difficult this task would be? (BFD supports both formats). (2) Generate an assembly language file and assemble it with NASM. (3) Rewriting the code in gas and using intel mmx macros. Am I missing something obvious? Any suggestions would be appreciated. The simpler the solution, the better. TIA Walt Dixon email: dixon@crd.ge.com us mail: Walt Dixon GE Corp. R&D PO Box 8 Schenectady, NY 12301 voice: 518-387-5798 fax: 518-387-5164 - -- Walt Dixon email: dixon@crd.ge.com us mail: Walt Dixon GE Corp. R&D PO Box 8 --------------------------- End of New-News digest ********************** From daemon@csg.lbl.gov Mon Sep 15 04:01:25 1997 From: daemon@csg.lbl.gov Date: Mon Sep 15 04:01:29 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Mon Sep 15 04:01:22 PDT 1997 Subject: Re: Big endian and Little endian( PC to PPC) Subject: Re: VxWorks won't close old socket ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: Re: Big endian and Little endian( PC to PPC) Date: Sun, 14 Sep 1997 08:08:40 GMT From: valentin@netcom.com (Valentin Pepelea) Organization: Netcom Communications, Santa Clara, CA Message-ID: References: <19970913215400.RAA10530@ladder02.news.aol.com> Sender: valentin@netcom11.netcom.com In article <19970913215400.RAA10530@ladder02.news.aol.com> caoy@aol.com (Caoy) writes: > >We need send a data structure down to the target through TCP/IP from the >PC host. On the PC side data format is little endian while big endian on >the target >side. > >My question: > >1) If we send the data structure as normal from PC side, what does the >byte order of the structures look like? For instances, there are four >integers in the structure. Are the four integers in the same order as they >are on the PC side but the byte order of the each integer is reversed OR >the byte order of the whole structure is reversed ? Yup, correct analysis. >2) Has anyone had this similiar situation? Oh yeah, very often. >How to get the correct data structure on the target side? What is the better >way to take care of this problem?? On the target side, we simply use a macro to convert the integers: #define ConvertEndian(x) ((x&0xFF)<<24 | (x&0xFF00)<<8 | (x&0xFF000000)>>8 | (x&0xFF00000000)>>24) I wrote this from memory, so double check my coding. Valentin - -- "An operating system without virtual memory Valentin Pepelea is an operating system without virtue." Embedded Systems Consultant (415) 516-7293 - Ancient Inca Proverb valentin@netcom.com --------------------------- Newsgroups: comp.os.vxworks Subject: Re: VxWorks won't close old socket Date: Fri, 12 Sep 1997 11:16:52 GMT From: fudath@ct.essd.northgrum.com (Thomas Fuda) Organization: Northrop Grumman Corporation, Baltimore, MD Message-ID: <3419217f.948714@155.104.79.99> References: <873835177.263688@wagasa.cts.com> Sender: news%mother@mdnews.md.essd.northgrum.com On 9 Sep 1997 19:59:37 GMT, mathue@king.cts.com (Mathue Moyer) wrote: >So the question is... has anybody else experienced problems with VxWorks >failing to properly terminate network connections? > Yes, I have has some problems with TCP sockets under VxWorks. Overcame our problems by enabling the TCP_KEEPALIVE option on our sockets and messing with the timers that control the keepalive function. The timers are initialized with the default values specified in tcp_timer.h (TCPTV_KEEP_INIT, TCPTV_KEEP_IDLE, TCPTV_KEEPINTVL, and TCPTV_KEEPCNT determine the defaults). These default values are extremely long in duration. Rather than changing the values in tcp_timer.h and recompiling the socket libraries I changed the global variables tcp_keepinit, tcp_keepidle, tcp_keepintvl and tcp_keepcnt (which are initialized with the values specified in tcp_timer.h). Maybe this would help you. Thomas Fuda Sr. Software Engineer Northrop Grumman Norden Systems --------------------------- End of New-News digest ********************** From cruz_nojunk@xyplex.com Mon Sep 15 09:02:21 1997 From: Roger Cruz Date: Mon Sep 15 09:02:24 PDT 1997 Subject: Relocating the bootrom file in a PC 486 --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Has anyone been successful in relocating the starting address of the bootrom code in a PC? The documentation says I should changed RAM_HIGH_ADRS in the VxWorks makefile and config.h. I changed mine from the default of 0x8000 to 0x10000 and I can't make it work. The system reboots. I'm using vxload.com and I have verified that the bootrom image fits in the free space. Thanks. PS. Remove "_nojunk" from email address to contact me. -- Roger Cruz cruz_nojunk@xyplex.com Xyplex Networks w: 508-952-4783 295 Foster Street f: 508-952-4887 Littleton, MA 01460 --MimeMultipartBoundary-- From goldappl@services.aminvestnet.com Mon Sep 15 21:03:15 1997 From: "Golden Apples LP" Date: Mon Sep 15 21:03:22 PDT 1997 Subject: ANIMATION -- Cels -- Giveaway Dear Art Enthusiast: The Golden Apples Group of Los Angeles is proud to offer original art "cels" (value $100.00*) at no cost to you in exchange for completing an industry survey at the Golden Apples web site. These hand-painted, one-of-a-kind cels were used used in the hit TV series Teenage Mutant Ninja Turtles. These cels are the last of an art form soon to be extinct, because most of today's animation is computer generated. This is an excellent gift or art investment and includes a Certificate of Authenticity. We are giving away this special piece of art because we hope this limited-time promotion will attract art enthusiasts, such as you, as well as your friends and family. In addition we want to create awareness of Golden Apples' new web sites. The principals of our company have been in the art business for many years and are one of the larger galleries buying and selling decorative, fine and collectible art. To see our cel samples and take the survey, click here: http://www.goldenapples.com Thank you. * Only one per person. Each cel measures 7 1/2 x 9 1/2 inches Shipping and handling are not included Note this is a one time mailing To be deleted from our mailing list, please click here: http://www.goldenapples.com/remove_me.html From daemon@csg.lbl.gov Tue Sep 16 04:00:37 1997 From: daemon@csg.lbl.gov Date: Tue Sep 16 04:00:40 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Tue Sep 16 04:00:33 PDT 1997 Subject: http://home.fia.net/~crespom ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: http://home.fia.net/~crespom Date: 10 Sep 1997 12:03:13 GMT From: merlito Organization: paper air machines Message-ID: <5v62a1$aqc$11164@usenet88.supernews.com> http://home.fia.net/~crespom --------------------------- End of New-News digest ********************** From wbrown@csg.lbl.gov Tue Sep 16 07:30:55 1997 From: wbrown@csg.lbl.gov (Bill Brown) Date: Tue Sep 16 07:30:58 PDT 1997 Subject: This is a test of the vxWorks exploder and broad-scattering system. This message was sent to the exploder via email. You do not need to adjust your set at this time. Disclaimer: Any opinions are my own and have | -bill nothing to do with the official policy or the | wlbrown@lbl.gov management of L.B.N.L, who probably couldn't | Berkeley, CA care less about employees who play with trains. | aka wlbrown@netcom.com From jlao@yurie.com Tue Sep 16 09:56:45 1997 From: jlao@yurie.com Date: Tue Sep 16 09:57:01 PDT 1997 Subject: VxWorks Needed --MimeMultipartBoundary Content-type: text/plain; charset=US-ASCII James Lao 09/16/97 12:54 PM We are looking for experience VxWorks with an excellent background in telecommunications in area of SS7 and ISDN. If you are interested, please contact me. James Lao Yurie Systems, Inc. 8301 Professional Place Landover, MD 20785-2237 jlao@yurie.com tel.(301)-352-4606 fax.(301)-352-4680 --MimeMultipartBoundary-- From R.Kirnum@dialogic.com Tue Sep 16 11:37:02 1997 From: "Kirnum, Robert" Date: Tue Sep 16 11:37:06 PDT 1997 Subject: Startup Script --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Is there some way to have a startup script (as in VxWorks 5.2) using Tornado? ________________________________________________________________________ D I A L Bob Kirnum _|_|_|O Software Engineer _|_|_|G 973-993-3000 Ext. 6589 _|_|_|I 973-993-5916 Fax _|_|_|C mailto:R.Kirnum@dialogic.com Get the Dialogic Edge at http://www.dialogic.com Dialogic Corp. / 1515 Rt. 10 / Parsippany, NJ 07054 ________________________________________________________________________ Did you know . . . that FireWire was developed by Apple? . . . that Apple is 150 in the Fortune 500 and Microsoft is 172? Believe in Macintosh? Check out http://www.evangelist.macaddict.com/ Join by emailing mailto:evangelist@macway.com --MimeMultipartBoundary-- From 82162060@msn.com Tue Sep 16 19:17:38 1997 From: 82162060@msn.com Date: Tue Sep 16 19:17:41 PDT 1997 Subject: WARNING - Major Investigation May Involve You ARE YOU BEING INVESTIGATED???? Learn the Internet tools that are used to investigate you, your friends, neighbors, enemies, Employees or anyone else! My huge report "SNOOPING THE INTERNET" of Internet sites will give you... * Thousands of Internet locations to look up people, credit, Social security, current or past employment, Driving records, medical information, addresses, phone numbers, Maps to city locations... Every day the media (television, radio, and newspapers) are full of stories about PERSONAL INFORMATION being used, traded, and sold over the Internet... usually without your permission or knowledge. With my report I show you HOW IT'S DONE!!! It's amazing.. Locate a debtor that is hiding, or get help in finding hidden assets. * Find that old romantic interest. * Find e-mail, telephone or address information on just about anyone! Unlisted phone numbers can often be found through some of these sites!! Perhaps you're working on a family "tree" or history. The Internet turns what once was years of work into hours of DISCOVERY & INFORMATION. Check birth, death, adoption or social security records. MILITARY Check service records of Army, Navy, Air Force or Marine Corps. Find out who's been telling the truth and who's been lying. Perhaps you can uncover the next lying politician!!! FELLOW EMPLOYEES; * Find out if your fellow employee was jailed on sex charges, or has other "skeletons" in the closet!! PERFORM BACKGROUD CHECKS; Check credit, driving or criminal records, Verify income or educational claims, Find out Military history and discipline, previous political affiliations, etc. YOUR KID'S FRIENDS; Find out the background of your children's friends & dates. WHATS THE LAW? STOP GUESSING!! * Look up laws, direct from law libraries around the world. Is that new business plan legal?? NEW JOB? NEW TOWN? NEW LIFE? Employment ads from around the world can be found on the Internet. Get a new job and disappear! The Internet can tell you just about ANYTHING, if you know WHERE to look. BONUS REPORT!!!! Check your credit report and use the Internet to force credit bureaus to remove derogatory information. My special BONUS REPORT included as part of the "SNOOPING THE INTERNET" collection reveals all sorts of credit tricks, legal and for "information purposes only" some of the ILLEGAL tricks. Research YOURSELF first! What you find will scare you. If you believe that the information that is compiled on you should be as easily available to you as it is to those who compile it, then. . . You want to order the SNOOPING THE INTERNET report I've put together. This huge report is WHERE YOU START! Once you locate these FREE private, college and government web sites, you'll find even MORE links to information search engines! YOU CAN FIND OUT ANYTHING ABOUT ANYBODY ANY TIME using the Internet!!!! 1) WE TAKE: AMERICAN EXPRESS <> VISA <> MASTERCARD TYPE OF CARD AMX/VISA/MC??____________________ NAME ON CREDIT CARD___________________________ CREDIT CARD #___________________________________ BILLING ADDRESS ________________________________ CITY_____________________________________________ STATE________________ZIP________________________ PHONE INCLUDE AREA CODE______________________ WE WILL BILL 39.95 to your account SALES TAX (2.90) added to CA residents >>> Send $39.95 ($42.85 in CA) cash, check or money order to: >>> CASINO CHICO >>> Background Investigations Division >>> 305 Nord Ave. >>> P.O. Box 4191 >>> Chico, CA 95927-4191 2) Send the same above requested credit card information to above address. 3) Fax the same above credit card information to 916-895-8470 4) Call phone # 916-876-4285. This is a 24 hour phone number to place a CREDIT CARD order. I will RUSH back to you SAME DAY my "SNOOPING THE INTERNET" report! Log on to the Internet and in moments you will fully understand... What information is available -- and exact Internet site to get there! 2nd BONUS!!!! Along with the report we will send a 3 1/2" disk with sites already "HOT LINKED". No need to type in those addresses. Simply click on the URL address and "PRESTO" you are at the web site!!! Personal ads, logs of personal e-mail, mention of individuals anywhere on the Internet are "your for the taking" with this report. Lists of resources to find even more information (private Investigation companies, etc..) Order surveillance equipment (if legal in your state) Send anonymous e-mail Research companies Research technology Locate military records FIND INFORMATION ON CRIMINALS Find Wanted fugitives - perhaps even a close associate! ABSOLUTE SATISIFACTION GUARANTEED: Your satisfaction is 100% guaranteed, just return the material For a full refund within 30 days if you aren't 100% satisfied. Copyright 1997 All Rights Reserved R Jon Scott Hall Publications. THIS IS NOT A FRANK MULLEN / CYBER PRODUCTS / SPYMASTER PUBLICATION From wbrown@csg.lbl.gov Wed Sep 17 07:10:58 1997 From: wbrown@csg.lbl.gov (Bill Brown) Date: Wed Sep 17 07:11:01 PDT 1997 Subject: This is a test of the vxworks exploder. -bill From mwbrown@dragon Wed Sep 17 08:45:36 1997 From: mwbrown@dragon.sandia.gov (Matthew W. K. Brown) Date: Wed Sep 17 08:45:38 PDT 1997 Subject: Problems upgrading to Tornado --MimeMultipartBoundary Content-Type: text/plain Content-Transfer-Encoding: 7bit vxWorkers: I am upgrading some S/W from 5.2 to Tornado and I have run into some problems. The code seems to work fine under 5.2. I took this same code and recompiled it with the compilers shipped with 5.3.1. It compiles fine, loads fine, and seems to run fine--except, it doesn't handle interrupts anymore. The device that generates the interrupts is a non-vxWorks device and shouldn't be effected by the upgrade. I did stick a VMETRO bus analyser board in to verify this and sure enough, interrupts are being generated (as expected) ... my code just isn't handling them under Tornado. This *exact* same code did handle them under 5.2! Just to make sure, I switched back to my 5.2 PROMs, recompiled the code again ... and it still works under 5.2. The only thing unusual that I am doing is that I built my 5.3 vxWorks with 5.2 support using the following includes: #define INCLUDE_LOADER /* object module loading */ #define INCLUDE_NET_SYM_TBL /* load symbol table from network */ #define INCLUDE_RLOGIN /* remote login */ #define INCLUDE_SHELL /* interactive c-expression interpreter */ #define INCLUDE_SHOW_ROUTINES /* show routines for system