From 3612b49598c303cfb22a4b609427f829828e2427 Mon Sep 17 00:00:00 2001
From: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Date: Tue, 25 Jan 2011 02:51:38 +0000
Subject: btrfs: fix return value check of btrfs_join_transaction()

The error check of btrfs_join_transaction()/btrfs_join_transaction_nolock()
is added, and the mistake of the error check in several places is
corrected.

For more stable Btrfs, I think that we should reduce BUG_ON().
But, I think that long time is necessary for this.
So, I propose this patch as a short-term solution.

With this patch:
 - To more stable Btrfs, the part that should be corrected is clarified.
 - The panic isn't done by the NULL pointer reference etc. (even if
   BUG_ON() is increased temporarily)
 - The error code is returned in the place where the error can be easily
   returned.

As a long-term plan:
 - BUG_ON() is reduced by using the forced-readonly framework, etc.

Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index bcf303204f7f..98ee139885cc 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -7478,7 +7478,7 @@ int btrfs_drop_dead_reloc_roots(struct btrfs_root *root)
 		BUG_ON(reloc_root->commit_root != NULL);
 		while (1) {
 			trans = btrfs_join_transaction(root, 1);
-			BUG_ON(!trans);
+			BUG_ON(IS_ERR(trans));
 
 			mutex_lock(&root->fs_info->drop_mutex);
 			ret = btrfs_drop_snapshot(trans, reloc_root);
-- 
cgit v1.2.3


From e9e22899de661af94cb9995885fd04e4c738838b Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 24 Jan 2011 21:43:19 +0000
Subject: Btrfs: do not release more reserved bytes to the global_block_rsv
 than we need

When we do btrfs_block_rsv_release, if global_block_rsv is not full we will
release all the extra bytes to global_block_rsv, even if it's only a little
short of the amount of space that we need to reserve.  This causes us to starve
ourselves of reservable space during the transaction which will force us to
shrink delalloc bytes and commit the transaction more often than we should.  So
instead just add the amount of bytes we need to add to the global reserve so
reserved == size, and then add the rest back into the space_info for general
use.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 98ee139885cc..7af618dcf2c0 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3589,8 +3589,20 @@ void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv,
 
 	if (num_bytes > 0) {
 		if (dest) {
-			block_rsv_add_bytes(dest, num_bytes, 0);
-		} else {
+			spin_lock(&dest->lock);
+			if (!dest->full) {
+				u64 bytes_to_add;
+
+				bytes_to_add = dest->size - dest->reserved;
+				bytes_to_add = min(num_bytes, bytes_to_add);
+				dest->reserved += bytes_to_add;
+				if (dest->reserved >= dest->size)
+					dest->full = 1;
+				num_bytes -= bytes_to_add;
+			}
+			spin_unlock(&dest->lock);
+		}
+		if (num_bytes) {
 			spin_lock(&space_info->lock);
 			space_info->bytes_reserved -= num_bytes;
 			spin_unlock(&space_info->lock);
-- 
cgit v1.2.3


From 68a82277b8619e6d0f2738b1d9b160b627e81e92 Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 24 Jan 2011 21:43:20 +0000
Subject: Btrfs: use the global block reserve if we cannot reserve space

We call use_block_rsv right before we make an allocation in order to make sure
we have enough space.  Now normally people have called btrfs_start_transaction()
with the appropriate amount of space that we need, so we just use some of that
pre-reserved space and move along happily.  The problem is where people use
btrfs_join_transaction(), which doesn't actually reserve any space.  So we try
and reserve space here, but we cannot flush delalloc, so this forces us to
return -ENOSPC when in reality we have plenty of space.  The most common symptom
is seeing a bunch of "couldn't dirty inode" messages in syslog.  With
xfstests 224 we end up falling back to start_transaction and then doing all the
flush delalloc stuff which causes to hang for a very long time.

