[patch 00/35] btrfs: Error Handling Patchset

[patch 00/35] btrfs: Error Handling Patchset

[Jeff Mahoney]

 This is the error handling patchset for btrfs.

  The series is comprised of several groups of patches:
  - Adding new btrfs_panic and btrfs_error routines
  - Adding the ability to abort transactions
  - Converting functions that don't have error conditions to
    return void
  - Eliminating the gfp_t argument to lock_ and unlock_extent,
    as they are always called with GFP_NOFS
  - Pushing up errors to handle them and eliminate BUG_ON sites
  - Several other bug fixes for issues we encountered during testing

 I'd like to thank David Sterba for a putting a ton of effort into testing
 and review.

 -Jeff

[patch 01/35] btrfs: Add btrfs_panic()

[Jeff Mahoney]

As part of the effort to eliminate BUG_ON as an error handling
technique, we need to determine which errors are actual logic errors,
which are on-disk corruption, and which are normal runtime errors
e.g. -ENOMEM.

Annotating these error cases is helpful to understand and report them.

This patch adds a btrfs_panic() routine that will either panic
or BUG depending on the new -ofatal_errors={panic,bug} mount option.
Since there are still so many BUG_ONs, it defaults to BUG for now but I
expect that to change once the error handling effort has made
significant progress.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ctree.h |   11 +++++++++++
 fs/btrfs/super.c |   50 +++++++++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 60 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 80b6486..a97a670 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1503,6 +1503,7 @@ struct btrfs_ioctl_defrag_range_args {
 #define BTRFS_MOUNT_SKIP_BALANCE	(1 << 19)
 #define BTRFS_MOUNT_CHECK_INTEGRITY	(1 << 20)
 #define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 21)
+#define BTRFS_MOUNT_PANIC_ON_FATAL_ERROR	(1 << 22)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2970,6 +2971,16 @@ do {								\
 		__btrfs_std_error((fs_info), __func__, __LINE__, (errno));\
 } while (0)
 
+void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
+		   unsigned int line, int errno, const char *fmt, ...);
+
+#define btrfs_panic(fs_info, errno, fmt, args...)			\
+do {									\
+	struct btrfs_fs_info *_i = (fs_info);				\
+	__btrfs_panic(_i, __func__, __LINE__, errno, fmt, ##args);	\
+	BUG_ON(!(_i->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR));	\
+} while (0)
+
 /* acl.c */
 #ifdef CONFIG_BTRFS_FS_POSIX_ACL
 struct posix_acl *btrfs_get_acl(struct inode *inode, int type);
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 3ce97b2..9774e38 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -76,6 +76,9 @@ static const char *btrfs_decode_error(struct btrfs_fs_info *fs_info, int errno,
 	case -EROFS:
 		errstr = "Readonly filesystem";
 		break;
+	case -EEXIST:
+		errstr = "Object already exists";
+		break;
 	default:
 		if (nbuf) {
 			if (snprintf(nbuf, 16, "error %d", -errno) >= 0)
@@ -145,6 +148,36 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 	btrfs_handle_error(fs_info);
 }
 
+/*
+ * __btrfs_panic decodes unexpected, fatal errors from the caller,
+ * issues an alert, and either panics or BUGs, depending on mount options.
+ */
+void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
+		   unsigned int line, int errno, const char *fmt, ...)
+{
+	char nbuf[16];
+	char *s_id = "<unknown>";
+	const char *errstr;
+	struct va_format vaf = { .fmt = fmt };
+	va_list args;
+
+	if (fs_info)
+		s_id = fs_info->sb->s_id;
+
+	va_start(args, fmt);
+	vaf.va = &args;
+
+	errstr = btrfs_decode_error(fs_info, errno, nbuf);
+	if (fs_info->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR)
+		panic(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (%s)\n",
+			s_id, function, line, &vaf, errstr);
+
+	printk(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (%s)\n",
+	       s_id, function, line, &vaf, errstr);
+	va_end(args);
+	/* Caller calls BUG() */
+}
+
 static void btrfs_put_super(struct super_block *sb)
 {
 	(void)close_ctree(btrfs_sb(sb)->tree_root);
@@ -166,7 +199,7 @@ enum {
 	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_inode_cache,
 	Opt_no_space_cache, Opt_recovery, Opt_skip_balance,
 	Opt_check_integrity, Opt_check_integrity_including_extent_data,
-	Opt_check_integrity_print_mask,
+	Opt_check_integrity_print_mask, Opt_fatal_errors,
 	Opt_err,
 };
 
@@ -206,6 +239,7 @@ static match_table_t tokens = {
 	{Opt_check_integrity, "check_int"},
 	{Opt_check_integrity_including_extent_data, "check_int_data"},
 	{Opt_check_integrity_print_mask, "check_int_print_mask=%d"},
+	{Opt_fatal_errors, "fatal_errors=%s"},
 	{Opt_err, NULL},
 };
 
@@ -438,6 +472,18 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			ret = -EINVAL;
 			goto out;
 #endif
+		case Opt_fatal_errors:
+			if (strcmp(args[0].from, "panic") == 0)
+				btrfs_set_opt(info->mount_opt,
+					      PANIC_ON_FATAL_ERROR);
+			else if (strcmp(args[0].from, "bug") == 0)
+				btrfs_clear_opt(info->mount_opt,
+					      PANIC_ON_FATAL_ERROR);
+			else {
+				ret = -EINVAL;
+				goto out;
+			}
+			break;
 		case Opt_err:
 			printk(KERN_INFO "btrfs: unrecognized mount option "
 			       "'%s'\n", p);
@@ -766,6 +812,8 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
 		seq_puts(seq, ",inode_cache");
 	if (btrfs_test_opt(root, SKIP_BALANCE))
 		seq_puts(seq, ",skip_balance");
+	if (btrfs_test_opt(root, PANIC_ON_FATAL_ERROR))
+		seq_puts(seq, ",fatal_errors=panic");
 	return 0;
 }
 
-- 
1.7.9

[patch 02/35] btrfs: Catch locking failures in {set,clear,convert}_extent_bit

[Jeff Mahoney]

The *_state functions can only return 0 or -EEXIST. This patch addresses
the cases where those functions returning -EEXIST represent a locking
failure. It handles them by panicking with an appropriate error message.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/extent_io.c |   58 ++++++++++++++++++++++++++++++++-----------------
 1 files changed, 38 insertions(+), 20 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index a55fbe6..601f23b 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -53,6 +53,12 @@ struct extent_page_data {
 	unsigned int sync_io:1;
 };
 
+static inline struct btrfs_fs_info *
+tree_fs_info(struct extent_io_tree *tree)
+{
+	return btrfs_sb(tree->mapping->host->i_sb);
+}
+
 int __init extent_io_init(void)
 {
 	extent_state_cache = kmem_cache_create("extent_state",
@@ -439,6 +445,13 @@ alloc_extent_state_atomic(struct extent_state *prealloc)
 	return prealloc;
 }
 
+void extent_io_tree_panic(struct extent_io_tree *tree, int err)
+{
+	btrfs_panic(tree_fs_info(tree), err, "Locking error: "
+		    "Extent tree was modified by another "
+		    "thread while locked.");
+}
+
 /*
  * clear some bits on a range in the tree.  This may require splitting
  * or inserting elements in the tree, so the gfp mask is used to
@@ -542,7 +555,9 @@ hit_next:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, start);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
+
 		prealloc = NULL;
 		if (err)
 			goto out;
@@ -564,7 +579,9 @@ hit_next:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, end + 1);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
+
 		if (wake)
 			wake_up(&state->wq);
 
@@ -742,8 +759,10 @@ again:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = insert_state(tree, prealloc, start, end, &bits);
+		if (err)
+			extent_io_tree_panic(tree, err);
+
 		prealloc = NULL;
-		BUG_ON(err == -EEXIST);
 		goto out;
 	}
 	state = rb_entry(node, struct extent_state, rb_node);
@@ -809,7 +828,9 @@ hit_next:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, start);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
+
 		prealloc = NULL;
 		if (err)
 			goto out;
@@ -846,12 +867,9 @@ hit_next:
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
 				   &bits);
-		BUG_ON(err == -EEXIST);
-		if (err) {
-			free_extent_state(prealloc);
-			prealloc = NULL;
-			goto out;
-		}
+		if (err)
+			extent_io_tree_panic(tree, err);
+
 		cache_state(prealloc, cached_state);
 		prealloc = NULL;
 		start = this_end + 1;
@@ -873,7 +891,8 @@ hit_next:
 		prealloc = alloc_extent_state_atomic(prealloc);
 		BUG_ON(!prealloc);
 		err = split_state(tree, state, prealloc, end + 1);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
 
 		set_state_bits(tree, prealloc, &bits);
 		cache_state(prealloc, cached_state);
@@ -946,7 +965,8 @@ again:
 		}
 		err = insert_state(tree, prealloc, start, end, &bits);
 		prealloc = NULL;
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
 		goto out;
 	}
 	state = rb_entry(node, struct extent_state, rb_node);
@@ -1002,7 +1022,8 @@ hit_next:
 			goto out;
 		}
 		err = split_state(tree, state, prealloc, start);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
 		prealloc = NULL;
 		if (err)
 			goto out;
@@ -1041,12 +1062,8 @@ hit_next:
 		 */
 		err = insert_state(tree, prealloc, start, this_end,
 				   &bits);
-		BUG_ON(err == -EEXIST);
-		if (err) {
-			free_extent_state(prealloc);
-			prealloc = NULL;
-			goto out;
-		}
+		if (err)
+			extent_io_tree_panic(tree, err);
 		prealloc = NULL;
 		start = this_end + 1;
 		goto search_again;
@@ -1065,7 +1082,8 @@ hit_next:
 		}
 
 		err = split_state(tree, state, prealloc, end + 1);
-		BUG_ON(err == -EEXIST);
+		if (err)
+			extent_io_tree_panic(tree, err);
 
 		set_state_bits(tree, prealloc, &bits);
 		clear_state_bit(tree, prealloc, &clear_bits, 0);
-- 
1.7.9

[patch 03/35] btrfs: Panic on bad rbtree operations

[Jeff Mahoney]

The ordered data and relocation trees have BUG_ONs to protect against
bad tree operations.

This patch replaces them with a panic that will report the problem.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ordered-data.c |   12 ++++++++++--
 fs/btrfs/relocation.c   |   36 +++++++++++++++++++++++++++++-------
 2 files changed, 39 insertions(+), 9 deletions(-)

diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index a1c9404..2857f28 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -59,6 +59,14 @@ static struct rb_node *tree_insert(struct rb_root *root, u64 file_offset,
 	return NULL;
 }
 
+static void ordered_data_tree_panic(struct inode *inode, int errno,
+					       u64 offset)
+{
+	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
+	btrfs_panic(fs_info, errno, "Inconsistency in ordered tree at offset "
+		    "%llu\n", (unsigned long long)offset);
+}
+
 /*
  * look for a given offset in the tree, and if it can't be found return the
  * first lesser offset
@@ -207,7 +215,8 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
 	spin_lock(&tree->lock);
 	node = tree_insert(&tree->tree, file_offset,
 			   &entry->rb_node);
-	BUG_ON(node);
+	if (node)
+		ordered_data_tree_panic(inode, -EEXIST, file_offset);
 	spin_unlock(&tree->lock);
 
 	spin_lock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
@@ -215,7 +224,6 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
 		      &BTRFS_I(inode)->root->fs_info->ordered_extents);
 	spin_unlock(&BTRFS_I(inode)->root->fs_info->ordered_extent_lock);
 
-	BUG_ON(node);
 	return 0;
 }
 
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 8c1aae2..e5996ff 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -326,6 +326,19 @@ static struct rb_node *tree_search(struct rb_root *root, u64 bytenr)
 	return NULL;
 }
 
+void backref_tree_panic(struct rb_node *rb_node, int errno,
+					  u64 bytenr)
+{
+
+	struct btrfs_fs_info *fs_info = NULL;
+	struct backref_node *bnode = rb_entry(rb_node, struct backref_node,
+					      rb_node);
+	if (bnode->root)
+		fs_info = bnode->root->fs_info;
+	btrfs_panic(fs_info, errno, "Inconsistency in backref cache "
+		    "found at offset %llu\n", (unsigned long long)bytenr);
+}
+
 /*
  * walk up backref nodes until reach node presents tree root
  */
@@ -452,7 +465,8 @@ static void update_backref_node(struct backref_cache *cache,
 	rb_erase(&node->rb_node, &cache->rb_root);
 	node->bytenr = bytenr;
 	rb_node = tree_insert(&cache->rb_root, node->bytenr, &node->rb_node);
-	BUG_ON(rb_node);
+	if (rb_node)
+		backref_tree_panic(rb_node, -EEXIST, bytenr);
 }
 
 /*
@@ -999,7 +1013,8 @@ next:
 	if (!cowonly) {
 		rb_node = tree_insert(&cache->rb_root, node->bytenr,
 				      &node->rb_node);
-		BUG_ON(rb_node);
+		if (rb_node)
+			backref_tree_panic(rb_node, -EEXIST, node->bytenr);
 		list_add_tail(&node->lower, &cache->leaves);
 	}
 
@@ -1034,7 +1049,9 @@ next:
 		if (!cowonly) {
 			rb_node = tree_insert(&cache->rb_root, upper->bytenr,
 					      &upper->rb_node);
-			BUG_ON(rb_node);
+			if (rb_node)
+				backref_tree_panic(rb_node, -EEXIST,
+						   upper->bytenr);
 		}
 
 		list_add_tail(&edge->list[UPPER], &upper->lower);
@@ -1180,7 +1197,8 @@ static int clone_backref_node(struct btrfs_trans_handle *trans,
 
 	rb_node = tree_insert(&cache->rb_root, new_node->bytenr,
 			      &new_node->rb_node);
-	BUG_ON(rb_node);
+	if (rb_node)
+		backref_tree_panic(rb_node, -EEXIST, new_node->bytenr);
 
 	if (!new_node->lowest) {
 		list_for_each_entry(new_edge, &new_node->lower, list[UPPER]) {
@@ -1252,7 +1270,8 @@ static int __update_reloc_root(struct btrfs_root *root, int del)
 		rb_node = tree_insert(&rc->reloc_root_tree.rb_root,
 				      node->bytenr, &node->rb_node);
 		spin_unlock(&rc->reloc_root_tree.lock);
-		BUG_ON(rb_node);
+		if (rb_node)
+			backref_tree_panic(rb_node, -EEXIST, node->bytenr);
 	} else {
 		list_del_init(&root->root_list);
 		kfree(node);
@@ -3154,7 +3173,8 @@ static int add_tree_block(struct reloc_control *rc,
 	block->key_ready = 0;
 
 	rb_node = tree_insert(blocks, block->bytenr, &block->rb_node);
-	BUG_ON(rb_node);
+	if (rb_node)
+		backref_tree_panic(rb_node, -EEXIST, block->bytenr);
 
 	return 0;
 }
@@ -3426,7 +3446,9 @@ static int find_data_references(struct reloc_control *rc,
 			block->key_ready = 1;
 			rb_node = tree_insert(blocks, block->bytenr,
 					      &block->rb_node);
-			BUG_ON(rb_node);
+			if (rb_node)
+				backref_tree_panic(rb_node, -EEXIST,
+						   block->bytenr);
 		}
 		if (counted)
 			added = 1;
-- 
1.7.9

[patch 04/35] btrfs: Fix kfree of member instead of structure

[Jeff Mahoney]

Correctness fix: The kfree calls in the add_delayed_* functions free
the node that's passed into it, but the node is a member of another
structure. It works because it's always the first member of the
containing structure, but it should really be using the containing
structure itself.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/delayed-ref.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c
index 66e4f29..29ecd54 100644
--- a/fs/btrfs/delayed-ref.c
+++ b/fs/btrfs/delayed-ref.c
@@ -487,7 +487,7 @@ static noinline int add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kfree(head_ref);
 	} else {
 		delayed_refs->num_heads++;
 		delayed_refs->num_heads_ready++;
@@ -549,7 +549,7 @@ static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kfree(full_ref);
 	} else {
 		delayed_refs->num_entries++;
 		trans->delayed_ref_updates++;
@@ -611,7 +611,7 @@ static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 		 * we've updated the existing ref, free the newly
 		 * allocated ref
 		 */
-		kfree(ref);
+		kfree(full_ref);
 	} else {
 		delayed_refs->num_entries++;
 		trans->delayed_ref_updates++;
-- 
1.7.9

[patch 05/35] btrfs: Simplify btrfs_insert_root

[Jeff Mahoney]

btrfs_insert_root is just a wrapper for btrfs_insert_item. Just return
the error directly.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/root-tree.c |    9 +++------
 1 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c
index f409990..1fd93d6 100644
--- a/fs/btrfs/root-tree.c
+++ b/fs/btrfs/root-tree.c
@@ -116,13 +116,10 @@ out:
 	return ret;
 }
 
-int btrfs_insert_root(struct btrfs_trans_handle *trans, struct btrfs_root
-		      *root, struct btrfs_key *key, struct btrfs_root_item
-		      *item)
+int btrfs_insert_root(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+		      struct btrfs_key *key, struct btrfs_root_item *item)
 {
-	int ret;
-	ret = btrfs_insert_item(trans, root, key, item, sizeof(*item));
-	return ret;
+	return btrfs_insert_item(trans, root, key, item, sizeof(*item));
 }
 
 /*
-- 
1.7.9

[patch 06/35] btrfs: clean_tree_block should panic on observed memory corruption and return void

[Jeff Mahoney]

 The only error condition in clean_tree_block is an accounting bug.
 Returning without modifying dirty_metadata_bytes and as if the cleaning
 as been performed may cause problems later so it should panic instead.

 It should probably be a BUG_ON but we have btrfs_panic now.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/disk-io.c |   15 ++++++++++-----
 fs/btrfs/disk-io.h |    4 ++--
 2 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 534266f..d52ec11 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1109,8 +1109,8 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
 
 }
 
-int clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root,
-		     struct extent_buffer *buf)
+void clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+		      struct extent_buffer *buf)
 {
 	struct inode *btree_inode = root->fs_info->btree_inode;
 	if (btrfs_header_generation(buf) ==
@@ -1121,8 +1121,14 @@ int clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 			spin_lock(&root->fs_info->delalloc_lock);
 			if (root->fs_info->dirty_metadata_bytes >= buf->len)
 				root->fs_info->dirty_metadata_bytes -= buf->len;
-			else
-				WARN_ON(1);
+			else {
+				spin_unlock(&root->fs_info->delalloc_lock);
+				btrfs_panic(root->fs_info, -EOVERFLOW,
+					  "Can't clear %lu bytes from "
+					  " dirty_mdatadata_bytes (%lu)",
+					  buf->len,
+					  root->fs_info->dirty_metadata_bytes);
+			}
 			spin_unlock(&root->fs_info->delalloc_lock);
 		}
 
@@ -1131,7 +1137,6 @@ int clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		clear_extent_buffer_dirty(&BTRFS_I(btree_inode)->io_tree,
 					  buf);
 	}
-	return 0;
 }
 
 static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index e4bc474..ca34237 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -44,8 +44,8 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
 			 int mirror_num, struct extent_buffer **eb);
 struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
 						   u64 bytenr, u32 blocksize);
-int clean_tree_block(struct btrfs_trans_handle *trans,
-		     struct btrfs_root *root, struct extent_buffer *buf);
+void clean_tree_block(struct btrfs_trans_handle *trans,
+		      struct btrfs_root *root, struct extent_buffer *buf);
 int open_ctree(struct super_block *sb,
 	       struct btrfs_fs_devices *fs_devices,
 	       char *options);
-- 
1.7.9

[patch 07/35] btrfs: avoid NULL deref in btrfs_reserve_extent with DEBUG_ENOSPC

[Jeff Mahoney]

 __find_space_info can return NULL but we don't check it before calling
 dump_space_info().

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/extent-tree.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 37e0a80..c32a9bf 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -5838,7 +5838,8 @@ again:
 			printk(KERN_ERR "btrfs allocation failed flags %llu, "
 			       "wanted %llu\n", (unsigned long long)data,
 			       (unsigned long long)num_bytes);
-			dump_space_info(sinfo, num_bytes, 1);
+			if (sinfo)
+				dump_space_info(sinfo, num_bytes, 1);
 		}
 	}
 
-- 
1.7.9

[patch 08/35] btrfs: Remove set bits return from clear_extent_bit

[Jeff Mahoney]

There is only one caller of clear_extent_bit that checks the return value
and it only checks if it's negative. Since there are no users of the
returned bits functionality of clear_extent_bit, stop returning it
and avoid complicating error handling.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/extent_io.c |   12 +++++-------
 1 files changed, 5 insertions(+), 7 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 601f23b..a4a7a18 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -462,8 +462,7 @@ void extent_io_tree_panic(struct extent_io_tree *tree, int err)
  *
  * the range [start, end] is inclusive.
  *
- * This takes the tree lock, and returns < 0 on error, > 0 if any of the
- * bits were already set, or zero if none of the bits were already set.
+ * This takes the tree lock, and returns 0 on success and < 0 on error.
  */
 int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 		     int bits, int wake, int delete,
@@ -477,7 +476,6 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 	struct rb_node *node;
 	u64 last_end;
 	int err;
-	int set = 0;
 	int clear = 0;
 
 	if (delete)
@@ -562,7 +560,7 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			set |= clear_state_bit(tree, state, &bits, wake);
+			clear_state_bit(tree, state, &bits, wake);
 			if (last_end == (u64)-1)
 				goto out;
 			start = last_end + 1;
@@ -585,13 +583,13 @@ hit_next:
 		if (wake)
 			wake_up(&state->wq);
 
-		set |= clear_state_bit(tree, prealloc, &bits, wake);
+		clear_state_bit(tree, prealloc, &bits, wake);
 
 		prealloc = NULL;
 		goto out;
 	}
 
-	set |= clear_state_bit(tree, state, &bits, wake);
+	clear_state_bit(tree, state, &bits, wake);
 next:
 	if (last_end == (u64)-1)
 		goto out;
@@ -608,7 +606,7 @@ out:
 	if (prealloc)
 		free_extent_state(prealloc);
 
-	return set;
+	return 0;
 
 search_again:
 	if (start > end)
-- 
1.7.9

Re: [patch 08/35] btrfs: Remove set bits return from clear_extent_bit

[Liu Bo]

On 03/22/2012 09:11 AM, Jeff Mahoney wrote:
> There is only one caller of clear_extent_bit that checks the return value
> and it only checks if it's negative. Since there are no users of the
> returned bits functionality of clear_extent_bit, stop returning it
> and avoid complicating error handling.
> 

A similar one from Li:
http://www.spinics.net/lists/linux-btrfs/msg15458.html

thanks,
liubo

> Signed-off-by: Jeff Mahoney <jeffm@suse.com>
> ---
>  fs/btrfs/extent_io.c |   12 +++++-------
>  1 files changed, 5 insertions(+), 7 deletions(-)
> 
> diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
> index 601f23b..a4a7a18 100644
> --- a/fs/btrfs/extent_io.c
> +++ b/fs/btrfs/extent_io.c
> @@ -462,8 +462,7 @@ void extent_io_tree_panic(struct extent_io_tree *tree, int err)
>   *
>   * the range [start, end] is inclusive.
>   *
> - * This takes the tree lock, and returns < 0 on error, > 0 if any of the
> - * bits were already set, or zero if none of the bits were already set.
> + * This takes the tree lock, and returns 0 on success and < 0 on error.
>   */
>  int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
>  		     int bits, int wake, int delete,
> @@ -477,7 +476,6 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
>  	struct rb_node *node;
>  	u64 last_end;
>  	int err;
> -	int set = 0;
>  	int clear = 0;
>  
>  	if (delete)
> @@ -562,7 +560,7 @@ hit_next:
>  		if (err)
>  			goto out;
>  		if (state->end <= end) {
> -			set |= clear_state_bit(tree, state, &bits, wake);
> +			clear_state_bit(tree, state, &bits, wake);
>  			if (last_end == (u64)-1)
>  				goto out;
>  			start = last_end + 1;
> @@ -585,13 +583,13 @@ hit_next:
>  		if (wake)
>  			wake_up(&state->wq);
>  
> -		set |= clear_state_bit(tree, prealloc, &bits, wake);
> +		clear_state_bit(tree, prealloc, &bits, wake);
>  
>  		prealloc = NULL;
>  		goto out;
>  	}
>  
> -	set |= clear_state_bit(tree, state, &bits, wake);
> +	clear_state_bit(tree, state, &bits, wake);
>  next:
>  	if (last_end == (u64)-1)
>  		goto out;
> @@ -608,7 +606,7 @@ out:
>  	if (prealloc)
>  		free_extent_state(prealloc);
>  
> -	return set;
> +	return 0;
>  
>  search_again:
>  	if (start > end)

Re: [patch 08/35] btrfs: Remove set bits return from clear_extent_bit

[Jeff Mahoney]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03/21/2012 10:03 PM, Liu Bo wrote:
> On 03/22/2012 09:11 AM, Jeff Mahoney wrote:
>> There is only one caller of clear_extent_bit that checks the
>> return value and it only checks if it's negative. Since there are
>> no users of the returned bits functionality of clear_extent_bit,
>> stop returning it and avoid complicating error handling.
>> 
> 
> A similar one from Li: 
> http://www.spinics.net/lists/linux-btrfs/msg15458.html

Yep, I saw that one fly by. This one has been part of this patchset
for months.

- -Jeff

>> Signed-off-by: Jeff Mahoney <jeffm@suse.com> --- 
>> fs/btrfs/extent_io.c |   12 +++++------- 1 files changed, 5
>> insertions(+), 7 deletions(-)
>> 
>> diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index
>> 601f23b..a4a7a18 100644 --- a/fs/btrfs/extent_io.c +++
>> b/fs/btrfs/extent_io.c @@ -462,8 +462,7 @@ void
>> extent_io_tree_panic(struct extent_io_tree *tree, int err) * *
>> the range [start, end] is inclusive. * - * This takes the tree
>> lock, and returns < 0 on error, > 0 if any of the - * bits were
>> already set, or zero if none of the bits were already set. + *
>> This takes the tree lock, and returns 0 on success and < 0 on
>> error. */ int clear_extent_bit(struct extent_io_tree *tree, u64
>> start, u64 end, int bits, int wake, int delete, @@ -477,7 +476,6
>> @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start,
>> u64 end, struct rb_node *node; u64 last_end; int err; -	int set =
>> 0; int clear = 0;
>> 
>> if (delete) @@ -562,7 +560,7 @@ hit_next: if (err) goto out; if
>> (state->end <= end) { -			set |= clear_state_bit(tree, state,
>> &bits, wake); +			clear_state_bit(tree, state, &bits, wake); if
>> (last_end == (u64)-1) goto out; start = last_end + 1; @@ -585,13
>> +583,13 @@ hit_next: if (wake) wake_up(&state->wq);
>> 
>> -		set |= clear_state_bit(tree, prealloc, &bits, wake); +
>> clear_state_bit(tree, prealloc, &bits, wake);
>> 
>> prealloc = NULL; goto out; }
>> 
>> -	set |= clear_state_bit(tree, state, &bits, wake); +
>> clear_state_bit(tree, state, &bits, wake); next: if (last_end ==
>> (u64)-1) goto out; @@ -608,7 +606,7 @@ out: if (prealloc) 
>> free_extent_state(prealloc);
>> 
>> -	return set; +	return 0;
>> 
>> search_again: if (start > end)
> 
> 


- -- 
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=iuYq
-----END PGP SIGNATURE-----

[patch 09/35] btrfs: find_and_setup_root error push-up

[Jeff Mahoney]

find_and_setup_root BUGs when it encounters an error from
btrfs_find_last_root, which can occur if a path can't be allocated.

