On 2 Feb 2020, at 17:53, Trond Myklebust wrote:
> Now that the page cache locking is repaired, we should be able to
> switch to using iterate_shared() for improved concurrency when
> doing readdir().
>
> Signed-off-by: Trond Myklebust <trond.myklebust@xxxxxxxxxxxxxxx>
> ---
> fs/nfs/dir.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
> index 60387dec9032..803e6fec0266 100644
> --- a/fs/nfs/dir.c
> +++ b/fs/nfs/dir.c
> @@ -58,7 +58,7 @@ static void nfs_readdir_clear_array(struct page*);
> const struct file_operations nfs_dir_operations = {
> .llseek = nfs_llseek_dir,
> .read = generic_read_dir,
> - .iterate = nfs_readdir,
> + .iterate_shared = nfs_readdir,
> .open = nfs_opendir,
> .release = nfs_closedir,
> .fsync = nfs_fsync_dir,
Jeez, it makes a lot of sense just to use the page lock. Wish I could have
thought of that a long time ago. Thanks for this, we'll send it through
some testing, but it looks good to me.
Reviewed-by: Benjamin Coddington <bcodding@xxxxxxxxxx>
Ben