facilities*/ #define INCLUDE_SPY /* spyLib for task monitoring */ #define INCLUDE_STARTUP_SCRIPT /* execute start-up script */ #define INCLUDE_STAT_SYM_TBL /* create user-readable error status */ #define INCLUDE_SYM_TBL /* symbol table package */ #define INCLUDE_TELNET /* telnet-style remote login */ #define INCLUDE_UNLOADER /* object module unloading */ #define INCLUDE_DEBUG /* native debugging */ Any ideas or suggestions? Thanks, Matthew ------------------------------------------------------------------------- Matthew Brown, SMTS Voice: (505) 844-1160 Sandia National Laboratories Fax: (505) 844-7059 P. O. Box 5800 M/S 0965 email: mwbrown@sandia.gov Albuquerque, NM 87185-0965 --MimeMultipartBoundary-- From page@rocko.mit.edu Wed Sep 17 09:02:57 1997 From: "Russell J. Page, Sr." Date: Wed Sep 17 09:03:00 PDT 1997 Subject: Joerger VDACM software --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Folks: I am looking for anyone in the EPICS community that may have the appropriate record and device support software for the Joerger VDACM 8-channel, 16-bit waveform generator. Thanks in advance Russ -- ----------------------------------------------------------------- |Russell J. Page, Sr. Phone (617) 253-9585 | |MIT Bates Linear Accelerator Ctr. FAX (617) 253-9799 | |PO Box 846 email :page@rocko.mit.edu | |Middleton, MA 01949 http://mitbates.mit.edu/ | ----------------------------------------------------------------- --MimeMultipartBoundary-- From budh@venus.gvg.tek.com Wed Sep 17 10:01:27 1997 From: budh@venus.gvg.tek.com (Bud Hensley) Date: Wed Sep 17 10:01:31 PDT 1997 Please unsubscribed. From chenyc@fnal.gov Wed Sep 17 16:35:43 1997 From: Yenchu Chen Date: Wed Sep 17 16:35:46 PDT 1997 Subject: platform supported by VxWorks 5.3 --MimeMultipartBoundary Content-Type: TEXT/PLAIN; charset=US-ASCII To whom it may concern, I learned that the new version, VxWorks 5.3, doesn't support IRIX 6. Could someone tell me what kinds of platform are supported by the VxWorks 5.3? By the way I could not find VxWorks 5.3 in Wind River's web pages!? Best regards, Yen-Chu Chen chenyc@fnal.gov (630) 840-8871 (experiment) (886)-(2) 789-9681 (Inst. of Phys., Academia Sinica) --MimeMultipartBoundary-- From daemon@csg.lbl.gov Thu Sep 18 04:00:40 1997 From: daemon@csg.lbl.gov Date: Thu Sep 18 04:00:44 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Thu Sep 18 04:00:37 PDT 1997 Subject: VxWorks Timing Subject: Power Off function ? ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: VxWorks Timing Date: Wed, 17 Sep 1997 13:29:07 +0200 From: Sam Michiels Organization: Department of Computer Science, K.U.Leuven Message-ID: <341FBF03.3912@cs.kuleuven.ac.be> Hello, I did some basic timing performance tests on our VxWorks target (Motorola 68360 25 Mhz). These tests included semaphorecreate, delete, take, give and the spawning of tasks... In the VxW FAQ, I found figures in the order of a few, to a few hundred MICROseconds... My tests resulted in much much faster performance figures (In the order of picoseconds :-/ (e.g. to create 1000 Mutual Exclusion Semaphores : 15 nanoseconds...) What am I doing wrong, or are the FAQ performance figures outdated (FAQ tests were done on a mv167 25Mhz)? I'm using POSIX 'clock_gettime()' method to do the timing... Can anyone help me with this? I've put part of the program below. Thnx, Sam. - ------------------------------------------------------------------ Here is part of the test program: #include #include #include SEM_ID SemPrio[1000]; void semTakeTiming( void ) { int s; struct timespec start, end; printf("** Timing of Semaphore TAKE **\n\n"); clock_gettime(CLOCK_REALTIME, &start); for(s=0 ; s<1000 ; s++){ semTake( SemPrio[s], WAIT_FOREVER); } clock_gettime(CLOCK_REALTIME, &end); printf("Delay to TAKE 1000 Mutual Exclusion Semaphores (Priority Queueing) : %ld nsecs\n", end.tv_nsec - start.tv_nsec); } --------------------------- Newsgroups: comp.os.vxworks Subject: Power Off function ? Date: 17 Sep 1997 10:48:13 GMT From: Thomas.Eriksson@lu.erisoft.se (Thomas Eriksson) Organization: Erisoft AB, Sweden Message-ID: <5vochd$pui$1@antares.lu.erisoft.se> Reply-To: Thomas.Eriksson@lu.erisoft.se Need a "Power Off" function on a SPARC CPU-5V ( in a Teraforce/Target-32 rack ) with Tornado 1.0.1 and BSP CPU-5V/-5VT/-7V/-8VT Rel. 1.1/3 Got this information: - --------------------------------------------------------------------- The software controlled down logic works as follows: The power down program starts a write access to the VME address $FCFFFFFE Hex. The specially designed decoding logic on the MOTH-32 responds with the DTACK signal on the VMEbus and this signal remains low. After this action, the logic generates the shutdown signal to the power supply. After detecting the shutdown signal, the power supply powers off the DC voltage. - ---------------------------------------------------------------------- Need an example for it, or more information. Have anyone done this or seen it somewhere ? Thanks / TE !~~~~~~~~~~~~~~~~~~~!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!~~~~~~~~~~~~~~~~~~~~~~~~~~~~! ! Thomas Eriksson ! E-mail: ! Home adress: ! !Ericsson Erisoft AB! Thomas.Eriksson@lu.erisoft.se ! Norra Pitholmsvagen 129 ! ! Box 920 ! Tel.company +46 920 202000 ! 941 46 Pitea, SWEDEN ! ! 971 28 LULEA ! Tel.direct +46 920 202773 ! Tel.home +46 911 69164 ! ! SWEDEN ! Tel.ECN 857 2773 ! Tel.GSM +46 70 5958170 ! ! Apt.LU/EPL/L/M ! Telefax +46 920 202099 ! Telefax +46 911 69164 ! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --------------------------- End of New-News digest ********************** From info@triada.si Thu Sep 18 07:26:23 1997 From: info@triada.si Date: Thu Sep 18 07:26:27 PDT 1997 Subject: Re: VxWorks Timing >My tests resulted in much much faster performance figures (In the order >of picoseconds :-/ >(e.g. to create 1000 Mutual Exclusion Semaphores : 15 nanoseconds...) Sam, I do not see semCreate in your source. If this is really missing all your semaphores are uninitialised. Jure ----------------------------------------------------------- info@triada.si Triada d.o.o. Rozna dolina 9/3 1000 Ljubljana Slovenia Europe fax: +386-61-1234-166 tel: +386-61-1234-891 ----------------------------------------------------------- From mea@mclean.sparta.com Thu Sep 18 07:44:28 1997 From: Mike Anderson Date: Thu Sep 18 07:44:31 PDT 1997 Subject: RE: Problems upgrading to Tornado --MimeMultipartBoundary Content-Type: text/plain > Submitted-by mwbrown@dragon Wed Sep 17 08:45:36 1997 > Submitted-by: mwbrown@dragon.sandia.gov (Matthew W. K. Brown) > > vxWorkers: > > I am upgrading some S/W from 5.2 to Tornado and I have run into some > problems. > > The code seems to work fine under 5.2. I took this same code and > recompiled it with the compilers shipped with 5.3.1. It compiles > fine, > loads fine, and seems to run fine--except, it doesn't handle > interrupts > anymore. The device that generates the interrupts is a non-vxWorks > device > and shouldn't be effected by the upgrade. I did stick a VMETRO bus > analyser board in to verify this and sure enough, interrupts are being > > generated (as expected) ... my code just isn't handling them under > Tornado. > This *exact* same code did handle them under 5.2! Just to make sure, > I > switched back to my 5.2 PROMs, recompiled the code again ... and it > still > works under 5.2. > > [snip] > > Hi Matt, > > I suspect that your problem may be your 5.3.1 BSP. You don't > mention which architecture you're using, so I assume that it's > a 68K flavor of some sort. Check out the sysLib.c and verify > that sysIntEnable still works. Just to make sure that the BSP > is still working, try the following: > > -> intConnect(0xa0 << 2, logMsg, "this is an interrupt\n") > -> sysIntEnable(5) > -> sysBusIntGen(5, 0xa0) > > This hooks logMsg up to vector 0xA0 (160 decimal) and should print > the message when you issue the sysBusIntGen (using level 5 in this > example). If you don't get anything with this, then it's > probably a bug in the 5.3.1 BSP. > > It's not that you have the 5.2 code enabled. In fact, I've > found Tornado particularly troublesome WRT local file systems > (just try to do a "cd yourBootHostName:/" from the Wind Shell > and watch the error message occur). Ram disks don't work, NFS > drives with local drive names that aren't visible to the boot > host (e.g., you've mounted them from another VxWorks client) > don't work, etc. Tornado is fine for developing code for that > deeply embedded gizmo that barely has enough RAM for the O/S, > but it really bites for complex, multiprocessor development > projects. Let's all hope that WRS *NEVER* eliminates the > ability to shut Tornado off and get back to good ol' VxWorks. > > Regards, > --MimeMultipartBoundary-- From mmcquade@pop.aa.net Thu Sep 18 11:02:04 1997 From: Mike McQuade Date: Thu Sep 18 11:02:07 PDT 1997 Subject: PMC / PCI Bandwidth --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" Hello, We are using an mv-2603 board and Tornado 1.01. Does anyone have any experience writing to a PMC card ? Im curious what kind of data rates you have seen writing from CPU DRAM to PMC space. How does one achieve the 132Mbyte / second figure ? Thanks. --MimeMultipartBoundary-- From mmcquade@pop.aa.net Thu Sep 18 13:16:37 1997 From: Mike McQuade Date: Thu Sep 18 13:16:40 PDT 1997 Subject: RamDisks, was-> RE: Problems upgrading to Tornado --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" At 07:44 AM 9/18/97 PDT, you wrote: >Submitted-by mea@mclean.sparta.com Thu Sep 18 07:44:28 1997 >Submitted-by: Mike Anderson > snip... >> Ram disks don't work, NFS >> drives with local drive names that aren't visible to the boot >> host .... Mike, What do you mean by "Ram Disks Don't work", we have been using a 2meg ram disk in our system for quite some time, we use it as a place to FTP a file when we update the FLASH memory. We have not had problems with the Ram Disk, yet... So, Id like to hear what problems you have had, to know what to look for. Thanks. Mike. --MimeMultipartBoundary-- From cgrames@mdc.com Thu Sep 18 13:47:05 1997 From: Charlie Grames Date: Thu Sep 18 13:47:09 PDT 1997 Subject: Re: VxWorks Timing Sam, I'm not sure what your clock source on a 68360 is, but, typically, the POSIX timing routines use the resolution of the system clock, which is, by default, 16.67 ms (60 Hz). That being the case, it would be impossible to get a smaller delta time than 16.67 ms. If you want to get finer resolution with your timings, you will need to use the timestamp clock. I don't know whether this is available by default with the BSP for your board. If it is, and assuming your timestamp clock runs as fast as the processor clock (which is doubtful), your tick resolution will be 40 ns. I don't see anything wrong with your code, assuming you are initializing your semaphores elsewhere. You might want to make a call to clock_getres() to determine the resolution of the POSIX clock. I think I can say with great certainty, however, that Wind River currently supports no processor that can complete 1000 calls to semTake in 15 ns. Charlie Grames The Boeing Company (314) 233-1956 Charles.R.Grames@boeing.com From mea@mclean.sparta.com Thu Sep 18 14:55:10 1997 From: Mike Anderson Date: Thu Sep 18 14:55:17 PDT 1997 Subject: RE: RamDisks, was-> RE: Problems upgrading to Tornado --MimeMultipartBoundary Content-Type: text/plain [snip] > What do you mean by "Ram Disks Don't work", we have been using a 2meg > ram disk in our system for quite some time, we use it as a place to > FTP a file when we update the FLASH memory. > > We have not had problems with the Ram Disk, yet... > > > So, Id like to hear what problems you have had, to > know what to look for. > > > So sorry for the confusion. Ram Disks *do* work. You just > can't "cd" to them and then "ls" from the WindShell. That's > what I meant when I said that. The whole issue stems from the > WindShell running locally on the host and the "ls" command on > WindShell executing locally. For instance, connect to the > target via WindShell and do an "ls". That works. Now, > shut your target off and do an "ls". It *still* works! The > only time the WindShell will detect that the target has gone > away is if you execute a command that *forces* the WindShell to > talk to the target. Also, the WindShell only has a subset of > the commands from usrLib.c. E.g., printErrno runs from the > WindShell, but try to load a piece of code that calls printErrno > and you'll get an invalid symbol error for printErrno! Again, > this is probably not a problem for the "skinny" client. But, > for porting code from older version of VxWorks, these problems > are just the tip of the iceberg. > > Regards, > --MimeMultipartBoundary-- From mea@mclean.sparta.com Thu Sep 18 14:59:29 1997 From: Mike Anderson Date: Thu Sep 18 14:59:32 PDT 1997 Subject: RE: PMC / PCI Bandwidth --MimeMultipartBoundary Content-Type: text/plain > -----Original Message----- > From: the vxWorks Users Group Exploder [SMTP:vxwexplo@lbl.gov] > Sent: Thursday, September 18, 1997 5:44 PM > To: vxworks_users@csg.lbl.gov > Subject: PMC / PCI Bandwidth > > Submitted-by mmcquade@pop.aa.net Thu Sep 18 11:02:04 1997 > Submitted-by: Mike McQuade > > --MimeMultipartBoundary > Content-Type: text/plain; charset="us-ascii" > > > Hello, > > We are using an mv-2603 board and Tornado 1.01. > > Does anyone have any experience writing to a PMC card ? > > Im curious what kind of data rates you have seen writing > from CPU DRAM to PMC space. > > How does one achieve the 132Mbyte / second figure ? > > > One uses a *lot* of really powerful drugs and then > washes it down with a fifth of bourbon, then you > squint your eyes and turn your head sideways ;-). > > 132MBs/sec is a myth of the PCI consortium. You'll > maybe see half of that in practice for long transfers and > maybe 75% for 1-2 word bursts. > > Regards, > --MimeMultipartBoundary-- From nrd1jrb@nrd.ups.com Fri Sep 19 05:34:57 1997 From: Jim Bucciferro Date: Fri Sep 19 05:35:00 PDT 1997 Subject: 486 RAM disk --------------17BDB2E9551636B996C8994B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi. I am starting a new project using a I486 system. I need to use a RAM disk. I was able to set up the RAM disk but I am not sure how to access it. Here is the code I used to create the RAM disk: STATUS setramdrive(int size) { BLK_DEV *pBlkDev; DOS_VOL_DESC *pVolDesc; if(size == 0) size = 512; pBlkDev = ramDevCreate(0,size,400,400,0) ; if( pBlkDev == NULL ) { printf("Creation of RAM Drive failed\n"); return(ERROR); } pVolDesc = dosFsMkfs("DEV1:", pBlkDev ); if( pVolDesc == NULL ) { printf("Creation of RAM Drive File System failed\n"); return(ERROR); } printf("Creation of RAM disk complete!\n"); return (OK); } This works fine because I can see the device when I issue the "devs" command from the shell. However, I cannot access it. How do I access this drive? Thanks. Jim Bucciferro --------------17BDB2E9551636B996C8994B Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi.
  I am starting a new project using a I486 system.  I need to use a RAM disk.
I was able to set up the RAM disk but I am not sure how to access it.

