llvm/tools/llvm-exegesis/lib/Analysis.cpp

//===-- Analysis.cpp --------------------------------------------*- C++ -*-===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//

#include "Analysis.h"
#include "BenchmarkResult.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/Support/FormatVariadic.h"
#include <unordered_set>
#include <vector>

namespace exegesis {

static const char kCsvSep = ',';

namespace {

enum EscapeTag { kEscapeCsv, kEscapeHtml };

template <EscapeTag Tag>
void writeEscaped(llvm::raw_ostream &OS, const llvm::StringRef S);

template <>
void writeEscaped<kEscapeCsv>(llvm::raw_ostream &OS, const llvm::StringRef S) {
  if (std::find(S.begin(), S.end(), kCsvSep) == S.end()) {
    OS << S;
  } else {
    // Needs escaping.
    OS << '"';
    for (const char C : S) {
      if (C == '"')
        OS << "\"\"";
      else
        OS << C;
    }
    OS << '"';
  }
}

template <>
void writeEscaped<kEscapeHtml>(llvm::raw_ostream &OS, const llvm::StringRef S) {
  for (const char C : S) {
    if (C == '<')
      OS << "&lt;";
    else if (C == '>')
      OS << "&gt;";
    else if (C == '&')
      OS << "&amp;";
    else
      OS << C;
  }
}

} // namespace

template <EscapeTag Tag>
static void
writeClusterId(llvm::raw_ostream &OS,
               const InstructionBenchmarkClustering::ClusterId &CID) {
  if (CID.isNoise())
    writeEscaped<Tag>(OS, "[noise]");
  else if (CID.isError())
    writeEscaped<Tag>(OS, "[error]");
  else
    OS << CID.getId();
}

template <EscapeTag Tag>
static void writeMeasurementValue(llvm::raw_ostream &OS, const double Value) {
  writeEscaped<Tag>(OS, llvm::formatv("{0:F}", Value).str());
}

// Prints a row representing an instruction, along with scheduling info and
// point coordinates (measurements).
void Analysis::printInstructionRowCsv(const size_t PointId,
                                      llvm::raw_ostream &OS) const {
  const InstructionBenchmark &Point = Clustering_.getPoints()[PointId];
  writeClusterId<kEscapeCsv>(OS, Clustering_.getClusterIdForPoint(PointId));
  OS << kCsvSep;
  writeEscaped<kEscapeCsv>(OS, Point.Key.OpcodeName);
  OS << kCsvSep;
  writeEscaped<kEscapeCsv>(OS, Point.Key.Config);
  OS << kCsvSep;
  const auto OpcodeIt = MnemonicToOpcode_.find(Point.Key.OpcodeName);
  if (OpcodeIt != MnemonicToOpcode_.end()) {
    const unsigned SchedClassId =
        InstrInfo_->get(OpcodeIt->second).getSchedClass();
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
    const auto &SchedModel = SubtargetInfo_->getSchedModel();
    const llvm::MCSchedClassDesc *const SCDesc =
        SchedModel.getSchedClassDesc(SchedClassId);
    writeEscaped<kEscapeCsv>(OS, SCDesc->Name);
#else
    OS << SchedClassId;
#endif
  }
  // FIXME: Print the sched class once InstructionBenchmark separates key into
  // (mnemonic, mode, opaque).
  for (const auto &Measurement : Point.Measurements) {
    OS << kCsvSep;
    writeMeasurementValue<kEscapeCsv>(OS, Measurement.Value);
  }
  OS << "\n";
}

Analysis::Analysis(const llvm::Target &Target,
                   const InstructionBenchmarkClustering &Clustering)
    : Clustering_(Clustering) {
  if (Clustering.getPoints().empty())
    return;

  InstrInfo_.reset(Target.createMCInstrInfo());
  const InstructionBenchmark &FirstPoint = Clustering.getPoints().front();
  SubtargetInfo_.reset(Target.createMCSubtargetInfo(FirstPoint.LLVMTriple,
                                                    FirstPoint.CpuName, ""));

