[PATCH v2.6.34.y 01/28] ext4: check missed return value in ext4_sync_file()

[PATCH v2.6.34.y 01/28] ext4: check missed return value in ext4_sync_file()

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 0671e704658b9f26f85e78d51176daa861f955c7 upstream (as of v2.6.34-git13)

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/fsync.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index 0d0c323..42bd94a 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -101,7 +101,7 @@ int ext4_sync_file(struct file *file, struct dentry *dentry, int datasync)
 		    (journal->j_fs_dev != journal->j_dev) &&
 		    (journal->j_flags & JBD2_BARRIER))
 			blkdev_issue_flush(inode->i_sb->s_bdev, NULL);
-		jbd2_log_wait_commit(journal, commit_tid);
+		ret = jbd2_log_wait_commit(journal, commit_tid);
 	} else if (journal->j_flags & JBD2_BARRIER)
 		blkdev_issue_flush(inode->i_sb->s_bdev, NULL);
 	return ret;
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 02/28] ext4: fix memory leaks in error path handling of ext4_ext_zeroout()

[Theodore Ts'o]

From: Jing Zhang <zj.barak@gmail.com>

commit b720303df7352d4a7a1f61e467e0a124913c0d41 upstream (as of v2.6.34-git13)

When EIO occurs after bio is submitted, there is no memory free
operation for bio, which results in memory leakage. And there is also
no check against bio_alloc() for bio.

Acked-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Signed-off-by: Jing Zhang <zj.barak@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c |   15 ++++++++-------
 1 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 236b834..228eeaf 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2544,7 +2544,7 @@ static void bi_complete(struct bio *bio, int error)
 /* FIXME!! we need to try to merge to left or right after zero-out  */
 static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
 {
-	int ret = -EIO;
+	int ret;
 	struct bio *bio;
 	int blkbits, blocksize;
 	sector_t ee_pblock;
@@ -2568,6 +2568,9 @@ static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
 			len = ee_len;
 
 		bio = bio_alloc(GFP_NOIO, len);
+		if (!bio)
+			return -ENOMEM;
+
 		bio->bi_sector = ee_pblock;
 		bio->bi_bdev   = inode->i_sb->s_bdev;
 
@@ -2595,17 +2598,15 @@ static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
 		submit_bio(WRITE, bio);
 		wait_for_completion(&event);
 
-		if (test_bit(BIO_UPTODATE, &bio->bi_flags))
-			ret = 0;
-		else {
-			ret = -EIO;
-			break;
+		if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
+			bio_put(bio);
+			return -EIO;
 		}
 		bio_put(bio);
 		ee_len    -= done;
 		ee_pblock += done  << (blkbits - 9);
 	}
-	return ret;
+	return 0;
 }
 
 #define EXT4_EXT_ZERO_LEN 7
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 03/28] ext4: Remove unnecessary call to ext4_get_group_desc() in mballoc

[Theodore Ts'o]

From: Jing Zhang <zj.barak@gmail.com>

commit 62e823a2cba18509ee826d775270e8ef9071b5bc upstream (as of v2.6.34-git13)

Signed-off-by: Jing Zhang <zj.barak@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |    2 --
 1 files changed, 0 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index b423a36..0550ea3 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2025,7 +2025,6 @@ repeat:
 
 		for (i = 0; i < ngroups; group++, i++) {
 			struct ext4_group_info *grp;
-			struct ext4_group_desc *desc;
 
 			if (group == ngroups)
 				group = 0;
@@ -2048,7 +2047,6 @@ repeat:
 			}
 
 			ac->ac_groups_scanned++;
-			desc = ext4_get_group_desc(sb, group, NULL);
 			if (cr == 0)
 				ext4_mb_simple_scan_group(ac, &e4b);
 			else if (cr == 1 &&
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 04/28] ext4: rename ext4_mb_release_desc() to ext4_mb_unload_buddy()

[Theodore Ts'o]

From: Jing Zhang <zj.barak@gmail.com>

commit e39e07fdfd98be8650385f12a7b81d6adc547510 upstream (as of v2.6.34-git13)

This function cleans up after ext4_mb_load_buddy(), so the renaming
makes the code clearer.

Signed-off-by: Jing Zhang <zj.barak@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |   24 ++++++++++++------------
 1 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 0550ea3..8639d5a 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -1150,7 +1150,7 @@ err:
 	return ret;
 }
 
-static void ext4_mb_release_desc(struct ext4_buddy *e4b)
+static void ext4_mb_unload_buddy(struct ext4_buddy *e4b)
 {
 	if (e4b->bd_bitmap_page)
 		page_cache_release(e4b->bd_bitmap_page);
@@ -1617,7 +1617,7 @@ int ext4_mb_try_best_found(struct ext4_allocation_context *ac,
 	}
 
 	ext4_unlock_group(ac->ac_sb, group);
-	ext4_mb_release_desc(e4b);
+	ext4_mb_unload_buddy(e4b);
 
 	return 0;
 }
@@ -1672,7 +1672,7 @@ int ext4_mb_find_by_goal(struct ext4_allocation_context *ac,
 		ext4_mb_use_best_found(ac, e4b);
 	}
 	ext4_unlock_group(ac->ac_sb, group);
-	ext4_mb_release_desc(e4b);
+	ext4_mb_unload_buddy(e4b);
 
 	return 0;
 }
@@ -2042,7 +2042,7 @@ repeat:
 			if (!ext4_mb_good_group(ac, group, cr)) {
 				/* someone did allocation from this group */
 				ext4_unlock_group(sb, group);
-				ext4_mb_release_desc(&e4b);
+				ext4_mb_unload_buddy(&e4b);
 				continue;
 			}
 
@@ -2056,7 +2056,7 @@ repeat:
 				ext4_mb_complex_scan_group(ac, &e4b);
 
 			ext4_unlock_group(sb, group);
-			ext4_mb_release_desc(&e4b);
+			ext4_mb_unload_buddy(&e4b);
 
 			if (ac->ac_status != AC_STATUS_CONTINUE)
 				break;
@@ -2146,7 +2146,7 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v)
 	ext4_lock_group(sb, group);
 	memcpy(&sg, ext4_get_group_info(sb, group), i);
 	ext4_unlock_group(sb, group);
-	ext4_mb_release_desc(&e4b);
+	ext4_mb_unload_buddy(&e4b);
 
 	seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg.info.bb_free,
 			sg.info.bb_fragments, sg.info.bb_first_free);
@@ -2566,7 +2566,7 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
 		}
 		ext4_unlock_group(sb, entry->group);
 		kmem_cache_free(ext4_free_ext_cachep, entry);
-		ext4_mb_release_desc(&e4b);
+		ext4_mb_unload_buddy(&e4b);
 	}
 
 	mb_debug(1, "freed %u blocks in %u structures\n", count, count2);
@@ -3695,7 +3695,7 @@ out:
 	ext4_unlock_group(sb, group);
 	if (ac)
 		kmem_cache_free(ext4_ac_cachep, ac);
-	ext4_mb_release_desc(&e4b);
+	ext4_mb_unload_buddy(&e4b);
 	put_bh(bitmap_bh);
 	return free;
 }
@@ -3799,7 +3799,7 @@ repeat:
 		if (bitmap_bh == NULL) {
 			ext4_error(sb, "Error reading block bitmap for %u",
 					group);
-			ext4_mb_release_desc(&e4b);
+			ext4_mb_unload_buddy(&e4b);
 			continue;
 		}
 
@@ -3808,7 +3808,7 @@ repeat:
 		ext4_mb_release_inode_pa(&e4b, bitmap_bh, pa, ac);
 		ext4_unlock_group(sb, group);
 
-		ext4_mb_release_desc(&e4b);
+		ext4_mb_unload_buddy(&e4b);
 		put_bh(bitmap_bh);
 
 		list_del(&pa->u.pa_tmp_list);
@@ -4072,7 +4072,7 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb,
 		ext4_mb_release_group_pa(&e4b, pa, ac);
 		ext4_unlock_group(sb, group);
 
-		ext4_mb_release_desc(&e4b);
+		ext4_mb_unload_buddy(&e4b);
 		list_del(&pa->u.pa_tmp_list);
 		call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback);
 	}
@@ -4608,7 +4608,7 @@ do_more:
 		atomic_add(count, &sbi->s_flex_groups[flex_group].free_blocks);
 	}
 
-	ext4_mb_release_desc(&e4b);
+	ext4_mb_unload_buddy(&e4b);
 
 	freed += count;
 
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 05/28] ext4: allow defrag (EXT4_IOC_MOVE_EXT) in 32bit compat mode

[Theodore Ts'o]

From: Christian Borntraeger <borntraeger@de.ibm.com>

commit b684b2ee9409f2890a8b3aea98525bbe5f84e276 upstream (as of v2.6.34-git13)

I have an x86_64 kernel with i386 userspace. e4defrag fails on the
EXT4_IOC_MOVE_EXT ioctl because it is not wired up for the compat
case. It seems that struct move_extent is compat save, only types
with fixed widths are used:
{
        __u32 reserved;         /* should be zero */
        __u32 donor_fd;         /* donor file descriptor */
        __u64 orig_start;       /* logical start offset in block for orig */
        __u64 donor_start;      /* logical start offset in block for donor */
        __u64 len;              /* block length to be moved */
        __u64 moved_len;        /* moved block length */
};

Lets just wire up EXT4_IOC_MOVE_EXT for the compat case.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
CC: Akira Fujita <a-fujita@rs.jp.nec.com>
---
 fs/ext4/ioctl.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 016d024..66fa0b0 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -375,6 +375,8 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 		break;
 	case EXT4_IOC_GROUP_ADD:
 		break;
+	case EXT4_IOC_MOVE_EXT:
+		break;
 	default:
 		return -ENOIOCTLCMD;
 	}
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 06/28] ext4: fix quota accounting in case of fallocate

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 35121c9860316d7799cea0fbc359a9186e7c2747 upstream (as of v2.6.34-git13)

allocated_meta_data is already included in 'used' variable.

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/inode.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 81d6054..55bfcd9 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1126,7 +1126,8 @@ void ext4_da_update_reserve_space(struct inode *inode,
 		 */
 		if (allocated_meta_blocks)
 			dquot_claim_block(inode, allocated_meta_blocks);
-		dquot_release_reservation_block(inode, mdb_free + used);
+		dquot_release_reservation_block(inode, mdb_free + used -
+						allocated_meta_blocks);
 	}
 
 	/*
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 07/28] ext4: check s_log_groups_per_flex in online resize code

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit 42007efd569f1cf3bfb9a61da60ef6c2179508ca upstream (as of v2.6.34-git13)

If groups_per_flex < 2, sbi->s_flex_groups[] doesn't get filled out,
and every other access to this first tests s_log_groups_per_flex;
same thing needs to happen in resize or we'll wander off into
a null pointer when doing an online resize of the file system.

Thanks to Christoph Biedl, who came up with the trivial testcase:

# truncate --size 128M fsfile
# mkfs.ext3 -F fsfile
# tune2fs -O extents,uninit_bg,dir_index,flex_bg,huge_file,dir_nlink,extra_isize fsfile
# e2fsck -yDf -C0 fsfile
# truncate --size 132M fsfile
# losetup /dev/loop0 fsfile
# mount /dev/loop0 mnt
# resize2fs -p /dev/loop0

	https://bugzilla.kernel.org/show_bug.cgi?id=13549

Reported-by: Alessandro Polverini <alex@nibbles.it>
Test-case-by: Christoph Biedl  <bugzilla.kernel.bpeb@manchmal.in-ulm.de>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/resize.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 5692c48..6df797e 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -911,7 +911,8 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input)
 	percpu_counter_add(&sbi->s_freeinodes_counter,
 			   EXT4_INODES_PER_GROUP(sb));
 
-	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
+	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FLEX_BG) &&
+	    sbi->s_log_groups_per_flex) {
 		ext4_group_t flex_group;
 		flex_group = ext4_flex_group(sbi, input->group);
 		atomic_add(input->free_blocks_count,
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 08/28] ext4: don't return to userspace after freezing the fs with a mutex held

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit 6b0310fbf087ad6e9e3b8392adca97cd77184084 upstream (as of v2.6.34-git13)

ext4_freeze() used jbd2_journal_lock_updates() which takes
the j_barrier mutex, and then returns to userspace.  The
kernel does not like this:

================================================
[ BUG: lock held when returning to user space! ]
------------------------------------------------
lvcreate/1075 is leaving the kernel with locks still held!
1 lock held by lvcreate/1075:
 #0:  (&journal->j_barrier){+.+...}, at: [<ffffffff811c6214>]
jbd2_journal_lock_updates+0xe1/0xf0

Use vfs_check_frozen() added to ext4_journal_start_sb() and
ext4_force_commit() instead.