Here is the code I used to create the RAM disk:
STATUS setramdrive(int size)
{
 BLK_DEV   *pBlkDev;
 DOS_VOL_DESC *pVolDesc;

 if(size == 0)
  size = 512;

 pBlkDev = ramDevCreate(0,size,400,400,0) ;
 
 if( pBlkDev == NULL )
 {
  printf("Creation of RAM Drive failed\n");
  return(ERROR);
 }

 pVolDesc = dosFsMkfs("DEV1:", pBlkDev );
 if( pVolDesc == NULL )
 {
  printf("Creation of RAM Drive File System failed\n");
  return(ERROR);
 }

 printf("Creation of RAM disk complete!\n");
 return (OK);
}

This works fine because I can see the device when I issue the "devs" command from the shell.

However, I cannot access it.
How do I access this drive?

Thanks.
Jim Bucciferro --------------17BDB2E9551636B996C8994B-- From blohm@vs.dasa.de Fri Sep 19 07:04:30 1997 From: "Gundula Blohm" Date: Fri Sep 19 07:04:33 PDT 1997 Subject: RE: windShell, was-> RE: Problems upgrading to Tornado On 18 Sep 97 at 14:55, the vxWorks Users Group Explo wrote: > > What do you mean by "Ram Disks Don't work", we have been using a 2meg > > ram disk in our system for quite some time, we use it as a place to > > FTP a file when we update the FLASH memory. > > > > So sorry for the confusion. Ram Disks *do* work. You just > > can't "cd" to them and then "ls" from the WindShell. That's > > what I meant when I said that. The whole issue stems from the > > WindShell running locally on the host and the "ls" command on > > WindShell executing locally. For instance, connect to the > > target via WindShell and do an "ls". That works. Now, > > shut your target off and do an "ls". It *still* works! The > > only time the WindShell will detect that the target has gone > > away is if you execute a command that *forces* the WindShell to > > talk to the target. You can force the shell to give the command to the target by typing @command e.g. @ls > > Also, the WindShell only has a subset of > > the commands from usrLib.c. E.g., printErrno runs from the > > WindShell, but try to load a piece of code that calls printErrno > > and you'll get an invalid symbol error for printErrno! It just depends on how you configured your BSP. I think it s INCLUDE_SHOW_ROUTINES. regards, dula =================================================== Gundula Blohm Tel. +49 (731) 392-4731 Fax +49 (731) 392-4958 mailto:blohm@vs.dasa.de =================================================== From dleviner@HiWAAY.net Fri Sep 19 07:06:28 1997 From: David Leviner Date: Fri Sep 19 07:06:31 PDT 1997 Subject: scsiLib interface for scsi adapter driver??? --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit VxWorks Gurus, I'm writing a driver for a scsi host bus vme adapter card. This scsi initiator board has some board specific functionality for acquiring status/identification data. It also has scsi command "pass through" capability for interfacing to disk arrays. I'm trying to find out how to interface to the scsi library in order to do scsi transactions under the "pass through" functionality of the adapter board. I need some help with the scsiLib's data structures in order to use them correctly in my xxScsiTransact() function. I've talked to Tech Support (Vinh Du - TSR #47581) at Wind River and they sent me some top level info from the device driver training notes. This info was somewhat vague and generalized. I asked Vinh Du if there were any source code templates or example code for writing this type of driver. He responded that there is but it is not available to me - WindRiver proprietary. Does anyone have any sample code or even template code for this type of driver? If not, does anyone know the details of the scsiLib data structure interface. Thanks, David -- ____________________________ | | | David Leviner | | Real-Time Systems Engineer | | Chandler/May, Inc. (CMI) | | 125 West Park Loop | | Huntsville, AL 35806 | | phone: (205)722-0175 | /) fax: (205)722-0144 (\ / ) email: dleviner@HiWAAY.net ( \ _( (|____________________________) ) /> (((\ \) / ) / ) / //))/ (\\\\ \_/ / \ \_/ ///// \ / \ / \ _/ \_ / ----/ /-------------------------\ \---- / / \ \ --MimeMultipartBoundary-- From tkb@mclean.sparta.com Fri Sep 19 07:15:58 1997 From: "Thomas K. Buchanan" Date: Fri Sep 19 07:16:00 PDT 1997 Subject: Re: 486 RAM disk --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" > I am starting a new project using a I486 system. I need to use a RAM >disk. >I was able to set up the RAM disk but I am not sure how to access it. > You will be able to access it programatically. You cannot access it from the WindShell command line. Tornado intercepts all file system related commands and executes some of them on the development host. The last time I tried it, Tornado could not tell that RAM disk commands had to be executed on the target. adios -------------------------------------------- /\ Thomas Keith Buchanan /##\ Principal Engineer /####\ /####/-- /####/---- SPARTA, Inc. -\####\__--- Computer Products Division ---\##\ /---- 7926 Jones Branch Drive ----\##\/\---- Suite 900 ----\####\-- McLean, VA 22102 ----\####/ ---/###/ Voice: 703 448 0210 -/###/ FAX: 703 893 5494 \##/ WWW: www.mclean.sparta.com \/ --MimeMultipartBoundary-- From karstens@wg.com Fri Sep 19 09:59:35 1997 From: Alexander Karstens Date: Fri Sep 19 09:59:38 PDT 1997 Subject: Re: 486 RAM disk Hi, I am guessing, that you can't read or write files. I ran into the same problem and I had to start the dos filesystem. Hope that helps. From mmcquade@pop.aa.net Fri Sep 19 10:04:33 1997 From: Mike McQuade Date: Fri Sep 19 10:04:37 PDT 1997 Subject: Re: 486 RAM disk >> here is how you do it. --MimeMultipartBoundary Content-Type: text/plain; charset="us-ascii" At 05:35 AM 9/19/97 PDT, Jim Bucciferro wrote: >I was able to set up the RAM disk but I am not sure how to access it. > We use a Ram disk on a PowerPC (mv2603) and we access it from the Windshell as follows. @cd "/ramdisk/" @ls Here is how you do it, this comes from a TSR that I had with WRS, this is a "workaround" that we have found to work very well in our application. Include the following at the end of usrConfig.c #ifdef INCLUDE_USRLIB #include "usrLib.h" /*************************************************************************** **** * * usrLibInit - pull in usrLib.o finctions * * This dummy routine references cd () in usrLib.c so that all usrLib * functions can be pulled into vxWorks. */ void usrLibInit (void) { cd (""); /* never actually called */ } #endif /* INCLUDE_USRLIB */ Now define INCLUDE_USRLIB in config.h This will allow you to do the following from windsh: -> @ls "DISKA:" -> @cd "DISKA: -> @pwd etc The "@" sign is important, as it passes the command to the target instead of interpreting it on the host side. --MimeMultipartBoundary-- From blohm@vs.dasa.de Fri Sep 19 10:30:55 1997 From: "Gundula Blohm" Date: Fri Sep 19 10:30:59 PDT 1997 Subject: Re: 486 RAM disk > > I am starting a new project using a I486 system. I need to use a RAM > >disk. > >I was able to set up the RAM disk but I am not sure how to access it. > > > > You will be able to access it programatically. You cannot access it from the > WindShell command line. Tornado intercepts all file system related commands > and executes some of them on the development host. The last time I tried it, > Tornado could not tell that RAM disk commands had to be executed on the > target. You can access it from WindShell, see my last message :-) (@command) Dula =================================================== Gundula Blohm Tel. +49 (731) 392-4731 Fax +49 (731) 392-4958 mailto:blohm@vs.dasa.de =================================================== From rtp.co.uk!ihw@rtp.co.uk Fri Sep 19 10:31:34 1997 From: Ian Willats Date: Fri Sep 19 10:31:37 PDT 1997 Subject: scsiLib interface for scsi adapter driver??? David Leviner asked: > Does anyone have > any sample code or even template code for this type of driver? If not, > does anyone know the details of the scsiLib data structure interface. I have written several drivers for both the `old' (pre-Tornado) and current scsiLib interfaces, and I'm very familiar with what goes on in and around the data structures used by scsiLib and the drivers underneath it. I can't supply any source code unless you purchase a source licence for scsiLib and/or a SCSI driver from WRS. However, if you'd like to call or e-mail me with your queries I'd be happy to help. BTW I will be out of the office Tuesday and Wednesday next week. Regards, Ian ---------------------------------------------------------------- Ian Willats E-mail: ihw@rtp.co.uk Real-Time Products Ltd. Unit 8, Holt Court, Telephone: +44 121 333 6955 Aston Science Park, Fax: +44 121 333 5433 Jennens Road, Birmingham B7 4EJ. England. Web: www.rtp.co.uk ---------------------------------------------------------------- From nrd1jrb@nrd.ups.com Fri Sep 19 10:51:33 1997 From: Jim Bucciferro Date: Fri Sep 19 10:51:36 PDT 1997 Subject: Re: 486 RAM disk --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit the vxWorks Users Group Exploder wrote: > > I am starting a new project using a I486 system. I need to use a > RAM > >disk. > >I was able to set up the RAM disk but I am not sure how to access it. > > > > > You will be able to access it programatically. You cannot access it > from > the WindShell command line. Tornado intercepts all file system > related > commands and executes some of them on the development host. The last > time > I tried it, Tornado could not tell that RAM disk commands had to be > executed on the target. > > adios Is it better to use the raw I/O, open(),creat(),read(),write() for a RAM disk or can I use the stdio functions - fopen(), fread(), fwrite()? Thanks. JB --MimeMultipartBoundary-- From sbatchu@holontech.com Fri Sep 19 13:12:19 1997 From: Suresh Batchu Date: Fri Sep 19 13:12:22 PDT 1997 Subject: Serial Driver for PC87334 I/O --MimeMultipartBoundary Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I am looking for a serial driver for the PC87334 Super I/O chip ( which is software compatible with PC16550A and PC16450) to work under vxWorks 5.3. Does anyone know where I can find them? I heard that these kind of drivers are available for free. Any help or advice is appreciated. Thanks in advance, Suresh Batchu HolonTech Corporation sbatchu@holontech.com --MimeMultipartBoundary-- From RHOWELLS@racalrec.co.uk Sat Sep 20 01:59:59 1997 From: RICHARD HOWELLS Date: Sat Sep 20 02:00:02 PDT 1997 Subject: C++ class libraries --MimeMultipartBoundary Content-Type: text/plain Content-Disposition: inline Hi, Is anybody out there using third party (commercial) class libraries (other than tools.h) in their application with VxWorks?? Which ones are you using? What about the STL? Is anybody using it in a embedded application & where can you get a version for use with VxWorks from? Any other comments on the use of C++, class libraries, STL welcome. Many thanks, Richard --MimeMultipartBoundary-- From daemon@csg.lbl.gov Sat Sep 20 04:00:33 1997 From: daemon@csg.lbl.gov Date: Sat Sep 20 04:00:39 PDT 1997 Subject: comp.os.vxworks newsdigest Comp.Os.Vxworks Daily Digest Sat Sep 20 04:00:30 PDT 1997 Subject: i2o OSM Support? Subject: Multicast support in VxWorks ------------------------------------------------------- Newsgroups: comp.os.vxworks Subject: i2o OSM Support? Date: Thu, 18 Sep 1997 18:46:37 -0600 From: kevin.russell@tek.com Organization: Tektronix, Inc. Message-ID: <874621491.28862@dejanews.com> Reply-To: kevin.russell@tek.com Has anyone implemented the host side of i2o under vxWorks? Does anyone know of any third party companies that sell libraries etc. for the host side of i2o. Thanks, Kevin Russell kevin.russell@tek.com 503-627-3548 - -------------------==== Posted via Deja News ====----------------------- http://www.dejanews.com/ Search, Read, Post to Usenet --------------------------- Newsgroups: comp.os.vxworks Subject: Multicast support in VxWorks Date: Fri, 19 Sep 1997 10:18:07 GMT From: jmorris@netcom.com (Jeffrey L. Morris) Organization: Netcom On-Line Services Message-ID: Sender: jmorris@netcom19.netcom.com Hello all: I'm using VxWorks with a PC386 target, NE2000 NIC. Does anyone know if the 'ene' driver supports multicast? If so, how do we activate it? How do we set the multicast address(es)? Thanks in advance for any information you can provide! JLM --------------------------- End of New-News digest ********************** From mailhost@aol.com Sat Sep 20 10:30:42 1997 From: mailhost@aol.com Date: Sat Sep 20 10:30:46 PDT 1997 Subject: Your Credit Report 


///////////////////////////////////////////////////////////////////////
                     STOP COLLECTION AGENCIES COLD!
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

If you have received this electronic mailing by mistake, please pardon our intrusion.

Greetings,

Have you or somebody that you know experienced the humiliation of a collection agency calling at work or at home? They always seem to know when somebody is sitting at your desk or visiting your home when they call.

Millions of collection telephone calls and letters are made or sent each week to people who owe money. Those of us who owe money will pay our debts and do not need the help of a collection agency hounding us at our homes or place of business. Consumers have been helpless to stop the harassment of these predators, UNTIL NOW!

Our hard-hitting report 'Stop Collection Agencies Cold (C 1996)' will show you how to put an end to those telephone calls and letters, INSTANTLY! At the same time, the collectors will be forbidden to EVER call or write you again!!

'Stop Collectors Cold (C 1996)' is a brief, concise report that will tell you everything you need to know to stop the harrassment forever and comes to you with a 100% money-back guarantee! This report is normally $49.95, but if you act WITHIN THE NEXT 24 HOURS, you will pay only $18.00 plus $1.95 S&H.
-----------------------------------------------------------------------

ORDER INFORMATION

IMS
5201 Kingston Pike, Suite 6-216
Knoxville, Tennessee 37919

YES, please rush me the special report 'Stop Collectors Cold (C 1996)' to:

Name:___________________________________________________

Address:________________________________________________

City:___________________________________________________

State:____________________  Zip:________________________

Your E-Mail address (Needed for confirmation):__________
///////////////////////////////////////////////////////////////////////

I have enclosed $18.00 + $1.95 shipping and handling ($19.95 TOTAL).

___Personal Check  ___Money Order  ___Cashiers Check

___I am faxing my order NOW! Please send as a bonus the collector letter package (A $9.95 value).
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

			Fax Your Order NOW!

Please tape your signed personal check in the space below and fax to IMS. The 24 hour facsimile number is 1-423-681-6520.

*****************************************************************

FAX: 1-423-681-6520

PLEASE TAPE YOUR CHECK HERE

PLEASE SIGN YOUR CHECK

*****************************************************************

Thank You for your order!






From 04470848@14923.com  Sat Sep 20 17:22:23 1997
From: 04470848@14923.com
Date: Sat Sep 20 17:22:26 PDT 1997
Subject: Your Web Page
Now you can submit your Web Page to over 620 Internet indexes in a few 
short minutes and watch your traffic soar!

The Central Registry has helped thousands of sites like yours greatly
improve the performance of their Web Page.  The more indexes you are
registered in, the more traffic your Web Page is going to receive.
Being registered in "the top" indexes is not enough.  Why not have your
Web Page registered in every index throughout the world that will 
generate extra traffic for you?

The Central Registry is the leader in Web Page registration and promotion.
We are the oldest and most well established service of this type on the
Internet.  We process more Web Page registrations than all of our
competitors combined.  The Central Registry is the only service that
dynamically asks you for the exact same information each index requests,
assuring accurate and thorough registration.  Unlike most services, we do
not exclude indexes with larger input forms and various detailed 
categories.  No other registration service can compare to our value,
professionalism and performance.

Please come visit our web page today at: http://www.CentralRegistry.com
and see what some of our customers have had to say about the overwhelming 
traffic increases they have seen as a result of being registered 
everywhere.  Thank you.


Sincerely,


The Central Registry
http://www.CentralRegistry.com



From daemon@csg.lbl.gov  Mon Sep 22 04:00:24 1997
From: daemon@csg.lbl.gov
Date: Mon Sep 22 04:00:39 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Mon Sep 22 04:00:20 PDT 1997

        Subject: Re: VxWorks Timing

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: Re: VxWorks Timing
Date: Fri, 19 Sep 1997 09:45:30 +0200
From: Sam Michiels 
Organization: Department of Computer Science, K.U.Leuven
Message-ID: <34222D9A.3C1F@cs.kuleuven.ac.be>
References: 

This is a multi-part message in MIME format.

- --------------7A6E8742254
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Charlie Grames wrote:
> 
> Sam,
> 
> I'm not sure what your clock source on a 68360 is, but, typically, the POSIX timing routines use the
> resolution of the system clock, which is, by default, 16.67 ms (60 Hz).  That being the case, it
> would be impossible to get a smaller delta time than 16.67 ms.
> 
> If you want to get finer resolution with your timings, you will need to use the timestamp clock.  I
> don't know whether this is available by default with the BSP for your board.  If it is, and assuming
> your timestamp clock runs as fast as the processor clock (which is doubtful), your tick resolution
> will be 40 ns.
> 
> I don't see anything wrong with your code, assuming you are initializing your semaphores
> elsewhere.  You might want to make a call to clock_getres() to determine the resolution of the
> POSIX clock.  I think I can say with great certainty, however, that Wind River currently supports no
> processor that can complete 1000 calls to semTake in 15 ns.
> 
> Charlie Grames
> The Boeing Company
> (314) 233-1956
> Charles.R.Grames@boeing.com

I don't think so either...
I attached the complete file with this mail... I did initialize the
semaphores, and I read the clock resolution too :
  (Clock resolution = 0 seconds  16666666 nanoseconds)

Is it ok to say then that the result I got was the number of
clock ticks...so when I multiply this by 16.67 msec
(= time/tick), I get the time it took (approximately of course)?

Thanks again for your help!

Sam.

- -- 


- -----------------------------------------------------------------------
 Sam Michiels                                Tel : +32 (0) 16 32.70.68	
 Kuleuven, Dept. Computer Science            Fax : +32 (0) 16 32.79.96
 Celestijnenlaan 200A 03.40
 B-3001 Heverlee			
 e-mail : Sam.Michiels@cs.kuleuven.ac.be
- -----------------------------------------------------------------------

- --------------7A6E8742254
Content-Type: text/plain; charset=us-ascii; name="semtim~1.c"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="semtim~1.c"

/* Timing program....

	 Creation of Mutual Exclusion Semaphores
		Esp. (Non)Inversion Safe, Deletion Safe, Fifo/Priority Queued

	 Taking/Giving of Mutual Exclusion Semaphores

	 Deletion of Semaphores

*/
  
/* includes */
#include 
#include 
#include 
#include 
#include 
#include 
#include 

/****************************************************************************/

const N=1000;
		
SEM_ID SemPrio[ 1000], SemFifo[ 1000], SemInvSafe[ 1000], SemNonInvSafe[ 1000];
/*					 SemDelSafe[ N];*/

/****************************************************************************/

void semCreateTiming( void )
{
		int s;
		struct timespec start, end;

		printf("** Timing of Creation of Mutual Exclusion Semaphores **\n");

		if( clock_gettime(CLOCK_REALTIME, &start) == -1){
		 printf("Error...%d\n", errno);
		 exit(-1);
		}

		for(s=0 ; s< N ; s++){
			SemPrio[s] = semMCreate (SEM_Q_PRIORITY);
		}

		if( clock_gettime(CLOCK_REALTIME, &end) == -1){
		 printf("Error...%d\n", errno);
		 exit(-1);
		}

		printf("Delay to create 1000 Default Mut Excl Semaphore with Priority Queueing: %ld nsecs\n", end.tv_nsec - start.tv_nsec);


/******************************************************************************/

		clock_gettime(CLOCK_REALTIME, &start);
		for(s=0 ; s
Date: Mon Sep 22 05:16:58 PDT 1997
Subject: PPP on a Synchronous HDLC Serial Device
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I am attempting to get PPP to work on a Synchronous HDLC Serial
interface.  Does anyone have any experience doing this?  Any help would
be greatly appreciated.

