lkml.org 
[lkml]   [2014]   [Feb]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: kmemleak or crc32_le bug?
On Thu, Feb 06, 2014 at 05:04:49PM +0100, Bernd Schubert wrote:
> I'm frequently getting
>
> UG: unable to handle kernel paging request at ffff880f87550dc0
> IP: [<ffffffff813016d0>] crc32_le+0x30/0x110
>
> called from kmemleak, see bottom of the message.
...
> With the "Cannot allocate a kmemleak_object structure" messages,

Just curious, is the free memory low when this happens?

> somehow looks like object is not proper initialized, but update_checksum()
> checks for that. Hmm, I'm not sure about kmemcheck_shadow_lookup(),
> especially about
>
> > if (!virt_addr_valid(address))
> > return NULL;
>
> So is the test
>
> > shadow = kmemcheck_shadow_lookup(addr);
> > if (!shadow)
> > return true;
>
> right here? Shouldn't that be 'return false'?

Are you using kmemcheck and kmemleak together?

I don't think update_checksum() is called on the object being allocated
but possibly on an object being freed when kmemleak_scan() is running.
This is generally a safe operation because of the object locks in
kmemleak_scan() but when an error condition just occurred (like kmemleak
not being able to allocate memory), kmemleak gets disabled and
kmemleak_free() no longer passes the information down to update the
object's flags. At this point, the running kmemleak_scan() potentially
reads unmmapped objects.

I need to think a bit more about this. Thanks for reporting.

--
Catalin


\
 
 \ /
  Last update: 2014-02-10 21:01    [W:0.873 / U:0.388 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site