[MLGO] Delete InlineSizeEstimatorAnalysis #167271
Merged
+11
−428
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
IR2Native was used initially to provide rewards for RL algorithms that preferred more immediate rewards. Now we have moved away from such training methodologies and this code has been sitting dead in tree providing a bit of a maintenance burden for a couple of years. Remove it given it is unused and requires some maintenance.
llvmbot
added
the
llvm:analysis
Member
|
@llvm/pr-subscribers-llvm-analysis Author: Aiden Grossman (boomanaiden154) ChangesInlineSizeEstimatorAnalysis was used initially to provide rewards for RL algorithms that preferred more immediate rewards. Now we have moved away from such training methodologies and this code has been sitting dead in tree providing a bit of a maintenance burden for a couple of years. Remove it given it is unused and requires some maintenance. This leaves around the actual IR2Native model for now as it's used in Patch is 23.49 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/167271.diff 6 Files Affected:
diff --git a/llvm/include/llvm/Analysis/InlineSizeEstimatorAnalysis.h b/llvm/include/llvm/Analysis/InlineSizeEstimatorAnalysis.h
deleted file mode 100644
index b44edd370dd1c..0000000000000
--- a/llvm/include/llvm/Analysis/InlineSizeEstimatorAnalysis.h
+++ /dev/null
@@ -1,47 +0,0 @@
-//===- InlineSizeEstimatorAnalysis.h - ML size estimator --------*- C++ -*-===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-//
-
-#ifndef LLVM_ANALYSIS_INLINESIZEESTIMATORANALYSIS_H
-#define LLVM_ANALYSIS_INLINESIZEESTIMATORANALYSIS_H
-
-#include "llvm/IR/PassManager.h"
-
-namespace llvm {
-class Function;
-
-class TFModelEvaluator;
-class InlineSizeEstimatorAnalysis
- : public AnalysisInfoMixin<InlineSizeEstimatorAnalysis> {
-public:
- InlineSizeEstimatorAnalysis();
- InlineSizeEstimatorAnalysis(InlineSizeEstimatorAnalysis &&);
- ~InlineSizeEstimatorAnalysis();
-
- static AnalysisKey Key;
- using Result = std::optional<size_t>;
- Result run(const Function &F, FunctionAnalysisManager &FAM);
- static bool isEvaluatorRequested();
-
-private:
- std::unique_ptr<TFModelEvaluator> Evaluator;
-};
-
-class InlineSizeEstimatorAnalysisPrinterPass
- : public PassInfoMixin<InlineSizeEstimatorAnalysisPrinterPass> {
- raw_ostream &OS;
-
-public:
- explicit InlineSizeEstimatorAnalysisPrinterPass(raw_ostream &OS) : OS(OS) {}
-
- PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM);
-
- static bool isRequired() { return true; }
-};
-} // namespace llvm
-#endif // LLVM_ANALYSIS_INLINESIZEESTIMATORANALYSIS_H
diff --git a/llvm/lib/Analysis/CMakeLists.txt b/llvm/lib/Analysis/CMakeLists.txt
index 88ebd65ec46af..bff9b62d98e06 100644
--- a/llvm/lib/Analysis/CMakeLists.txt
+++ b/llvm/lib/Analysis/CMakeLists.txt
@@ -89,7 +89,6 @@ add_llvm_component_library(LLVMAnalysis
InlineCost.cpp
InlineAdvisor.cpp
InlineOrder.cpp
- InlineSizeEstimatorAnalysis.cpp
InstCount.cpp
InstructionPrecedenceTracking.cpp
InstructionSimplify.cpp
diff --git a/llvm/lib/Analysis/DevelopmentModeInlineAdvisor.cpp b/llvm/lib/Analysis/DevelopmentModeInlineAdvisor.cpp
index 67e38ab8b35aa..d2be805a6f7a5 100644
--- a/llvm/lib/Analysis/DevelopmentModeInlineAdvisor.cpp
+++ b/llvm/lib/Analysis/DevelopmentModeInlineAdvisor.cpp
@@ -16,7 +16,6 @@
#include "llvm/ADT/BitVector.h"
#include "llvm/Analysis/CallGraph.h"
-#include "llvm/Analysis/InlineSizeEstimatorAnalysis.h"
#include "llvm/Analysis/MLInlineAdvisor.h"
#include "llvm/Analysis/ModelUnderTrainingRunner.h"
#include "llvm/Analysis/NoInferenceModelRunner.h"
@@ -89,9 +88,6 @@ struct InlineEvent {
/// error, even if AdvisedDecision were true, otherwise it agrees with
/// AdvisedDecision.
bool Effect = false;
-
- /// What the change in size was: size_after - size_before
- int64_t Reward = 0;
};
/// Collect data we may use for training a model.
