Use mutex instead of busy wait when installing optimized function.

src/optimizing-compiler-thread.h

Index: src/optimizing-compiler-thread.h
diff --git a/src/optimizing-compiler-thread.h b/src/optimizing-compiler-thread.h
index 8cb5e2dd59dec0e26d81984e8d929d3e613ae8c0..004fce7adac9253c5088dab83e13d46b0d3e2b60 100644
--- a/src/optimizing-compiler-thread.h
+++ b/src/optimizing-compiler-thread.h
@@ -50,6 +50,7 @@ class OptimizingCompilerThread : public Thread {
       isolate_(isolate),
       stop_semaphore_(OS::CreateSemaphore(0)),
       input_queue_semaphore_(OS::CreateSemaphore(0)),
+      install_mutex_(OS::CreateMutex()),
       time_spent_compiling_(0),
       time_spent_total_(0) {
     NoBarrier_Store(&stop_thread_, static_cast<AtomicWord>(false));
@@ -95,6 +96,7 @@ class OptimizingCompilerThread : public Thread {
   Semaphore* input_queue_semaphore_;
   UnboundQueue<OptimizingCompiler*> input_queue_;
   UnboundQueue<OptimizingCompiler*> output_queue_;
+  Mutex* install_mutex_;
   volatile AtomicWord stop_thread_;
   volatile Atomic32 queue_length_;
   int64_t time_spent_compiling_;