[lttng-dev] [PATCH 03/11] urcu/compiler: Use atomic builtins if configured

[Olivier Dion]

Use __atomic_signal_fence(__ATOMIC_SEQ_CST) for cmm_barrier() if
configured to use atomic builtins.

Change-Id: Ib168b50f1e97a8da861b92d6882c56db230ebb2c
Co-authored-by: Mathieu Desnoyers <mathieu.desnoyers at efficios.com>
Signed-off-by: Olivier Dion <odion at efficios.com>
---
 include/urcu/compiler.h | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/include/urcu/compiler.h b/include/urcu/compiler.h
index 2f32b38..3604488 100644
--- a/include/urcu/compiler.h
+++ b/include/urcu/compiler.h
@@ -25,10 +25,16 @@
 # include <type_traits>	/* for std::remove_cv */
 #endif
 
+#include <urcu/config.h>
+
 #define caa_likely(x)	__builtin_expect(!!(x), 1)
 #define caa_unlikely(x)	__builtin_expect(!!(x), 0)
 
-#define	cmm_barrier()	__asm__ __volatile__ ("" : : : "memory")
+#ifdef CONFIG_RCU_USE_ATOMIC_BUILTINS
+#  define cmm_barrier() __atomic_signal_fence(__ATOMIC_SEQ_CST)
+#else
+#  define cmm_barrier() asm volatile ("" : : : "memory")
+#endif
 
 /*
  * Instruct the compiler to perform only a single access to a variable
-- 
2.39.2