Btrfs: improve fsync by filtering extents that we want

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

The above Josef's patch performs very good in random sync write test,
because we won't have too much extents to merge.

However, it does not performs good on the test:
dd if=/dev/zero of=foobar bs=4k count=12500 oflag=sync

The reason is when we do sequencial sync write, we need to merge the
current extent just with the previous one, so that we can get accumulated
extents to log:

A(4k) --> AA(8k) --> AAA(12k) --> AAAA(16k) ...

So we'll have to flush more and more checksum into log tree, which is the
bottleneck according to my tests.

But we can avoid this by telling fsync the real extents that are needed
to be logged.

With this, I did the above dd sync write test (size=50m),

         w/o (orig)   w/ (josef's)   w/ (this)
SATA      104KB/s       109KB/s       121KB/s
ramdisk   1.5MB/s       1.5MB/s       10.7MB/s (613%)

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>

1 files changed, 3 insertions, 3 deletions

diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 58075d711d24..71e71539ffb7 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2833,8 +2833,8 @@ static int log_one_extent(struct btrfs_trans_handle *trans,
         struct btrfs_root *log = root->log_root;
         struct btrfs_file_extent_item *fi;
         struct btrfs_key key;
-        u64 start = em->start;
-        u64 len = em->len;
+        u64 start = em->mod_start;
+        u64 len = em->mod_len;
         u64 num_bytes;
         int nritems;
         int ret;
@@ -2970,7 +2970,7 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
                  * sequential then we need to copy the items we have and redo
                  * our search
                  */
-                if (args.nr && em->start != args.next_offset) {
+                if (args.nr && em->mod_start != args.next_offset) {
                         ret = copy_items(trans, log, dst_path, args.src,
                                          args.start_slot, args.nr,
                                          LOG_INODE_ALL);