So instead steal from the global reserve, which is what this is meant for
anyway.  With this patch and the other 2 I have sent xfstests 224 now passes
successfully.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 7af618dcf2c0..ff6bbfd75cf7 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -5646,6 +5646,7 @@ use_block_rsv(struct btrfs_trans_handle *trans,
 	      struct btrfs_root *root, u32 blocksize)
 {
 	struct btrfs_block_rsv *block_rsv;
+	struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
 	int ret;
 
 	block_rsv = get_block_rsv(trans, root);
@@ -5653,14 +5654,39 @@ use_block_rsv(struct btrfs_trans_handle *trans,
 	if (block_rsv->size == 0) {
 		ret = reserve_metadata_bytes(trans, root, block_rsv,
 					     blocksize, 0);
-		if (ret)
+		/*
+		 * If we couldn't reserve metadata bytes try and use some from
+		 * the global reserve.
+		 */
+		if (ret && block_rsv != global_rsv) {
+			ret = block_rsv_use_bytes(global_rsv, blocksize);
+			if (!ret)
+				return global_rsv;
+			return ERR_PTR(ret);
+		} else if (ret) {
 			return ERR_PTR(ret);
+		}
 		return block_rsv;
 	}
 
 	ret = block_rsv_use_bytes(block_rsv, blocksize);
 	if (!ret)
 		return block_rsv;
+	if (ret) {
+		WARN_ON(1);
+		ret = reserve_metadata_bytes(trans, root, block_rsv, blocksize,
+					     0);
+		if (!ret) {
+			spin_lock(&block_rsv->lock);
+			block_rsv->size += blocksize;
+			spin_unlock(&block_rsv->lock);
+			return block_rsv;
+		} else if (ret && block_rsv != global_rsv) {
+			ret = block_rsv_use_bytes(global_rsv, blocksize);
+			if (!ret)
+				return global_rsv;
+		}
+	}
 
 	return ERR_PTR(-ENOSPC);
 }
-- 
cgit v1.2.3


From b1953bcec95c189b1eea690a08e89646d7750bda Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Fri, 21 Jan 2011 21:10:01 +0000
Subject: Btrfs: make shrink_delalloc a little friendlier

Xfstests 224 will just sit there and spin for ever until eventually we give up
flushing delalloc and exit.  On my box this took several hours.  I could not
interrupt this process either, even though we use INTERRUPTIBLE.  So do 2 things

1) Keep us from looping over and over again without reclaiming anything
2) If we get interrupted exit the loop

I tested this and the test now exits in a reasonable amount of time, and can be
interrupted with ctrl+c.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index ff6bbfd75cf7..f96641a93fc9 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3345,8 +3345,10 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 	u64 reserved;
 	u64 max_reclaim;
 	u64 reclaimed = 0;
+	long time_left;
 	int pause = 1;
 	int nr_pages = (2 * 1024 * 1024) >> PAGE_CACHE_SHIFT;
+	int loops = 0;
 
 	block_rsv = &root->fs_info->delalloc_block_rsv;
 	space_info = block_rsv->space_info;
@@ -3359,7 +3361,7 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 
 	max_reclaim = min(reserved, to_reclaim);
 
-	while (1) {
+	while (loops < 1024) {
 		/* have the flusher threads jump in and do some IO */
 		smp_mb();
 		nr_pages = min_t(unsigned long, nr_pages,
@@ -3367,8 +3369,12 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 		writeback_inodes_sb_nr_if_idle(root->fs_info->sb, nr_pages);
 
 		spin_lock(&space_info->lock);
-		if (reserved > space_info->bytes_reserved)
+		if (reserved > space_info->bytes_reserved) {
+			loops = 0;
 			reclaimed += reserved - space_info->bytes_reserved;
+		} else {
+			loops++;
+		}
 		reserved = space_info->bytes_reserved;
 		spin_unlock(&space_info->lock);
 
@@ -3379,7 +3385,12 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans,
 			return -EAGAIN;
 
 		__set_current_state(TASK_INTERRUPTIBLE);
-		schedule_timeout(pause);
+		time_left = schedule_timeout(pause);
+
+		/* We were interrupted, exit */
+		if (time_left)
+			break;
+
 		pause <<= 1;
 		if (pause > HZ / 10)
 			pause = HZ / 10;
-- 
cgit v1.2.3


From 5df67083488ccbad925f583b698ab38f8629a016 Mon Sep 17 00:00:00 2001
From: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Date: Tue, 1 Feb 2011 09:17:35 +0000
Subject: btrfs: checking NULL or not in some functions

Because NULL is returned when the memory allocation fails,
it is checked whether it is NULL.

Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 2 ++
 fs/btrfs/extent_io.c   | 2 ++
 fs/btrfs/tree-log.c    | 6 ++++++
 3 files changed, 10 insertions(+)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index f96641a93fc9..9de4ff03882a 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -6496,6 +6496,8 @@ static noinline int relocate_inode_pages(struct inode *inode, u64 start,
 	int ret = 0;
 
 	ra = kzalloc(sizeof(*ra), GFP_NOFS);
+	if (!ra)
+		return -ENOMEM;
 
 	mutex_lock(&inode->i_mutex);
 	first_index = start >> PAGE_CACHE_SHIFT;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 6411ed6ca449..8862dda46ff6 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -1920,6 +1920,8 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 		nr = bio_get_nr_vecs(bdev);
 
 	bio = btrfs_bio_alloc(bdev, sector, nr, GFP_NOFS | __GFP_HIGH);
+	if (!bio)
+		return -ENOMEM;
 
 	bio_add_page(bio, page, page_size, offset);
 	bio->bi_end_io = end_io_func;
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 42dfc3077040..6d66e5caff97 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2751,7 +2751,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 	log = root->log_root;
 
 	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
 	dst_path = btrfs_alloc_path();
+	if (!dst_path) {
+		btrfs_free_path(path);
+		return -ENOMEM;
+	}
 
 	min_key.objectid = inode->i_ino;
 	min_key.type = BTRFS_INODE_ITEM_KEY;
-- 
cgit v1.2.3


From 98d5dc13e7e74b77ca3b4c3cbded9f48d2dbbbb7 Mon Sep 17 00:00:00 2001
From: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Date: Thu, 20 Jan 2011 06:19:37 +0000
Subject: btrfs: fix return value check of btrfs_start_transaction()

The error check of btrfs_start_transaction() is added, and the mistake
of the error check on several places is corrected.

Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c |  7 +++++--
 fs/btrfs/inode.c       |  1 +
 fs/btrfs/ioctl.c       | 10 ++++++++--
 fs/btrfs/relocation.c  |  3 +++
 fs/btrfs/super.c       |  2 ++
 fs/btrfs/tree-log.c    |  1 +
 fs/btrfs/volumes.c     | 19 +++++++++++++++++--
 7 files changed, 37 insertions(+), 6 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 9de4ff03882a..f07ba21cbf06 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -6271,6 +6271,8 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 	BUG_ON(!wc);
 
 	trans = btrfs_start_transaction(tree_root, 0);
+	BUG_ON(IS_ERR(trans));
+
 	if (block_rsv)
 		trans->block_rsv = block_rsv;
 
@@ -6368,6 +6370,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 
 			btrfs_end_transaction_throttle(trans, tree_root);
 			trans = btrfs_start_transaction(tree_root, 0);
+			BUG_ON(IS_ERR(trans));
 			if (block_rsv)
 				trans->block_rsv = block_rsv;
 		}
@@ -7587,7 +7590,7 @@ int btrfs_cleanup_reloc_trees(struct btrfs_root *root)
 
 	if (found) {
 		trans = btrfs_start_transaction(root, 1);
-		BUG_ON(!trans);
+		BUG_ON(IS_ERR(trans));
 		ret = btrfs_commit_transaction(trans, root);
 		BUG_ON(ret);
 	}
@@ -7831,7 +7834,7 @@ static noinline int relocate_one_extent(struct btrfs_root *extent_root,
 
 
 	trans = btrfs_start_transaction(extent_root, 1);
-	BUG_ON(!trans);
+	BUG_ON(IS_ERR(trans));
 
 	if (extent_key->objectid == 0) {
 		ret = del_extent_zero(trans, extent_root, path, extent_key);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 5621818921f8..36bc3f49ebf9 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -2357,6 +2357,7 @@ void btrfs_orphan_cleanup(struct btrfs_root *root)
 		 */
 		if (is_bad_inode(inode)) {
 			trans = btrfs_start_transaction(root, 0);
+			BUG_ON(IS_ERR(trans));
 			btrfs_orphan_del(trans, inode);
 			btrfs_end_transaction(trans, root);
 			iput(inode);
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 12dabe28cf54..02d224e8c83f 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -907,6 +907,10 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root,
 
