From: Miao Xie Date: Tue, 22 Jan 2013 10:50:35 +0000 (+0000) Subject: Btrfs: traverse and flush the delalloc inodes once X-Git-Url: https://git.karo-electronics.de/?a=commitdiff_plain;h=63607cc86ab808e077a895be2d7895dcbf6f4cf3;p=linux-beck.git Btrfs: traverse and flush the delalloc inodes once btrfs_start_delalloc_inodes() needn't traverse and flush the delalloc inodes repeatedly. It is because we can regard the data that the users write after we start delalloc inodes flush as the one which is after the delalloc inodes flush is done, and we can flush it next time. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 280e4d487636..d1e3a2d15f91 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7614,7 +7614,7 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput) INIT_LIST_HEAD(&works); INIT_LIST_HEAD(&splice); -again: + spin_lock(&root->fs_info->delalloc_lock); list_splice_init(&root->fs_info->delalloc_inodes, &splice); while (!list_empty(&splice)) { @@ -7650,13 +7650,6 @@ again: btrfs_wait_and_free_delalloc_work(work); } - spin_lock(&root->fs_info->delalloc_lock); - if (!list_empty(&root->fs_info->delalloc_inodes)) { - spin_unlock(&root->fs_info->delalloc_lock); - goto again; - } - spin_unlock(&root->fs_info->delalloc_lock); - /* the filemap_flush will queue IO into the worker threads, but * we have to make sure the IO is actually started and that * ordered extents get created before we return