GPUTPCCFCheckPadBaseline.h

Go to the documentation of this file.

// Copyright 2019-2020 CERN and copyright holders of ALICE O2.
// See https://alice-o2.web.cern.ch/copyright for details of the copyright holders.
// All rights not expressly granted are reserved.
//
// This software is distributed under the terms of the GNU General Public
// License v3 (GPL Version 3), copied verbatim in the file "COPYING".
//
// In applying this license CERN does not waive the privileges and immunities
// granted to it by virtue of its status as an Intergovernmental Organization
// or submit itself to any jurisdiction.
 
 
#ifndef O2_GPU_GPU_TPC_CF_CHECK_PAD_BASELINE_H
#define O2_GPU_GPU_TPC_CF_CHECK_PAD_BASELINE_H
 
#include "GPUGeneralKernels.h"
#include "GPUConstantMem.h"
#include "GPUTPCGeometry.h"
 
#include "clusterFinderDefs.h"
#include "CfArray2D.h"
 
namespace o2::gpu
{
 
struct HIPTailDescriptor {
  uint32_t iPrev;
  uint32_t iNext;
  uint16_t pad;
  uint16_t tailStart;
  uint16_t tailEnd;
  float qTot;
  float qMax;
};
 
class GPUTPCCFCheckPadBaseline : public GPUKernelTemplate
{
 
 public:
  enum {
    PadsPerCacheline = TPCMapMemoryLayout<uint16_t>::Width,
    TimebinsPerCacheline = TPCMapMemoryLayout<uint16_t>::Height,
    EntriesPerCacheline = PadsPerCacheline * TimebinsPerCacheline,
    NumOfCachedPads = GPUCA_WARP_SIZE / TimebinsPerCacheline,
    NumCLsPerWarp = GPUCA_WARP_SIZE / EntriesPerCacheline,
    NumOfCachedTBs = TimebinsPerCacheline * 8,
    // Threads index shared memory as [iThread / MaxNPadsPerRow][iThread % MaxNPadsPerRow].
    // Rounding up to a multiple of PadsPerCacheline ensures iThread / MaxNPadsPerRow < NumOfCachedTBs
    // for all threads, avoiding out-of-bounds access.
    MaxNPadsPerRow = CAMath::nextMultipleOf<PadsPerCacheline>(GPUTPCGeometry::MaxNPadsPerRow()),
 
    MaxADC = 1023,
 
    NThreads = GPUCA_GET_THREAD_COUNT(GPUCA_LB_GPUTPCCFCheckPadBaseline),
    SSClusterPadWidth = 5,
  };
 
  union HipTailRange {
    struct {
      int16_t start;
      int16_t end;
    };
 
    // Be careful with using default initialized values.
    // Need default constructor, so can be placed in shared memory.
    // Might be zero initialized, but invalid tail needs start = end = -1 instead.
    GPUdDefault() HipTailRange() = default;
    GPUdi() HipTailRange(int16_t st, int16_t e) : start(st), end(e) {}
 
    GPUdi() bool HasValue() const { return start > -1; }
    GPUdi() bool IsOpen() const { return start > -1 && end < 0; }
 
    GPUdi() void SetOpen(int16_t st)
    {
      start = st;
      end = -1;
    }
 
    GPUdi() int16_t Length() const { return end - start; }
 
    GPUdi() void Reset() { start = end = -1; }
  };
 
  struct GPUSharedMemory : public GPUKernelTemplate::GPUSharedMemoryScan64<int16_t, NThreads> {
    tpccf::Charge charges[NumOfCachedTBs][MaxNPadsPerRow];
    HipTailRange tails[MaxNPadsPerRow];
    uint8_t tailsClosedPad[MaxNPadsPerRow];
    HipTailRange tailsClosed[MaxNPadsPerRow];
    uint32_t tailsClosedStoreIdx[MaxNPadsPerRow];
    tpccf::Charge tailQTotScratch[NThreads];
    tpccf::Charge tailQMaxScratch[NThreads];
    uint32_t tailStoreBase;
  };
 
