Kernel 2.6.0-1.37 hang on boot w/ Fusion MPT SCSI driver

[mark]

Dave Jones wrote:

>On Sat, 2004-01-17 at 19:03, mark wrote:
>
>  
>
>>mptscsih: ioc0: >>  Attempting task abort! (sc=f7170864, numIOs=1)
>>mptscsih: ioc0: >>  Attempting target reset! (sc=f7170864, numIOs=1)
>>...
>>...
>>
>>etc. and requiring a power cycle to reset.  Same result w/  'noapic'
>>
>>Any insights / work arounds before I bugzilla?
>>    
>>
>
>Try and repeat with the latest kernel. Yesterdays update merged -bk4
>which had a merge from the scsi maintiainers repository.
>
>
>  
>
Ok will do as soon as possible.   Im on the road away from the box for a 
couple days.  BTW though, I see this particular LSI / MPT driver is 
under the drivers/message/fusion  tree and not under scsi

>>Unable to handle kernel NULL pointer dereference at virtual address 
>>00000000
>>printing eip:
>>c0226f3e
>>*pde = 00000000
>>Oops: 0000 [#1]
>>CPU:    0
>>EIP:    0060:[<c0226f3e>]    Not tainted
>>EFLAGS: 00010282
>>EIP is at vt_ioctl+0x1e/0x20c0
>>eax: 00000000   ebx: f6772000   ecx: f0e3be38   edx: bff6528c
>>esi: 00005401   edi: f6772000   ebp: f478ff68   esp: f0e3be98
>>ds: 007b   es: 007b   ss: 0068
>>Process rc (pid: 1239, threadinfo=f0e3a000 task=f49029b0)
>>Stack: 00000000 00000000 c01234f5 f7f67854 f444a2e4 f0e3bf14 00000001 
>>c01dc53c
>>     f49029b0 f4aabe88 00000001 f444a2f8 f0e3bef8 f45c6db0 f4aabe88 
>>f4706e10
>>     f49029b0 0052c9a0 00000000 fffe54b0 f45c5004 c01238d1 f45c6dec 
>>00000001
>>Call Trace:
>>[<c01234f5>] __change_page_attr+0x25/0x1e0
>>[<c01dc53c>] selinux_file_ioctl+0x10c/0x370
>>[<c01238d1>] kernel_map_pages+0x31/0x6c
>>[<c0226f20>] vt_ioctl+0x0/0x20c0
>>[<c0220c79>] tty_ioctl+0x409/0x510
>>[<c0187450>] sys_ioctl+0x170/0x330
>>[<c017011e>] sys_open+0x7e/0x90
>>[<c010beef>] syscall_call+0x7/0xb
>>
>>Code: 8b 18 89 1c 24 8b 04 9d 80 ef 44 c0 89 44 24 50 e8 0d 73 00
>>    
>>
>
>This is also interesting btw, can you bugzilla this please ?
>Add jmorris at redhat.com to the Cc: line too.
>
>  
>
Will do.