From owner-freebsd-current Mon Nov 29 15:24:36 1999 Delivered-To: freebsd-current@freebsd.org Received: from barracuda.aquarium.rtci.com (barracuda.aquarium.rtci.com [208.11.247.5]) by hub.freebsd.org (Postfix) with ESMTP id 0E29815397 for ; Mon, 29 Nov 1999 15:24:30 -0800 (PST) (envelope-from tstromberg@rtci.com) Received: from rtci.com (karma.afterthought.org [208.11.244.6]) by barracuda.aquarium.rtci.com (8.9.3+Sun/8.9.3) with ESMTP id SAA07309 for ; Mon, 29 Nov 1999 18:24:40 -0500 (EST) Message-ID: <38430B2D.2E7A5F0D@rtci.com> Date: Mon, 29 Nov 1999 18:24:29 -0500 From: Thomas Stromberg Organization: Research Triangle Commerce, Inc. X-Mailer: Mozilla 4.7 [en] (X11; I; FreeBSD 4.0-CURRENT i386) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-current@freebsd.org Subject: -CURRENT crash under high exec() loads.. (vm_map_insert?) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG While doing some testing (and actually, in the middle of me trying to post some results) for the FreeBSD Auditing project, my 4.0-CURRENT box crashed. These tests involved a barrage of automated exec() calls which I suspect is what tore it down. I had a similar crash two weeks ago, but did not have the kernel set for debug mode, so I wrote just a quick summary on the bottom of the trace. This may of course have absolutely nothing to do with the testing, as it was (at time of crash) testing /usr/bin/bc, which is mostly waiting on the timeout to kill it off because of the interactive mode. Otherwise there was ~960,000 accumulated exec()'s altogether in a 5 hour period. As you can see, cron is the process that crashed it. Two weeks ago it was an eggdrop. I've not had any crashes in -CURRENT while this program was not running however. If you notice, both times it crashed on vm_map_insert. Enviroment: ----------- Machine: PIII-500, 96M RAM uname: FreeBSD karma.afterthought.org 4.0-CURRENT FreeBSD 4.0-CURRENT #0: Tue Nov 23 07:31:52 EST 1999 root@:/usr/src/sys/compile/KARMA i386 load: ~1.20 Conditions: ----------- X w/ Netscape 4.7 was running 'breakwidgets' was pumping out up to 5000 exec() calls a minute. (depending on timeouts). At time of crash it was probably more like 30-100 execs a minute. (I suspect the latter, naturally). Dump: ----- IdlePTD 3022848 initial pcb at 26d3e0 panicstr: page fault panic messages: --- Fatal trap 12: page fault while in kernel mode fault virtual address = 0x8 fault code = supervisor write, page not present instruction pointer = 0x8:0xc01d7a67 stack pointer = 0x10:0xccd67df0 frame pointer = 0x10:0xccd67e0c code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 0, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 128 (cron) interrupt mask = none trap number = 12 panic: page fault syncing disks... 32 9 done dumping to dev #wd/0x20001, offset 1413889 dump 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 --- #0 0xc0131e80 in boot () #1 0xc013221d in panic () #2 0xc0215102 in trap_fatal () #3 0xc0214db5 in trap_pfault () #4 0xc0214987 in trap () #5 0xc01d7a67 in vm_map_insert () <=== #6 0xc01d7c88 in vm_map_find () #7 0xc01d6e93 in kmem_alloc_pageable () #8 0xc0211a3e in pmap_pinit () #9 0xc01d7650 in vmspace_alloc () #10 0xc01d950c in vmspace_fork () #11 0xc01d6a28 in vm_fork () #12 0xc012c96f in fork1 () #13 0xc012c1f6 in fork () #14 0xc021533a in syscall () #15 0xc0209d36 in Xint0x80_syscall () #16 0x804aeee in ?? () #17 0x8049c4d in ?? () #18 0x8049a15 in ?? () #19 0x80497d9 in ?? () Two weeks beforehand this was the crash trace (I wrote it down, if need be, I can re-type all the nasty details) vm_map_insert c0272568, 0, 0, 0, cf294000 @ 0x18c <=== vm_map_find kmem_alloc_pageable pmap_pinit vmspace_alloc vmspace_exec exec_new_vmspace eec_elf_imgact execve syscall xint0x80_syscall -- ====================================================================== thomas r. stromberg smtp://tstromberg@rtci.com assistant is manager / systems guru http://thomas.stromberg.org research triangle commerce, inc. finger://thomas@stromberg.org 'om mani pedme hung' pots://1.919.380.9771:3210 ================================================================[eof]= To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message