bob@rush.cts.com (Bob Ames) (02/19/89)
Anybody seen this? We've got a 3B1 which panics with NMI parity errors after 2-10 lines have printed through the parallel Port. Starlan printing works fine. We've tried other computers with this printer/cable and they print OK. I developed a theory that somehow the parallel port drains some power and causes memory to forget. I'll probably get 20 letters from people telling me how silly this theory is. |-) Bob --- National Organization for the | Bob Ames \ Marijuana: "The safest drug Reform of Marijuana Laws, NORML | 619-743-2546 \ in the world" (Source Drug "Marijuana is safer than tobacco"| bob@rush.cts.com \ Enforcement Admin - 1987) Surgeon General C Everett Koop'88 ##### ..nosc! )..!crash!rush.cts.com!bob "Pot is best fire de-erosion seed| ..hplabs!hp-sdd! / "We each pay a fabulous & 1 acre of pot equals 4 acres of| ..rutgers!ucsd! / price - for our visions trees for paper products." - USDA| / of paradise" -- Rush 1987
ditto@cbmvax.UUCP (Michael "Ford" Ditto) (02/20/89)
In article <788@rush.cts.com> bob@rush.cts.com (Bob Ames) writes: >We've got a 3B1 which panics with NMI parity errors after 2-10 lines >have printed through the parallel Port. Have you tried the same printer & cable on another Unix PC? My 3B1 has always done the same thing; I'm glad to hear I'm not the only one, since that seems to point to a cabling problem or something. Since I only discovered the problem after the warrantee expired, I just resigned myself to sending files to the Amiga for printing. But now, I'll investigate it further. I suspect a bug in the lp device driver which only shows up under some rare timing conditions which our printers somehow make happen. The one I tried is an Epson FX-85, with a "straight-through" cable with every pin connected. >I developed a theory that somehow the parallel port drains some power >and causes memory to forget. I'll probably get 20 letters from people >telling me how silly this theory is. |-) Hmmmm... doesn't seem too likely unless you have a marginal ram chip to start with. Also, I would expect that just having the printer connected would be enough to cause such a problem, even if it wasn't printing. -- -=] Ford [=- "The number of Unix installations (In Real Life: Mike Ditto) has grown to 10, with more expected." ford@kenobi.cts.com - The Unix Programmer's Manual, ...!sdcsvax!crash!kenobi!ford 2nd Edition, June, 1972. ditto@cbmvax.commodore.com
jay@banzai.UUCP (Jay Schuster) (02/22/89)
In article <788@rush.cts.com> bob@rush.cts.com (Bob Ames) writes: >We've got a 3B1 which panics with NMI parity errors after 2-10 lines >have printed through the parallel Port. We have seen a lot of 3B1's go through our doors, and just this week we had exactly this problem with one of them. Freshly loaded 3.51. This is with an Okidata 390 printer, but we have dozens of other 3B1's in the exact same configuration running with no problems. We loaded 3.51 again (actually, we swapped hard drives) and the problem persisted. We got AT&T to come and replace the motherboard, and now, instead of NMI's, we get completely garbled output, as if no handshaking is going on with the parallel port. I think we're going to open this baby up and check the wiring to the parallel port. Even though we bought it recently, it has an old serial number, so who knows? -- Jay Schuster uunet!uvm-gen!banzai!jay, attmail!banzai!jay The People's Computer Company `Revolutionary Programming'
ditto@cbmvax.UUCP (Michael "Ford" Ditto) (02/24/89)
In article <1409@mtunb.ATT.COM> jcm@mtunb.UUCP (was-John McMillan) writes: >PLEASE, > WHEN REPORTING ANY PROBLEMS: > > ***** IDENTIFY YOUR SOFTWARE VERSION ***** > >For those of us with several source trees -- about 6, in my case -- >it provides SOME focus to the issue. For some reason, the thought did not occur to me at the time that anyone on the net might be interested in investigating the problem ... silly me! (Also, I was still suspecting a hardware problem.) In fact, John McMillan's posting inspired me to attempt to reproduce the problem. I went out and bought an appropriate printer cable, since I don't have the one I was using before, and hooked up the printer. I pressed Shift-Print (which was the sure-fire way to cause a panic in my earlier attempts) and out came a perfect screen dump. I printed several pages of listings just fine also. Hmm. I'm pretty sure that the cable is the same; I know that the printer and computer are the same. Perhaps some software changed. I don't remember for sure whether I was running the 3.51 kernel or the "3.51.1" kernel AT&T sent me when I reported the tty000-break bug. The latter is what I am running now. >Also: If you are getting recurring information in the register >dumps, these registers -- in conjunction with the SOFTWARE VERSION -- >may pinpoint the problem. Well, the good news is that I have the panic message and register dump saved in a file. The bad news is that I have about 10 of these panic records, and I very stupidly did not keep track of which message corresponded to what event. Oh Well. I'll try to make it crash again later, and if I succeed, I promise to mail the panic message and a complete kernel core dump to anyone who wants it! Did I ever post my kernel core dump driver? -- -=] Ford [=- "The number of Unix installations (In Real Life: Mike Ditto) has grown to 10, with more expected." ford@kenobi.cts.com - The Unix Programmer's Manual, ...!sdcsvax!crash!kenobi!ford 2nd Edition, June, 1972. ditto@cbmvax.commodore.com
ebd@cup.portal.com (Elliot B Dierksen) (02/24/89)
One thing to check on 3B1/7300's is that in my experience with my own and numerous others is that they seem to eat parallel cables. I know this sounds really stupid but i have seen all kind of printer problems develop on Unix PC's rapidly and then mysteriously disappear when the cable is replaced... ???? Like i said, i know it sounds irrational but i have seen it happen at least 3 different times on 3 different boxes. Any good answers to that one?? :-) Elliot Dierksen ebd@cup.portal.com
rjg@sialis.mn.org (Robert J. Granvin) (02/26/89)
>One thing to check on 3B1/7300's is that in my experience with my own and >numerous others is that they seem to eat parallel cables. I know this sounds >really stupid but i have seen all kind of printer problems develop on Unix PC's >rapidly and then mysteriously disappear when the cable is replaced... ???? A very common problem with the 3b1/7300 and parallel cables isn't the cable itself, but rather how the cable connects to the 3b1. For some reason, the parts normally available on the cable and 3b1 end don't deal all that well with each other. When you plug it in, it may look and feel like it has a solid connection, but it often doesn't. Many people have had parallel printer problems, ranging from odd printing to complete failures. Solution? Forcibly press in the cable into the 3b1 while testing the printer. If this improves printer performance (or brings it back to life), then get some screws and bolt the cable in place. You'd be surprised, but it does wonders. -- Robert J. Granvin National Information Services Gorn base 2 just appeared. rjg@sialis.mn.org {amdahl,hpda}!bungia!sialis!rjg