Addresses-Red-Hat-Bugzilla: #568503

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/super.c |   20 ++++++++++----------
 1 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index e14d22c..00d09f5 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -241,6 +241,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
 	if (sb->s_flags & MS_RDONLY)
 		return ERR_PTR(-EROFS);
 
+	vfs_check_frozen(sb, SB_FREEZE_WRITE);
 	/* Special case here: if the journal has aborted behind our
 	 * backs (eg. EIO in the commit thread), then we still need to
 	 * take the FS itself readonly cleanly. */
@@ -3485,8 +3486,10 @@ int ext4_force_commit(struct super_block *sb)
 		return 0;
 
 	journal = EXT4_SB(sb)->s_journal;
-	if (journal)
+	if (journal) {
+		vfs_check_frozen(sb, SB_FREEZE_WRITE);
 		ret = ext4_journal_force_commit(journal);
+	}
 
 	return ret;
 }
@@ -3535,18 +3538,16 @@ static int ext4_freeze(struct super_block *sb)
 	 * the journal.
 	 */
 	error = jbd2_journal_flush(journal);
-	if (error < 0) {
-	out:
-		jbd2_journal_unlock_updates(journal);
-		return error;
-	}
+	if (error < 0)
+		goto out;
 
 	/* Journal blocked and flushed, clear needs_recovery flag. */
 	EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
 	error = ext4_commit_super(sb, 1);
-	if (error)
-		goto out;
-	return 0;
+out:
+	/* we rely on s_frozen to stop further updates */
+	jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
+	return error;
 }
 
 /*
@@ -3563,7 +3564,6 @@ static int ext4_unfreeze(struct super_block *sb)
 	EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
 	ext4_commit_super(sb, 1);
 	unlock_super(sb);
-	jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
 	return 0;
 }
 
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 09/28] ext4: stop issuing discards if not supported by device

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit a30eec2a8650a77f754e84b2e15f062fe652baa7 upstream (as of v2.6.34-git13)

Turn off issuance of discard requests if the device does
not support it - similar to the action we take for barriers.
This will save a little computation time if a non-discardable
device is mounted with -o discard, and also makes it obvious
that it's not doing what was asked at mount time ...

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |    8 +++++++-
 1 files changed, 7 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 8639d5a..9c6a40d 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2534,6 +2534,7 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
 			 entry->count, entry->group, entry);
 
 		if (test_opt(sb, DISCARD)) {
+			int ret;
 			ext4_fsblk_t discard_block;
 
 			discard_block = entry->start_blk +
@@ -2541,7 +2542,12 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn)
 			trace_ext4_discard_blocks(sb,
 					(unsigned long long)discard_block,
 					entry->count);
-			sb_issue_discard(sb, discard_block, entry->count);
+			ret = sb_issue_discard(sb, discard_block, entry->count);
+			if (ret == EOPNOTSUPP) {
+				ext4_warning(sb, __func__,
+					"discard not supported, disabling");
+				clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD);
+			}
 		}
 
 		err = ext4_mb_load_buddy(sb, entry->group, &e4b);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 10/28] ext4: don't scan/accumulate more pages than mballoc will allocate

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit c445e3e0a5c2804524dec6e55f66d63f6bc5bc3e upstream (as of v2.6.34-git13)

There was a bug reported on RHEL5 that a 10G dd on a 12G box
had a very, very slow sync after that.

At issue was the loop in write_cache_pages scanning all the way
to the end of the 10G file, even though the subsequent call
to mpage_da_submit_io would only actually write a smallish amt; then
we went back to the write_cache_pages loop ... wasting tons of time
in calling __mpage_da_writepage for thousands of pages we would
just revisit (many times) later.

Upstream it's not such a big issue for sys_sync because we get
to the loop with a much smaller nr_to_write, which limits the loop.

However, talking with Aneesh he realized that fsync upstream still
gets here with a very large nr_to_write and we face the same problem.

This patch makes mpage_add_bh_to_extent stop the loop after we've
accumulated 2048 pages, by setting mpd->io_done = 1; which ultimately
causes the write_cache_pages loop to break.

Repeating the test with a dirty_ratio of 80 (to leave something for
fsync to do), I don't see huge IO performance gains, but the reduction
in cpu usage is striking: 80% usage with stock, and 2% with the
below patch.  Instrumenting the loop in write_cache_pages clearly
shows that we are wasting time here.

Eventually we need to change mpage_da_map_pages() also submit its I/O
to the block layer, subsuming mpage_da_submit_io(), and then change it
call ext4_get_blocks() multiple times.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/inode.c |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 55bfcd9..89a31e8 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2350,6 +2350,15 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd,
 	sector_t next;
 	int nrblocks = mpd->b_size >> mpd->inode->i_blkbits;
 
+	/*
+	 * XXX Don't go larger than mballoc is willing to allocate
+	 * This is a stopgap solution.  We eventually need to fold
+	 * mpage_da_submit_io() into this function and then call
+	 * ext4_get_blocks() multiple times in a loop
+	 */
+	if (nrblocks >= 8*1024*1024/mpd->inode->i_sb->s_blocksize)
+		goto flush_it;
+
 	/* check if thereserved journal credits might overflow */
 	if (!(EXT4_I(mpd->inode)->i_flags & EXT4_EXTENTS_FL)) {
 		if (nrblocks >= EXT4_MAX_TRANS_DATA) {
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 11/28] ext4: Do not zero out uninitialized extents beyond i_size

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 21ca087a3891efab4d45488db8febee474d26c68 upstream (as of v2.6.34-git13)

The extents code will sometimes zero out blocks and mark them as
initialized instead of splitting an extent into several smaller ones.
This optimization however, causes problems if the extent is beyond
i_size because fsck will complain if there are uninitialized blocks
after i_size as this can not be distinguished from an inode that has
an incorrect i_size field.

https://bugzilla.kernel.org/show_bug.cgi?id=15742

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c |   67 ++++++++++++++++++++++++++++++++++++++++------------
 1 files changed, 51 insertions(+), 16 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 228eeaf..ee611da 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2631,11 +2631,21 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 	struct ext4_extent *ex2 = NULL;
 	struct ext4_extent *ex3 = NULL;
 	struct ext4_extent_header *eh;
-	ext4_lblk_t ee_block;
+	ext4_lblk_t ee_block, eof_block;
 	unsigned int allocated, ee_len, depth;
 	ext4_fsblk_t newblock;
 	int err = 0;
 	int ret = 0;
+	int may_zeroout;
+
+	ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
+		"block %llu, max_blocks %u\n", inode->i_ino,
+		(unsigned long long)iblock, max_blocks);
+
+	eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
+		inode->i_sb->s_blocksize_bits;
+	if (eof_block < iblock + max_blocks)
+		eof_block = iblock + max_blocks;
 
 	depth = ext_depth(inode);
 	eh = path[depth].p_hdr;
@@ -2644,16 +2654,23 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 	ee_len = ext4_ext_get_actual_len(ex);
 	allocated = ee_len - (iblock - ee_block);
 	newblock = iblock - ee_block + ext_pblock(ex);
+
 	ex2 = ex;
 	orig_ex.ee_block = ex->ee_block;
 	orig_ex.ee_len   = cpu_to_le16(ee_len);
 	ext4_ext_store_pblock(&orig_ex, ext_pblock(ex));
 
+	/*
+	 * It is safe to convert extent to initialized via explicit
+	 * zeroout only if extent is fully insde i_size or new_size.
+	 */
+	may_zeroout = ee_block + ee_len <= eof_block;
+
 	err = ext4_ext_get_access(handle, inode, path + depth);
 	if (err)
 		goto out;
 	/* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
-	if (ee_len <= 2*EXT4_EXT_ZERO_LEN) {
+	if (ee_len <= 2*EXT4_EXT_ZERO_LEN && may_zeroout) {
 		err =  ext4_ext_zeroout(inode, &orig_ex);
 		if (err)
 			goto fix_extent_len;
@@ -2684,7 +2701,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 	if (allocated > max_blocks) {
 		unsigned int newdepth;
 		/* If extent has less than EXT4_EXT_ZERO_LEN zerout directly */
-		if (allocated <= EXT4_EXT_ZERO_LEN) {
+		if (allocated <= EXT4_EXT_ZERO_LEN && may_zeroout) {
 			/*
 			 * iblock == ee_block is handled by the zerouout
 			 * at the beginning.
@@ -2760,7 +2777,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 		ex3->ee_len = cpu_to_le16(allocated - max_blocks);
 		ext4_ext_mark_uninitialized(ex3);
 		err = ext4_ext_insert_extent(handle, inode, path, ex3, 0);
-		if (err == -ENOSPC) {
+		if (err == -ENOSPC && may_zeroout) {
 			err =  ext4_ext_zeroout(inode, &orig_ex);
 			if (err)
 				goto fix_extent_len;
@@ -2784,8 +2801,10 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 		 * update the extent length after successful insert of the
 		 * split extent
 		 */
-		orig_ex.ee_len = cpu_to_le16(ee_len -
-						ext4_ext_get_actual_len(ex3));
+		ee_len -= ext4_ext_get_actual_len(ex3);
+		orig_ex.ee_len = cpu_to_le16(ee_len);
+		may_zeroout = ee_block + ee_len <= eof_block;
+
 		depth = newdepth;
 		ext4_ext_drop_refs(path);
 		path = ext4_ext_find_extent(inode, iblock, path);
@@ -2809,7 +2828,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 		 * otherwise give the extent a chance to merge to left
 		 */
 		if (le16_to_cpu(orig_ex.ee_len) <= EXT4_EXT_ZERO_LEN &&
-							iblock != ee_block) {
+			iblock != ee_block && may_zeroout) {
 			err =  ext4_ext_zeroout(inode, &orig_ex);
 			if (err)
 				goto fix_extent_len;
@@ -2878,7 +2897,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
 	goto out;
 insert:
 	err = ext4_ext_insert_extent(handle, inode, path, &newex, 0);
-	if (err == -ENOSPC) {
+	if (err == -ENOSPC && may_zeroout) {
 		err =  ext4_ext_zeroout(inode, &orig_ex);
 		if (err)
 			goto fix_extent_len;
@@ -2938,14 +2957,21 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 	struct ext4_extent *ex2 = NULL;
 	struct ext4_extent *ex3 = NULL;
 	struct ext4_extent_header *eh;
-	ext4_lblk_t ee_block;
+	ext4_lblk_t ee_block, eof_block;
 	unsigned int allocated, ee_len, depth;
 	ext4_fsblk_t newblock;
 	int err = 0;
+	int may_zeroout;
+
+	ext_debug("ext4_split_unwritten_extents: inode %lu, logical"
+		"block %llu, max_blocks %u\n", inode->i_ino,
+		(unsigned long long)iblock, max_blocks);
+
+	eof_block = (inode->i_size + inode->i_sb->s_blocksize - 1) >>
+		inode->i_sb->s_blocksize_bits;
+	if (eof_block < iblock + max_blocks)
+		eof_block = iblock + max_blocks;
 
-	ext_debug("ext4_split_unwritten_extents: inode %lu,"
-		  "iblock %llu, max_blocks %u\n", inode->i_ino,
-		  (unsigned long long)iblock, max_blocks);
 	depth = ext_depth(inode);
 	eh = path[depth].p_hdr;
 	ex = path[depth].p_ext;
@@ -2953,12 +2979,19 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 	ee_len = ext4_ext_get_actual_len(ex);
 	allocated = ee_len - (iblock - ee_block);
 	newblock = iblock - ee_block + ext_pblock(ex);
+
 	ex2 = ex;
 	orig_ex.ee_block = ex->ee_block;
 	orig_ex.ee_len   = cpu_to_le16(ee_len);
 	ext4_ext_store_pblock(&orig_ex, ext_pblock(ex));
 
 	/*
+	 * It is safe to convert extent to initialized via explicit
+	 * zeroout only if extent is fully insde i_size or new_size.
+	 */
+	may_zeroout = ee_block + ee_len <= eof_block;
+
+	/*
  	 * If the uninitialized extent begins at the same logical
  	 * block where the write begins, and the write completely
  	 * covers the extent, then we don't need to split it.
@@ -2992,7 +3025,7 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 		ex3->ee_len = cpu_to_le16(allocated - max_blocks);
 		ext4_ext_mark_uninitialized(ex3);
 		err = ext4_ext_insert_extent(handle, inode, path, ex3, flags);
-		if (err == -ENOSPC) {
+		if (err == -ENOSPC && may_zeroout) {
 			err =  ext4_ext_zeroout(inode, &orig_ex);
 			if (err)
 				goto fix_extent_len;
@@ -3016,8 +3049,10 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 		 * update the extent length after successful insert of the
 		 * split extent
 		 */
-		orig_ex.ee_len = cpu_to_le16(ee_len -
-						ext4_ext_get_actual_len(ex3));
+		ee_len -= ext4_ext_get_actual_len(ex3);
+		orig_ex.ee_len = cpu_to_le16(ee_len);
+		may_zeroout = ee_block + ee_len <= eof_block;
+
 		depth = newdepth;
 		ext4_ext_drop_refs(path);
 		path = ext4_ext_find_extent(inode, iblock, path);
@@ -3063,7 +3098,7 @@ static int ext4_split_unwritten_extents(handle_t *handle,
 	goto out;
 insert:
 	err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
-	if (err == -ENOSPC) {
+	if (err == -ENOSPC && may_zeroout) {
 		err =  ext4_ext_zeroout(inode, &orig_ex);
 		if (err)
 			goto fix_extent_len;
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 12/28] ext4: clean up inode bitmaps manipulation in ext4_free_inode

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit d17413c08cd2b1dd2bf2cfdbb0f7b736b2b2b15c upstrea (as of v2..34-git13)

- Reorganize locking scheme to batch two atomic operation in to one.
  This also allow us to state what healthy group must obey following rule
  ext4_free_inodes_count(sb, gdp) == ext4_count_free(inode_bitmap, NUM);
- Fix possible undefined pointer dereference.
- Even if group descriptor stats aren't accessible we have to update
  inode bitmaps.
- Move non-group members update out of group_lock.

