Re: [PATCH 4/3] memcg: call uncharge_swapcache outside of tree_lock (Re: [PATCH 0/3] fix stale swap cache account leak in memcg v7)

[Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>]

> I understand the problem, but, wait a bit. NACK to this patch itself.
> 
> 1. I placed _uncharge_ inside tree_lock because __remove_from_page_cache() does.
>    (i.e. using the same logic.)
>    So, plz change both logic at once.(change caller of  mem_cgroup_uncharge_cache_page())
> 
hmm, I see.
cache_charge is outside of tree_lock, so moving uncharge would make sense.
IMHO, we should make the period of spinlock as small as possible,
and charge/uncharge of pagecache/swapcache is protected by page lock, not tree_lock.

> 2. Shouldn't we disable IRQ while __mem_cgroup_uncharge_common() rather than moving
>    function ?
> 
Yes, this is another choise.
But, isn't it better to disable IRQ at all users of lock_page_cgroup..unlock_page_cgroup
to avoid this dead lock ?

Anyway, I'll postpone this fix for a while.
We should fix stale swap swapcache first.


Thanks,
Daisuke Nishimura.

--
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>