diff mbox series

fuse: cache readdir calls if filesystem opts out of opendir

Message ID 20190129003434.316288-1-chadaustin@fb.com (mailing list archive)
State New, archived
Headers show
Series fuse: cache readdir calls if filesystem opts out of opendir | expand

Commit Message

Chad Austin Jan. 29, 2019, 12:34 a.m. UTC 
If a filesystem returns ENOSYS from opendir and thus opts out of
opendir and releasedir requests, it almost certainly would also like
readdir results cached. Default open_flags to FOPEN_KEEP_CACHE and
FOPEN_CACHE_DIR in that case.

With this patch, I've measured recursive directory enumeration across
large FUSE mounts to be faster than native mounts.

Signed-off-by: Chad Austin <chadaustin@fb.com>
---
 fs/fuse/file.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

Miklos Szeredi Feb. 4, 2019, 7:43 a.m. UTC | #1 
On Tue, Jan 29, 2019 at 1:34 AM Chad Austin <chadaustin@fb.com> wrote:
>
> If a filesystem returns ENOSYS from opendir and thus opts out of
> opendir and releasedir requests, it almost certainly would also like
> readdir results cached. Default open_flags to FOPEN_KEEP_CACHE and
> FOPEN_CACHE_DIR in that case.
>
> With this patch, I've measured recursive directory enumeration across
> large FUSE mounts to be faster than native mounts.
>
> Signed-off-by: Chad Austin <chadaustin@fb.com>

Thanks, applied.

Miklos
diff mbox series

Patch

diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index cc6ffd23b80f..06096b60f1df 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -124,7 +124,8 @@  int fuse_do_open(struct fuse_conn *fc, u64 nodeid, struct file *file,
 		return -ENOMEM;
 
 	ff->fh = 0;
-	ff->open_flags = FOPEN_KEEP_CACHE; /* Default for no-open */
+	/* Default for no-open */
+	ff->open_flags = FOPEN_KEEP_CACHE | (isdir ? FOPEN_CACHE_DIR : 0);
 	if (isdir ? !fc->no_opendir : !fc->no_open) {
 		struct fuse_open_out outarg;
 		int err;