This patch pushes it up to its callers where it is already handled.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/disk-io.c |   11 ++++++-----
 1 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index d52ec11..155c4e3 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1197,10 +1197,10 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	return 0;
 }
 
-static int find_and_setup_root(struct btrfs_root *tree_root,
-			       struct btrfs_fs_info *fs_info,
-			       u64 objectid,
-			       struct btrfs_root *root)
+static int __must_check find_and_setup_root(struct btrfs_root *tree_root,
+					    struct btrfs_fs_info *fs_info,
+					    u64 objectid,
+					    struct btrfs_root *root)
 {
 	int ret;
 	u32 blocksize;
@@ -1213,7 +1213,8 @@ static int find_and_setup_root(struct btrfs_root *tree_root,
 				   &root->root_item, &root->root_key);
 	if (ret > 0)
 		return -ENOENT;
-	BUG_ON(ret);
+	else if (ret < 0)
+		return ret;
 
 	generation = btrfs_root_generation(&root->root_item);
 	blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
-- 
1.7.9

[patch 10/35] btrfs: btrfs_update_root error push-up

[Jeff Mahoney]

btrfs_update_root BUG's when it can't alloc a path, yet it can recover
from a search error. This patch returns -ENOMEM instead.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ctree.h     |    7 ++++---
 fs/btrfs/root-tree.c |    4 +++-
 2 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index a97a670..339e637 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2724,9 +2724,10 @@ int btrfs_del_root(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 int btrfs_insert_root(struct btrfs_trans_handle *trans, struct btrfs_root
 		      *root, struct btrfs_key *key, struct btrfs_root_item
 		      *item);
-int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
-		      *root, struct btrfs_key *key, struct btrfs_root_item
-		      *item);
+int __must_check btrfs_update_root(struct btrfs_trans_handle *trans,
+				   struct btrfs_root *root,
+				   struct btrfs_key *key,
+				   struct btrfs_root_item *item);
 int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, struct
 			 btrfs_root_item *item, struct btrfs_key *key);
 int btrfs_find_dead_roots(struct btrfs_root *root, u64 objectid);
diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c
index 1fd93d6..1486cf9 100644
--- a/fs/btrfs/root-tree.c
+++ b/fs/btrfs/root-tree.c
@@ -93,7 +93,9 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root
 	unsigned long ptr;
 
 	path = btrfs_alloc_path();
-	BUG_ON(!path);
+	if (!path)
+		return -ENOMEM;
+
 	ret = btrfs_search_slot(trans, root, key, path, 0, 1);
 	if (ret < 0)
 		goto out;
-- 
1.7.9

[patch 11/35] btrfs: Simplify btrfs_submit_bio_hook

[Jeff Mahoney]

btrfs_submit_bio_hook currently calls btrfs_bio_wq_end_io in either case
of an if statement that determines one of the arguments.

This patch moves the function call outside of the if statement and uses it
to only determine the different argument. This allows us to catch an
error in one place in a more visually obvious way.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/inode.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 892b347..f0dfd7d 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1479,13 +1479,14 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 	int ret = 0;
 	int skip_sum;
+	int metadata = 0;
 
 	skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
 	if (btrfs_is_free_space_inode(root, inode))
-		ret = btrfs_bio_wq_end_io(root->fs_info, bio, 2);
-	else
-		ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
+		metadata = 2;
+
+	ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
 	BUG_ON(ret);
 
 	if (!(rw & REQ_WRITE)) {
-- 
1.7.9

[patch 12/35] btrfs: Factor out tree->ops->merge_bio_hook call

[Jeff Mahoney]

In submit_extent_page, there's a visually noisy if statement that, in
the midst of other conditions, does the tree dependency for tree->ops
and tree->ops->merge_bio_hook before calling it, and then another
condition afterwards. If an error is returned from merge_bio_hook,
there's no way to catch it. It's considered a routine "1" return
value instead of a failure.

This patch factors out the dependency check into a new local merge_bio
routine and BUG's on an error. The if statement is less noisy as a side-
effect.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/extent_io.c |   17 ++++++++++++++---
 fs/btrfs/inode.c     |    5 +++--
 2 files changed, 17 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index a4a7a18..c342e92 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -2429,6 +2429,19 @@ static int submit_one_bio(int rw, struct bio *bio, int mirror_num,
 	return ret;
 }
 
+static int merge_bio(struct extent_io_tree *tree, struct page *page,
+		     unsigned long offset, size_t size, struct bio *bio,
+		     unsigned long bio_flags)
+{
+	int ret = 0;
+	if (tree->ops && tree->ops->merge_bio_hook)
+		ret = tree->ops->merge_bio_hook(page, offset, size, bio,
+						bio_flags);
+	BUG_ON(ret < 0);
+	return ret;
+
+}
+
 static int submit_extent_page(int rw, struct extent_io_tree *tree,
 			      struct page *page, sector_t sector,
 			      size_t size, unsigned long offset,
@@ -2457,9 +2470,7 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 				sector;
 
 		if (prev_bio_flags != bio_flags || !contig ||
-		    (tree->ops && tree->ops->merge_bio_hook &&
-		     tree->ops->merge_bio_hook(page, offset, page_size, bio,
-					       bio_flags)) ||
+		    merge_bio(tree, page, offset, page_size, bio, bio_flags) ||
 		    bio_add_page(bio, page, page_size, offset) < page_size) {
 			ret = submit_one_bio(rw, bio, mirror_num,
 					     prev_bio_flags);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index f0dfd7d..ad1e88d 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1425,10 +1425,11 @@ int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
 	map_length = length;
 	ret = btrfs_map_block(map_tree, READ, logical,
 			      &map_length, NULL, 0);
-
+	/* Will always return 0 or 1 with map_multi == NULL */
+	BUG_ON(ret < 0);
 	if (map_length < length + size)
 		return 1;
-	return ret;
+	return 0;
 }
 
 /*
-- 
1.7.9

[patch 13/35] btrfs: ->submit_bio_hook error push-up

[Jeff Mahoney]

This pushes failures from the submit_bio_hook callbacks,
btrfs_submit_bio_hook and btree_submit_bio_hook into the callers, including
callers of submit_one_bio where it catches the failures with BUG_ON.

It also pushes up through the ->readpage_io_failed_hook to
end_bio_extent_writepage where the error is already caught with BUG_ON.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/disk-io.c   |    6 +++---
 fs/btrfs/extent_io.c |   37 ++++++++++++++++++++++++++-----------
 fs/btrfs/inode.c     |    3 ++-
 3 files changed, 31 insertions(+), 15 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 155c4e3..2e4428b 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -847,9 +847,9 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 {
 	int ret;
 
-	ret = btrfs_bio_wq_end_io(BTRFS_I(inode)->root->fs_info,
-					  bio, 1);
-	BUG_ON(ret);
+	ret = btrfs_bio_wq_end_io(BTRFS_I(inode)->root->fs_info, bio, 1);
+	if (ret)
+		return ret;
 
 	if (!(rw & REQ_WRITE)) {
 		/*
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index c342e92..8368baa 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -2199,6 +2199,7 @@ int end_extent_writepage(struct page *page, int err, u64 start, u64 end)
 		/* Writeback already completed */
 		if (ret == 0)
 			return 1;
+		BUG_ON(ret < 0);
 	}
 
 	if (!uptodate) {
@@ -2351,6 +2352,7 @@ error_handled:
 				if (ret == 0)
 					goto error_handled;
 			}
+			BUG_ON(ret < 0);
 		}
 
 		if (uptodate) {
@@ -2402,8 +2404,8 @@ btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs,
 	return bio;
 }
 
-static int submit_one_bio(int rw, struct bio *bio, int mirror_num,
-			  unsigned long bio_flags)
+static int __must_check submit_one_bio(int rw, struct bio *bio,
+				       int mirror_num, unsigned long bio_flags)
 {
 	int ret = 0;
 	struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1;
@@ -2474,6 +2476,7 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 		    bio_add_page(bio, page, page_size, offset) < page_size) {
 			ret = submit_one_bio(rw, bio, mirror_num,
 					     prev_bio_flags);
+			BUG_ON(ret < 0);
 			bio = NULL;
 		} else {
 			return 0;
@@ -2494,8 +2497,10 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 
 	if (bio_ret)
 		*bio_ret = bio;
-	else
+	else {
 		ret = submit_one_bio(rw, bio, mirror_num, bio_flags);
+		BUG_ON(ret < 0);
+	}
 
 	return ret;
 }
@@ -2707,8 +2712,10 @@ int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 
 	ret = __extent_read_full_page(tree, page, get_extent, &bio, mirror_num,
 				      &bio_flags);
-	if (bio)
+	if (bio) {
 		ret = submit_one_bio(READ, bio, mirror_num, bio_flags);
+		BUG_ON(ret < 0);
+	}
 	return ret;
 }
 
@@ -3126,10 +3133,14 @@ retry:
 static void flush_epd_write_bio(struct extent_page_data *epd)
 {
 	if (epd->bio) {
+		int rw = WRITE;
+		int ret;
+
 		if (epd->sync_io)
-			submit_one_bio(WRITE_SYNC, epd->bio, 0, 0);
-		else
-			submit_one_bio(WRITE, epd->bio, 0, 0);
+			rw = WRITE_SYNC;
+
+		ret = submit_one_bio(rw, epd->bio, 0, 0);
+		BUG_ON(ret < 0);
 		epd->bio = NULL;
 	}
 }
@@ -3245,8 +3256,10 @@ int extent_readpages(struct extent_io_tree *tree,
 		page_cache_release(page);
 	}
 	BUG_ON(!list_empty(pages));
-	if (bio)
-		submit_one_bio(READ, bio, 0, bio_flags);
+	if (bio) {
+		int ret = submit_one_bio(READ, bio, 0, bio_flags);
+		BUG_ON(ret < 0);
+	}
 	return 0;
 }
 
@@ -4075,8 +4088,10 @@ int read_extent_buffer_pages(struct extent_io_tree *tree,
 		}
 	}
 
-	if (bio)
-		submit_one_bio(READ, bio, mirror_num, bio_flags);
+	if (bio) {
+		err = submit_one_bio(READ, bio, mirror_num, bio_flags);
+		BUG_ON(err < 0);
+	}
 
 	if (ret || wait != WAIT_COMPLETE)
 		return ret;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index ad1e88d..5394bdb 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1488,7 +1488,8 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 		metadata = 2;
 
 	ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
-	BUG_ON(ret);
+	if (ret)
+		return ret;
 
 	if (!(rw & REQ_WRITE)) {
 		if (bio_flags & EXTENT_BIO_COMPRESSED) {
-- 
1.7.9

[patch 14/35] btrfs: __add_reloc_root error push-up

[Jeff Mahoney]

This patch pushes kmalloc errors up to the caller and BUGs in the caller.

The BUG_ON for duplicate reloc tree root insertion is replaced with a
panic explaining the issue.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/relocation.c |   22 ++++++++++++++++------
 1 files changed, 16 insertions(+), 6 deletions(-)

diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index e5996ff..974b0df 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -1221,14 +1221,15 @@ fail:
 /*
  * helper to add 'address of tree root -> reloc tree' mapping
  */
-static int __add_reloc_root(struct btrfs_root *root)
+static int __must_check __add_reloc_root(struct btrfs_root *root)
 {
 	struct rb_node *rb_node;
 	struct mapping_node *node;
 	struct reloc_control *rc = root->fs_info->reloc_ctl;
 
 	node = kmalloc(sizeof(*node), GFP_NOFS);
-	BUG_ON(!node);
+	if (!node)
+		return -ENOMEM;
 
 	node->bytenr = root->node->start;
 	node->data = root;
@@ -1237,7 +1238,12 @@ static int __add_reloc_root(struct btrfs_root *root)
 	rb_node = tree_insert(&rc->reloc_root_tree.rb_root,
 			      node->bytenr, &node->rb_node);
 	spin_unlock(&rc->reloc_root_tree.lock);
-	BUG_ON(rb_node);
+	if (rb_node) {
+		kfree(node);
+		btrfs_panic(root->fs_info, -EEXIST, "Duplicate root found "
+			    "for start=%llu while inserting into relocation "
+			    "tree\n");
+	}
 
 	list_add_tail(&root->root_list, &rc->reloc_roots);
 	return 0;
@@ -1353,6 +1359,7 @@ int btrfs_init_reloc_root(struct btrfs_trans_handle *trans,
 	struct btrfs_root *reloc_root;
 	struct reloc_control *rc = root->fs_info->reloc_ctl;
 	int clear_rsv = 0;
+	int ret;
 
 	if (root->reloc_root) {
 		reloc_root = root->reloc_root;
@@ -1372,7 +1379,8 @@ int btrfs_init_reloc_root(struct btrfs_trans_handle *trans,
 	if (clear_rsv)
 		trans->block_rsv = NULL;
 
-	__add_reloc_root(reloc_root);
+	ret = __add_reloc_root(reloc_root);
+	BUG_ON(ret < 0);
 	root->reloc_root = reloc_root;
 	return 0;
 }
@@ -4226,7 +4234,8 @@ int btrfs_recover_relocation(struct btrfs_root *root)
 				       reloc_root->root_key.offset);
 		BUG_ON(IS_ERR(fs_root));
 
-		__add_reloc_root(reloc_root);
+		err = __add_reloc_root(reloc_root);
+		BUG_ON(err < 0);
 		fs_root->reloc_root = reloc_root;
 	}
 
@@ -4428,7 +4437,8 @@ void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 	reloc_root = create_reloc_root(trans, root->reloc_root,
 				       new_root->root_key.objectid);
 
-	__add_reloc_root(reloc_root);
+	ret = __add_reloc_root(reloc_root);
+	BUG_ON(ret < 0);
 	new_root->reloc_root = reloc_root;
 
 	if (rc->create_reloc_tree) {
-- 
1.7.9

[patch 15/35] btrfs: return void in functions without error conditions

[Jeff Mahoney]

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---

 fs/btrfs/async-thread.c      |   15 +--
 fs/btrfs/async-thread.h      |    4 
 fs/btrfs/compression.c       |    8 -
 fs/btrfs/compression.h       |    2 
 fs/btrfs/ctree.c             |  200 ++++++++++++++-----------------------------
 fs/btrfs/ctree.h             |   45 ++++-----
 fs/btrfs/delayed-inode.c     |    6 -
 fs/btrfs/delayed-ref.c       |   27 +----
 fs/btrfs/dir-item.c          |    9 -
 fs/btrfs/disk-io.c           |   39 +++-----
 fs/btrfs/disk-io.h           |    2 
 fs/btrfs/extent-tree.c       |   60 +++++-------
 fs/btrfs/extent_io.c         |   35 +++----
 fs/btrfs/extent_io.h         |    4 
 fs/btrfs/file-item.c         |   25 ++---
 fs/btrfs/inode-item.c        |    4 
 fs/btrfs/inode.c             |    7 -
 fs/btrfs/locking.c           |    6 -
 fs/btrfs/locking.h           |    4 
 fs/btrfs/ordered-data.c      |   48 +++-------
 fs/btrfs/ordered-data.h      |   24 ++---
 fs/btrfs/scrub.c             |   36 ++-----
 fs/btrfs/super.c             |    5 -
 fs/btrfs/transaction.c       |    3 
 fs/btrfs/tree-log.c          |   17 +--
 fs/btrfs/tree-log.h          |    2 
 fs/btrfs/volumes.c           |   18 +--
 fs/btrfs/volumes.h           |    4 
 include/trace/events/btrfs.h |   44 +++++++++
 29 files changed, 293 insertions(+), 410 deletions(-)

diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 0cc20b3..4270414 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -171,11 +171,11 @@ out:
 	spin_unlock_irqrestore(&workers->lock, flags);
 }
 
-static noinline int run_ordered_completions(struct btrfs_workers *workers,
+static noinline void run_ordered_completions(struct btrfs_workers *workers,
 					    struct btrfs_work *work)
 {
 	if (!workers->ordered)
-		return 0;
+		return;
 
 	set_bit(WORK_DONE_BIT, &work->flags);
 
@@ -213,7 +213,6 @@ static noinline int run_ordered_completions(struct btrfs_workers *workers,
 	}
 
 	spin_unlock(&workers->order_lock);
-	return 0;
 }
 
 static void put_worker(struct btrfs_worker_thread *worker)
@@ -399,7 +398,7 @@ again:
 /*
  * this will wait for all the worker threads to shutdown
  */
-int btrfs_stop_workers(struct btrfs_workers *workers)
+void btrfs_stop_workers(struct btrfs_workers *workers)
 {
 	struct list_head *cur;
 	struct btrfs_worker_thread *worker;
@@ -427,7 +426,6 @@ int btrfs_stop_workers(struct btrfs_workers *workers)
 		put_worker(worker);
 	}
 	spin_unlock_irq(&workers->lock);
-	return 0;
 }
 
 /*
@@ -615,14 +613,14 @@ found:
  * it was taken from.  It is intended for use with long running work functions
  * that make some progress and want to give the cpu up for others.
  */
-int btrfs_requeue_work(struct btrfs_work *work)
+void btrfs_requeue_work(struct btrfs_work *work)
 {
 	struct btrfs_worker_thread *worker = work->worker;
 	unsigned long flags;
 	int wake = 0;
 
 	if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
-		goto out;
+		return;
 
 	spin_lock_irqsave(&worker->lock, flags);
 	if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
@@ -649,9 +647,6 @@ int btrfs_requeue_work(struct btrfs_work *work)
 	if (wake)
 		wake_up_process(worker->task);
 	spin_unlock_irqrestore(&worker->lock, flags);
-out:
-
-	return 0;
 }
 
 void btrfs_set_work_high_prio(struct btrfs_work *work)
diff --git a/fs/btrfs/async-thread.h b/fs/btrfs/async-thread.h
index f34cc31..063698b 100644
--- a/fs/btrfs/async-thread.h
+++ b/fs/btrfs/async-thread.h
@@ -111,9 +111,9 @@ struct btrfs_workers {
 
 void btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work);
 int btrfs_start_workers(struct btrfs_workers *workers);
-int btrfs_stop_workers(struct btrfs_workers *workers);
+void btrfs_stop_workers(struct btrfs_workers *workers);
 void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
 			struct btrfs_workers *async_starter);
-int btrfs_requeue_work(struct btrfs_work *work);
+void btrfs_requeue_work(struct btrfs_work *work);
 void btrfs_set_work_high_prio(struct btrfs_work *work);
 #endif
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index d02c27c..0ed1ed2 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -226,8 +226,8 @@ out:
  * Clear the writeback bits on all of the file
  * pages for a compressed write
  */
-static noinline int end_compressed_writeback(struct inode *inode, u64 start,
-					     unsigned long ram_size)
+static noinline void end_compressed_writeback(struct inode *inode, u64 start,
+					      unsigned long ram_size)
 {
 	unsigned long index = start >> PAGE_CACHE_SHIFT;
 	unsigned long end_index = (start + ram_size - 1) >> PAGE_CACHE_SHIFT;
@@ -253,7 +253,6 @@ static noinline int end_compressed_writeback(struct inode *inode, u64 start,
 		index += ret;
 	}
 	/* the inode may be gone now */
-	return 0;
 }
 
 /*
@@ -734,7 +733,7 @@ struct btrfs_compress_op *btrfs_compress_op[] = {
 	&btrfs_lzo_compress,
 };
 
-int __init btrfs_init_compress(void)
+void __init btrfs_init_compress(void)
 {
 	int i;
 
@@ -744,7 +743,6 @@ int __init btrfs_init_compress(void)
 		atomic_set(&comp_alloc_workspace[i], 0);
 		init_waitqueue_head(&comp_workspace_wait[i]);
 	}
-	return 0;
 }
 
 /*
diff --git a/fs/btrfs/compression.h b/fs/btrfs/compression.h
index a12059f..9afb0a6 100644
--- a/fs/btrfs/compression.h
+++ b/fs/btrfs/compression.h
@@ -19,7 +19,7 @@
 #ifndef __BTRFS_COMPRESSION_
 #define __BTRFS_COMPRESSION_
 
-int btrfs_init_compress(void);
+void btrfs_init_compress(void);
 void btrfs_exit_compress(void);
 
 int btrfs_compress_pages(int type, struct address_space *mapping,
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 0639a55..9d472c2 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -36,7 +36,7 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root,
 			      struct extent_buffer *dst_buf,
 			      struct extent_buffer *src_buf);
-static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		   struct btrfs_path *path, int level, int slot);
 
 struct btrfs_path *btrfs_alloc_path(void)
@@ -1010,10 +1010,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		if (btrfs_header_nritems(right) == 0) {
 			clean_tree_block(trans, root, right);
 			btrfs_tree_unlock(right);
-			wret = del_ptr(trans, root, path, level + 1, pslot +
-				       1);
-			if (wret)
-				ret = wret;
+			del_ptr(trans, root, path, level + 1, pslot + 1);
 			root_sub_used(root, right->len);
 			btrfs_free_tree_block(trans, root, right, 0, 1, 0);
 			free_extent_buffer(right);
@@ -1051,9 +1048,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 	if (btrfs_header_nritems(mid) == 0) {
 		clean_tree_block(trans, root, mid);
 		btrfs_tree_unlock(mid);
-		wret = del_ptr(trans, root, path, level + 1, pslot);
-		if (wret)
-			ret = wret;
+		del_ptr(trans, root, path, level + 1, pslot);
 		root_sub_used(root, mid->len);
 		btrfs_free_tree_block(trans, root, mid, 0, 1, 0);
 		free_extent_buffer(mid);
@@ -1881,15 +1876,12 @@ done:
  * fixing up pointers when a given leaf/node is not in slot 0 of the
  * higher levels
  *
- * If this fails to write a tree block, it returns -1, but continues
- * fixing up the blocks in ram so the tree is consistent.
  */
-static int fixup_low_keys(struct btrfs_trans_handle *trans,
-			  struct btrfs_root *root, struct btrfs_path *path,
-			  struct btrfs_disk_key *key, int level)
+static void fixup_low_keys(struct btrfs_trans_handle *trans,
+			   struct btrfs_root *root, struct btrfs_path *path,
+			   struct btrfs_disk_key *key, int level)
 {
 	int i;
-	int ret = 0;
 	struct extent_buffer *t;
 
 	for (i = level; i < BTRFS_MAX_LEVEL; i++) {
@@ -1902,7 +1894,6 @@ static int fixup_low_keys(struct btrfs_trans_handle *trans,
 		if (tslot != 0)
 			break;
 	}
-	return ret;
 }
 
 /*
@@ -1911,9 +1902,9 @@ static int fixup_low_keys(struct btrfs_trans_handle *trans,
  * This function isn't completely safe. It's the caller's responsibility
  * that the new key won't break the order
  */
-int btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
-			    struct btrfs_root *root, struct btrfs_path *path,
-			    struct btrfs_key *new_key)
+void btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
+			     struct btrfs_root *root, struct btrfs_path *path,
+			     struct btrfs_key *new_key)
 {
 	struct btrfs_disk_key disk_key;
 	struct extent_buffer *eb;
@@ -1923,13 +1914,11 @@ int btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
 	slot = path->slots[0];
 	if (slot > 0) {
 		btrfs_item_key(eb, &disk_key, slot - 1);
-		if (comp_keys(&disk_key, new_key) >= 0)
-			return -1;
+		BUG_ON(comp_keys(&disk_key, new_key) >= 0);
 	}
 	if (slot < btrfs_header_nritems(eb) - 1) {
 		btrfs_item_key(eb, &disk_key, slot + 1);
-		if (comp_keys(&disk_key, new_key) <= 0)
-			return -1;
+		BUG_ON(comp_keys(&disk_key, new_key) <= 0);
 	}
 
 	btrfs_cpu_key_to_disk(&disk_key, new_key);
@@ -1937,7 +1926,6 @@ int btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(eb);
 	if (slot == 0)
 		fixup_low_keys(trans, root, path, &disk_key, 1);
-	return 0;
 }
 
 /*
@@ -2140,12 +2128,11 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
  *
  * slot and level indicate where you want the key to go, and
  * blocknr is the block the key points to.
- *
- * returns zero on success and < 0 on any error
  */
-static int insert_ptr(struct btrfs_trans_handle *trans, struct btrfs_root
-		      *root, struct btrfs_path *path, struct btrfs_disk_key
-		      *key, u64 bytenr, int slot, int level)
+static void insert_ptr(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root, struct btrfs_path *path,
+		       struct btrfs_disk_key *key, u64 bytenr,
+		       int slot, int level)
 {
 	struct extent_buffer *lower;
 	int nritems;
@@ -2155,8 +2142,7 @@ static int insert_ptr(struct btrfs_trans_handle *trans, struct btrfs_root
 	lower = path->nodes[level];
 	nritems = btrfs_header_nritems(lower);
 	BUG_ON(slot > nritems);
-	if (nritems == BTRFS_NODEPTRS_PER_BLOCK(root))
-		BUG();
+	BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(root));
 	if (slot != nritems) {
 		memmove_extent_buffer(lower,
 			      btrfs_node_key_ptr_offset(slot + 1),
@@ -2169,7 +2155,6 @@ static int insert_ptr(struct btrfs_trans_handle *trans, struct btrfs_root
 	btrfs_set_node_ptr_generation(lower, slot, trans->transid);
 	btrfs_set_header_nritems(lower, nritems + 1);
 	btrfs_mark_buffer_dirty(lower);
-	return 0;
 }
 
 /*
@@ -2190,7 +2175,6 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	struct btrfs_disk_key disk_key;
 	int mid;
 	int ret;
-	int wret;
 	u32 c_nritems;
 
 	c = path->nodes[level];
@@ -2247,11 +2231,8 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(c);
 	btrfs_mark_buffer_dirty(split);
 
