LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* Linux 2.6.19.2 : Oops 
@ 2007-01-17 10:21 Nicolas Bareil
  2007-01-17 23:35 ` Greg KH
  0 siblings, 1 reply; 4+ messages in thread
From: Nicolas Bareil @ 2007-01-17 10:21 UTC (permalink / raw)
  To: linux-kernel


Hello,

Since 2.6.19, I get the following Oops once a day, always with the same
process, newspipe[1] which use a lot of CPU, threads and I/O.

The kernel is patched by Grsecurity. The ext3 filesystem is on a
software RAID device (the two disks are SATA2). I tested the 
hardware (RAM, SMART disks) but nothing seem problematic.

kernel: BUG: unable to handle kernel NULL pointer dereference at virtual address 00000004
kernel: printing eip:
kernel: b0176344
kernel: *pgd =    0
kernel: *pmd =    0
kernel: Oops: 0000 [#1]
kernel: CPU:    0
kernel: EIP:    0060:[<b0176344>]    Not tainted VLI
kernel: EFLAGS: 00210286   (2.6.19.2-grsec #1)
kernel: EIP is at do_generic_mapping_read+0x34/0x4e0
kernel: eax: d8deeb50   ebx: 0000001b   ecx: 00000000   edx: ed6e1df4
kernel: esi: 00000000   edi: ed6e1e1c   ebp: e6235a40   esp: ed6e1dac
kernel: ds: 0068   es: 0068   ss: 0068
kernel: Process python (pid: 13360, ti=ed6e0000 task=daf1f030 task.ti=ed6e0000)
kernel: Stack: d8deeb50 00000001 00000000 00001000 00000010 00000000 00000000 ed6e1df4 
kernel: d8deeb50 00000010 00000000 00000010 00000010 0000000e 000105a5 00000000 
kernel: 00000000 00001000 00000000 00000020 00000000 00000020 0000000f 00000000 
kernel: Call Trace:
kernel: [<b01788f3>] generic_file_aio_read+0xf3/0x230
kernel: [<b0175a80>] file_read_actor+0x0/0x130
kernel: [<b019444d>] do_sync_read+0xed/0x130
kernel: [<b016a6d0>] autoremove_wake_function+0x0/0x60
kernel: [<b01848fc>] __vma_link+0x3c/0x70
kernel: [<b0185318>] vma_link+0x38/0xc0
kernel: [<b0194ed4>] vfs_read+0xd4/0x1c0
kernel: [<b019539b>] sys_read+0x4b/0x80
kernel: [<b0140047>] syscall_call+0x7/0xb
kernel: [<b014005f>] restore_all+0x0/0x18
kernel: =======================
kernel: Code: 83 ec 70 8b 84 24 84 00 00 00 8d 54 24 48 89 d7 8b b4 24 88 00 00 00 fc 8b 00 89 54 24 1c 89 44 24 20 f3 a5 8b b4 24 90 00 00 00 <8b> 5e 04 8b 0e 8b b4 24 94 00 00 00 89 da c1 fa 0c 89 c8 89 54 
kernel: EIP: [<b0176344>] do_generic_mapping_read+0x34/0x4e0 SS:ESP 0068:ed6e1dac

My config is available on http://chdir.org/~nbareil/config-2.6.19.2-grsec.gz

Thank you!

Footnotes: 
[1]  http://newspipe.sourceforge.net/

-- 
Nicolas Bareil                                  http://chdir.org/~nico/
OpenPGP=0xAE4F7057 Fingerprint=34DB22091049FB2F33E6B71580F314DAAE4F7057


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Linux 2.6.19.2 : Oops
  2007-01-17 10:21 Linux 2.6.19.2 : Oops Nicolas Bareil
@ 2007-01-17 23:35 ` Greg KH
  2007-01-18  7:48   ` Nicolas Bareil
  2007-01-22 19:37   ` Nicolas Bareil
  0 siblings, 2 replies; 4+ messages in thread
From: Greg KH @ 2007-01-17 23:35 UTC (permalink / raw)
  To: Nicolas Bareil; +Cc: linux-kernel

On Wed, Jan 17, 2007 at 11:21:59AM +0100, Nicolas Bareil wrote:
> 
> Hello,
> 
> Since 2.6.19, I get the following Oops once a day, always with the same
> process, newspipe[1] which use a lot of CPU, threads and I/O.
> 
> The kernel is patched by Grsecurity. The ext3 filesystem is on a
> software RAID device (the two disks are SATA2). I tested the 
> hardware (RAM, SMART disks) but nothing seem problematic.

Can you reproduce it without the grsec patch applied?

thanks,

greg k-h

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Linux 2.6.19.2 : Oops
  2007-01-17 23:35 ` Greg KH
@ 2007-01-18  7:48   ` Nicolas Bareil
  2007-01-22 19:37   ` Nicolas Bareil
  1 sibling, 0 replies; 4+ messages in thread
From: Nicolas Bareil @ 2007-01-18  7:48 UTC (permalink / raw)
  To: linux-kernel


On 18 Jan 2007, Greg KH wrote:
> > Since 2.6.19, I get the following Oops once a day, always with the same
> > process, newspipe[1] which use a lot of CPU, threads and I/O.
> > ...
> Can you reproduce it without the grsec patch applied?

I'm compiling a new kernel and will try it soon!

-- 
Nicolas Bareil                                  http://chdir.org/~nico/
OpenPGP=0xAE4F7057 Fingerprint=34DB22091049FB2F33E6B71580F314DAAE4F7057


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Linux 2.6.19.2 : Oops
  2007-01-17 23:35 ` Greg KH
  2007-01-18  7:48   ` Nicolas Bareil
@ 2007-01-22 19:37   ` Nicolas Bareil
  1 sibling, 0 replies; 4+ messages in thread
From: Nicolas Bareil @ 2007-01-22 19:37 UTC (permalink / raw)
  To: linux-kernel

Greg KH <greg@kroah.com> writes:
>> Since 2.6.19, I get the following Oops once a day, always with the same
>> process, newspipe[1] which use a lot of CPU, threads and I/O.
>> 
>> The kernel is patched by Grsecurity. The ext3 filesystem is on a
>> software RAID device (the two disks are SATA2). I tested the 
>> hardware (RAM, SMART disks) but nothing seem problematic.
>
> Can you reproduce it without the grsec patch applied?

After 4 days of tests without grsec, I didn't had any Oops. So it seems
I will have to find this bug with the grsecurity developpers...

Thank you!
-- 
Nicolas Bareil                                  http://chdir.org/~nico/
OpenPGP=0xAE4F7057 Fingerprint=34DB22091049FB2F33E6B71580F314DAAE4F7057


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2007-01-22 19:38 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-01-17 10:21 Linux 2.6.19.2 : Oops Nicolas Bareil
2007-01-17 23:35 ` Greg KH
2007-01-18  7:48   ` Nicolas Bareil
2007-01-22 19:37   ` Nicolas Bareil

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).