lockref: use arch_mutex_cpu_relax() in CMPXCHG_LOOP()
authorHeiko Carstens <heiko.carstens@de.ibm.com>
Mon, 23 Sep 2013 10:59:56 +0000 (12:59 +0200)
committerHeiko Carstens <heiko.carstens@de.ibm.com>
Sat, 28 Sep 2013 10:46:24 +0000 (12:46 +0200)
commit491f6f8e5fd9a57aaf03b6d6e3e153f1c27d8a46
tree81cb77f31a841e87cecb9329f26e6524cc44f3a2
parent083986e8248d978b6c961d3da6beb0c921c68220
lockref: use arch_mutex_cpu_relax() in CMPXCHG_LOOP()

Make use of arch_mutex_cpu_relax() so architectures can override the
default cpu_relax() semantics.
This is especially useful for s390, where cpu_relax() means that we
yield() the current (virtual) cpu and therefore is very expensive,
and would contradict the whole purpose of the lockless cmpxchg loop.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
lib/lockref.c