@@ -150,31 +146,15 @@ class DevelopmentModeMLInlineAdvisor : public MLInlineAdvisor {
GetModelRunner,
std::function<bool(CallBase &)> GetDefaultAdvice);
- size_t getTotalSizeEstimate();
-
- void updateNativeSizeEstimate(int64_t Change) {
- *CurrentNativeSize += Change;
- }
- void resetNativeSize(Function *F) {
- PreservedAnalyses PA = PreservedAnalyses::all();
- PA.abandon<InlineSizeEstimatorAnalysis>();
- FAM.invalidate(*F, PA);
- }
-
std::unique_ptr<MLInlineAdvice>
getAdviceFromModel(CallBase &CB, OptimizationRemarkEmitter &ORE) override;
- std::optional<size_t> getNativeSizeEstimate(const Function &F) const;
-
private:
bool isLogging() const { return !!Logger; }
std::unique_ptr<MLInlineAdvice> getMandatoryAdviceImpl(CallBase &CB) override;
const bool IsDoingInference;
std::unique_ptr<TrainingLogger> Logger;
-
- const std::optional<int32_t> InitialNativeSize;
- std::optional<int32_t> CurrentNativeSize;
};
/// A variant of MLInlineAdvice that tracks all non-trivial inlining
@@ -183,13 +163,9 @@ class LoggingMLInlineAdvice : public MLInlineAdvice {
public:
LoggingMLInlineAdvice(DevelopmentModeMLInlineAdvisor *Advisor, CallBase &CB,
OptimizationRemarkEmitter &ORE, bool Recommendation,
- TrainingLogger &Logger,
- std::optional<size_t> CallerSizeEstimateBefore,
- std::optional<size_t> CalleeSizeEstimateBefore,
- bool DefaultDecision, bool Mandatory = false)
+ TrainingLogger &Logger, bool DefaultDecision,
+ bool Mandatory = false)
: MLInlineAdvice(Advisor, CB, ORE, Recommendation), Logger(Logger),
- CallerSizeEstimateBefore(CallerSizeEstimateBefore),
- CalleeSizeEstimateBefore(CalleeSizeEstimateBefore),
DefaultDecision(DefaultDecision), Mandatory(Mandatory) {}
virtual ~LoggingMLInlineAdvice() = default;
@@ -200,59 +176,35 @@ class LoggingMLInlineAdvice : public MLInlineAdvice {
}
void recordInliningImpl() override {
MLInlineAdvice::recordInliningImpl();
- getAdvisor()->resetNativeSize(Caller);
- int Reward = std::numeric_limits<int>::max();
- if (InlineSizeEstimatorAnalysis::isEvaluatorRequested() &&
- !getAdvisor()->isForcedToStop()) {
- int NativeSizeAfter = *getAdvisor()->getNativeSizeEstimate(*Caller) +
- *CalleeSizeEstimateBefore;
- Reward = NativeSizeAfter -
- (*CallerSizeEstimateBefore + *CalleeSizeEstimateBefore);
- getAdvisor()->updateNativeSizeEstimate(Reward);
- }
- log(Reward, /*Success=*/true);
+ log(/*Success=*/true);
}
void recordInliningWithCalleeDeletedImpl() override {
MLInlineAdvice::recordInliningWithCalleeDeletedImpl();
- getAdvisor()->resetNativeSize(Caller);
- if (InlineSizeEstimatorAnalysis::isEvaluatorRequested() &&
- !getAdvisor()->isForcedToStop()) {
- int NativeSizeAfter = *getAdvisor()->getNativeSizeEstimate(*Caller);
- int Reward = NativeSizeAfter -
- (*CallerSizeEstimateBefore + *CalleeSizeEstimateBefore);
- getAdvisor()->updateNativeSizeEstimate(Reward);
- log(Reward, /*Success=*/true);
- } else {
- log(NoReward, /*Success=*/true);
- }
+ log(/*Success=*/true);
}
void recordUnsuccessfulInliningImpl(const InlineResult &Result) override {
MLInlineAdvice::recordUnsuccessfulInliningImpl(Result);
- log(NoReward, /*Success=*/false);
+ log(/*Success=*/false);
}
void recordUnattemptedInliningImpl() override {
MLInlineAdvice::recordUnattemptedInliningImpl();
- log(NoReward, /*Success=*/false);
+ log(/*Success=*/false);
}
- void log(int64_t Reward, bool Success) {
+ void log(bool Success) {
if (Mandatory)
return;
InlineEvent Event;
Event.AdvisedDecision = isInliningRecommended();
Event.DefaultDecision = DefaultDecision;
Event.Effect = Success;
- Event.Reward = Reward;
Logger.logInlineEvent(Event, getAdvisor()->getModelRunner());
}
- static const int64_t NoReward = 0;
TrainingLogger &Logger;
- const std::optional<size_t> CallerSizeEstimateBefore;
