AMDGPUTargetMachine.cpp File Reference

#include "AMDGPUTargetMachine.h"
#include "AMDGPU.h"
#include "AMDGPUAliasAnalysis.h"
#include "AMDGPUExportClustering.h"
#include "AMDGPUMacroFusion.h"
#include "AMDGPUTargetObjectFile.h"
#include "AMDGPUTargetTransformInfo.h"
#include "GCNIterativeScheduler.h"
#include "GCNSchedStrategy.h"
#include "R600.h"
#include "R600TargetMachine.h"
#include "SIMachineFunctionInfo.h"
#include "SIMachineScheduler.h"
#include "TargetInfo/AMDGPUTargetInfo.h"
#include "llvm/Analysis/CGSCCPassManager.h"
#include "llvm/CodeGen/GlobalISel/IRTranslator.h"
#include "llvm/CodeGen/GlobalISel/InstructionSelect.h"
#include "llvm/CodeGen/GlobalISel/Legalizer.h"
#include "llvm/CodeGen/GlobalISel/Localizer.h"
#include "llvm/CodeGen/GlobalISel/RegBankSelect.h"
#include "llvm/CodeGen/MIRParser/MIParser.h"
#include "llvm/CodeGen/Passes.h"
#include "llvm/CodeGen/RegAllocRegistry.h"
#include "llvm/CodeGen/TargetPassConfig.h"
#include "llvm/IR/IntrinsicsAMDGPU.h"
#include "llvm/IR/LegacyPassManager.h"
#include "llvm/IR/PassManager.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/InitializePasses.h"
#include "llvm/MC/TargetRegistry.h"
#include "llvm/Passes/PassBuilder.h"
#include "llvm/Transforms/IPO.h"
#include "llvm/Transforms/IPO/AlwaysInliner.h"
#include "llvm/Transforms/IPO/GlobalDCE.h"
#include "llvm/Transforms/IPO/Internalize.h"
#include "llvm/Transforms/IPO/PassManagerBuilder.h"
#include "llvm/Transforms/Scalar.h"
#include "llvm/Transforms/Scalar/GVN.h"
#include "llvm/Transforms/Scalar/InferAddressSpaces.h"
#include "llvm/Transforms/Utils.h"
#include "llvm/Transforms/Utils/SimplifyLibCalls.h"
#include "llvm/Transforms/Vectorize.h"

Include dependency graph for AMDGPUTargetMachine.cpp:

Go to the source code of this file.

Functions
LLVM_EXTERNAL_VISIBILITY void	LLVMInitializeAMDGPUTarget ()
static std::unique_ptr< TargetLoweringObjectFile >	createTLOF (const Triple &TT)
static ScheduleDAGInstrs *	createSIMachineScheduler (MachineSchedContext *C)
static ScheduleDAGInstrs *	createGCNMaxOccupancyMachineScheduler (MachineSchedContext *C)
static ScheduleDAGInstrs *	createIterativeGCNMaxOccupancyMachineScheduler (MachineSchedContext *C)
static ScheduleDAGInstrs *	createMinRegScheduler (MachineSchedContext *C)
static ScheduleDAGInstrs *	createIterativeILPMachineScheduler (MachineSchedContext *C)
static StringRef	computeDataLayout (const Triple &TT)
static LLVM_READNONE StringRef	getGPUOrDefault (const Triple &TT, StringRef GPU)
static Reloc::Model	getEffectiveRelocModel (Optional< Reloc::Model > RM)
static bool	mustPreserveGV (const GlobalValue &GV)
	Predicate for Internalize pass. More...