  // Build an index of mnemonic->opcode.
  for (int I = 0, E = InstrInfo_->getNumOpcodes(); I < E; ++I)
    MnemonicToOpcode_.emplace(InstrInfo_->getName(I), I);
}

template <>
llvm::Error
Analysis::run<Analysis::PrintClusters>(llvm::raw_ostream &OS) const {
  if (Clustering_.getPoints().empty())
    return llvm::Error::success();

  // Write the header.
  OS << "cluster_id" << kCsvSep << "opcode_name" << kCsvSep << "config"
     << kCsvSep << "sched_class";
  for (const auto &Measurement : Clustering_.getPoints().front().Measurements) {
    OS << kCsvSep;
    writeEscaped<kEscapeCsv>(OS, Measurement.Key);
  }
  OS << "\n";

  // Write the points.
  const auto &Clusters = Clustering_.getValidClusters();
  for (size_t I = 0, E = Clusters.size(); I < E; ++I) {
    for (const size_t PointId : Clusters[I].PointIndices) {
      printInstructionRowCsv(PointId, OS);
    }
    OS << "\n\n";
  }
  return llvm::Error::success();
}

std::unordered_map<unsigned, std::vector<size_t>>
Analysis::makePointsPerSchedClass() const {
  std::unordered_map<unsigned, std::vector<size_t>> PointsPerSchedClass;
  const auto &Points = Clustering_.getPoints();
  for (size_t PointId = 0, E = Points.size(); PointId < E; ++PointId) {
    const InstructionBenchmark &Point = Points[PointId];
    if (!Point.Error.empty())
      continue;
    const auto OpcodeIt = MnemonicToOpcode_.find(Point.Key.OpcodeName);
    if (OpcodeIt == MnemonicToOpcode_.end())
      continue;
    const unsigned SchedClassId =
        InstrInfo_->get(OpcodeIt->second).getSchedClass();
    PointsPerSchedClass[SchedClassId].push_back(PointId);
  }
  return PointsPerSchedClass;
}

void Analysis::printSchedClassClustersHtml(
    const std::vector<SchedClassCluster> &Clusters, const SchedClass &SC,
    llvm::raw_ostream &OS) const {
  const auto &Points = Clustering_.getPoints();
  OS << "<table class=\"sched-class-clusters\">";
  OS << "<tr><th>ClusterId</th><th>Opcode/Config</th>";
  assert(!Clusters.empty());
  for (const auto &Measurement :
       Points[Clusters[0].getPointIds()[0]].Measurements) {
    OS << "<th>";
    if (Measurement.DebugString.empty())
      writeEscaped<kEscapeHtml>(OS, Measurement.Key);
    else
      writeEscaped<kEscapeHtml>(OS, Measurement.DebugString);
    OS << "</th>";
  }
  OS << "</tr>";
  for (const SchedClassCluster &Cluster : Clusters) {
    OS << "<tr class=\""
       << (Cluster.measurementsMatch(*SubtargetInfo_, SC, Clustering_)
               ? "good-cluster"
               : "bad-cluster")
       << "\"><td>";
    writeClusterId<kEscapeHtml>(OS, Cluster.id());
    OS << "</td><td><ul>";
    for (const size_t PointId : Cluster.getPointIds()) {
      const auto &Point = Points[PointId];
      OS << "<li><span class=\"mono\">";
      writeEscaped<kEscapeHtml>(OS, Point.Key.OpcodeName);
      OS << "</span> <span class=\"mono\">";
      writeEscaped<kEscapeHtml>(OS, Point.Key.Config);
      OS << "</span></li>";
    }
    OS << "</ul></td>";
    for (const auto &Stats : Cluster.getRepresentative()) {
      OS << "<td class=\"measurement\">";
      writeMeasurementValue<kEscapeHtml>(OS, Stats.avg());
      OS << "<br><span class=\"minmax\">[";
      writeMeasurementValue<kEscapeHtml>(OS, Stats.min());
      OS << ";";
      writeMeasurementValue<kEscapeHtml>(OS, Stats.max());
      OS << "]</span></td>";
    }
    OS << "</tr>";
  }
  OS << "</table>";
}

