Patch is wrong, I missed that FL_FLOCK are handled here too. On 27.12.2021 18:50, Vasily Averin wrote: > nfsd and lockd use F_SETLK cmd with the FL_SLEEP flag set to request > asynchronous processing of blocking locks. > > Currently nfs mounted with 'local_lock' option use locks_lock_file_wait() > function blocked on such requests. > > To handle such requests properly, non-blocking posix_file_lock() > function should be used instead. > > https://bugzilla.kernel.org/show_bug.cgi?id=215383 > Signed-off-by: Vasily Averin <vvs@xxxxxxxxxxxxx> > --- > v2: fixed 'fl_flags && FL_SLEEP' => 'fl_flags & FL_SLEEP' > --- > fs/nfs/file.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/fs/nfs/file.c b/fs/nfs/file.c > index 24e7dccce355..38e1821cff5d 100644 > --- a/fs/nfs/file.c > +++ b/fs/nfs/file.c > @@ -769,9 +769,11 @@ do_setlk(struct file *filp, int cmd, struct file_lock *fl, int is_local) > */ > if (!is_local) > status = NFS_PROTO(inode)->lock(filp, cmd, fl); > + else if ((fl->fl_flags & FL_SLEEP) && IS_SETLK(cmd)) > + status = posix_lock_file(filp, fl, NULL); > else > status = locks_lock_file_wait(filp, fl); > - if (status < 0) > + if (status) > goto out; > > /* >
