From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756454AbaEaLTf (ORCPT ); Sat, 31 May 2014 07:19:35 -0400 Received: from mx2.parallels.com ([199.115.105.18]:50248 "EHLO mx2.parallels.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754558AbaEaLTe (ORCPT ); Sat, 31 May 2014 07:19:34 -0400 Date: Sat, 31 May 2014 15:19:23 +0400 From: Vladimir Davydov To: Christoph Lameter CC: , , , , Subject: Re: [PATCH -mm 8/8] slab: reap dead memcg caches aggressively Message-ID: <20140531111922.GD25076@esperanza> References: <23a736c90a81e13a2252d35d9fc3dc04a9ed7d7c.1401457502.git.vdavydov@parallels.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline In-Reply-To: X-Originating-IP: [81.5.110.170] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, May 30, 2014 at 10:01:26AM -0500, Christoph Lameter wrote: > On Fri, 30 May 2014, Vladimir Davydov wrote: > > > We don't disable free objects caching for SLAB, because it would force > > kfree to always take a spin lock, which would degrade performance > > significantly. > > You can use a similar approach than in SLUB. Reduce the size of the per > cpu array objects to zero. Then SLAB will always fall back to its slow > path in cache_flusharray() where you may be able to do something with less > of an impact on performace. In contrast to SLUB, for SLAB this will slow down kfree significantly. Fast path for SLAB is just putting an object to a per cpu array, while the slow path requires taking a per node lock, which is much slower even with no contention. There still can be lots of objects in a dead memcg cache (e.g. hundreds of megabytes of dcache), so such performance degradation is not acceptable, IMO. OTOH, we already have cache_reap running periodically for each cache. Making it drain all free objects in dead caches won't impact performance at all, neither will it complicate the code. The only downside is a dead cache won't be destroyed immediately after it becomes unused, but since cache_reap runs pretty often (each several secs), it shouldn't result in any problems, I guess. Thanks.