[PATCH 1/2] Btrfs: fix a bug in checking whether a inode is already in log

[PATCH 1/2] Btrfs: fix a bug in checking whether a inode is already in log

[Liu Bo]

This is based on Josef's "Btrfs: turbo charge fsync".

The current btrfs checks if an inode is in log by comparing
root's last_log_commit to inode's last_sub_trans[2].

But the problem is that this root->last_log_commit is shared among
inodes.

Say we have N inodes to be logged, after the first inode,
root's last_log_commit is updated and the N-1 remained files will
be skipped.

This fixes the bug by keeping a local copy of root's last_log_commit
inside each inode and this local copy will be maintained itself.

[1]: we regard each log transaction as a subset of btrfs's transaction,
i.e. sub_trans

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
 fs/btrfs/btrfs_inode.h |   14 ++++++--------
 fs/btrfs/inode.c       |    2 ++
 fs/btrfs/transaction.h |    1 +
 fs/btrfs/tree-log.c    |    1 +
 4 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 7c7bf81..ed8ca7c 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -144,6 +144,9 @@ struct btrfs_inode {
 	/* flags field from the on disk inode */
 	u32 flags;
 
+	/* a local copy of root's last_log_commit */
+	unsigned long last_log_commit;
+
 	/*
 	 * Counters to keep track of the number of extent item's we may use due
 	 * to delalloc and such.  outstanding_extents is the number of extent
@@ -203,15 +206,10 @@ static inline bool btrfs_is_free_space_inode(struct inode *inode)
 
 static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
 {
-	struct btrfs_root *root = BTRFS_I(inode)->root;
-	int ret = 0;
-
-	mutex_lock(&root->log_mutex);
 	if (BTRFS_I(inode)->logged_trans == generation &&
-	    BTRFS_I(inode)->last_sub_trans <= root->last_log_commit)
-		ret = 1;
-	mutex_unlock(&root->log_mutex);
-	return ret;
+	    BTRFS_I(inode)->last_sub_trans <= BTRFS_I(inode)->last_log_commit)
+		return 1;
+	return 0;
 }
 
 #endif
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 8879e46e..8fb6cb7 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -6757,6 +6757,7 @@ again:
 
 	BTRFS_I(inode)->last_trans = root->fs_info->generation;
 	BTRFS_I(inode)->last_sub_trans = BTRFS_I(inode)->root->log_transid;
+	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->root->last_log_commit;
 
 	unlock_extent_cached(io_tree, page_start, page_end, &cached_state, GFP_NOFS);
 
@@ -7010,6 +7011,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	ei->csum_bytes = 0;
 	ei->index_cnt = (u64)-1;
 	ei->last_unlink_trans = 0;
+	ei->last_log_commit = 0;
 
 	spin_lock_init(&ei->lock);
 	ei->outstanding_extents = 0;
diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h
index e8b8416..1a138bf 100644
--- a/fs/btrfs/transaction.h
+++ b/fs/btrfs/transaction.h
@@ -88,6 +88,7 @@ static inline void btrfs_set_inode_last_trans(struct btrfs_trans_handle *trans,
 {
 	BTRFS_I(inode)->last_trans = trans->transaction->transid;
 	BTRFS_I(inode)->last_sub_trans = BTRFS_I(inode)->root->log_transid;
+	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->root->last_log_commit;
 }
 
 int btrfs_end_transaction(struct btrfs_trans_handle *trans,
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index e70cdad..03af7f7 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3185,6 +3185,7 @@ next_slot:
 		}
 	}
 	BTRFS_I(inode)->logged_trans = trans->transid;
+	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans;
 out_unlock:
 	mutex_unlock(&BTRFS_I(inode)->log_mutex);
 
-- 
1.7.7.6

[PATCH 2/2] Btrfs: check if an inode has no checksum when logging it

[Liu Bo]

This is based on Josef's "Btrfs: turbo charge fsync".