-	wret = insert_ptr(trans, root, path, &disk_key, split->start,
-			  path->slots[level + 1] + 1,
-			  level + 1);
-	if (wret)
-		ret = wret;
+	insert_ptr(trans, root, path, &disk_key, split->start,
+		   path->slots[level + 1] + 1, level + 1);
 
 	if (path->slots[level] >= mid) {
 		path->slots[level] -= mid;
@@ -2537,7 +2518,6 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 	u32 old_left_nritems;
 	u32 nr;
 	int ret = 0;
-	int wret;
 	u32 this_item_size;
 	u32 old_left_item_size;
 
@@ -2643,9 +2623,7 @@ static noinline int __push_leaf_left(struct btrfs_trans_handle *trans,
 		clean_tree_block(trans, root, right);
 
 	btrfs_item_key(right, &disk_key, 0);
-	wret = fixup_low_keys(trans, root, path, &disk_key, 1);
-	if (wret)
-		ret = wret;
+	fixup_low_keys(trans, root, path, &disk_key, 1);
 
 	/* then fixup the leaf pointer in the path */
 	if (path->slots[0] < push_items) {
@@ -2738,21 +2716,17 @@ out:
 /*
  * split the path's leaf in two, making sure there is at least data_size
  * available for the resulting leaf level of the path.
- *
- * returns 0 if all went well and < 0 on failure.
  */
-static noinline int copy_for_split(struct btrfs_trans_handle *trans,
-			       struct btrfs_root *root,
-			       struct btrfs_path *path,
-			       struct extent_buffer *l,
-			       struct extent_buffer *right,
-			       int slot, int mid, int nritems)
+static noinline void copy_for_split(struct btrfs_trans_handle *trans,
+				    struct btrfs_root *root,
+				    struct btrfs_path *path,
+				    struct extent_buffer *l,
+				    struct extent_buffer *right,
+				    int slot, int mid, int nritems)
 {
 	int data_copy_size;
 	int rt_data_off;
 	int i;
-	int ret = 0;
-	int wret;
 	struct btrfs_disk_key disk_key;
 
 	nritems = nritems - mid;
@@ -2780,12 +2754,9 @@ static noinline int copy_for_split(struct btrfs_trans_handle *trans,
 	}
 
 	btrfs_set_header_nritems(l, mid);
-	ret = 0;
 	btrfs_item_key(right, &disk_key, 0);
-	wret = insert_ptr(trans, root, path, &disk_key, right->start,
-			  path->slots[1] + 1, 1);
-	if (wret)
-		ret = wret;
+	insert_ptr(trans, root, path, &disk_key, right->start,
+		   path->slots[1] + 1, 1);
 
 	btrfs_mark_buffer_dirty(right);
 	btrfs_mark_buffer_dirty(l);
@@ -2803,8 +2774,6 @@ static noinline int copy_for_split(struct btrfs_trans_handle *trans,
 	}
 
 	BUG_ON(path->slots[0] < 0);
-
-	return ret;
 }
 
 /*
@@ -2993,12 +2962,8 @@ again:
 	if (split == 0) {
 		if (mid <= slot) {
 			btrfs_set_header_nritems(right, 0);
-			wret = insert_ptr(trans, root, path,
-					  &disk_key, right->start,
-					  path->slots[1] + 1, 1);
-			if (wret)
-				ret = wret;
-
+			insert_ptr(trans, root, path, &disk_key, right->start,
+				   path->slots[1] + 1, 1);
 			btrfs_tree_unlock(path->nodes[0]);
 			free_extent_buffer(path->nodes[0]);
 			path->nodes[0] = right;
@@ -3006,29 +2971,21 @@ again:
 			path->slots[1] += 1;
 		} else {
 			btrfs_set_header_nritems(right, 0);
-			wret = insert_ptr(trans, root, path,
-					  &disk_key,
-					  right->start,
+			insert_ptr(trans, root, path, &disk_key, right->start,
 					  path->slots[1], 1);
-			if (wret)
-				ret = wret;
 			btrfs_tree_unlock(path->nodes[0]);
 			free_extent_buffer(path->nodes[0]);
 			path->nodes[0] = right;
 			path->slots[0] = 0;
-			if (path->slots[1] == 0) {
-				wret = fixup_low_keys(trans, root,
-						path, &disk_key, 1);
-				if (wret)
-					ret = wret;
-			}
+			if (path->slots[1] == 0)
+				fixup_low_keys(trans, root, path,
+					       &disk_key, 1);
 		}
 		btrfs_mark_buffer_dirty(right);
 		return ret;
 	}
 
-	ret = copy_for_split(trans, root, path, l, right, slot, mid, nritems);
-	BUG_ON(ret);
+	copy_for_split(trans, root, path, l, right, slot, mid, nritems);
 
 	if (split == 2) {
 		BUG_ON(num_doubles != 0);
@@ -3036,7 +2993,7 @@ again:
 		goto again;
 	}
 
-	return ret;
+	return 0;
 
 push_for_double:
 	push_for_double_split(trans, root, path, data_size);
@@ -3238,11 +3195,9 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
 		return ret;
 
 	path->slots[0]++;
-	ret = setup_items_for_insert(trans, root, path, new_key, &item_size,
-				     item_size, item_size +
-				     sizeof(struct btrfs_item), 1);
-	BUG_ON(ret);
-
+	setup_items_for_insert(trans, root, path, new_key, &item_size,
+			       item_size, item_size +
+			       sizeof(struct btrfs_item), 1);
 	leaf = path->nodes[0];
 	memcpy_extent_buffer(leaf,
 			     btrfs_item_ptr_offset(leaf, path->slots[0]),
@@ -3257,10 +3212,10 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
  * off the end of the item or if we shift the item to chop bytes off
  * the front.
  */
-int btrfs_truncate_item(struct btrfs_trans_handle *trans,
-			struct btrfs_root *root,
-			struct btrfs_path *path,
-			u32 new_size, int from_end)
+void btrfs_truncate_item(struct btrfs_trans_handle *trans,
+			 struct btrfs_root *root,
+			 struct btrfs_path *path,
+			 u32 new_size, int from_end)
 {
 	int slot;
 	struct extent_buffer *leaf;
@@ -3277,7 +3232,7 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans,
 
 	old_size = btrfs_item_size_nr(leaf, slot);
 	if (old_size == new_size)
-		return 0;
+		return;
 
 	nritems = btrfs_header_nritems(leaf);
 	data_end = leaf_data_end(root, leaf);
@@ -3350,15 +3305,14 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans,
 		btrfs_print_leaf(root, leaf);
 		BUG();
 	}
-	return 0;
 }
 
 /*
  * make the item pointed to by the path bigger, data_size is the new size.
  */
-int btrfs_extend_item(struct btrfs_trans_handle *trans,
-		      struct btrfs_root *root, struct btrfs_path *path,
-		      u32 data_size)
+void btrfs_extend_item(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root, struct btrfs_path *path,
+		       u32 data_size)
 {
 	int slot;
 	struct extent_buffer *leaf;
@@ -3416,7 +3370,6 @@ int btrfs_extend_item(struct btrfs_trans_handle *trans,
 		btrfs_print_leaf(root, leaf);
 		BUG();
 	}
-	return 0;
 }
 
 /*
@@ -3544,7 +3497,7 @@ int btrfs_insert_some_items(struct btrfs_trans_handle *trans,
 	ret = 0;
 	if (slot == 0) {
 		btrfs_cpu_key_to_disk(&disk_key, cpu_key);
-		ret = fixup_low_keys(trans, root, path, &disk_key, 1);
+		fixup_low_keys(trans, root, path, &disk_key, 1);
 	}
 
 	if (btrfs_leaf_free_space(root, leaf) < 0) {
@@ -3562,17 +3515,16 @@ out:
  * to save stack depth by doing the bulk of the work in a function
  * that doesn't call btrfs_search_slot
  */
-int setup_items_for_insert(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *root, struct btrfs_path *path,
-			   struct btrfs_key *cpu_key, u32 *data_size,
-			   u32 total_data, u32 total_size, int nr)
+void setup_items_for_insert(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, struct btrfs_path *path,
+			    struct btrfs_key *cpu_key, u32 *data_size,
+			    u32 total_data, u32 total_size, int nr)
 {
 	struct btrfs_item *item;
 	int i;
 	u32 nritems;
 	unsigned int data_end;
 	struct btrfs_disk_key disk_key;
-	int ret;
 	struct extent_buffer *leaf;
 	int slot;
 
@@ -3633,10 +3585,9 @@ int setup_items_for_insert(struct btrfs_trans_handle *trans,
 
 	btrfs_set_header_nritems(leaf, nritems + nr);
 
-	ret = 0;
 	if (slot == 0) {
 		btrfs_cpu_key_to_disk(&disk_key, cpu_key);
-		ret = fixup_low_keys(trans, root, path, &disk_key, 1);
+		fixup_low_keys(trans, root, path, &disk_key, 1);
 	}
 	btrfs_unlock_up_safe(path, 1);
 	btrfs_mark_buffer_dirty(leaf);
@@ -3645,7 +3596,6 @@ int setup_items_for_insert(struct btrfs_trans_handle *trans,
 		btrfs_print_leaf(root, leaf);
 		BUG();
 	}
-	return ret;
 }
 
 /*
@@ -3672,16 +3622,14 @@ int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
 	if (ret == 0)
 		return -EEXIST;
 	if (ret < 0)
-		goto out;
+		return ret;
 
 	slot = path->slots[0];
 	BUG_ON(slot < 0);
 
-	ret = setup_items_for_insert(trans, root, path, cpu_key, data_size,
+	setup_items_for_insert(trans, root, path, cpu_key, data_size,
 			       total_data, total_size, nr);
-
-out:
-	return ret;
+	return 0;
 }
 
 /*
@@ -3717,13 +3665,11 @@ int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
  * the tree should have been previously balanced so the deletion does not
  * empty a node.
  */
-static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
-		   struct btrfs_path *path, int level, int slot)
+static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+		    struct btrfs_path *path, int level, int slot)
 {
 	struct extent_buffer *parent = path->nodes[level];
 	u32 nritems;
-	int ret = 0;
-	int wret;
 
 	nritems = btrfs_header_nritems(parent);
 	if (slot != nritems - 1) {
@@ -3743,12 +3689,9 @@ static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		struct btrfs_disk_key disk_key;
 
 		btrfs_node_key(parent, &disk_key, 0);
-		wret = fixup_low_keys(trans, root, path, &disk_key, level + 1);
-		if (wret)
-			ret = wret;
+		fixup_low_keys(trans, root, path, &disk_key, level + 1);
 	}
 	btrfs_mark_buffer_dirty(parent);
-	return ret;
 }
 
 /*
@@ -3761,17 +3704,13 @@ static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
  * The path must have already been setup for deleting the leaf, including
  * all the proper balancing.  path->nodes[1] must be locked.
  */
-static noinline int btrfs_del_leaf(struct btrfs_trans_handle *trans,
-				   struct btrfs_root *root,
-				   struct btrfs_path *path,
-				   struct extent_buffer *leaf)
+static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
+				    struct btrfs_root *root,
+				    struct btrfs_path *path,
+				    struct extent_buffer *leaf)
 {
-	int ret;
-
 	WARN_ON(btrfs_header_generation(leaf) != trans->transid);
-	ret = del_ptr(trans, root, path, 1, path->slots[1]);
-	if (ret)
-		return ret;
+	del_ptr(trans, root, path, 1, path->slots[1]);
 
 	/*
 	 * btrfs_free_extent is expensive, we want to make sure we
@@ -3782,7 +3721,6 @@ static noinline int btrfs_del_leaf(struct btrfs_trans_handle *trans,
 	root_sub_used(root, leaf->len);
 
 	btrfs_free_tree_block(trans, root, leaf, 0, 1, 0);
-	return 0;
 }
 /*
  * delete the item at the leaf level in path.  If that empties
@@ -3839,8 +3777,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		} else {
 			btrfs_set_path_blocking(path);
 			clean_tree_block(trans, root, leaf);
-			ret = btrfs_del_leaf(trans, root, path, leaf);
-			BUG_ON(ret);
+			btrfs_del_leaf(trans, root, path, leaf);
 		}
 	} else {
 		int used = leaf_space_used(leaf, 0, nritems);
@@ -3848,10 +3785,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 			struct btrfs_disk_key disk_key;
 
 			btrfs_item_key(leaf, &disk_key, 0);
-			wret = fixup_low_keys(trans, root, path,
-					      &disk_key, 1);
-			if (wret)
-				ret = wret;
+			fixup_low_keys(trans, root, path, &disk_key, 1);
 		}
 
 		/* delete the leaf if it is mostly empty */
@@ -3879,9 +3813,9 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 
 			if (btrfs_header_nritems(leaf) == 0) {
 				path->slots[1] = slot;
-				ret = btrfs_del_leaf(trans, root, path, leaf);
-				BUG_ON(ret);
+				btrfs_del_leaf(trans, root, path, leaf);
 				free_extent_buffer(leaf);
+				ret = 0;
 			} else {
 				/* if we're still in the path, make sure
 				 * we're dirty.  Otherwise, one of the
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 339e637..30c5a24 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2485,8 +2485,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans,
 int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len);
 int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root,
 				       u64 start, u64 len);
-int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root);
+void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
+				 struct btrfs_root *root);
 int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root);
 int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
@@ -2549,8 +2549,8 @@ void btrfs_block_rsv_release(struct btrfs_root *root,
 			     u64 num_bytes);
 int btrfs_set_block_group_ro(struct btrfs_root *root,
 			     struct btrfs_block_group_cache *cache);
-int btrfs_set_block_group_rw(struct btrfs_root *root,
-			     struct btrfs_block_group_cache *cache);
+void btrfs_set_block_group_rw(struct btrfs_root *root,
+			      struct btrfs_block_group_cache *cache);
 void btrfs_put_block_group_cache(struct btrfs_fs_info *info);
 u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo);
 int btrfs_error_unpin_extent_range(struct btrfs_root *root,
@@ -2569,9 +2569,9 @@ int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2);
 int btrfs_previous_item(struct btrfs_root *root,
 			struct btrfs_path *path, u64 min_objectid,
 			int type);
-int btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
-			    struct btrfs_root *root, struct btrfs_path *path,
-			    struct btrfs_key *new_key);
+void btrfs_set_item_key_safe(struct btrfs_trans_handle *trans,
+			     struct btrfs_root *root, struct btrfs_path *path,
+			     struct btrfs_key *new_key);
 struct extent_buffer *btrfs_root_node(struct btrfs_root *root);
 struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root);
 int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path,
@@ -2591,12 +2591,13 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
 		      struct extent_buffer **cow_ret, u64 new_root_objectid);
 int btrfs_block_can_be_shared(struct btrfs_root *root,
 			      struct extent_buffer *buf);
-int btrfs_extend_item(struct btrfs_trans_handle *trans, struct btrfs_root
-		      *root, struct btrfs_path *path, u32 data_size);
-int btrfs_truncate_item(struct btrfs_trans_handle *trans,
-			struct btrfs_root *root,
-			struct btrfs_path *path,
-			u32 new_size, int from_end);
+void btrfs_extend_item(struct btrfs_trans_handle *trans,
+		       struct btrfs_root *root, struct btrfs_path *path,
+		       u32 data_size);
+void btrfs_truncate_item(struct btrfs_trans_handle *trans,
+			 struct btrfs_root *root,
+			 struct btrfs_path *path,
+			 u32 new_size, int from_end);
 int btrfs_split_item(struct btrfs_trans_handle *trans,
 		     struct btrfs_root *root,
 		     struct btrfs_path *path,
@@ -2630,10 +2631,10 @@ static inline int btrfs_del_item(struct btrfs_trans_handle *trans,
 	return btrfs_del_items(trans, root, path, path->slots[0], 1);
 }
 
-int setup_items_for_insert(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *root, struct btrfs_path *path,
-			   struct btrfs_key *cpu_key, u32 *data_size,
-			   u32 total_data, u32 total_size, int nr);
+void setup_items_for_insert(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, struct btrfs_path *path,
+			    struct btrfs_key *cpu_key, u32 *data_size,
+			    u32 total_data, u32 total_size, int nr);
 int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
 		      *root, struct btrfs_key *key, void *data, u32 data_size);
 int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
@@ -2911,7 +2912,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root);
 void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root);
 int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size);
-int btrfs_invalidate_inodes(struct btrfs_root *root);
+void btrfs_invalidate_inodes(struct btrfs_root *root);
 void btrfs_add_delayed_iput(struct inode *inode);
 void btrfs_run_delayed_iputs(struct btrfs_root *root);
 int btrfs_prealloc_file_range(struct inode *inode, int mode,
@@ -3021,10 +3022,10 @@ void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 /* scrub.c */
 int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end,
 		    struct btrfs_scrub_progress *progress, int readonly);
-int btrfs_scrub_pause(struct btrfs_root *root);
-int btrfs_scrub_pause_super(struct btrfs_root *root);
-int btrfs_scrub_continue(struct btrfs_root *root);
-int btrfs_scrub_continue_super(struct btrfs_root *root);
+void btrfs_scrub_pause(struct btrfs_root *root);
+void btrfs_scrub_pause_super(struct btrfs_root *root);
+void btrfs_scrub_continue(struct btrfs_root *root);
+void btrfs_scrub_continue_super(struct btrfs_root *root);
 int btrfs_scrub_cancel(struct btrfs_root *root);
 int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev);
 int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid);
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index fe4cd0f..6829590 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -836,10 +836,8 @@ static int btrfs_batch_insert_items(struct btrfs_trans_handle *trans,
 	btrfs_clear_path_blocking(path, NULL, 0);
 
 	/* insert the keys of the items */
-	ret = setup_items_for_insert(trans, root, path, keys, data_size,
-				     total_data_size, total_size, nitems);
-	if (ret)
-		goto error;
+	setup_items_for_insert(trans, root, path, keys, data_size,
+			       total_data_size, total_size, nitems);
 
 	/* insert the dir index items */
 	slot = path->slots[0];
diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c
index 29ecd54..69f22e3 100644
--- a/fs/btrfs/delayed-ref.c
+++ b/fs/btrfs/delayed-ref.c
@@ -420,7 +420,7 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing,
  * this does all the dirty work in terms of maintaining the correct
  * overall modification count.
  */
-static noinline int add_delayed_ref_head(struct btrfs_fs_info *fs_info,
+static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 					struct btrfs_trans_handle *trans,
 					struct btrfs_delayed_ref_node *ref,
 					u64 bytenr, u64 num_bytes,
@@ -494,13 +494,12 @@ static noinline int add_delayed_ref_head(struct btrfs_fs_info *fs_info,
 		delayed_refs->num_entries++;
 		trans->delayed_ref_updates++;
 	}
-	return 0;
 }
 
 /*
  * helper to insert a delayed tree ref into the rbtree.
  */
-static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
+static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 					 struct btrfs_trans_handle *trans,
 					 struct btrfs_delayed_ref_node *ref,
 					 u64 bytenr, u64 num_bytes, u64 parent,
@@ -554,13 +553,12 @@ static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 		delayed_refs->num_entries++;
 		trans->delayed_ref_updates++;
 	}
-	return 0;
 }
 
 /*
  * helper to insert a delayed data ref into the rbtree.
  */
-static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info,
+static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 					 struct btrfs_trans_handle *trans,
 					 struct btrfs_delayed_ref_node *ref,
 					 u64 bytenr, u64 num_bytes, u64 parent,
@@ -616,7 +614,6 @@ static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 		delayed_refs->num_entries++;
 		trans->delayed_ref_updates++;
 	}
-	return 0;
 }
 
 /*
@@ -634,7 +631,6 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	struct btrfs_delayed_tree_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
 
 	BUG_ON(extent_op && extent_op->is_data);
 	ref = kmalloc(sizeof(*ref), GFP_NOFS);
@@ -656,14 +652,12 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
+	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
 				   num_bytes, action, 0);
-	BUG_ON(ret);
 
-	ret = add_delayed_tree_ref(fs_info, trans, &ref->node, bytenr,
+	add_delayed_tree_ref(fs_info, trans, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, level, action,
 				   for_cow);
-	BUG_ON(ret);
 	if (!need_ref_seq(for_cow, ref_root) &&
 	    waitqueue_active(&delayed_refs->seq_wait))
 		wake_up(&delayed_refs->seq_wait);
@@ -685,7 +679,6 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	struct btrfs_delayed_data_ref *ref;
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
 
 	BUG_ON(extent_op && !extent_op->is_data);
 	ref = kmalloc(sizeof(*ref), GFP_NOFS);
@@ -707,14 +700,12 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
 	 * insert both the head node and the new ref without dropping
 	 * the spin lock
 	 */
-	ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
+	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
 				   num_bytes, action, 1);
-	BUG_ON(ret);
 
-	ret = add_delayed_data_ref(fs_info, trans, &ref->node, bytenr,
+	add_delayed_data_ref(fs_info, trans, &ref->node, bytenr,
 				   num_bytes, parent, ref_root, owner, offset,
 				   action, for_cow);
-	BUG_ON(ret);
 	if (!need_ref_seq(for_cow, ref_root) &&
 	    waitqueue_active(&delayed_refs->seq_wait))
 		wake_up(&delayed_refs->seq_wait);
@@ -729,7 +720,6 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 {
 	struct btrfs_delayed_ref_head *head_ref;
 	struct btrfs_delayed_ref_root *delayed_refs;
-	int ret;
 
 	head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS);
 	if (!head_ref)
@@ -740,10 +730,9 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info,
 	delayed_refs = &trans->transaction->delayed_refs;
 	spin_lock(&delayed_refs->lock);
 
-	ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
+	add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr,
 				   num_bytes, BTRFS_UPDATE_DELAYED_HEAD,
 				   extent_op->is_data);
-	BUG_ON(ret);
 
 	if (waitqueue_active(&delayed_refs->seq_wait))
 		wake_up(&delayed_refs->seq_wait);
diff --git a/fs/btrfs/dir-item.c b/fs/btrfs/dir-item.c
index 31d84e7..7674330 100644
--- a/fs/btrfs/dir-item.c
+++ b/fs/btrfs/dir-item.c
@@ -49,9 +49,8 @@ static struct btrfs_dir_item *insert_with_overflow(struct btrfs_trans_handle
 		di = btrfs_match_dir_item_name(root, path, name, name_len);
 		if (di)
 			return ERR_PTR(-EEXIST);
-		ret = btrfs_extend_item(trans, root, path, data_size);
-	}
-	if (ret < 0)
+		btrfs_extend_item(trans, root, path, data_size);
+	} else if (ret < 0)
 		return ERR_PTR(ret);
 	WARN_ON(ret > 0);
 	leaf = path->nodes[0];
@@ -383,8 +382,8 @@ int btrfs_delete_one_dir_name(struct btrfs_trans_handle *trans,
 		start = btrfs_item_ptr_offset(leaf, path->slots[0]);
 		memmove_extent_buffer(leaf, ptr, ptr + sub_item_len,
 			item_len - (ptr + sub_item_len - start));
-		ret = btrfs_truncate_item(trans, root, path,
-					  item_len - sub_item_len, 1);
+		btrfs_truncate_item(trans, root, path,
+				    item_len - sub_item_len, 1);
 	}
 	return ret;
 }
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 2e4428b..73ccadc 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -50,12 +50,12 @@ static void end_workqueue_fn(struct btrfs_work *work);
 static void free_fs_root(struct btrfs_root *root);
 static void btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
 				    int read_only);
-static int btrfs_destroy_ordered_operations(struct btrfs_root *root);
-static int btrfs_destroy_ordered_extents(struct btrfs_root *root);
+static void btrfs_destroy_ordered_operations(struct btrfs_root *root);
+static void btrfs_destroy_ordered_extents(struct btrfs_root *root);
 static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 				      struct btrfs_root *root);
-static int btrfs_destroy_pending_snapshots(struct btrfs_transaction *t);
-static int btrfs_destroy_delalloc_inodes(struct btrfs_root *root);
+static void btrfs_destroy_pending_snapshots(struct btrfs_transaction *t);
+static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root);
 static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 					struct extent_io_tree *dirty_pages,
 					int mark);
@@ -1139,10 +1139,10 @@ void clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	}
 }
 
-static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
-			u32 stripesize, struct btrfs_root *root,
-			struct btrfs_fs_info *fs_info,
-			u64 objectid)
+static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
+			 u32 stripesize, struct btrfs_root *root,
+			 struct btrfs_fs_info *fs_info,
+			 u64 objectid)
 {
 	root->node = NULL;
 	root->commit_root = NULL;
@@ -1194,7 +1194,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	root->defrag_running = 0;
 	root->root_key.objectid = objectid;
 	root->anon_dev = 0;
-	return 0;
 }
 
 static int __must_check find_and_setup_root(struct btrfs_root *tree_root,
@@ -2897,7 +2896,7 @@ int write_ctree_super(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
-int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
+void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 {
 	spin_lock(&fs_info->fs_roots_radix_lock);
 	radix_tree_delete(&fs_info->fs_roots_radix,
@@ -2910,7 +2909,6 @@ int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 	__btrfs_remove_free_space_cache(root->free_ino_pinned);
 	__btrfs_remove_free_space_cache(root->free_ino_ctl);
 	free_fs_root(root);
-	return 0;
 }
 
 static void free_fs_root(struct btrfs_root *root)
@@ -2927,7 +2925,7 @@ static void free_fs_root(struct btrfs_root *root)
 	kfree(root);
 }
 
-static int del_fs_roots(struct btrfs_fs_info *fs_info)
+static void del_fs_roots(struct btrfs_fs_info *fs_info)
 {
 	int ret;
 	struct btrfs_root *gang[8];
@@ -2956,7 +2954,6 @@ static int del_fs_roots(struct btrfs_fs_info *fs_info)
 		for (i = 0; i < ret; i++)
 			btrfs_free_fs_root(fs_info, gang[i]);
 	}
-	return 0;
 }
 
 int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info)
@@ -3299,7 +3296,7 @@ int btrfs_error_commit_super(struct btrfs_root *root)
 	return ret;
 }
 
-static int btrfs_destroy_ordered_operations(struct btrfs_root *root)
+static void btrfs_destroy_ordered_operations(struct btrfs_root *root)
 {
 	struct btrfs_inode *btrfs_inode;
 	struct list_head splice;
@@ -3321,11 +3318,9 @@ static int btrfs_destroy_ordered_operations(struct btrfs_root *root)
 
 	spin_unlock(&root->fs_info->ordered_extent_lock);
 	mutex_unlock(&root->fs_info->ordered_operations_mutex);
-
-	return 0;
 }
 
-static int btrfs_destroy_ordered_extents(struct btrfs_root *root)
+static void btrfs_destroy_ordered_extents(struct btrfs_root *root)
 {
 	struct list_head splice;
 	struct btrfs_ordered_extent *ordered;
@@ -3357,8 +3352,6 @@ static int btrfs_destroy_ordered_extents(struct btrfs_root *root)
 	}
 
 	spin_unlock(&root->fs_info->ordered_extent_lock);
-
-	return 0;
 }
 
 static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
@@ -3413,7 +3406,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 	return ret;
 }
 