- const std::optional<size_t> CalleeSizeEstimateBefore;
const int64_t DefaultDecision;
const int64_t Mandatory;
};
@@ -296,9 +248,9 @@ TrainingLogger::TrainingLogger(StringRef LogFileName,
if (EC)
dbgs() << (EC.message() + ":" + TrainingLog);
- L = std::make_unique<Logger>(
- std::move(OS), FT, TensorSpec::createSpec<int64_t>(RewardName, {1}),
- InlineSizeEstimatorAnalysis::isEvaluatorRequested());
+ L = std::make_unique<Logger>(std::move(OS), FT,
+ TensorSpec::createSpec<int64_t>(RewardName, {1}),
+ false);
L->switchContext("");
}
@@ -326,8 +278,6 @@ void TrainingLogger::logInlineEvent(const InlineEvent &Event,
L->logTensorValue(DecisionPos,
reinterpret_cast<const char *>(&Event.AdvisedDecision));
L->endObservation();
- if (InlineSizeEstimatorAnalysis::isEvaluatorRequested())
- L->logReward(Event.Reward);
// For debugging / later use
Effects.push_back(Event.Effect);
@@ -340,9 +290,7 @@ DevelopmentModeMLInlineAdvisor::DevelopmentModeMLInlineAdvisor(
GetModelRunner,
std::function<bool(CallBase &)> GetDefaultAdvice)
: MLInlineAdvisor(M, MAM, GetModelRunner, GetDefaultAdvice),
- IsDoingInference(isa<ModelUnderTrainingRunner>(getModelRunner())),
- InitialNativeSize(isLogging() ? getTotalSizeEstimate() : 0),
- CurrentNativeSize(InitialNativeSize) {
+ IsDoingInference(isa<ModelUnderTrainingRunner>(getModelRunner())) {
// We cannot have the case of neither inference nor logging.
if (!TrainingLog.empty())
Logger = std::make_unique<TrainingLogger>(
@@ -351,29 +299,12 @@ DevelopmentModeMLInlineAdvisor::DevelopmentModeMLInlineAdvisor(
assert(IsDoingInference || isLogging());
}
-std::optional<size_t>
-DevelopmentModeMLInlineAdvisor::getNativeSizeEstimate(const Function &F) const {
- if (!InlineSizeEstimatorAnalysis::isEvaluatorRequested())
- return std::nullopt;
- auto &R =
- FAM.getResult<InlineSizeEstimatorAnalysis>(const_cast<Function &>(F));
- if (!R) {
- F.getParent()->getContext().emitError(
- "Native size estimator is not present.");
- return 0;
- }
- return *R;
-}
-
std::unique_ptr<MLInlineAdvice>
DevelopmentModeMLInlineAdvisor::getMandatoryAdviceImpl(CallBase &CB) {
return std::make_unique<LoggingMLInlineAdvice>(
/*Advisor=*/this,
/*CB=*/CB, /*ORE=*/getCallerORE(CB), /*Recommendation=*/true,
/*Logger=*/*Logger,
- /*CallerSizeEstimateBefore=*/getNativeSizeEstimate(*CB.getCaller()),
- /*CalleeSizeEstimateBefore=*/
- getNativeSizeEstimate(*CB.getCalledFunction()),
/*DefaultDecision=*/true, /*Mandatory*/ true);
}
@@ -391,24 +322,9 @@ DevelopmentModeMLInlineAdvisor::getAdviceFromModel(
/*Advisor=*/this,
/*CB=*/CB, /*ORE=*/ORE, /*Recommendation=*/Recommendation,
/*Logger=*/*Logger,
- /*CallerSizeEstimateBefore=*/getNativeSizeEstimate(*CB.getCaller()),
- /*CalleeSizeEstimateBefore=*/
- getNativeSizeEstimate(*CB.getCalledFunction()),
/*DefaultDecision=*/DefaultAdvice);
}
-size_t DevelopmentModeMLInlineAdvisor::getTotalSizeEstimate() {
- if (!InlineSizeEstimatorAnalysis::isEvaluatorRequested())
- return 0;
- size_t Ret = 0;
- for (auto &F : M) {
- if (F.isDeclaration())
- continue;
- Ret += *getNativeSizeEstimate(F);
- }
- return Ret;
-}
-
std::unique_ptr<InlineAdvisor> llvm::getDevelopmentModeAdvisor(
Module &M, ModuleAnalysisManager &MAM,
std::function<bool(CallBase &)> GetDefaultAdvice) {
diff --git a/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp b/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp
deleted file mode 100644
index fc635726a6aa4..0000000000000
--- a/llvm/lib/Analysis/InlineSizeEstimatorAnalysis.cpp
+++ /dev/null
@@ -1,281 +0,0 @@
-//===- InlineSizeEstimatorAnalysis.cpp - IR to native size from ML model --===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-//
-// This implements feature and label extraction for offline supervised learning
-// of a IR to native size model.
