I sent the first patch in this series the other day, but didn't get any
responses. Since then I've had time to follow up on the client-side part
of this problem, which eventually also pointed out yet another bug on
the server side. There are also a couple of cleanup patches in here too,
and a patch to add some tracepoints that I found useful while diagnosing
this.

With this set on both client and server, I'm now able to run Yongcheng's
test for an hour straight with no stuck locks.

Jeff Layton (7):
  lockd: purge resources held on behalf of nlm clients when shutting
    down
  lockd: remove 2 unused helper functions
  lockd: move struct nlm_wait to lockd.h
  lockd: fix races in client GRANTED_MSG wait logic
  lockd: server should unlock lock if client rejects the grant
  nfs: move nfs_fhandle_hash to common include file
  lockd: add some client-side tracepoints

 fs/lockd/Makefile           |  6 ++-
 fs/lockd/clntlock.c         | 58 +++++++++++---------------
 fs/lockd/clntproc.c         | 42 ++++++++++++++-----
 fs/lockd/host.c             |  1 +
 fs/lockd/svclock.c          | 21 ++++++++--
 fs/lockd/trace.c            |  3 ++
 fs/lockd/trace.h            | 83 +++++++++++++++++++++++++++++++++++++
 fs/nfs/internal.h           | 15 -------
 include/linux/lockd/lockd.h | 29 ++++++-------
 include/linux/nfs.h         | 20 +++++++++
 10 files changed, 200 insertions(+), 78 deletions(-)
 create mode 100644 fs/lockd/trace.c
 create mode 100644 fs/lockd/trace.h

-- 
2.39.2