Note: this commit has been observed to fix fs corruption problems
under heavy fs load

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ialloc.c |   81 ++++++++++++++++++++++++-----------------------------
 1 files changed, 37 insertions(+), 44 deletions(-)

diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 57f6eef..52618d5 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -240,56 +240,49 @@ void ext4_free_inode(handle_t *handle, struct inode *inode)
 	if (fatal)
 		goto error_return;
 
-	/* Ok, now we can actually update the inode bitmaps.. */
-	cleared = ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group),
-					bit, bitmap_bh->b_data);
-	if (!cleared)
-		ext4_error(sb, "bit already cleared for inode %lu", ino);
-	else {
-		gdp = ext4_get_group_desc(sb, block_group, &bh2);
-
+	fatal = -ESRCH;
+	gdp = ext4_get_group_desc(sb, block_group, &bh2);
+	if (gdp) {
 		BUFFER_TRACE(bh2, "get_write_access");
 		fatal = ext4_journal_get_write_access(handle, bh2);
-		if (fatal) goto error_return;
-
-		if (gdp) {
-			ext4_lock_group(sb, block_group);
-			count = ext4_free_inodes_count(sb, gdp) + 1;
-			ext4_free_inodes_set(sb, gdp, count);
-			if (is_directory) {
-				count = ext4_used_dirs_count(sb, gdp) - 1;
-				ext4_used_dirs_set(sb, gdp, count);
-				if (sbi->s_log_groups_per_flex) {
-					ext4_group_t f;
-
-					f = ext4_flex_group(sbi, block_group);
-					atomic_dec(&sbi->s_flex_groups[f].used_dirs);
-				}
+	}
+	ext4_lock_group(sb, block_group);
+	cleared = ext4_clear_bit(bit, bitmap_bh->b_data);
+	if (fatal || !cleared) {
+		ext4_unlock_group(sb, block_group);
+		goto out;
+	}
 
-			}
-			gdp->bg_checksum = ext4_group_desc_csum(sbi,
-							block_group, gdp);
-			ext4_unlock_group(sb, block_group);
-			percpu_counter_inc(&sbi->s_freeinodes_counter);
-			if (is_directory)
-				percpu_counter_dec(&sbi->s_dirs_counter);
-
-			if (sbi->s_log_groups_per_flex) {
-				ext4_group_t f;
-
-				f = ext4_flex_group(sbi, block_group);
-				atomic_inc(&sbi->s_flex_groups[f].free_inodes);
-			}
-		}
-		BUFFER_TRACE(bh2, "call ext4_handle_dirty_metadata");
-		err = ext4_handle_dirty_metadata(handle, NULL, bh2);
-		if (!fatal) fatal = err;
+	count = ext4_free_inodes_count(sb, gdp) + 1;
+	ext4_free_inodes_set(sb, gdp, count);
+	if (is_directory) {
+		count = ext4_used_dirs_count(sb, gdp) - 1;
+		ext4_used_dirs_set(sb, gdp, count);
+		percpu_counter_dec(&sbi->s_dirs_counter);
 	}
-	BUFFER_TRACE(bitmap_bh, "call ext4_handle_dirty_metadata");
-	err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh);
-	if (!fatal)
-		fatal = err;
-	sb->s_dirt = 1;
+	gdp->bg_checksum = ext4_group_desc_csum(sbi, block_group, gdp);
+	ext4_unlock_group(sb, block_group);
+
+	percpu_counter_inc(&sbi->s_freeinodes_counter);
+	if (sbi->s_log_groups_per_flex) {
+		ext4_group_t f = ext4_flex_group(sbi, block_group);
+
+		atomic_inc(&sbi->s_flex_groups[f].free_inodes);
+		if (is_directory)
+			atomic_dec(&sbi->s_flex_groups[f].used_dirs);
+	}
+	BUFFER_TRACE(bh2, "call ext4_handle_dirty_metadata");
+	fatal = ext4_handle_dirty_metadata(handle, NULL, bh2);
+out:
+	if (cleared) {
+		BUFFER_TRACE(bitmap_bh, "call ext4_handle_dirty_metadata");
+		err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh);
+		if (!fatal)
+			fatal = err;
+		sb->s_dirt = 1;
+	} else
+		ext4_error(sb, "bit already cleared for inode %lu", ino);
+
 error_return:
 	brelse(bitmap_bh);
 	ext4_std_error(sb, fatal);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 13/28] ext4: init statistics after journal recovery

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 84061e07c5fbbbf9dc8aef8fb750fc3a2dfc31f3 upstream (as of v2.6.34-git13)

Currently block/inode/dir counters initialized before journal was
recovered. In fact after journal recovery this info will probably
change. And freeblocks it critical for correct delalloc mode
accounting.

https://bugzilla.kernel.org/show_bug.cgi?id=15768

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Acked-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/super.c |   40 ++++++++++++++++++----------------------
 1 files changed, 18 insertions(+), 22 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 00d09f5..d83cdcb 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2794,24 +2794,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	get_random_bytes(&sbi->s_next_generation, sizeof(u32));
 	spin_lock_init(&sbi->s_next_gen_lock);
 
-	err = percpu_counter_init(&sbi->s_freeblocks_counter,
-			ext4_count_free_blocks(sb));
-	if (!err) {
-		err = percpu_counter_init(&sbi->s_freeinodes_counter,
-				ext4_count_free_inodes(sb));
-	}
-	if (!err) {
-		err = percpu_counter_init(&sbi->s_dirs_counter,
-				ext4_count_dirs(sb));
-	}
-	if (!err) {
-		err = percpu_counter_init(&sbi->s_dirtyblocks_counter, 0);
-	}
-	if (err) {
-		ext4_msg(sb, KERN_ERR, "insufficient memory");
-		goto failed_mount3;
-	}
-
 	sbi->s_stripe = ext4_get_stripe_size(sbi);
 	sbi->s_max_writeback_mb_bump = 128;
 
@@ -2911,6 +2893,20 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	set_task_ioprio(sbi->s_journal->j_task, journal_ioprio);
 
 no_journal:
+	err = percpu_counter_init(&sbi->s_freeblocks_counter,
+				  ext4_count_free_blocks(sb));
+	if (!err)
+		err = percpu_counter_init(&sbi->s_freeinodes_counter,
+					  ext4_count_free_inodes(sb));
+	if (!err)
+		err = percpu_counter_init(&sbi->s_dirs_counter,
+					  ext4_count_dirs(sb));
+	if (!err)
+		err = percpu_counter_init(&sbi->s_dirtyblocks_counter, 0);
+	if (err) {
+		ext4_msg(sb, KERN_ERR, "insufficient memory");
+		goto failed_mount_wq;
+	}
 	if (test_opt(sb, NOBH)) {
 		if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) {
 			ext4_msg(sb, KERN_WARNING, "Ignoring nobh option - "
@@ -3060,6 +3056,10 @@ failed_mount_wq:
 		jbd2_journal_destroy(sbi->s_journal);
 		sbi->s_journal = NULL;
 	}
+	percpu_counter_destroy(&sbi->s_freeblocks_counter);
+	percpu_counter_destroy(&sbi->s_freeinodes_counter);
+	percpu_counter_destroy(&sbi->s_dirs_counter);
+	percpu_counter_destroy(&sbi->s_dirtyblocks_counter);
 failed_mount3:
 	if (sbi->s_flex_groups) {
 		if (is_vmalloc_addr(sbi->s_flex_groups))
@@ -3067,10 +3067,6 @@ failed_mount3:
 		else
 			kfree(sbi->s_flex_groups);
 	}
-	percpu_counter_destroy(&sbi->s_freeblocks_counter);
-	percpu_counter_destroy(&sbi->s_freeinodes_counter);
-	percpu_counter_destroy(&sbi->s_dirs_counter);
-	percpu_counter_destroy(&sbi->s_dirtyblocks_counter);
 failed_mount2:
 	for (i = 0; i < db_count; i++)
 		brelse(sbi->s_group_desc[i]);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 14/28] quota: use flags interface for dquot alloc/free space

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit 56246f9ae4cfa95b460f9dfbcfb1b772d85db046 upstream (as of 2.6.34-git13)

Switch __dquot_alloc_space and __dquot_free_space to take flags
to indicate whether to warn and/or to reserve (or free reserve).

This is slightly more readable at the callpoints, and makes it
cleaner to add a "nofail" option in the next patch.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/quota/dquot.c         |    8 +++++---
 include/linux/quotaops.h |   25 ++++++++++++++-----------
 2 files changed, 19 insertions(+), 14 deletions(-)

diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 788b580..8c40c07 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -1488,11 +1488,12 @@ static void inode_decr_space(struct inode *inode, qsize_t number, int reserve)
 /*
  * This operation can block, but only after everything is updated
  */
-int __dquot_alloc_space(struct inode *inode, qsize_t number,
-		int warn, int reserve)
+int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags)
 {
 	int cnt, ret = 0;
 	char warntype[MAXQUOTAS];
+	int warn = flags & DQUOT_SPACE_WARN;
+	int reserve = flags & DQUOT_SPACE_RESERVE;
 
 	/*
 	 * First test before acquiring mutex - solves deadlocks when we
@@ -1612,10 +1613,11 @@ EXPORT_SYMBOL(dquot_claim_space_nodirty);
 /*
  * This operation can block, but only after everything is updated
  */
-void __dquot_free_space(struct inode *inode, qsize_t number, int reserve)
+void __dquot_free_space(struct inode *inode, qsize_t number, int flags)
 {
 	unsigned int cnt;
 	char warntype[MAXQUOTAS];
+	int reserve = flags & DQUOT_SPACE_RESERVE;
 
 	/* First test before acquiring mutex - solves deadlocks when we
          * re-enter the quota code and are already holding the mutex */
diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h
index e6fa7ac..9edd53c 100644
--- a/include/linux/quotaops.h
+++ b/include/linux/quotaops.h
@@ -9,6 +9,9 @@
 
 #include <linux/fs.h>
 
+#define DQUOT_SPACE_WARN	0x1
+#define DQUOT_SPACE_RESERVE	0x2
+
 static inline struct quota_info *sb_dqopt(struct super_block *sb)
 {
 	return &sb->s_dquot;
@@ -33,9 +36,8 @@ int dquot_scan_active(struct super_block *sb,
 struct dquot *dquot_alloc(struct super_block *sb, int type);
 void dquot_destroy(struct dquot *dquot);
 
-int __dquot_alloc_space(struct inode *inode, qsize_t number,
-		int warn, int reserve);
-void __dquot_free_space(struct inode *inode, qsize_t number, int reserve);
+int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags);
+void __dquot_free_space(struct inode *inode, qsize_t number, int flags);
 
 int dquot_alloc_inode(const struct inode *inode);
 
@@ -231,17 +233,17 @@ static inline int dquot_transfer(struct inode *inode, struct iattr *iattr)
 }
 
 static inline int __dquot_alloc_space(struct inode *inode, qsize_t number,
-		int warn, int reserve)
+		int flags)
 {
-	if (!reserve)
+	if (!(flags & DQUOT_SPACE_RESERVE))
 		inode_add_bytes(inode, number);
 	return 0;
 }
 
 static inline void __dquot_free_space(struct inode *inode, qsize_t number,
-		int reserve)
+		int flags)
 {
-	if (!reserve)
+	if (!(flags & DQUOT_SPACE_RESERVE))
 		inode_sub_bytes(inode, number);
 }
 
@@ -257,7 +259,7 @@ static inline int dquot_claim_space_nodirty(struct inode *inode, qsize_t number)
 
 static inline int dquot_alloc_space_nodirty(struct inode *inode, qsize_t nr)
 {
-	return __dquot_alloc_space(inode, nr, 1, 0);
+	return __dquot_alloc_space(inode, nr, DQUOT_SPACE_WARN);
 }
 
 static inline int dquot_alloc_space(struct inode *inode, qsize_t nr)
@@ -282,7 +284,7 @@ static inline int dquot_alloc_block(struct inode *inode, qsize_t nr)
 
 static inline int dquot_prealloc_block_nodirty(struct inode *inode, qsize_t nr)
 {
-	return __dquot_alloc_space(inode, nr << inode->i_blkbits, 0, 0);
+	return __dquot_alloc_space(inode, nr << inode->i_blkbits, 0);
 }
 
 static inline int dquot_prealloc_block(struct inode *inode, qsize_t nr)
@@ -297,7 +299,8 @@ static inline int dquot_prealloc_block(struct inode *inode, qsize_t nr)
 
 static inline int dquot_reserve_block(struct inode *inode, qsize_t nr)
 {
-	return __dquot_alloc_space(inode, nr << inode->i_blkbits, 1, 1);
+	return __dquot_alloc_space(inode, nr << inode->i_blkbits,
+				DQUOT_SPACE_WARN|DQUOT_SPACE_RESERVE);
 }
 
 static inline int dquot_claim_block(struct inode *inode, qsize_t nr)
@@ -334,7 +337,7 @@ static inline void dquot_free_block(struct inode *inode, qsize_t nr)
 static inline void dquot_release_reservation_block(struct inode *inode,
 		qsize_t nr)
 {
-	__dquot_free_space(inode, nr << inode->i_blkbits, 1);
+	__dquot_free_space(inode, nr << inode->i_blkbits, DQUOT_SPACE_RESERVE);
 }
 
 #endif /* _LINUX_QUOTAOPS_ */
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 15/28] quota: add the option to not fail with EDQUOT in block

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit 0e05842bc117ea70ceb979cca798fd026879951b upstream (as of 2.6.34-git13)

To simplify metadata tracking for delalloc writes, ext4
will simply claim metadata blocks at allocation time, without
first speculatively reserving the worst case and then freeing
what was not used.

To do this, we need a mechanism to track allocations in
the quota subsystem, but potentially allow that allocation
to actually go over quota.

