summaryrefslogtreecommitdiff
path: root/fs/ceph
diff options
context:
space:
mode:
authorJeff Layton <jlayton@kernel.org>2021-09-10 22:36:29 +0300
committerJeff Layton <jlayton@kernel.org>2021-09-10 23:21:44 +0300
commit90f7d7a0d0d68623b5f7df5621a8d54d9518fcc4 (patch)
tree45d9848be89c64360b82aa49fbb19af46eb46164 /fs/ceph
parentbf9f243f23e6623f310ba03fbb14e10ec3a61290 (diff)
downloadlinux-90f7d7a0d0d68623b5f7df5621a8d54d9518fcc4.tar.xz
locks: remove LOCK_MAND flock lock support
As best I can tell, the logic for these has been broken for a long time (at least before the move to git), such that they never conflict with anything. Also, nothing checks for these flags and prevented opens or read/write behavior on the files. They don't seem to do anything. Given that, we can rip these symbols out of the kernel, and just make flock(2) return 0 when LOCK_MAND is set in order to preserve existing behavior. Cc: Matthew Wilcox <willy@infradead.org> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Jeff Layton <jlayton@kernel.org>
Diffstat (limited to 'fs/ceph')
-rw-r--r--fs/ceph/locks.c3
1 files changed, 0 insertions, 3 deletions
diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c
index bdeb271f47d9..d8c31069fbf2 100644
--- a/fs/ceph/locks.c
+++ b/fs/ceph/locks.c
@@ -302,9 +302,6 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
if (!(fl->fl_flags & FL_FLOCK))
return -ENOLCK;
- /* No mandatory locks */
- if (fl->fl_type & LOCK_MAND)
- return -EOPNOTSUPP;
dout("ceph_flock, fl_file: %p\n", fl->fl_file);