arch/tile: use atomic exchange in arch_write_unlock()

This idiom is used elsewhere when we do an unlock by writing a zero,
but I missed it here.  Using an atomic operation avoids waiting
on the write buffer for the unlocking write to be sent to the home cache.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

diff --git a/arch/tile/include/asm/spinlock_64.h b/arch/tile/include/asm/spinlock_64.h
index 72be5904e020e7df9b4923edb92bf25fbf8670a5..5f8b6a095fd84a0748ea645c96a4b3908ba6d6fb 100644 (file)
--- a/arch/tile/include/asm/spinlock_64.h
+++ b/arch/tile/include/asm/spinlock_64.h
@@ -137,7 +137,7 @@ static inline void arch_read_unlock(arch_rwlock_t *rw)
 static inline void arch_write_unlock(arch_rwlock_t *rw)
 {
        __insn_mf();
-       rw->lock = 0;
+       __insn_exch4(&rw->lock, 0);  /* Avoid waiting in the write buffer. */
 }
 
 static inline int arch_read_trylock(arch_rwlock_t *rw)