[Patch] ARM define atomic_exchange_acq/atomic_exchange_rel to __atomic_exchange_n

[Dinar Temirbulatov <dinar@kugelworks.com>]

[-- Attachment #1: Type: text/plain, Size: 457 bytes --]

Hello,
Following patch redefines atomic_exchange_acq/atomic_exchange_rel to
__atomic_exchange_n for ARM, that allows for example to reduce number
of instruction sequence for lll_unlock
from:
ldex, cmp, bne, stex, cmp, bne
to
ldex, stex, cmp, bne
, more on the issue here:
http://sourceware.org/bugzilla/show_bug.cgi?id=15640

This patch was tested on ARM a9 with glibc testsuite with no new
regressions. OK to commit?

                       Thanks, Dinar.

[-- Attachment #2: arm_atomic.patch --]
[-- Type: application/octet-stream, Size: 1131 bytes --]

diff --git a/ports/sysdeps/arm/bits/atomic.h b/ports/sysdeps/arm/bits/atomic.h
index 6e20df7..78d4a6e 100644
--- a/ports/sysdeps/arm/bits/atomic.h
+++ b/ports/sysdeps/arm/bits/atomic.h
@@ -79,3 +79,26 @@ void __arm_link_error (void);
 # define __arm_assisted_compare_and_exchange_val_32_acq(mem, newval, oldval) \
   ({ __arm_link_error (); oldval; })
 #endif
+
+#if __GNUC_PREREQ (4, 7)
+# define atomic_exchange_acq(mem, value)                                \
+  __atomic_val_bysize (__arch_exchange, int, mem, value, __ATOMIC_ACQUIRE)
+
+# define atomic_exchange_rel(mem, value)                                \
+  __atomic_val_bysize (__arch_exchange, int, mem, value, __ATOMIC_RELEASE)
+
+/* Atomic exchange (without compare).  */
+
+# define __arch_exchange_8_int(mem, newval, model)      \
+  (abort (), (typeof(*mem)) 0)
+
+# define __arch_exchange_16_int(mem, newval, model)     \
+  (abort (), (typeof(*mem)) 0)
+
+# define __arch_exchange_32_int(mem, newval, model)     \
+  __atomic_exchange_n (mem, newval, model)
+
+# define __arch_exchange_64_int(mem, newval, model)     \
+  (abort (), (typeof(*mem)) 0)
+
+#endif