Variables
static cl::opt< bool >	EnableSROA ("amdgpu-sroa", cl::desc("Run SROA after promote alloca pass"), cl::ReallyHidden, cl::init(true))
static cl::opt< bool >	EnableEarlyIfConversion ("amdgpu-early-ifcvt", cl::Hidden, cl::desc("Run early if-conversion"), cl::init(false))
static cl::opt< bool >	OptExecMaskPreRA ("amdgpu-opt-exec-mask-pre-ra", cl::Hidden, cl::desc("Run pre-RA exec mask optimizations"), cl::init(true))
static cl::opt< bool >	EnableLoadStoreVectorizer ("amdgpu-load-store-vectorizer", cl::desc("Enable load store vectorizer"), cl::init(true), cl::Hidden)
static cl::opt< bool >	ScalarizeGlobal ("amdgpu-scalarize-global-loads", cl::desc("Enable global load scalarization"), cl::init(true), cl::Hidden)
static cl::opt< bool >	InternalizeSymbols ("amdgpu-internalize-symbols", cl::desc("Enable elimination of non-kernel functions and unused globals"), cl::init(false), cl::Hidden)
static cl::opt< bool >	EarlyInlineAll ("amdgpu-early-inline-all", cl::desc("Inline all functions early"), cl::init(false), cl::Hidden)
static cl::opt< bool >	EnableSDWAPeephole ("amdgpu-sdwa-peephole", cl::desc("Enable SDWA peepholer"), cl::init(true))
static cl::opt< bool >	EnableDPPCombine ("amdgpu-dpp-combine", cl::desc("Enable DPP combiner"), cl::init(true))
static cl::opt< bool >	EnableAMDGPUAliasAnalysis ("enable-amdgpu-aa", cl::Hidden, cl::desc("Enable AMDGPU Alias Analysis"), cl::init(true))
static cl::opt< bool, true >	LateCFGStructurize ("amdgpu-late-structurize", cl::desc("Enable late CFG structurization"), cl::location(AMDGPUTargetMachine::EnableLateStructurizeCFG), cl::Hidden)
static cl::opt< bool >	EnableLibCallSimplify ("amdgpu-simplify-libcall", cl::desc("Enable amdgpu library simplifications"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableLowerKernelArguments ("amdgpu-ir-lower-kernel-arguments", cl::desc("Lower kernel argument loads in IR pass"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableRegReassign ("amdgpu-reassign-regs", cl::desc("Enable register reassign optimizations on gfx10+"), cl::init(true), cl::Hidden)
static cl::opt< bool >	OptVGPRLiveRange ("amdgpu-opt-vgpr-liverange", cl::desc("Enable VGPR liverange optimizations for if-else structure"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableAtomicOptimizations ("amdgpu-atomic-optimizations", cl::desc("Enable atomic optimizations"), cl::init(false), cl::Hidden)
static cl::opt< bool >	EnableSIModeRegisterPass ("amdgpu-mode-register", cl::desc("Enable mode register pass"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableDCEInRA ("amdgpu-dce-in-ra", cl::init(true), cl::Hidden, cl::desc("Enable machine DCE inside regalloc"))
static cl::opt< bool >	EnableScalarIRPasses ("amdgpu-scalar-ir-passes", cl::desc("Enable scalar IR passes"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableStructurizerWorkarounds ("amdgpu-enable-structurizer-workarounds", cl::desc("Enable workarounds for the StructurizeCFG pass"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnableLDSReplaceWithPointer ("amdgpu-enable-lds-replace-with-pointer", cl::desc("Enable LDS replace with pointer pass"), cl::init(false), cl::Hidden)
static cl::opt< bool, true >	EnableLowerModuleLDS ("amdgpu-enable-lower-module-lds", cl::desc("Enable lower module lds pass"), cl::location(AMDGPUTargetMachine::EnableLowerModuleLDS), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnablePreRAOptimizations ("amdgpu-enable-pre-ra-optimizations", cl::desc("Enable Pre-RA optimizations pass"), cl::init(true), cl::Hidden)
static cl::opt< bool >	EnablePromoteKernelArguments ("amdgpu-enable-promote-kernel-arguments", cl::desc("Enable promotion of flat kernel pointer arguments to global"), cl::Hidden, cl::init(true))
static MachineSchedRegistry	SISchedRegistry ("si", "Run SI's custom scheduler", createSIMachineScheduler)
static MachineSchedRegistry	GCNMaxOccupancySchedRegistry ("gcn-max-occupancy", "Run GCN scheduler to maximize occupancy", createGCNMaxOccupancyMachineScheduler)
static MachineSchedRegistry	IterativeGCNMaxOccupancySchedRegistry ("gcn-max-occupancy-experimental", "Run GCN scheduler to maximize occupancy (experimental)", createIterativeGCNMaxOccupancyMachineScheduler)
static MachineSchedRegistry	GCNMinRegSchedRegistry ("gcn-minreg", "Run GCN iterative scheduler for minimal register usage (experimental)", createMinRegScheduler)
static MachineSchedRegistry	GCNILPSchedRegistry ("gcn-ilp", "Run GCN iterative scheduler for ILP scheduling (experimental)", createIterativeILPMachineScheduler)
static const char	RegAllocOptNotSupportedMessage []

