ata workqueue corruption

From: Johannes Berg
Date: Fri Feb 23 2007 - 06:07:26 EST


While debugging why I keep having problems with suspend to disk on my
quad with more recent kernels (2.6.20 worked fine), I managed to capture
this weird ata workqueue corruption via netconsole, just a bit later the
system froze (I was in X but this seems like a reason to panic):

[ 243.981215] ata/0 S 0000000000000000 0 80 19 81 77 (L-TLB)
[ 243.981225] Call Trace:
[ 243.981231] [C00000007EC63AA0] [C000000000529B18] devinet_sysctl+0xb8/0x968 (unreliable)
[ 243.981245] [C00000007EC63C70] [C00000000000FB28] .__switch_to+0xe4/0x118
[ 243.981257] [C00000007EC63D00] [C00000000035BEA4] .schedule+0x73c/0x894
[ 243.981267] [C00000007EC63E00] [C0000000000597C8] .worker_thread+0x120/0x1b4
[ 243.981278] [C00000007EC63EE0] [C00000000005E4D8] .kthread+0x11c/0x16c
[ 243.981289] [C00000007EC63F90] [C000000000021780] .kernel_thread+0x4c/0x68
[ 243.981298] ata/1 S 0000000000000000 0 81 19 82 80 (L-TLB)
[ 243.981308] Call Trace:
[ 243.981313] [C00000007EC67AA0] [5A5A5A5A5A5A5A5A] 0x5a5a5a5a5a5a5a5a (unreliable)
[ 243.981325] [C00000007EC67C70] [C00000000000FB28] .__switch_to+0xe4/0x118
[ 243.981336] [C00000007EC67D00] [C00000000035BEA4] .schedule+0x73c/0x894
[ 243.981347] [C00000007EC67E00] [C0000000000597C8] .worker_thread+0x120/0x1b4
[ 243.981357] [C00000007EC67EE0] [C00000000005E4D8] .kthread+0x11c/0x16c
[ 243.981368] [C00000007EC67F90] [C000000000021780] .kernel_thread+0x4c/0x68
[ 243.981377] ata/2 S 0000000000000000 0 82 19 84 81 (L-TLB)
[ 243.981387] Call Trace:
[ 243.981393] [C00000007EC6BAA0] [5A5A5A5A5A5A5A5A] 0x5a5a5a5a5a5a5a5a (unreliable)
[ 243.981404] [C00000007EC6BC70] [C00000000000FB28] .__switch_to+0xe4/0x118
[ 243.981415] [C00000007EC6BD00] [C00000000035BEA4] .schedule+0x73c/0x894
[ 243.981426] [C00000007EC6BE00] [C0000000000597C8] .worker_thread+0x120/0x1b4
[ 243.981437] [C00000007EC6BEE0] [C00000000005E4D8] .kthread+0x11c/0x16c
[ 243.981448] [C00000007EC6BF90] [C000000000021780] .kernel_thread+0x4c/0x68
[ 243.981456] ata_aux S 0000000000000000 0 84 19 107 82 (L-TLB)
[ 243.981466] Call Trace:
[ 243.981472] [C00000007EC77AA0] [5A5A5A5A5A5A5AA5] 0x5a5a5a5a5a5a5aa5 (unreliable)
[ 243.981483] [C00000007EC77C70] [C00000000000FB28] .__switch_to+0xe4/0x118
[ 243.981495] [C00000007EC77D00] [C00000000035BEA4] .schedule+0x73c/0x894
[ 243.981505] [C00000007EC77E00] [C0000000000597C8] .worker_thread+0x120/0x1b4
[ 243.981516] [C00000007EC77EE0] [C00000000005E4D8] .kthread+0x11c/0x16c
[ 243.981527] [C00000007EC77F90] [C000000000021780] .kernel_thread+0x4c/0x68

Unfortunately I have no more than this to offer and it doesn't seem to
be reproducible.

johannes

Attachment: signature.asc
Description: This is a digitally signed message part