From patchwork Thu Nov 22 08:36:28 2012 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lukas Czerner X-Patchwork-Id: 201172 Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id B56982C008A for ; Fri, 23 Nov 2012 06:39:57 +1100 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756957Ab2KVTbr (ORCPT ); Thu, 22 Nov 2012 14:31:47 -0500 Received: from mx1.redhat.com ([209.132.183.28]:12257 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755449Ab2KVTbp (ORCPT ); Thu, 22 Nov 2012 14:31:45 -0500 Received: from int-mx09.intmail.prod.int.phx2.redhat.com (int-mx09.intmail.prod.int.phx2.redhat.com [10.5.11.22]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id qAM8ae5E014464 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Thu, 22 Nov 2012 03:36:40 -0500 Received: from dhcp-1-104.brq.redhat.com (dhcp-1-104.brq.redhat.com [10.34.1.104]) by int-mx09.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id qAM8aawe025686; Thu, 22 Nov 2012 03:36:38 -0500 From: Lukas Czerner To: linux-ext4@vger.kernel.org Cc: gnehzuil.liu@gmail.com, tytso@mit.edu, dmonakhov@openvz.org, bergwolf@gmail.com, Lukas Czerner Subject: [PATCH 1/2 v5] ext4: Prevent race while walking extent tree Date: Thu, 22 Nov 2012 09:36:28 +0100 Message-Id: <1353573389-5388-2-git-send-email-lczerner@redhat.com> In-Reply-To: <1353573389-5388-1-git-send-email-lczerner@redhat.com> References: <1353573389-5388-1-git-send-email-lczerner@redhat.com> X-Scanned-By: MIMEDefang 2.68 on 10.5.11.22 Sender: linux-ext4-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-ext4@vger.kernel.org Currently ext4_ext_walk_space() only takes i_data_sem for read when searching for the extent at given block with ext4_ext_find_extent(). Then it drops the lock and the extent tree can be changed at will. However later on we're searching for the 'next' extent, but the extent tree might already have changed, so the information might not be accurate. In fact we can hit BUG_ON(end <= start) if the extent got inserted into the tree after the one we found and before the block we were searching for. This has been reproduced by running xfstests 225 in loop on s390x architecture, but theoretically we could hit this on any other architecture as well, but probably not as often. Moreover the extent currently in delayed allocation might be allocated after we search the extent tree and before we search extent status tree delayed buffers resulting in those delayed buffers being completely missed, even though completely written and allocated. We fix all those problems in several steps: 1. remove unnecessary callback indirection 2. rename functions ext4_ext_walk_space -> ext4_fill_fiemap_extents ext4_ext_fiemap_cb -> ext4_find_delayed_extent 3. move fiemap_fill_next_extent() into ext4_fill_fiemap_extents() 4. hold the i_data_sem for: ext4_ext_find_extent() ext4_ext_next_allocated_block() ext4_find_delayed_extent() 5. call fiemap_fill_next_extent after releasing the i_data_sem 6. move path reinitialization into the critical section. Signed-off-by: Lukas Czerner --- v4: Reworked v5: Change check for existing delayed extent. Update comments fs/ext4/ext4_extents.h | 14 ----- fs/ext4/extents.c | 136 +++++++++++++++++++++++++++--------------------- 2 files changed, 76 insertions(+), 74 deletions(-) diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h index 603bb11..173b6c5 100644 --- a/fs/ext4/ext4_extents.h +++ b/fs/ext4/ext4_extents.h @@ -144,20 +144,6 @@ struct ext4_ext_path { */ /* - * to be called by ext4_ext_walk_space() - * negative retcode - error - * positive retcode - signal for ext4_ext_walk_space(), see below - * callback must return valid extent (passed or newly created) - */ -typedef int (*ext_prepare_callback)(struct inode *, ext4_lblk_t, - struct ext4_ext_cache *, - struct ext4_extent *, void *); - -#define EXT_CONTINUE 0 -#define EXT_BREAK 1 -#define EXT_REPEAT 2 - -/* * Maximum number of logical blocks in a file; ext4_extent's ee_block is * __le32. */ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index d3dd618..fbe7dc2 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -109,6 +109,9 @@ static int ext4_split_extent_at(handle_t *handle, int split_flag, int flags); +static int ext4_find_delayed_extent(struct inode *inode, + struct ext4_ext_cache *newex); + static int ext4_ext_truncate_extend_restart(handle_t *handle, struct inode *inode, int needed) @@ -1959,27 +1962,33 @@ cleanup: return err; } -static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, - ext4_lblk_t num, ext_prepare_callback func, - void *cbdata) +static int ext4_fill_fiemap_extents(struct inode *inode, + ext4_lblk_t block, ext4_lblk_t num, + struct fiemap_extent_info *fieinfo) { struct ext4_ext_path *path = NULL; struct ext4_ext_cache cbex; struct ext4_extent *ex; - ext4_lblk_t next, start = 0, end = 0; + ext4_lblk_t next, next_del, start = 0, end = 0; ext4_lblk_t last = block + num; - int depth, exists, err = 0; - - BUG_ON(func == NULL); - BUG_ON(inode == NULL); + int exists, depth = 0, err = 0; + unsigned int flags = 0; + unsigned char blksize_bits = inode->i_sb->s_blocksize_bits; while (block < last && block != EXT_MAX_BLOCKS) { num = last - block; /* find extent for this block */ down_read(&EXT4_I(inode)->i_data_sem); + + if (path && ext_depth(inode) != depth) { + /* depth was changed. we have to realloc path */ + kfree(path); + path = NULL; + } + path = ext4_ext_find_extent(inode, block, path); - up_read(&EXT4_I(inode)->i_data_sem); if (IS_ERR(path)) { + up_read(&EXT4_I(inode)->i_data_sem); err = PTR_ERR(path); path = NULL; break; @@ -1987,13 +1996,16 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, depth = ext_depth(inode); if (unlikely(path[depth].p_hdr == NULL)) { + up_read(&EXT4_I(inode)->i_data_sem); EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); err = -EIO; break; } ex = path[depth].p_ext; next = ext4_ext_next_allocated_block(path); + ext4_ext_drop_refs(path); + flags = 0; exists = 0; if (!ex) { /* there is no extent yet, so try to allocate @@ -2037,30 +2049,54 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, cbex.ec_block = le32_to_cpu(ex->ee_block); cbex.ec_len = ext4_ext_get_actual_len(ex); cbex.ec_start = ext4_ext_pblock(ex); + if (ext4_ext_is_uninitialized(ex)) + flags |= FIEMAP_EXTENT_UNWRITTEN; } + /* + * Find delayed extent and update cbex accordingly. We call + * it even in !exists case to find out whether cbex is the + * last existing extent or not. + */ + next_del = ext4_find_delayed_extent(inode, &cbex); + if (!exists && next_del) { + exists = 1; + flags |= FIEMAP_EXTENT_DELALLOC; + } + up_read(&EXT4_I(inode)->i_data_sem); + if (unlikely(cbex.ec_len == 0)) { EXT4_ERROR_INODE(inode, "cbex.ec_len == 0"); err = -EIO; break; } - err = func(inode, next, &cbex, ex, cbdata); - ext4_ext_drop_refs(path); - - if (err < 0) - break; - if (err == EXT_REPEAT) - continue; - else if (err == EXT_BREAK) { - err = 0; - break; + /* This is possible iff next == next_del == EXT_MAX_BLOCKS */ + if (next == next_del) { + flags |= FIEMAP_EXTENT_LAST; + if (unlikely(next_del != EXT_MAX_BLOCKS || + next != EXT_MAX_BLOCKS)) { + EXT4_ERROR_INODE(inode, + "next extent == %u, next " + "delalloc extent = %u", + next, next_del); + err = -EIO; + break; + } } - if (ext_depth(inode) != depth) { - /* depth was changed. we have to realloc path */ - kfree(path); - path = NULL; + if (exists) { + err = fiemap_fill_next_extent(fieinfo, + (__u64)cbex.ec_block << blksize_bits, + (__u64)cbex.ec_start << blksize_bits, + (__u64)cbex.ec_len << blksize_bits, + flags); + if (err < 0) + break; + if (err == 1) { + err = 0; + break; + } } block = cbex.ec_block + cbex.ec_len; @@ -4493,26 +4529,23 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, } /* - * Callback function called for each extent to gather FIEMAP information. + * If newex is not existing extent (newex->ec_start equals zero) find + * delayed extent at start of newex and update newex accordingly and + * return start of the next delayed extent. + * + * If newex is existing extent (newex->ec_start is not equal zero) + * return start of next delayed extent or EXT_MAX_BLOCKS if no delayed + * extent found. Leave newex unmodified. */ -static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next, - struct ext4_ext_cache *newex, struct ext4_extent *ex, - void *data) +static int ext4_find_delayed_extent(struct inode *inode, + struct ext4_ext_cache *newex) { struct extent_status es; - __u64 logical; - __u64 physical; - __u64 length; - __u32 flags = 0; ext4_lblk_t next_del; - int ret = 0; - struct fiemap_extent_info *fieinfo = data; - unsigned char blksize_bits; es.start = newex->ec_block; next_del = ext4_es_find_extent(inode, &es); - next = min(next_del, next); if (newex->ec_start == 0) { /* * No extent in extent-tree contains block @newex->ec_start, @@ -4520,37 +4553,19 @@ static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next, */ if (es.len == 0) /* A hole found. */ - return EXT_CONTINUE; + return 0; if (es.start > newex->ec_block) { /* A hole found. */ newex->ec_len = min(es.start - newex->ec_block, newex->ec_len); - return EXT_CONTINUE; + return 0; } - flags |= FIEMAP_EXTENT_DELALLOC; newex->ec_len = es.start + es.len - newex->ec_block; } - if (ex && ext4_ext_is_uninitialized(ex)) - flags |= FIEMAP_EXTENT_UNWRITTEN; - - if (next == EXT_MAX_BLOCKS) - flags |= FIEMAP_EXTENT_LAST; - - blksize_bits = inode->i_sb->s_blocksize_bits; - logical = (__u64)newex->ec_block << blksize_bits; - physical = (__u64)newex->ec_start << blksize_bits; - length = (__u64)newex->ec_len << blksize_bits; - - ret = fiemap_fill_next_extent(fieinfo, logical, physical, - length, flags); - if (ret < 0) - return ret; - if (ret == 1) - return EXT_BREAK; - return EXT_CONTINUE; + return next_del; } /* fiemap flags we can handle specified here */ #define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) @@ -4772,6 +4787,7 @@ out_mutex: mutex_unlock(&inode->i_mutex); return err; } + int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { @@ -4799,11 +4815,11 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1; /* - * Walk the extent tree gathering extent information. - * ext4_ext_fiemap_cb will push extents back to user. + * Walk the extent tree gathering extent information + * and pushing extents back to the user. */ - error = ext4_ext_walk_space(inode, start_blk, len_blks, - ext4_ext_fiemap_cb, fieinfo); + error = ext4_fill_fiemap_extents(inode, start_blk, + len_blks, fieinfo); } return error;