From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752875AbbFHJAV (ORCPT ); Mon, 8 Jun 2015 05:00:21 -0400 Received: from mail-pd0-f171.google.com ([209.85.192.171]:36179 "EHLO mail-pd0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752511AbbFHI7u (ORCPT ); Mon, 8 Jun 2015 04:59:50 -0400 From: Tejun Heo To: axboe@kernel.dk Cc: linux-kernel@vger.kernel.org, cgroups@vger.kernel.org, vgoyal@redhat.com, avanzini.arianna@gmail.com, Tejun Heo Subject: [PATCH 3/8] cfq-iosched: fix oom cfq_queue ref leak in cfq_set_request() Date: Mon, 8 Jun 2015 17:59:28 +0900 Message-Id: <1433753973-23684-4-git-send-email-tj@kernel.org> X-Mailer: git-send-email 2.4.2 In-Reply-To: <1433753973-23684-1-git-send-email-tj@kernel.org> References: <1433753973-23684-1-git-send-email-tj@kernel.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org If the cfq_queue cached in cfq_io_cq is the oom one, cfq_set_request() replaces it by invoking cfq_get_queue() again without putting the oom queue leaking the reference it was holding. While oom queues are not released through reference counting, they're still reference counted and this can theoretically lead to the reference count overflowing and incorrectly invoke the usual release path on it. Fix it by making cfq_set_request() put the ref it was holding. Signed-off-by: Tejun Heo Cc: Vivek Goyal Cc: Arianna Avanzini --- block/cfq-iosched.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index c7b33aa..e0a34ba 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -4231,6 +4231,8 @@ cfq_set_request(struct request_queue *q, struct request *rq, struct bio *bio, new_queue: cfqq = cic_to_cfqq(cic, is_sync); if (!cfqq || cfqq == &cfqd->oom_cfqq) { + if (cfqq) + cfq_put_queue(cfqq); cfqq = cfq_get_queue(cfqd, is_sync, cic, bio, gfp_mask); cic_set_cfqq(cic, cfqq, is_sync); } else { -- 2.4.2