[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

[Linux-cachefs] Another bug with fscache ?



Dear David,

I don't know if it is completely related to fscache but today I found was of 
my server in a bad shape with the following stack-trace in the kern.log :


[336115.332070] ------------[ cut here ]------------                                                                                          
[336115.332102] kernel BUG at 
/usr/src/linux-2.6-2.6.33/debian/build/source_i386_none/lib/radix-tree.c:602!                                   
[336115.332154] invalid opcode: 0000 [#1] SMP                                                                                                 
[336115.332186] last sysfs file: /sys/devices/virtual/block/md1/md/mismatch_cnt                                                               
[336115.332216] CPU 5                                                                                                                         
[336115.332239] Pid: 19372, comm: lighttpd Tainted: G    B      2.6.33-2-amd64 
#1 0P010H/PowerEdge M600                                       
[336115.332287] RIP: 0010:[<ffffffff81182040>]  [<ffffffff81182040>] 
radix_tree_tag_get+0xbc/0xe3                                             
[336115.332341] RSP: 0018:ffff8800cfcc1c28  EFLAGS: 00210202                                                                                  
[336115.332367] RAX: 0000000000000001 RBX: 000000000000003f RCX: 
0000000000000000                                                             
[336115.332411] RDX: 000000000000003f RSI: 0000000000000001 RDI: 
ffff880347d253b0                                                             
[336115.332454] RBP: ffff880347d253b0 R08: 0000000000000efe R09: 
0000000000000000                                                             
[336115.332498] R10: 0000000000000001 R11: ffffffffa0350cf6 R12: 0000000000000001
[336115.332541] R13: 0000000000000000 R14: 0000000000000001 R15: 
0000000000000eff
[336115.332585] FS:  0000000000000000(0000) GS:ffff88000ff40000(0063) 
knlGS:00000000f731b6d0
[336115.332630] CS:  0010 DS: 002b ES: 002b CR0: 000000008005003b
[336115.332667] CR2: 0000000009652000 CR3: 00000001a601f000 CR4: 
00000000000006e0
[336115.332710] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 
0000000000000000
[336115.332758] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[336115.332807] Process lighttpd (pid: 19372, threadinfo ffff8800cfcc0000, task 
ffff8803e4307000)
[336115.332852] Stack:
[336115.332872]  0000000100000001 ffff88042c586850 ffffea0001b620c8 
00000000000000d0
[336115.332906] <0> ffff88042c586820 000000000000000b ffff8800cfcc1d50 
ffffffffa0247b67
[336115.332957] <0> 0000000200000002 ffff88042c586820 ffffea0001b620c8 
00000000000000d0
[336115.333025] Call Trace:
[336115.333052]  [<ffffffffa0247b67>] ? __fscache_maybe_release_page+0x42/0x115 
[fscache]
[336115.333116]  [<ffffffffa0372e7d>] ? nfs_fscache_release_page+0x66/0x99 [nfs]
[336115.333148]  [<ffffffff810b6dee>] ? invalidate_inode_pages2_range+0x15a/0x262
[336115.333197]  [<ffffffffa035312f>] ? nfs_invalidate_mapping_nolock+0x18/0xb4 
[nfs]
[336115.333246]  [<ffffffffa0354097>] ? nfs_revalidate_mapping+0x85/0x99 [nfs]
[336115.333273]  [<ffffffffa0351158>] ? nfs_file_splice_read+0x5b/0x8e [nfs]
[336115.333273]  [<ffffffff811043d3>] ? splice_direct_to_actor+0xbe/0x188
[336115.333273]  [<ffffffff81104a1c>] ? direct_splice_actor+0x0/0x1e
[336115.333273]  [<ffffffff81113274>] ? ep_scan_ready_list+0x132/0x151
[336115.333273]  [<ffffffff811044e7>] ? do_splice_direct+0x4a/0x64
[336115.333273]  [<ffffffff810e8fa8>] ? do_sendfile+0x12d/0x1a8
[336115.333273]  [<ffffffff8106685b>] ? getnstimeofday+0x55/0xaf
[336115.333273]  [<ffffffff810e906c>] ? sys_sendfile64+0x49/0x88
[336115.333273]  [<ffffffff8103145f>] ? sysenter_dispatch+0x7/0x2e
[336115.333273] Code: 00 00 44 0f 44 f0 41 83 fc 01 75 23 8b 74 24 04 89 da 48 
89 ef e8 be fb ff ff 85 c0 0f 95 c0 0f b6 c0 85 c0 74 1f 45 85 f6 74 1a <0f> 0b 
eb fe 48 63 c3 41 83 ed 06 41 ff cc 48 8d 44 c5 18 48 8b
[336115.333273] RIP  [<ffffffff81182040>] radix_tree_tag_get+0xbc/0xe3
[336115.333273]  RSP <ffff8800cfcc1c28>
[336115.334487] ---[ end trace 2da6bc1f424cd8a1 ]---

The lighttpd process (which is the one accessing file over nfs cached by 
fscache) was stalled and impossible to kill.

I had to reboot the server.

Maybe it rings a bell to you ?

Regards,

-- 
RD


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]