-//
-//===----------------------------------------------------------------------===//
-#include "llvm/Analysis/InlineSizeEstimatorAnalysis.h"
-
-#ifdef LLVM_HAVE_TFLITE
-#include "llvm/Analysis/Utils/TFUtils.h"
-#endif
-#include "llvm/IR/Function.h"
-#include "llvm/IR/PassManager.h"
-#include "llvm/Support/raw_ostream.h"
-
-using namespace llvm;
-
-AnalysisKey InlineSizeEstimatorAnalysis::Key;
-
-#ifdef LLVM_HAVE_TFLITE
-#include "llvm/Analysis/LoopInfo.h"
-#include "llvm/Analysis/TargetLibraryInfo.h"
-#include "llvm/Analysis/TargetTransformInfo.h"
-#include "llvm/IR/BasicBlock.h"
-#include "llvm/IR/Dominators.h"
-#include "llvm/IR/Instructions.h"
-#include "llvm/Support/Casting.h"
-#include "llvm/Support/CommandLine.h"
-#include <algorithm>
-#include <deque>
-#include <optional>
-
-static cl::opt<std::string> TFIR2NativeModelPath(
- "ml-inliner-ir2native-model", cl::Hidden,
- cl::desc("Path to saved model evaluating native size from IR."));
-
-#define DEBUG_TYPE "inline-size-estimator"
-namespace {
-unsigned getMaxInstructionID() {
-#define LAST_OTHER_INST(NR) return NR;
-#include "llvm/IR/Instruction.def"
-}
-
-class IRToNativeSizeLearning {
-public:
- enum class NamedFeatureIndex : size_t {
- InitialSize,
- Blocks,
- Calls,
- IsLocal,
- IsLinkOnceODR,
- IsLinkOnce,
- Loops,
- MaxLoopDepth,
- MaxDomTreeLevel,
-
- NumNamedFeatures
- };
- static const size_t NumNamedFeatures =
- static_cast<size_t>(NamedFeatureIndex::NumNamedFeatures);
- struct FunctionFeatures {
- static const size_t FeatureCount;
-
- std::array<int32_t, NumNamedFeatures> NamedFeatures = {0};
- std::vector<int32_t> InstructionHistogram;
- std::vector<int32_t> InstructionPairHistogram;
-
- void fillTensor(int32_t *Ptr) const;
- int32_t &operator[](NamedFeatureIndex Pos) {
- return NamedFeatures[static_cast<size_t>(Pos)];
- }
- };
- IRToNativeSizeLearning() = default;
-
- static FunctionFeatures getFunctionFeatures(Function &F,
- FunctionAnalysisManager &FAM);
-};
-
-// This is a point in time - we determined including these pairs of
-// consecutive instructions (in the IR layout available at inline time) as
-// features improves the model performance. We want to move away from manual
-// feature selection.
-// The array is given in opcode pairs rather than labels because 1) labels
-// weren't readily available, and 2) the successions were hand - extracted.
-//
-// This array must be sorted.
