Skip to contentSkip to conversation

btrfs: Rename btrfs_join_transaction_nolock

PATCH In review Nikolay Borisov 3 participants 3 messages Oct 8, 2019 – Oct 10, 2019 Btrfs file system

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 btrfs: Rename btrfs_join_transaction_nolock

This function is used only during the final phase of freespace cache writeout. This is necessary since using the plain btrfs_join_transaction api is deadlock prone. The deadlock looks like:

T1:
btrfs_commit_Transaction
commit_cowonly_roots
btrfs_write_dirty_block_groups
btrfs_wait_cache_io
__btrfs_wait_cache_io
btrfs_wait_ordered_range <-- Triggers ordered IO for freespace inode and blocks transaction commit until freespace cache writeout.

T2: <-- after T1 has triggered the writeout finish_ordered_fn
btrfs_finish_ordered_io
btrfs_join_transaction <--- this would block waiting for current transaction to commit, but since trans commit is waiting for this writeout to finish.

The special purpose functions prevents it by simply skipping the "wait for writeout" since it's guaranteed the transaction won't proceed until we are done.

---
 fs/btrfs/inode.c       | 12 ++++++------
 fs/btrfs/transaction.c |  2 +-
 fs/btrfs/transaction.h |  2 +-
 3 files changed, 8 insertions(+), 8 deletions(-)
fs/btrfs/inode.c +6 -6
index 16415815217c..1c6f7c0a2b1d 100644
@@ -3053,7 +3053,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	int compress_type = 0;
 	int ret = 0;
 	u64 logical_len = ordered_extent->len;
-	bool nolock;
+	bool freespace_inode;
 	bool truncated = false;
 	bool range_locked = false;
 	bool clear_new_delalloc_bytes = false;
@@ -3064,7 +3064,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	    !test_bit(BTRFS_ORDERED_DIRECT, &ordered_extent->flags))
 		clear_new_delalloc_bytes = true;
 
-	nolock = btrfs_is_free_space_inode(BTRFS_I(inode));
+	freespace_inode = btrfs_is_free_space_inode(BTRFS_I(inode));
 
 	if (test_bit(BTRFS_ORDERED_IOERR, &ordered_extent->flags)) {
 		ret = -EIO;
@@ -3095,8 +3095,8 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		btrfs_qgroup_free_data(inode, NULL, ordered_extent->file_offset,
 				       ordered_extent->len);
 		btrfs_ordered_update_i_size(inode, 0, ordered_extent);
-		if (nolock)
-			trans = btrfs_join_transaction_nolock(root);
+		if (freespace_inode)
+			trans = btrfs_join_transaction_spacecache(root);
 		else
 			trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans)) {
@@ -3130,8 +3130,8 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 			EXTENT_DEFRAG, 0, 0, &cached_state);
 	}
 
-	if (nolock)
-		trans = btrfs_join_transaction_nolock(root);
+	if (freespace_inode)
+		trans = btrfs_join_transaction_spacecache(root);
 	else
 		trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans)) {
fs/btrfs/transaction.c +1 -1
index 54b8718054ce..6f133906c862 100644
@@ -729,7 +729,7 @@ struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root)
 				 true);
 }
 
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root)
+struct btrfs_trans_handle *btrfs_join_transaction_spacecache(struct btrfs_root *root)
 {
 	return start_transaction(root, 0, TRANS_JOIN_NOLOCK,
 				 BTRFS_RESERVE_NO_FLUSH, true);
fs/btrfs/transaction.h +1 -1
index 2ac89fb0d709..49f7196368f5 100644
@@ -183,7 +183,7 @@ struct btrfs_trans_handle *btrfs_start_transaction_fallback_global_rsv(
 					unsigned int num_items,
 					int min_factor);
 struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root);
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root);
+struct btrfs_trans_handle *btrfs_join_transaction_spacecache(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_join_transaction_nostart(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_attach_transaction(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_attach_transaction_barrier(
Signature · 1 line
2.17.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 21 lines · Nikolay Borisov

This function is used only during the final phase of freespace cache writeout. This is necessary since using the plain btrfs_join_transaction api is deadlock prone. The deadlock looks like:

T1:
btrfs_commit_Transaction
commit_cowonly_roots
btrfs_write_dirty_block_groups
btrfs_wait_cache_io
__btrfs_wait_cache_io
btrfs_wait_ordered_range <-- Triggers ordered IO for freespace inode and blocks transaction commit until freespace cache writeout.

