[PATCH 0/4] locks: locks related cleanups for v4.1

[Jeff Layton <jeff.layton@primarydata.com>]

Nothing terribly earth-shattering here. Just a few locking-related
cleanup patches that I'm planning to queue up for v4.1. The only one
that really warrants close attention is the 4th one that gets rid of the
i_lock protection around the i_flctx pointer.

One of the earlier patchsets that added the i_flctx field used cmpxchg
to assign it, but I ended up taking that out due to a problem that Sasha
Levin reported. I now think that I misunderstood the problem and that
using cmpxchg for that should be ok.

Jeff Layton (4):
  locks: don't allocate a lock context for an F_UNLCK request
  locks: change lm_get_owner and lm_put_owner prototypes
  locks: get rid of WE_CAN_BREAK_LSLK_NOW dead code
  locks: use cmpxchg to assign i_flctx pointer

 fs/locks.c          | 45 +++++++++++++++++++--------------------------
 fs/nfsd/nfs4state.c | 18 ++++++++++--------
 include/linux/fs.h  |  4 ++--
 3 files changed, 31 insertions(+), 36 deletions(-)

-- 
2.1.0