  // Accumulated values from scanning cached charges in a pad
  struct PadChargeAccu {
    int32_t totalCharges = 0;
    int32_t consecCharges = 0;
    int32_t maxConsecCharges = 0;
    tpccf::Charge maxCharge = 0;
    int16_t HIPtb = -1;
    int16_t aboveThresholdStart = -1; // first TB of current above-hipTailThreshold streak; used to extend the tail back over the rising edge before saturation
    HipTailRange activeHIPTail{-1, -1};
    tpccf::Charge tailFilterCharge = 0;
  };
 
  typedef GPUTPCClusterFinder processorType;
  GPUhdi() static processorType* Processor(GPUConstantMem& processors)
  {
    return processors.tpcClusterer;
  }
 
  GPUhdi() constexpr static gpudatatypes::RecoStep GetRecoStep()
  {
    return gpudatatypes::RecoStep::TPCClusterFinding;
  }
 
  static int32_t GetNBlocks(bool isGPU)
  {
    // Important to exclude rightmost padding from Pad Filter.
    // There's nothing to filter there and padding is counted as start of a row, so it causes an overflow in the row count.
    const int32_t nBlocksCPU = (TPC_CLUSTERER_STRIDED_PAD_COUNT - GPUCF_PADDING_PAD) / PadsPerCacheline;
    return isGPU ? GPUTPCGeometry::NROWS : nBlocksCPU;
  }
 
  template <int32_t iKernel = defaultKernel>
  GPUd() static void Thread(int32_t nBlocks, int32_t nThreads, int32_t iBlock, int32_t iThread, GPUSharedMemory& smem, processorType& clusterer);
 
 private:
  GPUd() static void CheckBaselineGPU(int32_t nBlocks, int32_t nThreads, int32_t iBlock, int32_t iThread, GPUSharedMemory& smem, processorType& clusterer);
  GPUd() static void CheckBaselineCPU(int32_t nBlocks, int32_t nThreads, int32_t iBlock, int32_t iThread, GPUSharedMemory& smem, processorType& clusterer);
 
  GPUd() static void updatePadBaseline(int32_t pad, const GPUTPCClusterFinder&, int32_t totalCharges, int32_t consecCharges, tpccf::Charge maxCharge);
};
 
class GPUTPCCFHIPTailConnector : public GPUKernelTemplate
{
 public:
  enum {
    MaxHIPTails = 1 << 15,
    MaxHIPTailsPerRow = MaxHIPTails,
  };
 
  struct GPUSharedMemory {
  };
 
  typedef GPUTPCClusterFinder processorType;
  GPUhdi() static processorType* Processor(GPUConstantMem& processors)
  {
    return processors.tpcClusterer;
  }
 
  GPUhdi() constexpr static gpudatatypes::RecoStep GetRecoStep()
  {
    return gpudatatypes::RecoStep::TPCClusterFinding;
  }
 
  template <int32_t iKernel = defaultKernel>
  GPUd() static void Thread(int32_t nBlocks, int32_t nThreads, int32_t iBlock, int32_t iThread, GPUSharedMemory& smem, processorType& clusterer);
};
 
class GPUTPCCFHIPClusterizer : public GPUKernelTemplate
{
 public:
  enum {
    MaxHIPTails = GPUTPCCFHIPTailConnector::MaxHIPTails,
    MaxHIPTailsPerRow = GPUTPCCFHIPTailConnector::MaxHIPTailsPerRow,
  };
 
  struct GPUSharedMemory {
  };
 
  typedef GPUTPCClusterFinder processorType;
  GPUhdi() static processorType* Processor(GPUConstantMem& processors)
  {
    return processors.tpcClusterer;
  }
 
  GPUhdi() constexpr static gpudatatypes::RecoStep GetRecoStep()
  {
    return gpudatatypes::RecoStep::TPCClusterFinding;
  }
 
  template <int32_t iKernel = defaultKernel>
  GPUd() static void Thread(int32_t nBlocks, int32_t nThreads, int32_t iBlock, int32_t iThread, GPUSharedMemory& smem, processorType& clusterer);
};
 
} // namespace o2::gpu
 
#endif