// Return the non-redundant list of WriteProcRes used by the given sched class.
// The scheduling model for LLVM is such that each instruction has a certain
// number of uops which consume resources which are described by WriteProcRes
// entries. Each entry describe how many cycles are spent on a specific ProcRes
// kind.
// For example, an instruction might have 3 uOps, one dispatching on P0
// (ProcResIdx=1) and two on P06 (ProcResIdx = 7).
// Note that LLVM additionally denormalizes resource consumption to include
// usage of super resources by subresources. So in practice if there exists a
// P016 (ProcResIdx=10), then the cycles consumed by P0 are also consumed by
// P06 (ProcResIdx = 7) and P016 (ProcResIdx = 10), and the resources consumed
// by P06 are also consumed by P016. In the figure below, parenthesized cycles
// denote implied usage of superresources by subresources:
//            P0      P06    P016
//     uOp1    1      (1)     (1)
//     uOp2            1      (1)
//     uOp3            1      (1)
//     =============================
//             1       3       3
// Eventually we end up with three entries for the WriteProcRes of the
// instruction:
//    {ProcResIdx=1,  Cycles=1}  // P0
//    {ProcResIdx=7,  Cycles=3}  // P06
//    {ProcResIdx=10, Cycles=3}  // P016
//
// Note that in this case, P016 does not contribute any cycles, so it would
// be removed by this function.
// FIXME: Move this to MCSubtargetInfo and use it in llvm-mca.
static llvm::SmallVector<llvm::MCWriteProcResEntry, 8>
getNonRedundantWriteProcRes(const llvm::MCSchedClassDesc &SCDesc,
                            const llvm::MCSubtargetInfo &STI) {
  llvm::SmallVector<llvm::MCWriteProcResEntry, 8> Result;
  const auto &SM = STI.getSchedModel();
  const unsigned NumProcRes = SM.getNumProcResourceKinds();

  // This assumes that the ProcResDescs are sorted in topological order, which
  // is guaranteed by the tablegen backend.
  llvm::SmallVector<float, 32> ProcResUnitUsage(NumProcRes);
  for (const auto *WPR = STI.getWriteProcResBegin(&SCDesc),
                  *const WPREnd = STI.getWriteProcResEnd(&SCDesc);
       WPR != WPREnd; ++WPR) {
    const llvm::MCProcResourceDesc *const ProcResDesc =
        SM.getProcResource(WPR->ProcResourceIdx);
    if (ProcResDesc->SubUnitsIdxBegin == nullptr) {
      // This is a ProcResUnit.
      Result.push_back({WPR->ProcResourceIdx, WPR->Cycles});
      ProcResUnitUsage[WPR->ProcResourceIdx] += WPR->Cycles;
    } else {
      // This is a ProcResGroup. First see if it contributes any cycles or if
      // it has cycles just from subunits.
      float RemainingCycles = WPR->Cycles;
      for (const auto *SubResIdx = ProcResDesc->SubUnitsIdxBegin;
           SubResIdx != ProcResDesc->SubUnitsIdxBegin + ProcResDesc->NumUnits;
           ++SubResIdx) {
        RemainingCycles -= ProcResUnitUsage[*SubResIdx];
      }
      if (RemainingCycles < 0.01f) {
        // The ProcResGroup contributes no cycles of its own.
        continue;
      }
      // The ProcResGroup contributes `RemainingCycles` cycles of its own.
      Result.push_back({WPR->ProcResourceIdx,
                        static_cast<uint16_t>(std::round(RemainingCycles))});
      // Spread the remaining cycles over all subunits.
      for (const auto *SubResIdx = ProcResDesc->SubUnitsIdxBegin;
           SubResIdx != ProcResDesc->SubUnitsIdxBegin + ProcResDesc->NumUnits;
           ++SubResIdx) {
        ProcResUnitUsage[*SubResIdx] += RemainingCycles / ProcResDesc->NumUnits;
      }
    }
  }
  return Result;
}

