Re: 2.6.20-mm2

["Rafael J. Wysocki" <rjw@sisk.pl>]

On Monday, 19 February 2007 12:45, Michal Piotrowski wrote:
> On 19/02/07, Rafael J. Wysocki <rjw@sisk.pl> wrote:
> > On Monday, 19 February 2007 01:00, Andrew Morton wrote:
> > > On Mon, 19 Feb 2007 00:25:48 +0100 "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > >
> > > > > netconsole is good.
> > > >
> > > > I know. :-)
> > > >
> > > > In the meantime, I've got something worse on another x86_64 box:
> > > >
> > > > Asus Laptop ACPI Extras version 0.30
> > > >   L5D model detected, supported
> > > > audit(1171831698.918:2): audit_pid=4281 old=0 by auid=4294967295
> > > > general protection fault: 0000 [2] PREEMPT
> > > > last sysfs file: /class/net/eth2/carrier
> > > > CPU 0
> > > > Modules linked in: af_packet ipv6 snd_pcm_oss snd_mixer_oss snd_seq snd_seq_device asus_acpi backlight button battery ac dm_mod pcmr
> > > > Pid: 178, comm: pdflush Not tainted 2.6.20-mm2 #1
> > > > RIP: 0010:[<ffffffff8034bce4>]  [<ffffffff8034bce4>] __make_request+0x134/0x370
> > > > RSP: 0000:ffff81005ed659a0  EFLAGS: 00010297
> > > > RAX: 00000000ffffffff RBX: 6b6b6b6b6b6b6b6b RCX: 000000000203396a
> > > > RDX: 0000000100000000 RSI: ffff810037b4dbb0 RDI: ffff81004683d8c0
> > > > RBP: ffff81005ed659f0 R08: ffff81004683d070 R09: ffff81003d333cc0
> > > > R10: 0000000000000000 R11: 0000000000000000 R12: ffff810037b4dbb0
> > > > R13: ffff81005daba3f0 R14: ffff810037daca90 R15: ffff81005daba3d0
> > > > FS:  00002ad4a29e6d00(0000) GS:ffffffff805db000(0000) knlGS:0000000000000000
> > > > CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b
> > > > CR2: 00002b6a345aa000 CR3: 0000000056585000 CR4: 00000000000006e0
> > > > Process pdflush (pid: 178, threadinfo ffff81005ed64000, task ffff810037b060c0)
> > > > Stack:  ffff810002852540 0000000000000001 ffff810037b4dbb0 ffffffff8026be21
> > > >  ffff81005ed65a40 0000000000000008 ffff810037b4dbb0 0000000000000800
> > > >  0000000000000008 ffff8100021d94e0 ffff81005ed65a40 ffffffff80348e7c
> > > > Call Trace:
> > > >  [<ffffffff8026be21>] mempool_alloc_slab+0x11/0x20
> > > >  [<ffffffff80348e7c>] generic_make_request+0x1ec/0x230
> > >
> > > yeah. everyone except me is hitting that.
> >
> > FWIW, I don't see it on an SMP machine.
> >
> 
> I can reproduce this on my SMT P4.
> 
> CONFIG_SMP=y
> CONFIG_X86_PC=y
> CONFIG_MPENTIUM4=y
> CONFIG_NR_CPUS=2
> CONFIG_SCHED_SMT=y

It may be related to preemption.  The box I'm not seeing it on runs a
non-preemptible kernel (CONFIG_PREEMPT_VOLUNTARY is set).

BTW, on the box where I'm able to reproduce it, I have

(gdb) l *__make_request+0x134
0xffffffff8034b764 is in __make_request (include/asm/processor.h:411).
406     #define cpu_has_fpu 1
407
408     #define ARCH_HAS_PREFETCH
409     static inline void prefetch(void *x)
410     {
411             asm volatile("prefetcht0 %0" :: "m" (*(unsigned long *)x));
412     }
413
414     #define ARCH_HAS_PREFETCHW 1
415     static inline void prefetchw(void *x)

So I guess x is NULL somewhere ...