MIPS: Replace smp_mb with release barrier function in unlocks.

Repleace smp_mb() in arch_write_unlock() and __clear_bit_unlock() to
smp_mb__before_llsc() call which does "release" barrier functionality.

It seems like it was missed in commit f252ffd5
during introduction of "acquire" and "release" semantics.

[ralf@linux-mips: The original patch submission was labelled a fix but
actually it replaces a barrier with another less restrictive type of
barrier so it doesn't fix any ill behaviour but rather squeezes out a
tad better performance.  Further improvments will be possible once
smp_release() has been merged.]
Signed-off-by: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: benh@kernel.crashing.org
Cc: will.deacon@arm.com
Cc: linux-kernel@vger.kernel.org
Cc: markos.chandras@imgtec.com
Cc: macro@linux-mips.org
Cc: Steven.Hill@imgtec.com
Cc: alexander.h.duyck@redhat.com
Cc: davem@davemloft.net
Patchwork: https://patchwork.linux-mips.org/patch/10507/Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

arch/mips/include/asm/bitops.h

@@ -469,7 +469,7 @@ static inline int test_and_change_bit(unsigned long nr,
  */
 static inline void __clear_bit_unlock(unsigned long nr, volatile unsigned long *addr)
 {
-	smp_mb();
+	smp_mb__before_llsc();
 	__clear_bit(nr, addr);
 }
 

arch/mips/include/asm/spinlock.h

@@ -317,7 +317,7 @@ static inline void arch_write_lock(arch_rwlock_t *rw)
 
 static inline void arch_write_unlock(arch_rwlock_t *rw)
 {
-	smp_mb();
+	smp_mb__before_llsc();
 
 	__asm__ __volatile__(
 	"				# arch_write_unlock	\n"