Re: Oops at __bio_clone with 2.6.16-rc6 anyone??????

[Andrew Morton]

"Don Dupuis" <dondster@xxxxxxxxx> wrote:
>
> Yes it does also happen on 2.6.16
> 

(please don't top-post).

> 
> On 3/27/06, Andrew Morton <akpm@xxxxxxxx> wrote:
> > "Don Dupuis" <dondster@xxxxxxxxx> wrote:
> > >
> > > I will get this oops during reboots. It doesn't happen everytime, but
> > > It happens on this system at least 1 to 2 out of 10 reboots. The
> > > machine is a Dell Powervault 745n. Here is the oops output:
> > >
> > > Mar 20 22:27:49 (none) kernel: EXT3-fs: mounted filesystem with
> > > journal data mode.
> > > Mar 20 22:27:49 (none) kernel: Unable to handle kernel paging request
> > > at virtual address f8000000
> > > Mar 20 22:27:49 (none) kernel: printing eip:
> > > Mar 20 22:27:49 (none) kernel: c0156db1
> > > Mar 20 22:27:49 (none) kernel: *pde = 00000000
> > > Mar 20 22:27:49 (none) kernel: Oops: 0000 [#1]
> > > Mar 20 22:27:49 (none) kernel: SMP
> > > Mar 20 22:27:49 (none) kernel: Modules linked in:
> > > Mar 20 22:27:49 (none) kernel: CPU: 0
> > > Mar 20 22:27:50 (none) kernel: EIP: 0060:[<c0156db1>] Not tainted VLI
> > > Mar 20 22:27:50 (none) kernel: EFLAGS: 00010206 (2.6.16-rc6 #3)
> > > Mar 20 22:27:50 (none) kernel: EIP is at __bio_clone+0x29/0x9b
> > > Mar 20 22:27:50 (none) kernel: eax: 00000300 ebx: f68f3700 ecx:
> > > 00000002 edx: f7fffc80
> > > Mar 20 22:27:50 (none) kernel: esi: f8000000 edi: f7f3d378 ebp:
> > > f7c44b98 esp: f7c44b84
> > > Mar 20 22:27:50 (none) kernel: ds: 007b es: 007b ss: 0068
> > > Mar 20 22:27:50 (none) kernel: Process ldconfig (pid: 581,
> > > threadinfo=f7c44000 task=f7db9070)
> > > Mar 20 22:27:50 (none) kernel: Stack: <0>f7d3b458 f68f3700 f68f3700
> > > f7fffc80 f65b4640 f7c44ba8 c0156e4e f7d4c664
> > > Mar 20 22:27:50 (none) kernel: 00000010 f7c44bf4 c02c8346
> > > 00000080 00000000 00000e00 c0154b1b 00000000
> > > Mar 20 22:27:50 (none) kernel: 0000007f 00000080 f7fffc80
> > > f7d4a740 f7d44400 f7fffc80 f7d3b458 c01579c3
> > > Mar 20 22:27:50 (none) kernel: Call Trace:
> > > Mar 20 22:27:50 (none) kernel: [<c0104260>] show_stack_log_lvl+0xa8/0xb0
> > > Mar 20 22:27:50 (none) kernel: [<c0104397>] show_registers+0x109/0x171
> > > Mar 20 22:27:50 (none) kernel: [<c010456e>] die+0xfb/0x16f
> > > Mar 20 22:27:50 (none) kernel: [<c0114750>] do_page_fault+0x359/0x48b
> > > Mar 20 22:27:50 (none) kernel: [<c0103f0b>] error_code+0x4f/0x54
> > > Mar 20 22:27:50 (none) kernel: [<c0156e4e>] bio_clone+0x2b/0x31
> > > Mar 20 22:27:50 (none) kernel: [<c02c8346>] make_request+0x208/0x3d4
> > > Mar 20 22:27:50 (none) kernel: [<c02c8211>] make_request+0xd3/0x3d4
> > > Mar 20 22:27:50 (none) kernel: [<c01d3b68>] generic_make_request+0xf5/0x105
> > > Mar 20 22:27:50 (none) kernel: [<c01d3c19>] submit_bio+0xa1/0xa9
> > > Mar 20 22:27:50 (none) kernel: [<c0170453>] mpage_bio_submit+0x1c/0x21
> > > Mar 20 22:27:50 (none) kernel: [<c017085c>] do_mpage_readpage+0x30b/0x44d
> > > Mar 20 22:27:50 (none) kernel: [<c0170a2b>] mpage_readpages+0x8d/0xf1
> > > Mar 20 22:27:50 (none) kernel: [<c01a7ee7>] ext3_readpages+0x14/0x16
> > > Mar 20 22:27:50 (none) kernel: [<c013e92f>] read_pages+0x26/0xc6

Jens points out that the "dm: bio split bvec fix" patch which went into
2.6.16.1 might fix this.  Can you try it please?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/