Bug#632678: linux-image-2.6.32-5-amd64: el sistema se cuelga de forma aleatoria

[Jonathan Nieder <jrnieder@gmail.com>]

retitle 632678 [radeon] NULL pointer dereference in vt_console_device, beep, screen goes black
tags 632678 + moreinfo
quit

Hi,

Simón C. A. wrote:

> De vez en cuando el sistema emite un pitido corto
[...]

translate.google.com comes to the rescue:

| Occasionally the system emits a short beep through the internal PC speaker
| and the display turns off.  It's possible to reboot using the combination
| ALT+SysRQ+REISUB.
|
| The time of the crash is random, but the probability increases with
| more applications running or applications consuming more resources.
| At the time of the crash, the kernel.log file indicates that there
| was a null pointer reference:

> Jun 22 22:59:20 debian kernel: [ 5392.581979] BUG: unable to handle kernel NULL pointer dereference at 0000000000000292
> Jun 22 22:59:20 debian kernel: [ 5392.581993] IP: [<ffffffff812072ca>] vt_console_device+0x14/0x1e

So, the .device method of vt_console_driver was called with one of its
arguments equal to NULL.  The only caller I know of is
kernel/printk.c::console_device, which suggests that the latter was
called with its argument NULL...

[...]
> Jun 22 22:59:20 debian kernel: [ 5392.582282] Call Trace:
> Jun 22 22:59:20 debian kernel: [ 5392.582292]  [<ffffffff811fbc22>] ? n_tty_poll+0x124/0x138
> Jun 22 22:59:20 debian kernel: [ 5392.582300]  [<ffffffff811f854a>] ? tty_poll+0x56/0x6d
> Jun 22 22:59:20 debian kernel: [ 5392.582311]  [<ffffffff810fc436>] ? do_select+0x37b/0x57a
> Jun 22 22:59:20 debian kernel: [ 5392.582320]  [<ffffffff810fcaaf>] ? __pollwait+0x0/0xd6
> Jun 22 22:59:20 debian kernel: [ 5392.582328]  [<ffffffff810fcb85>] ? pollwake+0x0/0x5b
> Jun 22 22:59:20 debian kernel: [ 5392.582336]  [<ffffffff810fcb85>] ? pollwake+0x0/0x5b
> Jun 22 22:59:20 debian kernel: [ 5392.582344]  [<ffffffff810fcb85>] ? pollwake+0x0/0x5b
> Jun 22 22:59:20 debian kernel: [ 5392.582352]  [<ffffffff810fcb85>] ? pollwake+0x0/0x5b
> Jun 22 22:59:20 debian kernel: [ 5392.582362]  [<ffffffff810c2c71>] ? shmem_delete_inode+0x0/0xf8
> Jun 22 22:59:20 debian kernel: [ 5392.582371]  [<ffffffff81064e3b>] ? bit_waitqueue+0x10/0xa0
> Jun 22 22:59:20 debian kernel: [ 5392.582379]  [<ffffffff81116dbe>] ? fsnotify_clear_marks_by_inode+0x20/0xcd
> Jun 22 22:59:20 debian kernel: [ 5392.582390]  [<ffffffff810e5bbd>] ? virt_to_head_page+0x9/0x2a
> Jun 22 22:59:20 debian kernel: [ 5392.582398]  [<ffffffff8103fa3a>] ? __wake_up+0x30/0x44
> Jun 22 22:59:20 debian kernel: [ 5392.582441]  [<ffffffffa0388d3f>] ? radeon_bo_wait+0xb1/0xbd [radeon]
> Jun 22 22:59:20 debian kernel: [ 5392.582451]  [<ffffffff812fbab6>] ? mutex_lock+0xd/0x31
> Jun 22 22:59:20 debian kernel: [ 5392.582460]  [<ffffffff810fc7b9>] ? core_sys_select+0x184/0x21e

... but it is not in the backtrace.  So much for that guess.  Could
you try a kernel from unstable or experimental and let us know how it
goes?

Thanks for reporting,
Jonathan

> Jun 22 22:59:20 debian kernel: [ 5392.582471]  [<ffffffff8101657d>] ? read_tsc+0xa/0x20
> Jun 22 22:59:20 debian kernel: [ 5392.582479]  [<ffffffff8106c48f>] ? ktime_get_ts+0x68/0xb2
> Jun 22 22:59:20 debian kernel: [ 5392.582487]  [<ffffffff810fca86>] ? sys_select+0x92/0xbb
> Jun 22 22:59:20 debian kernel: [ 5392.582495]  [<ffffffff81010b42>] ? system_call_fastpath+0x16/0x1b
> Jun 22 22:59:20 debian kernel: [ 5392.582500] Code: 48 6b c0 28 48 8b 80 70 c3
> 67 81 8a 80 23 01 00 00 d0 e8 83 e0 03 c3 66 8b 47 42 66 85 c0 74 05 98 ff c8
> eb 06 8b 05 7e 5a 47 00 <89> 06 48 8b 05 8d 6a 47 00 c3 8a 87 ec 00 00 00 83 e0
> 01 3c 01
> Jun 22 22:59:20 debian kernel: [ 5392.582563] RIP  [<ffffffff812072ca>] vt_console_device+0x14/0x1e
> Jun 22 22:59:20 debian kernel: [ 5392.582572]  RSP <ffff88006cf659b0>
> Jun 22 22:59:20 debian kernel: [ 5392.582576] CR2: 0000000000000292
> Jun 22 22:59:20 debian kernel: [ 5392.582581] ---[ end trace 1d363e716392f765 ]---