Date: Tue, 24 Nov 1998 01:11:01 -0500 (EST) From: techs@obfuscation.org To: freebsd-bugs@FreeBSD.ORG Cc: techs@obfuscation.org Subject: Re: kern/4859: SMP kernel panics with timeout table full msg when running two cpu intensive programs Message-ID: <19981124061101.3264.qmail@3jane.obfuscation.org>
index | next in thread | raw e-mail
Kenneth Merry <ken@plutotech.com> wrote: ivar@hosteng.org wrote... > > >Number: 4859 > >Category: kern > >Synopsis: SMP kernel panics with timeout table full msg when running two cpu intensive programs > >Confidential: no > >Severity: serious > >Priority: medium > >Responsible: freebsd-bugs > >State: open > >Class: sw-bug > >Submitter-Id: current-users > >Arrival-Date: Sun Oct 26 11:10:00 PST 1997 > >Last-Modified: > >Originator: Ivar E. Hosteng > >Organization: > Merkantildata ASA > >Release: 3.0-971002-SNAP > >Environment: > FreeBSD romulus.hosteng.org 3.0-971002-SNAP FreeBSD 3.0-971002-SNAP #0: Sun Oct 26 18:22:38 CET 1997 root@romulus.hosteng.org:/usr/src/sys/compile/erik i386 > >Description: > The kernel is bulilt with SMP enabled and is running on a 2 CPU 200Mhz PPRo system (Intel Providence motherboard) with 128 MB ram. > > When i run the rc5 key cracking program for the rc5 cracking contest (ftp://ftp.distributed.net/pub/rc5-64/v2.6401/rc56401-freebsd-x86-cli.tar.gz) the system panics with a 'timeout table full' followed by a 'page fault' panic. > > Othervise the system have been solid. > >How-To-Repeat: > > Just let two copies of the above mentioned program run concurrently for a period of time. My record (3 trys) is 5 minutes until it panics) > I tried to reproduce your problem with a -current SMP kernel from > October 26th. I ran two processes simultaneously for about 20 minutes or > so without problems. > > Try upgrading your kernel to something current, and see if that > makes any difference. > > In case it matters, I'm running: > ASUS P/I-P65UP5 w/ C-P6ND cpu card with 2 Pentium Pro 200's (256K cache) > 128MB parity RAM For what it's worth, the distributed.net RC5DES Client v2.7100.416 running two instances of itself hangs my system completely. no panic message. no dump. no nothing. just hang. one instance runs for hours and hours. adding a second instance hangs the system anywhere from 15 minutes to an hour later. FreeBSD hologram-rose.obfuscation.org 3.0-RELEASE FreeBSD 3.0-RELEASE #3: Thu Nov 19 21:02:49 EST 1998 techs@hologram-rose.obfuscation.org:/usr/src/sys/compile/HOLOGRAM-ROSE i386 the system is a Micronics W6LI Lightning with two stepping 9 Pentium Pro 180's and 160mb of edo ram. If someone can tell me what further diagnostic steps would be useful in getting this tracked down, I'm more than happy to do it. Like Ivar noted, the system has been stable otherwise. Is this a PPro only issue? -- Erik Fichtner; Warrior SysAdmin (emf|techs) http://www.obfuscation.org/~techs N 39 10.409' W 77 11.750' "Any attempt to brew coffee with a teapot should result in the error code "418 I'm a teapot". The resulting entity body MAY be short and stout." - Sec 2.3.2, RFC 2324 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the messagehelp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981124061101.3264.qmail>