[bug] shrink_slab shrinkersize handling

[bug] shrink_slab shrinkersize handling

[solmac john]

[-- Attachment #1: Type: text/plain, Size: 1506 bytes --]

Hi All,

During mm performance testing sometimes I observed below kernel messages

[   80.776000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
delete nr=-2133936901
[   80.784000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
delete nr=-2139256767
[   80.796000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
delete nr=-2079333971
[   80.804000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
delete nr=-2096156269
[   80.812000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
delete nr=-20658392

 After debugging is I fount these prints from mm/vmscan.c
Kernel Error msg point

   unsigned long total_scan;

               unsigned long max_pass;

                max_pass = do_shrinker_shrink(shrinker, shrink, 0);

                delta = (4 * nr_pages_scanned) / shrinker->seeks;

                delta *= max_pass;

                do_div(delta, lru_pages + 1);

                shrinker->nr += delta;

                if (shrinker->nr < 0) {

                        printk(KERN_ERR "shrink_slab: %pF negative objects
to "

                               "delete nr=%ld\n",

                               shrinker->shrink, shrinker->nr);

                        shrinker->nr = max_pass;
                }



I found one patch  http://lkml.org/lkml/2011/8/22/80    for this fix
Please let me know reason why I am getting above error and above is really
fix for this problem.  ?
I am working on ARM cortex A9  linux-3.0.20 kernel.

Thanks,
John

[-- Attachment #2: Type: text/html, Size: 4737 bytes --]

Re: [bug] shrink_slab shrinkersize handling

[Andrew Morton]

On Wed, 16 May 2012 14:33:18 +0530
solmac john <johnsolmac@gmail.com> wrote:

> Hi All,
> 
> During mm performance testing sometimes I observed below kernel messages
> 
> [   80.776000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
> delete nr=-2133936901
> [   80.784000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
> delete nr=-2139256767
> [   80.796000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
> delete nr=-2079333971
> [   80.804000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
> delete nr=-2096156269
> [   80.812000] shrink_slab: ashmem_shrink+0x0/0x114 negative objects to
> delete nr=-20658392
> 
> ...
>
> I found one patch  http://lkml.org/lkml/2011/8/22/80    for this fix
> Please let me know reason why I am getting above error and above is really
> fix for this problem.  ?

Yes, that patch should fix it.

Aside: I spent some time trying to work out the reason why local
variable `max_pass' in shrink_slab() is called `max_pass' and failed.


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>