Detailed Description

The AMDGPU target machine contains all of the hardware specific information needed to emit code for SI+ GPUs.

Definition in file AMDGPUTargetMachine.cpp.

Function Documentation

computeDataLayout()

static StringRef computeDataLayout ( const Triple &  TT )

static

Definition at line 447 of file AMDGPUTargetMachine.cpp.

References llvm::Triple::r600.

createGCNMaxOccupancyMachineScheduler()

static ScheduleDAGInstrs* createGCNMaxOccupancyMachineScheduler ( MachineSchedContext *  C )

static

Definition at line 392 of file AMDGPUTargetMachine.cpp.

References llvm::ScheduleDAGMI::addMutation(), llvm::createAMDGPUExportClusteringDAGMutation(), llvm::createAMDGPUMacroFusionDAGMutation(), llvm::createLoadClusterDAGMutation(), llvm::ScheduleDAG::TII, and llvm::ScheduleDAG::TRI.

createIterativeGCNMaxOccupancyMachineScheduler()

static ScheduleDAGInstrs* createIterativeGCNMaxOccupancyMachineScheduler ( MachineSchedContext *  C )

static

Definition at line 402 of file AMDGPUTargetMachine.cpp.

References llvm::createLoadClusterDAGMutation(), and llvm::GCNIterativeScheduler::SCHEDULE_LEGACYMAXOCCUPANCY.

createIterativeILPMachineScheduler()

static ScheduleDAGInstrs* createIterativeILPMachineScheduler ( MachineSchedContext *  C )

static

Definition at line 415 of file AMDGPUTargetMachine.cpp.

References llvm::createAMDGPUMacroFusionDAGMutation(), llvm::createLoadClusterDAGMutation(), and llvm::GCNIterativeScheduler::SCHEDULE_ILP.

createMinRegScheduler()

static ScheduleDAGInstrs* createMinRegScheduler ( MachineSchedContext *  C )

static

Definition at line 409 of file AMDGPUTargetMachine.cpp.

References llvm::GCNIterativeScheduler::SCHEDULE_MINREGFORCED.

createSIMachineScheduler()

static ScheduleDAGInstrs* createSIMachineScheduler ( MachineSchedContext *  C )

static

Definition at line 387 of file AMDGPUTargetMachine.cpp.

createTLOF()

static std::unique_ptr<TargetLoweringObjectFile> createTLOF ( const Triple &  TT )

static

Definition at line 383 of file AMDGPUTargetMachine.cpp.

getEffectiveRelocModel()

static Reloc::Model getEffectiveRelocModel ( Optional< Reloc::Model >  RM )

static

Definition at line 474 of file AMDGPUTargetMachine.cpp.

References llvm::Reloc::PIC_.

getGPUOrDefault()

static LLVM_READNONE StringRef getGPUOrDefault ( const Triple &  TT, StringRef  GPU  )

static

Definition at line 463 of file AMDGPUTargetMachine.cpp.

