GPU Clusterizer: Fix should not release an unused event

davidrohr · davidrohr · commit 4728f06ed333 · 2024-11-16T13:16:36.000+01:00
diff --git a/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx b/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx
@@ -865,8 +865,12 @@ int32_t GPUChainTracking::RunTPCClusterizer(bool synchronizeOutput)
         }
 
         if (fragment.index == 0) {
-          runKernel<GPUMemClean16>({GetGridAutoStep(lane, RecoStep::TPCClusterFinding), krnlRunRangeNone, {nullptr, transferRunning[lane] == 1 ? &mEvents->stream[lane] : nullptr}}, clustererShadow.mPclusterInRow, GPUCA_ROW_COUNT * sizeof(*clustererShadow.mPclusterInRow));
-          transferRunning[lane] = 2;
+          deviceEvent* waitEvent = nullptr;
+          if (transferRunning[lane] == 1) {
+            waitEvent = &mEvents->stream[lane];
+            transferRunning[lane] = 2;
+          }
+          runKernel<GPUMemClean16>({GetGridAutoStep(lane, RecoStep::TPCClusterFinding), krnlRunRangeNone, {nullptr, waitEvent}}, clustererShadow.mPclusterInRow, GPUCA_ROW_COUNT * sizeof(*clustererShadow.mPclusterInRow));
         }
 
         if (clusterer.mPmemory->counters.nClusters == 0) {

Original file line number	Diff line number	Diff line change
`@@ -865,8 +865,12 @@ int32_t GPUChainTracking::RunTPCClusterizer(bool synchronizeOutput)`
`865`	`865`	`}`
`866`	`866`
`867`	`867`	`if (fragment.index == 0) {`
`868`		`- runKernel<GPUMemClean16>({GetGridAutoStep(lane, RecoStep::TPCClusterFinding), krnlRunRangeNone, {nullptr, transferRunning[lane] == 1 ? &mEvents->stream[lane] : nullptr}}, clustererShadow.mPclusterInRow, GPUCA_ROW_COUNT * sizeof(*clustererShadow.mPclusterInRow));`
`869`		`- transferRunning[lane] = 2;`
	`868`	`+ deviceEvent* waitEvent = nullptr;`
	`869`	`+ if (transferRunning[lane] == 1) {`
	`870`	`+ waitEvent = &mEvents->stream[lane];`
	`871`	`+ transferRunning[lane] = 2;`
	`872`	`+ }`
	`873`	`+ runKernel<GPUMemClean16>({GetGridAutoStep(lane, RecoStep::TPCClusterFinding), krnlRunRangeNone, {nullptr, waitEvent}}, clustererShadow.mPclusterInRow, GPUCA_ROW_COUNT * sizeof(*clustererShadow.mPclusterInRow));`
`870`	`874`	`}`
`871`	`875`
`872`	`876`	`if (clusterer.mPmemory->counters.nClusters == 0) {`