Re: [PATCH v4] memcg: fix leak on wrong LRU with FUSE

[Daisuke Nishimura]

On Thu, 10 Mar 2011 14:47:52 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:

> On Thu, 10 Mar 2011 08:36:59 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
> 
> > will add. Thank you !
> > 
> 
> Here is v4 based on feedbacks.
> ==
> 
> fs/fuse/dev.c::fuse_try_move_page() does
> 
>    (1) remove a page by ->steal()
>    (2) re-add the page to page cache 
>    (3) link the page to LRU if it was not on LRU at (1)
> 
> This implies the page is _on_ LRU when it's added to radix-tree.
> So, the page is added to  memory cgroup while it's on LRU and
> the pave will remain in the old(wrong) memcg.
> By this bug, force_empty()'s LRU scan cannot find the page and
> rmdir() will never ends.
> 
> This is the same behavior as SwapCache and needs special care as
>  - remove page from LRU before overwrite pc->mem_cgroup.
>  - add page to LRU after overwrite pc->mem_cgroup.
> 
> This will fixes memcg's rmdir() hang issue with FUSE.
> 
> Changelog v3=v4:
>   - moved PageLRU() check into the leaf function.
>   - added comments
> 
> Changelog v2=>v3:
>   - fixed double accounting.
> 
> Changelog v1=>v2:
>   - clean up.
>   - cover !PageLRU() by pagevec case.
> 
> Reviewed-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>

Acked-by: Daisuke Nishimura <nishimura@xxxxxxxxxxxxxxxxx>

I hope this can fix the original BZ case.

Thanks,
Daisuke Nishimura.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/