[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181105130544.GJ4361@dhcp22.suse.cz>
Date: Mon, 5 Nov 2018 14:05:44 +0100
From: Michal Hocko <mhocko@...nel.org>
To: Kuo-Hsin Yang <vovoy@...omium.org>
Cc: linux-kernel@...r.kernel.org, intel-gfx@...ts.freedesktop.org,
linux-mm@...ck.org, Chris Wilson <chris@...is-wilson.co.uk>,
Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
Peter Zijlstra <peterz@...radead.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Dave Hansen <dave.hansen@...el.com>
Subject: Re: [PATCH v4] mm, drm/i915: mark pinned shmemfs pages as unevictable
On Mon 05-11-18 14:02:09, Michal Hocko wrote:
> On Mon 05-11-18 19:13:48, Kuo-Hsin Yang wrote:
> > The i915 driver uses shmemfs to allocate backing storage for gem
> > objects. These shmemfs pages can be pinned (increased ref count) by
> > shmem_read_mapping_page_gfp(). When a lot of pages are pinned, vmscan
> > wastes a lot of time scanning these pinned pages. In some extreme case,
> > all pages in the inactive anon lru are pinned, and only the inactive
> > anon lru is scanned due to inactive_ratio, the system cannot swap and
> > invokes the oom-killer. Mark these pinned pages as unevictable to speed
> > up vmscan.
> >
> > Export pagevec API check_move_unevictable_pages().
>
> Thanks for reworking the patch. This looks much more to my taste. At
> least the mm part. I haven't really looked at the the drm part.
One side note. Longterm we probably want a better pinning API. It would
hide this LRU manipulation implementation detail + enforce some limiting
and provide a good way that the pin is longterm. People are working on
this already but it is a PITA and long time to get there.
--
Michal Hocko
SUSE Labs
Powered by blists - more mailing lists