Re: [PATCH 4.20 71/92] Revert "mm: slowly shrink slabs with a relatively small number of objects"

[Wolfgang Walter]

Am Montag, 18. Februar 2019, 14:34:52 schrieb Sasha Levin:
> On Mon, Feb 18, 2019 at 10:30:44AM -0500, Rik van Riel wrote:
> >On Mon, 2019-02-18 at 14:43 +0100, Greg Kroah-Hartman wrote:
> >> 4.20-stable review patch.  If anyone has any objections, please let
> >> me know.
> >> 
> >> ------------------
> >> 
> >> From: Dave Chinner <dchinner@xxxxxxxxxx>
> >> 
> >> commit a9a238e83fbb0df31c3b9b67003f8f9d1d1b6c96 upstream.
> >> 
> >> This reverts commit 172b06c32b9497 ("mm: slowly shrink slabs with a
> >> relatively small number of objects").
> >
> >This revert will result in the slab caches of dead
> >cgroups with a small number of remaining objects never
> >getting reclaimed, which can be a memory leak in some
> >configurations.
> 
> What's the "right" choice though? we get either leaky cgroups or hanging
> xfs.

This not only made xfs hang. (Since 4.19.3) on two of our servers ext4 got so 
sluggish that login with ssh needed a minute and more. bash tab completion was 
unusable. Sometimes systemd-journald was restarted with "Watchdog timeout 
(limit 3min)!", (journal also was on ext4).

> 
> >But hey, that's your tradeoff to make.
> 
> I don't think that any decision was made here, the stable tree simply
> follows upstream with regards to fixes and bugs such as these: we remain
> "bug compatible" with upstream in these scenarios, there was no decision
> made to prefer either bug.
> 
> --
> Thanks,
> Sasha

Regards
-- 
Wolfgang Walter
Studentenwerk München
Anstalt des öffentlichen Rechts