Analysis::SchedClass::SchedClass(const llvm::MCSchedClassDesc &SD,
                                 const llvm::MCSubtargetInfo &STI)
    : SCDesc(SD),
      NonRedundantWriteProcRes(getNonRedundantWriteProcRes(SD, STI)),
      IdealizedProcResPressure(computeIdealizedProcResPressure(
          STI.getSchedModel(), NonRedundantWriteProcRes)) {}

void Analysis::SchedClassCluster::addPoint(
    size_t PointId, const InstructionBenchmarkClustering &Clustering) {
  PointIds.push_back(PointId);
  const auto &Point = Clustering.getPoints()[PointId];
  if (ClusterId.isUndef()) {
    ClusterId = Clustering.getClusterIdForPoint(PointId);
    Representative.resize(Point.Measurements.size());
  }
  for (size_t I = 0, E = Point.Measurements.size(); I < E; ++I) {
    Representative[I].push(Point.Measurements[I]);
  }
  assert(ClusterId == Clustering.getClusterIdForPoint(PointId));
}

bool Analysis::SchedClassCluster::measurementsMatch(
    const llvm::MCSubtargetInfo &STI, const SchedClass &SC,
    const InstructionBenchmarkClustering &Clustering) const {
  const size_t NumMeasurements = Representative.size();
  std::vector<BenchmarkMeasure> ClusterCenterPoint(NumMeasurements);
  std::vector<BenchmarkMeasure> SchedClassPoint(NumMeasurements);
  // Latency case.
  assert(!Clustering.getPoints().empty());
  const InstructionBenchmarkKey::ModeE Mode =
      Clustering.getPoints()[0].Key.Mode;
  if (Mode == InstructionBenchmarkKey::Latency) {
    if (NumMeasurements != 1) {
      llvm::errs()
          << "invalid number of measurements in latency mode: expected 1, got "
          << NumMeasurements << "\n";
      return false;
    }
    // Find the latency.
    SchedClassPoint[0].Value = 0.0;
    for (unsigned I = 0; I < SC.SCDesc.NumWriteLatencyEntries; ++I) {
      const llvm::MCWriteLatencyEntry *const WLE =
          STI.getWriteLatencyEntry(&SC.SCDesc, I);
      SchedClassPoint[0].Value =
          std::max<double>(SchedClassPoint[0].Value, WLE->Cycles);
    }
    ClusterCenterPoint[0].Value = Representative[0].avg();
  } else if (Mode == InstructionBenchmarkKey::Uops) {
    for (int I = 0, E = Representative.size(); I < E; ++I) {
      // Find the pressure on ProcResIdx `Key`.
      uint16_t ProcResIdx = 0;
      if (!llvm::to_integer(Representative[I].key(), ProcResIdx, 10)) {
        llvm::errs() << "expected ProcResIdx key, got "
                     << Representative[I].key() << "\n";
        return false;
      }
      const auto ProcResPressureIt =
          std::find_if(SC.IdealizedProcResPressure.begin(),
                       SC.IdealizedProcResPressure.end(),
                       [ProcResIdx](const std::pair<uint16_t, float> &WPR) {
                         return WPR.first == ProcResIdx;
                       });
      SchedClassPoint[I].Value =
          ProcResPressureIt == SC.IdealizedProcResPressure.end()
              ? 0.0
              : ProcResPressureIt->second;
      ClusterCenterPoint[I].Value = Representative[I].avg();
    }
  } else {
    llvm::errs() << "unimplemented measurement matching for mode " << Mode
                 << "\n";
    return false;
  }
  return Clustering.isNeighbour(ClusterCenterPoint, SchedClassPoint);
}