 	if (new_size > old_size) {
 		trans = btrfs_start_transaction(root, 0);
+		if (IS_ERR(trans)) {
+			ret = PTR_ERR(trans);
+			goto out_unlock;
+		}
 		ret = btrfs_grow_device(trans, device, new_size);
 		btrfs_commit_transaction(trans, root);
 	} else {
@@ -2141,9 +2145,9 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
 	path->leave_spinning = 1;
 
 	trans = btrfs_start_transaction(root, 1);
-	if (!trans) {
+	if (IS_ERR(trans)) {
 		btrfs_free_path(path);
-		return -ENOMEM;
+		return PTR_ERR(trans);
 	}
 
 	dir_id = btrfs_super_root_dir(&root->fs_info->super_copy);
@@ -2337,6 +2341,8 @@ static noinline long btrfs_ioctl_start_sync(struct file *file, void __user *argp
 	u64 transid;
 
 	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
 	transid = trans->transid;
 	btrfs_commit_transaction_async(trans, root, 0);
 
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index ea9965430241..1f5556acb530 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2028,6 +2028,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc,
 
 	while (1) {
 		trans = btrfs_start_transaction(root, 0);
+		BUG_ON(IS_ERR(trans));
 		trans->block_rsv = rc->block_rsv;
 
 		ret = btrfs_block_rsv_check(trans, root, rc->block_rsv,
@@ -3665,6 +3666,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 
 	while (1) {
 		trans = btrfs_start_transaction(rc->extent_root, 0);
+		BUG_ON(IS_ERR(trans));
 
 		if (update_backref_cache(trans, &rc->backref_cache)) {
 			btrfs_end_transaction(trans, rc->extent_root);
@@ -4033,6 +4035,7 @@ static noinline_for_stack int mark_garbage_root(struct btrfs_root *root)
 	int ret;
 
 	trans = btrfs_start_transaction(root->fs_info->tree_root, 0);
+	BUG_ON(IS_ERR(trans));
 
 	memset(&root->root_item.drop_progress, 0,
 		sizeof(root->root_item.drop_progress));
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index f4e45fdded30..0209b5fc772c 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -623,6 +623,8 @@ int btrfs_sync_fs(struct super_block *sb, int wait)
 	btrfs_wait_ordered_extents(root, 0, 0);
 
 	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
 	ret = btrfs_commit_transaction(trans, root);
 	return ret;
 }
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 6d66e5caff97..a4bbb854dfd2 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3112,6 +3112,7 @@ int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
 	BUG_ON(!path);
 
 	trans = btrfs_start_transaction(fs_info->tree_root, 0);
+	BUG_ON(IS_ERR(trans));
 
 	wc.trans = trans;
 	wc.pin = 1;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index f2d2f4ccc738..7cad59353b09 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1212,6 +1212,10 @@ static int btrfs_rm_dev_item(struct btrfs_root *root,
 		return -ENOMEM;
 
 	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans)) {
+		btrfs_free_path(path);
+		return PTR_ERR(trans);
+	}
 	key.objectid = BTRFS_DEV_ITEMS_OBJECTID;
 	key.type = BTRFS_DEV_ITEM_KEY;
 	key.offset = device->devid;
@@ -1604,6 +1608,12 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	}
 
 	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans)) {
+		kfree(device);
+		ret = PTR_ERR(trans);
+		goto error;
+	}
+
 	lock_chunks(root);
 
 	device->barriers = 1;
@@ -1872,7 +1882,7 @@ static int btrfs_relocate_chunk(struct btrfs_root *root,
 		return ret;
 
 	trans = btrfs_start_transaction(root, 0);
-	BUG_ON(!trans);
+	BUG_ON(IS_ERR(trans));
 
 	lock_chunks(root);
 
@@ -2046,7 +2056,7 @@ int btrfs_balance(struct btrfs_root *dev_root)
 		BUG_ON(ret);
 
 		trans = btrfs_start_transaction(dev_root, 0);
-		BUG_ON(!trans);
+		BUG_ON(IS_ERR(trans));
 
 		ret = btrfs_grow_device(trans, device, old_size);
 		BUG_ON(ret);
@@ -2212,6 +2222,11 @@ again:
 
