Re: [PATCH] mm: preallocate page before lock_page at filemap COW.(WasRe: [PATCH V2] mm: Do not keep page locked during page fault whilecharging it for memcg

[Hiroyuki Kamezawa]

2011/6/24 Michal Hocko <mhocko@xxxxxxx>:
> Sorry, forgot to send my
> Reviewed-by: Michal Hocko <mhocko@suse>
>

Thanks.

> I still have concerns about this way to handle the issue. See the follow
> up discussion in other thread (https://lkml.org/lkml/2011/6/23/135).
>
> Anyway I think that we do not have many other options to handle this.
> Either we unlock, charge, lock&restes or we preallocate, fault in
>
I agree.

> Or am I missing some other ways how to do it? What do others think about
> these approaches?
>

Yes, I'd like to hear other mm specialists' suggestion. and I'll think
other way, again.
Anyway, memory reclaim with holding a lock_page() can cause big latency
or starvation especially when memcg is used. It's better to avoid it.

Thanks,
-Kame
--
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/