-static const std::array<std::pair<size_t, size_t>, 137>
- ImportantInstructionSuccessions{
- {{1, 1}, {1, 4}, {1, 5}, {1, 7}, {1, 8}, {1, 9}, {1, 11},
- {1, 12}, {1, 13}, {1, 14}, {1, 18}, {1, 20}, {1, 22}, {1, 24},
- {1, 25}, {1, 26}, {1, 27}, {1, 28}, {1, 29}, {1, 30}, {1, 31},
- {1, 32}, {1, 33}, {1, 34}, {1, 39}, {1, 40}, {1, 42}, {1, 45},
- {2, 1}, {2, 2}, {2, 13}, {2, 28}, {2, 29}, {2, 32}, {2, 33},
- {2, 34}, {2, 38}, {2, 48}, {2, 49}, {2, 53}, {2, 55}, {2, 56},
- {13, 2}, {13, 13}, {13, 26}, {13, 33}, {13, 34}, {13, 56}, {15, 27},
- {28, 2}, {28, 48}, {28, 53}, {29, 2}, {29, 33}, {29, 56}, {31, 31},
- {31, 33}, {31, 34}, {31, 49}, {32, 1}, {32, 2}, {32, 13}, {32, 15},
- {32, 28}, {32, 29}, {32, 32}, {32, 33}, {32, 34}, {32, 39}, {32, 40},
- {32, 48}, {32, 49}, {32, 53}, {32, 56}, {33, 1}, {33, 2}, {33, 32},
- {33, 33}, {33, 34}, {33, 49}, {33, 53}, {33, 56}, {34, 1}, {34, 2},
- {34, 32}, {34, 33}, {34, 34}, {34, 49}, {34, 53}, {34, 56}, {38, 34},
- {39, 57}, {40, 34}, {47, 15}, {47, 49}, {48, 2}, {48, 34}, {48, 56},
- {49, 1}, {49, 2}, {49, 28}, {49, 32}, {49, 33}, {49, 34}, {49, 39},
- {49, 49}, {49, 56}, {53, 1}, {53, 2}, {53, 28}, {53, 34}, {53, 53},
- {53, 57}, {55, 1}, {55, 28}, {55, 34}, {55, 53}, {55, 55}, {55, 56},
- {56, 1}, {56, 2}, {56, 7}, {56, 13}, {56, 32}, {56, 33}, {56, 34},
- {56, 49}, {56, 53}, {56, 56}, {56, 64}, {57, 34}, {57, 56}, {57, 57},
- {64, 1}, {64, 64}, {65, 1}, {65, 65}}};
-
-// We have: 9 calculated features (the features here); 1 feature for each
-// instruction opcode; and 1 feature for each manually-identified sequence.
-// For the latter 2, we build a histogram: we count the number of
-// occurrences of each instruction opcode or succession of instructions,
-// respectively.
-// Note that instruction opcodes start from 1. For convenience, we also have an
-// always 0 feature for the '0' opcode, hence the extra 1.
-const size_t IRToNativeSizeLearning::FunctionFeatures::FeatureCount =
- ImportantInstructionSuccessions.size() + getMaxInstructionID() + 1 +
- IRToNativeSizeLearning::NumNamedFeatures;
-
-size_t getSize(Function &F, TargetTransformInfo &TTI) {
- size_t Ret = 0;
- for (const auto &BB : F)
- for (const auto &I : BB)
- Ret += TTI.getInstructionCost(
- &I, TargetTransformInfo::TargetCostKind::TCK_CodeSize)
- .getValue();
- return Ret;
-}
-
-size_t getSize(Function &F, FunctionAnalysisManager &FAM) {
- auto &TTI = FAM.getResult<TargetIRAnalysis>(F);
- return getSize(F, TTI);
-}
-
-unsigned getMaxDominatorTreeDepth(const Function &F,
- const DominatorTree &Tree) {
- unsigned Ret = 0;
- for (const auto &BB : F)
- if (const auto *TN = Tree.getNode(&BB))
- Ret = std::max(Ret, TN->getLevel());
- return Ret;
-}
-} // namespace
-
-IRToNativeSizeLearning::FunctionFeatures
-IRToNativeSizeLearning::getFunctionFeatures(Function &F,
- FunctionAnalysisManager &FAM) {
- assert(llvm::is_sorted(ImportantInstructionSuccessions) &&
- "expected function features are sorted");
-
- auto &DomTree = FAM.getResult<DominatorTreeAnalysis>(F);
- FunctionFeatures FF;
- size_t InstrCount = getMaxInstructionID() + 1;
- FF.InstructionHistogram.resize(InstrCount);
-
- FF.InstructionPairHistogram.resize(ImportantInstructionSuccessions.size());
-
- int StartID = 0;
- int LastID = StartID;
- auto getPairIndex = [](size_t a, size_t b) {
- auto I = llvm::find(ImportantInstructionSuccessions, std::make_pair(a, b));
- if (I == ImportantInstructionSuccessions.end())
- return -1;
- return static_cast<int>(
- std::distance(ImportantInstructionSuccessions.begin(), I));
- };
-
- // We don't want debug calls, because they'd just add noise.
