Next May 11 : BUG during scsi initialization

[Matthew Wilcox]

On Mon, May 11, 2009 at 05:34:07PM +0530, Sachin Sant wrote:
> Matthew Wilcox wrote:
>> On Mon, May 11, 2009 at 05:16:10PM +0530, Sachin Sant wrote:
>>   
>>> Today's Next tree failed to boot on a Power6 box with following BUG :
>>
>> This doesn't actually appear to be a SCSI bug ... it looks like SCSI tried
>> to allocate memory and things went wrong in the memory allocator:
>>
>> [c0000000c7d038b0] [c0000000005d67d8] ._spin_lock+0x10/0x24
>> [c0000000c7d03920] [c00000000013fbdc] .__slab_alloc_page+0x344/0x3cc
>> [c0000000c7d039e0] [c000000000141168] .kmem_cache_alloc+0x13c/0x21c
>> [c0000000c7d03aa0] [c000000000141b04] .kmem_cache_create+0x294/0x2a8
>> [c0000000c7d03b90] [d000000000ea14cc] .scsi_init_queue+0x38/0x170 [scsi_mod]
>>
>> Which memory allocator did you have selected (SLAB, SLUB, SLOB, SLQB)?
>>   
> Default one. SLQB
>
> CONFIG_SLQB_ALLOCATOR=y
> CONFIG_SLQB=y
>
> Page size is 64K with Config DEBUG_PAGEALLOC set.
>
> CONFIG_PPC_HAS_HASH_64K=y
> CONFIG_PPC_64K_PAGES=y
> CONFIG_DEBUG_PAGEALLOC=y

Hm.  We've seen some similar problems at Intel while doing database
performance tests with SLQB.  Any ideas, Nick?

-- 
Matthew Wilcox				Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."