-static int btrfs_destroy_pending_snapshots(struct btrfs_transaction *t)
+static void btrfs_destroy_pending_snapshots(struct btrfs_transaction *t)
 {
 	struct btrfs_pending_snapshot *snapshot;
 	struct list_head splice;
@@ -3431,11 +3424,9 @@ static int btrfs_destroy_pending_snapshots(struct btrfs_transaction *t)
 
 		kfree(snapshot);
 	}
-
-	return 0;
 }
 
-static int btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
+static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
 {
 	struct btrfs_inode *btrfs_inode;
 	struct list_head splice;
@@ -3455,8 +3446,6 @@ static int btrfs_destroy_delalloc_inodes(struct btrfs_root *root)
 	}
 
 	spin_unlock(&root->fs_info->delalloc_lock);
-
-	return 0;
 }
 
 static int btrfs_destroy_marked_extents(struct btrfs_root *root,
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index ca34237..1608492 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -64,7 +64,7 @@ struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
 int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info);
 void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr);
 void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr);
-int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root);
+void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root);
 void btrfs_mark_buffer_dirty(struct extent_buffer *buf);
 int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid);
 int btrfs_set_buffer_uptodate(struct extent_buffer *buf);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index c32a9bf..0daa1df 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1010,7 +1010,7 @@ static int convert_extent_item_v0(struct btrfs_trans_handle *trans,
 		return ret;
 	BUG_ON(ret);
 
-	ret = btrfs_extend_item(trans, root, path, new_size);
+	btrfs_extend_item(trans, root, path, new_size);
 
 	leaf = path->nodes[0];
 	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
@@ -1592,13 +1592,13 @@ out:
  * helper to add new inline back ref
  */
 static noinline_for_stack
-int setup_inline_extent_backref(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root,
-				struct btrfs_path *path,
-				struct btrfs_extent_inline_ref *iref,
-				u64 parent, u64 root_objectid,
-				u64 owner, u64 offset, int refs_to_add,
-				struct btrfs_delayed_extent_op *extent_op)
+void setup_inline_extent_backref(struct btrfs_trans_handle *trans,
+				 struct btrfs_root *root,
+				 struct btrfs_path *path,
+				 struct btrfs_extent_inline_ref *iref,
+				 u64 parent, u64 root_objectid,
+				 u64 owner, u64 offset, int refs_to_add,
+				 struct btrfs_delayed_extent_op *extent_op)
 {
 	struct extent_buffer *leaf;
 	struct btrfs_extent_item *ei;
@@ -1608,7 +1608,6 @@ int setup_inline_extent_backref(struct btrfs_trans_handle *trans,
 	u64 refs;
 	int size;
 	int type;
-	int ret;
 
 	leaf = path->nodes[0];
 	ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
@@ -1617,7 +1616,7 @@ int setup_inline_extent_backref(struct btrfs_trans_handle *trans,
 	type = extent_ref_type(parent, owner);
 	size = btrfs_extent_inline_ref_size(type);
 
-	ret = btrfs_extend_item(trans, root, path, size);
+	btrfs_extend_item(trans, root, path, size);
 
 	ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item);
 	refs = btrfs_extent_refs(leaf, ei);
@@ -1652,7 +1651,6 @@ int setup_inline_extent_backref(struct btrfs_trans_handle *trans,
 		btrfs_set_extent_inline_ref_offset(leaf, iref, root_objectid);
 	}
 	btrfs_mark_buffer_dirty(leaf);
-	return 0;
 }
 
 static int lookup_extent_backref(struct btrfs_trans_handle *trans,
@@ -1687,12 +1685,12 @@ static int lookup_extent_backref(struct btrfs_trans_handle *trans,
  * helper to update/remove inline back ref
  */
 static noinline_for_stack
-int update_inline_extent_backref(struct btrfs_trans_handle *trans,
-				 struct btrfs_root *root,
-				 struct btrfs_path *path,
-				 struct btrfs_extent_inline_ref *iref,
-				 int refs_to_mod,
-				 struct btrfs_delayed_extent_op *extent_op)
+void update_inline_extent_backref(struct btrfs_trans_handle *trans,
+				  struct btrfs_root *root,
+				  struct btrfs_path *path,
+				  struct btrfs_extent_inline_ref *iref,
+				  int refs_to_mod,
+				  struct btrfs_delayed_extent_op *extent_op)
 {
 	struct extent_buffer *leaf;
 	struct btrfs_extent_item *ei;
@@ -1703,7 +1701,6 @@ int update_inline_extent_backref(struct btrfs_trans_handle *trans,
 	u32 item_size;
 	int size;
 	int type;
-	int ret;
 	u64 refs;
 
 	leaf = path->nodes[0];
@@ -1745,10 +1742,9 @@ int update_inline_extent_backref(struct btrfs_trans_handle *trans,
 			memmove_extent_buffer(leaf, ptr, ptr + size,
 					      end - ptr - size);
 		item_size -= size;
-		ret = btrfs_truncate_item(trans, root, path, item_size, 1);
+		btrfs_truncate_item(trans, root, path, item_size, 1);
 	}
 	btrfs_mark_buffer_dirty(leaf);
-	return 0;
 }
 
 static noinline_for_stack
@@ -1768,13 +1764,13 @@ int insert_inline_extent_backref(struct btrfs_trans_handle *trans,
 					   root_objectid, owner, offset, 1);
 	if (ret == 0) {
 		BUG_ON(owner < BTRFS_FIRST_FREE_OBJECTID);
-		ret = update_inline_extent_backref(trans, root, path, iref,
-						   refs_to_add, extent_op);
+		update_inline_extent_backref(trans, root, path, iref,
+					     refs_to_add, extent_op);
 	} else if (ret == -ENOENT) {
-		ret = setup_inline_extent_backref(trans, root, path, iref,
-						  parent, root_objectid,
-						  owner, offset, refs_to_add,
-						  extent_op);
+		setup_inline_extent_backref(trans, root, path, iref, parent,
+					    root_objectid, owner, offset,
+					    refs_to_add, extent_op);
+		ret = 0;
 	}
 	return ret;
 }
@@ -1804,12 +1800,12 @@ static int remove_extent_backref(struct btrfs_trans_handle *trans,
 				 struct btrfs_extent_inline_ref *iref,
 				 int refs_to_drop, int is_data)
 {
-	int ret;
+	int ret = 0;
 
 	BUG_ON(!is_data && refs_to_drop != 1);
 	if (iref) {
-		ret = update_inline_extent_backref(trans, root, path, iref,
-						   -refs_to_drop, NULL);
+		update_inline_extent_backref(trans, root, path, iref,
+					     -refs_to_drop, NULL);
 	} else if (is_data) {
 		ret = remove_extent_data_ref(trans, root, path, refs_to_drop);
 	} else {
@@ -4734,7 +4730,7 @@ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache,
 	return ret;
 }
 
-int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
+void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
@@ -4764,7 +4760,6 @@ int btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans,
 	up_write(&fs_info->extent_commit_sem);
 
 	update_global_block_rsv(fs_info);
-	return 0;
 }
 
 static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end)
@@ -7189,7 +7184,7 @@ u64 btrfs_account_ro_block_groups_free_space(struct btrfs_space_info *sinfo)
 	return free_bytes;
 }
 
-int btrfs_set_block_group_rw(struct btrfs_root *root,
+void btrfs_set_block_group_rw(struct btrfs_root *root,
 			      struct btrfs_block_group_cache *cache)
 {
 	struct btrfs_space_info *sinfo = cache->space_info;
@@ -7205,7 +7200,6 @@ int btrfs_set_block_group_rw(struct btrfs_root *root,
 	cache->ro = 0;
 	spin_unlock(&cache->lock);
 	spin_unlock(&sinfo->lock);
-	return 0;
 }
 
 /*
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 8368baa..65216fa 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -142,6 +142,7 @@ static struct extent_state *alloc_extent_state(gfp_t mask)
 #endif
 	atomic_set(&state->refs, 1);
 	init_waitqueue_head(&state->wq);
+	trace_alloc_extent_state(state, mask, _RET_IP_);
 	return state;
 }
 
@@ -159,6 +160,7 @@ void free_extent_state(struct extent_state *state)
 		list_del(&state->leak_list);
 		spin_unlock_irqrestore(&leak_lock, flags);
 #endif
+		trace_free_extent_state(state, _RET_IP_);
 		kmem_cache_free(extent_state_cache, state);
 	}
 }
@@ -617,8 +619,8 @@ search_again:
 	goto again;
 }
 
-static int wait_on_state(struct extent_io_tree *tree,
-			 struct extent_state *state)
+static void wait_on_state(struct extent_io_tree *tree,
+			  struct extent_state *state)
 		__releases(tree->lock)
 		__acquires(tree->lock)
 {
@@ -628,7 +630,6 @@ static int wait_on_state(struct extent_io_tree *tree,
 	schedule();
 	spin_lock(&tree->lock);
 	finish_wait(&state->wq, &wait);
-	return 0;
 }
 
 /*
@@ -636,7 +637,7 @@ static int wait_on_state(struct extent_io_tree *tree,
  * The range [start, end] is inclusive.
  * The tree lock is taken by this function
  */
-int wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits)
+void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits)
 {
 	struct extent_state *state;
 	struct rb_node *node;
@@ -673,7 +674,6 @@ again:
 	}
 out:
 	spin_unlock(&tree->lock);
-	return 0;
 }
 
 static void set_state_bits(struct extent_io_tree *tree,
@@ -1359,9 +1359,9 @@ out:
 	return found;
 }
 
-static noinline int __unlock_for_delalloc(struct inode *inode,
-					  struct page *locked_page,
-					  u64 start, u64 end)
+static noinline void __unlock_for_delalloc(struct inode *inode,
+					   struct page *locked_page,
+					   u64 start, u64 end)
 {
 	int ret;
 	struct page *pages[16];
@@ -1371,7 +1371,7 @@ static noinline int __unlock_for_delalloc(struct inode *inode,
 	int i;
 
 	if (index == locked_page->index && end_index == index)
-		return 0;
+		return;
 
 	while (nr_pages > 0) {
 		ret = find_get_pages_contig(inode->i_mapping, index,
@@ -1386,7 +1386,6 @@ static noinline int __unlock_for_delalloc(struct inode *inode,
 		index += ret;
 		cond_resched();
 	}
-	return 0;
 }
 
 static noinline int lock_delalloc_pages(struct inode *inode,
@@ -1777,39 +1776,34 @@ int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,
  * helper function to set a given page up to date if all the
  * extents in the tree for that page are up to date
  */
-static int check_page_uptodate(struct extent_io_tree *tree,
-			       struct page *page)
+static void check_page_uptodate(struct extent_io_tree *tree, struct page *page)
 {
 	u64 start = (u64)page->index << PAGE_CACHE_SHIFT;
 	u64 end = start + PAGE_CACHE_SIZE - 1;
 	if (test_range_bit(tree, start, end, EXTENT_UPTODATE, 1, NULL))
 		SetPageUptodate(page);
-	return 0;
 }
 
 /*
  * helper function to unlock a page if all the extents in the tree
  * for that page are unlocked
  */
-static int check_page_locked(struct extent_io_tree *tree,
-			     struct page *page)
+static void check_page_locked(struct extent_io_tree *tree, struct page *page)
 {
 	u64 start = (u64)page->index << PAGE_CACHE_SHIFT;
 	u64 end = start + PAGE_CACHE_SIZE - 1;
 	if (!test_range_bit(tree, start, end, EXTENT_LOCKED, 0, NULL))
 		unlock_page(page);
-	return 0;
 }
 
 /*
  * helper function to end page writeback if all the extents
  * in the tree for that page are done with writeback
  */
-static int check_page_writeback(struct extent_io_tree *tree,
-			     struct page *page)
+static void check_page_writeback(struct extent_io_tree *tree,
+				 struct page *page)
 {
 	end_page_writeback(page);
-	return 0;
 }
 
 /*
@@ -3835,7 +3829,7 @@ void free_extent_buffer(struct extent_buffer *eb)
 	WARN_ON(1);
 }
 
-int clear_extent_buffer_dirty(struct extent_io_tree *tree,
+void clear_extent_buffer_dirty(struct extent_io_tree *tree,
 			      struct extent_buffer *eb)
 {
 	unsigned long i;
@@ -3867,7 +3861,6 @@ int clear_extent_buffer_dirty(struct extent_io_tree *tree,
 		ClearPageError(page);
 		unlock_page(page);
 	}
-	return 0;
 }
 
 int set_extent_buffer_dirty(struct extent_io_tree *tree,
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index cecc351..f7f321e 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -287,8 +287,8 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset,
 			   unsigned long src_offset, unsigned long len);
 void memset_extent_buffer(struct extent_buffer *eb, char c,
 			  unsigned long start, unsigned long len);
-int wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits);
-int clear_extent_buffer_dirty(struct extent_io_tree *tree,
+void wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits);
+void clear_extent_buffer_dirty(struct extent_io_tree *tree,
 			      struct extent_buffer *eb);
 int set_extent_buffer_dirty(struct extent_io_tree *tree,
 			     struct extent_buffer *eb);
diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c
index c7fb3a4..edb69b4 100644
--- a/fs/btrfs/file-item.c
+++ b/fs/btrfs/file-item.c
@@ -483,18 +483,17 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
  * This calls btrfs_truncate_item with the correct args based on the
  * overlap, and fixes up the key as required.
  */
-static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
-				      struct btrfs_root *root,
-				      struct btrfs_path *path,
-				      struct btrfs_key *key,
-				      u64 bytenr, u64 len)
+static noinline void truncate_one_csum(struct btrfs_trans_handle *trans,
+				       struct btrfs_root *root,
+				       struct btrfs_path *path,
+				       struct btrfs_key *key,
+				       u64 bytenr, u64 len)
 {
 	struct extent_buffer *leaf;
 	u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
 	u64 csum_end;
 	u64 end_byte = bytenr + len;
 	u32 blocksize_bits = root->fs_info->sb->s_blocksize_bits;
-	int ret;
 
 	leaf = path->nodes[0];
 	csum_end = btrfs_item_size_nr(leaf, path->slots[0]) / csum_size;
@@ -510,7 +509,7 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
 		 */
 		u32 new_size = (bytenr - key->offset) >> blocksize_bits;
 		new_size *= csum_size;
-		ret = btrfs_truncate_item(trans, root, path, new_size, 1);
+		btrfs_truncate_item(trans, root, path, new_size, 1);
 	} else if (key->offset >= bytenr && csum_end > end_byte &&
 		   end_byte > key->offset) {
 		/*
@@ -522,15 +521,13 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans,
 		u32 new_size = (csum_end - end_byte) >> blocksize_bits;
 		new_size *= csum_size;
 
-		ret = btrfs_truncate_item(trans, root, path, new_size, 0);
+		btrfs_truncate_item(trans, root, path, new_size, 0);
 
 		key->offset = end_byte;
-		ret = btrfs_set_item_key_safe(trans, root, path, key);
-		BUG_ON(ret);
+		btrfs_set_item_key_safe(trans, root, path, key);
 	} else {
 		BUG();
 	}
-	return 0;
 }
 
 /*
@@ -639,9 +636,7 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans,
 
 			key.offset = end_byte - 1;
 		} else {
-			ret = truncate_one_csum(trans, root, path,
-						&key, bytenr, len);
-			BUG_ON(ret);
+			truncate_one_csum(trans, root, path, &key, bytenr, len);
 			if (key.offset < bytenr)
 				break;
 		}
@@ -772,7 +767,7 @@ again:
 		if (diff != csum_size)
 			goto insert;
 
-		ret = btrfs_extend_item(trans, root, path, diff);
+		btrfs_extend_item(trans, root, path, diff);
 		goto csum;
 	}
 
diff --git a/fs/btrfs/inode-item.c b/fs/btrfs/inode-item.c
index baa74f3..c079240 100644
--- a/fs/btrfs/inode-item.c
+++ b/fs/btrfs/inode-item.c
@@ -128,7 +128,7 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans,
 	item_start = btrfs_item_ptr_offset(leaf, path->slots[0]);
 	memmove_extent_buffer(leaf, ptr, ptr + sub_item_len,
 			      item_size - (ptr + sub_item_len - item_start));
-	ret = btrfs_truncate_item(trans, root, path,
+	btrfs_truncate_item(trans, root, path,
 				  item_size - sub_item_len, 1);
 out:
 	btrfs_free_path(path);
@@ -165,7 +165,7 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans,
 			goto out;
 
 		old_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]);
-		ret = btrfs_extend_item(trans, root, path, ins_len);
+		btrfs_extend_item(trans, root, path, ins_len);
 		ref = btrfs_item_ptr(path->nodes[0], path->slots[0],
 				     struct btrfs_inode_ref);
 		ref = (struct btrfs_inode_ref *)((unsigned long)ref + old_size);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 5394bdb..dc800ee 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3164,8 +3164,8 @@ search_again:
 				}
 				size =
 				    btrfs_file_extent_calc_inline_size(size);
-				ret = btrfs_truncate_item(trans, root, path,
-							  size, 1);
+				btrfs_truncate_item(trans, root, path,
+						    size, 1);
 			} else if (root->ref_cows) {
 				inode_sub_bytes(inode, item_end + 1 -
 						found_key.offset);
@@ -3782,7 +3782,7 @@ static void inode_tree_del(struct inode *inode)
 	}
 }
 
-int btrfs_invalidate_inodes(struct btrfs_root *root)
+void btrfs_invalidate_inodes(struct btrfs_root *root)
 {
 	struct rb_node *node;
 	struct rb_node *prev;
@@ -3842,7 +3842,6 @@ again:
 		node = rb_next(node);
 	}
 	spin_unlock(&root->inode_lock);
-	return 0;
 }
 
 static int btrfs_init_locked_inode(struct inode *inode, void *p)
diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c
index 5e178d8..272f911 100644
--- a/fs/btrfs/locking.c
+++ b/fs/btrfs/locking.c
@@ -208,7 +208,7 @@ void btrfs_tree_read_unlock_blocking(struct extent_buffer *eb)
  * take a spinning write lock.  This will wait for both
  * blocking readers or writers
  */
-int btrfs_tree_lock(struct extent_buffer *eb)
+void btrfs_tree_lock(struct extent_buffer *eb)
 {
 again:
 	wait_event(eb->read_lock_wq, atomic_read(&eb->blocking_readers) == 0);
@@ -230,13 +230,12 @@ again:
 	atomic_inc(&eb->spinning_writers);
 	atomic_inc(&eb->write_locks);
 	eb->lock_owner = current->pid;
-	return 0;
 }
 
 /*
  * drop a spinning or a blocking write lock.
  */
-int btrfs_tree_unlock(struct extent_buffer *eb)
+void btrfs_tree_unlock(struct extent_buffer *eb)
 {
 	int blockers = atomic_read(&eb->blocking_writers);
 
@@ -255,7 +254,6 @@ int btrfs_tree_unlock(struct extent_buffer *eb)
 		atomic_dec(&eb->spinning_writers);
 		write_unlock(&eb->lock);
 	}
-	return 0;
 }
 
 void btrfs_assert_tree_locked(struct extent_buffer *eb)
diff --git a/fs/btrfs/locking.h b/fs/btrfs/locking.h
index 17247dd..ca52681 100644
--- a/fs/btrfs/locking.h
+++ b/fs/btrfs/locking.h
@@ -24,8 +24,8 @@
 #define BTRFS_WRITE_LOCK_BLOCKING 3
 #define BTRFS_READ_LOCK_BLOCKING 4
 
-int btrfs_tree_lock(struct extent_buffer *eb);
-int btrfs_tree_unlock(struct extent_buffer *eb);
+void btrfs_tree_lock(struct extent_buffer *eb);
+void btrfs_tree_unlock(struct extent_buffer *eb);
 int btrfs_try_spin_lock(struct extent_buffer *eb);
 
 void btrfs_tree_read_lock(struct extent_buffer *eb);
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index 2857f28..bbf6d0d 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -257,9 +257,9 @@ int btrfs_add_ordered_extent_compress(struct inode *inode, u64 file_offset,
  * when an ordered extent is finished.  If the list covers more than one
  * ordered extent, it is split across multiples.
  */
-int btrfs_add_ordered_sum(struct inode *inode,
-			  struct btrfs_ordered_extent *entry,
-			  struct btrfs_ordered_sum *sum)
+void btrfs_add_ordered_sum(struct inode *inode,
+			   struct btrfs_ordered_extent *entry,
+			   struct btrfs_ordered_sum *sum)
 {
 	struct btrfs_ordered_inode_tree *tree;
 
@@ -267,7 +267,6 @@ int btrfs_add_ordered_sum(struct inode *inode,
 	spin_lock(&tree->lock);
 	list_add_tail(&sum->list, &entry->list);
 	spin_unlock(&tree->lock);
-	return 0;
 }
 
 /*
@@ -392,7 +391,7 @@ out:
  * used to drop a reference on an ordered extent.  This will free
  * the extent if the last reference is dropped
  */
-int btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry)
+void btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry)
 {
 	struct list_head *cur;
 	struct btrfs_ordered_sum *sum;
@@ -408,7 +407,6 @@ int btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry)
 		}
 		kfree(entry);
 	}
-	return 0;
 }
 
 /*
@@ -416,8 +414,8 @@ int btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry)
  * and you must wake_up entry->wait.  You must hold the tree lock
  * while you call this function.
  */
-static int __btrfs_remove_ordered_extent(struct inode *inode,
-				struct btrfs_ordered_extent *entry)
+static void __btrfs_remove_ordered_extent(struct inode *inode,
+					  struct btrfs_ordered_extent *entry)
 {
 	struct btrfs_ordered_inode_tree *tree;
 	struct btrfs_root *root = BTRFS_I(inode)->root;
@@ -444,35 +442,30 @@ static int __btrfs_remove_ordered_extent(struct inode *inode,
 		list_del_init(&BTRFS_I(inode)->ordered_operations);
 	}
 	spin_unlock(&root->fs_info->ordered_extent_lock);
-
-	return 0;
 }
 
 /*
  * remove an ordered extent from the tree.  No references are dropped
  * but any waiters are woken.
  */
-int btrfs_remove_ordered_extent(struct inode *inode,
-				struct btrfs_ordered_extent *entry)
+void btrfs_remove_ordered_extent(struct inode *inode,
+				 struct btrfs_ordered_extent *entry)
 {
 	struct btrfs_ordered_inode_tree *tree;
-	int ret;
 
 	tree = &BTRFS_I(inode)->ordered_tree;
 	spin_lock(&tree->lock);
-	ret = __btrfs_remove_ordered_extent(inode, entry);
+	__btrfs_remove_ordered_extent(inode, entry);
 	spin_unlock(&tree->lock);
 	wake_up(&entry->wait);
-
-	return ret;
 }
 
 /*
  * wait for all the ordered extents in a root.  This is done when balancing
  * space between drives.
  */
-int btrfs_wait_ordered_extents(struct btrfs_root *root,
-			       int nocow_only, int delay_iput)
+void btrfs_wait_ordered_extents(struct btrfs_root *root,
+				int nocow_only, int delay_iput)
 {
 	struct list_head splice;
 	struct list_head *cur;
@@ -520,7 +513,6 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root,
 		spin_lock(&root->fs_info->ordered_extent_lock);
 	}
 	spin_unlock(&root->fs_info->ordered_extent_lock);
-	return 0;
 }
 
 /*
@@ -533,7 +525,7 @@ int btrfs_wait_ordered_extents(struct btrfs_root *root,
  * extra check to make sure the ordered operation list really is empty
  * before we return
  */
-int btrfs_run_ordered_operations(struct btrfs_root *root, int wait)
+void btrfs_run_ordered_operations(struct btrfs_root *root, int wait)
 {
 	struct btrfs_inode *btrfs_inode;
 	struct inode *inode;
@@ -581,8 +573,6 @@ again:
 
 	spin_unlock(&root->fs_info->ordered_extent_lock);
 	mutex_unlock(&root->fs_info->ordered_operations_mutex);
-
-	return 0;
 }
 
 /*
@@ -617,7 +607,7 @@ void btrfs_start_ordered_extent(struct inode *inode,
 /*
  * Used to wait on ordered extents across a large range of bytes.
  */
-int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
+void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len)
 {
 	u64 end;
 	u64 orig_end;
@@ -672,7 +662,6 @@ again:
 		schedule_timeout(1);
 		goto again;
 	}
-	return 0;
 }
 
 /*
@@ -956,9 +945,8 @@ out:
  * If trans is not null, we'll do a friendly check for a transaction that
  * is already flushing things and force the IO down ourselves.
  */
-int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root,
-				struct inode *inode)
+void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
+				 struct btrfs_root *root, struct inode *inode)
 {
 	u64 last_mod;
 
@@ -969,7 +957,7 @@ int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
 	 * commit, we can safely return without doing anything
 	 */
 	if (last_mod < root->fs_info->last_trans_committed)
-		return 0;
+		return;
 
 	/*
 	 * the transaction is already committing.  Just start the IO and
@@ -977,7 +965,7 @@ int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
 	 */
 	if (trans && root->fs_info->running_transaction->blocked) {
 		btrfs_wait_ordered_range(inode, 0, (u64)-1);
-		return 0;
+		return;
 	}
 
 	spin_lock(&root->fs_info->ordered_extent_lock);
@@ -986,6 +974,4 @@ int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
 			      &root->fs_info->ordered_operations);
 	}
 	spin_unlock(&root->fs_info->ordered_extent_lock);
-
-	return 0;
 }
diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h
index ff1f69a..c355ad4 100644
--- a/fs/btrfs/ordered-data.h
+++ b/fs/btrfs/ordered-data.h
@@ -138,8 +138,8 @@ btrfs_ordered_inode_tree_init(struct btrfs_ordered_inode_tree *t)
 	t->last = NULL;
 }
 