void Analysis::printSchedClassDescHtml(const SchedClass &SC,
                                       llvm::raw_ostream &OS) const {
  OS << "<table class=\"sched-class-desc\">";
  OS << "<tr><th>Valid</th><th>Variant</th><th>uOps</th><th>Latency</"
        "th><th>WriteProcRes</th><th title=\"This is the idealized unit "
        "resource (port) pressure assuming ideal distribution\">Idealized "
        "Resource Pressure</th></tr>";
  if (SC.SCDesc.isValid()) {
    const auto &SM = SubtargetInfo_->getSchedModel();
    OS << "<tr><td>&#10004;</td>";
    OS << "<td>" << (SC.SCDesc.isVariant() ? "&#10004;" : "&#10005;")
       << "</td>";
    OS << "<td>" << SC.SCDesc.NumMicroOps << "</td>";
    // Latencies.
    OS << "<td><ul>";
    for (int I = 0, E = SC.SCDesc.NumWriteLatencyEntries; I < E; ++I) {
      const auto *const Entry =
          SubtargetInfo_->getWriteLatencyEntry(&SC.SCDesc, I);
      OS << "<li>" << Entry->Cycles;
      if (SC.SCDesc.NumWriteLatencyEntries > 1) {
        // Dismabiguate if more than 1 latency.
        OS << " (WriteResourceID " << Entry->WriteResourceID << ")";
      }
      OS << "</li>";
    }
    OS << "</ul></td>";
    // WriteProcRes.
    OS << "<td><ul>";
    for (const auto &WPR : SC.NonRedundantWriteProcRes) {
      OS << "<li><span class=\"mono\">";
      writeEscaped<kEscapeHtml>(OS,
                                SM.getProcResource(WPR.ProcResourceIdx)->Name);
      OS << "</span>: " << WPR.Cycles << "</li>";
    }
    OS << "</ul></td>";
    // Idealized port pressure.
    OS << "<td><ul>";
    for (const auto &Pressure : SC.IdealizedProcResPressure) {
      OS << "<li><span class=\"mono\">";
      writeEscaped<kEscapeHtml>(OS, SubtargetInfo_->getSchedModel()
                                        .getProcResource(Pressure.first)
                                        ->Name);
      OS << "</span>: ";
      writeMeasurementValue<kEscapeHtml>(OS, Pressure.second);
      OS << "</li>";
    }
    OS << "</ul></td>";
    OS << "</tr>";
  } else {
    OS << "<tr><td>&#10005;</td><td></td><td></td></tr>";
  }
  OS << "</table>";
}

static constexpr const char kHtmlHead[] = R"(
<head>
<title>llvm-exegesis Analysis Results</title>
<style>
body {
  font-family: sans-serif
}
span.sched-class-name {
  font-weight: bold;
  font-family: monospace;
}
span.opcode {
  font-family: monospace;
}
span.config {
  font-family: monospace;
}
div.inconsistency {
  margin-top: 50px;
}
table {
  margin-left: 50px;
  border-collapse: collapse;
}
table, table tr,td,th {
  border: 1px solid #444;
}
table ul {
  padding-left: 0px;
  margin: 0px;
  list-style-type: none;
}
table.sched-class-clusters td {
  padding-left: 10px;
  padding-right: 10px;
  padding-top: 10px;
  padding-bottom: 10px;
}
table.sched-class-desc td {
  padding-left: 10px;
  padding-right: 10px;
  padding-top: 2px;
  padding-bottom: 2px;
}
span.mono {
  font-family: monospace;
}
td.measurement {
  text-align: center;
}
tr.good-cluster td.measurement {
  color: #292
}
tr.bad-cluster td.measurement {
  color: #922
}
tr.good-cluster td.measurement span.minmax {
  color: #888;
}
tr.bad-cluster td.measurement span.minmax {
  color: #888;
}
</style>
</head>
)";

template <>
llvm::Error Analysis::run<Analysis::PrintSchedClassInconsistencies>(
    llvm::raw_ostream &OS) const {
  const auto &FirstPoint = Clustering_.getPoints()[0];
  // Print the header.
  OS << "<!DOCTYPE html><html>" << kHtmlHead << "<body>";
  OS << "<h1><span class=\"mono\">llvm-exegesis</span> Analysis Results</h1>";
  OS << "<h3>Triple: <span class=\"mono\">";
  writeEscaped<kEscapeHtml>(OS, FirstPoint.LLVMTriple);
  OS << "</span></h3><h3>Cpu: <span class=\"mono\">";
  writeEscaped<kEscapeHtml>(OS, FirstPoint.CpuName);
  OS << "</span></h3>";

  for (const auto &SchedClassAndPoints : makePointsPerSchedClass()) {
    const auto SchedClassId = SchedClassAndPoints.first;
    const std::vector<size_t> &SchedClassPoints = SchedClassAndPoints.second;
    const auto &SchedModel = SubtargetInfo_->getSchedModel();
    const llvm::MCSchedClassDesc *const SCDesc =
        SchedModel.getSchedClassDesc(SchedClassId);
    if (!SCDesc)
      continue;
    const SchedClass SC(*SCDesc, *SubtargetInfo_);