This patch adds a DQUOT_SPACE_NOFAIL flag and function
variants for this purpose.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/quota/dquot.c         |    3 ++-
 include/linux/quotaops.h |   12 ++++++++++++
 2 files changed, 14 insertions(+), 1 deletions(-)

diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 8c40c07..df6832e 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -1494,6 +1494,7 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags)
 	char warntype[MAXQUOTAS];
 	int warn = flags & DQUOT_SPACE_WARN;
 	int reserve = flags & DQUOT_SPACE_RESERVE;
+	int nofail = flags & DQUOT_SPACE_NOFAIL;
 
 	/*
 	 * First test before acquiring mutex - solves deadlocks when we
@@ -1514,7 +1515,7 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags)
 			continue;
 		ret = check_bdq(inode->i_dquot[cnt], number, !warn,
 				warntype+cnt);
-		if (ret) {
+		if (ret && !nofail) {
 			spin_unlock(&dq_data_lock);
 			goto out_flush_warn;
 		}
diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h
index 9edd53c..f8dbeb0 100644
--- a/include/linux/quotaops.h
+++ b/include/linux/quotaops.h
@@ -11,6 +11,7 @@
 
 #define DQUOT_SPACE_WARN	0x1
 #define DQUOT_SPACE_RESERVE	0x2
+#define DQUOT_SPACE_NOFAIL	0x4
 
 static inline struct quota_info *sb_dqopt(struct super_block *sb)
 {
@@ -262,6 +263,12 @@ static inline int dquot_alloc_space_nodirty(struct inode *inode, qsize_t nr)
 	return __dquot_alloc_space(inode, nr, DQUOT_SPACE_WARN);
 }
 
+static inline void dquot_alloc_space_nofail(struct inode *inode, qsize_t nr)
+{
+	__dquot_alloc_space(inode, nr, DQUOT_SPACE_WARN|DQUOT_SPACE_NOFAIL);
+	mark_inode_dirty(inode);
+}
+
 static inline int dquot_alloc_space(struct inode *inode, qsize_t nr)
 {
 	int ret;
@@ -277,6 +284,11 @@ static inline int dquot_alloc_block_nodirty(struct inode *inode, qsize_t nr)
 	return dquot_alloc_space_nodirty(inode, nr << inode->i_blkbits);
 }
 
+static inline void dquot_alloc_block_nofail(struct inode *inode, qsize_t nr)
+{
+	dquot_alloc_space_nofail(inode, nr << inode->i_blkbits);
+}
+
 static inline int dquot_alloc_block(struct inode *inode, qsize_t nr)
 {
 	return dquot_alloc_space(inode, nr << inode->i_blkbits);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 16/28] ext4: don't use quota reservation for speculative metadata

[Theodore Ts'o]

From: Eric Sandeen <sandeen@redhat.com>

commit 72b8ab9dde211ea518ff27e631b2046ef90c29a2 upstream (as of 2.6.34-git13)

Because we can badly over-reserve metadata when we
calculate worst-case, it complicates things for quota, since
we must reserve and then claim later, retry on EDQUOT, etc.
Quota is also a generally smaller pool than fs free blocks,
so this over-reservation hurts more, and more often.

I'm of the opinion that it's not the worst thing to allow
metadata to push a user slightly over quota.  This simplifies
the code and avoids the false quota rejections that result
from worst-case speculation.

This patch stops the speculative quota-charging for
worst-case metadata requirements, and just charges quota
when the blocks are allocated at writeout.  It also is
able to remove the try-again loop on EDQUOT.

This patch has been tested indirectly by running the xfstests
suite with a hack to mount & enable quota prior to the test.

I also did a more specific test of fragmenting freespace
and then doing a large delalloc write under quota; quota
stopped me at the right amount of file IO, and then the
writeout generated enough metadata (due to the fragmentation)
that it put me slightly over quota, as expected.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/balloc.c |    5 ++-
 fs/ext4/inode.c  |   69 ++++++++++++++++++------------------------------------
 2 files changed, 26 insertions(+), 48 deletions(-)

diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index d2f37a5..95b7594 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -591,14 +591,15 @@ ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode,
 	ret = ext4_mb_new_blocks(handle, &ar, errp);
 	if (count)
 		*count = ar.len;
-
 	/*
-	 * Account for the allocated meta blocks
+	 * Account for the allocated meta blocks.  We will never
+	 * fail EDQUOT for metdata, but we do account for it.
 	 */
 	if (!(*errp) && EXT4_I(inode)->i_delalloc_reserved_flag) {
 		spin_lock(&EXT4_I(inode)->i_block_reservation_lock);
 		EXT4_I(inode)->i_allocated_meta_blocks += ar.len;
 		spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
+		dquot_alloc_block_nofail(inode, ar.len);
 	}
 	return ret;
 }
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 89a31e8..df43217 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1076,7 +1076,6 @@ void ext4_da_update_reserve_space(struct inode *inode,
 {
 	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 	struct ext4_inode_info *ei = EXT4_I(inode);
-	int mdb_free = 0, allocated_meta_blocks = 0;
 
 	spin_lock(&ei->i_block_reservation_lock);
 	trace_ext4_da_update_reserve_space(inode, used);
@@ -1091,11 +1090,10 @@ void ext4_da_update_reserve_space(struct inode *inode,
 
 	/* Update per-inode reservations */
 	ei->i_reserved_data_blocks -= used;
-	used += ei->i_allocated_meta_blocks;
 	ei->i_reserved_meta_blocks -= ei->i_allocated_meta_blocks;
-	allocated_meta_blocks = ei->i_allocated_meta_blocks;
+	percpu_counter_sub(&sbi->s_dirtyblocks_counter,
+			   used + ei->i_allocated_meta_blocks);
 	ei->i_allocated_meta_blocks = 0;
-	percpu_counter_sub(&sbi->s_dirtyblocks_counter, used);
 
 	if (ei->i_reserved_data_blocks == 0) {
 		/*
@@ -1103,31 +1101,23 @@ void ext4_da_update_reserve_space(struct inode *inode,
 		 * only when we have written all of the delayed
 		 * allocation blocks.
 		 */
-		mdb_free = ei->i_reserved_meta_blocks;
+		percpu_counter_sub(&sbi->s_dirtyblocks_counter,
+				   ei->i_reserved_meta_blocks);
 		ei->i_reserved_meta_blocks = 0;
 		ei->i_da_metadata_calc_len = 0;
-		percpu_counter_sub(&sbi->s_dirtyblocks_counter, mdb_free);
 	}
 	spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
 
-	/* Update quota subsystem */
-	if (quota_claim) {
+	/* Update quota subsystem for data blocks */
+	if (quota_claim)
 		dquot_claim_block(inode, used);
-		if (mdb_free)
-			dquot_release_reservation_block(inode, mdb_free);
-	} else {
+	else {
 		/*
 		 * We did fallocate with an offset that is already delayed
 		 * allocated. So on delayed allocated writeback we should
-		 * not update the quota for allocated blocks. But then
-		 * converting an fallocate region to initialized region would
-		 * have caused a metadata allocation. So claim quota for
-		 * that
+		 * not re-claim the quota for fallocated blocks.
 		 */
-		if (allocated_meta_blocks)
-			dquot_claim_block(inode, allocated_meta_blocks);
-		dquot_release_reservation_block(inode, mdb_free + used -
-						allocated_meta_blocks);
+		dquot_release_reservation_block(inode, used);
 	}
 
 	/*
@@ -1861,7 +1851,7 @@ static int ext4_da_reserve_space(struct inode *inode, sector_t lblock)
 	int retries = 0;
 	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 	struct ext4_inode_info *ei = EXT4_I(inode);
-	unsigned long md_needed, md_reserved;
+	unsigned long md_needed;
 	int ret;
 
 	/*
@@ -1871,22 +1861,24 @@ static int ext4_da_reserve_space(struct inode *inode, sector_t lblock)
 	 */
 repeat:
 	spin_lock(&ei->i_block_reservation_lock);
-	md_reserved = ei->i_reserved_meta_blocks;
 	md_needed = ext4_calc_metadata_amount(inode, lblock);
 	trace_ext4_da_reserve_space(inode, md_needed);
 	spin_unlock(&ei->i_block_reservation_lock);
 
 	/*
-	 * Make quota reservation here to prevent quota overflow
-	 * later. Real quota accounting is done at pages writeout
-	 * time.
+	 * We will charge metadata quota at writeout time; this saves
+	 * us from metadata over-estimation, though we may go over by
+	 * a small amount in the end.  Here we just reserve for data.
 	 */
-	ret = dquot_reserve_block(inode, md_needed + 1);
+	ret = dquot_reserve_block(inode, 1);
 	if (ret)
 		return ret;
-
+	/*
+	 * We do still charge estimated metadata to the sb though;
+	 * we cannot afford to run out of free blocks.
+	 */
 	if (ext4_claim_free_blocks(sbi, md_needed + 1)) {
-		dquot_release_reservation_block(inode, md_needed + 1);
+		dquot_release_reservation_block(inode, 1);
 		if (ext4_should_retry_alloc(inode->i_sb, &retries)) {
 			yield();
 			goto repeat;
@@ -1933,12 +1925,13 @@ static void ext4_da_release_space(struct inode *inode, int to_free)
 		 * only when we have written all of the delayed
 		 * allocation blocks.
 		 */
-		to_free += ei->i_reserved_meta_blocks;
+		percpu_counter_sub(&sbi->s_dirtyblocks_counter,
+				   ei->i_reserved_meta_blocks);
 		ei->i_reserved_meta_blocks = 0;
 		ei->i_da_metadata_calc_len = 0;
 	}
 
-	/* update fs dirty blocks counter */
+	/* update fs dirty data blocks counter */
 	percpu_counter_sub(&sbi->s_dirtyblocks_counter, to_free);
 
 	spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
@@ -3086,7 +3079,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 			       loff_t pos, unsigned len, unsigned flags,
 			       struct page **pagep, void **fsdata)
 {
-	int ret, retries = 0, quota_retries = 0;
+	int ret, retries = 0;
 	struct page *page;
 	pgoff_t index;
 	unsigned from, to;
@@ -3145,22 +3138,6 @@ retry:
 
 	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
 		goto retry;

[PATCH v2.6.34.y 17/28] ext4: Remove extraneous newlines in ext4_msg() calls

[Theodore Ts'o]

From: Curt Wohlgemuth <curtw@google.com>

commit fbe845ddf368f77f86aa7500f8fd2690f54c66a8 upstream (as of v2.6.34-git13)

Addresses-Google-Bug: #2562325

Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/inode.c |    6 +++---
 fs/ext4/super.c |    2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index df43217..6aa0442 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2276,7 +2276,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
 		ext4_msg(mpd->inode->i_sb, KERN_CRIT,
 			 "delayed block allocation failed for inode %lu at "
 			 "logical offset %llu with max blocks %zd with "
-			 "error %d\n", mpd->inode->i_ino,
+			 "error %d", mpd->inode->i_ino,
 			 (unsigned long long) next,
 			 mpd->b_size >> mpd->inode->i_blkbits, err);
 		printk(KERN_CRIT "This should not happen!!  "
@@ -2944,7 +2944,7 @@ retry:
 		if (IS_ERR(handle)) {
 			ret = PTR_ERR(handle);
 			ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: "
-			       "%ld pages, ino %lu; err %d\n", __func__,
+			       "%ld pages, ino %lu; err %d", __func__,
 				wbc->nr_to_write, inode->i_ino, ret);
 			goto out_writepages;
 		}
@@ -3019,7 +3019,7 @@ retry:
 	if (pages_skipped != wbc->pages_skipped)
 		ext4_msg(inode->i_sb, KERN_CRIT,
 			 "This should not happen leaving %s "
-			 "with nr_to_write = %ld ret = %d\n",
+			 "with nr_to_write = %ld ret = %d",
 			 __func__, wbc->nr_to_write, ret);
 
 	/* Update index */
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index d83cdcb..f9e57e1 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2998,7 +2998,7 @@ no_journal:
 	err = ext4_setup_system_zone(sb);
 	if (err) {
 		ext4_msg(sb, KERN_ERR, "failed to initialize system "
-			 "zone (%d)\n", err);
+			 "zone (%d)", err);
 		goto failed_mount4;
 	}
 
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 18/28] ext4: Prevent creation of files larger than RLIMIT_FSIZE using fallocate

[Theodore Ts'o]

From: Nikanth Karthikesan <knikanth@suse.de>

commit 6d19c42b7cf81c39632b6d4dbc514e8449bcd346 upstream (as of v2.6.34-git13)

Currently using posix_fallocate one can bypass an RLIMIT_FSIZE limit
and create a file larger than the limit. Add a check for that.

Signed-off-by: Nikanth Karthikesan <knikanth@suse.de>
Signed-off-by: Amit Arora <aarora@in.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index ee611da..8a8f9f0 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -3708,6 +3708,11 @@ long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
 	 */
 	credits = ext4_chunk_trans_blocks(inode, max_blocks);
 	mutex_lock(&inode->i_mutex);
+	ret = inode_newsize_ok(inode, (len + offset));
+	if (ret) {
+		mutex_unlock(&inode->i_mutex);
+		return ret;
+	}
 retry:
 	while (ret >= 0 && ret < max_blocks) {
 		block = block + ret;
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 19/28] ext4: check for a good block group before loading buddy pages

[Theodore Ts'o]

From: Curt Wohlgemuth <curtw@google.com>

commit 8a57d9d61a6e361c7bb159dda797672c1df1a691 upstream (as of v2.6.34-git13)

This adds a new field in ext4_group_info to cache the largest available
block range in a block group; and don't load the buddy pages until *after*
we've done a sanity check on the block group.