- for (const auto &BB : F) {
- for (const auto &I : BB.instructionsWithoutDebug()) {
- auto ID = I.getOpcode();
-
- ++FF.InstructionHistogram[ID];
- int PairIndex = getPairIndex(LastID, ID);
- if (PairIndex >= 0)
- ++FF.InstructionPairHistogram[PairIndex];
- LastID = ID;
- if (isa<CallBase>(I))
- ++FF[NamedFeatureIndex::Calls];
- }
- }
-
- FF[NamedFeatureIndex::InitialSize] = getSize(F, FAM);
- FF[NamedFeatureIndex::IsLocal] = F.hasLocalLinkage();
- FF[NamedFeatureIndex::IsLinkOnceODR] = F.hasLinkOnceODRLinkage();
- FF[NamedFeatureIndex::IsLinkOnce] = F.hasLinkOnceLinkage();
- FF[NamedFeatureIndex::Blocks] = F.size();
- auto &LI = FAM.getResult<LoopAnalysis>(F);
- FF[NamedFeatureIndex::Loops] = std::distance(LI.begin(), LI.end());
- for (auto &L : LI)
- FF[NamedFeatureIndex::MaxLoopDepth] =
- std::max(FF[NamedFeatureIndex::MaxLoopDepth],
- static_cast<int32_t>(L->getLoopDepth()));
- FF[NamedFeatureIndex::MaxDomTreeLevel] = getMaxDominatorTreeDepth(F, DomTree);
- return FF;
-}
-
-void IRToNativeSizeLearning::FunctionFeatures::fillTensor(int32_t *Ptr) const {
- std::copy(NamedFeatures.begin(), NamedFeatures.end(), Ptr);
- Ptr += NamedFeatures.size();
- std::copy(InstructionHistogram.begin(), InstructionHistogram.end(), Ptr);
- Ptr += InstructionHistogram.size();
- std::copy(InstructionPairHistogram.begin(), InstructionPairHistogram.end(),
- Ptr);
-}
-
-bool InlineSizeEstimatorAnalysis::isEvaluatorRequested() {
- return !TFIR2NativeModelPath.empty();
-}
-
-InlineSizeEstimatorAnalysis::InlineSizeEstimatorAnalysis() {
- if (!isEvaluatorRequested()) {
- return;
- }
- std::vector<TensorSpec> InputSpecs{TensorSpec::createSpec<int32_t>(
- "serving_default_input_1",
- {1, static_cast<int64_t>(
- IRToNativeSizeLearning::FunctionFeatures::FeatureCount)})};
- std::vector<TensorSpec> OutputSpecs{
- TensorSpec::createSpec<float>("StatefulPartitionedCall", {1})};
- Evaluator = std::make_unique<TFModelEvaluator>(
- TFIR2NativeModelPath.getValue().c_str(), InputSpecs, OutputSpecs);
- if (!Evaluator || !Evaluator->isValid()) {
- Evaluator.reset();
- return;
- }
-}
-
-InlineSizeEstimatorAn...
[truncated]
|
mtrofin
reviewed
Nov 10, 2025
Member
mtrofin
left a comment
mtrofin
left a comment
There was a problem hiding this comment.
LGTM, but have you done a build with tflite and ran tests? I'm surprised there are no .ll files that needed changing.
Contributor
Author
Yeah, I did a local build and everything passed. The only thing that failed was the TFLite unit test because I also tried to delete the |
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/72/builds/15721 Here is the relevant piece of the build log for the reference |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
InlineSizeEstimatorAnalysis was used initially to provide rewards for RL algorithms that preferred more immediate rewards. Now we have moved away from such training methodologies and this code has been sitting dead in tree providing a bit of a maintenance burden for a couple of years. Remove it given it is unused and requires some maintenance. This leaves around the actual IR2Native model for now as it's used in
TFUtilsTest.cpp. Eventually we should probably remove that too.