lkml.org 
[lkml]   [2010]   [Mar]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: unable to handle kernel paging request on resume with 2.6.33-00001-gbaac35c
Date
On Monday 01 March 2010, Michal Hocko wrote:
> [Let's CC mm guys]

I guess it's rather architecture-related than a genering mm issue.

> On Mon, Mar 01, 2010 at 10:07:37PM +0100, Rafael J. Wysocki wrote:
> > On Monday 01 March 2010, Michal Hocko wrote:
> > > Hi,
> > >
> > > I have experienced the following kernel BUG on resume from suspend from
> > > disk (the whole log from hibarnation to suspend along with kernel
> > > config are attached):
> > >
> > > BUG: unable to handle kernel paging request at 00aaaaaa
> > > IP: [<c019e28c>] anon_vma_link+0x2c/0x39
> > > *pde = 00000000
> > > Oops: 0002 [#1] PREEMPT SMP
> > > last sysfs file: /sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0003:00/power_supply/AC/type
> > > Modules linked in: aes_i586 aes_generic iwl3945 iwlcore mac80211 cfg80211 fbcon font bitblit softcursor i915 drm_kms_helper drm fb i2c_algo_bit cfbcopyarea i2c_core cfbimgblt cfbfillrect fuse tun coretemp hwmon snd_hda_codec_realtek snd_hda_intel snd_hda_codec arc4 ecb snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_oss snd_seq_midi_event snd_seq snd_timer fujitsu_laptop snd_seq_device rtc_cmos rtc_core led_class rtc_lib snd snd_page_alloc video backlight output [last unloaded: cfg80211]
> > >
> > > Pid: 3942, comm: kxkb Not tainted 2.6.33-00001-gbaac35c #11 FJNB1B5/LIFEBOOK S7110
> > > EIP: 0060:[<c019e28c>] EFLAGS: 00010246 CPU: 1
> > > EIP is at anon_vma_link+0x2c/0x39
> > > EAX: 00aaaaaa EBX: f69c6410 ECX: f69c6414 EDX: f63e4df4
> > > ESI: f63e4dc0 EDI: f63e4e14 EBP: f6901ec0 ESP: f6901eb8
> > > DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> > > Process kxkb (pid: 3942, ti=f6901000 task=f6aa6ff0 task.ti=f6901000)
> > > Stack:
> > > f63e4dc0 f23fc7e4 f6901efc c012fc28 f6aa6ff0 f63e4e30 f63e4e34 f63e4e24
> > > <0> ca4656f4 f6ace734 f6aa6ff0 f6ace700 ca4656c0 f23fc790 ca560000 fffffff4
> > > <0> f659ef94 f6901f38 c0130821 f6aa6ff0 f6901fb4 bff441f0 ca560208 00000000
> > > Call Trace:
> > > [<c012fc28>] ? dup_mm+0x1c7/0x3d3
> > > [<c0130821>] ? copy_process+0x98e/0xf26
> > > [<c0130ed6>] ? do_fork+0x11d/0x2a1
> > > [<c0434547>] ? _raw_spin_unlock+0x14/0x28
> > > [<c01b6795>] ? set_close_on_exec+0x45/0x4b
> > > [<c01b6e98>] ? do_fcntl+0x15f/0x3f1
> > > [<c0108678>] ? sys_clone+0x20/0x25
> > > [<c010291d>] ? ptregs_clone+0x15/0x38
> > > [<c0102850>] ? sysenter_do_call+0x12/0x26
> > > Code: 89 e5 56 53 0f 1f 44 00 00 8b 58 3c 89 c6 85 db 74 22 89 d8 e8 54 65 29 00 8b 43 08 8d 56 34 8d 4b 04 89 53 08 89 4e 34 89 46 38 <89> 10 89 d8 e8 9e 62 29 00 5b 5e 5d c3 55 89 e5 0f 1f 44 00 00
> > > EIP: [<c019e28c>] anon_vma_link+0x2c/0x39 SS:ESP 0068:f6901eb8
> > > CR2: 0000000000aaaaaa
> > > ---[ end trace b7f008b0e5aa7c65 ]---
> >
> > This looks like a low-level memory management issue of some sort.
>
> Yes, it really looks strange. dup_mm+0x1c7 matches to:
> c102fc0e: 81 60 14 ff df ff ff andl $0xffffdfff,0x14(%eax)
> c102fc15: 8b 45 ec mov -0x14(%ebp),%eax
> c102fc18: c7 43 0c 00 00 00 00 movl $0x0,0xc(%ebx)
> c102fc1f: 89 03 mov %eax,(%ebx)
> c102fc21: 89 d8 mov %ebx,%eax
> c102fc23: e8 38 e6 06 00 call c109e260 <anon_vma_link>
> c102fc28: 8b 43 48 mov 0x48(%ebx),%eax <<< BANG
>
> which corresponds to:
> kernel/fork.c:336
> tmp->vm_flags &= ~VM_LOCKED;
> tmp->vm_mm = mm;
> tmp->vm_next = NULL;
> anon_vma_link(tmp);
> file = tmp->vm_file; <<< BANG
>
> ebx is tmp which somehow got deallocated. I cannot see how this could happened.

Through a page tables corruption or a TLB issue, for example.

> > What's the HEAD commit in this kernel tree?
>
> $ git describe
> v2.6.33-1-gbaac35c

I can't find gbaac35c anywhere post 2.6.33. Can you just send the output
of "git show | head -1", please?

> > Also, is the problem reproducible?
>
> As I've already mentioned. This is the first time I have seen this problem.
> I am using suspend to disk and wake up quite often (several times a day). I
> haven't tried suspend/resume loop test yet.

OK

Given the apparent nature of the problem it will be extremely difficult to
track down without a reliable way to reproduce it.

Rafael


\
 
 \ /
  Last update: 2010-03-02 01:07    [W:0.048 / U:0.492 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site