With large allocation requests (e.g., fallocate(), 8MiB) and relatively full
partitions, it's easy to have no block groups with a block extent large
enough to satisfy the input request length.  This currently causes the loop
during cr == 0 in ext4_mb_regular_allocator() to load the buddy bitmap pages
for EVERY block group.  That can be a lot of pages.  The patch below allows
us to call ext4_mb_good_group() BEFORE we load the buddy pages (although we
have check again after we lock the block group).

Addresses-Google-Bug: #2578108
Addresses-Google-Bug: #2704453

Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ext4.h    |    1 +
 fs/ext4/mballoc.c |   70 +++++++++++++++++++++++++++++++++++++++++++----------
 2 files changed, 58 insertions(+), 13 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index bf938cf..d266003 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1678,6 +1678,7 @@ struct ext4_group_info {
 	ext4_grpblk_t	bb_first_free;	/* first free block */
 	ext4_grpblk_t	bb_free;	/* total free blocks */
 	ext4_grpblk_t	bb_fragments;	/* nr of freespace fragments */
+	ext4_grpblk_t	bb_largest_free_order;/* order of largest frag in BG */
 	struct          list_head bb_prealloc_list;
 #ifdef DOUBLE_CHECK
 	void            *bb_bitmap;
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 9c6a40d..086135d 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -658,6 +658,27 @@ static void ext4_mb_mark_free_simple(struct super_block *sb,
 	}
 }
 
+/*
+ * Cache the order of the largest free extent we have available in this block
+ * group.
+ */
+static void
+mb_set_largest_free_order(struct super_block *sb, struct ext4_group_info *grp)
+{
+	int i;
+	int bits;
+
+	grp->bb_largest_free_order = -1; /* uninit */
+
+	bits = sb->s_blocksize_bits + 1;
+	for (i = bits; i >= 0; i--) {
+		if (grp->bb_counters[i] > 0) {
+			grp->bb_largest_free_order = i;
+			break;
+		}
+	}
+}
+
 static noinline_for_stack
 void ext4_mb_generate_buddy(struct super_block *sb,
 				void *buddy, void *bitmap, ext4_group_t group)
@@ -700,6 +721,7 @@ void ext4_mb_generate_buddy(struct super_block *sb,
 		 */
 		grp->bb_free = free;
 	}
+	mb_set_largest_free_order(sb, grp);
 
 	clear_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state));
 
@@ -725,6 +747,9 @@ void ext4_mb_generate_buddy(struct super_block *sb,
  * contain blocks_per_page (PAGE_CACHE_SIZE / blocksize)  blocks.
  * So it can have information regarding groups_per_page which
  * is blocks_per_page/2
+ *
+ * Locking note:  This routine takes the block group lock of all groups
+ * for this page; do not hold this lock when calling this routine!
  */
 
 static int ext4_mb_init_cache(struct page *page, char *incore)
@@ -910,6 +935,11 @@ out:
 	return err;
 }
 
+/*
+ * Locking note:  This routine calls ext4_mb_init_cache(), which takes the
+ * block group lock of all groups for this page; do not hold the BG lock when
+ * calling this routine!
+ */
 static noinline_for_stack
 int ext4_mb_init_group(struct super_block *sb, ext4_group_t group)
 {
@@ -1004,6 +1034,11 @@ err:
 	return ret;
 }
 
+/*
+ * Locking note:  This routine calls ext4_mb_init_cache(), which takes the
+ * block group lock of all groups for this page; do not hold the BG lock when
+ * calling this routine!
+ */
 static noinline_for_stack int
 ext4_mb_load_buddy(struct super_block *sb, ext4_group_t group,
 					struct ext4_buddy *e4b)
@@ -1299,6 +1334,7 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b,
 			buddy = buddy2;
 		} while (1);
 	}
+	mb_set_largest_free_order(sb, e4b->bd_info);
 	mb_check_buddy(e4b);
 }
 
@@ -1427,6 +1463,7 @@ static int mb_mark_used(struct ext4_buddy *e4b, struct ext4_free_extent *ex)
 		e4b->bd_info->bb_counters[ord]++;
 		e4b->bd_info->bb_counters[ord]++;
 	}
+	mb_set_largest_free_order(e4b->bd_sb, e4b->bd_info);
 
 	mb_set_bits(EXT4_MB_BITMAP(e4b), ex->fe_start, len0);
 	mb_check_buddy(e4b);
@@ -1821,16 +1858,22 @@ void ext4_mb_scan_aligned(struct ext4_allocation_context *ac,
 	}
 }
 
+/* This is now called BEFORE we load the buddy bitmap. */
 static int ext4_mb_good_group(struct ext4_allocation_context *ac,
 				ext4_group_t group, int cr)
 {
 	unsigned free, fragments;
-	unsigned i, bits;
 	int flex_size = ext4_flex_bg_size(EXT4_SB(ac->ac_sb));
 	struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group);
 
 	BUG_ON(cr < 0 || cr >= 4);
-	BUG_ON(EXT4_MB_GRP_NEED_INIT(grp));
+
+	/* We only do this if the grp has never been initialized */
+	if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) {
+		int ret = ext4_mb_init_group(ac->ac_sb, group);
+		if (ret)
+			return 0;
+	}
 
 	free = grp->bb_free;
 	fragments = grp->bb_fragments;
@@ -1843,17 +1886,16 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac,
 	case 0:
 		BUG_ON(ac->ac_2order == 0);
 
+		if (grp->bb_largest_free_order < ac->ac_2order)
+			return 0;
+
 		/* Avoid using the first bg of a flexgroup for data files */
 		if ((ac->ac_flags & EXT4_MB_HINT_DATA) &&
 		    (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) &&
 		    ((group % flex_size) == 0))
 			return 0;
 
-		bits = ac->ac_sb->s_blocksize_bits + 1;
-		for (i = ac->ac_2order; i <= bits; i++)
-			if (grp->bb_counters[i] > 0)
-				return 1;
-		break;
+		return 1;
 	case 1:
 		if ((free / fragments) >= ac->ac_g_ex.fe_len)
 			return 1;
@@ -2024,14 +2066,11 @@ repeat:
 		group = ac->ac_g_ex.fe_group;
 
 		for (i = 0; i < ngroups; group++, i++) {
-			struct ext4_group_info *grp;
-
 			if (group == ngroups)
 				group = 0;
 
-			/* quick check to skip empty groups */
-			grp = ext4_get_group_info(sb, group);
-			if (grp->bb_free == 0)
+			/* This now checks without needing the buddy page */
+			if (!ext4_mb_good_group(ac, group, cr))
 				continue;
 
 			err = ext4_mb_load_buddy(sb, group, &e4b);
@@ -2039,8 +2078,12 @@ repeat:
 				goto out;
 
 			ext4_lock_group(sb, group);
+
+			/*
+			 * We need to check again after locking the
+			 * block group
+			 */
 			if (!ext4_mb_good_group(ac, group, cr)) {
-				/* someone did allocation from this group */
 				ext4_unlock_group(sb, group);
 				ext4_mb_unload_buddy(&e4b);
 				continue;
@@ -2253,6 +2296,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 	INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list);
 	init_rwsem(&meta_group_info[i]->alloc_sem);
 	meta_group_info[i]->bb_free_root = RB_ROOT;
+	meta_group_info[i]->bb_largest_free_order = -1;  /* uninit */
 
 #ifdef DOUBLE_CHECK
 	{
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 20/28] ext4: Show journal_checksum option

[Theodore Ts'o]

From: Jan Kara <jack@suse.cz>

commit 39a4bade8c1826b658316d66ee81c09b0a4d7d42 upstream (as of v2.6.34-git13)

We failed to show journal_checksum option in /proc/mounts. Fix it.

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/super.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index f9e57e1..aa947e0 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -942,6 +942,8 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	seq_puts(seq, test_opt(sb, BARRIER) ? "1" : "0");
 	if (test_opt(sb, JOURNAL_ASYNC_COMMIT))
 		seq_puts(seq, ",journal_async_commit");
+	else if (test_opt(sb, JOURNAL_CHECKSUM))
+		seq_puts(seq, ",journal_checksum");
 	if (test_opt(sb, NOBH))
 		seq_puts(seq, ",nobh");
 	if (test_opt(sb, I_VERSION))
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 21/28] ext4: Use our own write_cache_pages()

[Theodore Ts'o]

Make a copy of write_cache_pages() for the benefit of
ext4_da_writepages().  This allows us to simplify the code some, and
will allow us to further customize the code in future patches.

There are some nasty hacks in write_cache_pages(), which Linus has
(correctly) characterized as vile.  I've just copied it into
write_cache_pages_da(), without trying to clean those bits up lest I
break something in the ext4's delalloc implementation, which is a bit
fragile right now.  This will allow Dave Chinner to clean up
write_cache_pages() in mm/page-writeback.c, without worrying about
breaking ext4.  Eventually write_cache_pages_da() will go away when I
rewrite ext4's delayed allocation and create a general
ext4_writepages() which is used for all of ext4's writeback.  Until
now this is the lowest risk way to clean up the core
write_cache_pages() function.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: Dave Chinner <david@fromorbit.com>
---
 fs/ext4/inode.c |  141 ++++++++++++++++++++++++++++++++++++++++++++++---------
 1 files changed, 119 insertions(+), 22 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 6aa0442..830336d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2426,17 +2426,6 @@ static int __mpage_da_writepage(struct page *page,
 	struct buffer_head *bh, *head;
 	sector_t logical;
 
-	if (mpd->io_done) {
-		/*
-		 * Rest of the page in the page_vec
-		 * redirty then and skip then. We will
-		 * try to write them again after
-		 * starting a new transaction
-		 */
-		redirty_page_for_writepage(wbc, page);
-		unlock_page(page);
-		return MPAGE_DA_EXTENT_TAIL;
-	}
 	/*
 	 * Can we merge this page to current extent?
 	 */
@@ -2831,6 +2820,124 @@ static int ext4_da_writepages_trans_blocks(struct inode *inode)
 	return ext4_chunk_trans_blocks(inode, max_blocks);
 }
 
+/*
+ * write_cache_pages_da - walk the list of dirty pages of the given
+ * address space and call the callback function (which usually writes
+ * the pages).
+ *
+ * This is a forked version of write_cache_pages().  Differences:
+ *	Range cyclic is ignored.
+ *	no_nrwrite_index_update is always presumed true
+ */
+static int write_cache_pages_da(struct address_space *mapping,
+				struct writeback_control *wbc,
+				struct mpage_da_data *mpd)
+{
+	int ret = 0;
+	int done = 0;
+	struct pagevec pvec;
+	int nr_pages;
+	pgoff_t index;
+	pgoff_t end;		/* Inclusive */
+	long nr_to_write = wbc->nr_to_write;
+
+	pagevec_init(&pvec, 0);
+	index = wbc->range_start >> PAGE_CACHE_SHIFT;
+	end = wbc->range_end >> PAGE_CACHE_SHIFT;
+
+	while (!done && (index <= end)) {
+		int i;
+
+		nr_pages = pagevec_lookup_tag(&pvec, mapping, &index,
+			      PAGECACHE_TAG_DIRTY,
+			      min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1);
+		if (nr_pages == 0)
+			break;
+
+		for (i = 0; i < nr_pages; i++) {
+			struct page *page = pvec.pages[i];
+
+			/*
+			 * At this point, the page may be truncated or
+			 * invalidated (changing page->mapping to NULL), or
+			 * even swizzled back from swapper_space to tmpfs file
+			 * mapping. However, page->index will not change
+			 * because we have a reference on the page.
+			 */
+			if (page->index > end) {
+				done = 1;
+				break;
+			}
+
+			lock_page(page);
+
+			/*
+			 * Page truncated or invalidated. We can freely skip it
+			 * then, even for data integrity operations: the page
+			 * has disappeared concurrently, so there could be no
+			 * real expectation of this data interity operation
+			 * even if there is now a new, dirty page at the same
+			 * pagecache address.
+			 */
+			if (unlikely(page->mapping != mapping)) {
+continue_unlock:
+				unlock_page(page);
+				continue;
+			}
+
+			if (!PageDirty(page)) {
+				/* someone wrote it for us */
+				goto continue_unlock;
+			}
+
+			if (PageWriteback(page)) {
+				if (wbc->sync_mode != WB_SYNC_NONE)
+					wait_on_page_writeback(page);
+				else
+					goto continue_unlock;
+			}
+
+			BUG_ON(PageWriteback(page));
+			if (!clear_page_dirty_for_io(page))
+				goto continue_unlock;
+
+			ret = __mpage_da_writepage(page, wbc, mpd);
+			if (unlikely(ret)) {
+				if (ret == AOP_WRITEPAGE_ACTIVATE) {
+					unlock_page(page);
+					ret = 0;
+				} else {
+					done = 1;
+					break;
+				}
+			}
+
+			if (nr_to_write > 0) {
+				nr_to_write--;
+				if (nr_to_write == 0 &&
+				    wbc->sync_mode == WB_SYNC_NONE) {
+					/*
+					 * We stop writing back only if we are
+					 * not doing integrity sync. In case of
+					 * integrity sync we have to keep going
+					 * because someone may be concurrently
+					 * dirtying pages, and we might have
+					 * synced a lot of newly appeared dirty
+					 * pages, but have not synced all of the
+					 * old dirty pages.
+					 */
+					done = 1;
+					break;
+				}
+			}
+		}
+		pagevec_release(&pvec);
+		cond_resched();
+	}
+	return ret;
+}
+
+
 static int ext4_da_writepages(struct address_space *mapping,
 			      struct writeback_control *wbc)
 {
@@ -2839,7 +2946,6 @@ static int ext4_da_writepages(struct address_space *mapping,
 	handle_t *handle = NULL;
 	struct mpage_da_data mpd;
 	struct inode *inode = mapping->host;
-	int no_nrwrite_index_update;
 	int pages_written = 0;
 	long pages_skipped;
 	unsigned int max_pages;
@@ -2919,12 +3025,6 @@ static int ext4_da_writepages(struct address_space *mapping,
 	mpd.wbc = wbc;
 	mpd.inode = mapping->host;
 
