Re: 2.6.33-rc2+ bug in fs/btrfs/ordered-data.c:672

["Yan, Zheng " <yanzheng@21cn.com>]

On Mon, Jan 4, 2010 at 8:22 AM, Johannes Hirte
<johannes.hirte@fem.tu-ilmenau.de> wrote:
> Am Samstag 02 Januar 2010 23:34:17 schrieb Carlos R. Mafra:
>> On Sa =A02.Jan'10 at 11:32:27 +0100, Carlos R. Mafra wrote:
>> > I started testing btrfs for my /home a few days ago and yesterday
>> > I hit a kernel bug, using 2.6.33-rc2-00187-g08d869a.
>> >
>> > I wasn't doing any stress test with it, I was simply watching a
>> > DVD with xine while chrome was open in another workspace.
>>
>> Today I hit the bug twice, and it was always with chrome. I've
>> seen the other two reports in kerneloops.org involving
>> btrfs_ordered_update_i_size() and chrome was there too.
>>
>> And now I saved the dmesg in the ext3 partition, so now there
>> is no need for the photo,
>>
>>
>> [13450.613952] ------------[ cut here ]------------
>> [13450.613957] kernel BUG at fs/btrfs/ordered-data.c:672!
>> [13450.613960] invalid opcode: 0000 [#1] SMP
>> [13450.613963] last sysfs file:
>> =A0/sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq [13450.6139=
66] CPU 1
>> [13450.613970] Pid: 3372, comm: chrome Not tainted
>> =A02.6.33-rc2-fs-00187-g08d869a #293 VAIO/VGN-FZ240E [13450.613973] =
RIP:
>> =A00010:[<ffffffff811d9537>] =A0[<ffffffff811d9537>]
>> =A0btrfs_ordered_update_i_size+0x237/0x3d0 [13450.613982] RSP:
>> =A00018:ffff880071355da8 =A0EFLAGS: 00010287
>> [13450.613984] RAX: ffff88007f0c0728 RBX: 0000000000000000 RCX:
>> =A0ffff880077c0c6a8 [13450.613986] RDX: 00000000002b5000 RSI:
>> =A00000000000000000 RDI: ffff880077c0c6b8 [13450.613989] RBP:
>> =A0ffff880071355e18 R08: 0000000000000000 R09: 0000000000000000
>> =A0[13450.613991] R10: 0000000000000000 R11: 0000000000000000 R12:
>> =A0ffff8800729dd260 [13450.613993] R13: 00000000002b4b24 R14:
>> =A00000000000000000 R15: 00000000002b4b24 [13450.613996] FS:
>> =A000007f2b3db68910(0000) GS:ffff880001b00000(0000) knlGS:0000000000=
000000
>> =A0[13450.613999] CS: =A00010 DS: 0000 ES: 0000 CR0: 000000008005003=
3
>> [13450.614001] CR2: 00007f2b3c2a4000 CR3: 0000000071167000 CR4:
>> =A000000000000006e0 [13450.614003] DR0: 0000000000000000 DR1:
>> =A00000000000000000 DR2: 0000000000000000 [13450.614005] DR3:
>> =A00000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
>> =A0[13450.614008] Process chrome (pid: 3372, threadinfo ffff88007135=
4000,
>> =A0task ffff88007d380640) [13450.614010] Stack:
>> [13450.614011] =A0ffff8800729dd0c0 0000000000000000 fffffffffffff000
>> =A00000000000000fff [13450.614015] <0> ffff8800729dd0f0 ffff88007c7c=
c000
>> =A0ffff8800729dd0c0 ffff8800729dd170 [13450.614019] <0> ffff88007135=
5e18
>> =A0ffff880071355ee8 ffff8800729dd260 0000000000000000 [13450.614023]=
 Call
>> =A0Trace:
>> [13450.614028] =A0[<ffffffff811c915d>] btrfs_setattr+0x17d/0x270
>> [13450.614033] =A0[<ffffffff810e5fe4>] notify_change+0x104/0x2e0
>> [13450.614037] =A0[<ffffffff810ce84f>] do_truncate+0x5f/0x90
>> [13450.614041] =A0[<ffffffff810d0002>] ? vfs_write+0x132/0x180
>> [13450.614044] =A0[<ffffffff810cef79>] sys_ftruncate+0xe9/0x130
>> [13450.614049] =A0[<ffffffff81002dab>] system_call_fastpath+0x16/0x1=
b
>> [13450.614050] Code: 0f 1f 40 00 eb 88 66 0f 1f 44 00 00 49 8b 84 24=
 38 ff
>> =A0ff ff 48 85 c0 74 1b 48 8b 50 98 49 39 d5 72 12 48 03 50 a8 49 39=
 d5 73 09
>> =A0<0f> 0b eb fe 0f 1f 44 00 00 49 8b 94 24 30 ff ff ff 48 89 55 b8
>> =A0[13450.614082] RIP =A0[<ffffffff811d9537>]
>> =A0btrfs_ordered_update_i_size+0x237/0x3d0 [13450.614087] =A0RSP
>> =A0<ffff880071355da8>
>> [13450.614090] ---[ end trace 74172209f4d15206 ]---
>
> Sounds like the bug reported here: http://article.gmane.org/gmane.com=
p.file-
> systems.btrfs/4332/match=3Dbtrfs+fails+randomly. Can you try the patc=
h provided
> in that thread?
>
I have sent a patch for this.
http://www.mail-archive.com/linux-btrfs@vger.kernel.org/msg03686.html

Regards
Yan, Zheng