Re: [PATCH 2/3] slab_common: clear pointers to per memcg caches on destroy

[Vladimir Davydov <vdavydov@virtuozzo.com>]

On Thu, Oct 08, 2015 at 02:17:35PM -0700, Andrew Morton wrote:
> On Thu, 8 Oct 2015 19:02:40 +0300 Vladimir Davydov <vdavydov@virtuozzo.com> wrote:
> 
> > Currently, we do not clear pointers to per memcg caches in the
> > memcg_params.memcg_caches array when a global cache is destroyed with
> > kmem_cache_destroy. It is fine if the global cache does get destroyed.
> > However, a cache can be left on the list if it still has active objects
> > when kmem_cache_destroy is called (due to a memory leak). If this
> > happens, the entries in the array will point to already freed areas,
> > which is likely to result in data corruption when the cache is reused
> > (via slab merging).
> 
> It's important that we report these leaks so the kernel bug can get
> fixed.  The patch doesn't add such detection and reporting, but it
> could do so?

Reporting individual leaks is up to the slab implementation, we simply
can't do it from the generic code, so we just warn that there is a leak
there. SLUB already dumps addresses of all leaked objects to the log
(see kmem_cache_close -> free_partial -> list_slab_objects).

Thanks,
Vladimir

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>