From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753561AbaGLP4m (ORCPT ); Sat, 12 Jul 2014 11:56:42 -0400 Received: from mail-wg0-f48.google.com ([74.125.82.48]:57551 "EHLO mail-wg0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751435AbaGLP4l (ORCPT ); Sat, 12 Jul 2014 11:56:41 -0400 Message-ID: <53C15AB6.9000605@kernel.dk> Date: Sat, 12 Jul 2014 17:56:38 +0200 From: Jens Axboe User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.6.0 MIME-Version: 1.0 To: Tejun Heo CC: linux-kernel@vger.kernel.org, Shirish Pargaonkar , Sasha Levin , Jet Chen Subject: Re: [PATCH block/for-linus] blkcg: don't call into policy draining if root_blkg is already gone References: <20140705224321.GA9701@mtj.dyndns.org> In-Reply-To: <20140705224321.GA9701@mtj.dyndns.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2014-07-06 00:43, Tejun Heo wrote: > While a queue is being destroyed, all the blkgs are destroyed and its > ->root_blkg pointer is set to NULL. If someone else starts to drain > while the queue is in this state, the following oops happens. > > NULL pointer dereference at 0000000000000028 > IP: [] blk_throtl_drain+0x84/0x230 > PGD e4a1067 PUD b773067 PMD 0 > Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC > Modules linked in: cfq_iosched(-) [last unloaded: cfq_iosched] > CPU: 1 PID: 537 Comm: bash Not tainted 3.16.0-rc3-work+ #2 > Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 > task: ffff88000e222250 ti: ffff88000efd4000 task.ti: ffff88000efd4000 > RIP: 0010:[] [] blk_throtl_drain+0x84/0x230 > RSP: 0018:ffff88000efd7bf0 EFLAGS: 00010046 > RAX: 0000000000000000 RBX: ffff880015091450 RCX: 0000000000000001 > RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 > RBP: ffff88000efd7c10 R08: 0000000000000000 R09: 0000000000000001 > R10: ffff88000e222250 R11: 0000000000000000 R12: ffff880015091450 > R13: ffff880015092e00 R14: ffff880015091d70 R15: ffff88001508fc28 > FS: 00007f1332650740(0000) GS:ffff88001fa80000(0000) knlGS:0000000000000000 > CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b > CR2: 0000000000000028 CR3: 0000000009446000 CR4: 00000000000006e0 > Stack: > ffffffff8144e8f6 ffff880015091450 0000000000000000 ffff880015091d80 > ffff88000efd7c28 ffffffff8144ae2f ffff880015091450 ffff88000efd7c58 > ffffffff81427641 ffff880015091450 ffffffff82401f00 ffff880015091450 > Call Trace: > [] blkcg_drain_queue+0x1f/0x60 > [] __blk_drain_queue+0x71/0x180 > [] blk_queue_bypass_start+0x6e/0xb0 > [] blkcg_deactivate_policy+0x38/0x120 > [] blk_throtl_exit+0x34/0x50 > [] blkcg_exit_queue+0x35/0x40 > [] blk_release_queue+0x26/0xd0 > [] kobject_cleanup+0x38/0x70 > [] kobject_put+0x28/0x60 > [] blk_put_queue+0x15/0x20 > [] scsi_device_dev_release_usercontext+0x16b/0x1c0 > [] execute_in_process_context+0x89/0xa0 > [] scsi_device_dev_release+0x1c/0x20 > [] device_release+0x32/0xa0 > [] kobject_cleanup+0x38/0x70 > [] kobject_put+0x28/0x60 > [] put_device+0x17/0x20 > [] __scsi_remove_device+0xa9/0xe0 > [] scsi_remove_device+0x2b/0x40 > [] sdev_store_delete+0x27/0x30 > [] dev_attr_store+0x18/0x30 > [] sysfs_kf_write+0x3e/0x50 > [] kernfs_fop_write+0xe7/0x170 > [] vfs_write+0xaf/0x1d0 > [] SyS_write+0x4d/0xc0 > [] system_call_fastpath+0x16/0x1b > > 776687bce42b ("block, blk-mq: draining can't be skipped even if > bypass_depth was non-zero") made it easier to trigger this bug by > making blk_queue_bypass_start() drain even when it loses the first > bypass test to blk_cleanup_queue(); however, the bug has always been > there even before the commit as blk_queue_bypass_start() could race > against queue destruction, win the initial bypass test but perform the > actual draining after blk_cleanup_queue() already destroyed all blkgs. > > Fix it by skippping calling into policy draining if all the blkgs are > already gone. Applied, thanks Tejun. -- Jens Axboe