I am using vxWorks 5.3.1 on a CPU32 CPU.

Regards

Pat

--------------------------------------------------
      /\   Patrick T. Pinkowski, Senior Engineer
     /##\        ptp@mclean.sparta.com
    /####\         703.448.1683 x228
   /####/--
  /####/----    SPARTA, Inc.
 -\####\__---   7926 Jones Branch Drive, Suite 900
---\##\  /----  McLean, Virginia  22102
----\##\/\----  (703)448-0210 (Main)
 ----\####\--   (703)893-5494 (Facsimile)
  ----\####/    http://www.mclean.sparta.com
   ---/###/
    -/###/ SPARTA
     \##/  ~~~~~~
      \/   Pride In Performance


--MimeMultipartBoundary--


From Richard_W_Mattox@ccmail.orl.lmco.com  Mon Sep 22 07:04:59 1997
From: Richard_W_Mattox@ccmail.orl.lmco.com
Date: Mon Sep 22 07:07:45 PDT 1997
Subject: Rational C++ and VxWorks
--MimeMultipartBoundary
Content-type: TEXT/PLAIN
Content-transfer-encoding: 7BIT

Hello fellow VxWorkers,

The new application I am working on is to be written in C++ hosted on an HP 
workstation targeted for the 604e PowerPC.  I am currently evaluating Rational 
Rose/C++ and Apex.  Has anyone developed under this environment with Rational?  
If so, then can you address the pros and cons of using their tools.  If Rational
was not used, then can anyone recommend a set of tools that they did use.  I do 
not want to use the gnu tools that are supplied with VxWorks.


TIA,
Rick Mattox
richard.w.mattox@lmco.com
--MimeMultipartBoundary--


From t_wmoles@qualcomm.com  Mon Sep 22 09:59:22 1997
From: Hugh Molesworth 
Date: Mon Sep 22 09:59:26 PDT 1997
Subject: ld68k -M; List of error numbers
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Anyone kindly point me to a list of the linker error numbers?

We are using a 68060 target from a Sun host VxWorks 5.3. and I can't
find any reference to the error numbers in the WindRiver documentation
set and didn't particularly want to search the GNU sources (our
installation of info does not have the numbers).

My specific problem is that when I enable the -M map option I get a core
dump (about 15Mbytes) with a numeric error message part way through the
link. The output binary is something over 4Mbyte and compiles ok without
the -M option.

Maybe it's just too big ...

Hugh
--MimeMultipartBoundary--


From t_wmoles@qualcomm.com  Mon Sep 22 11:00:37 1997
From: Hugh Molesworth 
Date: Mon Sep 22 11:00:40 PDT 1997
Subject: ld68k -M; List of error numbers
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"

Anyone kindly point me to a list of the linker error numbers?

We are using a 68060 target from a Sun host VxWorks 5.3. and I can't
find any reference to the error numbers in the WindRiver documentation
set and didn't particularly want to search the GNU sources (our
installation of info does not have the numbers).

My specific problem is that when I enable the -M map option I get a core
dump (about 15Mbytes) with a numeric error message part way through the
link. The output binary is something over 4Mbyte and compiles ok without
the -M option.

Maybe it's just too big ...

Hugh

--MimeMultipartBoundary--


From jonggonl@lgis.lg.co.kr  Mon Sep 22 16:33:11 1997
From: =?EUC-KR?B?wMzBvrDv?= 
Date: Mon Sep 22 16:33:15 PDT 1997
Subject: explosion
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit


explosion
>
--MimeMultipartBoundary--


From daemon@csg.lbl.gov  Tue Sep 23 04:01:23 1997
From: daemon@csg.lbl.gov
Date: Tue Sep 23 04:01:27 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Tue Sep 23 04:01:19 PDT 1997

        Subject: What if . . .
        Subject: Re: C++ class libraries
        Subject: FS: MVME167 and MVME177
        Subject: Re: Download vxWorks with 162BUG

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: What if . . .
Date: 16 Sep 1997 04:17:41 GMT
From: goforit@savetrees.com
Organization: The Alliance
Message-ID: <5vl195$nn9$315@news.insnet.com>

Ten years ago, you could have bought 36 shares of Microsoft for 28.01 
per share. That $1,008.36 would be $198,126.00 now. What if you had 
another chance to invest in a company with that much potential?  If you 
leave your email at "http://theprofitcenter.net/stock" you could get 
the chance to consider an Internet investment with tremendous potential. 


---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: C++ class libraries
Date: Mon, 22 Sep 1997 23:02:30 -0400
From: wotsizname@somethingorother.com
Organization: None
Message-ID: <342830b1.3338470@news.tiac.net>
References: 

> Any other comments on the use of C++, class libraries, STL
> welcome.

Am I the only one who cannot get CrossWind to find the right source line(s)
when debugging C++ code under VxSim for Solaris?



---------------------------

Newsgroups: comp.os.vxworks
Subject: FS: MVME167 and MVME177
Date: Mon, 22 Sep 1997 21:48:58 -0500
From: Innovative Research Technologies 
Organization: Netcom
Message-ID: <34272E1A.5418@innovative-research.com>
Reply-To: Marty@innovative-research.com

We currently have the following boards available.  

MFG: MOTOROLA                                                 QTY
- -----------------------------------------------------------------

MVME167   68040 CPU, 33Mhz, 64MB Memory mezz card             02
          4 SIO, 1PIO, Ethernet, SCSI
          DMA and A32/D32/BLT64 VMEbus I/F
          167bug

MVME177   68060 CPU, 50Mhz, 32MB Memory mezz card             02
          4 SIO, 1PIO, Ethernet, SCSI
          DMA and A32/D32/BLT64 VMEbus I/F
          177bug

MVME952   3-slot desktop VME chassis                          01



Still under warranty.  E-mail if interested or for more
information.


Marty Norton
E-mail: Marty@innovative-research.com
INNOVATIVE RESEARCH TECH.



---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Download vxWorks with 162BUG
Date: 23 Sep 97 02:21:30 GMT
From: steved@pacifier.com (Steve Doiel)
Message-ID: <342727aa.0@news.pacifier.com>
References: <01bcc71a$e95e9be0$6ae42299@raymondy>

In article <01bcc71a$e95e9be0$6ae42299@raymondy>, ryeung@earthlink.net says...
>
>I'm trying to download vxWorks to DRAM on our MVME162LC board
>from PC (with Windows NT).  NIOP in 162 Bug is used.  On PC side, I have
>TFTP Daemon running and security disabled (so that authentication is not
>performed).  NIOT is used to set up some of the parameters.
>I'm getting error code of 0500, meaning "TFTP retry count exceeded".
>Turning on tracing, I got the following decoded results (messages not in
>order):
>Unsupported ETHERNET Type
>Unsupported UDP Type
>ARP Request
>ARP Reply
>Unsupported ICMP Type
>
>Can anyone shed some light here?  Am I using the right command here?
>
A couple of things:
  Did you do the NIOT command first in order to set up the server and
client IP addresses?  These do not come from the same place as the ones
used by VxWorks (apparently).
  Did you start up the system clock?  162's are shipped with the clock turned
off in order to save the battery, but you must start them in order for the
ethernet communication to work under 162bug.

I hope this helps,
SteveD


---------------------------

End of New-News digest
**********************


From nrd1jrb@nrd.ups.com  Tue Sep 23 06:44:21 1997
From: Jim Bucciferro 
Date: Tue Sep 23 06:44:24 PDT 1997
Subject: Low level FTP
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I need to use FTP.  However, ftpXfer doesn't work for some remote
servers so I have to resort to using the lower level FTP commands.  I
got the current chain of commands from the VxWorks Manual but I'm not
sure if I need all of them.
Here is my code:

 int ctrlSock;
 int dataSock;
 int status = OK;
...

if( (ctrlSock = ftpHookup( server )) == ERROR )  {...}

 if( ftpLogin( ctrlSock, username, password,"") == ERROR ) {...}

 if( ftpCommand(ctrlSock, "TYPE I", 0,0,0,0,0,0 ) != FTP_COMPLETE )
{...}

 if( ( dataSock = ftpDataConnInit( ctrlSock )) == ERROR ) {...}

 if( ftpCommand( ctrlSock, "RETR %s/%s",serverpath, serverfile,0,0,0,0 )
!= FTP_PRELIM ) {...}

 if( (dataSock = ftpDataConnGet( dataSock )) == ERROR ) {...}
....
If all of these commands pass, then I retrieve the file and copy it to a
ram disk, which works fine.

The question is whether I need all of these commands, and is the order
correct.

Thanks.
Jim Bucciferro
UPS Research & Development



--MimeMultipartBoundary--


From Thomas.Arand@vs.siemens.de  Tue Sep 23 06:57:42 1997
From: "Arand, Thomas OEN BN EF3" 
Date: Tue Sep 23 06:57:45 PDT 1997
Subject: AW: comp.os.vxworks newsdigest


> ----------
> Von: 	vxwexplo@lbl.gov[SMTP:vxwexplo@lbl.gov]
> Gesendet: 	Dienstag, 23. September 1997 13:01
> An: 	vxworks_users@csg.lbl.gov
> Betreff: 	comp.os.vxworks newsdigest
> 
> Submitted-by daemon@csg.lbl.gov  Tue Sep 23 04:01:23 1997
> Submitted-by: daemon@csg.lbl.gov
> 
> Comp.Os.Vxworks Daily Digest    Tue Sep 23 04:01:19 PDT 1997
> 
>         Subject: What if . . .
>         Subject: Re: C++ class libraries
>         Subject: FS: MVME167 and MVME177
>         Subject: Re: Download vxWorks with 162BUG
> 
> -------------------------------------------------------
> 
> Newsgroups: comp.os.vxworks
> Subject: What if . . .
> Date: 16 Sep 1997 04:17:41 GMT
> From: goforit@savetrees.com
> Organization: The Alliance
> Message-ID: <5vl195$nn9$315@news.insnet.com>
> 
> Ten years ago, you could have bought 36 shares of Microsoft for 28.01 
> per share. That $1,008.36 would be $198,126.00 now. What if you had 
> another chance to invest in a company with that much potential?  If
> you 
> leave your email at "http://theprofitcenter.net/stock" you could get 
> the chance to consider an Internet investment with tremendous
> potential. 
> 
> 
> ---------------------------
> 
> Newsgroups: comp.os.vxworks
> Subject: Re: C++ class libraries
> Date: Mon, 22 Sep 1997 23:02:30 -0400
> From: wotsizname@somethingorother.com
> Organization: None
> Message-ID: <342830b1.3338470@news.tiac.net>
> References: 
> 
> > Any other comments on the use of C++, class libraries, STL
> > welcome.
> 
> Am I the only one who cannot get CrossWind to find the right source
> line(s)
> when debugging C++ code under VxSim for Solaris?
> 
> 
No, You aren't. However, as You did not include very much information
(which versions do You use, how do You see the error), I only can guess
that it is the same error as occured here.

I had this problem with Tornado 1.0. In the meantime, I changed to
1.0.1, and everything is allright.

Ciao,
	Thomas Arand


> ---------------------------
> 
> Newsgroups: comp.os.vxworks
> Subject: FS: MVME167 and MVME177
> Date: Mon, 22 Sep 1997 21:48:58 -0500
> From: Innovative Research Technologies 
> Organization: Netcom
> Message-ID: <34272E1A.5418@innovative-research.com>
> Reply-To: Marty@innovative-research.com
> 
> We currently have the following boards available.  
> 
> MFG: MOTOROLA                                                 QTY
> - -----------------------------------------------------------------
> 
> MVME167   68040 CPU, 33Mhz, 64MB Memory mezz card             02
>           4 SIO, 1PIO, Ethernet, SCSI
>           DMA and A32/D32/BLT64 VMEbus I/F
>           167bug
> 
> MVME177   68060 CPU, 50Mhz, 32MB Memory mezz card             02
>           4 SIO, 1PIO, Ethernet, SCSI
>           DMA and A32/D32/BLT64 VMEbus I/F
>           177bug
> 
> MVME952   3-slot desktop VME chassis                          01
> 
> 
> 
> Still under warranty.  E-mail if interested or for more
> information.
> 
> 
> Marty Norton
> E-mail: Marty@innovative-research.com
> INNOVATIVE RESEARCH TECH.
> 
> 
> 
> ---------------------------
> 
> Newsgroups: comp.os.vxworks
> Subject: Re: Download vxWorks with 162BUG
> Date: 23 Sep 97 02:21:30 GMT
> From: steved@pacifier.com (Steve Doiel)
> Message-ID: <342727aa.0@news.pacifier.com>
> References: <01bcc71a$e95e9be0$6ae42299@raymondy>
> 
> In article <01bcc71a$e95e9be0$6ae42299@raymondy>, ryeung@earthlink.net
> says...
> >
> >I'm trying to download vxWorks to DRAM on our MVME162LC board
> >from PC (with Windows NT).  NIOP in 162 Bug is used.  On PC side, I
> have
> >TFTP Daemon running and security disabled (so that authentication is
> not
> >performed).  NIOT is used to set up some of the parameters.
> >I'm getting error code of 0500, meaning "TFTP retry count exceeded".
> >Turning on tracing, I got the following decoded results (messages not
> in
> >order):
> >Unsupported ETHERNET Type
> >Unsupported UDP Type
> >ARP Request
> >ARP Reply
> >Unsupported ICMP Type
> >
> >Can anyone shed some light here?  Am I using the right command here?
> >
> A couple of things:
>   Did you do the NIOT command first in order to set up the server and
> client IP addresses?  These do not come from the same place as the
> ones
> used by VxWorks (apparently).
>   Did you start up the system clock?  162's are shipped with the clock
> turned
> off in order to save the battery, but you must start them in order for
> the
> ethernet communication to work under 162bug.
> 
> I hope this helps,
> SteveD
> 
> 
> ---------------------------
> 
> End of New-News digest
> **********************
> 
> 


From ugema10@bonsai.de  Tue Sep 23 07:42:41 1997
From: Dirk Hilger 
Date: Tue Sep 23 07:42:48 PDT 1997
Subject: starting target from flash
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"

Hallo,

I am using a PEP vsbc32 board (CPU32) with 1MB flash.
How can I write my object code into flash and start the target from there?
How can I write variables into flash and read these variables out of the flash?

--MimeMultipartBoundary--


From tony.mayes@tecra.it  Tue Sep 23 08:12:43 1997
From: tony.mayes@tecra.it
Date: Tue Sep 23 08:12:47 PDT 1997
Subject: Tornado 1.0.1 Crosswind Debugger
--MimeMultipartBoundary
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT

What experiences have people had with the 1.0,1.0.1 releases of 
Tornado on Win95 or NT hosts?

I am particularly interested in the host debugger, crosswind, which 
is giving us a lot of problems which WRS seem not to be able to give 
us solutions on at this time.

We run on NT 4.0 hosts to PC based targets, using ethernet connection 
between host and target. So far we have removed all of the anti virus 
software, rebuilt NT 4.0 from scratch and reinstalled from scratch 
the Tornado tools, oh and we got rid of all the NT 4.0 service packs 
as well. Still however we see lots of lock ups in the debugger.

I am interested to know if we are alone in this problem or not and 
solutions if anyone has been here and overcome it.

Thanks in advance

Tony








Tony Mayes

Email address          tony.mayes@epid.eurotherm.co.uk

Phone 031 975243
Fax   031 977512

Eurotherm SpA, 
Via XXIV Maggio, 22070 Guanzate (CO), Italy

(NOTE: The opinions expressed within are my 
       own and are not necessarily  those 
       of my employer)

--MimeMultipartBoundary--


From jwc@ssd.ray.com  Tue Sep 23 09:25:05 1997
From: Cosgrove 
Date: Tue Sep 23 09:25:08 PDT 1997
Subject: VME Interrupts on a 2604
Hi, fellow "VxWorkers"

I am having trouble connecting the VME interrupts to a 2604 based handler.
I have done this in both the 68k and 1604 PPC families (but both use the VMEchip2)
but in the Universe/Raven chipset, I am having a little trouble.

Could someone offer a code fragment to connect a VME generated interrupt to a 
handler on a 2604(3)??? (including any needed enabling)

I must be missing some crucial because the VMEbus analyser shows the IACK occurred,
but I never get the interrupt...

