Add system specific 'relax cpu' instruction.

lib/pls/include/pls/internal/base/system_details.h

@@ -24,9 +24,28 @@ constexpr std::uintptr_t CACHE_LINE_SIZE = 64;
  * Choose one of the following ways to store thread specific data.
  * Try to choose the fastest available on this processor/system.
  */
-//                #define PLS_THREAD_SPECIFIC_PTHREAD
+//#define PLS_THREAD_SPECIFIC_PTHREAD
 #define PLS_THREAD_SPECIFIC_COMPILER
 
+/**
+ * When spinning one wants to 'relax' the CPU from some task,
+ * e.g. disabling speculative execution/branch prediction
+ * or reducing its clock speed.
+ * This is both good for power draw, as well as for hyperthreading.
+ *
+ * Choose the implementation appropriate for your compiler-cpu combination.
+ */
+#if (COMPILER == MVCC)
+#include <emmintrin.h>
+inline void relax_cpu() {
+  _mm_pause();
+}
+#elif (COMPILER == GCC || COMPILER == LLVM)
+inline void relax_cpu() {
+  asm("pause");
+}
+#endif
+
 }
 }
 }