-int btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry);
-int btrfs_remove_ordered_extent(struct inode *inode,
+void btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry);
+void btrfs_remove_ordered_extent(struct inode *inode,
 				struct btrfs_ordered_extent *entry);
 int btrfs_dec_test_ordered_pending(struct inode *inode,
 				   struct btrfs_ordered_extent **cached,
@@ -154,14 +154,14 @@ int btrfs_add_ordered_extent_dio(struct inode *inode, u64 file_offset,
 int btrfs_add_ordered_extent_compress(struct inode *inode, u64 file_offset,
 				      u64 start, u64 len, u64 disk_len,
 				      int type, int compress_type);
-int btrfs_add_ordered_sum(struct inode *inode,
-			  struct btrfs_ordered_extent *entry,
-			  struct btrfs_ordered_sum *sum);
+void btrfs_add_ordered_sum(struct inode *inode,
+			   struct btrfs_ordered_extent *entry,
+			   struct btrfs_ordered_sum *sum);
 struct btrfs_ordered_extent *btrfs_lookup_ordered_extent(struct inode *inode,
 							 u64 file_offset);
 void btrfs_start_ordered_extent(struct inode *inode,
 				struct btrfs_ordered_extent *entry, int wait);
-int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len);
+void btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len);
 struct btrfs_ordered_extent *
 btrfs_lookup_first_ordered_extent(struct inode * inode, u64 file_offset);
 struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode,
@@ -170,10 +170,10 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode,
 int btrfs_ordered_update_i_size(struct inode *inode, u64 offset,
 				struct btrfs_ordered_extent *ordered);
 int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, u32 *sum);
-int btrfs_run_ordered_operations(struct btrfs_root *root, int wait);
-int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
-				struct btrfs_root *root,
-				struct inode *inode);
-int btrfs_wait_ordered_extents(struct btrfs_root *root,
-			       int nocow_only, int delay_iput);
+void btrfs_run_ordered_operations(struct btrfs_root *root, int wait);
+void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
+				 struct btrfs_root *root,
+				 struct inode *inode);
+void btrfs_wait_ordered_extents(struct btrfs_root *root,
+				int nocow_only, int delay_iput);
 #endif
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index abc0fbf..a2e8aa4 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -948,12 +948,12 @@ static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer)
 	return fail;
 }
 
-static int scrub_submit(struct scrub_dev *sdev)
+static void scrub_submit(struct scrub_dev *sdev)
 {
 	struct scrub_bio *sbio;
 
 	if (sdev->curr == -1)
-		return 0;
+		return;
 
 	sbio = sdev->bios[sdev->curr];
 	sbio->err = 0;
@@ -961,8 +961,6 @@ static int scrub_submit(struct scrub_dev *sdev)
 	atomic_inc(&sdev->in_flight);
 
 	btrfsic_submit_bio(READ, sbio->bio);
-
-	return 0;
 }
 
 static int scrub_page(struct scrub_dev *sdev, u64 logical, u64 len,
@@ -1008,9 +1006,7 @@ again:
 		sbio->bio = bio;
 	} else if (sbio->physical + sbio->count * PAGE_SIZE != physical ||
 		   sbio->logical + sbio->count * PAGE_SIZE != logical) {
-		ret = scrub_submit(sdev);
-		if (ret)
-			return ret;
+		scrub_submit(sdev);
 		goto again;
 	}
 	sbio->spag[sbio->count].flags = flags;
@@ -1025,9 +1021,7 @@ again:
 	ret = bio_add_page(sbio->bio, page, PAGE_SIZE, 0);
 	if (!ret) {
 		__free_page(page);
-		ret = scrub_submit(sdev);
-		if (ret)
-			return ret;
+		scrub_submit(sdev);
 		goto again;
 	}
 
@@ -1036,13 +1030,8 @@ again:
 		memcpy(sbio->spag[sbio->count].csum, csum, sdev->csum_size);
 	}
 	++sbio->count;
-	if (sbio->count == SCRUB_PAGES_PER_BIO || force) {
-		int ret;
-
-		ret = scrub_submit(sdev);
-		if (ret)
-			return ret;
-	}
+	if (sbio->count == SCRUB_PAGES_PER_BIO || force)
+		scrub_submit(sdev);
 
 	return 0;
 }
@@ -1656,7 +1645,7 @@ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end,
 	return ret;
 }
 
-int btrfs_scrub_pause(struct btrfs_root *root)
+void btrfs_scrub_pause(struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
@@ -1671,29 +1660,24 @@ int btrfs_scrub_pause(struct btrfs_root *root)
 		mutex_lock(&fs_info->scrub_lock);
 	}
 	mutex_unlock(&fs_info->scrub_lock);
-
-	return 0;
 }
 
-int btrfs_scrub_continue(struct btrfs_root *root)
+void btrfs_scrub_continue(struct btrfs_root *root)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 
 	atomic_dec(&fs_info->scrub_pause_req);
 	wake_up(&fs_info->scrub_pause_wait);
-	return 0;
 }
 
-int btrfs_scrub_pause_super(struct btrfs_root *root)
+void btrfs_scrub_pause_super(struct btrfs_root *root)
 {
 	down_write(&root->fs_info->scrub_super_lock);
-	return 0;
 }
 
-int btrfs_scrub_continue_super(struct btrfs_root *root)
+void btrfs_scrub_continue_super(struct btrfs_root *root)
 {
 	up_write(&root->fs_info->scrub_super_lock);
-	return 0;
 }
 
 int btrfs_scrub_cancel(struct btrfs_root *root)
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 9774e38..ae7963b 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -1408,9 +1408,7 @@ static int __init init_btrfs_fs(void)
 	if (err)
 		return err;
 
-	err = btrfs_init_compress();
-	if (err)
-		goto free_sysfs;
+	btrfs_init_compress();
 
 	err = btrfs_init_cachep();
 	if (err)
@@ -1451,7 +1449,6 @@ free_cachep:
 	btrfs_destroy_cachep();
 free_compress:
 	btrfs_exit_compress();
-free_sysfs:
 	btrfs_exit_sysfs();
 	return err;
 }
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 04b77e3..6e256d9 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -1214,8 +1214,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 		if (flush_on_commit || snap_pending) {
 			btrfs_start_delalloc_inodes(root, 1);
-			ret = btrfs_wait_ordered_extents(root, 0, 1);
-			BUG_ON(ret);
+			btrfs_wait_ordered_extents(root, 0, 1);
 		}
 
 		ret = btrfs_run_delayed_items(trans, root);
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 966cc74..37b52b8 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -212,14 +212,13 @@ int btrfs_pin_log_trans(struct btrfs_root *root)
  * indicate we're done making changes to the log tree
  * and wake up anyone waiting to do a sync
  */
-int btrfs_end_log_trans(struct btrfs_root *root)
+void btrfs_end_log_trans(struct btrfs_root *root)
 {
 	if (atomic_dec_and_test(&root->log_writers)) {
 		smp_mb();
 		if (waitqueue_active(&root->log_writer_wait))
 			wake_up(&root->log_writer_wait);
 	}
-	return 0;
 }
 
 
@@ -378,12 +377,11 @@ insert:
 		u32 found_size;
 		found_size = btrfs_item_size_nr(path->nodes[0],
 						path->slots[0]);
-		if (found_size > item_size) {
+		if (found_size > item_size)
 			btrfs_truncate_item(trans, root, path, item_size, 1);
-		} else if (found_size < item_size) {
-			ret = btrfs_extend_item(trans, root, path,
-						item_size - found_size);
-		}
+		else if (found_size < item_size)
+			btrfs_extend_item(trans, root, path,
+					  item_size - found_size);
 	} else if (ret) {
 		return ret;
 	}
@@ -1963,8 +1961,8 @@ static int wait_log_commit(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
-static int wait_for_writer(struct btrfs_trans_handle *trans,
-			   struct btrfs_root *root)
+static void wait_for_writer(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root)
 {
 	DEFINE_WAIT(wait);
 	while (root->fs_info->last_trans_log_full_commit !=
@@ -1978,7 +1976,6 @@ static int wait_for_writer(struct btrfs_trans_handle *trans,
 		mutex_lock(&root->log_mutex);
 		finish_wait(&root->log_writer_wait, &wait);
 	}
-	return 0;
 }
 
 /*
diff --git a/fs/btrfs/tree-log.h b/fs/btrfs/tree-log.h
index 2270ac5..862ac81 100644
--- a/fs/btrfs/tree-log.h
+++ b/fs/btrfs/tree-log.h
@@ -38,7 +38,7 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
 			       struct btrfs_root *root,
 			       const char *name, int name_len,
 			       struct inode *inode, u64 dirid);
-int btrfs_end_log_trans(struct btrfs_root *root);
+void btrfs_end_log_trans(struct btrfs_root *root);
 int btrfs_pin_log_trans(struct btrfs_root *root);
 int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
 		    struct btrfs_root *root, struct inode *inode,
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index ef41f28..43778b5 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -67,7 +67,7 @@ static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
 	kfree(fs_devices);
 }
 
-int btrfs_cleanup_fs_uuids(void)
+void btrfs_cleanup_fs_uuids(void)
 {
 	struct btrfs_fs_devices *fs_devices;
 
@@ -77,7 +77,6 @@ int btrfs_cleanup_fs_uuids(void)
 		list_del(&fs_devices->list);
 		free_fs_devices(fs_devices);
 	}
-	return 0;
 }
 
 static noinline struct btrfs_device *__find_device(struct list_head *head,
@@ -130,7 +129,7 @@ static void requeue_list(struct btrfs_pending_bios *pending_bios,
  * the list if the block device is congested.  This way, multiple devices
  * can make progress from a single worker thread.
  */
-static noinline int run_scheduled_bios(struct btrfs_device *device)
+static noinline void run_scheduled_bios(struct btrfs_device *device)
 {
 	struct bio *pending;
 	struct backing_dev_info *bdi;
@@ -316,7 +315,6 @@ loop_lock:
 
 done:
 	blk_finish_plug(&plug);
-	return 0;
 }
 
 static void pending_bios_fn(struct btrfs_work *work)
@@ -455,7 +453,7 @@ error:
 	return ERR_PTR(-ENOMEM);
 }
 
-int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices)
+void btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices)
 {
 	struct btrfs_device *device, *next;
 
@@ -503,7 +501,6 @@ again:
 	fs_devices->latest_trans = latest_transid;
 
 	mutex_unlock(&uuid_mutex);
-	return 0;
 }
 
 static void __free_device(struct work_struct *work)
@@ -3967,7 +3964,7 @@ struct async_sched {
  * This will add one bio to the pending list for a device and make sure
  * the work struct is scheduled.
  */
-static noinline int schedule_bio(struct btrfs_root *root,
+static noinline void schedule_bio(struct btrfs_root *root,
 				 struct btrfs_device *device,
 				 int rw, struct bio *bio)
 {
@@ -3979,7 +3976,7 @@ static noinline int schedule_bio(struct btrfs_root *root,
 		bio_get(bio);
 		btrfsic_submit_bio(rw, bio);
 		bio_put(bio);
-		return 0;
+		return;
 	}
 
 	/*
@@ -4013,7 +4009,6 @@ static noinline int schedule_bio(struct btrfs_root *root,
 	if (should_queue)
 		btrfs_queue_worker(&root->fs_info->submit_workers,
 				   &device->work);
-	return 0;
 }
 
 int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
@@ -4215,7 +4210,7 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 	return 0;
 }
 
-static int fill_device_from_item(struct extent_buffer *leaf,
+static void fill_device_from_item(struct extent_buffer *leaf,
 				 struct btrfs_dev_item *dev_item,
 				 struct btrfs_device *device)
 {
@@ -4232,8 +4227,6 @@ static int fill_device_from_item(struct extent_buffer *leaf,
 
 	ptr = (unsigned long)btrfs_device_uuid(dev_item);
 	read_extent_buffer(leaf, device->uuid, ptr, BTRFS_UUID_SIZE);
-
-	return 0;
 }
 
 static int open_seed_devices(struct btrfs_root *root, u8 *fsid)
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index 19ac950..bb6b03f 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -260,12 +260,12 @@ int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 			  struct btrfs_fs_devices **fs_devices_ret);
 int btrfs_close_devices(struct btrfs_fs_devices *fs_devices);
-int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices);
+void btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices);
 int btrfs_add_device(struct btrfs_trans_handle *trans,
 		     struct btrfs_root *root,
 		     struct btrfs_device *device);
 int btrfs_rm_device(struct btrfs_root *root, char *device_path);
-int btrfs_cleanup_fs_uuids(void);
+void btrfs_cleanup_fs_uuids(void);
 int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len);
 int btrfs_grow_device(struct btrfs_trans_handle *trans,
 		      struct btrfs_device *device, u64 new_size);
diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h
index 84f3001..91b91e8 100644
--- a/include/trace/events/btrfs.h
+++ b/include/trace/events/btrfs.h
@@ -6,6 +6,7 @@
 
 #include <linux/writeback.h>
 #include <linux/tracepoint.h>
+#include <trace/events/gfpflags.h>
 
 struct btrfs_root;
 struct btrfs_fs_info;
@@ -862,6 +863,49 @@ TRACE_EVENT(btrfs_setup_cluster,
 		  __entry->size, __entry->max_size, __entry->bitmap)
 );
 
+struct extent_state;
+TRACE_EVENT(alloc_extent_state,
+
+	TP_PROTO(struct extent_state *state, gfp_t mask, unsigned long IP),
+
+	TP_ARGS(state, mask, IP),
+
+	TP_STRUCT__entry(
+		__field(struct extent_state *, state)
+		__field(gfp_t, mask)
+		__field(unsigned long, ip)
+	),
+
+	TP_fast_assign(
+		__entry->state	= state,
+		__entry->mask	= mask,
+		__entry->ip	= IP
+	),
+
+	TP_printk("state=%p; mask = %s; caller = %pF", __entry->state,
+		  show_gfp_flags(__entry->mask), (void *)__entry->ip)
+);
+
+TRACE_EVENT(free_extent_state,
+
+	TP_PROTO(struct extent_state *state, unsigned long IP),
+
+	TP_ARGS(state, IP),
+
+	TP_STRUCT__entry(
+		__field(struct extent_state *, state)
+		__field(unsigned long, ip)
+	),
+
+	TP_fast_assign(
+		__entry->state	= state,
+		__entry->ip = IP
+	),
+
+	TP_printk(" state=%p; caller = %pF", __entry->state,
+		  (void *)__entry->ip)
+);
+
 #endif /* _TRACE_BTRFS_H */
 
 /* This part must be outside protection */
-- 
1.7.9

[patch 16/35] btrfs: drop gfp_t from lock_extent

[Jeff Mahoney]

 lock_extent and unlock_extent are always called with GFP_NOFS, drop the
 argument and use GFP_NOFS consistently.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/compression.c      |    6 +++---
 fs/btrfs/disk-io.c          |    2 +-
 fs/btrfs/extent_io.c        |   41 +++++++++++++++++++----------------------
 fs/btrfs/extent_io.h        |    9 ++++-----
 fs/btrfs/file.c             |    7 +++----
 fs/btrfs/free-space-cache.c |    2 +-
 fs/btrfs/inode.c            |   33 ++++++++++++++-------------------
 fs/btrfs/ioctl.c            |   17 ++++++++---------
 fs/btrfs/relocation.c       |   22 ++++++++++------------
 9 files changed, 63 insertions(+), 76 deletions(-)

diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index 0ed1ed2..cfd158d 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -496,7 +496,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 		 * sure they map to this compressed extent on disk.
 		 */
 		set_page_extent_mapped(page);
-		lock_extent(tree, last_offset, end, GFP_NOFS);
+		lock_extent(tree, last_offset, end);
 		read_lock(&em_tree->lock);
 		em = lookup_extent_mapping(em_tree, last_offset,
 					   PAGE_CACHE_SIZE);
@@ -506,7 +506,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 		    (last_offset + PAGE_CACHE_SIZE > extent_map_end(em)) ||
 		    (em->block_start >> 9) != cb->orig_bio->bi_sector) {
 			free_extent_map(em);
-			unlock_extent(tree, last_offset, end, GFP_NOFS);
+			unlock_extent(tree, last_offset, end);
 			unlock_page(page);
 			page_cache_release(page);
 			break;
@@ -534,7 +534,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 			nr_pages++;
 			page_cache_release(page);
 		} else {
-			unlock_extent(tree, last_offset, end, GFP_NOFS);
+			unlock_extent(tree, last_offset, end);
 			unlock_page(page);
 			page_cache_release(page);
 			break;
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 73ccadc..69ef456 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -332,7 +332,7 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 		return 0;
 
 	lock_extent_bits(io_tree, eb->start, eb->start + eb->len - 1,
-			 0, &cached_state, GFP_NOFS);
+			 0, &cached_state);
 	if (extent_buffer_uptodate(io_tree, eb, cached_state) &&
 	    btrfs_header_generation(eb) == parent_transid) {
 		ret = 0;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 65216fa..0112c02 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -1171,42 +1171,40 @@ static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
  * us if waiting is desired.
  */
 int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, struct extent_state **cached_state, gfp_t mask)
+		     int bits, struct extent_state **cached_state)
 {
 	int err;
 	u64 failed_start;
 	while (1) {
 		err = set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
 				     EXTENT_LOCKED, &failed_start,
-				     cached_state, mask);
-		if (err == -EEXIST && (mask & __GFP_WAIT)) {
+				     cached_state, GFP_NOFS);
+		if (err == -EEXIST) {
 			wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED);
 			start = failed_start;
-		} else {
+		} else
 			break;
-		}
 		WARN_ON(start > end);
 	}
 	return err;
 }
 
-int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask)
+int lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
 {
-	return lock_extent_bits(tree, start, end, 0, NULL, mask);
+	return lock_extent_bits(tree, start, end, 0, NULL);
 }
 
-int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end,
-		    gfp_t mask)
+int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
 {
 	int err;
 	u64 failed_start;
 
 	err = set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
-			     &failed_start, NULL, mask);
+			     &failed_start, NULL, GFP_NOFS);
 	if (err == -EEXIST) {
 		if (failed_start > start)
 			clear_extent_bit(tree, start, failed_start - 1,
-					 EXTENT_LOCKED, 1, 0, NULL, mask);
+					 EXTENT_LOCKED, 1, 0, NULL, GFP_NOFS);
 		return 0;
 	}
 	return 1;
@@ -1219,10 +1217,10 @@ int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end,
 				mask);
 }
 
-int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask)
+int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end)
 {
 	return clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, NULL,
-				mask);
+				GFP_NOFS);
 }
 
 /*
@@ -1518,8 +1516,7 @@ again:
 	BUG_ON(ret);
 
 	/* step three, lock the state bits for the whole range */
-	lock_extent_bits(tree, delalloc_start, delalloc_end,
-			 0, &cached_state, GFP_NOFS);
+	lock_extent_bits(tree, delalloc_start, delalloc_end, 0, &cached_state);
 
 	/* then test to make sure it is all still delalloc */
 	ret = test_range_bit(tree, delalloc_start, delalloc_end,
@@ -2557,11 +2554,11 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 
 	end = page_end;
 	while (1) {
-		lock_extent(tree, start, end, GFP_NOFS);
+		lock_extent(tree, start, end);
 		ordered = btrfs_lookup_ordered_extent(inode, start);
 		if (!ordered)
 			break;
-		unlock_extent(tree, start, end, GFP_NOFS);
+		unlock_extent(tree, start, end);
 		btrfs_start_ordered_extent(inode, ordered, 1);
 		btrfs_put_ordered_extent(ordered);
 	}
@@ -2598,7 +2595,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 				end - cur + 1, 0);
 		if (IS_ERR_OR_NULL(em)) {
 			SetPageError(page);
-			unlock_extent(tree, cur, end, GFP_NOFS);
+			unlock_extent(tree, cur, end);
 			break;
 		}
 		extent_offset = cur - em->start;
@@ -2650,7 +2647,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		if (test_range_bit(tree, cur, cur_end,
 				   EXTENT_UPTODATE, 1, NULL)) {
 			check_page_uptodate(tree, page);
-			unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS);
+			unlock_extent(tree, cur, cur + iosize - 1);
 			cur = cur + iosize;
 			pg_offset += iosize;
 			continue;
@@ -2660,7 +2657,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree,
 		 */
 		if (block_start == EXTENT_MAP_INLINE) {
 			SetPageError(page);
-			unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS);
+			unlock_extent(tree, cur, cur + iosize - 1);
 			cur = cur + iosize;
 			pg_offset += iosize;
 			continue;
@@ -3274,7 +3271,7 @@ int extent_invalidatepage(struct extent_io_tree *tree,
 	if (start > end)
 		return 0;
 
-	lock_extent_bits(tree, start, end, 0, &cached_state, GFP_NOFS);
+	lock_extent_bits(tree, start, end, 0, &cached_state);
 	wait_on_page_writeback(page);
 	clear_extent_bit(tree, start, end,
 			 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
@@ -3488,7 +3485,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	}
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, start, start + len, 0,
-			 &cached_state, GFP_NOFS);
+			 &cached_state);
 
 	em = get_extent_skip_holes(inode, start, last_for_get_extent,
 				   get_extent);
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index f7f321e..439e183 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -182,14 +182,13 @@ int try_release_extent_buffer(struct extent_io_tree *tree, struct page *page);
 int try_release_extent_state(struct extent_map_tree *map,
 			     struct extent_io_tree *tree, struct page *page,
 			     gfp_t mask);
-int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask);
+int lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
-		     int bits, struct extent_state **cached, gfp_t mask);
-int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask);
+		     int bits, struct extent_state **cached);
+int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end,
 			 struct extent_state **cached, gfp_t mask);
-int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end,
-		    gfp_t mask);
+int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end);
 int extent_read_full_page(struct extent_io_tree *tree, struct page *page,
 			  get_extent_t *get_extent, int mirror_num);
 int __init extent_io_init(void);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index e8d06b6..0eb80cc 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1105,8 +1105,7 @@ again:
 	if (start_pos < inode->i_size) {
 		struct btrfs_ordered_extent *ordered;
 		lock_extent_bits(&BTRFS_I(inode)->io_tree,
-				 start_pos, last_pos - 1, 0, &cached_state,
-				 GFP_NOFS);
+				 start_pos, last_pos - 1, 0, &cached_state);
 		ordered = btrfs_lookup_first_ordered_extent(inode,
 							    last_pos - 1);
 		if (ordered &&
@@ -1638,7 +1637,7 @@ static long btrfs_fallocate(struct file *file, int mode,
 		 * transaction
 		 */
 		lock_extent_bits(&BTRFS_I(inode)->io_tree, alloc_start,
-				 locked_end, 0, &cached_state, GFP_NOFS);
+				 locked_end, 0, &cached_state);
 		ordered = btrfs_lookup_first_ordered_extent(inode,
 							    alloc_end - 1);
 		if (ordered &&
@@ -1737,7 +1736,7 @@ static int find_desired_extent(struct inode *inode, loff_t *offset, int origin)
 		return -ENXIO;
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend, 0,
-			 &cached_state, GFP_NOFS);
+			 &cached_state);
 
 	/*
 	 * Delalloc is such a pain.  If we have a hole and we have pending
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 710ea38..ecbcc87 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -869,7 +869,7 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
 	io_ctl_prepare_pages(&io_ctl, inode, 0);
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1,
-			 0, &cached_state, GFP_NOFS);
+			 0, &cached_state);
 
 	node = rb_first(&ctl->free_space_offset);
 	if (!node && cluster) {
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index dc800ee..d16bf3f 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -597,7 +597,7 @@ retry:
 
 			lock_extent(io_tree, async_extent->start,
 					 async_extent->start +
-					 async_extent->ram_size - 1, GFP_NOFS);
+					 async_extent->ram_size - 1);
 
 			/* allocate blocks */
 			ret = cow_file_range(inode, async_cow->locked_page,
@@ -625,8 +625,7 @@ retry:
 		}
 
 		lock_extent(io_tree, async_extent->start,
-			    async_extent->start + async_extent->ram_size - 1,
-			    GFP_NOFS);
+			    async_extent->start + async_extent->ram_size - 1);
 
 		trans = btrfs_join_transaction(root);
 		BUG_ON(IS_ERR(trans));
@@ -649,7 +648,7 @@ retry:
 			async_extent->pages = NULL;
 			unlock_extent(io_tree, async_extent->start,
 				      async_extent->start +
-				      async_extent->ram_size - 1, GFP_NOFS);
+				      async_extent->ram_size - 1);
 			goto retry;
 		}
 
@@ -1574,7 +1573,7 @@ again:
 	page_end = page_offset(page) + PAGE_CACHE_SIZE - 1;
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end, 0,
-			 &cached_state, GFP_NOFS);
+			 &cached_state);
 
 	/* already ordered? We're done */
 	if (PagePrivate2(page))
@@ -1765,7 +1764,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 
 	lock_extent_bits(io_tree, ordered_extent->file_offset,
 			 ordered_extent->file_offset + ordered_extent->len - 1,
-			 0, &cached_state, GFP_NOFS);
+			 0, &cached_state);
 
 	if (nolock)
 		trans = btrfs_join_transaction_nolock(root);
@@ -3285,8 +3284,7 @@ again:
 	}
 	wait_on_page_writeback(page);
 
-	lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state,
-			 GFP_NOFS);
+	lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state);
 	set_page_extent_mapped(page);
 
 	ordered = btrfs_lookup_ordered_extent(inode, page_start);
@@ -3362,7 +3360,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size)
 		btrfs_wait_ordered_range(inode, hole_start,
 					 block_end - hole_start);
 		lock_extent_bits(io_tree, hole_start, block_end - 1, 0,
-				 &cached_state, GFP_NOFS);
+				 &cached_state);
 		ordered = btrfs_lookup_ordered_extent(inode, hole_start);
 		if (!ordered)
 			break;
@@ -5604,7 +5602,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock,
 		free_extent_map(em);
 		/* DIO will do one hole at a time, so just unlock a sector */
 		unlock_extent(&BTRFS_I(inode)->io_tree, start,
-			      start + root->sectorsize - 1, GFP_NOFS);
+			      start + root->sectorsize - 1);
 		return 0;
 	}
 
@@ -5745,7 +5743,7 @@ static void btrfs_endio_direct_read(struct bio *bio, int err)
 	} while (bvec <= bvec_end);
 
 	unlock_extent(&BTRFS_I(inode)->io_tree, dip->logical_offset,
-		      dip->logical_offset + dip->bytes - 1, GFP_NOFS);
+		      dip->logical_offset + dip->bytes - 1);
 	bio->bi_private = dip->private;
 
 	kfree(dip->csums);
@@ -5796,7 +5794,7 @@ again:
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, ordered->file_offset,
 			 ordered->file_offset + ordered->len - 1, 0,
-			 &cached_state, GFP_NOFS);
+			 &cached_state);
 
 	if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags)) {
 		ret = btrfs_mark_extent_written(trans, inode,
@@ -6211,7 +6209,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 
 	while (1) {
 		lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend,
-				 0, &cached_state, GFP_NOFS);
+				 0, &cached_state);
 		/*
 		 * We're concerned with the entire range that we're going to be
 		 * doing DIO to, so we need to make sure theres no ordered
@@ -6365,8 +6363,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 		btrfs_releasepage(page, GFP_NOFS);
 		return;
 	}
-	lock_extent_bits(tree, page_start, page_end, 0, &cached_state,
-			 GFP_NOFS);
+	lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
 	ordered = btrfs_lookup_ordered_extent(page->mapping->host,
 					   page_offset(page));
 	if (ordered) {
@@ -6388,8 +6385,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 		}
 		btrfs_put_ordered_extent(ordered);
 		cached_state = NULL;
-		lock_extent_bits(tree, page_start, page_end, 0, &cached_state,
-				 GFP_NOFS);
+		lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
 	}
 	clear_extent_bit(tree, page_start, page_end,
 		 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
@@ -6464,8 +6460,7 @@ again:
 	}
 	wait_on_page_writeback(page);
 
