From 225f744de8523575f06897410c1b35000db0f9c0 Mon Sep 17 00:00:00 2001
From: Christian Sonnabend <sonnabendch@gmail.com>
Date: Thu, 31 Jul 2025 23:50:54 +0200
Subject: [PATCH] bug-fix for memory allocation

---
 .../Global/GPUChainTrackingClusterizer.cxx    | 34 ++++++++++---------
 1 file changed, 18 insertions(+), 16 deletions(-)

diff --git a/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx b/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx
index 07b332db1fc12..846df352d3a34 100644
--- a/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx
+++ b/GPU/GPUTracking/Global/GPUChainTrackingClusterizer.cxx
@@ -643,26 +643,28 @@ int32_t GPUChainTracking::RunTPCClusterizer(bool synchronizeOutput)
 
   // Maximum of 4 lanes supported
   HighResTimer* nnTimers[12];
-  if (GetProcessingSettings().nn.applyNNclusterizer && GetProcessingSettings().debugLevel >= 1) {
-    nnTimers[0] = &getTimer<GPUTPCNNClusterizer, 0>("GPUTPCNNClusterizer_ONNXClassification_0_", 0);
-    nnTimers[1] = &getTimer<GPUTPCNNClusterizer, 1>("GPUTPCNNClusterizer_ONNXRegression_1_", 1);
-    nnTimers[2] = &getTimer<GPUTPCNNClusterizer, 2>("GPUTPCNNClusterizer_ONNXRegression2_2_", 2);
-    nnTimers[3] = &getTimer<GPUTPCNNClusterizer, 3>("GPUTPCNNClusterizer_ONNXClassification_0_", 3);
-    nnTimers[4] = &getTimer<GPUTPCNNClusterizer, 4>("GPUTPCNNClusterizer_ONNXRegression_1_", 4);
-    nnTimers[5] = &getTimer<GPUTPCNNClusterizer, 5>("GPUTPCNNClusterizer_ONNXRegression2_2_", 5);
-    nnTimers[6] = &getTimer<GPUTPCNNClusterizer, 6>("GPUTPCNNClusterizer_ONNXClassification_0_", 6);
-    nnTimers[7] = &getTimer<GPUTPCNNClusterizer, 7>("GPUTPCNNClusterizer_ONNXRegression_1_", 7);
-    nnTimers[8] = &getTimer<GPUTPCNNClusterizer, 8>("GPUTPCNNClusterizer_ONNXRegression2_2_", 8);
-    nnTimers[9] = &getTimer<GPUTPCNNClusterizer, 9>("GPUTPCNNClusterizer_ONNXClassification_0_", 9);
-    nnTimers[10] = &getTimer<GPUTPCNNClusterizer, 10>("GPUTPCNNClusterizer_ONNXRegression_1_", 10);
-    nnTimers[11] = &getTimer<GPUTPCNNClusterizer, 11>("GPUTPCNNClusterizer_ONNXRegression2_2_", 11);
-  }
 
   if (GetProcessingSettings().nn.applyNNclusterizer) {
     int32_t deviceId = -1;
     int32_t numLanes = GetProcessingSettings().nTPCClustererLanes;
     int32_t maxThreads = mRec->getNKernelHostThreads(true);
     // bool recreateMemoryAllocator = false;
+
+    if (GetProcessingSettings().debugLevel >= 1) {
+      nnTimers[0] = &getTimer<GPUTPCNNClusterizer, 0>("GPUTPCNNClusterizer_ONNXClassification_0_", 0);
+      nnTimers[1] = &getTimer<GPUTPCNNClusterizer, 1>("GPUTPCNNClusterizer_ONNXRegression_1_", 1);
+      nnTimers[2] = &getTimer<GPUTPCNNClusterizer, 2>("GPUTPCNNClusterizer_ONNXRegression2_2_", 2);
+      nnTimers[3] = &getTimer<GPUTPCNNClusterizer, 3>("GPUTPCNNClusterizer_ONNXClassification_0_", 3);
+      nnTimers[4] = &getTimer<GPUTPCNNClusterizer, 4>("GPUTPCNNClusterizer_ONNXRegression_1_", 4);
+      nnTimers[5] = &getTimer<GPUTPCNNClusterizer, 5>("GPUTPCNNClusterizer_ONNXRegression2_2_", 5);
+      nnTimers[6] = &getTimer<GPUTPCNNClusterizer, 6>("GPUTPCNNClusterizer_ONNXClassification_0_", 6);
+      nnTimers[7] = &getTimer<GPUTPCNNClusterizer, 7>("GPUTPCNNClusterizer_ONNXRegression_1_", 7);
+      nnTimers[8] = &getTimer<GPUTPCNNClusterizer, 8>("GPUTPCNNClusterizer_ONNXRegression2_2_", 8);
+      nnTimers[9] = &getTimer<GPUTPCNNClusterizer, 9>("GPUTPCNNClusterizer_ONNXClassification_0_", 9);
+      nnTimers[10] = &getTimer<GPUTPCNNClusterizer, 10>("GPUTPCNNClusterizer_ONNXRegression_1_", 10);
+      nnTimers[11] = &getTimer<GPUTPCNNClusterizer, 11>("GPUTPCNNClusterizer_ONNXRegression2_2_", 11);
+    }
+
     mRec->runParallelOuterLoop(doGPU, numLanes, [&](uint32_t lane) {
       nnApplications[lane].init(nn_settings, GetProcessingSettings().deterministicGPUReconstruction);
       if (nnApplications[lane].mModelsUsed[0]) {
@@ -708,7 +710,7 @@ int32_t GPUChainTracking::RunTPCClusterizer(bool synchronizeOutput)
         LOG(info) << "(ORT) Allocated ONNX stream for lane " << lane << " and device " << deviceId;
       }
     });
-    mRec->runParallelOuterLoop(doGPU, NSECTORS, [&](uint32_t sector) {
+    for (int32_t sector = 0; sector < NSECTORS; sector++) {
       GPUTPCNNClusterizer& clustererNN = processors()->tpcNNClusterer[sector];
       GPUTPCNNClusterizer& clustererNNShadow = doGPU ? processorsShadow()->tpcNNClusterer[sector] : clustererNN;
       int32_t lane = sector % numLanes;
@@ -725,7 +727,7 @@ int32_t GPUChainTracking::RunTPCClusterizer(bool synchronizeOutput)
       AllocateRegisteredMemory(clustererNN.mMemoryId);
       // nnApplications[lane].createBoundary(clustererNNShadow);
       // nnApplications[lane].createIndexLookup(clustererNNShadow);
-    });
+    }
     if (doGPU) {
       WriteToConstantMemory(RecoStep::TPCClusterFinding, (char*)&processors()->tpcNNClusterer - (char*)processors(), &processorsShadow()->tpcNNClusterer, sizeof(GPUTPCNNClusterizer) * NSECTORS, mRec->NStreams() - 1, &mEvents->init);
     }