-	/*
-	 * we don't want write_cache_pages to update
-	 * nr_to_write and writeback_index
-	 */
-	no_nrwrite_index_update = wbc->no_nrwrite_index_update;
-	wbc->no_nrwrite_index_update = 1;
 	pages_skipped = wbc->pages_skipped;
 
 retry:
@@ -2966,8 +3066,7 @@ retry:
 		mpd.io_done = 0;
 		mpd.pages_written = 0;
 		mpd.retval = 0;
-		ret = write_cache_pages(mapping, wbc, __mpage_da_writepage,
-					&mpd);
+		ret = write_cache_pages_da(mapping, wbc, &mpd);
 		/*
 		 * If we have a contiguous extent of pages and we
 		 * haven't done the I/O yet, map the blocks and submit
@@ -3033,8 +3132,6 @@ retry:
 		mapping->writeback_index = index;
 
 out_writepages:
-	if (!no_nrwrite_index_update)
-		wbc->no_nrwrite_index_update = 0;
 	wbc->nr_to_write -= nr_to_writebump;
 	wbc->range_start = range_start;
 	trace_ext4_da_writepages_result(inode, wbc, ret, pages_written);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 22/28] ext4: Use bitops to read/modify i_flags in struct ext4_inode_info

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 12e9b892002d9af057655d35b44db8ee9243b0dc upstream (as of v2.6.34-git13)

At several places we modify EXT4_I(inode)->i_flags without holding
i_mutex (ext4_do_update_inode, ...). These modifications are racy and
we can lose updates to i_flags. So convert handling of i_flags to use
bitops which are atomic.

https://bugzilla.kernel.org/show_bug.cgi?id=15792

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/dir.c         |    4 +-
 fs/ext4/ext4.h        |  109 ++++++++++++++++++++++++++++++++++++++++++-------
 fs/ext4/ext4_jbd2.h   |    8 ++--
 fs/ext4/extents.c     |   10 ++--
 fs/ext4/file.c        |    2 +-
 fs/ext4/ialloc.c      |    4 +-
 fs/ext4/inode.c       |   30 +++++++-------
 fs/ext4/mballoc.c     |    4 +-
 fs/ext4/migrate.c     |    2 +-
 fs/ext4/move_extent.c |    4 +-
 fs/ext4/namei.c       |   10 ++--
 fs/ext4/super.c       |    1 +
 fs/ext4/xattr.c       |    4 +-
 13 files changed, 136 insertions(+), 56 deletions(-)

diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index 86cb6d8..e7ce97b 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -111,7 +111,7 @@ static int ext4_readdir(struct file *filp,
 
 	if (EXT4_HAS_COMPAT_FEATURE(inode->i_sb,
 				    EXT4_FEATURE_COMPAT_DIR_INDEX) &&
-	    ((EXT4_I(inode)->i_flags & EXT4_INDEX_FL) ||
+	    ((ext4_test_inode_flag(inode, EXT4_INODE_INDEX)) ||
 	     ((inode->i_size >> sb->s_blocksize_bits) == 1))) {
 		err = ext4_dx_readdir(filp, dirent, filldir);
 		if (err != ERR_BAD_DX_DIR) {
@@ -122,7 +122,7 @@ static int ext4_readdir(struct file *filp,
 		 * We don't set the inode dirty flag since it's not
 		 * critical that it get flushed back to the disk.
 		 */
-		EXT4_I(filp->f_path.dentry->d_inode)->i_flags &= ~EXT4_INDEX_FL;
+		ext4_clear_inode_flag(filp->f_path.dentry->d_inode, EXT4_INODE_INDEX);
 	}
 	stored = 0;
 	offset = filp->f_pos & (sb->s_blocksize - 1);
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index d266003..7fe80c5 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -321,6 +321,83 @@ static inline __u32 ext4_mask_flags(umode_t mode, __u32 flags)
 		return flags & EXT4_OTHER_FLMASK;
 }
 
+/*
+ * Inode flags used for atomic set/get
+ */
+enum {
+	EXT4_INODE_SECRM	= 0,	/* Secure deletion */
+	EXT4_INODE_UNRM		= 1,	/* Undelete */
+	EXT4_INODE_COMPR	= 2,	/* Compress file */
+	EXT4_INODE_SYNC		= 3,	/* Synchronous updates */
+	EXT4_INODE_IMMUTABLE	= 4,	/* Immutable file */
+	EXT4_INODE_APPEND	= 5,	/* writes to file may only append */
+	EXT4_INODE_NODUMP	= 6,	/* do not dump file */
+	EXT4_INODE_NOATIME	= 7,	/* do not update atime */
+/* Reserved for compression usage... */
+	EXT4_INODE_DIRTY	= 8,
+	EXT4_INODE_COMPRBLK	= 9,	/* One or more compressed clusters */
+	EXT4_INODE_NOCOMPR	= 10,	/* Don't compress */
+	EXT4_INODE_ECOMPR	= 11,	/* Compression error */
+/* End compression flags --- maybe not all used */
+	EXT4_INODE_INDEX	= 12,	/* hash-indexed directory */
+	EXT4_INODE_IMAGIC	= 13,	/* AFS directory */
+	EXT4_INODE_JOURNAL_DATA	= 14,	/* file data should be journaled */
+	EXT4_INODE_NOTAIL	= 15,	/* file tail should not be merged */
+	EXT4_INODE_DIRSYNC	= 16,	/* dirsync behaviour (directories only) */
+	EXT4_INODE_TOPDIR	= 17,	/* Top of directory hierarchies*/
+	EXT4_INODE_HUGE_FILE	= 18,	/* Set to each huge file */
+	EXT4_INODE_EXTENTS	= 19,	/* Inode uses extents */
+	EXT4_INODE_EA_INODE	= 21,	/* Inode used for large EA */
+	EXT4_INODE_EOFBLOCKS	= 22,	/* Blocks allocated beyond EOF */
+	EXT4_INODE_RESERVED	= 31,	/* reserved for ext4 lib */
+};
+
+#define TEST_FLAG_VALUE(FLAG) (EXT4_##FLAG##_FL == (1 << EXT4_INODE_##FLAG))
+#define CHECK_FLAG_VALUE(FLAG) if (!TEST_FLAG_VALUE(FLAG)) { \
+	printk(KERN_EMERG "EXT4 flag fail: " #FLAG ": %d %d\n", \
+		EXT4_##FLAG##_FL, EXT4_INODE_##FLAG); BUG_ON(1); }
+
+/*
+ * Since it's pretty easy to mix up bit numbers and hex values, and we
+ * can't do a compile-time test for ENUM values, we use a run-time
+ * test to make sure that EXT4_XXX_FL is consistent with respect to
+ * EXT4_INODE_XXX.  If all is well the printk and BUG_ON will all drop
+ * out so it won't cost any extra space in the compiled kernel image.
+ * But it's important that these values are the same, since we are
+ * using EXT4_INODE_XXX to test for the flag values, but EXT4_XX_FL
+ * must be consistent with the values of FS_XXX_FL defined in
+ * include/linux/fs.h and the on-disk values found in ext2, ext3, and
+ * ext4 filesystems, and of course the values defined in e2fsprogs.
+ *
+ * It's not paranoia if the Murphy's Law really *is* out to get you.  :-)
+ */
+static inline void ext4_check_flag_values(void)
+{
+	CHECK_FLAG_VALUE(SECRM);
+	CHECK_FLAG_VALUE(UNRM);
+	CHECK_FLAG_VALUE(COMPR);
+	CHECK_FLAG_VALUE(SYNC);
+	CHECK_FLAG_VALUE(IMMUTABLE);
+	CHECK_FLAG_VALUE(APPEND);
+	CHECK_FLAG_VALUE(NODUMP);
+	CHECK_FLAG_VALUE(NOATIME);
+	CHECK_FLAG_VALUE(DIRTY);
+	CHECK_FLAG_VALUE(COMPRBLK);
+	CHECK_FLAG_VALUE(NOCOMPR);
+	CHECK_FLAG_VALUE(ECOMPR);
+	CHECK_FLAG_VALUE(INDEX);
+	CHECK_FLAG_VALUE(IMAGIC);
+	CHECK_FLAG_VALUE(JOURNAL_DATA);
+	CHECK_FLAG_VALUE(NOTAIL);
+	CHECK_FLAG_VALUE(DIRSYNC);
+	CHECK_FLAG_VALUE(TOPDIR);
+	CHECK_FLAG_VALUE(HUGE_FILE);
+	CHECK_FLAG_VALUE(EXTENTS);
+	CHECK_FLAG_VALUE(EA_INODE);
+	CHECK_FLAG_VALUE(EOFBLOCKS);
+	CHECK_FLAG_VALUE(RESERVED);
+}
+
 /* Used to pass group descriptor data when online resize is done */
 struct ext4_new_group_input {
 	__u32 group;		/* Group number for this data */
@@ -616,9 +693,8 @@ struct ext4_ext_cache {
  */
 struct ext4_inode_info {
 	__le32	i_data[15];	/* unconverted */
-	__u32	i_flags;
-	ext4_fsblk_t	i_file_acl;
 	__u32	i_dtime;
+	ext4_fsblk_t	i_file_acl;
 
 	/*
 	 * i_block_group is the number of the block group which contains
@@ -629,6 +705,7 @@ struct ext4_inode_info {
 	 */
 	ext4_group_t	i_block_group;
 	unsigned long	i_state_flags;		/* Dynamic state flags */
+	unsigned long	i_flags;
 
 	ext4_lblk_t		i_dir_start_lookup;
 #ifdef CONFIG_EXT4_FS_XATTR
@@ -1064,20 +1141,22 @@ enum {
 	EXT4_STATE_DIO_UNWRITTEN,	/* need convert on dio done*/
 };
 
-static inline int ext4_test_inode_state(struct inode *inode, int bit)
-{
-	return test_bit(bit, &EXT4_I(inode)->i_state_flags);
+#define EXT4_INODE_BIT_FNS(name, field)					\
+static inline int ext4_test_inode_##name(struct inode *inode, int bit)	\
+{									\
+	return test_bit(bit, &EXT4_I(inode)->i_##field);		\
+}									\
+static inline void ext4_set_inode_##name(struct inode *inode, int bit)	\
+{									\
+	set_bit(bit, &EXT4_I(inode)->i_##field);			\
+}									\
+static inline void ext4_clear_inode_##name(struct inode *inode, int bit) \
+{									\
+	clear_bit(bit, &EXT4_I(inode)->i_##field);			\
 }
 
-static inline void ext4_set_inode_state(struct inode *inode, int bit)
-{
-	set_bit(bit, &EXT4_I(inode)->i_state_flags);
-}
-
-static inline void ext4_clear_inode_state(struct inode *inode, int bit)
-{
-	clear_bit(bit, &EXT4_I(inode)->i_state_flags);
-}
+EXT4_INODE_BIT_FNS(flag, flags)
+EXT4_INODE_BIT_FNS(state, state_flags)
 #else
 /* Assume that user mode programs are passing in an ext4fs superblock, not
  * a kernel struct super_block.  This will allow us to call the feature-test
@@ -1264,7 +1343,7 @@ struct ext4_dir_entry_2 {
 
 #define is_dx(dir) (EXT4_HAS_COMPAT_FEATURE(dir->i_sb, \
 				      EXT4_FEATURE_COMPAT_DIR_INDEX) && \
-		      (EXT4_I(dir)->i_flags & EXT4_INDEX_FL))
+		    ext4_test_inode_flag((dir), EXT4_INODE_INDEX))
 #define EXT4_DIR_LINK_MAX(dir) (!is_dx(dir) && (dir)->i_nlink >= EXT4_LINK_MAX)
 #define EXT4_DIR_LINK_EMPTY(dir) ((dir)->i_nlink == 2 || (dir)->i_nlink == 1)
 
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index b79ad51..dade0c0 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -273,7 +273,7 @@ static inline int ext4_should_journal_data(struct inode *inode)
 		return 1;
 	if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA)
 		return 1;
-	if (EXT4_I(inode)->i_flags & EXT4_JOURNAL_DATA_FL)
+	if (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA))
 		return 1;
 	return 0;
 }
@@ -284,7 +284,7 @@ static inline int ext4_should_order_data(struct inode *inode)
 		return 0;
 	if (!S_ISREG(inode->i_mode))
 		return 0;
-	if (EXT4_I(inode)->i_flags & EXT4_JOURNAL_DATA_FL)
+	if (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA))
 		return 0;
 	if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
 		return 1;
@@ -297,7 +297,7 @@ static inline int ext4_should_writeback_data(struct inode *inode)
 		return 0;
 	if (EXT4_JOURNAL(inode) == NULL)
 		return 1;
-	if (EXT4_I(inode)->i_flags & EXT4_JOURNAL_DATA_FL)
+	if (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA))
 		return 0;
 	if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)
 		return 1;
@@ -321,7 +321,7 @@ static inline int ext4_should_dioread_nolock(struct inode *inode)
 		return 0;
 	if (!S_ISREG(inode->i_mode))
 		return 0;
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 		return 0;
 	if (ext4_should_journal_data(inode))
 		return 0;
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 8a8f9f0..a1d3f7c 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -3509,7 +3509,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 			set_buffer_uninit(bh_result);
 	}
 