-	lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state,
-			 GFP_NOFS);
+	lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state);
 	set_page_extent_mapped(page);
 
 	/*
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index d8b5471..4e51494 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -797,9 +797,9 @@ static int should_defrag_range(struct inode *inode, u64 start, u64 len,
 
 	if (!em) {
 		/* get the big lock and read metadata off disk */
-		lock_extent(io_tree, start, start + len - 1, GFP_NOFS);
+		lock_extent(io_tree, start, start + len - 1);
 		em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
-		unlock_extent(io_tree, start, start + len - 1, GFP_NOFS);
+		unlock_extent(io_tree, start, start + len - 1);
 
 		if (IS_ERR(em))
 			return 0;
@@ -887,10 +887,10 @@ again:
 		page_start = page_offset(page);
 		page_end = page_start + PAGE_CACHE_SIZE - 1;
 		while (1) {
-			lock_extent(tree, page_start, page_end, GFP_NOFS);
+			lock_extent(tree, page_start, page_end);
 			ordered = btrfs_lookup_ordered_extent(inode,
 							      page_start);
-			unlock_extent(tree, page_start, page_end, GFP_NOFS);
+			unlock_extent(tree, page_start, page_end);
 			if (!ordered)
 				break;
 
@@ -946,8 +946,7 @@ again:
 	page_end = page_offset(pages[i_done - 1]) + PAGE_CACHE_SIZE;
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree,
-			 page_start, page_end - 1, 0, &cached_state,
-			 GFP_NOFS);
+			 page_start, page_end - 1, 0, &cached_state);
 	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
 			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
 			  EXTENT_DO_ACCOUNTING, 0, 0, &cached_state,
@@ -2326,13 +2325,13 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 	   another, and lock file content */
 	while (1) {
 		struct btrfs_ordered_extent *ordered;
-		lock_extent(&BTRFS_I(src)->io_tree, off, off+len, GFP_NOFS);
+		lock_extent(&BTRFS_I(src)->io_tree, off, off+len);
 		ordered = btrfs_lookup_first_ordered_extent(src, off+len);
 		if (!ordered &&
 		    !test_range_bit(&BTRFS_I(src)->io_tree, off, off+len,
 				   EXTENT_DELALLOC, 0, NULL))
 			break;
-		unlock_extent(&BTRFS_I(src)->io_tree, off, off+len, GFP_NOFS);
+		unlock_extent(&BTRFS_I(src)->io_tree, off, off+len);
 		if (ordered)
 			btrfs_put_ordered_extent(ordered);
 		btrfs_wait_ordered_range(src, off, len);
@@ -2551,7 +2550,7 @@ next:
 	ret = 0;
 out:
 	btrfs_release_path(path);
-	unlock_extent(&BTRFS_I(src)->io_tree, off, off+len, GFP_NOFS);
+	unlock_extent(&BTRFS_I(src)->io_tree, off, off+len);
 out_unlock:
 	mutex_unlock(&src->i_mutex);
 	mutex_unlock(&inode->i_mutex);
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 974b0df..5e89a5e 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -1604,15 +1604,14 @@ int replace_file_extents(struct btrfs_trans_handle *trans,
 				WARN_ON(!IS_ALIGNED(end, root->sectorsize));
 				end--;
 				ret = try_lock_extent(&BTRFS_I(inode)->io_tree,
-						      key.offset, end,
-						      GFP_NOFS);
+						      key.offset, end);
 				if (!ret)
 					continue;
 
 				btrfs_drop_extent_cache(inode, key.offset, end,
 							1);
 				unlock_extent(&BTRFS_I(inode)->io_tree,
-					      key.offset, end, GFP_NOFS);
+					      key.offset, end);
 			}
 		}
 
@@ -1983,9 +1982,9 @@ static int invalidate_extent_cache(struct btrfs_root *root,
 		}
 
 		/* the lock_extent waits for readpage to complete */
-		lock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+		lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		btrfs_drop_extent_cache(inode, start, end, 1);
-		unlock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+		unlock_extent(&BTRFS_I(inode)->io_tree, start, end);
 	}
 	return 0;
 }
@@ -2889,12 +2888,12 @@ int prealloc_file_extent_cluster(struct inode *inode,
 		else
 			end = cluster->end - offset;
 
-		lock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+		lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		num_bytes = end + 1 - start;
 		ret = btrfs_prealloc_file_range(inode, 0, start,
 						num_bytes, num_bytes,
 						end + 1, &alloc_hint);
-		unlock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+		unlock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		if (ret)
 			break;
 		nr++;
@@ -2926,7 +2925,7 @@ int setup_extent_mapping(struct inode *inode, u64 start, u64 end,
 	em->bdev = root->fs_info->fs_devices->latest_bdev;
 	set_bit(EXTENT_FLAG_PINNED, &em->flags);
 
-	lock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+	lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 	while (1) {
 		write_lock(&em_tree->lock);
 		ret = add_extent_mapping(em_tree, em);
@@ -2937,7 +2936,7 @@ int setup_extent_mapping(struct inode *inode, u64 start, u64 end,
 		}
 		btrfs_drop_extent_cache(inode, start, end, 0);
 	}
-	unlock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
+	unlock_extent(&BTRFS_I(inode)->io_tree, start, end);
 	return ret;
 }
 
@@ -3017,8 +3016,7 @@ static int relocate_file_extent_cluster(struct inode *inode,
 		page_start = (u64)page->index << PAGE_CACHE_SHIFT;
 		page_end = page_start + PAGE_CACHE_SIZE - 1;
 
-		lock_extent(&BTRFS_I(inode)->io_tree,
-			    page_start, page_end, GFP_NOFS);
+		lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end);
 
 		set_page_extent_mapped(page);
 
@@ -3034,7 +3032,7 @@ static int relocate_file_extent_cluster(struct inode *inode,
 		set_page_dirty(page);
 
 		unlock_extent(&BTRFS_I(inode)->io_tree,
-			      page_start, page_end, GFP_NOFS);
+			      page_start, page_end);
 		unlock_page(page);
 		page_cache_release(page);
 
-- 
1.7.9

[patch 17/35] btrfs: split extent_state ops

[Jeff Mahoney]

 set_extent_bit can do exclusive locking but only when called by lock_extent*,

 Drop the exclusive bits argument except when called by lock_extent.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/extent_io.c |   36 +++++++++++++++++++++++-------------
 fs/btrfs/extent_io.h |    2 +-
 fs/btrfs/inode.c     |    2 +-
 3 files changed, 25 insertions(+), 15 deletions(-)

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 0112c02..ffa7cc3 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -721,9 +721,10 @@ static void uncache_state(struct extent_state **cached_ptr)
  * [start, end] is inclusive This takes the tree lock.
  */
 
-int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		   int bits, int exclusive_bits, u64 *failed_start,
-		   struct extent_state **cached_state, gfp_t mask)
+static int __must_check
+__set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
+		 int bits, int exclusive_bits, u64 *failed_start,
+		 struct extent_state **cached_state, gfp_t mask)
 {
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
@@ -917,6 +918,15 @@ search_again:
 	goto again;
 }
 
+int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits,
+		   u64 *failed_start, struct extent_state **cached_state,
+		   gfp_t mask)
+{
+	return __set_extent_bit(tree, start, end, bits, 0, failed_start,
+				cached_state, mask);
+}
+
+
 /**
  * convert_extent - convert all bits in a given range from one bit to another
  * @tree:	the io tree to search
@@ -1111,14 +1121,14 @@ search_again:
 int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 		     gfp_t mask)
 {
-	return set_extent_bit(tree, start, end, EXTENT_DIRTY, 0, NULL,
+	return set_extent_bit(tree, start, end, EXTENT_DIRTY, NULL,
 			      NULL, mask);
 }
 
 int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 		    int bits, gfp_t mask)
 {
-	return set_extent_bit(tree, start, end, bits, 0, NULL,
+	return set_extent_bit(tree, start, end, bits, NULL,
 			      NULL, mask);
 }
 
@@ -1133,7 +1143,7 @@ int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end,
 {
 	return set_extent_bit(tree, start, end,
 			      EXTENT_DELALLOC | EXTENT_UPTODATE,
-			      0, NULL, cached_state, mask);
+			      NULL, cached_state, mask);
 }
 
 int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
@@ -1147,7 +1157,7 @@ int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
 int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end,
 		     gfp_t mask)
 {
-	return set_extent_bit(tree, start, end, EXTENT_NEW, 0, NULL,
+	return set_extent_bit(tree, start, end, EXTENT_NEW, NULL,
 			      NULL, mask);
 }
 
@@ -1155,7 +1165,7 @@ int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask)
 {
 	return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0,
-			      NULL, cached_state, mask);
+			      cached_state, mask);
 }
 
 static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
@@ -1176,9 +1186,9 @@ int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 	int err;
 	u64 failed_start;
 	while (1) {
-		err = set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
-				     EXTENT_LOCKED, &failed_start,
-				     cached_state, GFP_NOFS);
+		err = __set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
+				       EXTENT_LOCKED, &failed_start,
+				       cached_state, GFP_NOFS);
 		if (err == -EEXIST) {
 			wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED);
 			start = failed_start;
@@ -1199,8 +1209,8 @@ int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
 	int err;
 	u64 failed_start;
 
-	err = set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
-			     &failed_start, NULL, GFP_NOFS);
+	err = __set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
+			       &failed_start, NULL, GFP_NOFS);
 	if (err == -EEXIST) {
 		if (failed_start > start)
 			clear_extent_bit(tree, start, failed_start - 1,
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 439e183..3a171c2 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -209,7 +209,7 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
 int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
 		    int bits, gfp_t mask);
 int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-		   int bits, int exclusive_bits, u64 *failed_start,
+		   int bits, u64 *failed_start,
 		   struct extent_state **cached_state, gfp_t mask);
 int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
 			struct extent_state **cached_state, gfp_t mask);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index d16bf3f..593a2c3 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -6233,7 +6233,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 	if (writing) {
 		write_bits = EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING;
 		ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend,
-				     EXTENT_DELALLOC, 0, NULL, &cached_state,
+				     EXTENT_DELALLOC, NULL, &cached_state,
 				     GFP_NOFS);
 		if (ret) {
 			clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
-- 
1.7.9

Re: [patch 17/35] btrfs: split extent_state ops

[Liu Bo]

On 03/22/2012 09:11 AM, Jeff Mahoney wrote:
>  set_extent_bit can do exclusive locking but only when called by lock_extent*,
> 
>  Drop the exclusive bits argument except when called by lock_extent.
> 

Hi Jeff,

I have to say this is not a good cleanup, any benefits?

thanks,
liubo

> Signed-off-by: Jeff Mahoney <jeffm@suse.com>
> ---
>  fs/btrfs/extent_io.c |   36 +++++++++++++++++++++++-------------
>  fs/btrfs/extent_io.h |    2 +-
>  fs/btrfs/inode.c     |    2 +-
>  3 files changed, 25 insertions(+), 15 deletions(-)
> 
> diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
> index 0112c02..ffa7cc3 100644
> --- a/fs/btrfs/extent_io.c
> +++ b/fs/btrfs/extent_io.c
> @@ -721,9 +721,10 @@ static void uncache_state(struct extent_state **cached_ptr)
>   * [start, end] is inclusive This takes the tree lock.
>   */
>  
> -int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
> -		   int bits, int exclusive_bits, u64 *failed_start,
> -		   struct extent_state **cached_state, gfp_t mask)
> +static int __must_check
> +__set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
> +		 int bits, int exclusive_bits, u64 *failed_start,
> +		 struct extent_state **cached_state, gfp_t mask)
>  {
>  	struct extent_state *state;
>  	struct extent_state *prealloc = NULL;
> @@ -917,6 +918,15 @@ search_again:
>  	goto again;
>  }
>  
> +int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits,
> +		   u64 *failed_start, struct extent_state **cached_state,
> +		   gfp_t mask)
> +{
> +	return __set_extent_bit(tree, start, end, bits, 0, failed_start,
> +				cached_state, mask);
> +}
> +
> +
>  /**
>   * convert_extent - convert all bits in a given range from one bit to another
>   * @tree:	the io tree to search
> @@ -1111,14 +1121,14 @@ search_again:
>  int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
>  		     gfp_t mask)
>  {
> -	return set_extent_bit(tree, start, end, EXTENT_DIRTY, 0, NULL,
> +	return set_extent_bit(tree, start, end, EXTENT_DIRTY, NULL,
>  			      NULL, mask);
>  }
>  
>  int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
>  		    int bits, gfp_t mask)
>  {
> -	return set_extent_bit(tree, start, end, bits, 0, NULL,
> +	return set_extent_bit(tree, start, end, bits, NULL,
>  			      NULL, mask);
>  }
>  
> @@ -1133,7 +1143,7 @@ int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end,
>  {
>  	return set_extent_bit(tree, start, end,
>  			      EXTENT_DELALLOC | EXTENT_UPTODATE,
> -			      0, NULL, cached_state, mask);
> +			      NULL, cached_state, mask);
>  }
>  
>  int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
> @@ -1147,7 +1157,7 @@ int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
>  int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end,
>  		     gfp_t mask)
>  {
> -	return set_extent_bit(tree, start, end, EXTENT_NEW, 0, NULL,
> +	return set_extent_bit(tree, start, end, EXTENT_NEW, NULL,
>  			      NULL, mask);
>  }
>  
> @@ -1155,7 +1165,7 @@ int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
>  			struct extent_state **cached_state, gfp_t mask)
>  {
>  	return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0,
> -			      NULL, cached_state, mask);
> +			      cached_state, mask);
>  }
>  
>  static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
> @@ -1176,9 +1186,9 @@ int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
>  	int err;
>  	u64 failed_start;
>  	while (1) {
> -		err = set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
> -				     EXTENT_LOCKED, &failed_start,
> -				     cached_state, GFP_NOFS);
> +		err = __set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
> +				       EXTENT_LOCKED, &failed_start,
> +				       cached_state, GFP_NOFS);
>  		if (err == -EEXIST) {
>  			wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED);
>  			start = failed_start;
> @@ -1199,8 +1209,8 @@ int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end)
>  	int err;
>  	u64 failed_start;
>  
> -	err = set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
> -			     &failed_start, NULL, GFP_NOFS);
> +	err = __set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED,
> +			       &failed_start, NULL, GFP_NOFS);
>  	if (err == -EEXIST) {
>  		if (failed_start > start)
>  			clear_extent_bit(tree, start, failed_start - 1,
> diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
> index 439e183..3a171c2 100644
> --- a/fs/btrfs/extent_io.h
> +++ b/fs/btrfs/extent_io.h
> @@ -209,7 +209,7 @@ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
>  int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
>  		    int bits, gfp_t mask);
>  int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
> -		   int bits, int exclusive_bits, u64 *failed_start,
> +		   int bits, u64 *failed_start,
>  		   struct extent_state **cached_state, gfp_t mask);
>  int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
>  			struct extent_state **cached_state, gfp_t mask);
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index d16bf3f..593a2c3 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -6233,7 +6233,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
>  	if (writing) {
>  		write_bits = EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING;
>  		ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend,
> -				     EXTENT_DELALLOC, 0, NULL, &cached_state,
> +				     EXTENT_DELALLOC, NULL, &cached_state,
>  				     GFP_NOFS);
>  		if (ret) {
>  			clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,

Re: [patch 17/35] btrfs: split extent_state ops

[Jeff Mahoney]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03/21/2012 09:56 PM, Liu Bo wrote:
> On 03/22/2012 09:11 AM, Jeff Mahoney wrote:
>> set_extent_bit can do exclusive locking but only when called by
>> lock_extent*,
>> 
>> Drop the exclusive bits argument except when called by
>> lock_extent.
>> 
> 
> Hi Jeff,
> 
> I have to say this is not a good cleanup, any benefits?

Yeah, it eliminates an argument that's only used by a single caller.

- -Jeff

>> Signed-off-by: Jeff Mahoney <jeffm@suse.com> --- 
>> fs/btrfs/extent_io.c |   36 +++++++++++++++++++++++------------- 
>> fs/btrfs/extent_io.h |    2 +- fs/btrfs/inode.c     |    2 +- 3
>> files changed, 25 insertions(+), 15 deletions(-)
>> 
>> diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index
>> 0112c02..ffa7cc3 100644 --- a/fs/btrfs/extent_io.c +++
>> b/fs/btrfs/extent_io.c @@ -721,9 +721,10 @@ static void
>> uncache_state(struct extent_state **cached_ptr) * [start, end] is
>> inclusive This takes the tree lock. */
>> 
>> -int set_extent_bit(struct extent_io_tree *tree, u64 start, u64
>> end, -		   int bits, int exclusive_bits, u64 *failed_start, -
>> struct extent_state **cached_state, gfp_t mask) +static int
>> __must_check +__set_extent_bit(struct extent_io_tree *tree, u64
>> start, u64 end, +		 int bits, int exclusive_bits, u64
>> *failed_start, +		 struct extent_state **cached_state, gfp_t
>> mask) { struct extent_state *state; struct extent_state *prealloc
>> = NULL; @@ -917,6 +918,15 @@ search_again: goto again; }
>> 
>> +int set_extent_bit(struct extent_io_tree *tree, u64 start, u64
>> end, int bits, +		   u64 *failed_start, struct extent_state
>> **cached_state, +		   gfp_t mask) +{ +	return
>> __set_extent_bit(tree, start, end, bits, 0, failed_start, +
>> cached_state, mask); +} + + /** * convert_extent - convert all
>> bits in a given range from one bit to another * @tree:	the io
>> tree to search @@ -1111,14 +1121,14 @@ search_again: int
>> set_extent_dirty(struct extent_io_tree *tree, u64 start, u64
>> end, gfp_t mask) { -	return set_extent_bit(tree, start, end,
>> EXTENT_DIRTY, 0, NULL, +	return set_extent_bit(tree, start, end,
>> EXTENT_DIRTY, NULL, NULL, mask); }
>> 
>> int set_extent_bits(struct extent_io_tree *tree, u64 start, u64
>> end, int bits, gfp_t mask) { -	return set_extent_bit(tree, start,
>> end, bits, 0, NULL, +	return set_extent_bit(tree, start, end,
>> bits, NULL, NULL, mask); }
>> 
>> @@ -1133,7 +1143,7 @@ int set_extent_delalloc(struct
>> extent_io_tree *tree, u64 start, u64 end, { return
>> set_extent_bit(tree, start, end, EXTENT_DELALLOC |
>> EXTENT_UPTODATE, -			      0, NULL, cached_state, mask); +
>> NULL, cached_state, mask); }
>> 
>> int clear_extent_dirty(struct extent_io_tree *tree, u64 start,
>> u64 end, @@ -1147,7 +1157,7 @@ int clear_extent_dirty(struct
>> extent_io_tree *tree, u64 start, u64 end, int
>> set_extent_new(struct extent_io_tree *tree, u64 start, u64 end, 
>> gfp_t mask) { -	return set_extent_bit(tree, start, end,
>> EXTENT_NEW, 0, NULL, +	return set_extent_bit(tree, start, end,
>> EXTENT_NEW, NULL, NULL, mask); }
>> 
>> @@ -1155,7 +1165,7 @@ int set_extent_uptodate(struct
>> extent_io_tree *tree, u64 start, u64 end, struct extent_state
>> **cached_state, gfp_t mask) { return set_extent_bit(tree, start,
>> end, EXTENT_UPTODATE, 0, -			      NULL, cached_state, mask); +
>> cached_state, mask); }
>> 
>> static int clear_extent_uptodate(struct extent_io_tree *tree, u64
>> start, @@ -1176,9 +1186,9 @@ int lock_extent_bits(struct
>> extent_io_tree *tree, u64 start, u64 end, int err; u64
>> failed_start; while (1) { -		err = set_extent_bit(tree, start,
>> end, EXTENT_LOCKED | bits, -				     EXTENT_LOCKED,
>> &failed_start, -				     cached_state, GFP_NOFS); +		err =
>> __set_extent_bit(tree, start, end, EXTENT_LOCKED | bits, +
>> EXTENT_LOCKED, &failed_start, +				       cached_state,
>> GFP_NOFS); if (err == -EEXIST) { wait_extent_bit(tree,
>> failed_start, end, EXTENT_LOCKED); start = failed_start; @@
>> -1199,8 +1209,8 @@ int try_lock_extent(struct extent_io_tree
>> *tree, u64 start, u64 end) int err; u64 failed_start;
>> 
>> -	err = set_extent_bit(tree, start, end, EXTENT_LOCKED,
>> EXTENT_LOCKED, -			     &failed_start, NULL, GFP_NOFS); +	err =
>> __set_extent_bit(tree, start, end, EXTENT_LOCKED, EXTENT_LOCKED, 
>> +			       &failed_start, NULL, GFP_NOFS); if (err == -EEXIST) { 
>> if (failed_start > start) clear_extent_bit(tree, start,
>> failed_start - 1, diff --git a/fs/btrfs/extent_io.h
>> b/fs/btrfs/extent_io.h index 439e183..3a171c2 100644 ---
>> a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -209,7
>> +209,7 @@ int clear_extent_bit(struct extent_io_tree *tree, u64
>> start, u64 end, int set_extent_bits(struct extent_io_tree *tree,
>> u64 start, u64 end, int bits, gfp_t mask); int
>> set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, -
>> int bits, int exclusive_bits, u64 *failed_start, +		   int bits,
>> u64 *failed_start, struct extent_state **cached_state, gfp_t
>> mask); int set_extent_uptodate(struct extent_io_tree *tree, u64
>> start, u64 end, struct extent_state **cached_state, gfp_t mask); 
>> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index
>> d16bf3f..593a2c3 100644 --- a/fs/btrfs/inode.c +++
>> b/fs/btrfs/inode.c @@ -6233,7 +6233,7 @@ static ssize_t
>> btrfs_direct_IO(int rw, struct kiocb *iocb, if (writing) { 
>> write_bits = EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING; ret =
>> set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend, -
>> EXTENT_DELALLOC, 0, NULL, &cached_state, +
>> EXTENT_DELALLOC, NULL, &cached_state, GFP_NOFS); if (ret) { 
>> clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart,
> 
> 


- -- 
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=GDqG
-----END PGP SIGNATURE-----

[patch 18/35] btrfs: btrfs_drop_snapshot should return int

[Jeff Mahoney]

Commit cb1b69f4 (Btrfs: forced readonly when btrfs_drop_snapshot() fails)
made btrfs_drop_snapshot return void because there were no callers checking
the return value. That is the wrong order to handle error propogation since
the caller will have no idea that an error has occured and continue on
as if nothing went wrong.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ctree.h       |    6 +++---
 fs/btrfs/extent-tree.c |    4 ++--
 fs/btrfs/relocation.c  |    3 ++-
 fs/btrfs/transaction.c |    7 +++++--
 4 files changed, 12 insertions(+), 8 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 30c5a24..f6bca05 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2661,9 +2661,9 @@ static inline int btrfs_next_item(struct btrfs_root *root, struct btrfs_path *p)
 }
 int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path);
 int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf);
-void btrfs_drop_snapshot(struct btrfs_root *root,
-			 struct btrfs_block_rsv *block_rsv, int update_ref,
-			 int for_reloc);
+int __must_check btrfs_drop_snapshot(struct btrfs_root *root,
+				     struct btrfs_block_rsv *block_rsv,
+				     int update_ref, int for_reloc);
 int btrfs_drop_subtree(struct btrfs_trans_handle *trans,
 			struct btrfs_root *root,
 			struct extent_buffer *node,
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 0daa1df..cd6f8ae 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -6734,7 +6734,7 @@ static noinline int walk_up_tree(struct btrfs_trans_handle *trans,
  * also make sure backrefs for the shared block and all lower level
  * blocks are properly updated.
  */
-void btrfs_drop_snapshot(struct btrfs_root *root,
+int btrfs_drop_snapshot(struct btrfs_root *root,
 			 struct btrfs_block_rsv *block_rsv, int update_ref,
 			 int for_reloc)
 {
@@ -6902,7 +6902,7 @@ out_free:
 out:
 	if (err)
 		btrfs_std_error(root->fs_info, err);
-	return;
+	return err;
 }
 
 /*
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 5e89a5e..a87678e 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2272,7 +2272,8 @@ again:
 		} else {
 			list_del_init(&reloc_root->root_list);
 		}
-		btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0, 1);
+		ret = btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0, 1);
+		BUG_ON(ret < 0);
 	}
 
 	if (found) {
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 6e256d9..fb5cd5a 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -1387,6 +1387,8 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root)
 	spin_unlock(&fs_info->trans_lock);
 
 	while (!list_empty(&list)) {
+		int ret;
+
 		root = list_entry(list.next, struct btrfs_root, root_list);
 		list_del(&root->root_list);
 
@@ -1394,9 +1396,10 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root)
 
 		if (btrfs_header_backref_rev(root->node) <
 		    BTRFS_MIXED_BACKREF_REV)
-			btrfs_drop_snapshot(root, NULL, 0, 0);
+			ret = btrfs_drop_snapshot(root, NULL, 0, 0);
 		else
-			btrfs_drop_snapshot(root, NULL, 1, 0);
+			ret =btrfs_drop_snapshot(root, NULL, 1, 0);
+		BUG_ON(ret < 0);
 	}
 	return 0;
 }
-- 
1.7.9

[patch 19/35] btrfs: Dont BUG_ON errors from btrfs_create_subvol_root()

[Jeff Mahoney]

This is called from only one place - create_subvol() which passes errors
safely back out to it's caller, btrfs_mksubvol where they are handled.

Additionally, btrfs_create_subvol_root() itself bug's needlessly from error
return of btrfs_update_inode(). Since create_subvol() was fixed to catch
errors we can bubble this one up too.

Signed-off-by: Mark Fasheh <mfasheh@suse.com>
---
 fs/btrfs/inode.c |    3 +--
 fs/btrfs/ioctl.c |    5 +++++
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 593a2c3..4113233 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -6734,10 +6734,9 @@ int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
 	btrfs_i_size_write(inode, 0);
 
 	err = btrfs_update_inode(trans, new_root, inode);
-	BUG_ON(err);
 
 	iput(inode);
-	return 0;
+	return err;
 }
 
 struct inode *btrfs_alloc_inode(struct super_block *sb)
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 4e51494..de25e42 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -430,6 +430,11 @@ static noinline int create_subvol(struct btrfs_root *root,
 	btrfs_record_root_in_trans(trans, new_root);
 
 	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
+	if (ret) {
+		/* We potentially lose an unused inode item here */
+		goto fail;
+	}
+
 	/*
 	 * insert the directory item
 	 */
-- 
1.7.9

[patch 20/35] btrfs: Dont BUG_ON() errors in update_ref_for_cow()

[Jeff Mahoney]

The only caller of update_ref_for_cow() is __btrfs_cow_block() which was
originally ignoring any return values. update_ref_for_cow() however doesn't
look like a candidate to become a void function - there are a few places
where errors can occur.

