Re: Kswapd in 3.2.0-rc5 is a CPU hog

[KAMEZAWA Hiroyuki]

On Tue, 27 Dec 2011 21:33:04 +0800
Hillf Danton <dhillf@xxxxxxxxx> wrote:

> On Tue, Dec 27, 2011 at 8:20 AM, KAMEZAWA Hiroyuki
> <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
> > On Mon, 26 Dec 2011 20:35:46 +0800
> > Hillf Danton <dhillf@xxxxxxxxx> wrote:
> >
> >> On Sun, Dec 25, 2011 at 6:21 PM, Nikolay S. <nowhere@xxxxxxxxxxxxxxxx> wrote:
> >> >
> >> > Uhm.., is this patch against 3.2-rc4? I can not apply it. There's no
> >> > mem_cgroup_lru_del_list(), but void mem_cgroup_del_lru_list(). Should I
> >> > place changes there?
> >> >
> >> > And also, -rc7 is here. May the problem be addressed as part of some
> >> > ongoing work? Is there any reason to try -rc7 (the problem requires
> >> > several days of uptime to become obvious)?
> >> >
> >>
> >> Sorry, Nikolay, it is not based on the -next, nor on the -rc5(I assumed it was).
> >> The following is based on -next, and if you want to test -rc5, please
> >> grep MEM_CGROUP_ZSTAT mm/memcontrol.c and change it.
> >>
> >> Best regard
> >>
> >
> > Hmm ? memcg is used ? Why do you consider this will be a help ?
> >
> 
> Hi Kame
> 
> Please see the livelock at
>         https://lkml.org/lkml/2011/12/23/222
> 
> and if it is related to the hog here, please check if the following is
> in right direction.
> 

But linux-next and his 3.2-rc5 has totally different LRU design.
3.2-rc doesn't have unified memcg/global LRU and memcg doesn't affect
kswapd behavior if not used.

Thanks,
-Kame

> Thanks
> Hillf
> ---
> 
> --- a/mm/memcontrol.c	Mon Dec 26 20:34:38 2011
> +++ b/mm/memcontrol.c	Tue Dec 27 20:05:12 2011
> @@ -3637,6 +3637,7 @@ static int mem_cgroup_force_empty_list(s
>  	list = &mz->lruvec.lists[lru];
> 
>  	loop = MEM_CGROUP_ZSTAT(mz, lru);
> +	WARN_ON((long)loop < 0);
>  	/* give some margin against EBUSY etc...*/
>  	loop += 256;
>  	busy = NULL;
> --
> 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/
> 

--
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/