2.6.0-test8 oops in find_inode_fast

Just got this on my laptop. It's not too reproducable, as I've been running this kernel since it came out. I'll switch to 2.6.0-test8-bk in a bit.

Anything anyone wants me to try?

thanks,

greg k-h

Unable to handle kernel paging request at virtual address 0031314d
 printing eip:
c0162596
*pde = 00000000
Oops: 0000 [#1]
CPU:    0
EIP:    0060:[<c0162596>]    Not tainted
EFLAGS: 00010206

EIP is at find_inode_fast+0x26/0x60
eax: 0031314d ebx: cb70d164 ecx: 00007fff edx: 0031314d
esi: 000be80d edi: d650e000 ebp: d6f7f224 esp: c4f45d70
ds: 007b es: 007b ss: 0068
Process ls (pid: 7439, threadinfo=c4f44000 task=c817a0c0)

Stack: 000be80d d6f7f224 000be80d d650e000 c01629db d650e000 d6f7f224 000be80d 
       c36ea9d0 00000000 c744b900 00000246 fffffff4 c4f45f00 c013c225 c02a0507 
       000be80d d650e000 c744b720 cc472320 c017ed30 d650e000 000be80d cc67e438 
Call Trace:
 [<c01629db>] iget_locked+0x3b/0x80
 [<c013c225>] kmem_cache_alloc+0x25/0x60
 [<c017ed30>] ext3_lookup+0x50/0xb0
 [<c0160f3b>] d_alloc+0x1b/0x1b0
 [<c0161369>] d_lookup+0x29/0x50
 [<c0158652>] real_lookup+0x62/0xd0
 [<c015888a>] do_lookup+0x4a/0x90
 [<c0158dd6>] link_path_walk+0x506/0x830
 [<c0167b07>] __mark_inode_dirty+0x27/0xa0
 [<c0141c60>] handle_mm_fault+0x70/0x110
 [<c0178946>] ext3_readdir+0x486/0x4a0
 [<c013c225>] kmem_cache_alloc+0x25/0x60
 [<c015836e>] getname+0x5e/0xa0
 [<c01594e4>] __user_walk+0x24/0x40
 [<c0155014>] vfs_lstat+0x14/0x50
 [<c01556e1>] sys_lstat64+0x11/0x30
 [<c0108fb3>] syscall_call+0x7/0xb