xfs oops with full partition

[Dmitry Melekhov <dm@xxxxxxxxxxxxxxxx>]

Hello!

I don't know which xfs version is in suse kernel.

We get following oopses when users write to 100% full partition:

Unable to handle kernel NULL pointer dereference at virtual address 00000004
f945a5d4
*pde = 00000000
Oops: 0002 [#1]
CPU:    0
EIP:    0060:[<f945a5d4>]    Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010286   (2.6.5-7.201-smp SLES9_SP2_BRANCH-200508250620450000)
eax: 00000000   ebx: ffffffff   ecx: fffffc19   edx: ffffffff
esi: cafe2218   edi: e2f47648   ebp: f60bb018   esp: cdc83e10
ds: 007b   es: 007b   ss: 0068
Stack: d7c76804 e2f47648 f60bb000 000017d4 00001cea 00000000 00000000 
000017d4
   e2f47648 000017d4 00001cea f9459278 000017d4 00001cea 00000000 00000001
   f60bb000 ee7085f0 000017d4 00001cea ee7085e0 000017cc 00001cea 000017d4
Call Trace:
[<f9459278>] xfs_trans_chunk_committed+0x168/0x200 [xfs]
[<f945934c>] xfs_trans_committed+0x3c/0x100 [xfs]
[<f944c16c>] xlog_state_do_callback+0x24c/0x3e0 [xfs]
[<c012677a>] schedule+0x65a/0xe40
[<f944c39f>] xlog_iodone+0x9f/0x130 [xfs]
[<f9472257>] pagebuf_iodone_work+0x27/0x40 [xfs]
[<c013bfe7>] worker_thread+0x187/0x230
[<f9472230>] pagebuf_iodone_work+0x0/0x40 [xfs]
[<c0123e80>] default_wake_function+0x0/0x10
[<c0123e80>] default_wake_function+0x0/0x10
[<c013be60>] worker_thread+0x0/0x230
[<c013fc54>] kthread+0xd4/0x118
[<c013fb80>] kthread+0x0/0x118
[<c0107005>] kernel_thread_helper+0x5/0x10
Code: 89 78 04 8b 44 24 08 ff 40 20 8b 54 24 04 39 14 24 74 3e b0


>>EIP; f945a5d4 <__crc_dma_pool_free+4de965/69c753>   <=====

>>ebx; ffffffff <__kernel_rt_sigreturn+1bbf/????>
>>ecx; fffffc19 <__kernel_rt_sigreturn+17d9/????>
>>edx; ffffffff <__kernel_rt_sigreturn+1bbf/????>
>>esi; cafe2218 <__crc_dev_change_flags+33aa1/20bfbe>
>>edi; e2f47648 <__crc_class_simple_device_remove+2911c/4eed5b>
>>ebp; f60bb018 <__crc___ip_select_ident+461513/65b400>
>>esp; cdc83e10 <__crc_fsync_buffers_list+34509f/856d01>

Trace; f9459278 <__crc_dma_pool_free+4dd609/69c753>
Trace; f945934c <__crc_dma_pool_free+4dd6dd/69c753>
Trace; f944c16c <__crc_dma_pool_free+4d04fd/69c753>
Trace; c012677a <schedule+65a/e40>
Trace; f944c39f <__crc_dma_pool_free+4d0730/69c753>
Trace; f9472257 <__crc_dma_pool_free+4f65e8/69c753>
Trace; c013bfe7 <worker_thread+187/230>
Trace; f9472230 <__crc_dma_pool_free+4f65c1/69c753>
Trace; c0123e80 <default_wake_function+0/10>
Trace; c0123e80 <default_wake_function+0/10>
Trace; c013be60 <worker_thread+0/230>
Trace; c013fc54 <kthread+d4/118>
Trace; c013fb80 <kthread+0/118>
Trace; c0107005 <kernel_thread_helper+5/10>

Code;  f945a5d4 <__crc_dma_pool_free+4de965/69c753>
00000000 <_EIP>:
Code;  f945a5d4 <__crc_dma_pool_free+4de965/69c753>   <=====
  0:   89 78 04                  mov    %edi,0x4(%eax)   <=====
Code;  f945a5d7 <__crc_dma_pool_free+4de968/69c753>
  3:   8b 44 24 08               mov    0x8(%esp),%eax
Code;  f945a5db <__crc_dma_pool_free+4de96c/69c753>
  7:   ff 40 20                  incl   0x20(%eax)
Code;  f945a5de <__crc_dma_pool_free+4de96f/69c753>
  a:   8b 54 24 04               mov    0x4(%esp),%edx
Code;  f945a5e2 <__crc_dma_pool_free+4de973/69c753>
  e:   39 14 24                  cmp    %edx,(%esp)
Code;  f945a5e5 <__crc_dma_pool_free+4de976/69c753>
 11:   74 3e                     je     51 <_EIP+0x51>
Code;  f945a5e7 <__crc_dma_pool_free+4de978/69c753>
 13:   b0 00                     mov    $0x0,%al



Is this problem already fixed in lates xfs code?