    // Bucket sched class points into sched class clusters.
    std::vector<SchedClassCluster> SchedClassClusters;
    for (const size_t PointId : SchedClassPoints) {
      const auto &ClusterId = Clustering_.getClusterIdForPoint(PointId);
      if (!ClusterId.isValid())
        continue; // Ignore noise and errors. FIXME: take noise into account ?
      auto SchedClassClusterIt =
          std::find_if(SchedClassClusters.begin(), SchedClassClusters.end(),
                       [ClusterId](const SchedClassCluster &C) {
                         return C.id() == ClusterId;
                       });
      if (SchedClassClusterIt == SchedClassClusters.end()) {
        SchedClassClusters.emplace_back();
        SchedClassClusterIt = std::prev(SchedClassClusters.end());
      }
      SchedClassClusterIt->addPoint(PointId, Clustering_);
    }

    // Print any scheduling class that has at least one cluster that does not
    // match the checked-in data.
    if (std::all_of(SchedClassClusters.begin(), SchedClassClusters.end(),
                    [this, &SC](const SchedClassCluster &C) {
                      return C.measurementsMatch(*SubtargetInfo_, SC,
                                                 Clustering_);
                    }))
      continue; // Nothing weird.

    OS << "<div class=\"inconsistency\"><p>Sched Class <span "
          "class=\"sched-class-name\">";
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
    writeEscaped<kEscapeHtml>(OS, SCDesc->Name);
#else
    OS << SchedClassId;
#endif
    OS << "</span> contains instructions whose performance characteristics do"
          " not match that of LLVM:</p>";
    printSchedClassClustersHtml(SchedClassClusters, SC, OS);
    OS << "<p>llvm SchedModel data:</p>";
    printSchedClassDescHtml(SC, OS);
    OS << "</div>";
  }