Thanks

John Cosgrove
Raytheon
(401)842-4167 (desk)
(401)842-3511 (lab)
jwc@ssd.ray.com



From wbrown@csg.lbl.gov  Tue Sep 23 10:05:41 1997
From: wbrown@csg.lbl.gov (Bill Brown)
Date: Tue Sep 23 10:05:44 PDT 1997
Subject: Re: explosion

KAA-BOOM???????

F.Y.I.

As it happens, the vxWorks exploder is manually maintained by a (sub?)human.

There are three ways to read postings to the exploder and news group:

    1) Subscribe to the immediate mailing list.  This will get the messages
       mailed to you "as they happen."  New subscribers are placed on this
       list by default.

    2) Subscribe to the digest.  This will get all of the messages that were
       posted to the list within the last 24 hours mailed to you each morning
       as one message.

    3) Read the news group "comp.os.vxworks".  This contains all of the
       messages that are sent to the mailing list as well as messages posted
       to the news group.  In addition, all messages posted to the news group
       are automagically mailed to the immediate mailing list subscribers once
       each day, and are included in the next edition of the digest.

Messages to the list should be either emailed to vxwexplo@lbl.gov or posted
to the "comp.os.vxworks" news group.  When responding to messages, please
minimize the quoted section of the original message.  Messages which contain
the "trailer" added by the exploder are deleted in an attempt to avoid
"storms" of bounce messages.

Requests for subscriptions, removal from the list, address changes, etc.
should be sent to "vxwexplo-request@lbl.gov".  While requests can be sent
to me directly, sending them to "vxwexplo-request@lbl.gov" insures that
someone will receive and deal with them when I am not available.  Please
be sure to include the address that you used when you subscribed and note
which list (immediate list or digest) you receive when requesting service.

DO NOT SEND ADMINISTRATIVE REQUESTS TO "vxwexplo@lbl.gov."  Since these
request are of little interest to the rest of the world, filters in the
exploder will send these messages to some file that we'll probably forget
to check.  The correct addresses for posting and for administrative matters
are appended to each message mailed by the exploder.

Also - PLEASE - when "bounce" messages start to appear on the list, DON'T
SEND COMPLAINTS TO THE WHOLE LIST!!!  That just makes things worse!  If you
must complain, send a message to me at "vxwexplo-request@lbl.gov".  While
it may not result in my fixing the problem any sooner, at least it will
not make things worse.

Several sites maintain their own "sub-exploders" which redistribute postings
within the supporting orginization.  We do not maintain these lists.

The WRS web-site URL is www.wrs.com

Information on the vxWorks archive is available by sending e-mail to
	netlib@thor.atd.ucar.edu
with the text
	send index
as the body of the message.

The VxWorks exploder archive is located at 'csg.lbl.gov' in '/pub/vxwexplo'.
Anonymous login is permitted.

-bill
"duty mailing-list-grunt of the day"


From campiglia@ocegr.fr  Tue Sep 23 10:46:56 1997
From: Jean-Marie Campiglia 
Date: Tue Sep 23 10:46:59 PDT 1997
Subject: PCI boards experiences on mv26xx series
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I'm about to migrate my application (including PCI boards) from an
ultra604 to mtx604 board.
My software drives PCI boards which controller is an AMCC S5933 used
both in master R/W mode.
Has anybody got an experience of using this PCI controller with a mv260X
board since Host/Pci bridge (Raven asic) is also used on Mtx series.

Thanks,
JM Campiglia

OCE Industries SA
--MimeMultipartBoundary--


From mmcquade@pop.aa.net  Tue Sep 23 12:51:05 1997
From: Mike McQuade 
Date: Tue Sep 23 12:51:09 PDT 1997
Subject: Re: VME Interrupts on a 2604
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"

At 09:25 AM 9/23/97 PDT, jwc@ssd.ray.com  wrote:

>Hi, fellow "VxWorkers"
>
>I am having trouble connecting the VME interrupts to a 2604 based handler.
>I have done this in both the 68k and 1604 PPC families (but both use the
VMEchip2)
>but in the Universe/Raven chipset, I am having a little trouble.
>
>Could someone offer a code fragment to connect a VME generated interrupt
to a 
>handler on a 2604(3)??? (including any needed enabling)
>
>I must be missing some crucial because the VMEbus analyser shows the IACK
occurred,
>but I never get the interrupt...
>

here is an example used on a 2603, the hardware doc for the board tells us
the vector 
put on the VME bus is 0xe3.



Mike.


void my_isr()
{
   *(UINT *)UNIVERSE_LINT_STAT = 
      LONGSWAP (LINT_STAT_MASK | LINT_STAT_VIRQ3);          /*reset
interrupt*/
   
   logMsg("VME Level 3 Interrupt happened...\n");
}


void setup_rup()
{
   if( sysIntEnable(3) == ERROR)                            /* enable VME
level 3 */
      logMsg( "intEnable Failed\n");
   
   if( intConnect(INUM_TO_IVEC (0xe3), my_isr, 0) == ERROR)  	/* connect
ISR to VME Vector */
      logMsg ( "intConnect Failed\n");                          
}







--MimeMultipartBoundary--


From rd1@idm-associates.com  Wed Sep 24 00:39:20 1997
From: rd1@idm-associates.com
Date: Wed Sep 24 00:39:23 PDT 1997
Subject: You'll Never Get Ahead Unless...
Let's face it:  you'll never get ahead financially just working a job for  someone else.  You need to own your own business if you expect to make any  financial headway.  I've been in business for a long time.  While many businesses have good earnings potential, they are laden down with high costs of inventory, shrinking margins, risk factors, employees, overhead and marketing burdens...not to mention the owner's ball and chain..  I went on a search for two years to find the perfect HOME based business.  I finally found it!

 We represent an income opportunity that is unparalleled with anything else  in the USA.  We GIVE AWAY a wireless, two way state-of-the-art voice  technology home security system to homeowners and businesses.  The system is portable so it can be moved with the customer.  It is even available to renters.  NO other security company offers this option.  The customer signs a mandatory monitoring contract with a UL listed company at a competitive rate.  My compensation works similar to the cellular phone industry.  Each time I give away a system, I make at least $300.  Combine this with the ability to do network marketing, no risk, no employees, no overhead.... I've got a gold mine.
 Please take a moment to view our website at
 http://www.idm-associates.com/rd1/freedom.htm
 I promise it will be worth your while.  You, too, can be in business for a very minor investment.. You'll make at least $696 on your first four sales.  I made $1500 in my first two weeks. Your road to financial freedom is only a mouse click away.

 You will be only contact once with this proposal.  If you wish, you may reply this message with "remove" in the subject line.  We will respectfully honor your request.





From rtp.co.uk!jfd@rtp.co.uk  Wed Sep 24 03:25:37 1997
From: Julian Day 
Date: Wed Sep 24 03:25:44 PDT 1997
Subject: Re: VxWorks 5.3.1 FTP Server


-----Original Message-----
Subject:      VxWorks 5.3.1 FTP Server
From:         "Engineering group (1)" 
Date:         1997/09/23
Message-Id:   <875005830.12726.0.nnrp-04.c30b1810@news.demon.co.uk>
Newsgroups:   comp.os.vxworks

 We have code running under VxWorks 5.3 (x86) and I have just rebuilt
evrything for version 5.3.1. All is OK apart from the fact that the FTP
server will not let me log in. The documentation states that there is no
username or password checking but the error I get from the FTP client on the
host indicates that my password is invalid. Any ideas?

[Julian Day]  

FROM SPR 8602:

The reason is that Tornado 1.0.1 (VxWorks 5.3.1) is deliverd with an
ftpdLib.o compiled with INCLUDE_SECURITY enabled. Therefore the ftp deamon
always asks for a valid username and password.


RESOLUTION:

1. In target/config/all/configAll.h

   Place the following defines in the 'Included Software Facilities':

#define INCLUDE_SYM_TBL (or INCLUDE_CONFIGURATION_5_2)
#define INCLUDE_SECURITY

   You can see the default user name ("target") and password ("password")
   under
 
#define LOGIN_USER_NAME ...
#define LOGIN_PASSWORD ...

2. The option INCLUDE_CONFIGURATION_5_2 will give a target shell, plus
   target-based tools, whereas INCLUDE_SYM_TBL simply includes the symbol
   table facilities.

   If you use the INCLUDE_SYM_TBL option, then you will need to make an
   additional change. In the file target/config/all/usrConfig.c, there is
   the following block of code:

    /* initialize interactive shell */

#ifdef  INCLUDE_SHELL
#ifdef  INCLUDE_SECURITY                      /* include shell security */
   if ((sysFlags & SYSFLG_NO_SECURITY) == 0)
      {
      loginInit ();                           /* initialize login table */
      shellLoginInstall (loginPrompt, NULL);  /* install security program */

      /* add additional users here as required */

      loginUserAdd (LOGIN_USER_NAME, LOGIN_PASSWORD);
   }
#endif  /* INCLUDE_SECURITY */
...

You need to copy the code bracketed by INCLUDE_SECURITY to a position
outside the INCLUDE_SHELL brackets.


SPR / PATCH: 8602

KEYWORDS: FTPD, FTP, SECURITY, PASSWORD



From daemon@csg.lbl.gov  Wed Sep 24 04:01:32 1997
From: daemon@csg.lbl.gov
Date: Wed Sep 24 04:01:35 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Wed Sep 24 04:01:29 PDT 1997

        Subject: How to make symbols available to Wind Shell
        Subject: Re: Changing interrupt priorities on mv2604 PowerPC?

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: How to make symbols available to Wind Shell
Date: 23 Sep 1997 03:12:59 GMT
From: "Raymond K.S. Yeung" 
Organization: AdiCom Wireless
Message-ID: <01bcc7ce$e8ffcd40$df60a9ce@host.hooked.net>

I used the startup-script feature to have the 
target-resident shell to load in our applications.
How do I make the symbols accessible to
Windsh (the serial link won't be available to
us after the system comes up)?

Thanks,
Raymond

---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Changing interrupt priorities on mv2604 PowerPC?
Date: Wed, 24 Sep 1997 00:28:12 GMT
From: georg@sgl.ists.ca (Georg Feil)
Organization: Institute for Space and Terrestrial Science
Message-ID: 
References: 
Sender: news@newshub.ists.ca

...After a bit more looking around and suggestions from people here
is what I have come up with on my problem:

I figured out the hanging part can be corrected by also adjusting
the PCI to ISA interrupt routing as I suspected. These are hard-coded(!)
in the sysPciExtIbcInit() routine in sl82565IntrCtl.c:

    pciConfigOutWord (pciBusNo, pciDevNo, pciFuncNo, PCI_CFG_IBC_INTR_ROUTE,
                      0xabef);

I generalized this to the following and everything works:

    pciConfigOutWord (pciBusNo, pciDevNo, pciFuncNo, PCI_CFG_IBC_INTR_ROUTE,
                      (LN_INT_LVL << 12) |
                      (UNIV_INT_LVL <<  8) |
                      (SCSI_INT_LVL <<  4) |
                      0x000f );

The ISA interrupt lines do have an implicit priority, although what I'm after
is priorities along with actual interrupt nesting, and I'm not sure how/if
that's possible. Unfortuntately I was wrong about the priority order, as 
shipped the VxWorks assignments do make the Universe higher priority than
SCSI, although lower than the LAN. I have switched the universe to be higher
priority than the LAN, but with no obvious effect. I think the priority may
simply apply to simultaneous interrupts, not nested ones.

I've realized just what a hack the VxWorks mv2604 BSP is. Simple stuff like
the above should not need to be patched by users. The following comment 
says it all:

sl82565IntrCtl.c:
   INTERNAL
   ** HELP **
   The sysPciExtIbcInit function does not belong in this driver module.
   The excIntConnect function in sysIbcInit does not belong in this module.
   These should be removed and the change in initialization sequence should
   be updated.


No wonder there are so many problems...

Georg
- --
Georg Feil
                                             |  http://www.sgl.ists.ca/~georg
Space Geodynamics Laboratory                 | Email: georg@sgl.ists.ca
Institute for Space and Terrestrial Science  | Phone: (416) 665-5458
4850 Keele St./North York/Ont/Canada/M3J 3K1 |   Fax: (416) 665-1815

---------------------------

End of New-News digest
**********************


From pcrescio@dsrnet.com  Wed Sep 24 11:28:05 1997
From: "Crescioli, Phil" 
Date: Wed Sep 24 11:28:09 PDT 1997
Subject: RE: comp.os.vxworks newsdigest
>Subject: How to make symbols available to Wind Shell
>Date: 23 Sep 1997 03:12:59 GMT
>From: "Raymond K.S. Yeung" 
>Organization: AdiCom Wireless

>I used the startup-script feature to have the 
>target-resident shell to load in our applications.
>How do I make the symbols accessible to
>Windsh (the serial link won't be available to
>us after the system comes up)?

>Thanks,
>Raymond

You can redirect stdio to a windsh by placing the following in a windsh.tcl
file in your .wind dir:
                   if{[shParse {tstz = open ("/vio/0",2,0)}] !=-1} {
                       shParse {vf0 = tstz};
                       shParse {ioGlobalStdSet (0,vf0)};
                       shParse {ioGlobalStdSet (1,vf0)};
                       shParse {ioGlobalStdSet (2,vf0)};
                       shParse {logFdAdd (vf0) ;
                      }
PAC
          


From fenj@symmetrycomm.com  Wed Sep 24 12:38:13 1997
From: Fen Jain 
Date: Wed Sep 24 12:38:16 PDT 1997
Subject: experience on PPP and TCP/IP  
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I would like to get comments from anyone who has used TCP/IP from Wind
River and PPP from CoSystem or TeleNetwork (WindRiver partners) on your
experience in using the packages for Internet access.  We are interested
in kowning if you are able to use it successfully in real product and
how these two compared with PSOS.

Thank you in advance for your help.

Fen Jain
Symmetry Communications Systems

--MimeMultipartBoundary--


From daemon@csg.lbl.gov  Thu Sep 25 04:01:49 1997
From: daemon@csg.lbl.gov
Date: Thu Sep 25 04:01:52 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Thu Sep 25 04:01:45 PDT 1997

        Subject: VxWorks RPC
        Subject: IP broadcast to 255.255.255.255
        Subject: Re: flash memory programming: linker problem
        Subject: Re: How to make symbols available to Wind Shell
        Subject: VxWorks SE contract

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: VxWorks RPC
Date: 24 Sep 1997 17:04:25 GMT
From: wendy@sw.stratus.com (Wendy McNaughton)
Organization: Stratus Computer, Inc.
Message-ID: <60bh6p$f0h@transfer.stratus.com>

Does anyone have any experiences that they can share with regards to using
RPC on VxWorks?  Has anyone attempted to use RPC on VxWorks to talk to an
NT-based RPC client/server?  Is it possible to use the VxWorks NT-based
portmapper as a general purpose portmapper?

- ------------------------------------------------------------------
Wendy McNaughton		| wendy@seussnt.stratus.com              
Stratus Computer		| Wendy_McNaughton@vos.stratus.com
55 Fairbanks Blvd.
Marlboro, MA  01752
(508) 460-2930

---------------------------

Newsgroups: comp.os.vxworks
Subject: IP broadcast to 255.255.255.255
Date: Wed, 24 Sep 1997 10:31:27 -0700
From: Sudhakar Ravi 
Organization: Sonic Systems
Message-ID: <34294E6E.86A29613@sonicsys.com>

Does anyone know why VxWorks is unable to send an ip packet to
255.255.255.255?

The broadcast address on my ethernet port (using ifShow) comes up as
net.255 where net is the network part of my class C address.  If I call
ifBroadcastSet() with 255.255.255.255, I get a -1 error.

I have an application (DHCP Server) that needs to send to the all 255
address and not the subnet broadcast.  I'm using VxWorks 5.3.


---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: flash memory programming: linker problem
Date: Wed, 24 Sep 1997 10:41:55 -0700
From: John Fabiani 
Organization: Wind River Systems, Inc.
Message-ID: <342950E3.FE2E77CE@wrs.com>
References: 

At runtime, dynamically download the object containing the flash code.
This will place the flash code in ram.  This assumes you have at least
a serial port.

