Re: oops in bk pull (oct 03)

[Linus Torvalds]

In article <[e-mail]>,
Paul E. Erkkila <[e-mail]> wrote:
>
>Oops in drivers/pci/probe.c
>
>Oops (copied), ksymoops, and lspci -vv attached
>
>No modules in ksyms, skipping objects
>Unable to handle kernel paging request at virtual address f8000008
>c01c9d10
>*pde = 00000000
>Oops: 0002
>CPU:    0
>EIP:    0060:[<c01c9d10>]    Not tainted
>Using defaults from ksymoops -t elf32-i386 -a i386
>EFLAGS: 00010202
>eax: f8000008   ebx: 00000010     ecx: 00000000       edx: f8000008
>esi: c1523c00   edi: c1523d38     ebp: 00000001       esp: dffcdb60
>ds: 0068    es: 0068    ss: 0068
>Stack:  c01c9e91 f8000008 fffffff0 00000010 dffcdb78 c1523ef8 f8000008 00000008
>        d0000008 c1523c00 c1523f48 00000000 00000000 c01ca2a6 c1523c00 00000006
>        00000030 dffcdbac 00000000 00000600 c1523c00 dffcdc20 c03a1351 c1523c00
> [<c01c9e91>] pci_read_bases+0x161/0x340
> [<c01ca2a6>] pci_setup_device+0x1b6/0x3d0
> [<c0105109>] init+0x79/0x200
> [<c0105090>] init+0x0/0x200
> [<c01073e5>] kernel_thread_helper+0x5/0x10
>Code: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 48 c3 8d b4

Something has corrupted your kernel image. Those 16 0x00 bytes are
definitely not the right code, looks like an errant memset() through a
wild pointer cleared it or something.

Is this repeatable? Does it happen with current BK?

		Linus