From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753816AbbIETrw (ORCPT ); Sat, 5 Sep 2015 15:47:52 -0400 Received: from mail-yk0-f179.google.com ([209.85.160.179]:35540 "EHLO mail-yk0-f179.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752022AbbIETrm (ORCPT ); Sat, 5 Sep 2015 15:47:42 -0400 Date: Sat, 5 Sep 2015 15:47:36 -0400 From: Tejun Heo To: Jens Axboe Cc: Josh Boyer , Jeff Moyer , msnitzer@redhat.com, Li Zefan , Johannes Weiner , cgroups@vger.kernel.org, "Linux-Kernel@Vger. Kernel. Org" , "Richard W.M. Jones" Subject: [PATCH block/for-linus] block: blkg_destroy_all() should clear q->root_blkg and ->root_rl.blkg Message-ID: <20150905194736.GB8231@mtj.duckdns.org> References: <20150902145307.GG22326@mtj.duckdns.org> <20150902153255.GH22326@mtj.duckdns.org> <20150904104602.GN29283@redhat.com> <20150904171302.GE25329@mtj.duckdns.org> <20150904204244.GQ29283@redhat.com> <20150905153439.GA18461@redhat.com> <20150905154840.GA19460@redhat.com> <20150905183801.GA8231@mtj.duckdns.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20150905183801.GA8231@mtj.duckdns.org> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org While making the root blkg unconditional, ec13b1d6f0a0 ("blkcg: always create the blkcg_gq for the root blkcg") removed the part which clears q->root_blkg and ->root_rl.blkg during q exit. This leaves the two pointers dangling after blkg_destroy_all(). blk-throttle exit path performs blkg traversals and dereferences ->root_blkg and can lead to the following oops. BUG: unable to handle kernel NULL pointer dereference at 0000000000000558 IP: [] __blkg_lookup+0x26/0x70 ... task: ffff88001b4e2580 ti: ffff88001ac0c000 task.ti: ffff88001ac0c000 RIP: 0010:[] [] __blkg_lookup+0x26/0x70 ... Call Trace: [] blk_throtl_drain+0x5a/0x110 [] blkcg_drain_queue+0x18/0x20 [] __blk_drain_queue+0xc0/0x170 [] blk_queue_bypass_start+0x61/0x80 [] blkcg_deactivate_policy+0x39/0x100 [] blk_throtl_exit+0x38/0x50 [] blkcg_exit_queue+0x3e/0x50 [] blk_release_queue+0x1e/0xc0 ... While the bug is a straigh-forward use-after-free bug, it is tricky to reproduce because blkg release is RCU protected and the rest of exit path usually finishes before RCU grace period. This patch fixes the bug by updating blkg_destro_all() to clear q->root_blkg and ->root_rl.blkg. Signed-off-by: Tejun Heo Reported-by: "Richard W.M. Jones" Reported-by: Josh Boyer Link: http://lkml.kernel.org/g/CA+5PVA5rzQ0s4723n5rHBcxQa9t0cW8BPPBekr_9aMRoWt2aYg@mail.gmail.com Fixes: ec13b1d6f0a0 ("blkcg: always create the blkcg_gq for the root blkcg") Cc: stable@vger.kernel.org # v4.2+ --- Hello, Richard, can you please verify that this patch fixes the bug? Thanks a lot! block/blk-cgroup.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index d6283b3..9cc48d1d 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -387,6 +387,9 @@ static void blkg_destroy_all(struct request_queue *q) blkg_destroy(blkg); spin_unlock(&blkcg->lock); } + + q->root_blkg = NULL; + q->root_rl.blkg = NULL; } /*