And use mutex instead. You may never want to do this.
It is a workaround to run c++11 on fishtest where many
machiens have HTenabled and this can be a problem when
number of cores set is higher than number of physical cores.
To disable spinlocks, just compile with -DNO_SPINLOCK flag
No functional change.
const size_t MAX_SPLITPOINTS_PER_THREAD = 8;
const size_t MAX_SLAVES_PER_SPLITPOINT = 4;
-
+#if !defined(NO_SPINLOCK)
/// Spinlock class wraps low level atomic operations to provide a spin lock
class Spinlock {
void release() { lock.store(1, std::memory_order_release); }
};
+#else
+
+class Spinlock {
+
+ std::mutex mutex;
+
+public:
+ void acquire() { mutex.lock(); }
+ void release() { mutex.unlock(); }
+};
+
+#endif
/// SplitPoint struct stores information shared by the threads searching in
/// parallel below the same split point. It is populated at splitting time.
/// -DNO_PREFETCH | Disable use of prefetch asm-instruction. You may need this to
/// | run on some very old machines.
///
+/// -DNO_SPINLOCK | Use mutex instead of spinlocks. This is much slower, so you
+/// | really don't want to do this in general case.
+///
/// -DUSE_POPCNT | Add runtime support for use of popcnt asm-instruction. Works
/// | only in 64-bit mode and requires hardware with popcnt support.
///