 	/* Shrinking succeeded, else we would be at "done". */
 	trans = btrfs_start_transaction(root, 0);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto done;
+	}
+
 	lock_chunks(root);
 
 	device->disk_total_bytes = new_size;
-- 
cgit v1.2.3


From 3c14874acc71180553fb5aba528e3cf57c5b958b Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Wed, 2 Feb 2011 15:53:47 +0000
Subject: Btrfs: exclude super blocks when we read in block groups

This has been resulting in a BUT_ON(ret) after btrfs_reserve_extent in
btrfs_cow_file_range.  The reason is we don't actually calculate the bytes_super
for a block group until we go to cache it, which means that the space_info can
hand out reservations for space that it doesn't actually have, and we can run
out of data space.  This is also a problem if you are using space caching since
we don't ever calculate bytes_super for the block groups.  So instead everytime
we read a block group call exclude_super_stripes, which calculates the
bytes_super for the block group so it can be left out of the space_info.  Then
whenever caching completes we just call free_excluded_extents so that the super
excluded extents are freed up.  Also if we are unmounting and we hit any block
groups that haven't been cached we still need to call free_excluded_extents to
make sure things are cleaned up properly.  Thanks,

Reported-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 26 ++++++++++++++++++--------
 1 file changed, 18 insertions(+), 8 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index f07ba21cbf06..565e22d77b1b 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -320,11 +320,6 @@ static int caching_kthread(void *data)
 	if (!path)
 		return -ENOMEM;
 
-	exclude_super_stripes(extent_root, block_group);
-	spin_lock(&block_group->space_info->lock);
-	block_group->space_info->bytes_readonly += block_group->bytes_super;
-	spin_unlock(&block_group->space_info->lock);
-
 	last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET);
 
 	/*
@@ -467,8 +462,10 @@ static int cache_block_group(struct btrfs_block_group_cache *cache,
 			cache->cached = BTRFS_CACHE_NO;
 		}
 		spin_unlock(&cache->lock);
-		if (ret == 1)
+		if (ret == 1) {
+			free_excluded_extents(fs_info->extent_root, cache);
 			return 0;
+		}
 	}
 
 	if (load_cache_only)
@@ -4036,6 +4033,7 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes)
 
 	num_bytes = ALIGN(num_bytes, root->sectorsize);
 	atomic_dec(&BTRFS_I(inode)->outstanding_extents);
+	WARN_ON(atomic_read(&BTRFS_I(inode)->outstanding_extents) < 0);
 
 	spin_lock(&BTRFS_I(inode)->accounting_lock);
 	nr_extents = atomic_read(&BTRFS_I(inode)->outstanding_extents);
@@ -8325,6 +8323,13 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info)
 		if (block_group->cached == BTRFS_CACHE_STARTED)
 			wait_block_group_cache_done(block_group);
 
+		/*
+		 * We haven't cached this block group, which means we could
+		 * possibly have excluded extents on this block group.
+		 */
+		if (block_group->cached == BTRFS_CACHE_NO)
+			free_excluded_extents(info->extent_root, block_group);
+
 		btrfs_remove_free_space_cache(block_group);
 		btrfs_put_block_group(block_group);
 
@@ -8439,6 +8444,13 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		cache->flags = btrfs_block_group_flags(&cache->item);
 		cache->sectorsize = root->sectorsize;
 
+		/*
+		 * We need to exclude the super stripes now so that the space
+		 * info has super bytes accounted for, otherwise we'll think
+		 * we have more space than we actually do.
+		 */
+		exclude_super_stripes(root, cache);
+
 		/*
 		 * check for two cases, either we are full, and therefore
 		 * don't need to bother with the caching work since we won't
@@ -8447,12 +8459,10 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		 * time, particularly in the full case.
 		 */
 		if (found_key.offset == btrfs_block_group_used(&cache->item)) {
-			exclude_super_stripes(root, cache);
 			cache->last_byte_to_unpin = (u64)-1;
 			cache->cached = BTRFS_CACHE_FINISHED;
 			free_excluded_extents(root, cache);
 		} else if (btrfs_block_group_used(&cache->item) == 0) {
-			exclude_super_stripes(root, cache);
 			cache->last_byte_to_unpin = (u64)-1;
 			cache->cached = BTRFS_CACHE_FINISHED;
 			add_new_free_space(cache, root->fs_info,
-- 
cgit v1.2.3


