Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752501AbdFAQlJ (ORCPT ); Thu, 1 Jun 2017 12:41:09 -0400 Received: from shadbolt.e.decadent.org.uk ([88.96.1.126]:33092 "EHLO shadbolt.e.decadent.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751863AbdFAPof (ORCPT ); Thu, 1 Jun 2017 11:44:35 -0400 Content-Type: text/plain; charset="UTF-8" Content-Disposition: inline Content-Transfer-Encoding: 8bit MIME-Version: 1.0 From: Ben Hutchings To: linux-kernel@vger.kernel.org, stable@vger.kernel.org CC: akpm@linux-foundation.org, "Theodore Ts'o" , "Jan Kara" Date: Thu, 01 Jun 2017 16:43:15 +0100 Message-ID: X-Mailer: LinuxStableQueue (scripts by bwh) Subject: [PATCH 3.16 027/212] ext4: avoid deadlock when expanding inode size In-Reply-To: X-SA-Exim-Connect-IP: 82.70.136.246 X-SA-Exim-Mail-From: ben@decadent.org.uk X-SA-Exim-Scanned: No (on shadbolt.decadent.org.uk); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2590 Lines: 88 3.16.44-rc1 review patch. If anyone has any objections, please let me know. ------------------ From: Jan Kara commit 2e81a4eeedcaa66e35f58b81e0755b87057ce392 upstream. When we need to move xattrs into external xattr block, we call ext4_xattr_block_set() from ext4_expand_extra_isize_ea(). That may end up calling ext4_mark_inode_dirty() again which will recurse back into the inode expansion code leading to deadlocks. Protect from recursion using EXT4_STATE_NO_EXPAND inode flag and move its management into ext4_expand_extra_isize_ea() since its manipulation is safe there (due to xattr_sem) from possible races with ext4_xattr_set_handle() which plays with it as well. Signed-off-by: Jan Kara Signed-off-by: Theodore Ts'o [bwh: Backported to 3.16: adjust context[ Signed-off-by: Ben Hutchings --- fs/ext4/inode.c | 2 -- fs/ext4/xattr.c | 19 +++++++++++++------ 2 files changed, 13 insertions(+), 8 deletions(-) --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5062,8 +5062,6 @@ int ext4_mark_inode_dirty(handle_t *hand sbi->s_want_extra_isize, iloc, handle); if (ret) { - ext4_set_inode_state(inode, - EXT4_STATE_NO_EXPAND); if (mnt_count != le16_to_cpu(sbi->s_es->s_mnt_count)) { ext4_warning(inode->i_sb, --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -1277,11 +1277,13 @@ int ext4_expand_extra_isize_ea(struct in int s_min_extra_isize = le16_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_min_extra_isize); down_write(&EXT4_I(inode)->xattr_sem); + /* + * Set EXT4_STATE_NO_EXPAND to avoid recursion when marking inode dirty + */ + ext4_set_inode_state(inode, EXT4_STATE_NO_EXPAND); retry: - if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) { - up_write(&EXT4_I(inode)->xattr_sem); - return 0; - } + if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) + goto out; header = IHDR(inode, raw_inode); entry = IFIRST(header); @@ -1306,8 +1308,7 @@ retry: (void *)header, total_ino, inode->i_sb->s_blocksize); EXT4_I(inode)->i_extra_isize = new_extra_isize; - error = 0; - goto cleanup; + goto out; } /* @@ -1467,6 +1468,8 @@ retry: kfree(bs); } brelse(bh); +out: + ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND); up_write(&EXT4_I(inode)->xattr_sem); return 0; @@ -1478,6 +1481,10 @@ cleanup: kfree(is); kfree(bs); brelse(bh); + /* + * We deliberately leave EXT4_STATE_NO_EXPAND set here since inode + * size expansion failed. + */ up_write(&EXT4_I(inode)->xattr_sem); return error; }