-	if (unlikely(EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL)) {
+	if (unlikely(ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))) {
 		if (unlikely(!eh->eh_entries)) {
 			EXT4_ERROR_INODE(inode,
 					 "eh->eh_entries == 0 ee_block %d",
@@ -3520,7 +3520,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 		last_ex = EXT_LAST_EXTENT(eh);
 		if (iblock + ar.len > le32_to_cpu(last_ex->ee_block)
 		    + ext4_ext_get_actual_len(last_ex))
-			EXT4_I(inode)->i_flags &= ~EXT4_EOFBLOCKS_FL;
+			ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
 	}
 	err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
 	if (err) {
@@ -3661,7 +3661,7 @@ static void ext4_falloc_update_inode(struct inode *inode,
 		 * can proceed even if the new size is the same as i_size.
 		 */
 		if (new_size > i_size_read(inode))
-			EXT4_I(inode)->i_flags |= EXT4_EOFBLOCKS_FL;
+			ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
 	}
 
 }
@@ -3689,7 +3689,7 @@ long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
 	 * currently supporting (pre)allocate mode for extent-based
 	 * files _only_
 	 */
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 		return -EOPNOTSUPP;
 
 	/* preallocation to directories is currently not supported */
@@ -3939,7 +3939,7 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	int error = 0;
 
 	/* fallback to generic here if not in extents fmt */
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 		return generic_block_fiemap(inode, fieinfo, start, len,
 			ext4_get_block);
 
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index d0776e4..5313ae4 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -66,7 +66,7 @@ ext4_file_write(struct kiocb *iocb, const struct iovec *iov,
 	 * is smaller than s_maxbytes, which is for extent-mapped files.
 	 */
 
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)) {
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 		struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 		size_t length = iov_length(iov, nr_segs);
 
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 52618d5..7f6b582 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -492,7 +492,7 @@ static int find_group_orlov(struct super_block *sb, struct inode *parent,
 
 	if (S_ISDIR(mode) &&
 	    ((parent == sb->s_root->d_inode) ||
-	     (EXT4_I(parent)->i_flags & EXT4_TOPDIR_FL))) {
+	     (ext4_test_inode_flag(parent, EXT4_INODE_TOPDIR)))) {
 		int best_ndir = inodes_per_group;
 		int ret = -1;
 
@@ -1038,7 +1038,7 @@ got:
 	if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_EXTENTS)) {
 		/* set extent flag only for directory, file and normal symlink*/
 		if (S_ISDIR(mode) || S_ISREG(mode) || S_ISLNK(mode)) {
-			EXT4_I(inode)->i_flags |= EXT4_EXTENTS_FL;
+			ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS);
 			ext4_ext_tree_init(handle, inode);
 		}
 	}
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 830336d..97dec7c 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -933,7 +933,7 @@ static int ext4_ind_get_blocks(handle_t *handle, struct inode *inode,
 	int count = 0;
 	ext4_fsblk_t first_block = 0;
 
-	J_ASSERT(!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL));
+	J_ASSERT(!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)));
 	J_ASSERT(handle != NULL || (flags & EXT4_GET_BLOCKS_CREATE) == 0);
 	depth = ext4_block_to_path(inode, iblock, offsets,
 				   &blocks_to_boundary);
@@ -1061,7 +1061,7 @@ static int ext4_indirect_calc_metadata_amount(struct inode *inode,
  */
 static int ext4_calc_metadata_amount(struct inode *inode, sector_t lblock)
 {
-	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
 		return ext4_ext_calc_metadata_amount(inode, lblock);
 
 	return ext4_indirect_calc_metadata_amount(inode, lblock);
@@ -1241,7 +1241,7 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block,
 	 * file system block.
 	 */
 	down_read((&EXT4_I(inode)->i_data_sem));
-	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) {
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		retval =  ext4_ext_get_blocks(handle, inode, block, max_blocks,
 				bh, 0);
 	} else {
@@ -1303,7 +1303,7 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block,
 	 * We need to check for EXT4 here because migrate
 	 * could have changed the inode type in between
 	 */
-	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) {
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		retval =  ext4_ext_get_blocks(handle, inode, block, max_blocks,
 					      bh, flags);
 	} else {
@@ -2353,7 +2353,7 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd,
 		goto flush_it;
 
 	/* check if thereserved journal credits might overflow */
-	if (!(EXT4_I(mpd->inode)->i_flags & EXT4_EXTENTS_FL)) {
+	if (!(ext4_test_inode_flag(mpd->inode, EXT4_INODE_EXTENTS))) {
 		if (nrblocks >= EXT4_MAX_TRANS_DATA) {
 			/*
 			 * With non-extent format we are limited by the journal
@@ -2813,7 +2813,7 @@ static int ext4_da_writepages_trans_blocks(struct inode *inode)
 	 * number of contiguous block. So we will limit
 	 * number of contiguous block to a sane value
 	 */
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) &&
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) &&
 	    (max_blocks > EXT4_MAX_TRANS_DATA))
 		max_blocks = EXT4_MAX_TRANS_DATA;
 
@@ -4057,7 +4057,7 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb,
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
 
-	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
 		return ext4_ext_direct_IO(rw, iocb, iov, offset, nr_segs);
 
 	return ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs);
@@ -4696,12 +4696,12 @@ void ext4_truncate(struct inode *inode)
 	if (!ext4_can_truncate(inode))
 		return;
 
-	EXT4_I(inode)->i_flags &= ~EXT4_EOFBLOCKS_FL;
+	ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
 
 	if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC))
 		ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE);
 
-	if (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL) {
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) {
 		ext4_ext_truncate(inode);
 		return;
 	}
@@ -5539,7 +5539,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
-		if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)) {
+		if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 			struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 
 			if (attr->ia_size > sbi->s_bitmap_maxbytes) {
@@ -5552,7 +5552,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	if (S_ISREG(inode->i_mode) &&
 	    attr->ia_valid & ATTR_SIZE &&
 	    (attr->ia_size < inode->i_size ||
-	     (EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL))) {
+	     (ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS)))) {
 		handle_t *handle;
 
 		handle = ext4_journal_start(inode, 3);
@@ -5584,7 +5584,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 			}
 		}
 		/* ext4_truncate will clear the flag */
-		if ((EXT4_I(inode)->i_flags & EXT4_EOFBLOCKS_FL))
+		if ((ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS)))
 			ext4_truncate(inode);
 	}
 
@@ -5660,7 +5660,7 @@ static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks,
 
 static int ext4_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
 {
-	if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 		return ext4_indirect_trans_blocks(inode, nrblocks, chunk);
 	return ext4_ext_index_trans_blocks(inode, nrblocks, chunk);
 }
@@ -5995,9 +5995,9 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val)
 	 */
 
 	if (val)
-		EXT4_I(inode)->i_flags |= EXT4_JOURNAL_DATA_FL;
+		ext4_set_inode_flag(inode, EXT4_INODE_JOURNAL_DATA);
 	else
-		EXT4_I(inode)->i_flags &= ~EXT4_JOURNAL_DATA_FL;
+		ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA);
 	ext4_set_aops(inode);
 
 	jbd2_journal_unlock_updates(journal);
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 086135d..e1f32a5 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2006,7 +2006,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
 	sbi = EXT4_SB(sb);
 	ngroups = ext4_get_groups_count(sb);
 	/* non-extent files are limited to low blocks/groups */
-	if (!(EXT4_I(ac->ac_inode)->i_flags & EXT4_EXTENTS_FL))
+	if (!(ext4_test_inode_flag(ac->ac_inode, EXT4_INODE_EXTENTS)))
 		ngroups = sbi->s_blockfile_groups;
 
 	BUG_ON(ac->ac_status == AC_STATUS_FOUND);
@@ -3171,7 +3171,7 @@ ext4_mb_use_preallocated(struct ext4_allocation_context *ac)
 			continue;
 
 		/* non-extent files can't have physical blocks past 2^32 */
-		if (!(EXT4_I(ac->ac_inode)->i_flags & EXT4_EXTENTS_FL) &&
+		if (!(ext4_test_inode_flag(ac->ac_inode, EXT4_INODE_EXTENTS)) &&
 			pa->pa_pstart + pa->pa_len > EXT4_MAX_BLOCK_FILE_PHYS)
 			continue;
 
diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c
index 34dcfc5..6f3a27e 100644
--- a/fs/ext4/migrate.c
+++ b/fs/ext4/migrate.c
@@ -475,7 +475,7 @@ int ext4_ext_migrate(struct inode *inode)
 	 */
 	if (!EXT4_HAS_INCOMPAT_FEATURE(inode->i_sb,
 				       EXT4_FEATURE_INCOMPAT_EXTENTS) ||
-	    (EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+	    (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 		return -EINVAL;
 
 	if (S_ISLNK(inode->i_mode) && inode->i_blocks == 0)
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index d1fc662..e81fe7f 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -976,11 +976,11 @@ mext_check_arguments(struct inode *orig_inode,
 	}
 
 	/* Ext4 move extent supports only extent based file */
-	if (!(EXT4_I(orig_inode)->i_flags & EXT4_EXTENTS_FL)) {
+	if (!(ext4_test_inode_flag(orig_inode, EXT4_INODE_EXTENTS))) {
 		ext4_debug("ext4 move extent: orig file is not extents "
 			"based file [ino:orig %lu]\n", orig_inode->i_ino);
 		return -EOPNOTSUPP;
-	} else if (!(EXT4_I(donor_inode)->i_flags & EXT4_EXTENTS_FL)) {
+	} else if (!(ext4_test_inode_flag(donor_inode, EXT4_INODE_EXTENTS))) {
 		ext4_debug("ext4 move extent: donor file is not extents "
 			"based file [ino:donor %lu]\n", donor_inode->i_ino);
 		return -EOPNOTSUPP;
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 0c070fa..efab592 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -656,7 +656,7 @@ int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash,
 	dxtrace(printk(KERN_DEBUG "In htree_fill_tree, start hash: %x:%x\n", 
 		       start_hash, start_minor_hash));
 	dir = dir_file->f_path.dentry->d_inode;
-	if (!(EXT4_I(dir)->i_flags & EXT4_INDEX_FL)) {
+	if (!(ext4_test_inode_flag(dir, EXT4_INODE_INDEX))) {
 		hinfo.hash_version = EXT4_SB(dir->i_sb)->s_def_hash_version;
 		if (hinfo.hash_version <= DX_HASH_TEA)
 			hinfo.hash_version +=
@@ -801,7 +801,7 @@ static void ext4_update_dx_flag(struct inode *inode)
 {
 	if (!EXT4_HAS_COMPAT_FEATURE(inode->i_sb,
 				     EXT4_FEATURE_COMPAT_DIR_INDEX))
-		EXT4_I(inode)->i_flags &= ~EXT4_INDEX_FL;
+		ext4_clear_inode_flag(inode, EXT4_INODE_INDEX);
 }
 
 /*
@@ -1418,7 +1418,7 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry,
 		brelse(bh);
 		return retval;
 	}
-	EXT4_I(dir)->i_flags |= EXT4_INDEX_FL;
+	ext4_set_inode_flag(dir, EXT4_INODE_INDEX);
 	data1 = bh2->b_data;
 
 	memcpy (data1, de, len);
@@ -1491,7 +1491,7 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry,
 		retval = ext4_dx_add_entry(handle, dentry, inode);
 		if (!retval || (retval != ERR_BAD_DX_DIR))
 			return retval;
-		EXT4_I(dir)->i_flags &= ~EXT4_INDEX_FL;
+		ext4_clear_inode_flag(dir, EXT4_INODE_INDEX);
 		dx_fallback++;
 		ext4_mark_inode_dirty(handle, dir);
 	}
@@ -2297,7 +2297,7 @@ retry:
 		}
 	} else {
 		/* clear the extent format for fast symlink */
-		EXT4_I(inode)->i_flags &= ~EXT4_EXTENTS_FL;
+		ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS);
 		inode->i_op = &ext4_fast_symlink_inode_operations;
 		memcpy((char *)&EXT4_I(inode)->i_data, symname, l);
 		inode->i_size = l-1;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index aa947e0..2aa9314 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4139,6 +4139,7 @@ static int __init init_ext4_fs(void)
 {
 	int err;
 
+	ext4_check_flag_values();
 	err = init_ext4_system_zone();
 	if (err)
 		return err;
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index b4c5aa8..58f8bb6 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -820,7 +820,7 @@ inserted:
 						EXT4_I(inode)->i_block_group);
 
 			/* non-extent files can't have physical blocks past 2^32 */
-			if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 				goal = goal & EXT4_MAX_BLOCK_FILE_PHYS;
 
 			block = ext4_new_meta_blocks(handle, inode,
@@ -828,7 +828,7 @@ inserted:
 			if (error)
 				goto cleanup;
 
-			if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL))
+			if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
 				BUG_ON(block > EXT4_MAX_BLOCK_FILE_PHYS);
 
 			ea_idebug(inode, "creating block %d", block);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 23/28] ext4: Avoid crashing on NULL ptr dereference on a filesystem error

[Theodore Ts'o]

commit f70f362b4a6fe47c239dbfb3efc0cc2c10e4f09c upstream (as of v2.6.34-git13)

If the EOFBLOCK_FL flag is set when it should not be and the inode is
zero length, then eh_entries is zero, and ex is NULL, so dereferencing
ex to print ex->ee_block causes a kernel OOPS in
ext4_ext_map_blocks().

On top of that, the error message which is printed isn't very helpful.
So we fix this by printing something more explanatory which doesn't
involve trying to print ex->ee_block.