From c26a920373a983b52223eed5a13b97404d8b4158 Mon Sep 17 00:00:00 2001
From: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Date: Mon, 14 Feb 2011 00:45:29 +0000
Subject: Btrfs: check return value of alloc_extent_map()

I add the check on the return value of alloc_extent_map() to several places.
In addition, alloc_extent_map() returns only the address or NULL.
Therefore, check by IS_ERR() is unnecessary. So, I remove IS_ERR() checking.

Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent-tree.c | 2 +-
 fs/btrfs/extent_map.c  | 4 ++--
 fs/btrfs/file.c        | 1 +
 fs/btrfs/inode.c       | 3 +++
 4 files changed, 7 insertions(+), 3 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 565e22d77b1b..a7aaa10c5302 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -6584,7 +6584,7 @@ static noinline int relocate_data_extent(struct inode *reloc_inode,
 	u64 end = start + extent_key->offset - 1;
 
 	em = alloc_extent_map(GFP_NOFS);
-	BUG_ON(!em || IS_ERR(em));
+	BUG_ON(!em);
 
 	em->start = start;
 	em->len = extent_key->offset;
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index b0e1fce12530..2b6c12e983b3 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -51,8 +51,8 @@ struct extent_map *alloc_extent_map(gfp_t mask)
 {
 	struct extent_map *em;
 	em = kmem_cache_alloc(extent_map_cache, mask);
-	if (!em || IS_ERR(em))
-		return em;
+	if (!em)
+		return NULL;
 	em->in_tree = 0;
 	em->flags = 0;
 	em->compress_type = BTRFS_COMPRESS_NONE;
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index b0ff34b96607..65338a1d14ad 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -185,6 +185,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			split = alloc_extent_map(GFP_NOFS);
 		if (!split2)
 			split2 = alloc_extent_map(GFP_NOFS);
+		BUG_ON(!split || !split2);
 
 		write_lock(&em_tree->lock);
 		em = lookup_extent_mapping(em_tree, start, len);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index c9bc0afdbfc6..8d392ed73d57 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -644,6 +644,7 @@ retry:
 					async_extent->ram_size - 1, 0);
 
 		em = alloc_extent_map(GFP_NOFS);
+		BUG_ON(!em);
 		em->start = async_extent->start;
 		em->len = async_extent->ram_size;
 		em->orig_start = em->start;
@@ -820,6 +821,7 @@ static noinline int cow_file_range(struct inode *inode,
 		BUG_ON(ret);
 
 		em = alloc_extent_map(GFP_NOFS);
+		BUG_ON(!em);
 		em->start = start;
 		em->orig_start = em->start;
 		ram_size = ins.offset;
@@ -1169,6 +1171,7 @@ out_check:
 			struct extent_map_tree *em_tree;
 			em_tree = &BTRFS_I(inode)->extent_tree;
 			em = alloc_extent_map(GFP_NOFS);
+			BUG_ON(!em);
 			em->start = cur_offset;
 			em->orig_start = em->start;
 			em->len = num_bytes;
-- 
cgit v1.2.3


From 91435650c233b93e0da389db74f4b2c11c5ad2d4 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Wed, 16 Feb 2011 13:10:41 -0500
Subject: Btrfs: put ENOSPC debugging under a mount option

ENOSPC in btrfs is getting to the point where the extra debugging isn't
required.  I've put it under mount -o enospc_debug just in case someone
is having difficult problems.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/ctree.h       | 1 +
 fs/btrfs/extent-tree.c | 2 +-
 fs/btrfs/super.c       | 7 ++++++-
 3 files changed, 8 insertions(+), 2 deletions(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 72195378bef9..6297701bc19c 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1254,6 +1254,7 @@ struct btrfs_root {
 #define BTRFS_MOUNT_SPACE_CACHE		(1 << 12)
 #define BTRFS_MOUNT_CLEAR_CACHE		(1 << 13)
 #define BTRFS_MOUNT_USER_SUBVOL_RM_ALLOWED (1 << 14)
+#define BTRFS_MOUNT_ENOSPC_DEBUG	 (1 << 15)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index a7aaa10c5302..d375fc04a065 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -5377,7 +5377,7 @@ again:
 			       num_bytes, data, 1);
 		goto again;
 	}
-	if (ret == -ENOSPC) {
+	if (ret == -ENOSPC && btrfs_test_opt(root, ENOSPC_DEBUG)) {
 		struct btrfs_space_info *sinfo;
 
