ide-scsi bug hard locks machine

Hi all, heres the kernel logs from my machine which hard locks.

If you need more info email me directly I'm not subscribed.

Thanks.

ptelnet log 1: 13 Jan 2003 1:51 pm
ide-scsi: abort called for 2
bad: scheduling while atomic!

Call Trace: [<c0116941>]  [<c0109b76>]  [<c01169b0>]  [<c011a5db>]  [<c0109d8c>]  [<c0263100>]  [<c0262730>]  [<c026848c>]  [<c0268390>]  [<c0261e63>]  [<c0261cc0>]  [<c0261c80>]  [<c026225d>]  [<c02622d7>]  [<c0262b65>]  [<c0109d97>]  [<c0262c89>]  [<c0262bb0>]  [<c0108be9>] 

hde: lost interrupt
ide-scsi: CoD != 0 in idescsi_pc_intr
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }
ide-scsi: reset called for 2

------------[ cut here ]------------
kernel BUG at drivers/scsi/ide-scsi.c:483!
invalid operand: 0000
CPU:    0
EIP:    0060:[<c0267845>]    Not tainted
EFLAGS: 00010286
eax: c024ab50   ebx: c0448d84   ecx: edae15ed   edx: 0000d002
esi: d7c96800   edi: d7c93bc0   ebp: 00000000   esp: d7c4be5c
ds: 007b   es: 007b   ss: 0068
Process scsi_eh_0 (pid: 11, threadinfo=d7c4a000 task=d7c498c0)
Stack: 0000d002 c0448d84 00000008 00000080 0000001e c0448d84 c0448d84 00000000 
       d7d023c0 c0246bd9 c0448d84 d7c93bc0 00000000 00000088 0000001e d7c4a000 
       d7d023c0 c0448d84 d7d7eb80 c0246dd9 c0448d84 d7d023c0 c0448e2c d7c4a000 
Call Trace: [<c0246bd9>]  [<c0246dd9>]  [<c02474ff>]  [<c0247443>]  [<c026825b>]  [<c0261dbc>]  [<c0261cc0>]  [<c0261c80>]  [<c026225d>]  [<c0262417>]  [<c02622d7>]  [<c0262b79>]  [<c0109d97>]  [<c0262c89>]  [<c0262bb0>]  [<c0108be9>] 
Code: 0f 0b e3 01 2b 07 38 c0 8b 57 38 a1 c8 d0 42 c0 89 d1 29 c1 
 hde: ATAPI reset complete

hde: status error: status=0x48 { DriveReady DataRequest }
hde: drive not ready for command
hde: lost interrupt
ide-scsi: CoD != 0 in idescsi_pc_intr
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }
hde: ATAPI reset complete
hde: irq timeout: status=0xc0 { Busy }

Jason Thomas                           Phone:  +61 2 6257 7111
Unix System Administrator              Fax:    +61 2 6257 7311
tSA Consulting Group Pty. Ltd.         Mobile: 0418 29 66 81
1 Hall Street Lyneham ACT 2602         http://www.topic.com.au/

Someone broke ide-scsi in 2.5. I've not had time to investigate why yet. If you need ide-scsi use 2.4.

Some more info for those interested, it seems to have started in 2.5.53 with the implementation of the mid level api.

heres a trace from 2.5.53 maybe this will make it easier to find.

Unable to handle kernel NULL pointer dereference at virtual address 00000030
 printing eip:
c0252ba3
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0060:[<c0252ba3>]    Not tainted
EFLAGS: 00010086

EIP is at idescsi_abort+0x73/0x110
eax: 00000002 ebx: 00000086 ecx: d7cc5200 edx: 00000000
esi: d7c3e000 edi: d7c3e000 ebp: 00000007 esp: d7c3ff60
ds: 0068 es: 0068 ss: 0068
Process scsi_eh_0 (pid: 9, threadinfo=d7c3e000 task=d7d1a680)

Stack: c030dee0 00000002 d7c3e000 d7c8c980 c040e1ec d7c3e000 00000202 d7c3e000
       00000000 c024c7a2 d7cc5200 d7cc5200 d7cdf180 d7c3e000 c024c8c8 d7cc5200
       d7cdf180 d7cdf180 d7c3ffd4 c024d165 d7cc5200 d7cdf180 c0109d97 d7cc5200
Call Trace:
 [<c024c7a2>] scsi_try_to_abort_cmd+0x62/0x80
 [<c024c8c8>] scsi_eh_abort_cmd+0x38/0x70
 [<c024d165>] scsi_unjam_host+0xa5/0xf0
 [<c0109d97>] __down_failed_interruptible+0x7/0xc
 [<c024d289>] scsi_error_handler+0xd9/0x110
 [<c024d1b0>] scsi_error_handler+0x0/0x110
 [<c0108bfd>] kernel_thread_helper+0x5/0x18

Code: 39 42 30 74 67 8b 44 24 10 83 c0 10 89 04 24 e8 69 55 fc ff
 <6>note: scsi_eh_0[9] exited with preempt_count 2