George Pontis wrote:
> 
> I am in need of a solution to controlling the link process more finely
> than is usually done with vxWorks.
> 
> In this very small system, there is 512K of flash, and 256K RAM. The
> application runs nicely from ROM, and I normally build the ROMable image
> using the vxWorks.res_rom_nosym target. There is plenty of spare ROM and
> RAM in this configuration. However, the top 64K sector of flash ROM is
> supposed to hold specific calibration data for the product. I have
> written and tested the routines that erase and reprogram this area, but
> cannot execute them from the normal mode of operation.
> 
> The problem is that one cannot be fetching instructions from flash ROM
> while simultaneously programming the same part. A simple solution would
> be to locate the flash routines in RAM, then block interrupts while they
> execute. However, it does not appear to be possible to locate only some
> of the code in RAM. The vxWorks options seem to provide only for 100% in
> RAM, or 100% in ROM. (What a contrast this is to other embedded
> development tools, that provide fine grain control of the compiler-
> generated sections and the linker segment grouping/ordering)
> 
> I talked with tech support this morning, and they could only suggest that
> I check the archives maintained by the user group at csg.lbl.gov, which I
> have done without benefit.
> 
> Does anyone out there in netland have a semi-clean workaround to this
> problem ?

---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: How to make symbols available to Wind Shell
Date: 24 Sep 1997 18:14:20 GMT
From: "Raymond K.S. Yeung" 
Organization: AdiCom Wireless
Message-ID: <01bcc915$fc4667e0$df60a9ce@host.hooked.net>
References: <01bcc7ce$e8ffcd40$df60a9ce@host.hooked.net> <34286ad7.234186@news.clark.net>

I've figured out a way to resolve this.  Target symbol table
seems to be available to telnet (Windsh uses a symbol
table maintained by target server on the host.  So if
application image is loaded without going through tgt
svr, this host symbol table would not have the symbols).


M. Hayman  wrote in article
<34286ad7.234186@news.clark.net>...
> 1) Use Tornado 1.0.1
> 
> OR
> 
> 2) Rather than load your applications from the startup script,
> statically link them to the kernel when you build it.  You can do this
> by either adding them to your Vx Makefile (ADDED_MODULES = ), or by
> defining "added modules" from WindCFG prior to building the kernel.
> By doing this, their symbol information will be built into
> vxWorks.sym, which is available to both the target shell AND a Tornado
> target server (e.g. WindSh).

---------------------------

Newsgroups: comp.os.vxworks
Subject: VxWorks SE contract
Date: 24 Sep 1997 18:56:21 GMT
From: tadp0le@aol.com (Tadp0le)
Organization: AOL http://www.aol.com
Message-ID: <19970924185601.OAA12740@ladder02.news.aol.com>

Hi. My name is Peter Fuller.I have a HOT contract (possibly to perm) position
available in the West Chicago burbs for a premier company. If you are
interested/available please email tadp0le@aol.com/fax(630-505-0983)/or call
888-823-5627.
Thanks!

---------------------------

End of New-News digest
**********************


From Huang.Wu@nrc.ca  Thu Sep 25 06:55:49 1997
From: "Wu, Huang" 
Date: Thu Sep 25 06:55:52 PDT 1997
Subject: send file from vx
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

	send file from vx
	ex. - send usrTime.shar from vx
--MimeMultipartBoundary--


From chenyc@fnal.gov  Thu Sep 25 08:38:22 1997
From: Yenchu Chen 
Date: Thu Sep 25 08:38:25 PDT 1997
Subject: SCSI drivers in VxWorks 5.3
--MimeMultipartBoundary
Content-Type: TEXT/PLAIN; charset=US-ASCII

Hi,

   Does the VxWorks 5.3 supports disconnect feature, i.e. in case I have
multiple tape drives attached to one MVME167, can I send data block
through SCSI bus to one tape drive, disconnect it and go to the next one
before it finishes?

Yen-Chu

--MimeMultipartBoundary--


From rtp.co.uk!ihw@rtp.co.uk  Thu Sep 25 08:53:33 1997
From: Ian Willats 
Date: Thu Sep 25 08:53:36 PDT 1997
Subject: Flash memory programming
George Pontis asked:

> The problem is that one cannot be fetching instructions from flash ROM
> while simultaneously programming the same part. A simple solution would
> be to locate the flash routines in RAM, then block interrupts while they
> execute.
   ...
> Does anyone out there in netland have a semi-clean workaround to this
> problem ?

I had to do exactly this and ended up by having the flash programming 
routines copy themselves on-demand (i.e. the first time they're called) 
into dynamically-allocated RAM and executing from there, with interrupts 
disabled.  Whether this satisfies the `semi-clean' criterion is debatable, 
but it works and does not require the flash programming code to be loaded 
at run-time.

Let me know if you want to see the code I used, George.

Ian
----------------------------------------------------------------
Ian Willats                          E-mail:       ihw@rtp.co.uk
Real-Time Products Ltd.
Unit 8, Holt Court,                  Telephone: +44 121 333 6955
Aston Science Park,                  Fax:       +44 121 333 5433
Jennens Road,
Birmingham B7 4EJ. England.          Web:          www.rtp.co.uk
----------------------------------------------------------------




From Phil_Watson@masirv.com  Thu Sep 25 08:54:42 1997
From: Phil_Watson@masirv.com (Phil Watson)
Date: Thu Sep 25 08:54:49 PDT 1997
Subject: Re: comp.os.vxworks newsdigest
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

------------- Begin Included Message -------------
Date: 24 Sep 1997 17:04:25 GMT
From: wendy@sw.stratus.com (Wendy McNaughton)

Does anyone have any experiences that they can share with regards to using
RPC on VxWorks?  Has anyone attempted to use RPC on VxWorks to talk to an
NT-based RPC client/server?  Is it possible to use the VxWorks NT-based
portmapper as a general purpose portmapper?
------------- End Included Message -------------

We have been using RPC on VxWorks for almost three years now. It works fairly 
well, allowing for its task-specific nature in Vx. We have had some problems 
with using TCP as the transport between two Vx hosts; WRS is aware of the 
problems and had a patch or two. But we just use the UDP transport and avoid the 
question.

We have not used an NT-based server with a Vx box over RPC, but have used SunOS 
and Windows 95 and 3.x. For the latter we used FTP Software's portmapper and RPC 
library.

Any more questions for me, let me know.

PSW
----
Philip Watson, Phil_Watson@masirv.com
Manager, Embedded Software Development
Matsushita Avionics Systems Corp.
15253 Bake Parkway
Irvine, California, 92618  U.S.A.
--MimeMultipartBoundary--


From Huang.Wu@nrc.ca  Thu Sep 25 09:25:19 1997
From: "Wu, Huang" 
Date: Thu Sep 25 09:25:26 PDT 1997
Subject: send string.shar from vx
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

send string.shar from vx
--MimeMultipartBoundary--


From ZINGLEBJ@OCD231.md.essd.northgrum.com  Thu Sep 25 12:15:17 1997
From: "Zingler, Bill J." 
Date: Thu Sep 25 12:15:20 PDT 1997
Subject: Basic VxWorks Ethernet Questions

Hello,

We are late in our project and now need to add a new, separate ethernet   
network to our system to connect our CPU board to a single subsystem.  We   
are using a Radstone CPU-46 with an MXES-1 daughter card in our VMEBUS   
target.  This configuration comes with ethernet on both boards with   
separate connections to the backplane.  However, they share hardware on   
the board.  Since I don't have the schematics, it is not clear exactly   
what is shared.  What we would like to do is to use the MXES ethernet   
(SN) for our download, shut this off, and then use the CPU-46 (SN) during   
execution.  WRS provides documentation for setting up many types of   
networks (SLIP, SM, etc.) but I haven't found anything that describes how   
I would configure a standard ethernet interface (SN) beyond them telling   
me to #INCLUDE_SN, or if VxWorks supports multiple ethernet connections   
of the same type from one VxWorks image.  Also, since the boot SN   
ethernet setup is done by VxWorks as a part of boot, how do I shut down,   
or close this connection?

Thanks for your help,

Joe B.  


From john.cunningham@das.honeywell.com  Thu Sep 25 12:40:05 1997
From: "Cunningham, John S (NM75)" 
Date: Thu Sep 25 12:40:09 PDT 1997
Subject: floating point underflow errors

I am getting floating-point underflow errors on a MVME2603 target using
VxWorks.  I would like to disable the floating-point underflow error.
Will I have to manually edit code to write to the PC603e registers in my
>tasks to achive this or can I disable the underflow error in VxWorks?  Does
anyone know how I might do this in VxWorks?  

Thanks
John


From gretchen@wrs.com  Thu Sep 25 16:51:39 1997
From: Gretchen Maddox 
Date: Thu Sep 25 16:51:42 PDT 1997
Subject: WRS Developers Conference
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable



Wind River Systems Announces Its 1998 Developers Conference,=20
"IQ 1998: Smart Solutions for Smarter Products"
=20
Intel and Motorola to Co-Sponsor- May 11th - 13th 1998, San Francisco, CA=
=20


ALAMEDA, Calif., September 25, 1997 -- Wind River Systems (NSADAQ: WIND),
the leading provider of integrated embedded software systems, today
announced its premiere annual Developers Conference, IQ 1998: Smart
Solutions for Smarter Products(TM), to be held at the Fairmont Hotel, San
Francisco, CA on May 11th - 13th 1998. The conference will provide an
educational and interactive forum for the most influential members of the
embedded development community.=20

Co-sponsored by Intel and Motorola, and focusing on embedded real-time
solutions for software and hardware developers, the conference will open
with keynote addresses by Jerry Fiddler, co-founder and chairman of Wind
River Systems, and senior executives from Intel and Motorola.  In additio=
n
to numerous general sessions, over fifty educational and interactive
breakout session opportunities will be available for attendees, covering =
a
full range of developer topics.  The conference will also feature a speci=
al
San Francisco evening event.=20

Call For Papers
Developers, industry and technology experts in the embedded software
applications community are encourage to submit papers for presentation at
the conference. Topics may include, but are not limited to: Industry
Trends, Peripherals/Computing, Intelligent I/O, Development Tools, and
Communications (telecomm/datacomm). Panel presentations as well as
individual presentations are also accepted.=20

To submit your 100 word abstract, contact Tonia Miller of LKE Productions=
,
by phone at (415) 544-9300, or email at tonia@lke.com.  Submission deadli=
ne
is Friday, October 31st, 1997.

Wind River Systems=20
Wind River Systems, Inc., is the leading provider of integrated software
development tools for real-time embedded applications in the Internet,
telecommunications, datacommunications, office automation, networking,
computer peripherals, medical, automotive, industrial, aerospace, and
multimedia markets.  Through the industry=92s most comprehensive product =
line
and customer support, Wind River enables customers to develop and reuse
application software across a variety of products and platforms, shorteni=
ng
development cycles and improving time-to-market. Incorporated in 1983, Wi=
nd
River is headquartered in Alameda, California with operations in the UK,
France, Germany, Sweden, Italy, Israel, Korea, and Japan and representati=
on
worldwide  Wind River reported revenues of 64 million dollars in fiscal
year 1997, a 45% increase from fiscal year 1996.
					####			=20
								MCL-PR-DEV-97-9
Tornado, VxWorks, Wind River Systems, "IQ 1998: Smart Solutions for Smart=
er
Products" and the Wind River Systems logo are all trademarks or registere=
d
trademarks of Wind River Systems, Inc. All other trademarks are the
properties of their respective owners.


***************************************************************
Gretchen Maddox			 	        (510) 749-2396
Wind River Systems				   Fax: (510) 749-2010
1010 Atlantic Avenue				    http://www.wrs.com
Alameda, CA  94501		   =09
***************************************************************

--MimeMultipartBoundary--


From bob@seaweed.com  Thu Sep 25 19:56:02 1997
From: bob schulman 
Date: Thu Sep 25 19:56:05 PDT 1997
Subject: Re: memory fragmentation
Lori Smith  wrote:

> We are using VxWorks 5.3.1 on a powerpc.  Our system needs to operate
> for hours on end and we are worried about memory fragmentation.  If
> we do a lot of malloc/frees (or C++ new/deletes), we are concerned
> that we will end up with a completely fragmented memory space and
> that we will not be able to function after a while.  Have any of you
> run into this problem?  Do you know if there is any auto-deframenting
> that goes on in the system?  Or something that we can invoke
> periodically to defragment our memory?

Seaweed Systems has a written-from-scratch memLib replacement. It
has all the functionality which Wind River's memLib has...and more!

It has memory leak detection, sanity checks at malloc/realloc/free
time, user-callable memory arena sanity checker, constant-time allocation,
better fragmentation behavior, callbacks for various low-memory
conditions, and more! A real ginzu knife of an app.

Please call or email for information:
	info@seaweed.com
	510.482-3575


From rd1@idm-associates.com  Thu Sep 25 23:15:11 1997
From: rd1@idm-associates.com
Date: Thu Sep 25 23:15:14 PDT 1997
Subject: Crime Pays
Dear Fellow Citizen,

CRIME-Don't say "it will never happen to me." Because that's exactly
what I thought three months ago when my husband caught someone trying to
come through our bedroom window at 2:00 a.m. With myself, two small
girls, and my husband in the home what possibily could he have wanted
from us?  Rape, burglary, murder...THANK GOODNESS we didn't have to find
out.

According to the U.S. Justice Department you have a 1 in 4 chance of
being a victim in your own home.

An ABC affiliate ran a story recently on home security and their
conclusion was that your best defense is a home security system.

You can now receive a state-of-the-art home security system with NO
BASIC EQUIPMENT COST.  And YOU own the system.

Isn't it worth the PROTECTION of YOUR FAMILY?  You bet it is!

The product:  Wireless State-of-the-Art Home Security Systems with
two-way voice.  It is user installable, portable, high quality
technology.  The system is GIVEN away to home owners, business...EVEN apartment
dwellers anywhere in the USA.

There is also a business opportunity available to you.  Suppose we showed you how to make a full-time income at the same time.  We'll show you how to  GIVE AWAY (YES, GIVE AWAY) a FREE Home Security System and make $300 or more in doing so.

Would YOU be interested?

There is no inventory to carry or money to collect.  The customer signs
a monitoring agreement with a UL listed company at a competitive rate.

The opportunity our company represents is unmatched to any other company
in the USA.  Along with earning generous commissions, this
opportunity can also be networked.

To learn how you can receive a home security system please visit our web
site at: http://www.idm-associates.com/rd1/HA/HAa.htm
Ffor more about the business opportunity, please visit our web  
site at: http://www.idm-associates.com/rd1/HA/HAfreedom.htm


///////////////////////////////////////////////////////////////////////////////
This Message was Composed using Extractor Pro Bulk E- Mail Software. If 
you wish to be removed from this advertiser's future mailings, please reply 
with the subject "Remove" and this software will automatically block you 
from their future mailings.
////////////////////////////////////////////////////////////////////////////////



From rtp.co.uk!ihw@rtp.co.uk  Thu Sep 25 23:50:43 1997
From: Ian Willats 
Date: Thu Sep 25 23:50:47 PDT 1997
Subject: SCSI drivers in VxWorks v5.3
Yenchu Chen  asked:

>   Does the VxWorks 5.3 supports disconnect feature, 

Yes, it does provided you INCLUDE_SCSI2 in config.h.

Ian



From ABonfiglio@sligos.es  Fri Sep 26 02:05:35 1997
From: ABonfiglio@sligos.es
Date: Fri Sep 26 02:05:39 PDT 1997
Subject: AMD ELAN SC300
--MimeMultipartBoundary
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Dear all,


Does anybody know if there is a BSP for VxWorks, based in AMD ELAN =
SC300
microcontroller ?

Thanks.




=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=
=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D
=3D   Alberto BONFIGLIO
=3D   SLIGOS Servicios Inform=E1ticos, S.L.

=3D   Paseo de Castellana 141    planta 14

=3D   28046 Madrid - SPAIN

=3D
=3D   Tel.:  +34 1 567.22.52

=3D   Fax.: +34 1 571.49.74

=3D   e-mail: ABonfiglio@sligos.es
=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=
=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D+=3D
--MimeMultipartBoundary--


From daemon@csg.lbl.gov  Fri Sep 26 04:03:17 1997
From: daemon@csg.lbl.gov
Date: Fri Sep 26 04:03:20 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Fri Sep 26 04:03:14 PDT 1997

        Subject: Re: VxWorks RPC
        Subject: Testing...please ignore
        Subject: Re: Z85230 at MVME162
        Subject: Re: Trouble accessing VME from mv2604 PPC

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: Re: VxWorks RPC
Date: Thu, 25 Sep 1997 14:49:49 GMT
From: noblenet@world.std.com (NobleNet Inc.)
Organization: NobleNet, Inc.
Message-ID: 
References: <60bh6p$f0h@transfer.stratus.com>

In article <60bh6p$f0h@transfer.stratus.com>,
Wendy McNaughton  wrote:
>Does anyone have any experiences that they can share with regards to using
>RPC on VxWorks?  Has anyone attempted to use RPC on VxWorks to talk to an
>NT-based RPC client/server?  Is it possible to use the VxWorks NT-based
>portmapper as a general purpose portmapper?


