[PATCH -tip v3 0/5] locking: Use acquire/release semantics

[Davidlohr Bueso <dave@stgolabs.net>]

Changes from v2:
 o Redid changelog for patches 2-5 (tglx)

Changes from v1:
 o Fix bogus acquire in generic rwsem lock downgrade (Linus)
 o Fix bogus acquire in unlock_rt_mutex_safe() (tglx)

Hi,

This series continues porting users to Will's new _{acquire|release|relaxed}
optimizations for weakly ordered archs -- and in practice, arm64 being the only
actual user we have, thus could use a _lot_ more testing. While we have osq in
-tip recently, along with Waiman's one for native qspinlocks, the following
patches pretty much complete the move over for all things kernel/locking/*.
Of course, particular primitives can be found in the obvious patch titles. Note
that patch 1 adds the necessary machinery to fully support some locks, such as
rwsems.

Compiles on x86 and x86-64.

Thanks!

Davidlohr Bueso (5):
  asm-generic: Add _{relaxed|acquire|release}() variants for inc/dec
    atomics
  locking/mutex: Use acquire/release semantics
  locking/rtmutex: Use acquire/release semantics
  locking/mcs: Use acquire/release semantics
  locking/rwsem: Use acquire/release semantics

 include/asm-generic/atomic-long.h | 29 +++++++-----
 include/asm-generic/mutex-dec.h   |  8 ++--
 include/asm-generic/mutex-xchg.h  | 10 ++--
 include/asm-generic/rwsem.h       | 21 ++++++---
 include/linux/atomic.h            | 97 +++++++++++++++++++++++++++++++++++++++
 kernel/locking/mcs_spinlock.h     |  4 +-
 kernel/locking/mutex.c            |  9 ++--
 kernel/locking/rtmutex.c          | 30 ++++++++----
 kernel/locking/rwsem-xadd.c       |  5 +-
 9 files changed, 168 insertions(+), 45 deletions(-)

--
2.1.4