H2H Async memcpy fix

In this change, the cpu memcpy will wait until all the commands in the current stream are done.
Note that, it only waits on current stream. But not on other streams.

hipamd/src/hip_hcc.cpp

@@ -2405,6 +2405,11 @@ hipError_t hipMemcpyAsync(void* dst, const void* src, size_t sizeBytes, hipMemcp
         } else if (kind == hipMemcpyHostToHost) {
             tprintf (TRACE_COPY2, "H2H copy with memcpy");
 
+			/* As this is a CPU op, we need to wait until all
+			the commands in current stream are finished.
+			*/
+			stream->wait();
+
             memcpy(dst, src, sizeBytes);
 
         } else {