Interesting, thanks. Patch doesn't apply clean to 4.7.2, but I get the
gist, delete one version of should_compact_retry() and move ifdef so
it only has __alloc_pages_direct_compact(). I'll try it out if my
rsync gets OOM'd again.

On Fri, Aug 26, 2016 at 3:02 PM, Tomasz Torcz [off-list ref] wrote:

On Fri, Aug 26, 2016 at 03:00:48PM -0400, E V wrote:

quoted

Just upgraded from 4.6.5 to 4.7.2 for my btrfs backup server with 32GB
of ram. Only thing that run's on it is an rsync of an NFS filesystem
to the local btrfs. Cached mem tends to hang out around 26-30GB, but
with 4.7.2 the OOM is now going crazy and trying to kill whatever it
can including my ssh and rsync process. Anyone seen anything similar?

  That's probably OOM regression in 4.7. Could you please test this patch:
https://lkml.org/lkml/2016/8/23/145

--
Tomasz Torcz              ,,If you try to upissue this patchset I shall be seeking
xmpp: zdzichubg@chrome.pl   an IP-routable hand grenade.'' -- Andrew Morton (LKML)

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html