Skip to contentSkip to conversation

nfsd, lease changes for 2.6.37

DISCUSSION J. Bruce Fields 1 participant 1 message Nov 10, 2010

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Please pull from:

git://linux-nfs.org/~bfields/linux.git for-2.6.37

One nfsd 4.1 bugfix, and two more bits of mopup from the lease de-BKL work. (The last of which is really just cleanup--but obvious enough I hoped to slip it past the functional-regressions-only-post-rc1 filter....)

--b.

J. Bruce Fields (3):
nfsd4: fix 4.1 connection registration race
locks: fix leak on merging leases
locks: remove dead lease error-handling code

 fs/locks.c                      |   19 +++++++------------
 fs/nfsd/nfs4state.c             |   16 ++++++++++++----
 include/linux/sunrpc/svc_xprt.h |   18 ++++++++++++++----
 3 files changed, 33 insertions(+), 20 deletions(-)

commit 8896b93f42459b18b145c69d399b62870df48061

Author: J. Bruce Fields <[email protected]>
Date:   Wed Nov 3 18:09:18 2010 -0400

locks: remove dead lease error-handling code

A minor oversight from f7347ce4ee7c65415f84be915c018473e7076f31, "fasync: re-organize fasync entry insertion to allow it under a spinlock": this cleanup-on-error was only needed to handle -ENOMEM. Now that we're preallocating it's unneeded.

commit 3df057ac9afe83c4af84016df3baf3a0eb1d3d33

Author: J. Bruce Fields <[email protected]>
Date:   Wed Nov 3 16:49:44 2010 -0400

locks: fix leak on merging leases

We must also free the passed-in lease in the case it wasn't used because an existing lease was upgrade/downgraded or already existed.

Note the nfsd caller doesn't care because it's fl_change callback returns an error in those cases.

commit 21b75b019983dfa5c2dda588f4b60b4ca69844a4

Author: J. Bruce Fields <[email protected]>
Date:   Tue Oct 26 10:07:17 2010 -0400

nfsd4: fix 4.1 connection registration race

If a connection is closed just after a sequence or create_session is sent over it, we could end up trying to register a callback that will never get called since the xprt is already marked dead.