diff options
author | David Sterba <dsterba@suse.com> | 2017-01-06 13:27:00 +0100 |
---|---|---|
committer | David Sterba <dsterba@suse.com> | 2017-02-14 15:50:51 +0100 |
commit | 95995dbbe691c78cf78948e48b9778531bf1fb3c (patch) | |
tree | 5e267a315717c49bed416513c98a01715da456af /fs/btrfs/inode.c | |
parent | 26d30f85290723613a32e8f1d02504352502a224 (diff) | |
download | op-kernel-dev-95995dbbe691c78cf78948e48b9778531bf1fb3c.zip op-kernel-dev-95995dbbe691c78cf78948e48b9778531bf1fb3c.tar.gz |
btrfs: remove unused logic of limiting async delalloc pages
A proposed patch in https://marc.info/?l=linux-btrfs&m=147859791003837
pointed out bad limit threshold in cow_file_range_async, but it turned
out that the whole logic is not necessary and is done by writeback. We
agreed to remove it.
Signed-off-by: David Sterba <dsterba@suse.com>
Diffstat (limited to 'fs/btrfs/inode.c')
-rw-r--r-- | fs/btrfs/inode.c | 7 |
1 files changed, 0 insertions, 7 deletions
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index c0e982c..39bb0bb 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1167,7 +1167,6 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page, struct btrfs_root *root = BTRFS_I(inode)->root; unsigned long nr_pages; u64 cur_end; - int limit = 10 * SZ_1M; clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED, 1, 0, NULL, GFP_NOFS); @@ -1199,12 +1198,6 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page, btrfs_queue_work(fs_info->delalloc_workers, &async_cow->work); - if (atomic_read(&fs_info->async_delalloc_pages) > limit) { - wait_event(fs_info->async_submit_wait, - (atomic_read(&fs_info->async_delalloc_pages) < - limit)); - } - while (atomic_read(&fs_info->async_submit_draining) && atomic_read(&fs_info->async_delalloc_pages)) { wait_event(fs_info->async_submit_wait, |