NobleNet sells ONC/RPC tools that support NT/95 and VxWorks platforms.
We support VxWorks Host environments: Solaris, SunOS, HP-UX, ... , and
NT.  VxWorks Target environments: 68k, x86, Sparc, i960, alpha, mips.

Wind River Systems is our customer, and NobleNet is the supplier of the
ONC/RPC for the VxWorks NT Tornado environment.

NobleNet has also developed RPC server execution environment in VxWorks
that can launch server tasks on demand. This functionality can also
launch a server task per client.

The purchase of NobleNet RPC 3.0 allows redistribution of ONC/RPC DLL's
and portmapper executables for NT/95 without runtime fees. These
re-distributable components also include install tools which install
Portmapper as NT/Service or 95/Tray-Device to make your product
installation easy.

Steve Lemmo
CTO, NobleNet Inc.

NobleNet, Inc.
337 Turnpike Road
Southboro, MA  01772
Voice +1 508 460 8222
Fax +1 508 460 3456



---------------------------

Newsgroups: comp.os.vxworks
Subject: Testing...please ignore
Date: Thu, 25 Sep 1997 09:33:48 -0700
From: Greg Nichols 
Organization: Wind River Systems, Inc.
Message-ID: <342A926C.5440@wrs.com>
Reply-To: gregoryn@wrs.com

This is a test...please ignore

---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Z85230 at MVME162
Date: Thu, 25 Sep 1997 11:55:42 -0400
From: dwp 
Organization: theBorg Technology
Message-ID: <342A897E.6CBF@felix.kodak.com>
References: <342A72C8.C83@exp.bessy.de>
Reply-To: dwp 

Chris Hellwig wrote:
> 
> Hi,
> 
> I'm going to use the serial port 2 at a MVME162,
> which is driven by a Zilog 85230.
> 
> Unfortunately the chip driver  z8530Serial.c only
> supports FIOSETBAUDRATE in tyCoIoctl(), so I'm not
> able to set parity, char-lenght and stopbits via
> ioctl(). Is somebody out there who modified the
> z8530Serial.c to support these things.
> 
> What is the way to make use of the interrupt capabilitys
> of the chip???? I would like to have an interrupt if
> CR appears at the port to take the whole string from
> input buffer.

Here's all the hints you need.  Have fun!
dave

	serState.pChan = (I8250_CHAN*)sysSerialChanGet(SERIAL_PORT_ONE);

	if (!serState.pChan)
		return(FALSE);

	serState.fd = open("/tyCo/1", O_RDWR, 0);

	if (!serState.fd)
		return(FALSE);


/*	set baudrate */
	if (ioctl(serState.fd, FIOBAUDRATE, SIP_BAUD_RATE) == ERROR)
	{
		close(serState.fd);
		return(FALSE);
	}

/*	flush the buffers */
	if (ioctl(serState.fd, FIOFLUSH, 0) == ERROR)
	{
		close(serState.fd);
		return(FALSE);
	}

/*	set serial I/O to raw character mode  */
	if (ioctl(serState.fd, FIOSETOPTIONS, OPT_RAW) == ERROR)
	{
		close(serState.fd);
		return(FALSE);
	}

/*	input buffer callback function */
	if (ioctl(serState.fd, FIOPROTOHOOK, 
				(int)((FUNCPTR)cb_getData)) == ERROR)
	{
		close(serState.fd);
		return(FALSE);
	}
 
/*	callback function passed parameter */
	if (ioctl(serState.fd, FIOPROTOARG, 0) == ERROR)
	{
		close(serState.fd);
		return(FALSE);
	}

/* 
	Set the serial port to:
   	8 data bits, Even parity, 1 Stop bit, 
*/

	(*serState.pChan->outByte) (serState.pChan->lcr, 
				SIP_I8250_LCR_WRD8 | 
				SIP_I8250_LCR_SB1 |
				SIP_I8250_LCR_PEV);

---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Trouble accessing VME from mv2604 PPC
Date: 12 Sep 1997 06:02:40 GMT
From: caoy@aol.com (Caoy)
Organization: AOL http://www.aol.com
Message-ID: <19970912060200.CAA01381@ladder01.news.aol.com>
References: 


Hi,  I am new to VME as well as to VxWorks. 

I don't have the answer for this question, but I do have some related
questions. 

We have a MVME2604 board and a I/O board with dual port RAM 
which we need to map to CPU's VME A24 memory window. How am I supposed to it??

What is the difference in term of the purpose between mv2600.h
and confgi.h?

What is the difference between the Options: ???
config.h: VME_A24_MSTR_BUS
and
mv2600.h: VME_A24_MSTR_LOCAL

Any response is highly appreciated!!

Yipu


---------------------------

End of New-News digest
**********************


From nguyend2@ttc.com  Fri Sep 26 05:55:53 1997
From: "Dzung Nguyen" 
Date: Fri Sep 26 05:55:56 PDT 1997
Subject: Custom Driver
     
     
     
     Hi,
     
     Could someone be so kind to tell me how to attach a custom driver to 
     the TCP/IP stach in Tornado?  In Version 5.1 there is a function 
     "if_attach", but this function seems to disappear in 5.3.x.
     
     Your email response is greatly appreciated.
     
     Thank you very much in advance.  
     
     Dzung



From dbruce@targettech.com  Fri Sep 26 06:34:17 1997
From: David Bruce 
Date: Fri Sep 26 06:34:20 PDT 1997
Subject: Re: Crime Pays
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

remove
-- 
David R. Bruce                      phone: (508) 657-9940
Senior Software Engineer            fax  : (508) 694-1485
Target Technologies, Inc.           email: dbruce@targettech.com
                                    WWW  : http://www.targettech.com
--MimeMultipartBoundary--


From dbruce@targettech.com  Fri Sep 26 06:34:22 1997
From: David Bruce 
Date: Fri Sep 26 06:34:25 PDT 1997
Subject: remove
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

-- 
David R. Bruce                      phone: (508) 657-9940
Senior Software Engineer            fax  : (508) 694-1485
Target Technologies, Inc.           email: dbruce@targettech.com
                                    WWW  : http://www.targettech.com
--MimeMultipartBoundary--


From jwc@ssd.ray.com  Fri Sep 26 07:32:43 1997
From: Cosgrove 
Date: Fri Sep 26 07:32:46 PDT 1997
Subject: ioTaskStdSet( )
Hi, Folks,

Does ioTaskStdSet() act as a "tee" or should it move all I/O?
It my problem the fact I am running this thru windsh?

My ~/.wind/windsh.tcl contains:

        if { [shParse {tstz = open ("/vio/0",2,0)}] != -1 } {
              shParse {vf0 = tstz};
              shParse {ioTaskStdSet(0x17fc1a0, 0, vf0)};
              shParse {ioTaskStdSet(0x17fc1a0, 1, vf0)};
              shParse {ioTaskStdSet(0x17fc1a0, 2, vf0)};
              shParse {logFdAdd (vf0)};
              shParse {printf("Std I/O redirected for tLogTask\n")};
	}

where 0x17fc1a0 is the TID for tLogMsg...

Any help is greatly appreciated...

John Cosgrove
Raytheon
(401)842-4167 (desk)
(401)842-3511 (lab)
jwc@ssd.ray.com

 


From lspadafora@algw2.lucent.com  Fri Sep 26 07:59:20 1997
From: Louis Spadafora 
Date: Fri Sep 26 07:59:23 PDT 1997
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

Please remove me from the user group.

Thank you
Louis Spadafora

--MimeMultipartBoundary--


From lars.antback@pullmax.se  Fri Sep 26 09:46:51 1997
From: lars.antback@pullmax.se
Date: Fri Sep 26 09:46:54 PDT 1997
Subject: RE: floating point underflow errors


Question submitted by "Cunningham, John S (NM75)"   


As far as I have discovered you can not set some global initialization   
data for items like the
floating point registers in VxWorks. We ran in to a similar problem and   
had to add code for each
task to disable underflow and overflow errors. There is a couple of   
usefull functions and macros
defined in /h/arch/ppc/archPpc.h and /h/arch/ppc/vxPpcLib.h that   
simpifies the code writing.

>I am getting floating-point underflow errors on a MVME2603 target using
>VxWorks.  I would like to disable the floating-point underflow error.
>Will I have to manually edit code to write to the PC603e registers in my
>tasks to achive this or can I disable the underflow error in VxWorks?   
 Does
>anyone know how I might do this in VxWorks?
>
>Thanks
>John

/Ant

                              ////
                             (0 0)
 ________________________o00o_( )_o00o___________________________
/                              |                                 \
| Lars Antb„ck                 | tel: + 46 (0)31 57 07 00        |
| Pullmax AB                   | email: ant@pullmax.se           |
\________________________________________________________________/



From cgrames@mdc.com  Fri Sep 26 13:53:21 1997
From: Charlie Grames 
Date: Fri Sep 26 13:53:24 PDT 1997
Subject: floating point underflow errors -Reply
--MimeMultipartBoundary
Content-Type: text/plain
Content-Disposition: inline

John,

We ran into the same problem.  You can write your own floating-point initialization routine and
have it run at the creation of each new task by setting the fppCreateHookRtn global variable to that
routine.  Following is a short example of how you might do that:


#include "usrConfig.h"


LOCAL void usrFppCreateHook(FAST WIND_TCB *pTcb)
{

/*
**  Modify the the floating-point status and control register in the floating-
**  point context of the new task.
*/
    pTcb->pFpContext->fpcsr &= ~_PPC_FPSCR_UE;
}


void usrStartup(void)
{
    .
    .
    .
/*
**  Install a hook to modify the floating-point environment of all new tasks.
*/
    fppCreateHookRtn = (FUNCPTR)usrFppCreateHook;
    .
    .
    .
}

Hope this helps.

Charlie Grames
The Boeing Company
(314) 233-1956
Charles.R.Grames@boeing.com


Submitted-by john.cunningham@das.honeywell.com  Thu Sep 25 12:40:05 1997
Submitted-by: "Cunningham, John S (NM75)" 


I am getting floating-point underflow errors on a MVME2603 target using
VxWorks.  I would like to disable the floating-point underflow error.
Will I have to manually edit code to write to the PC603e registers in my
>tasks to achive this or can I disable the underflow error in VxWorks?  Does
anyone know how I might do this in VxWorks?  

--MimeMultipartBoundary--


From ZINGLEBJ@OCD231.md.essd.northgrum.com  Fri Sep 26 14:33:09 1997
From: "Zingler, Bill J." 
Date: Fri Sep 26 14:33:16 PDT 1997
Subject: Exception Handlers needed for CPU46

Hello out there,

I am looking for some generic exception handlers for use with the   
Radstone CPU-46 board that I could implement in place of the VxWorks   
exception handlers.  Specifically, the exception during an exception   
which causes the target to reboot.  In our application this would be   
utter (no, I do not work for Gateway) disaster, if it were ever to occur.   
 The other exception handler I need is for the less disasterous   
exceptions (bus error, etc) that causes VxWorks to suspend the offending   
Task.  Searching their literature has turned up what looks like a viable   
solution using setjmp/longjmp functions that need to be embedded into   
each of the Tasks.  Has anyone used this to try and keep things limping   
along after such a failure or is there a better way?

Thanks for the help.

Joe Blodgett  


From chip@Fiberlane.com  Fri Sep 26 17:57:37 1997
From: "Chip.Roberson" 
Date: Fri Sep 26 17:57:40 PDT 1997
Subject: WANTED: RTOS/firmware engineer
This is a cryptographically signed message in MIME format.

--------------ms544B16D6E99438E8F4EF53CF
Content-Type: multipart/mixed; boundary="------------E728B252CF31D27FC1197E76"

This is a multi-part message in MIME format.
--------------E728B252CF31D27FC1197E76
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

My apologies if this is considered inappropriate for this
list but we are looking explicitly for someone with VxWorks
experience.

OS and Firmware Engineer
------------------------

Position requires experience in specification, design, development and
testing of hardware and low-level software for networking products, mainly
switches, bridges and routers. Must have experience in getting Real-Time
kernels (vxWorks) working on a hardware platform and all required
hardware/software interactions. Must be well conversant with vxWorks,
UNIX/C development environment, GNU tools. Must have experience in
writing device-drivers, and low-level software. Must have sound
knowledge of Operating systems. Must have a BSCS/BSEE and 3-5 years of
experience in real time operating systems/kernels/device drivers/
diagnostics.

The position is located in Petaluma, CA which is 37 miles north of San
Francisco in the Sonoma wine country.  Fiberlane is a venture-funded
start up with handsome equity packages for employees.  If interested,
please send your resume to mailto:staffing@fiberlane.com.

Regards.
--------------E728B252CF31D27FC1197E76
Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Charles Roberson
Content-Disposition: attachment; filename="vcard.vcf"

begin:          vcard
fn:             Charles Roberson
n:              Roberson;Charles
org:            Fiberlane Communications, Inc.
adr:            Suite 200;;1318 Redwood Way;Petaluma;CA;94954;USA
email;internet: chip@fiberlane.com
title:          Manager, Software Development
tel;work:       (707) 793-1942
tel;fax:        (707) 793-9057
tel;home:       (707) 765-1675
x-mozilla-cpt:  ;0
x-mozilla-html: TRUE
version:        2.1
end:            vcard


--------------E728B252CF31D27FC1197E76--

--------------ms544B16D6E99438E8F4EF53CF
Content-Type: application/x-pkcs7-signature; name="smime.p7s"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="smime.p7s"
Content-Description: S/MIME Cryptographic Signature
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--------------ms544B16D6E99438E8F4EF53CF--



From daemon@csg.lbl.gov  Sat Sep 27 04:02:53 1997
From: daemon@csg.lbl.gov
Date: Sat Sep 27 04:02:56 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Sat Sep 27 04:02:49 PDT 1997

        Subject: Re-entrant version of telnetLib
        Subject: Re: Changing interrupt priorities on mv2604 PowerPC?

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: Re-entrant version of telnetLib
Date: Fri, 26 Sep 1997 09:30:40 -0400
From: John Bliss 
Organization: Cabletron Systems Inc.
Message-ID: <342BB900.3654@ctron.com>

Hey Guys,

The product I'm working on has the requirement to support multiple
telnet sessions.  TelnetLib, as provided by WRS, can not support
multiple telnetd sessions.

Has anyone gone through the process of modifying telnetLib to support
multiple sessions?

Any input would be greatly appreciated?

Thanks,

JDB

---------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Changing interrupt priorities on mv2604 PowerPC?
Date: Wed, 24 Sep 1997 06:46:47 GMT
From: mmcquade@aa.net (Mike)
Organization: Alternate Access Inc.
Message-ID: <3428b549.612679203@news.aa.net>
References: 

georg@sgl.ists.ca (Georg Feil) wrote:


>
>We are not using MPIC interrupts right now because it causes too many problems
>(spurious interrupts & crashes) even though we have all the latest patches
>from WindSurf. Therefore I assume the entries in mv2600.h to change priorities
>are:
>


Georg,


We are using MPIC interrupts on a 2603 board, we have not had any
problems with Jumbo patch #2 installed, we are running 8000+ VME
interrupts / second, +Mouse, Keyboard, and SCSI and LAN interrupts.


When we installed Jumbo Patch #3, we couldn't run SCSI any more, so we
had to drop back to jumbo patch #2.

Other than that, we have been running the MPIC interrupts since early
Febuary (before they we included in the BSP), and we have been very
happy with the MPIC.


*** Don't blame the "604" lockups on the MPIC ***


Changing priority levels with the MPIC/RAVEN is very easy, and I would
reccomend changing it from the way it comes in the BSP, with
interrupts set to the same level.



Mike.




---------------------------

End of New-News digest
**********************


From 35574487@aol.com  Sat Sep 27 15:51:01 1997
From: 35574487@aol.com
Date: Sat Sep 27 15:51:04 PDT 1997
Subject: Credit Card for you!!!!!


*************************************************************************************************
This is a one time message, please forgive me if you are offended by receiving this---you will automatically be removed if you do not order.  thank you
*************************************************************************************************


NEW CREDIT FILE LEGALLY!!!!

Bad credit? Judgements? Other problems?  THIS IS THE PROGRAM FOR YOU!!!!

Learn how to open a second (or more) SEPERATE credit file.  This unique and virtually unknown program is 100% legal, and can work every time!!!

This is the chance you have been looking for.  You no longer have to be embarrassed about your credit problems----THIS IS YOUR CHANCE FOR A NEW START!!!!----remember that it does not matter what your credit past is.