References llvm::Triple::amdgcn, llvm::Triple::AMDHSA, and llvm::StringRef::empty().

LLVMInitializeAMDGPUTarget()

LLVM_EXTERNAL_VISIBILITY void LLVMInitializeAMDGPUTarget

(

)

Definition at line 309 of file AMDGPUTargetMachine.cpp.

References llvm::PassRegistry::getPassRegistry(), llvm::getTheAMDGPUTarget(), llvm::getTheGCNTarget(), llvm::initializeAMDGPUAAWrapperPassPass(), llvm::initializeAMDGPUAlwaysInlinePass(), llvm::initializeAMDGPUAnnotateKernelFeaturesPass(), llvm::initializeAMDGPUAnnotateUniformValuesPass(), llvm::initializeAMDGPUArgumentUsageInfoPass(), llvm::initializeAMDGPUAtomicOptimizerPass(), llvm::initializeAMDGPUAttributorPass(), llvm::initializeAMDGPUCodeGenPreparePass(), llvm::initializeAMDGPUCtorDtorLoweringPass(), llvm::initializeAMDGPUDAGToDAGISelPass(), llvm::initializeAMDGPUExternalAAWrapperPass(), llvm::initializeAMDGPUFixFunctionBitcastsPass(), llvm::initializeAMDGPULateCodeGenPreparePass(), llvm::initializeAMDGPULowerIntrinsicsPass(), llvm::initializeAMDGPULowerKernelArgumentsPass(), llvm::initializeAMDGPULowerKernelAttributesPass(), llvm::initializeAMDGPULowerModuleLDSPass(), llvm::initializeAMDGPUOpenCLEnqueuedBlockLoweringPass(), llvm::initializeAMDGPUPostLegalizerCombinerPass(), llvm::initializeAMDGPUPreLegalizerCombinerPass(), llvm::initializeAMDGPUPrintfRuntimeBindingPass(), llvm::initializeAMDGPUPromoteAllocaPass(), llvm::initializeAMDGPUPromoteAllocaToVectorPass(), llvm::initializeAMDGPUPromoteKernelArgumentsPass(), llvm::initializeAMDGPUPropagateAttributesEarlyPass(), llvm::initializeAMDGPUPropagateAttributesLatePass(), llvm::initializeAMDGPURegBankCombinerPass(), llvm::initializeAMDGPUReplaceLDSUseWithPointerPass(), llvm::initializeAMDGPUResourceUsageAnalysisPass(), llvm::initializeAMDGPURewriteOutArgumentsPass(), llvm::initializeAMDGPUSimplifyLibCallsPass(), llvm::initializeAMDGPUUnifyDivergentExitNodesPass(), llvm::initializeAMDGPUUnifyMetadataPass(), llvm::initializeAMDGPUUseNativeCallsPass(), llvm::initializeGCNDPPCombinePass(), llvm::initializeGCNNSAReassignPass(), llvm::initializeGCNPreRAOptimizationsPass(), llvm::initializeGlobalISel(), llvm::initializeR600ClauseMergePassPass(), llvm::initializeR600ControlFlowFinalizerPass(), llvm::initializeR600ExpandSpecialInstrsPassPass(), llvm::initializeR600PacketizerPass(), llvm::initializeR600VectorRegMergerPass(), llvm::initializeSIAnnotateControlFlowPass(), llvm::initializeSIFixSGPRCopiesPass(), llvm::initializeSIFixVGPRCopiesPass(), llvm::initializeSIFoldOperandsPass(), llvm::initializeSIFormMemoryClausesPass(), llvm::initializeSIInsertHardClausesPass(), llvm::initializeSIInsertWaitcntsPass(), llvm::initializeSILateBranchLoweringPass(), llvm::initializeSILoadStoreOptimizerPass(), llvm::initializeSILowerControlFlowPass(), llvm::initializeSILowerI1CopiesPass(), llvm::initializeSILowerSGPRSpillsPass(), llvm::initializeSIMemoryLegalizerPass(), llvm::initializeSIModeRegisterPass(), llvm::initializeSIOptimizeExecMaskingPass(), llvm::initializeSIOptimizeExecMaskingPreRAPass(), llvm::initializeSIOptimizeVGPRLiveRangePass(), llvm::initializeSIPeepholeSDWAPass(), llvm::initializeSIPostRABundlerPass(), llvm::initializeSIPreAllocateWWMRegsPass(), llvm::initializeSIPreEmitPeepholePass(), llvm::initializeSIShrinkInstructionsPass(), llvm::initializeSIWholeQuadModePass(), X, and Y.