T2: <-- after T1 has triggered the writeout finish_ordered_fn
btrfs_finish_ordered_io
btrfs_join_transaction <--- this would block waiting for current transaction to commit, but since trans commit is waiting for this writeout to finish.

Quoted 5 lines

The special purpose functions prevents it by simply skipping the "wait for writeout" since it's guaranteed the transaction won't proceed until we are done.

Indeed the old _nolock() is pretty confusing. So the rename makes a lot of sense.

Thanks,
Qu

Quoted 77 lines
---
 fs/btrfs/inode.c       | 12 ++++++------
 fs/btrfs/transaction.c |  2 +-
 fs/btrfs/transaction.h |  2 +-
 3 files changed, 8 insertions(+), 8 deletions(-)
fs/btrfs/inode.c +6 -6
index 16415815217c..1c6f7c0a2b1d 100644
@@ -3053,7 +3053,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	int compress_type = 0;
 	int ret = 0;
 	u64 logical_len = ordered_extent->len;
-	bool nolock;
+	bool freespace_inode;
 	bool truncated = false;
 	bool range_locked = false;
 	bool clear_new_delalloc_bytes = false;
@@ -3064,7 +3064,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	    !test_bit(BTRFS_ORDERED_DIRECT, &ordered_extent->flags))
 		clear_new_delalloc_bytes = true;
 
-	nolock = btrfs_is_free_space_inode(BTRFS_I(inode));
+	freespace_inode = btrfs_is_free_space_inode(BTRFS_I(inode));
 
 	if (test_bit(BTRFS_ORDERED_IOERR, &ordered_extent->flags)) {
 		ret = -EIO;
@@ -3095,8 +3095,8 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 		btrfs_qgroup_free_data(inode, NULL, ordered_extent->file_offset,
 				       ordered_extent->len);
 		btrfs_ordered_update_i_size(inode, 0, ordered_extent);
-		if (nolock)
-			trans = btrfs_join_transaction_nolock(root);
+		if (freespace_inode)
+			trans = btrfs_join_transaction_spacecache(root);
 		else
 			trans = btrfs_join_transaction(root);
 		if (IS_ERR(trans)) {
@@ -3130,8 +3130,8 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 			EXTENT_DEFRAG, 0, 0, &cached_state);
 	}
 
-	if (nolock)
-		trans = btrfs_join_transaction_nolock(root);
+	if (freespace_inode)
+		trans = btrfs_join_transaction_spacecache(root);
 	else
 		trans = btrfs_join_transaction(root);
 	if (IS_ERR(trans)) {
fs/btrfs/transaction.c +1 -1
index 54b8718054ce..6f133906c862 100644
@@ -729,7 +729,7 @@ struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root)
 				 true);
 }
 
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root)
+struct btrfs_trans_handle *btrfs_join_transaction_spacecache(struct btrfs_root *root)
 {
 	return start_transaction(root, 0, TRANS_JOIN_NOLOCK,
 				 BTRFS_RESERVE_NO_FLUSH, true);
fs/btrfs/transaction.h +1 -1
index 2ac89fb0d709..49f7196368f5 100644
@@ -183,7 +183,7 @@ struct btrfs_trans_handle *btrfs_start_transaction_fallback_global_rsv(
 					unsigned int num_items,
 					int min_factor);
 struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root);
-struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root);
+struct btrfs_trans_handle *btrfs_join_transaction_spacecache(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_join_transaction_nostart(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_attach_transaction(struct btrfs_root *root);
 struct btrfs_trans_handle *btrfs_attach_transaction_barrier(
#2 lore raw
In reply to Nikolay Borisov
Quoted 21 lines · Nikolay Borisov

This function is used only during the final phase of freespace cache writeout. This is necessary since using the plain btrfs_join_transaction api is deadlock prone. The deadlock looks like:

T1:
btrfs_commit_Transaction
commit_cowonly_roots
btrfs_write_dirty_block_groups
btrfs_wait_cache_io
__btrfs_wait_cache_io
btrfs_wait_ordered_range <-- Triggers ordered IO for freespace inode and blocks transaction commit until freespace cache writeout.

T2: <-- after T1 has triggered the writeout finish_ordered_fn
btrfs_finish_ordered_io
btrfs_join_transaction <--- this would block waiting for current transaction to commit, but since trans commit is waiting for this writeout to finish.

Quoted 5 lines

The special purpose functions prevents it by simply skipping the "wait for writeout" since it's guaranteed the transaction won't proceed until we are done.

Added to misc-next, thanks.