[PATCH] v9fs: handle async processing of F_SETLK with FL_SLEEP flag

[Vasily Averin]

kernel export thread (nfsd/lockd/ksmbd) uses F_SETLK cmd with the FL_SLEEP
flag set to request asynchronous processing of blocking locks.

Currently v9fs does not support such requests and calls blocking
locks_lock_file_wait() function.

To work around the problem let's detect such request, drop FL_SLEEP
before execution of potentially blocking functions.

Dropped FL_SLEEP flag should be restored back because some calling
function (nfsd4_lock) require it.

https://bugzilla.kernel.org/show_bug.cgi?id=215383
Signed-off-by: Vasily Averin <vvs@xxxxxxxxxxxxx>
---
 fs/9p/vfs_file.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index 612e297f3763..81c98afdbb32 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -135,6 +135,7 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 	int res = 0;
 	unsigned char fl_type;
 	struct v9fs_session_info *v9ses;
+	bool async = false;
 
 	fid = filp->private_data;
 	BUG_ON(fid == NULL);
@@ -142,6 +143,10 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 	if ((fl->fl_flags & FL_POSIX) != FL_POSIX)
 		BUG();
 
+	async = (fl->fl_flags & FL_SLEEP) && IS_SETLK(cmd);
+	if (async)
+		fl->fl_flags &= ~FL_SLEEP;
+
 	res = locks_lock_file_wait(filp, fl);
 	if (res < 0)
 		goto out;
@@ -230,6 +235,8 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 	if (flock.client_id != fid->clnt->name)
 		kfree(flock.client_id);
 out:
+	if (async)
+		fl->fl_flags |= FL_SLEEP;
 	return res;
 }
 
-- 
2.25.1