ShiYong LI kirjoitti:
> Hi all,
>
> For OMAP3430 chip, while getting DEBUG_SLAB config enabled, found a bug
> that last user information is missed in slab corruption log dumped by
> kernel. Actually, It's caused by ignorance of redzone and last user tag
> while calling kmem_cache_create() function if cache alignment > 16 bytes
> (unsigned long long).
>
> Here is a patch to fix this problem. Already verified it on kernel 2.6.29.
The patch is badly whitespace damaged.
> From 26a5a8ad2a1d7612929a91f6866cea9d1bea6077 Mon Sep 17 00:00:00 2001
> From: Shiyong Li <[email protected]
> <mailto:[email protected]>>
> Date: Wed, 31 Mar 2010 10:09:35 +0800
> Subject: [PATCH] Fix missing of last user info while getting DEBUG_SLAB
> config enabled.
> As OMAP3 cache line is 64 byte long, while calling kmem_cache_create()
> funtion, some cases need 64 byte alignment of requested memory space.
> But, if cache line > 16 bytes, current kernel ignore redzone
> and last user debug head/trail tag to make sure this alignment is not
> broken.
> This fix removes codes that ignorance of redzone and last user tag.
> Instead, use "align" argument value as object offset to guarantee the
> alignment.
> Signed-off-by: Shiyong Li <[email protected]
> <mailto:[email protected]>>
> ---
> mm/slab.c | 7 ++-----
> 1 files changed, 2 insertions(+), 5 deletions(-)
> diff --git a/mm/slab.c b/mm/slab.c
> index a8a38ca..84af997 100644
> --- a/mm/slab.c
> +++ b/mm/slab.c
> @@ -2267,9 +2267,6 @@ kmem_cache_create (const char *name, size_t size,
> size_t align,
> if (ralign < align) {
> ralign = align;
> }
> - /* disable debug if necessary */
> - if (ralign > __alignof__(unsigned long long))
> - flags &= ~(SLAB_RED_ZONE | SLAB_STORE_USER);
> /*
> * 4) Store it.
> */
> @@ -2289,8 +2286,8 @@ kmem_cache_create (const char *name, size_t size,
> size_t align,
> */
> if (flags & SLAB_RED_ZONE) {
> /* add space for red zone words */
> - cachep->obj_offset += sizeof(unsigned long long);
> - size += 2 * sizeof(unsigned long long);
> + cachep->obj_offset += align;
> + size += align + sizeof(unsigned long long);
> }
I don't understand what you're trying to do here. What if align is less
han sizeof(unsigned long long)? What if SLAB_RED_ZONE is not enabled but
SLAB_STORE_USER is?
Pekka
Hi, Pekka
Thanks for your feedback.
The subject and comments are kind of misleading. Sorry about that.
It intends to make both Red-Zone and LastUser are available even the
alignment is larger than sizeof(long long).
The idea is to reserve certain room ("align") to let the obj's offset
is aligned to cache-line-size to address requirement of kmalloc().
Regarding align is less than sizeof(unsigned long long), I thin it is
handled in previous code.
Or you still expect to explicitly let align is no less than
sizeof(unsigned long long)?
if (flags & SLAB_RED_ZONE) {
ralign = REDZONE_ALIGN;
... ...
}
--
Best Regards
Hu Tao
On Thu, Apr 1, 2010 at 2:29 PM, Pekka Enberg <[email protected]> wrote:
> ShiYong LI kirjoitti:
>>
>> Hi all,
>> ?For OMAP3430 chip, while getting DEBUG_SLAB config enabled, found a bug
>> that last user information is missed in slab corruption log dumped by
>> kernel. Actually, It's caused by ignorance of redzone and last user tag
>> while calling kmem_cache_create() function if cache alignment > 16 bytes
>> (unsigned long long). ?Here is a patch to fix this problem. Already verified
>> it on kernel 2.6.29.
>
> The patch is badly whitespace damaged.
>
>> ?From 26a5a8ad2a1d7612929a91f6866cea9d1bea6077 Mon Sep 17 00:00:00 2001
>> From: Shiyong Li <[email protected]
>> <mailto:[email protected]>>
>> Date: Wed, 31 Mar 2010 10:09:35 +0800
>> Subject: [PATCH] Fix missing of last user info while getting DEBUG_SLAB
>> config enabled.
>> As OMAP3 cache line is 64 byte long, while calling kmem_cache_create()
>> funtion, some cases need 64 byte alignment of requested memory space.
>> But, if cache line > 16 bytes, current kernel ignore redzone
>> and last user debug head/trail tag to make sure this alignment is not
>> broken.
>> This fix removes codes that ignorance of redzone and last user tag.
>> Instead, use "align" argument value as object offset to guarantee the
>> alignment.
>> Signed-off-by: Shiyong Li <[email protected]
>> <mailto:[email protected]>>
>> ---
>> ?mm/slab.c | ? ?7 ++-----
>> ?1 files changed, 2 insertions(+), 5 deletions(-)
>> diff --git a/mm/slab.c b/mm/slab.c
>> index a8a38ca..84af997 100644
>> --- a/mm/slab.c
>> +++ b/mm/slab.c
>> @@ -2267,9 +2267,6 @@ kmem_cache_create (const char *name, size_t size,
>> size_t align,
>> ?if (ralign < align) {
>> ? ralign = align;
>> ?}
>> - /* disable debug if necessary */
>> - if (ralign > __alignof__(unsigned long long))
>> - ?flags &= ~(SLAB_RED_ZONE | SLAB_STORE_USER);
>> ?/*
>> ? * 4) Store it.
>> ? */
>> @@ -2289,8 +2286,8 @@ kmem_cache_create (const char *name, size_t size,
>> size_t align,
>> ? */
>> ?if (flags & SLAB_RED_ZONE) {
>> ? /* add space for red zone words */
>> - ?cachep->obj_offset += sizeof(unsigned long long);
>> - ?size += 2 * sizeof(unsigned long long);
>> + ?cachep->obj_offset += align;
>> + ?size += align + sizeof(unsigned long long);
>> ?}
>
> I don't understand what you're trying to do here. What if align is less han
> sizeof(unsigned long long)? What if SLAB_RED_ZONE is not enabled but
> ?SLAB_STORE_USER is?
>
> ? ? ? ? ? ? ? ? ? ? ? ?Pekka
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to [email protected]. ?For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
>