Addresses-Google-Bug: #2655740

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c |    9 +++++----
 1 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index a1d3f7c..8e8a7a4 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -3381,8 +3381,9 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 	 */
 	if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
 		EXT4_ERROR_INODE(inode, "bad extent address "
-				 "iblock: %d, depth: %d pblock %lld",
-				 iblock, depth, path[depth].p_block);
+				 "iblock: %lu, depth: %d pblock %lld",
+				 (unsigned long) iblock, depth,
+				 path[depth].p_block);
 		err = -EIO;
 		goto out2;
 	}
@@ -3512,8 +3513,8 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 	if (unlikely(ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))) {
 		if (unlikely(!eh->eh_entries)) {
 			EXT4_ERROR_INODE(inode,
-					 "eh->eh_entries == 0 ee_block %d",
-					 ex->ee_block);
+					 "eh->eh_entries == 0 and "
+					 "EOFBLOCKS_FL set");
 			err = -EIO;
 			goto out2;
 		}
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 24/28] ext4: Clear the EXT4_EOFBLOCKS_FL flag only when warranted

[Theodore Ts'o]

commit 786ec7915e530936b9eb2e3d12274145cab7aa7d upstream (as of v2.6.34-git13)

Dimitry Monakhov discovered an edge case where it was possible for the
EXT4_EOFBLOCKS_FL flag could get cleared unnecessarily.  This is true;
I have a test case that can be exercised via downloading and
decompressing the file:

wget ftp://ftp.kernel.org/pub/linux/kernel/people/tytso/ext4-testcases/eofblocks-fl-test-case.img.bz2
bunzip2 eofblocks-fl-test-case.img
dd if=/dev/zero of=eofblocks-fl-test-case.img bs=1k seek=17925 bs=1k count=1 conv=notrunc

However, triggering it in real life is highly unlikely since it
requires an extremely fragmented sparse file with a hole in exactly
the right place in the extent tree.  (It actually took quite a bit of
work to generate this test case.)  Still, it's nice to get even
extreme corner cases to be correct, so this patch makes sure that we
don't clear the EXT4_EOFBLOCKS_FL incorrectly even in this corner
case.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/extents.c |   18 +++++++++++++++---
 1 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 8e8a7a4..5e663a7 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -3329,7 +3329,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 	struct ext4_extent_header *eh;
 	struct ext4_extent newex, *ex, *last_ex;
 	ext4_fsblk_t newblock;
-	int err = 0, depth, ret, cache_type;
+	int i, err = 0, depth, ret, cache_type;
 	unsigned int allocated = 0;
 	struct ext4_allocation_request ar;
 	ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio;
@@ -3519,8 +3519,20 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 			goto out2;
 		}
 		last_ex = EXT_LAST_EXTENT(eh);
-		if (iblock + ar.len > le32_to_cpu(last_ex->ee_block)
-		    + ext4_ext_get_actual_len(last_ex))
+		/*
+		 * If the current leaf block was reached by looking at
+		 * the last index block all the way down the tree, and
+		 * we are extending the inode beyond the last extent
+		 * in the current leaf block, then clear the
+		 * EOFBLOCKS_FL flag.
+		 */
+		for (i = depth-1; i >= 0; i--) {
+			if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
+				break;
+		}
+		if ((i < 0) &&
+		    (iblock + ar.len > le32_to_cpu(last_ex->ee_block) +
+		     ext4_ext_get_actual_len(last_ex)))
 			ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
 	}
 	err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 25/28] ext4: restart ext4_ext_remove_space() after transaction restart

[Theodore Ts'o]

From: Dmitry Monakhov <dmonakhov@openvz.org>

commit 0617b83fa239db9743a18ce6cc0e556f4d0fd567 upstream (as of v2.6.34-git13)

If i_data_sem was internally dropped due to transaction restart, it is
necessary to restart path look-up because extents tree was possibly
modified by ext4_get_block().

https://bugzilla.kernel.org/show_bug.cgi?id=15827

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Acked-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/extents.c |   16 +++++++++-------
 1 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 5e663a7..bb03109 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -107,11 +107,8 @@ static int ext4_ext_truncate_extend_restart(handle_t *handle,
 	if (err <= 0)
 		return err;
 	err = ext4_truncate_restart_trans(handle, inode, needed);
-	/*
-	 * We have dropped i_data_sem so someone might have cached again
-	 * an extent we are going to truncate.
-	 */
-	ext4_ext_invalidate_cache(inode);
+	if (err == 0)
+		err = -EAGAIN;
 
 	return err;
 }
@@ -2361,7 +2358,7 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
 	int depth = ext_depth(inode);
 	struct ext4_ext_path *path;
 	handle_t *handle;
-	int i = 0, err = 0;
+	int i, err;
 
 	ext_debug("truncate since %u\n", start);
 
@@ -2370,23 +2367,26 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
 	if (IS_ERR(handle))
 		return PTR_ERR(handle);
 
+again:
 	ext4_ext_invalidate_cache(inode);
 
 	/*
 	 * We start scanning from right side, freeing all the blocks
 	 * after i_size and walking into the tree depth-wise.
 	 */
+	depth = ext_depth(inode);
 	path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
 	if (path == NULL) {
 		ext4_journal_stop(handle);
 		return -ENOMEM;
 	}
+	path[0].p_depth = depth;
 	path[0].p_hdr = ext_inode_hdr(inode);
 	if (ext4_ext_check(inode, path[0].p_hdr, depth)) {
 		err = -EIO;
 		goto out;
 	}
-	path[0].p_depth = depth;
+	i = err = 0;
 
 	while (i >= 0 && err == 0) {
 		if (i == depth) {
@@ -2480,6 +2480,8 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
 out:
 	ext4_ext_drop_refs(path);
 	kfree(path);
+	if (err == -EAGAIN)
+		goto again;
 	ext4_journal_stop(handle);
 
 	return err;
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 26/28] ext4: Conditionally define compat ioctl numbers

[Theodore Ts'o]

From: Ben Hutchings <ben@decadent.org.uk>

commit 899ad0cea6ad7ff4ba24b16318edbc3cbbe03fad upstream (as of v2.6.34-git13)

It is unnecessary, and in general impossible, to define the compat
ioctl numbers except when building the filesystem with CONFIG_COMPAT
defined.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ext4.h |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 7fe80c5..0419018 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -475,6 +475,7 @@ struct ext4_new_group_data {
 #define EXT4_IOC_ALLOC_DA_BLKS		_IO('f', 12)
 #define EXT4_IOC_MOVE_EXT		_IOWR('f', 15, struct move_extent)
 
+#if defined(__KERNEL__) && defined(CONFIG_COMPAT)
 /*
  * ioctl commands in 32 bit emulation
  */
@@ -490,6 +491,7 @@ struct ext4_new_group_data {
 #endif
 #define EXT4_IOC32_GETVERSION_OLD	FS_IOC32_GETVERSION
 #define EXT4_IOC32_SETVERSION_OLD	FS_IOC32_SETVERSION
+#endif
 
 
 /*
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 27/28] ext4: Fix compat EXT4_IOC_ADD_GROUP

[Theodore Ts'o]

From: Ben Hutchings <ben@decadent.org.uk>

commit 4d92dc0f00a775dc2e1267b0e00befb783902fe7 upstream (as of v2.6.34-git13)

struct ext4_new_group_input needs to be converted because u64 has
only 32-bit alignment on some 32-bit architectures, notably i386.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ext4.h  |   16 ++++++++++++++++
 fs/ext4/ioctl.c |   25 +++++++++++++++++++++++--
 2 files changed, 39 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 0419018..2e31640 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -29,6 +29,9 @@
 #include <linux/wait.h>
 #include <linux/blockgroup_lock.h>
 #include <linux/percpu_counter.h>
+#ifdef __KERNEL__
+#include <linux/compat.h>
+#endif
 
 /*
  * The fourth extended filesystem constants/structures
@@ -409,6 +412,18 @@ struct ext4_new_group_input {
 	__u16 unused;
 };
 
+#if defined(__KERNEL__) && defined(CONFIG_COMPAT)
+struct compat_ext4_new_group_input {
+	u32 group;
+	compat_u64 block_bitmap;
+	compat_u64 inode_bitmap;
+	compat_u64 inode_table;
+	u32 blocks_count;
+	u16 reserved_blocks;
+	u16 unused;
+};
+#endif
+
 /* The struct ext4_new_group_input in kernel space, with free_blocks_count */
 struct ext4_new_group_data {
 	__u32 group;
@@ -486,6 +501,7 @@ struct ext4_new_group_data {
 #define EXT4_IOC32_GETRSVSZ		_IOR('f', 5, int)
 #define EXT4_IOC32_SETRSVSZ		_IOW('f', 6, int)
 #define EXT4_IOC32_GROUP_EXTEND		_IOW('f', 7, unsigned int)
+#define EXT4_IOC32_GROUP_ADD		_IOW('f', 8, struct compat_ext4_new_group_input)
 #ifdef CONFIG_JBD2_DEBUG
 #define EXT4_IOC32_WAIT_FOR_READONLY	_IOR('f', 99, int)
 #endif
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 66fa0b0..6ddec84 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -373,8 +373,29 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case EXT4_IOC32_SETRSVSZ:
 		cmd = EXT4_IOC_SETRSVSZ;
 		break;
-	case EXT4_IOC_GROUP_ADD:
-		break;
+	case EXT4_IOC32_GROUP_ADD: {
+		struct compat_ext4_new_group_input __user *uinput;
+		struct ext4_new_group_input input;
+		mm_segment_t old_fs;
+		int err;
+
+		uinput = compat_ptr(arg);
+		err = get_user(input.group, &uinput->group);
+		err |= get_user(input.block_bitmap, &uinput->block_bitmap);
+		err |= get_user(input.inode_bitmap, &uinput->inode_bitmap);
+		err |= get_user(input.inode_table, &uinput->inode_table);
+		err |= get_user(input.blocks_count, &uinput->blocks_count);
+		err |= get_user(input.reserved_blocks,
+				&uinput->reserved_blocks);
+		if (err)
+			return -EFAULT;
+		old_fs = get_fs();
+		set_fs(KERNEL_DS);
+		err = ext4_ioctl(file, EXT4_IOC_GROUP_ADD,
+				 (unsigned long) &input);
+		set_fs(old_fs);
+		return err;
+	}
 	case EXT4_IOC_MOVE_EXT:
 		break;
 	default:
-- 
1.6.6.1.1.g974db.dirty

[PATCH v2.6.34.y 28/28] ext4: Make fsync sync new parent directories in no-journal mode

[Theodore Ts'o]

From: Frank Mayhar <fmayhar@google.com>

commit 14ece1028b3ed53ffec1b1213ffc6acaf79ad77c upstream (as of v2.6.34-git13)

Add a new ext4 state to tell us when a file has been newly created; use
that state in ext4_sync_file in no-journal mode to tell us when we need
to sync the parent directory as well as the inode and data itself.  This
fixes a problem in which a panic or power failure may lose the entire
file even when using fsync, since the parent directory entry is lost.

Addresses-Google-Bug: #2480057

Signed-off-by: Frank Mayhar <fmayhar@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ext4.h  |    1 +
 fs/ext4/fsync.c |   31 +++++++++++++++++++++++++++++--
 fs/ext4/namei.c |    2 ++
 3 files changed, 32 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 2e31640..650ef37 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1157,6 +1157,7 @@ enum {
 	EXT4_STATE_DA_ALLOC_CLOSE,	/* Alloc DA blks on close */
 	EXT4_STATE_EXT_MIGRATE,		/* Inode is migrating */
 	EXT4_STATE_DIO_UNWRITTEN,	/* need convert on dio done*/
+	EXT4_STATE_NEWENTRY,		/* File just added to dir */
 };
 
 #define EXT4_INODE_BIT_FNS(name, field)					\
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index 42bd94a..61da601 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -35,6 +35,29 @@
 #include <trace/events/ext4.h>
 
 /*
+ * If we're not journaling and this is a just-created file, we have to
+ * sync our parent directory (if it was freshly created) since
+ * otherwise it will only be written by writeback, leaving a huge
+ * window during which a crash may lose the file.  This may apply for
+ * the parent directory's parent as well, and so on recursively, if
+ * they are also freshly created.
+ */
+static void ext4_sync_parent(struct inode *inode)
+{
+	struct dentry *dentry = NULL;
+
+	while (inode && ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) {
+		ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY);
+		dentry = list_entry(inode->i_dentry.next,
+				    struct dentry, d_alias);
+		if (!dentry || !dentry->d_parent || !dentry->d_parent->d_inode)
+			break;
+		inode = dentry->d_parent->d_inode;
+		sync_mapping_buffers(inode->i_mapping);
+	}
+}
+
+/*
  * akpm: A new design for ext4_sync_file().
  *
  * This is only called from sys_fsync(), sys_fdatasync() and sys_msync().
@@ -67,8 +90,12 @@ int ext4_sync_file(struct file *file, struct dentry *dentry, int datasync)
 	if (ret < 0)
 		return ret;
 	
-	if (!journal)
-		return simple_fsync(file, dentry, datasync);
+	if (!journal) {
+		ret = simple_fsync(file, dentry, datasync);
+ 		if (!ret && !list_empty(&inode->i_dentry))
+ 			ext4_sync_parent(inode);
+		return ret;
+	}
 
 	/*
 	 * data=writeback,ordered:
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index efab592..2f31631 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -1519,6 +1519,8 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry,
 	de->rec_len = ext4_rec_len_to_disk(blocksize, blocksize);
 	retval = add_dirent_to_buf(handle, dentry, inode, de, bh);
 	brelse(bh);
+	if (retval == 0)
+		ext4_set_inode_state(inode, EXT4_STATE_NEWENTRY);
 	return retval;
 }
 
-- 
1.6.6.1.1.g974db.dirty