Introduce the PTI 0.11 into kineto

Signed-off-by: Chen, Zejun <[email protected]>

Author: zejun-chen

libkineto/src/plugin/xpupti/XpuptiActivityApi.cpp

@@ -170,6 +170,23 @@ void XpuptiActivityApi::bufferCompleted(
 }
 #endif
 
+static void enableSpecifcRuntimeAPIsTracing() {
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueUSMFill_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueUSMFill2D_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueUSMMemcpy_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueUSMMemcpy2D_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueKernelLaunch_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueKernelLaunchCustomExp_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueCooperativeKernelLaunchExp_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueMemBufferFill_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueMemBufferRead_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueMemBufferWrite_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urEnqueueMemBufferCopy_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urUSMHostAlloc_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urUSMSharedAlloc_id));
+    XPUPTI_CALL(ptiViewEnableRuntimeApi(1, pti_api_group_id::PTI_API_GROUP_SYCL, urUSMDeviceAlloc_id));
+}
+
 void XpuptiActivityApi::enableXpuptiActivities(
     const std::set<ActivityType>& selected_activities) {
 #ifdef HAS_XPUPTI
@@ -192,7 +209,8 @@ void XpuptiActivityApi::enableXpuptiActivities(
       externalCorrelationEnabled_ = true;
     }
     if (activity == ActivityType::XPU_RUNTIME) {
-      XPUPTI_CALL(ptiViewEnable(PTI_VIEW_SYCL_RUNTIME_CALLS));
+      XPUPTI_CALL(ptiViewEnable(PTI_VIEW_RUNTIME_API));
+      enableSpecifcRuntimeAPIsTracing();
     }
     if (activity == ActivityType::OVERHEAD) {
       XPUPTI_CALL(ptiViewEnable(PTI_VIEW_COLLECTION_OVERHEAD));
@@ -222,7 +240,7 @@ void XpuptiActivityApi::disablePtiActivities(
       XPUPTI_CALL(ptiViewDisable(PTI_VIEW_EXTERNAL_CORRELATION));
     }
     if (activity == ActivityType::XPU_RUNTIME) {
-      XPUPTI_CALL(ptiViewDisable(PTI_VIEW_SYCL_RUNTIME_CALLS));
+      XPUPTI_CALL(ptiViewDisable(PTI_VIEW_RUNTIME_API));
     }
     if (activity == ActivityType::OVERHEAD) {
       XPUPTI_CALL(ptiViewDisable(PTI_VIEW_COLLECTION_OVERHEAD));

libkineto/src/plugin/xpupti/XpuptiActivityHandlers.cpp

@@ -98,16 +98,18 @@ inline void XpuptiActivityProfilerSession::handleCorrelationActivity(
 }
 
 void XpuptiActivityProfilerSession::handleRuntimeActivity(
-    const pti_view_record_sycl_runtime* activity,
+    const pti_view_record_api* activity,
     ActivityLogger* logger) {
   traceBuffer_.span.opCount += 1;
   traceBuffer_.gpuOpCount += 1;
   const ITraceActivity* linked =
       linkedActivity(activity->_correlation_id, cpuCorrelationMap_);
+  const char* api_name = nullptr;
+  ptiViewGetApiIdName(activity->_api_group, activity->_api_id, &api_name);
   traceBuffer_.emplace_activity(
       traceBuffer_.span,
       ActivityType::XPU_RUNTIME,
-      std::string(activity->_name));
+      std::string(api_name));
   auto& runtime_activity = traceBuffer_.activities.back();
   runtime_activity->startTime = activity->_start_timestamp;
   runtime_activity->endTime = activity->_end_timestamp;
@@ -331,9 +333,9 @@ void XpuptiActivityProfilerSession::handlePtiActivity(
           reinterpret_cast<const pti_view_record_external_correlation*>(
               record));
       break;
-    case PTI_VIEW_SYCL_RUNTIME_CALLS:
+    case PTI_VIEW_RUNTIME_API:
       handleRuntimeActivity(
-          reinterpret_cast<const pti_view_record_sycl_runtime*>(record),
+          reinterpret_cast<const pti_view_record_api*>(record),
           logger);
       break;
     case PTI_VIEW_DEVICE_GPU_KERNEL:

libkineto/src/plugin/xpupti/XpuptiActivityProfiler.h

@@ -52,7 +52,7 @@ class XpuptiActivityProfilerSession
   void handleCorrelationActivity(
       const pti_view_record_external_correlation* correlation);
   void handleRuntimeActivity(
-      const pti_view_record_sycl_runtime* activity,
+      const pti_view_record_api* activity,
       ActivityLogger* logger);
   void handleKernelActivity(
       const pti_view_record_kernel* activity,