 		sinfo = __find_space_info(root->fs_info, data);
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 0209b5fc772c..db0a827252bd 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -155,7 +155,8 @@ enum {
 	Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl, Opt_compress,
 	Opt_compress_type, Opt_compress_force, Opt_compress_force_type,
 	Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard,
-	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, Opt_err,
+	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed,
+	Opt_enospc_debug, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -184,6 +185,7 @@ static match_table_t tokens = {
 	{Opt_space_cache, "space_cache"},
 	{Opt_clear_cache, "clear_cache"},
 	{Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"},
+	{Opt_enospc_debug, "enospc_debug"},
 	{Opt_err, NULL},
 };
 
@@ -358,6 +360,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_user_subvol_rm_allowed:
 			btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED);
 			break;
+		case Opt_enospc_debug:
+			btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG);
+			break;
 		case Opt_err:
 			printk(KERN_INFO "btrfs: unrecognized mount option "
 			       "'%s'\n", p);
-- 
cgit v1.2.3


From c87f08ca44e83b2c8d28f63f9c33f3a270a04bbe Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Wed, 16 Feb 2011 13:57:04 -0500
Subject: Btrfs: allow balance to explicitly allocate chunks as it relocates

Btrfs device shrinking and balancing ends up reallocating all the blocks
in order to allow COW to move them to new destinations.  It is somewhat
awkward in terms of ENOSPC because most of the enospc code is built
around the idea that some operation on a reference counted tree triggers
allocations in the non-reference counted trees.

This commit changes the balancing code to deal with enospc by trying to
allocate a new chunk.  If that allocation succeeds, we go ahead and
retry whatever failed due to enospc.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/ctree.h       |  2 ++
 fs/btrfs/extent-tree.c |  7 +++++++
 fs/btrfs/relocation.c  | 13 ++++++++++++-
 3 files changed, 21 insertions(+), 1 deletion(-)

(limited to 'fs/btrfs/extent-tree.c')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 6297701bc19c..28188a786da0 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2219,6 +2219,8 @@ int btrfs_error_unpin_extent_range(struct btrfs_root *root,
 				   u64 start, u64 end);
 int btrfs_error_discard_extent(struct btrfs_root *root, u64 bytenr,
 			       u64 num_bytes);
+int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, u64 type);
 
 /* ctree.c */
 int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key,
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index d375fc04a065..100e409e9053 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -8066,6 +8066,13 @@ out:
 	return ret;
 }
 
+int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, u64 type)
+{
+	u64 alloc_flags = get_alloc_profile(root, type);
+	return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1);
+}
+
 /*
  * helper to account the unused space of all the readonly block group in the
  * list. takes mirrors into account.
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 0825e4ed9447..31ade5802ae8 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -3654,6 +3654,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	u32 item_size;
 	int ret;
 	int err = 0;
+	int progress = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -3666,9 +3667,10 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	}
 
 	while (1) {
+		progress++;
 		trans = btrfs_start_transaction(rc->extent_root, 0);
 		BUG_ON(IS_ERR(trans));
-
+restart:
 		if (update_backref_cache(trans, &rc->backref_cache)) {
 			btrfs_end_transaction(trans, rc->extent_root);
 			continue;
@@ -3781,6 +3783,15 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 			}
 		}
 	}
+	if (trans && progress && err == -ENOSPC) {
+		ret = btrfs_force_chunk_alloc(trans, rc->extent_root,
+					      rc->block_group->flags);
+		if (ret == 0) {
+			err = 0;
+			progress = 0;
+			goto restart;
+		}
+	}
 
 	btrfs_release_path(rc->extent_root, path);
 	clear_extent_bits(&rc->processed_blocks, 0, (u64)-1, EXTENT_DIRTY,
-- 
cgit v1.2.3