From patchwork Sun May 29 08:15:34 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Rongwei Wang X-Patchwork-Id: 12864139 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 927DFC433FE for ; Sun, 29 May 2022 08:15:46 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id E13AF8D0002; Sun, 29 May 2022 04:15:44 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id D54DA8D0006; Sun, 29 May 2022 04:15:44 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id B51428D0005; Sun, 29 May 2022 04:15:44 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0011.hostedemail.com [216.40.44.11]) by kanga.kvack.org (Postfix) with ESMTP id 9B8608D0003 for ; Sun, 29 May 2022 04:15:44 -0400 (EDT) Received: from smtpin07.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay01.hostedemail.com (Postfix) with ESMTP id 6E91561225 for ; Sun, 29 May 2022 08:15:44 +0000 (UTC) X-FDA: 79518071808.07.2BA9E65 Received: from out30-131.freemail.mail.aliyun.com (out30-131.freemail.mail.aliyun.com [115.124.30.131]) by imf20.hostedemail.com (Postfix) with ESMTP id 934EB1C0004 for ; Sun, 29 May 2022 08:15:26 +0000 (UTC) X-Alimail-AntiSpam: AC=PASS;BC=-1|-1;BR=01201311R101e4;CH=green;DM=||false|;DS=||;FP=0|-1|-1|-1|0|-1|-1|-1;HT=e01e04423;MF=rongwei.wang@linux.alibaba.com;NM=1;PH=DS;RN=9;SR=0;TI=SMTPD_---0VEe4qOA_1653812137; Received: from localhost.localdomain(mailfrom:rongwei.wang@linux.alibaba.com fp:SMTPD_---0VEe4qOA_1653812137) by smtp.aliyun-inc.com(127.0.0.1); Sun, 29 May 2022 16:15:38 +0800 From: Rongwei Wang To: akpm@linux-foundation.org, vbabka@suse.cz, roman.gushchin@linux.dev, iamjoonsoo.kim@lge.com, rientjes@google.com, penberg@kernel.org, cl@linux.com Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: [PATCH 2/3] mm/slub: improve consistency of nr_slabs count Date: Sun, 29 May 2022 16:15:34 +0800 Message-Id: <20220529081535.69275-2-rongwei.wang@linux.alibaba.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220529081535.69275-1-rongwei.wang@linux.alibaba.com> References: <20220529081535.69275-1-rongwei.wang@linux.alibaba.com> MIME-Version: 1.0 X-Stat-Signature: i7moyhehx6ekjtnr5gx9ge6ruc9pecqw X-Rspam-User: Authentication-Results: imf20.hostedemail.com; dkim=none; spf=pass (imf20.hostedemail.com: domain of rongwei.wang@linux.alibaba.com designates 115.124.30.131 as permitted sender) smtp.mailfrom=rongwei.wang@linux.alibaba.com; dmarc=pass (policy=none) header.from=alibaba.com X-Rspamd-Server: rspam10 X-Rspamd-Queue-Id: 934EB1C0004 X-HE-Tag: 1653812126-797024 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: Currently, discard_slab() can change nr_slabs count without holding node's list_lock. This will lead some error messages print when scanning node's partial or full list, e.g. validate all slabs. Literally, it affects the consistency of nr_slabs count. Here, discard_slab() is abandoned, And dec_slabs_node() is called before releasing node's list_lock. dec_slabs_nodes() and free_slab() can be called separately to ensure consistency of nr_slabs count. Signed-off-by: Rongwei Wang --- mm/slub.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index 310e56d99116..bffb95bbb0ee 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2039,12 +2039,6 @@ static void free_slab(struct kmem_cache *s, struct slab *slab) __free_slab(s, slab); } -static void discard_slab(struct kmem_cache *s, struct slab *slab) -{ - dec_slabs_node(s, slab_nid(slab), slab->objects); - free_slab(s, slab); -} - /* * Management of partially allocated slabs. */ @@ -2413,6 +2407,7 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab, if (!new.inuse && n->nr_partial >= s->min_partial) { mode = M_FREE; + spin_lock_irqsave(&n->list_lock, flags); } else if (new.freelist) { mode = M_PARTIAL; /* @@ -2437,7 +2432,7 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab, old.freelist, old.counters, new.freelist, new.counters, "unfreezing slab")) { - if (mode == M_PARTIAL || mode == M_FULL) + if (mode != M_FULL_NOLIST) spin_unlock_irqrestore(&n->list_lock, flags); goto redo; } @@ -2449,7 +2444,10 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab, stat(s, tail); } else if (mode == M_FREE) { stat(s, DEACTIVATE_EMPTY); - discard_slab(s, slab); + dec_slabs_node(s, slab_nid(slab), slab->objects); + spin_unlock_irqrestore(&n->list_lock, flags); + + free_slab(s, slab); stat(s, FREE_SLAB); } else if (mode == M_FULL) { add_full(s, n, slab); @@ -2502,6 +2500,7 @@ static void __unfreeze_partials(struct kmem_cache *s, struct slab *partial_slab) if (unlikely(!new.inuse && n->nr_partial >= s->min_partial)) { slab->next = slab_to_discard; slab_to_discard = slab; + dec_slabs_node(s, slab_nid(slab), slab->objects); } else { add_partial(n, slab, DEACTIVATE_TO_TAIL); stat(s, FREE_ADD_PARTIAL); @@ -2516,7 +2515,7 @@ static void __unfreeze_partials(struct kmem_cache *s, struct slab *partial_slab) slab_to_discard = slab_to_discard->next; stat(s, DEACTIVATE_EMPTY); - discard_slab(s, slab); + free_slab(s, slab); stat(s, FREE_SLAB); } } @@ -3404,9 +3403,10 @@ static void __slab_free(struct kmem_cache *s, struct slab *slab, remove_full(s, n, slab); } + dec_slabs_node(s, slab_nid(slab), slab->objects); spin_unlock_irqrestore(&n->list_lock, flags); stat(s, FREE_SLAB); - discard_slab(s, slab); + free_slab(s, slab); } /* @@ -4265,6 +4265,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) if (!slab->inuse) { remove_partial(n, slab); list_add(&slab->slab_list, &discard); + dec_slabs_node(s, slab_nid(slab), slab->objects); } else { list_slab_objects(s, slab, "Objects remaining in %s on __kmem_cache_shutdown()"); @@ -4273,7 +4274,7 @@ static void free_partial(struct kmem_cache *s, struct kmem_cache_node *n) spin_unlock_irq(&n->list_lock); list_for_each_entry_safe(slab, h, &discard, slab_list) - discard_slab(s, slab); + free_slab(s, slab); } bool __kmem_cache_empty(struct kmem_cache *s) @@ -4595,6 +4596,7 @@ static int __kmem_cache_do_shrink(struct kmem_cache *s) if (free == slab->objects) { list_move(&slab->slab_list, &discard); n->nr_partial--; + dec_slabs_node(s, slab_nid(slab), slab->objects); } else if (free <= SHRINK_PROMOTE_MAX) list_move(&slab->slab_list, promote + free - 1); } @@ -4610,7 +4612,7 @@ static int __kmem_cache_do_shrink(struct kmem_cache *s) /* Release empty slabs */ list_for_each_entry_safe(slab, t, &discard, slab_list) - discard_slab(s, slab); + free_slab(s, slab); if (slabs_node(s, node)) ret = 1;