From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933257Ab2IROR4 (ORCPT ); Tue, 18 Sep 2012 10:17:56 -0400 Received: from mailhub.sw.ru ([195.214.232.25]:40636 "EHLO relay.sw.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933183Ab2IRORA (ORCPT ); Tue, 18 Sep 2012 10:17:00 -0400 From: Glauber Costa To: Cc: , , , Tejun Heo , , Suleiman Souhlal , Frederic Weisbecker , Mel Gorman , David Rientjes , Glauber Costa , Christoph Lameter , Pekka Enberg , Michal Hocko , Johannes Weiner Subject: [PATCH v3 13/16] slab: slab-specific propagation changes. Date: Tue, 18 Sep 2012 18:12:07 +0400 Message-Id: <1347977530-29755-14-git-send-email-glommer@parallels.com> X-Mailer: git-send-email 1.7.11.4 In-Reply-To: <1347977530-29755-1-git-send-email-glommer@parallels.com> References: <1347977530-29755-1-git-send-email-glommer@parallels.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org When a parent cache does tune_cpucache, we need to propagate that to the children as well. For that, we unfortunately need to tap into the slab core. Signed-off-by: Glauber Costa CC: Christoph Lameter CC: Pekka Enberg CC: Michal Hocko CC: Kamezawa Hiroyuki CC: Johannes Weiner CC: Suleiman Souhlal --- mm/slab.c | 28 +++++++++++++++++++++++++++- mm/slab_common.c | 1 + 2 files changed, 28 insertions(+), 1 deletion(-) diff --git a/mm/slab.c b/mm/slab.c index d5c196b..bd9928f 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -4104,7 +4104,7 @@ static void do_ccupdate_local(void *info) } /* Always called with the slab_mutex held */ -static int do_tune_cpucache(struct kmem_cache *cachep, int limit, +static int __do_tune_cpucache(struct kmem_cache *cachep, int limit, int batchcount, int shared, gfp_t gfp) { struct ccupdate_struct *new; @@ -4147,6 +4147,32 @@ static int do_tune_cpucache(struct kmem_cache *cachep, int limit, return alloc_kmemlist(cachep, gfp); } +static int do_tune_cpucache(struct kmem_cache *cachep, int limit, + int batchcount, int shared, gfp_t gfp) +{ + int ret; +#ifdef CONFIG_MEMCG_KMEM + struct kmem_cache *c; + struct mem_cgroup_cache_params *p; +#endif + + ret = __do_tune_cpucache(cachep, limit, batchcount, shared, gfp); +#ifdef CONFIG_MEMCG_KMEM + if (slab_state < FULL) + return ret; + + if ((ret < 0) || (cachep->memcg_params.id == -1)) + return ret; + + list_for_each_entry(p, &cachep->memcg_params.sibling_list, sibling_list) { + c = container_of(p, struct kmem_cache, memcg_params); + /* return value determined by the parent cache only */ + __do_tune_cpucache(c, limit, batchcount, shared, gfp); + } +#endif + return ret; +} + /* Called with slab_mutex held always */ static int enable_cpucache(struct kmem_cache *cachep, gfp_t gfp) { diff --git a/mm/slab_common.c b/mm/slab_common.c index c6fb4a7..341c3e1 100644 --- a/mm/slab_common.c +++ b/mm/slab_common.c @@ -124,6 +124,7 @@ kmem_cache_create_memcg(struct mem_cgroup *memcg, const char *name, size_t size, #ifdef CONFIG_MEMCG_KMEM s->memcg_params.memcg = memcg; s->memcg_params.parent = parent_cache; + s->memcg_params.id = -1; /* not registered yet */ #endif s->name = kstrdup(name, GFP_KERNEL); if (!s->name) { -- 1.7.11.4