If an inode is a BTRFS_INODE_NODATASUM one, we don't need to look for csum
items any more.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
 fs/btrfs/tree-log.c |   23 ++++++++++++-----------
 1 files changed, 12 insertions(+), 11 deletions(-)

diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 03af7f7..2df2b7a 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2680,13 +2680,14 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans,
 }
 
 static noinline int copy_items(struct btrfs_trans_handle *trans,
-			       struct btrfs_root *log,
+			       struct inode *inode,
 			       struct btrfs_path *dst_path,
 			       struct extent_buffer *src,
 			       int start_slot, int nr, int inode_only)
 {
 	unsigned long src_offset;
 	unsigned long dst_offset;
+	struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
 	struct btrfs_file_extent_item *extent;
 	struct btrfs_inode_item *inode_item;
 	int ret;
@@ -2695,6 +2696,7 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 	char *ins_data;
 	int i;
 	struct list_head ordered_sums;
+	int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
 
 	INIT_LIST_HEAD(&ordered_sums);
 
@@ -2745,7 +2747,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
 		 * or deletes of this inode don't have to relog the inode
 		 * again
 		 */
-		if (btrfs_key_type(ins_keys + i) == BTRFS_EXTENT_DATA_KEY) {
+		if (btrfs_key_type(ins_keys + i) == BTRFS_EXTENT_DATA_KEY &&
+		    !skip_csum) {
 			int found_type;
 			extent = btrfs_item_ptr(src, start_slot + i,
 						struct btrfs_file_extent_item);
@@ -2873,7 +2876,7 @@ next_slot:
 		    args->start_slot + args->nr == path->slots[0]) {
 			args->nr++;
 		} else if (args->nr) {
-			ret = copy_items(trans, log, dst_path, args->src,
+			ret = copy_items(trans, inode, dst_path, args->src,
 					 args->start_slot, args->nr,
 					 LOG_INODE_ALL);
 			if (ret)
@@ -2910,7 +2913,7 @@ next_slot:
 		}
 
 		if (args->nr) {
-			ret = copy_items(trans, log, dst_path, args->src,
+			ret = copy_items(trans, inode, dst_path, args->src,
 					 args->start_slot, args->nr,
 					 LOG_INODE_ALL);
 			if (ret)
@@ -2930,7 +2933,6 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
 				     struct btrfs_path *dst_path)
 {
 	struct log_args args;
-	struct btrfs_root *log = root->log_root;
 	struct extent_map *em, *n;
 	struct list_head extents;
 	struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree;
@@ -2971,7 +2973,7 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
 		 * our search
 		 */
 		if (args.nr && em->mod_start != args.next_offset) {
-			ret = copy_items(trans, log, dst_path, args.src,
+			ret = copy_items(trans, inode, dst_path, args.src,
 					 args.start_slot, args.nr,
 					 LOG_INODE_ALL);
 			if (ret)
@@ -2984,7 +2986,7 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
 	}
 
 	if (!ret && args.nr)
-		ret = copy_items(trans, log, dst_path, args.src,
+		ret = copy_items(trans, inode, dst_path, args.src,
 				 args.start_slot, args.nr, LOG_INODE_ALL);
 	btrfs_release_path(path);
 	WARN_ON(!list_empty(&extents));
@@ -3109,7 +3111,7 @@ again:
 			goto next_slot;
 		}
 
-		ret = copy_items(trans, log, dst_path, src, ins_start_slot,
+		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
 				 ins_nr, inode_only);
 		if (ret) {
 			err = ret;
@@ -3127,7 +3129,7 @@ next_slot:
 			goto again;
 		}
 		if (ins_nr) {
-			ret = copy_items(trans, log, dst_path, src,
+			ret = copy_items(trans, inode, dst_path, src,
 					 ins_start_slot,
 					 ins_nr, inode_only);
 			if (ret) {
@@ -3148,8 +3150,7 @@ next_slot:
 			break;
 	}
 	if (ins_nr) {
-		ret = copy_items(trans, log, dst_path, src,
-				 ins_start_slot,
+		ret = copy_items(trans, inode, dst_path, src, ins_start_slot,
 				 ins_nr, inode_only);
 		if (ret) {
 			err = ret;
-- 
1.7.7.6