So instead I changed update_ref_for_cow() to bubble all errors up (instead
of BUG_ON). __btrfs_cow_block() was then updated to catch and BUG_ON() any
errors from update_ref_for_cow(). The end effect is that we have no change
in behavior, but about 8 different places where a BUG_ON(ret) was removed.

Obviously a future patch will have to address the BUG_ON() in
__btrfs_cow_block().

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/ctree.c |   11 +++++++----
 1 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 9d472c2..e2e43c0 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -331,7 +331,8 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 	if (btrfs_block_can_be_shared(root, buf)) {
 		ret = btrfs_lookup_extent_info(trans, root, buf->start,
 					       buf->len, &refs, &flags);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 		BUG_ON(refs == 0);
 	} else {
 		refs = 1;
@@ -375,7 +376,8 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 							  buf->start,
 							  buf->len,
 							  new_flags, 0);
-			BUG_ON(ret);
+			if (ret)
+				return ret;
 		}
 	} else {
 		if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) {
@@ -415,7 +417,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 {
 	struct btrfs_disk_key disk_key;
 	struct extent_buffer *cow;
-	int level;
+	int level, ret;
 	int last_ref = 0;
 	int unlock_orig = 0;
 	u64 parent_start;
@@ -467,7 +469,8 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 			    (unsigned long)btrfs_header_fsid(cow),
 			    BTRFS_FSID_SIZE);
 
-	update_ref_for_cow(trans, root, buf, cow, &last_ref);
+	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
+	BUG_ON(ret);
 
 	if (root->ref_cows)
 		btrfs_reloc_cow_block(trans, root, buf, cow);
-- 
1.7.9

[patch 21/35] btrfs: Dont BUG_ON kzalloc error in btrfs_lookup_csums_range()

[Jeff Mahoney]

Unfortunately it isn't enough to just exit here - the kzalloc() happens in a
loop and the allocated items are added to a linked list whose head is passed
in from the caller.

To fix the BUG_ON() and also provide the semantic that the list passed in is
only modified on success, I create function-local temporary list that we add
items too. If no error is met, that list is spliced to the callers at the
end of the function. Otherwise the list will be walked and all items freed
before the error value is returned.

I did a simple test on this patch by forcing an error at the kzalloc() point
and verifying that when this hits (git clone seemed to exercise this), the
function throws the proper error. Unfortunately but predictably, we later
hit a BUG_ON(ret) type line that still hasn't been fixed up ;)

Signed-off-by: Mark Fasheh <mfasheh@suse.com>
---
 fs/btrfs/file-item.c |   15 +++++++++++++--
 1 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c
index edb69b4..89af104 100644
--- a/fs/btrfs/file-item.c
+++ b/fs/btrfs/file-item.c
@@ -284,6 +284,7 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 	struct btrfs_ordered_sum *sums;
 	struct btrfs_sector_sum *sector_sum;
 	struct btrfs_csum_item *item;
+	LIST_HEAD(tmplist);
 	unsigned long offset;
 	int ret;
 	size_t size;
@@ -358,7 +359,10 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 					MAX_ORDERED_SUM_BYTES(root));
 			sums = kzalloc(btrfs_ordered_sum_size(root, size),
 					GFP_NOFS);
-			BUG_ON(!sums);
+			if (!sums) {
+				ret = -ENOMEM;
+				goto fail;
+			}
 
 			sector_sum = sums->sums;
 			sums->bytenr = start;
@@ -380,12 +384,19 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 				offset += csum_size;
 				sector_sum++;
 			}
-			list_add_tail(&sums->list, list);
+			list_add_tail(&sums->list, &tmplist);
 		}
 		path->slots[0]++;
 	}
 	ret = 0;
 fail:
+	while (ret < 0 && !list_empty(&tmplist)) {
+		sums = list_entry(&tmplist, struct btrfs_ordered_sum, list);
+		list_del(&sums->list);
+		kfree(sums);
+	}
+	list_splice_tail(&tmplist, list);
+
 	btrfs_free_path(path);
 	return ret;
 }
-- 
1.7.9

[patch 22/35] btrfs: Dont BUG_ON errors in __finish_chunk_alloc()

[Jeff Mahoney]

All callers of __finish_chunk_alloc() BUG_ON() return value, so it's trivial
for us to always bubble up any errors caught in __finish_chunk_alloc() to be
caught there.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/volumes.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 43778b5..023aae0 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -3417,16 +3417,18 @@ static int __finish_chunk_alloc(struct btrfs_trans_handle *trans,
 	key.offset = chunk_offset;
 
 	ret = btrfs_insert_item(trans, chunk_root, &key, chunk, item_size);
-	BUG_ON(ret);
 
-	if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
+	if (ret == 0 && map->type & BTRFS_BLOCK_GROUP_SYSTEM) {
+		/*
+		 * TODO: Cleanup of inserted chunk root in case of
+		 * failure.
+		 */
 		ret = btrfs_add_system_chunk(chunk_root, &key, chunk,
 					     item_size);
-		BUG_ON(ret);
 	}
 
 	kfree(chunk);
-	return 0;
+	return ret;
 }
 
 /*
-- 
1.7.9

[patch 23/35] btrfs: Go readonly on bad extent refs in update_ref_for_cow()

[Jeff Mahoney]

update_ref_for_cow() will BUG_ON() after it's call to
btrfs_lookup_extent_info() if no existing references are found.  Since refs
are computed directly from disk, this should be treated as a corruption
instead of a logic error.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/ctree.c |    6 +++++-
 1 files changed, 5 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index e2e43c0..3b767d2 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -333,7 +333,11 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
 					       buf->len, &refs, &flags);
 		if (ret)
 			return ret;
-		BUG_ON(refs == 0);
+		if (refs == 0) {
+			ret = -EROFS;
+			btrfs_std_error(root->fs_info, ret);
+			return ret;
+		}
 	} else {
 		refs = 1;
 		if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID ||
-- 
1.7.9

[patch 24/35] btrfs: Dont BUG_ON errors from update_ref_for_cow()

[Jeff Mahoney]

__btrfs_cow_block(), the only caller of update_ref_for_cow() will BUG_ON()
any error return.  Instead, we can go read-only fs as update_ref_for_cow()
manipulates disk data in a way which doesn't look like it's easily rolled
back.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/ctree.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 3b767d2..36e16bd 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -474,7 +474,10 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
 			    BTRFS_FSID_SIZE);
 
 	ret = update_ref_for_cow(trans, root, buf, cow, &last_ref);
-	BUG_ON(ret);
+	if (ret) {
+		btrfs_std_error(root->fs_info, ret);
+		return ret;
+	}
 
 	if (root->ref_cows)
 		btrfs_reloc_cow_block(trans, root, buf, cow);
-- 
1.7.9

[patch 25/35] btrfs: Go readonly on tree errors in balance_level

[Jeff Mahoney]

balace_level() seems to deal with missing tree nodes by BUG_ON(). Instead,
we can easily just set the file system readonly and bubble -EROFS back up
the stack.

Signed-off-by: Mark Fasheh <mfasheh@suse.com>
---
 fs/btrfs/ctree.c |   13 +++++++++++--
 1 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 36e16bd..651a26a 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -944,7 +944,12 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 
 		/* promote the child to a root */
 		child = read_node_slot(root, mid, 0);
-		BUG_ON(!child);
+		if (!child) {
+			ret = -EROFS;
+			btrfs_std_error(root->fs_info, ret);
+			goto enospc;
+		}
+
 		btrfs_tree_lock(child);
 		btrfs_set_lock_blocking(child);
 		ret = btrfs_cow_block(trans, root, child, mid, 0, &child);
@@ -1042,7 +1047,11 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
 		 * otherwise we would have pulled some pointers from the
 		 * right
 		 */
-		BUG_ON(!left);
+		if (!left) {
+			ret = -EROFS;
+			btrfs_std_error(root->fs_info, ret);
+			goto enospc;
+		}
 		wret = balance_node_right(trans, root, mid, left);
 		if (wret < 0) {
 			ret = wret;
-- 
1.7.9

[patch 26/35] btrfs: Dont BUG_ON insert errors in btrfs_alloc_dev_extent()

[Jeff Mahoney]

The only caller of btrfs_alloc_dev_extent() is __btrfs_alloc_chunk() which
already bugs on any error returned. We can remove the BUG_ON's in
btrfs_alloc_dev_extent() then since __btrfs_alloc_chunk() will "catch" them
anyway.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/volumes.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 023aae0..8d4b7d4 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1075,7 +1075,8 @@ int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
 	key.type = BTRFS_DEV_EXTENT_KEY;
 	ret = btrfs_insert_empty_item(trans, root, path, &key,
 				      sizeof(*extent));
-	BUG_ON(ret);
+	if (ret)
+		goto out;
 
 	leaf = path->nodes[0];
 	extent = btrfs_item_ptr(leaf, path->slots[0],
@@ -1090,6 +1091,7 @@ int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans,
 
 	btrfs_set_dev_extent_length(leaf, extent, num_bytes);
 	btrfs_mark_buffer_dirty(leaf);
+out:
 	btrfs_free_path(path);
 	return ret;
 }
-- 
1.7.9

[patch 27/35] btrfs: Remove BUG_ON from __btrfs_alloc_chunk()

[Jeff Mahoney]

We BUG_ON() error from add_extent_mapping(), but that error looks pretty
easy to bubble back up - as far as I can tell there have not been any
permanent modifications to fs state at that point.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/volumes.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 8d4b7d4..8bf19e1 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -3327,8 +3327,9 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
 	write_lock(&em_tree->lock);
 	ret = add_extent_mapping(em_tree, em);
 	write_unlock(&em_tree->lock);
-	BUG_ON(ret);
 	free_extent_map(em);
+	if (ret)
+		goto error;
 
 	ret = btrfs_make_block_group(trans, extent_root, 0, type,
 				     BTRFS_FIRST_CHUNK_TREE_OBJECTID,
-- 
1.7.9

[patch 28/35] btrfs: Remove BUG_ON from __finish_chunk_alloc()

[Jeff Mahoney]

btrfs_alloc_chunk() unconditionally BUGs on any error returned from
__finish_chunk_alloc() so there's no need for two BUG_ON lines. Remove the
one from __finish_chunk_alloc().

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
---
 fs/btrfs/volumes.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 8bf19e1..8a30d63 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -3383,7 +3383,8 @@ static int __finish_chunk_alloc(struct btrfs_trans_handle *trans,
 		device = map->stripes[index].dev;
 		device->bytes_used += stripe_size;
 		ret = btrfs_update_device(trans, device);
-		BUG_ON(ret);
+		if (ret)
+			goto out_free;
 		index++;
 	}
 
@@ -3430,6 +3431,7 @@ static int __finish_chunk_alloc(struct btrfs_trans_handle *trans,
 					     item_size);
 	}
 
+out_free:
 	kfree(chunk);
 	return ret;
 }
-- 
1.7.9

[patch 29/35] btrfs: add varargs to btrfs_error

[Jeff Mahoney]

 btrfs currently handles most errors with BUG_ON. This patch is a work-in-
 progress but aims to handle most errors other than internal logic
 errors and ENOMEM more gracefully.

 This iteration prevents most crashes but can run into lockups with
 the page lock on occasion when the timing "works out."

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ctree.h |   12 ++++++++-
 fs/btrfs/super.c |   63 ++++++++++++++++++++++++++++++++++++++++++++++++------
 2 files changed, 66 insertions(+), 9 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index f6bca05..8829f80 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2964,13 +2964,21 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size);
 /* super.c */
 int btrfs_parse_options(struct btrfs_root *root, char *options);
 int btrfs_sync_fs(struct super_block *sb, int wait);
+void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...);
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
-		     unsigned int line, int errno);
+		     unsigned int line, int errno, const char *fmt, ...);
 
 #define btrfs_std_error(fs_info, errno)				\
 do {								\
 	if ((errno))						\
-		__btrfs_std_error((fs_info), __func__, __LINE__, (errno));\
+		__btrfs_std_error((fs_info), __func__,		\
+				   __LINE__, (errno), NULL);	\
+} while (0)
+
+#define btrfs_error(fs_info, errno, fmt, args...)		\
+do {								\
+	__btrfs_std_error((fs_info), __func__, __LINE__,	\
+			  (errno), fmt, ##args);		\
 } while (0)
 
 void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function,
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index ae7963b..7fe69ee 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -127,25 +127,74 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
  * invokes the approciate error response.
  */
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
-		     unsigned int line, int errno)
+		       unsigned int line, int errno, const char *fmt, ...)
 {
 	struct super_block *sb = fs_info->sb;
 	char nbuf[16];
 	const char *errstr;
+	va_list args;
+	va_start(args, fmt);
 
 	/*
 	 * Special case: if the error is EROFS, and we're already
 	 * under MS_RDONLY, then it is safe here.
 	 */
 	if (errno == -EROFS && (sb->s_flags & MS_RDONLY))
-		return;
+  		return;
 
-	errstr = btrfs_decode_error(fs_info, errno, nbuf);
-	printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s\n",
-		sb->s_id, function, line, errstr);
-	save_error_info(fs_info);
+  	errstr = btrfs_decode_error(fs_info, errno, nbuf);
+	if (fmt) {
+		struct va_format vaf = {
+			.fmt = fmt,
+			.va = &args,
+		};
+
+		printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s (%pV)\n",
+			sb->s_id, function, line, errstr, &vaf);
+	} else {
+		printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s\n",
+			sb->s_id, function, line, errstr);
+	}
+
+	/* Don't go through full error handling during mount */
+	if (sb->s_flags & MS_BORN) {
+		save_error_info(fs_info);
+		btrfs_handle_error(fs_info);
+	}
+	va_end(args);
+}
 
-	btrfs_handle_error(fs_info);
+const char *logtypes[] = {
+	"emergency",
+	"alert",
+	"critical",
+	"error",
+	"warning",
+	"notice",
+	"info",
+	"debug",
+};
+
+void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...)
+{
+	struct super_block *sb = fs_info->sb;
+	char lvl[4];
+	struct va_format vaf;
+	va_list args;
+	const char *type = logtypes[4];
+
+	va_start(args, fmt);
+
+	if (fmt[0] == '<' && isdigit(fmt[1]) && fmt[2] == '>') {
+		strncpy(lvl, fmt, 3);
+		fmt += 3;
+		type = logtypes[fmt[1] - '0'];
+	} else
+		*lvl = '\0';
+
+	vaf.fmt = fmt;
+	vaf.va = &args;
+	printk("%sBTRFS %s (device %s): %pV", lvl, type, sb->s_id, &vaf);
 }
 
 /*
-- 
1.7.9

[patch 30/35] btrfs: enhance transaction abort infrastructure

[Jeff Mahoney]

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 fs/btrfs/ctree.h       |   13 +++-
 fs/btrfs/disk-io.c     |   50 ++++++++++++-
 fs/btrfs/disk-io.h     |    4 +
 fs/btrfs/relocation.c  |   14 ++--
 fs/btrfs/scrub.c       |    8 ++-
 fs/btrfs/super.c       |   74 +++++++++++++++++--
 fs/btrfs/transaction.c |  190 +++++++++++++++++++++++++++++++++++++++---------
 fs/btrfs/transaction.h |    3 +
 8 files changed, 300 insertions(+), 56 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 8829f80..b6ebea5 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2968,6 +2968,16 @@ void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...);
 void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
 		     unsigned int line, int errno, const char *fmt, ...);
 
+void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root, const char *function,
+			       unsigned int line, int errno);
+
+#define btrfs_abort_transaction(trans, root, errno)		\
+do {								\
+	__btrfs_abort_transaction(trans, root, __func__,	\
+				  __LINE__, errno);		\
+} while (0)
+
 #define btrfs_std_error(fs_info, errno)				\
 do {								\
 	if ((errno))						\
@@ -3024,7 +3034,7 @@ void btrfs_reloc_cow_block(struct btrfs_trans_handle *trans,
 void btrfs_reloc_pre_snapshot(struct btrfs_trans_handle *trans,
 			      struct btrfs_pending_snapshot *pending,
 			      u64 *bytes_to_reserve);
-void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
+int btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 			      struct btrfs_pending_snapshot *pending);
 
 /* scrub.c */
@@ -3034,6 +3044,7 @@ void btrfs_scrub_pause(struct btrfs_root *root);
 void btrfs_scrub_pause_super(struct btrfs_root *root);
 void btrfs_scrub_continue(struct btrfs_root *root);
 void btrfs_scrub_continue_super(struct btrfs_root *root);
+int __btrfs_scrub_cancel(struct btrfs_fs_info *info);
 int btrfs_scrub_cancel(struct btrfs_root *root);
 int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev);
 int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid);
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 69ef456..6297a03 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -61,7 +61,6 @@ static int btrfs_destroy_marked_extents(struct btrfs_root *root,
 					int mark);
 static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
 				       struct extent_io_tree *pinned_extents);
-static int btrfs_cleanup_transaction(struct btrfs_root *root);
 
 /*
  * end_io_wq structs are used to do processing in task context when an IO is
@@ -2896,6 +2895,19 @@ int write_ctree_super(struct btrfs_trans_handle *trans,
 	return ret;
 }
 
+/* Kill all outstanding I/O */
+void btrfs_abort_devices(struct btrfs_root *root)
+{
+	struct list_head *head;
+	struct btrfs_device *dev;
+	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+	head = &root->fs_info->fs_devices->devices;
+	list_for_each_entry_rcu(dev, head, dev_list) {
+		blk_abort_queue(dev->bdev->bd_disk->queue);
+	}
+	mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+}
+
 void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
 {
 	spin_lock(&fs_info->fs_roots_radix_lock);
@@ -3536,13 +3548,43 @@ static int btrfs_destroy_pinned_extent(struct btrfs_root *root,
 	return 0;
 }
 
-static int btrfs_cleanup_transaction(struct btrfs_root *root)
+void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
+				   struct btrfs_root *root)
+{
+	btrfs_destroy_delayed_refs(cur_trans, root);
+	btrfs_block_rsv_release(root, &root->fs_info->trans_block_rsv,
+				cur_trans->dirty_pages.dirty_bytes);
+
+	/* FIXME: cleanup wait for commit */
+	cur_trans->in_commit = 1;
+	cur_trans->blocked = 1;
+	if (waitqueue_active(&root->fs_info->transaction_blocked_wait))
+		wake_up(&root->fs_info->transaction_blocked_wait);
+
+	cur_trans->blocked = 0;
+	if (waitqueue_active(&root->fs_info->transaction_wait))
+		wake_up(&root->fs_info->transaction_wait);
+
+	cur_trans->commit_done = 1;
+	if (waitqueue_active(&cur_trans->commit_wait))
+		wake_up(&cur_trans->commit_wait);
+
+	btrfs_destroy_pending_snapshots(cur_trans);
+
+	btrfs_destroy_marked_extents(root, &cur_trans->dirty_pages,
+				     EXTENT_DIRTY);
+
+	/*
+	memset(cur_trans, 0, sizeof(*cur_trans));
+	kmem_cache_free(btrfs_transaction_cachep, cur_trans);
+	*/
+}
+
+int btrfs_cleanup_transaction(struct btrfs_root *root)
 {
 	struct btrfs_transaction *t;
 	LIST_HEAD(list);
 
-	WARN_ON(1);
-
 	mutex_lock(&root->fs_info->transaction_kthread_mutex);
 
 	spin_lock(&root->fs_info->trans_lock);
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index 1608492..a7ace1a 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -85,6 +85,10 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
 int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root);
+int btrfs_cleanup_transaction(struct btrfs_root *root);
+void btrfs_cleanup_one_transaction(struct btrfs_transaction *trans,
+				  struct btrfs_root *root);
+void btrfs_abort_devices(struct btrfs_root *root);
 
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 void btrfs_init_lockdep(void);
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index a87678e..cba7a0b 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -4410,7 +4410,7 @@ void btrfs_reloc_pre_snapshot(struct btrfs_trans_handle *trans,
  * called after snapshot is created. migrate block reservation
  * and create reloc root for the newly created snapshot
  */
-void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
+int btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 			       struct btrfs_pending_snapshot *pending)
 {
 	struct btrfs_root *root = pending->root;
@@ -4420,7 +4420,7 @@ void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 	int ret;
 
 	if (!root->reloc_root)
-		return;
+		return 0;
 
 	rc = root->fs_info->reloc_ctl;
 	rc->merging_rsv_size += rc->nodes_relocated;
@@ -4429,19 +4429,21 @@ void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
 		ret = btrfs_block_rsv_migrate(&pending->block_rsv,
 					      rc->block_rsv,
 					      rc->nodes_relocated);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 	}
 
 	new_root = pending->snap;
 	reloc_root = create_reloc_root(trans, root->reloc_root,
 				       new_root->root_key.objectid);
+	if (IS_ERR(reloc_root))
+		return PTR_ERR(reloc_root);
 
 	ret = __add_reloc_root(reloc_root);
 	BUG_ON(ret < 0);
 	new_root->reloc_root = reloc_root;
 
-	if (rc->create_reloc_tree) {
+	if (rc->create_reloc_tree)
 		ret = clone_backref_node(trans, rc, root, reloc_root);
-		BUG_ON(ret);
-	}
+	return ret;
 }
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index a2e8aa4..794cbb5 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -1680,9 +1680,8 @@ void btrfs_scrub_continue_super(struct btrfs_root *root)
 	up_write(&root->fs_info->scrub_super_lock);
 }
 
-int btrfs_scrub_cancel(struct btrfs_root *root)
+int __btrfs_scrub_cancel(struct btrfs_fs_info *fs_info)
 {
-	struct btrfs_fs_info *fs_info = root->fs_info;
 
 	mutex_lock(&fs_info->scrub_lock);
 	if (!atomic_read(&fs_info->scrubs_running)) {
@@ -1703,6 +1702,11 @@ int btrfs_scrub_cancel(struct btrfs_root *root)
 	return 0;
 }
 
+int btrfs_scrub_cancel(struct btrfs_root *root)
+{
+	return __btrfs_scrub_cancel(root->fs_info);
+}
+
 int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 7fe69ee..0517bd7 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -119,6 +119,8 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
 	if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
 		sb->s_flags |= MS_RDONLY;
 		printk(KERN_INFO "btrfs is forced readonly\n");
+		__btrfs_scrub_cancel(fs_info);
+//		WARN_ON(1);
 	}
 }
 
@@ -198,6 +200,34 @@ void btrfs_printk(struct btrfs_fs_info *fs_info, const char *fmt, ...)
 }
 
 /*
+ * We only mark the transaction aborted and then set the file system read-only.
+ * This will prevent new transactions from starting or trying to join this
+ * one.
+ *
+ * This means that error recovery at the call site is limited to freeing
+ * any local memory allocations and passing the error code up without
+ * further cleanup. The transaction should complete as it normally would
+ * in the call path but will return -EIO.
+ *
+ * We'll complete the cleanup in btrfs_end_transaction and
+ * btrfs_commit_transaction.
+ */
+void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
+			       struct btrfs_root *root, const char *function,
+			       unsigned int line, int errno)
+{
+	WARN_ON_ONCE(1);
+	trans->aborted = errno;
+	/* Nothing used. The other threads that have joined this
+	 * transaction may be able to continue. */
+	if (!trans->blocks_used) {
+		btrfs_printk(root->fs_info, "Aborting unused transaction.\n");
+		return;
+	}
+	trans->transaction->aborted = errno;
+	__btrfs_std_error(root->fs_info, function, line, errno, NULL);
+}
+/*
  * __btrfs_panic decodes unexpected, fatal errors from the caller,
  * issues an alert, and either panics or BUGs, depending on mount options.
  */
@@ -295,6 +325,7 @@ static match_table_t tokens = {
 /*
  * Regular mount options parser.  Everything that is needed only when
  * reading in a new superblock is parsed here.
+ * XXX JDM: This needs to be cleaned up for remount.
  */
 int btrfs_parse_options(struct btrfs_root *root, char *options)
 {
@@ -1096,11 +1127,20 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(sb);
 	struct btrfs_root *root = fs_info->tree_root;
+	unsigned old_flags = sb->s_flags;
+	unsigned long old_opts = fs_info->mount_opt;
+	unsigned long old_compress_type = fs_info->compress_type;
+	u64 old_max_inline = fs_info->max_inline;
+	u64 old_alloc_start = fs_info->alloc_start;
+	int old_thread_pool_size = fs_info->thread_pool_size;
+	unsigned int old_metadata_ratio = fs_info->metadata_ratio;
 	int ret;
 
 	ret = btrfs_parse_options(root, data);
-	if (ret)
-		return -EINVAL;
+	if (ret) {
+		ret = -EINVAL;
+		goto restore;
+	}
 
 	if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))
 		return 0;
@@ -1108,26 +1148,44 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data)
 	if (*flags & MS_RDONLY) {
 		sb->s_flags |= MS_RDONLY;
 
-		ret =  btrfs_commit_super(root);
-		WARN_ON(ret);
+		ret = btrfs_commit_super(root);
+		if (ret)
+			goto restore;
 	} else {
 		if (fs_info->fs_devices->rw_devices == 0)
-			return -EACCES;
+			ret = -EACCES;
+			goto restore;
 
 		if (btrfs_super_log_root(fs_info->super_copy) != 0)
-			return -EINVAL;
+			ret = -EINVAL;
+			goto restore;
 
 		ret = btrfs_cleanup_fs_roots(fs_info);
-		WARN_ON(ret);
+		if (ret)
+			goto restore;
 
 		/* recover relocation */
 		ret = btrfs_recover_relocation(root);
-		WARN_ON(ret);
+		if (ret)
+			goto restore;
 
 		sb->s_flags &= ~MS_RDONLY;
 	}
 
 	return 0;
+
+restore:
+	/* We've hit an error - don't reset MS_RDONLY */
+	if (sb->s_flags & MS_RDONLY)
+		old_flags |= MS_RDONLY;
+	sb->s_flags = old_flags;
+	fs_info->mount_opt = old_opts;
+	fs_info->compress_type = old_compress_type;
+	fs_info->max_inline = old_max_inline;
+	fs_info->alloc_start = old_alloc_start;
+	fs_info->thread_pool_size = old_thread_pool_size;
+	fs_info->metadata_ratio = old_metadata_ratio;
+	return ret;
 }
 
