Fully initialize GPU agents before loading tools.

Required for debug agent requires copy API and trap handler to be initalized
prior to loading.  Existing tools do not make use of internal queue or scratch
memory intercept which is what PostToolsInit allows.

PostToolsInit() will be removed in a following cleanup change.

Change-Id: If43377843808e3eff0defd9204910a67a852902f

runtime/hsa-runtime/core/runtime/runtime.cpp

@@ -1218,9 +1218,7 @@ hsa_status_t Runtime::Load() {
   // Load extensions
   LoadExtensions();
 
-  // Load tools libraries
-  LoadTools();
-
+  // Initialize per GPU scratch, blits, and trap handler
   for (core::Agent* agent : gpu_agents_) {
     hsa_status_t status =
         reinterpret_cast<amd::GpuAgentInt*>(agent)->PostToolsInit();
@@ -1230,6 +1228,9 @@ hsa_status_t Runtime::Load() {
     }
   }
 
+  // Load tools libraries
+  LoadTools();
+
   return HSA_STATUS_SUCCESS;
 }