Date: Mon, 15 Dec 2008 16:13:04 -0600 From: Guy Helmer <ghelmer@palisadesys.com> To: Kostik Belousov <kostikbel@gmail.com> Cc: freebsd-stable@freebsd.org Subject: Re: 7.1RC1: system hang Message-ID: <4946D670.40402@palisadesys.com> In-Reply-To: <20081215213332.GQ2038@deviant.kiev.zoral.com.ua> References: <49419BE0.2070502@palisadesys.com> <3a142e750812111900u1d9264bh27d39092f5dcc4bf@mail.gmail.com> <49426880.40302@palisadesys.com> <3a142e750812120729v478d850do93e31e5d8502ef68@mail.gmail.com> <4942C178.2030206@palisadesys.com> <49468FE2.4040007@palisadesys.com> <20081215213332.GQ2038@deviant.kiev.zoral.com.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
Kostik Belousov wrote:
> On Mon, Dec 15, 2008 at 11:12:02AM -0600, Guy Helmer wrote:
>
>> I have a recurring hang on FreeBSD 7.1 roughly-RC1 (releng_7_1 checkout
>> as of 2008-12-08) on a dual-CPU hyperthreaded Xeon i386. Here is the
>> output from ps/m, show allpcpu, show locks, show alllocks, allt, show
>> allchains, and show lockedvnods commands in the debugger.
>>
> Can you show me the source line for vm_fault+0x1b1b ?
> Do the l *(vm_fault+0x1b1b) at the kgdb prompt, you do not need the core,
> only the same debugging kernel as was booted on deadlocked machine.
>
(kgdb) l *(vm_fault+0x1b1b)
0xc07a765b is in vm_fault (../../../vm/vm_fault.c:888).
883 if (((fault_flags & VM_FAULT_WIRE_MASK) == 0) && (wired
== 0)) {
884 vm_fault_prefault(fs.map->pmap, vaddr, fs.entry);
885 }
886 VM_OBJECT_LOCK(fs.object);
887 vm_page_lock_queues();
888 vm_page_flag_set(fs.m, PG_REFERENCED);
889
890 /*
891 * If the page is not wired down, then put it where the
pageout daemon
892 * can find it.
Thanks! Let me know if there is anything else you need,
Guy
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4946D670.40402>