 /* Used to sort the devices by max_avail(descending sort) */
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index fb5cd5a..5a4999a 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -31,7 +31,7 @@
 
 #define BTRFS_ROOT_TRANS_TAG 0
 
-static noinline void put_transaction(struct btrfs_transaction *transaction)
+void put_transaction(struct btrfs_transaction *transaction)
 {
 	WARN_ON(atomic_read(&transaction->use_count) == 0);
 	if (atomic_dec_and_test(&transaction->use_count)) {
@@ -58,6 +58,12 @@ static noinline int join_transaction(struct btrfs_root *root, int nofail)
 
 	spin_lock(&root->fs_info->trans_lock);
 loop:
+	/* The file system has been taken offline. No new transactions. */
+	if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
+		spin_unlock(&root->fs_info->trans_lock);
+		return -EROFS;
+	}
+
 	if (root->fs_info->trans_no_join) {
 		if (!nofail) {
 			spin_unlock(&root->fs_info->trans_lock);
@@ -67,6 +73,8 @@ loop:
 
 	cur_trans = root->fs_info->running_transaction;
 	if (cur_trans) {
+		if (cur_trans->aborted)
+			return cur_trans->aborted;
 		atomic_inc(&cur_trans->use_count);
 		atomic_inc(&cur_trans->num_writers);
 		cur_trans->num_joined++;
@@ -123,6 +131,7 @@ loop:
 	root->fs_info->generation++;
 	cur_trans->transid = root->fs_info->generation;
 	root->fs_info->running_transaction = cur_trans;
+	cur_trans->aborted = 0;
 	spin_unlock(&root->fs_info->trans_lock);
 
 	return 0;
@@ -318,6 +327,7 @@ again:
 	h->use_count = 1;
 	h->block_rsv = NULL;
 	h->orig_rsv = NULL;
+	h->aborted = 0;
 
 	smp_mb();
 	if (cur_trans->blocked && may_wait_transaction(root, type)) {
@@ -440,6 +450,7 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 	struct btrfs_transaction *cur_trans = trans->transaction;
 	struct btrfs_block_rsv *rsv = trans->block_rsv;
 	int updates;
+	int err;
 
 	smp_mb();
 	if (cur_trans->blocked || cur_trans->delayed_refs.flushing)
@@ -453,8 +464,11 @@ int btrfs_should_end_transaction(struct btrfs_trans_handle *trans,
 
 	updates = trans->delayed_ref_updates;
 	trans->delayed_ref_updates = 0;
-	if (updates)
-		btrfs_run_delayed_refs(trans, root, updates);
+	if (updates) {
+		err = btrfs_run_delayed_refs(trans, root, updates);
+		if (err) /* Error code will also eval true */
+			return err;
+	}
 
 	trans->block_rsv = rsv;
 
@@ -525,6 +539,11 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	if (throttle)
 		btrfs_run_delayed_iputs(root);
 
+	if (trans->aborted ||
+	    root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
+		return -EIO;
+	}
+
 	return 0;
 }
 
@@ -690,11 +709,13 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans,
 		ret = btrfs_update_root(trans, tree_root,
 					&root->root_key,
 					&root->root_item);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 
 		old_root_used = btrfs_root_used(&root->root_item);
 		ret = btrfs_write_dirty_block_groups(trans, root);
-		BUG_ON(ret);
+		if (ret)
+			return ret;
 	}
 
 	if (root != root->fs_info->extent_root)
@@ -705,6 +726,10 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans,
 
 /*
  * update all the cowonly tree roots on disk
+ *
+ * The error handling in this function may not be obvious. Any of the
+ * failures will cause the file system to go offline. We still need
+ * to clean up the delayed refs.
  */
 static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans,
 					 struct btrfs_root *root)
@@ -715,22 +740,30 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans,
 	int ret;
 
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
-	BUG_ON(ret);
+	if (ret)
+		return ret;
 
 	eb = btrfs_lock_root_node(fs_info->tree_root);
-	btrfs_cow_block(trans, fs_info->tree_root, eb, NULL, 0, &eb);
+	ret = btrfs_cow_block(trans, fs_info->tree_root, eb, NULL,
+			      0, &eb);
 	btrfs_tree_unlock(eb);
 	free_extent_buffer(eb);
 
+	if (ret)
+		return ret;
+
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
-	BUG_ON(ret);
+	if (ret)
+		return ret;
 
 	while (!list_empty(&fs_info->dirty_cowonly_roots)) {
 		next = fs_info->dirty_cowonly_roots.next;
 		list_del_init(next);
 		root = list_entry(next, struct btrfs_root, dirty_list);
 
-		update_cowonly_root(trans, root);
+		ret = update_cowonly_root(trans, root);
+		if (ret)
+			return ret;
 	}
 
 	down_write(&fs_info->extent_commit_sem);
@@ -874,7 +907,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 
 	new_root_item = kmalloc(sizeof(*new_root_item), GFP_NOFS);
 	if (!new_root_item) {
-		pending->error = -ENOMEM;
+		ret = pending->error = -ENOMEM;
 		goto fail;
 	}
 
@@ -911,7 +944,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	 * insert the directory item
 	 */
 	ret = btrfs_set_inode_index(parent_inode, &index);
-	BUG_ON(ret);
+	BUG_ON(ret); /* -ENOMEM */
 	ret = btrfs_insert_dir_item(trans, parent_root,
 				dentry->d_name.name, dentry->d_name.len,
 				parent_inode, &key,
@@ -920,12 +953,14 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 		pending->error = -EEXIST;
 		dput(parent);
 		goto fail;
-	}
+	} else if (ret)
+		goto abort_trans;
 
 	btrfs_i_size_write(parent_inode, parent_inode->i_size +
 					 dentry->d_name.len * 2);
 	ret = btrfs_update_inode(trans, parent_root, parent_inode);
-	BUG_ON(ret);
+	if (ret)
+		goto abort_trans;
 
 	/*
 	 * pull in the delayed directory update
@@ -934,7 +969,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	 * snapshot
 	 */
 	ret = btrfs_run_delayed_items(trans, root);
-	BUG_ON(ret);
+	if (ret) /* Transaction aborted */
+		goto fail;
 
 	record_root_in_trans(trans, root);
 	btrfs_set_root_last_snapshot(&root->root_item, trans->transid);
@@ -949,10 +985,16 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	btrfs_set_root_flags(new_root_item, root_flags);
 
 	old = btrfs_lock_root_node(root);
-	btrfs_cow_block(trans, root, old, NULL, 0, &old);
+	ret = btrfs_cow_block(trans, root, old, NULL, 0, &old);
+	if (ret)
+		goto abort_trans;
+
 	btrfs_set_lock_blocking(old);
 
-	btrfs_copy_root(trans, root, old, &tmp, objectid);
+	ret = btrfs_copy_root(trans, root, old, &tmp, objectid);
+	if (ret)
+		goto abort_trans;
+
 	btrfs_tree_unlock(old);
 	free_extent_buffer(old);
 
@@ -966,7 +1008,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	ret = btrfs_insert_root(trans, tree_root, &key, new_root_item);
 	btrfs_tree_unlock(tmp);
 	free_extent_buffer(tmp);
-	BUG_ON(ret);
+	if (ret)
+		goto abort_trans;
 
 	/*
 	 * insert root back/forward references
@@ -975,19 +1018,28 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 				 parent_root->root_key.objectid,
 				 btrfs_ino(parent_inode), index,
 				 dentry->d_name.name, dentry->d_name.len);
-	BUG_ON(ret);
+	if (ret)
+		goto fail;
 	dput(parent);
 
 	key.offset = (u64)-1;
 	pending->snap = btrfs_read_fs_root_no_name(root->fs_info, &key);
-	BUG_ON(IS_ERR(pending->snap));
+	if (IS_ERR(pending->snap))
+		goto abort_trans;
 
-	btrfs_reloc_post_snapshot(trans, pending);
+	ret = btrfs_reloc_post_snapshot(trans, pending);
+	if (ret)
+		goto abort_trans;
+	ret = 0;
 fail:
 	kfree(new_root_item);
 	trans->block_rsv = rsv;
 	btrfs_block_rsv_release(root, &pending->block_rsv, (u64)-1);
-	return 0;
+	return ret;
+
+abort_trans:
+	btrfs_abort_transaction(trans, root, ret);
+	goto fail;
 }
 
 /*
@@ -1124,6 +1176,33 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans,
 	return 0;
 }
 
+
+static void cleanup_transaction(struct btrfs_trans_handle *trans,
+				struct btrfs_root *root)
+{
+	struct btrfs_transaction *cur_trans = trans->transaction;
+
+	WARN_ON(trans->use_count > 1);
+
+	spin_lock(&root->fs_info->trans_lock);
+	list_del_init(&cur_trans->list);
+	spin_unlock(&root->fs_info->trans_lock);
+
+	btrfs_cleanup_one_transaction(trans->transaction, root);
+
+	put_transaction(cur_trans);
+	put_transaction(cur_trans);
+
+	trace_btrfs_transaction_commit(root);
+
+	btrfs_scrub_continue(root);
+
+	if (current->journal_info == trans)
+		current->journal_info = NULL;
+
+	kmem_cache_free(btrfs_trans_handle_cachep, trans);
+}
+
 /*
  * btrfs_transaction state sequence:
  *    in_commit = 0, blocked = 0  (initial)
@@ -1135,10 +1214,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root)
 {
 	unsigned long joined = 0;
-	struct btrfs_transaction *cur_trans;
+	struct btrfs_transaction *cur_trans = trans->transaction;
 	struct btrfs_transaction *prev_trans = NULL;
 	DEFINE_WAIT(wait);
-	int ret;
+	int ret = -EIO;
 	int should_grow = 0;
 	unsigned long now = get_seconds();
 	int flush_on_commit = btrfs_test_opt(root, FLUSHONCOMMIT);
@@ -1148,13 +1227,18 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	btrfs_trans_release_metadata(trans, root);
 	trans->block_rsv = NULL;
 
+	if (cur_trans->aborted)
+		goto cleanup_transaction;
+
 	/* make a pass through all the delayed refs we have so far
 	 * any runnings procs may add more while we are here
 	 */
 	ret = btrfs_run_delayed_refs(trans, root, 0);
-	BUG_ON(ret);
+	if (ret)
+		goto cleanup_transaction;
 
 	cur_trans = trans->transaction;
+
 	/*
 	 * set the flushing flag so procs in this transaction have to
 	 * start sending their work down.
@@ -1162,19 +1246,20 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	cur_trans->delayed_refs.flushing = 1;
 
 	ret = btrfs_run_delayed_refs(trans, root, 0);
-	BUG_ON(ret);
+	if (ret)
+		goto cleanup_transaction;
 
 	spin_lock(&cur_trans->commit_lock);
 	if (cur_trans->in_commit) {
 		spin_unlock(&cur_trans->commit_lock);
 		atomic_inc(&cur_trans->use_count);
-		btrfs_end_transaction(trans, root);
+		ret = btrfs_end_transaction(trans, root);
 
 		wait_for_commit(root, cur_trans);
 
 		put_transaction(cur_trans);
 
-		return 0;
+		return ret;
 	}
 
 	trans->transaction->in_commit = 1;
@@ -1218,7 +1303,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		}
 
 		ret = btrfs_run_delayed_items(trans, root);
-		BUG_ON(ret);
+		if (ret)
+			goto cleanup_transaction;
 
 		/*
 		 * rename don't use btrfs_join_transaction, so, once we
@@ -1260,13 +1346,22 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	mutex_lock(&root->fs_info->reloc_mutex);
 
 	ret = btrfs_run_delayed_items(trans, root);
-	BUG_ON(ret);
+	if (ret) {
+		mutex_unlock(&root->fs_info->reloc_mutex);
+		goto cleanup_transaction;
+	}
 
 	ret = create_pending_snapshots(trans, root->fs_info);
-	BUG_ON(ret);
+	if (ret) {
+		mutex_unlock(&root->fs_info->reloc_mutex);
+		goto cleanup_transaction;
+	}
 
 	ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
-	BUG_ON(ret);
+	if (ret) {
+		mutex_unlock(&root->fs_info->reloc_mutex);
+		goto cleanup_transaction;
+	}
 
 	/*
 	 * make sure none of the code above managed to slip in a
@@ -1293,7 +1388,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	mutex_lock(&root->fs_info->tree_log_mutex);
 
 	ret = commit_fs_roots(trans, root);
-	BUG_ON(ret);
+	if (ret) {
+		mutex_unlock(&root->fs_info->tree_log_mutex);
+		goto cleanup_transaction;
+	}
 
 	/* commit_fs_roots gets rid of all the tree log roots, it is now
 	 * safe to free the root of tree log roots
@@ -1301,7 +1399,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	btrfs_free_log_root_tree(trans, root->fs_info);
 
 	ret = commit_cowonly_roots(trans, root);
-	BUG_ON(ret);
+	if (ret) {
+		mutex_unlock(&root->fs_info->tree_log_mutex);
+		goto cleanup_transaction;
+	}
 
 	btrfs_prepare_extent_commit(trans, root);
 
@@ -1335,8 +1436,18 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 	wake_up(&root->fs_info->transaction_wait);
 
 	ret = btrfs_write_and_wait_transaction(trans, root);
-	BUG_ON(ret);
-	write_ctree_super(trans, root, 0);
+	if (ret) {
+		btrfs_error(root->fs_info, ret,
+			    "Error while writing out transaction.");
+		mutex_unlock(&root->fs_info->tree_log_mutex);
+		goto cleanup_transaction;
+	}
+
+	ret = write_ctree_super(trans, root, 0);
+	if (ret) {
+		mutex_unlock(&root->fs_info->tree_log_mutex);
+		goto cleanup_transaction;
+	}
 
 	/*
 	 * the super is written, we can safely allow the tree-loggers
@@ -1372,6 +1483,15 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 		btrfs_run_delayed_iputs(root);
 
 	return ret;
+
+cleanup_transaction:
+	btrfs_printk(root->fs_info, "Skipping commit of aborted transaction.\n");
+//	WARN_ON(1);
+	if (current->journal_info == trans)
+		current->journal_info = NULL;
+	cleanup_transaction(trans, root);
+
+	return ret;
 }
 
 /*
diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h
index 02564e6..fe27379 100644
--- a/fs/btrfs/transaction.h
+++ b/fs/btrfs/transaction.h
@@ -43,6 +43,7 @@ struct btrfs_transaction {
 	wait_queue_head_t commit_wait;
 	struct list_head pending_snapshots;
 	struct btrfs_delayed_ref_root delayed_refs;
+	int aborted;
 };
 
 struct btrfs_trans_handle {
@@ -55,6 +56,7 @@ struct btrfs_trans_handle {
 	struct btrfs_transaction *transaction;
 	struct btrfs_block_rsv *block_rsv;
 	struct btrfs_block_rsv *orig_rsv;
+	int aborted;
 };
 
 struct btrfs_pending_snapshot {
@@ -114,4 +116,5 @@ int btrfs_wait_marked_extents(struct btrfs_root *root,
 				struct extent_io_tree *dirty_pages, int mark);
 int btrfs_transaction_blocked(struct btrfs_fs_info *info);
 int btrfs_transaction_in_commit(struct btrfs_fs_info *info);
+void put_transaction(struct btrfs_transaction *transaction);
 #endif
-- 
1.7.9

[patch 32/35] btrfs: Fix busyloop in transaction_kthread()

[Jeff Mahoney]

When a filesystem got aborted due do error, transaction_kthread() will
busyloop.  Fix it by going to sleep in that case as well. Maybe we should
just stop transaction_kthread() when filesystem is aborted but that would be
more complex.

Signed-off-by: Jan Kara <jack@suse.cz>
---
 fs/btrfs/disk-io.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 16a0cad..438993e 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1640,8 +1640,10 @@ static int transaction_kthread(void *arg)
 	u64 transid;
 	unsigned long now;
 	unsigned long delay;
+	bool cannot_commit;
 
 	do {
+		cannot_commit = false;
 		delay = HZ * 30;
 		vfs_check_frozen(root->fs_info->sb, SB_FREEZE_WRITE);
 		mutex_lock(&root->fs_info->transaction_kthread_mutex);
@@ -1665,8 +1667,10 @@ static int transaction_kthread(void *arg)
 
 		/* If the file system is aborted, this will always fail. */
 		trans = btrfs_join_transaction(root);
-		if (IS_ERR(trans))
+		if (IS_ERR(trans)) {
+			cannot_commit = true;
 			goto sleep;
+		}
 		if (transid == trans->transid) {
 			btrfs_commit_transaction(trans, root);
 		} else {
@@ -1679,7 +1683,8 @@ sleep:
 		if (!try_to_freeze()) {
 			set_current_state(TASK_INTERRUPTIBLE);
 			if (!kthread_should_stop() &&
-			    !btrfs_transaction_blocked(root->fs_info))
+			    (!btrfs_transaction_blocked(root->fs_info) ||
+			     cannot_commit))
 				schedule_timeout(delay);
 			__set_current_state(TASK_RUNNING);
 		}
-- 
1.7.9

[patch 33/35] btrfs: handle errors when excluding super extents

[Jeff Mahoney]

Some images produced by btrfs-image cause a mount crash in rmap block
when excluding superblock extents. Handle the errors gracefully.

Signed-off-by: David Sterba <dsterba@suse.cz>
---
 fs/btrfs/extent-tree.c |   19 ++++++++++++++++---
 fs/btrfs/volumes.c     |   20 +++++++++++++++++---
 2 files changed, 33 insertions(+), 6 deletions(-)

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 0bdb9fa..8055d58 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -250,10 +250,17 @@ static int exclude_super_stripes(struct btrfs_root *root,
 
 	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
 		bytenr = btrfs_sb_offset(i);
+		if (bytenr <= cache->key.objectid + cache->key.offset)
+			break;
+		logical = NULL;
 		ret = btrfs_rmap_block(&root->fs_info->mapping_tree,
 				       cache->key.objectid, bytenr,
 				       0, &logical, &nr, &stripe_len);
-		BUG_ON(ret); /* -ENOMEM */
+		if (ret < 0) {
+			kfree(logical);
+			free_excluded_extents(root, cache);
+			goto out;
+		}
 
 		while (nr--) {
 			cache->bytes_super += stripe_len;
@@ -264,7 +271,9 @@ static int exclude_super_stripes(struct btrfs_root *root,
 
 		kfree(logical);
 	}
-	return 0;
+
+out:
+	return ret;
 }
 
 static struct btrfs_caching_control *
@@ -7665,7 +7674,11 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 		 * info has super bytes accounted for, otherwise we'll think
 		 * we have more space than we actually do.
 		 */
-		exclude_super_stripes(root, cache);
+		ret = exclude_super_stripes(root, cache);
+		if (ret < 0) {
+			kfree(cache);
+			goto error;
+		}
 
 		/*
 		 * check for two cases, either we are full, and therefore
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index f818f48..3400fbd 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -3911,12 +3911,22 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
 	u64 length;
 	u64 stripe_nr;
 	int i, j, nr = 0;
+	int ret = 0;
 
 	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, chunk_start, 1);
 	read_unlock(&em_tree->lock);
 
-	BUG_ON(!em || em->start != chunk_start);
+	if (!em)
+		return -ENOENT;
+
+	if (em->start != chunk_start) {
+		printk(KERN_DEBUG "btrfs: rmap em->start %llu != %llu\n",
+				em->start, chunk_start);
+		ret = -EINVAL;
+		goto out_error;
+	}
+
 	map = (struct map_lookup *)em->bdev;
 
 	length = em->len;
@@ -3926,7 +3936,10 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
 		do_div(length, map->num_stripes);
 
 	buf = kzalloc(sizeof(u64) * map->num_stripes, GFP_NOFS);
-	BUG_ON(!buf); /* -ENOMEM */
+	if (!buf) {
+		ret = -ENOMEM;
+		goto out_error;
+	}
 
 	for (i = 0; i < map->num_stripes; i++) {
 		if (devid && map->stripes[i].dev->devid != devid)
@@ -3960,8 +3973,9 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
 	*naddrs = nr;
 	*stripe_len = map->stripe_len;
 
+out_error:
 	free_extent_map(em);
-	return 0;
+	return ret;
 }
 
 static void btrfs_end_bio(struct bio *bio, int err)
-- 
1.7.9

Re: [patch 33/35] btrfs: handle errors when excluding super extents

[David Sterba]

On Wed, Mar 21, 2012 at 09:11:37PM -0400, Jeff Mahoney wrote:
> --- a/fs/btrfs/extent-tree.c
> +++ b/fs/btrfs/extent-tree.c
> @@ -250,10 +250,17 @@ static int exclude_super_stripes(struct btrfs_root *root,
>  
>  	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
>  		bytenr = btrfs_sb_offset(i);
> +		if (bytenr <= cache->key.objectid + cache->key.offset)
> +			break;

The condition is wrong. Exclude superblock if it's inside the block
group:

             if (cache->key.objectid + cache->key.offset <= bytenr)
                     break;
             if (bytenr < cache->key.objectid)
                     continue;

break if the remaining sb's are too far and skip to next one otherwise.

> +		logical = NULL;
>  		ret = btrfs_rmap_block(&root->fs_info->mapping_tree,
>  				       cache->key.objectid, bytenr,
>  				       0, &logical, &nr, &stripe_len);
> -		BUG_ON(ret); /* -ENOMEM */
> +		if (ret < 0) {
> +			kfree(logical);
> +			free_excluded_extents(root, cache);
> +			goto out;
> +		}
>  
>  		while (nr--) {
>  			cache->bytes_super += stripe_len;

[patch 34/35] btrfs: enhance superblock sanity checks

[Jeff Mahoney]

Validate checksum algorithm during mount and prevent BUG_ON later in
btrfs_super_csum_size.

Signed-off-by: David Sterba <dsterba@suse.cz>
---
 fs/btrfs/disk-io.c |   23 ++++++++++++++++++-----
 1 files changed, 18 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 438993e..38c2ee1 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -48,7 +48,7 @@
 static struct extent_io_ops btree_extent_io_ops;
 static void end_workqueue_fn(struct btrfs_work *work);
 static void free_fs_root(struct btrfs_root *root);
-static void btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
+static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
 				    int read_only);
 static void btrfs_destroy_ordered_operations(struct btrfs_root *root);
 static void btrfs_destroy_ordered_extents(struct btrfs_root *root);
@@ -2135,7 +2135,12 @@ int open_ctree(struct super_block *sb,
 	/* check FS state, whether FS is broken. */
 	fs_info->fs_state |= btrfs_super_flags(disk_super);
 
-	btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
+	ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
+	if (ret) {
+		printk(KERN_ERR "btrfs: superblock contains fatal errors\n");
+		err = ret;
+		goto fail_alloc;
+	}
 
 	/*
 	 * run through our array of backup supers and setup
@@ -3324,15 +3329,23 @@ out:
 	return 0;
 }
 
-static void btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
+static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info,
 			      int read_only)
 {
+	if (btrfs_super_csum_type(fs_info->super_copy) >= ARRAY_SIZE(btrfs_csum_sizes)) {
+		printk(KERN_ERR "btrfs: unsupported checksum algorithm\n");
+		return -EINVAL;
+	}
+
 	if (read_only)
-		return;
+		return 0;
 
-	if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR)
+	if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
 		printk(KERN_WARNING "warning: mount fs with errors, "
 		       "running btrfsck is recommended\n");
+	}
+
+	return 0;
 }
 
 int btrfs_error_commit_super(struct btrfs_root *root)
-- 
1.7.9

[patch 35/35] btrfs: disallow unequal data/metadata blocksize for mixed block groups

[Jeff Mahoney]

With support for bigger metadata blocks, we must avoid mounting a
filesystem with different block size for mixed block groups, this causes
corruption (found by xfstests/083).

Signed-off-by: David Sterba <dsterba@suse.cz>
---
 fs/btrfs/disk-io.c |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 38c2ee1..55272b0 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2302,6 +2302,14 @@ int open_ctree(struct super_block *sb,
 		goto fail_sb_buffer;
 	}
 
+	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
+			(leafsize != nodesize && sectorsize != nodesize)) {
+		printk(KERN_WARNING "btrfs: unequal leaf/node/sector sizes "
+				"are not allowed for mixed block groups on %s\n",
+				sb->s_id);
+		goto fail_sb_buffer;
+	}
+
 	mutex_lock(&fs_info->chunk_mutex);
 	ret = btrfs_read_sys_array(tree_root);
 	mutex_unlock(&fs_info->chunk_mutex);
-- 
1.7.9

Re: [patch 35/35] btrfs: disallow unequal data/metadata blocksize for mixed block groups

[Stefan Behrens]

On 3/22/2012 2:11 AM, Jeff Mahoney wrote:
> With support for bigger metadata blocks, we must avoid mounting a
> filesystem with different block size for mixed block groups, this causes
> corruption (found by xfstests/083).
> 
> Signed-off-by: David Sterba <dsterba@suse.cz>
> ---
>  fs/btrfs/disk-io.c |    8 ++++++++
>  1 files changed, 8 insertions(+), 0 deletions(-)
> 
> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
> index 38c2ee1..55272b0 100644
> --- a/fs/btrfs/disk-io.c
> +++ b/fs/btrfs/disk-io.c
> @@ -2302,6 +2302,14 @@ int open_ctree(struct super_block *sb,
>  		goto fail_sb_buffer;
>  	}
>  
> +	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
> +			(leafsize != nodesize && sectorsize != nodesize)) {

Dave, Shouldn't this be like this (you want all three sizes to be equal
in mixed mode, don't you?):
	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
			(leafsize != nodesize || sectorsize != nodesize)) {


> +		printk(KERN_WARNING "btrfs: unequal leaf/node/sector sizes "
> +				"are not allowed for mixed block groups on %s\n",
> +				sb->s_id);
> +		goto fail_sb_buffer;
> +	}
> +
>  	mutex_lock(&fs_info->chunk_mutex);
>  	ret = btrfs_read_sys_array(tree_root);
>  	mutex_unlock(&fs_info->chunk_mutex);

Re: [patch 35/35] btrfs: disallow unequal data/metadata blocksize for mixed block groups

[David Sterba]

On Thu, Mar 22, 2012 at 05:02:24PM +0100, Stefan Behrens wrote:
> > --- a/fs/btrfs/disk-io.c
> > +++ b/fs/btrfs/disk-io.c
> > @@ -2302,6 +2302,14 @@ int open_ctree(struct super_block *sb,
> >  		goto fail_sb_buffer;
> >  	}
> >  
> > +	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
> > +			(leafsize != nodesize && sectorsize != nodesize)) {
> 
> Dave, Shouldn't this be like this (you want all three sizes to be equal
> in mixed mode, don't you?):
> 	if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) &&
> 			(leafsize != nodesize || sectorsize != nodesize)) {

Yes, you're right, thanks for catching it.


david