From patchwork Thu Dec 1 05:54:56 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Qu Wenruo X-Patchwork-Id: 9455465 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id B865260585 for ; Thu, 1 Dec 2016 05:56:30 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id AB6B62847F for ; Thu, 1 Dec 2016 05:56:30 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 9ED4428489; Thu, 1 Dec 2016 05:56:30 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 43A802847F for ; Thu, 1 Dec 2016 05:56:30 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758391AbcLAF4Z (ORCPT ); Thu, 1 Dec 2016 00:56:25 -0500 Received: from cn.fujitsu.com ([222.73.24.84]:21600 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1750731AbcLAF4Y (ORCPT ); Thu, 1 Dec 2016 00:56:24 -0500 X-IronPort-AV: E=Sophos;i="5.20,367,1444665600"; d="scan'208";a="994964" Received: from unknown (HELO cn.fujitsu.com) ([10.167.250.3]) by song.cn.fujitsu.com with ESMTP; 01 Dec 2016 13:55:12 +0800 Received: from localhost.localdomain (unknown [10.167.226.34]) by cn.fujitsu.com (Postfix) with ESMTP id 6CE6642469E4; Thu, 1 Dec 2016 13:55:09 +0800 (CST) From: Qu Wenruo To: linux-btrfs@vger.kernel.org Cc: dsterba@suse.cz Subject: [PATCH 2/2] btrfs: qgroup: Fix qgroup reserved space underflow caused by buffered write and quota enable Date: Thu, 1 Dec 2016 13:54:56 +0800 Message-Id: <20161201055456.32062-2-quwenruo@cn.fujitsu.com> X-Mailer: git-send-email 2.10.2 In-Reply-To: <20161201055456.32062-1-quwenruo@cn.fujitsu.com> References: <20161201055456.32062-1-quwenruo@cn.fujitsu.com> MIME-Version: 1.0 X-yoursite-MailScanner-ID: 6CE6642469E4.ACE3A X-yoursite-MailScanner: Found to be clean X-yoursite-MailScanner-From: quwenruo@cn.fujitsu.com Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP [BUG] Under the following case, we can underflow qgroup reserved space. Task A | Task B --------------------------------------------------------------- Quota disabled | Buffered write | |- btrfs_check_data_free_space() | | *NO* qgroup space is reserved | | since quota is *DISABLED* | |- All pages are copied to page | cache | | Enable quota | Quota scan finished | | Sync_fs | |- run_delalloc_range | |- Write pages | |- btrfs_finish_ordered_io | |- insert_reserved_file_extent | |- btrfs_qgroup_release_data() | Since no qgroup space is reserved in Task A, we underflow qgroup reserved space This can be detected by fstest btrfs/104. [CAUSE] In insert_reserved_file_extent() we info qgroup to release the @ram_bytes size of qgroup reserved_space under all case. And btrfs_qgroup_release_data() will check if qgroup is enabled. However in above case, the buffered write happens before quota is enabled, so we don't havee reserved space for that range. [FIX] In insert_reserved_file_extent(), we info qgroup to release the acctual byte number it released. In above case, since we don't have reserved space, we info qgroup to release 0 byte, so the problem can be fixed. And thanks to the @reserved parameter introduced by qgroup rework, and previous patch to return release bytes, the fix can be as small as less than 10 lines. Signed-off-by: Qu Wenruo --- To David: These 2 patches, with updated extra WARN_ON() patch(V5), btrfs on x86_64 is good for current qgroup test group. But the bug reported by Chadan still exists, and the fix for that will be delayed for a while, as the fix involves large interface change (add struct extent_changeset for every qgroup reserve caller). I'll run extra tests for that fix to ensure they are OK and won't cause extra bugs. Thanks, Qu --- fs/btrfs/inode.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 242dc7e..3db58d9 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2256,6 +2256,7 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans, struct btrfs_path *path; struct extent_buffer *leaf; struct btrfs_key ins; + u64 qg_released; int extent_inserted = 0; int ret; @@ -2311,15 +2312,19 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans, ins.objectid = disk_bytenr; ins.offset = disk_num_bytes; ins.type = BTRFS_EXTENT_ITEM_KEY; - ret = btrfs_alloc_reserved_file_extent(trans, root, - root->root_key.objectid, - btrfs_ino(inode), file_pos, - ram_bytes, &ins); + /* * Release the reserved range from inode dirty range map, as it is * already moved into delayed_ref_head */ - btrfs_qgroup_release_data(inode, file_pos, ram_bytes); + ret = btrfs_qgroup_release_data(inode, file_pos, ram_bytes); + if (ret < 0) + goto out; + qg_released = ret; + ret = btrfs_alloc_reserved_file_extent(trans, root, + root->root_key.objectid, + btrfs_ino(inode), file_pos, + qg_released, &ins); out: btrfs_free_path(path);