You will now be able to get:
***CREDIT CARDS
***CAR LOANS
***BANK LOANS
***MORTGAGE LOANS
***RENT AN APARTMENT

This program will even show you how to REPAIR your old credit file, WHILE you enjoy your AAA credit, in as little as 30 days.  

REMEMBER----YOU CAN ONCE AGAIN HAVE GOOD (AAA) CREDIT!!!!

This is a first time promotional experiment on the internet.  Our program normally sells for $79, but act now to get our limited discount offer!!!
ONLY $49 (PLUS $3 FOR S&H)

ACT NOW!!!!!  ONLY A LIMITED NUMBER TO BE SOLD AT THIS PRICE (250) AND ONCE THEY ARE GONE--THAT IS IT!!!!!!

SEND $52 ---CHECK, MONEY ORDER, OR CASH TO:

D W C
P.O. Box 54789
Phoenix, AZ 85078

Be sure to include your:

NAME_________________________________
ADDRESS_____________________________
E-MAIL ADDRESS_______________________
PHONE NUBER_________________________

If you would like to learn how to make extra income write "I WANT TO EARN AN EXTRA $50,000 PER YEAR" on your order

 






From altan@venus.aselsan.com.tr  Mon Sep 29 05:52:53 1997
From: Cagatay ALTAN 
Date: Mon Sep 29 05:53:09 PDT 1997
Subject: Flash memory programming problem
--MimeMultipartBoundary
Content-Type: TEXT/PLAIN; charset=US-ASCII

  Dear All,
  I am trying to program the flash memory of an mvme162, but I receive an
access fault every time it begins to write.
  Before starting to write to the flash, I read the address 0xfffcc000 to
a dummy variable which I hope gives me the write access to the flash, but
it does not. Is there another thing to do to provide a write access?
  I wonder if someone can find what I'm doing wrong or canv provide me
with a C routine that performs this.
  Thanks to all.
						Cagatay Altan,
						ASELSAN

--MimeMultipartBoundary--


From raman@net.com  Mon Sep 29 09:38:25 1997
From: raman@net.com (Raj Raman)
Date: Mon Sep 29 09:38:28 PDT 1997
Subject: Re: comp.os.vxworks newsdigest
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


We've made changes to the "telnetLib" and "rlogLib" source code
to ensure we could have multple concurrent sessions.

The problem (Im sure you've figure out too) is the target-resident
vxWorks shell-paser which is NOT reentrant. Since the telnet session
brings up the shell, there could be only one session at a time.

We've modified the source code to "listen/accept" connections and
spawn a task associated with each connection that handshakes with
an "User Interface" task that has the menu options. In gist, we've
modified code to have a session-specific task be the server-side
Interface task (as opposed to the Shell). 

Feel freee to email me if you have any more questions.

- raj 

------------------------------------------------------------------------

email: raman@net.com
-------------------------------------------------------------------------

% Hey Guys,
% 
% The product I'm working on has the requirement to support multiple
% telnet sessions.  TelnetLib, as provided by WRS, can not support
% multiple telnetd sessions.
% 
% Has anyone gone through the process of modifying telnetLib to support
% multiple sessions?
% 
% Any input would be greatly appreciated?
% 
% Thanks,
% 
% JDB
% 
%
--MimeMultipartBoundary--


From stan@rti.com  Mon Sep 29 12:36:55 1997
From: stan@rti.com (Stan Schneider)
Date: Mon Sep 29 12:36:58 PDT 1997
Subject: Re: Re-entrant version of telnetLib

>> Date: Fri, 26 Sep 1997 09:30:40 -0400
>> From: John Bliss 

>> Has anyone gone through the process of modifying telnetLib to support
>> multiple sessions?

Hi John,

We have a simple reentrant shell called "rshell" that ships with RTILib.  It
will support multiple connections, although you do it through a host terminal
program, not telnet.  See www.rti.com for details.

        -- Stan


=============================================================================
=                                           =                               =
=   Stan Schneider                          =   email: stan@rti.com         =
=   Real-Time Innovations, Inc.             =   Phone: (408) 720-8312 x104  =
=   155A Moffett Park Drive, Suite 111      =   Fax:   (408) 734-5009       =
=   Sunnyvale, CA 94089                     =   http://www.rti.com          =
=                                           =                               =
=============================================================================



From pbrett@hathaway-systems.com  Tue Sep 30 03:29:16 1997
From: pbrett@hathaway-systems.com (Pete Brett)
Date: Tue Sep 30 03:29:20 PDT 1997
Subject: Re: experience on PPP and TCP/IP  
--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable

On Wed, 24 Sep 97 12:38:18 PDT, you wrote:

>Submitted-by fenj@symmetrycomm.com  Wed Sep 24 12:38:13 1997
>Submitted-by: Fen Jain 
>
>--MimeMultipartBoundary
>Content-Type: text/plain; charset=3Dus-ascii
>Content-Transfer-Encoding: 7bit
>
>I would like to get comments from anyone who has used TCP/IP from Wind
>River and PPP from CoSystem or TeleNetwork (WindRiver partners) on your
>experience in using the packages for Internet access.  We are interested
>in kowning if you are able to use it successfully in real product and
>how these two compared with PSOS.

I have used PPP to connect not to the Internet but to a small internet
comprising up to 16 data acquisition systems running vxWorks on a LAN.
These machines send data to a 32-bit Windows machine for processing
and viewing.
With a patch applied to LCP.C, the vxWorks implementation appears to
connect to Win95 or Unix machines with no problems, and maintain a
solid connection.
Unfortunately the same cannot be said of Windows NT machines. The
problem is that Microsoft use two new authentication methods (CHAP80
which is also known as MSCHAP and Shiva PAP). When vxWorks attempts to
make a connection to Windows NT, the NT machine tries to negotiate for
these two authentication methods, even if the "use any authentication
method including clear text" option is selected. The vxWorks PPP
implementation can authenticate using CHAP MD5 or PAP but not CHAP80
or SPAP. Ideally, vxWorks should send a request for CHAP or PAP with
each configure NAK sent in response to requests for CHAP80 or SPAP,
but instead it sends only one configure request for an authentication
method near the start of negotiation. The result is that the
negotiation does not converge and the session is terminated.
This problem does not occur if the NT machine initiates the connection
to vxWorks.
I have reported the problem to Microsoft and WindRiver, but so far
have not got a solution.
I have not used PSOS and so cannot compare it with vxWorks.

Regards,
	Pete
--MimeMultipartBoundary--


From daemon@csg.lbl.gov  Tue Sep 30 04:01:10 1997
From: daemon@csg.lbl.gov
Date: Tue Sep 30 04:01:14 PDT 1997
Subject: comp.os.vxworks newsdigest
Comp.Os.Vxworks Daily Digest    Tue Sep 30 04:01:07 PDT 1997

        Subject: Re: Big endian and Little endian( PC to PPC)
        Subject: Help on C++ cout

-------------------------------------------------------

Newsgroups: comp.os.vxworks
Subject: Re: Big endian and Little endian( PC to PPC)
Date: 19 Sep 1997 12:50:31 -0600
From: bgeer@xmission.xmission.com (bgeer)
Organization: XMission Internet (801 539 0900)
Message-ID: <5vuhhn$3ii$1@xmission.xmission.com>
References: <19970913215400.RAA10530@ladder02.news.aol.com>

caoy@aol.com (Caoy) writes:

 >We need send a data structure down to the target through TCP/IP from the
 >PC host. On the PC side data format is little endian while big endian on
 >the target 
 >side.

 >My question:

 >1) If we send the data structure as normal from PC side,  what does the
 >byte order of the structures look like?

Assuming you are using a network driver, the payload (data) will be in
the wrong byte order for the receiving processor.  Check out "man
htonl" family of functions for host-to-network byte order conversion.

I'm not sure about such things as socket headers.  An application must
offer up a sockaddr struct when calling socket().  I assume the app
would have to "htonl" (or whatever) any data it puts in the struct.

I've been sloppy about this, having only worked on big endian systems
to-date.  If (when?) I need to port my apps to a little endian
machine, I'm sure I'll learn the answer...:-)

Cheers, Bob
- -- 
<> Robert Geer & Donna Tomky /              *             <>
<>    bgeer@xmission.com    |    _o      *   o *      o   <>
<>   Salt Lake City, Utah   |   -\<,      * <\      
<>          U S A           |   O/ O     __ /__,    />    <>

---------------------------

Newsgroups: comp.os.vxworks
Subject: Help on C++ cout
Date: 29 Sep 1997 22:54:06 GMT
From: "Raymond K.S. Yeung" 
Organization: AdiCom Wireless
Message-ID: <01bccd2a$ed562ea0$df60a9ce@host.hooked.net>

This was originally "C++ cout output disappeared."
I'm providing a code sample here and combination
of things that I tried.  One (desired) combination
led to the problem, while the other two has no
problem.  I hope someone can shed some light
here.  Thanks,
Raymond

C++ Code
========
void Dbg (int dbg_lvl, char * str1, int data1, int data2, int data3) {
   if (DEBUG_LEVEL >= dbg_lvl) {
      if (semTake (output_sema, sema_tmo) == ERROR)
         return;

      cout << "[" << taskName ( taskIdSelf () ) << "]: ";
      cout << str1 << " " << data1 << " " << data2 << " " << data3 << endl;

      semGive (output_sema);
   }
}
void CP_DISPATCHER_CLASS::ShowCpTaskStat (void) {
   int  inst_id;

   for (inst_id=0; inst_id < CP_MAX_NUM_OF_CALLS; inst_id++) {
      Dbg (LOW_DBG_LVL, "Inst ", inst_id, (int) cp_task [inst_id].state,
(int) cp_task [inst_id].call_id);
   }
}

Expected output is, for instance,   [CpDispatch]: Inst 1 0 45
Instead, actual output is,               1045
That is, all strings are missing, including the white space.

Combinations to reproduce the problem
=============================
I've tried 3 combinations, one of them (the desired combination) has the
problem:

#1. Set CONSOLE_TTY=NONE and INCLUDE_STARTUP_SCRIPT   => problem
#2. Set CONSOLE_TTY=0         and don't INCLUDE_STARTUP_SCRIPT => no
problem
#3. Set CONSOLE_TTY=NONE and don't INCLUDE_STARTUP_SCRIPT => no problem

When STARTUP script is used, C/C++ application is brought up vi script by
target shell.
When STARTUP script is not used, C/C++ application is brought up manually
via WindSh.


---------------------------

End of New-News digest
**********************


From naivar@telomere.lanl.gov  Tue Sep 30 07:58:30 1997
From: Mark Naivar 
Date: Tue Sep 30 07:58:33 PDT 1997
Subject: Re: Flash memory programming problem
--MimeMultipartBoundary
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Mon, 29 Sep 1997, the vxWorks Users Group Exploder wrote:

> Submitted-by altan@venus.aselsan.com.tr  Mon Sep 29 05:52:53 1997
> Submitted-by: Cagatay ALTAN 
> 
> --MimeMultipartBoundary
> Content-Type: TEXT/PLAIN; charset=US-ASCII
> 
>   Dear All,
>   I am trying to program the flash memory of an mvme162, but I receive an
> access fault every time it begins to write.
>   Before starting to write to the flash, I read the address 0xfffcc000 to
> a dummy variable which I hope gives me the write access to the flash, but
> it does not. Is there another thing to do to provide a write access?
>   I wonder if someone can find what I'm doing wrong or canv provide me
> with a C routine that performs this.
>   Thanks to all.
> 						Cagatay Altan,
> 						ASELSAN
>

You might try looking in $WIND_HOME/target/src/drv/mem, on my system 
there is a file called flashMem.c which has some routines for reading and 
writing to FLASH.  Reading from FLASH can be done with a pointer, but 
writing involves some extra steps.  Hope this helps.

-Mark 

/-------------------------+--------------------------\
| Mark Naivar             | mnaivar@lanl.gov         |
| Los Alamos National Lab |                          |
| LS-5,  MS M888          | (505) 665-2034 (office)  |
| P.O. Box 1663           | (505) 665-6738 (lab)     |
| Los Alamos, NM 87545    | (505) 667-1321 (FAX)     |
\-------------------------+--------------------------/

--MimeMultipartBoundary--


From Eli_Wittner@cti-pet.com  Tue Sep 30 10:33:11 1997
From: "Eli Wittner"
Date: Tue Sep 30 10:33:14 PDT 1997
Subject: Little endian vs Big Endian
--MimeMultipartBoundary
Content-type: text/plain; charset=US-ASCII





>>caoy@aol.com (Caoy) writes:
> >We need send a data structure down to the target through TCP/IP from the
>>PC host. On the PC side data format is little endian while big endian on
>>the target side.
 >>My question:
>>1) If we send the data structure as normal from PC side,  what does the
>>byte order of the structures look like?
I'm not sure if this will help, but we transfer data between a 68040
vxWorks client and a PC running NT4.0.  We use Winsock for the TCP/IP end
on the PC side and the (psuedo?) standard Berkeley Sockets on the vxWorks
side.  The appropriate sockaddr structures must be set up appropriately on
both sides, and there is a host of documentation for this.  We mainly
transfer data TO the PC.  The following code snippits are a couple of
routines at the vxWorks end that rearrange the bytes for both an integer
array and a long integer array to a form readily useable on the PC side.
We don't do the reverse, but this should give you a feel for the byte
ordering.  It's not elegant, but it works.  It is possible that the canned
routines (htonl, etc.) may also work, but I haven't used them except in
conversion of internet addresses, etc.

Good luck -- Eli Wittner -- CTI Cyclotron Systems, Inc.

/**************************************************************************
****
 *  ROUTINE: intMemCpy
                                                        *
 *

 *
 *  This routine copies an integer array into a character buffer, swapping
    *
 *    bytes as we go so that the "little/big endian" differences between
the  *
 *    PC and Motorola computers are accounted for.
                            *
 *

 *

***************************************************************************
***/

static void intMemcpy ( unsigned char * szDest,
                        unsigned int *szSource,
                        int iNumBytes )
{
   register int i, j ;
   unsigned int iT ;

   for ( i=0; i<(iNumBytes/2); i++ )
      {
      j = i*2 ;
      iT = *(szSource+i) ;

      *(szDest+j)   = (unsigned char)(iT & 0x00ff ) ;
      *(szDest+j+1) = (unsigned char)(iT>>8) ;
      }

}  /*  End function intMemcpy  */



/**************************************************************************
****
 *  ROUTINE: longMemCpy
                                                       *
 *

 *
 *  This routine copies a long integer array into a character buffer,
swapping    *
 *    bytes as we go so that the "little/big endian" differences between
the  *
 *    PC and Motorola computers are accounted for.
                            *
 *

 *

***************************************************************************
***/

static void longMemcpy ( unsigned char * szDest,
                         unsigned long *szSource,
                         int iNumBytes )
{
   register int i, j ;
   unsigned long iT ;

   for ( i=0; i<(iNumBytes/4); i++ )
      {
      iT = *(szSource+i) ;
      for ( j=0; j<4; j++ )
         {
         *(szDest+(i*4)+j) = iT & 0xff ;
         iT = iT >>8 ;

         }
      }

}  /*  End function longMemcpy  */



--MimeMultipartBoundary--


From pat.shaw@canada.cdev.com  Tue Sep 30 12:43:55 1997
From: "Shaw, Pat" 
Date: Tue Sep 30 12:43:58 PDT 1997
Subject: AMD 79C970A
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------ =_NextPart_000_01BCCDB7.0DF4FD80
Content-Type: text/plain

Hi:
	I was wondering if anybody has ported the VxWorks AMD 79C970
driver to a Pentium based BSD.  I am not totally familiar with the the
various driver interaction that take place between VxWorks and PCI based
devices.



Thanks for your help,

Patrick Shaw
Computing Devices Canada

------ =_NextPart_000_01BCCDB7.0DF4FD80
Content-Type: application/ms-tnef
Content-Transfer-Encoding: base64
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------ =_NextPart_000_01BCCDB7.0DF4FD80--


From mmcquade@aa.net  Tue Sep 30 20:33:29 1997
From: Mike McQuade 
Date: Tue Sep 30 20:33:32 PDT 1997
Subject: mv2603/4 and PCI bursting
--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"



We are working with an mv-2603 board, I have a Vmetro PCI analyzer
on eval, one thing that Im curious about is PCI burst transactions,
I don't see _ANY_ on this system.

We have a PMC video board on the mv2603, in addition to the DEC21140,
and the SCSI chip on the PCI bus.


I have enabled burst transfers in the video chip, and the Raven
chip has write posting on, yet no write bursts.


Does anyone know if its possible to do PCI burst transactions ?


OR do write bursts only occur when you have "WriteBack" cache ?


Thanks


Mike.







--MimeMultipartBoundary--