  OS << "</body></html>";
  return llvm::Error::success();
}

// Distributes a pressure budget as evenly as possible on the provided subunits
// given the already existing port pressure distribution.
//
// The algorithm is as follows: while there is remaining pressure to
// distribute, find the subunits with minimal pressure, and distribute
// remaining pressure equally up to the pressure of the unit with
// second-to-minimal pressure.
// For example, let's assume we want to distribute 2*P1256
// (Subunits = [P1,P2,P5,P6]), and the starting DensePressure is:
//     DensePressure =        P0   P1   P2   P3   P4   P5   P6   P7
//                           0.1  0.3  0.2  0.0  0.0  0.5  0.5  0.5
//     RemainingPressure = 2.0
// We sort the subunits by pressure:
//     Subunits = [(P2,p=0.2), (P1,p=0.3), (P5,p=0.5), (P6, p=0.5)]
// We'll first start by the subunits with minimal pressure, which are at
// the beginning of the sorted array. In this example there is one (P2).
// The subunit with second-to-minimal pressure is the next one in the
// array (P1). So we distribute 0.1 pressure to P2, and remove 0.1 cycles
// from the budget.
//     Subunits = [(P2,p=0.3), (P1,p=0.3), (P5,p=0.5), (P5,p=0.5)]
//     RemainingPressure = 1.9
// We repeat this process: distribute 0.2 pressure on each of the minimal
// P2 and P1, decrease budget by 2*0.2:
//     Subunits = [(P2,p=0.5), (P1,p=0.5), (P5,p=0.5), (P5,p=0.5)]
//     RemainingPressure = 1.5
// There are no second-to-minimal subunits so we just share the remaining
// budget (1.5 cycles) equally:
//     Subunits = [(P2,p=0.875), (P1,p=0.875), (P5,p=0.875), (P5,p=0.875)]
//     RemainingPressure = 0.0
// We stop as there is no remaining budget to distribute.
void distributePressure(float RemainingPressure,
                        llvm::SmallVector<uint16_t, 32> Subunits,
                        llvm::SmallVector<float, 32> &DensePressure) {
  // Find the number of subunits with minimal pressure (they are at the
  // front).
  llvm::sort(Subunits.begin(), Subunits.end(),
             [&DensePressure](const uint16_t A, const uint16_t B) {
               return DensePressure[A] < DensePressure[B];
             });
  const auto getPressureForSubunit = [&DensePressure,
                                      &Subunits](size_t I) -> float & {
    return DensePressure[Subunits[I]];
  };
  size_t NumMinimalSU = 1;
  while (NumMinimalSU < Subunits.size() &&
         getPressureForSubunit(NumMinimalSU) == getPressureForSubunit(0)) {
    ++NumMinimalSU;
  }
  while (RemainingPressure > 0.0f) {
    if (NumMinimalSU == Subunits.size()) {
      // All units are minimal, just distribute evenly and be done.
      for (size_t I = 0; I < NumMinimalSU; ++I) {
        getPressureForSubunit(I) += RemainingPressure / NumMinimalSU;
      }
      return;
    }
    // Distribute the remaining pressure equally.
    const float MinimalPressure = getPressureForSubunit(NumMinimalSU - 1);
    const float SecondToMinimalPressure = getPressureForSubunit(NumMinimalSU);
    assert(MinimalPressure < SecondToMinimalPressure);
    const float Increment = SecondToMinimalPressure - MinimalPressure;
    if (RemainingPressure <= NumMinimalSU * Increment) {
      // There is not enough remaining pressure.
      for (size_t I = 0; I < NumMinimalSU; ++I) {
        getPressureForSubunit(I) += RemainingPressure / NumMinimalSU;
      }
      return;
    }
    // Bump all minimal pressure subunits to `SecondToMinimalPressure`.
    for (size_t I = 0; I < NumMinimalSU; ++I) {
      getPressureForSubunit(I) = SecondToMinimalPressure;
      RemainingPressure -= SecondToMinimalPressure;
    }
    while (NumMinimalSU < Subunits.size() &&
           getPressureForSubunit(NumMinimalSU) == SecondToMinimalPressure) {
      ++NumMinimalSU;
    }
  }
}

std::vector<std::pair<uint16_t, float>> computeIdealizedProcResPressure(
    const llvm::MCSchedModel &SM,
    llvm::SmallVector<llvm::MCWriteProcResEntry, 8> WPRS) {
  // DensePressure[I] is the port pressure for Proc Resource I.
  llvm::SmallVector<float, 32> DensePressure(SM.getNumProcResourceKinds());
  llvm::sort(WPRS.begin(), WPRS.end(),
             [](const llvm::MCWriteProcResEntry &A,
                const llvm::MCWriteProcResEntry &B) {
               return A.ProcResourceIdx < B.ProcResourceIdx;
             });
  for (const llvm::MCWriteProcResEntry &WPR : WPRS) {
    // Get units for the entry.
    const llvm::MCProcResourceDesc *const ProcResDesc =
        SM.getProcResource(WPR.ProcResourceIdx);
    if (ProcResDesc->SubUnitsIdxBegin == nullptr) {
      // This is a ProcResUnit.
      DensePressure[WPR.ProcResourceIdx] += WPR.Cycles;
    } else {
      // This is a ProcResGroup.
      llvm::SmallVector<uint16_t, 32> Subunits(ProcResDesc->SubUnitsIdxBegin,
                                               ProcResDesc->SubUnitsIdxBegin +
                                                   ProcResDesc->NumUnits);
      distributePressure(WPR.Cycles, Subunits, DensePressure);
    }
  }
  // Turn dense pressure into sparse pressure by removing zero entries.
  std::vector<std::pair<uint16_t, float>> Pressure;
  for (unsigned I = 0, E = SM.getNumProcResourceKinds(); I < E; ++I) {
    if (DensePressure[I] > 0.0f)
      Pressure.emplace_back(I, DensePressure[I]);
  }
  return Pressure;
}

} // namespace exegesis