Re: [PATCH] kmemcheck: Test the full object in kmemcheck_is_obj_initialized()

[Pekka Enberg <penberg@cs.helsinki.fi>]

On Mon, Feb 8, 2010 at 2:58 PM, Catalin Marinas <catalin.marinas@arm.com> wrote:
> On Mon, 2010-02-08 at 12:54 +0000, Pekka Enberg wrote:
>> On Mon, Feb 8, 2010 at 1:16 PM, Catalin Marinas <catalin.marinas@arm.com> wrote:
>> > This is a fix for bug #14845 (bugzilla.kernel.org). The
>> > update_checksum() function in mm/kmemleak.c calls
>> > kmemcheck_is_obj_initialised() before scanning an object. When
>> > KMEMCHECK_PARTIAL_OK is enabled, this function returns true. However,
>> > the crc32_le() reads smaller intervals (32-bit) for which
>> > kmemleak_is_obj_initialised() may be false leading to a kmemcheck
>> > warning.
>> >
>> > Note that kmemcheck_is_obj_initialized() is currently only used by
>> > kmemleak before scanning a memory location.
>> >
>> > Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
>> > Cc: Andrew Morton <akpm@linux-foundation.org>
>> > Cc: Christian Casteyde <casteyde.christian@free.fr>
>> > Cc: Pekka Enberg <penberg@cs.helsinki.fi>
>> > Cc: Vegard Nossum <vegardno@ifi.uio.no>
>>
>> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
>>
>> Andrew, Vegard has been rather quiet lately so do you mind picking up
>> this patch? Alternatively, I can pick it up in slab.git if you so
>> prefer.
>
> It's been pretty late indeed, sorry about that (I mostly work on ARM
> hardware where kmemcheck isn't available and couldn't test the patch
> earlier).
>
> Anyway, IMHO it's not critical to be merged in 2.6.33 but it's up to
> you.

I think it's definitely .34 material but probably should go to -stable.

                        Pekka