mustPreserveGV()

static bool mustPreserveGV ( const GlobalValue &  GV )

static

Predicate for Internalize pass.

Definition at line 518 of file AMDGPUTargetMachine.cpp.

References F, llvm::AMDGPU::isEntryFunctionCC(), llvm::Constant::removeDeadConstantUsers(), and llvm::Value::use_empty().

Referenced by llvm::AMDGPUTargetMachine::adjustPassManager(), and llvm::AMDGPUTargetMachine::registerPassBuilderCallbacks().

Variable Documentation

EarlyInlineAll

cl::opt<bool> EarlyInlineAll("amdgpu-early-inline-all", cl::desc("Inline all functions early"), cl::init(false), cl::Hidden)

static

Referenced by llvm::AMDGPUTargetMachine::adjustPassManager(), and llvm::AMDGPUTargetMachine::registerPassBuilderCallbacks().

EnableAMDGPUAliasAnalysis

cl::opt<bool> EnableAMDGPUAliasAnalysis("enable-amdgpu-aa", cl::Hidden, cl::desc("Enable AMDGPU Alias Analysis"), cl::init(true))

static

Referenced by llvm::AMDGPUPassConfig::addIRPasses(), and llvm::AMDGPUTargetMachine::adjustPassManager().

EnableAtomicOptimizations

cl::opt<bool> EnableAtomicOptimizations("amdgpu-atomic-optimizations", cl::desc("Enable atomic optimizations"), cl::init(false), cl::Hidden)

static

EnableDCEInRA

cl::opt<bool> EnableDCEInRA("amdgpu-dce-in-ra", cl::init(true), cl::Hidden, cl::desc("Enable machine DCE inside regalloc"))

static

EnableDPPCombine

cl::opt<bool> EnableDPPCombine("amdgpu-dpp-combine", cl::desc("Enable DPP combiner"), cl::init(true))

static

EnableEarlyIfConversion

cl::opt<bool> EnableEarlyIfConversion("amdgpu-early-ifcvt", cl::Hidden, cl::desc("Run early if-conversion"), cl::init(false))

static

EnableLDSReplaceWithPointer

cl::opt<bool> EnableLDSReplaceWithPointer("amdgpu-enable-lds-replace-with-pointer", cl::desc("Enable LDS replace with pointer pass"), cl::init(false), cl::Hidden)

static

Referenced by llvm::AMDGPUPassConfig::addIRPasses().

EnableLibCallSimplify

cl::opt<bool> EnableLibCallSimplify("amdgpu-simplify-libcall", cl::desc("Enable amdgpu library simplifications"), cl::init(true), cl::Hidden)

static

Referenced by llvm::AMDGPUTargetMachine::adjustPassManager(), and llvm::AMDGPUTargetMachine::registerPassBuilderCallbacks().

EnableLoadStoreVectorizer

cl::opt<bool> EnableLoadStoreVectorizer("amdgpu-load-store-vectorizer", cl::desc("Enable load store vectorizer"), cl::init(true), cl::Hidden)

static

Referenced by llvm::AMDGPUPassConfig::addCodeGenPrepare().

EnableLowerKernelArguments

cl::opt<bool> EnableLowerKernelArguments("amdgpu-ir-lower-kernel-arguments", cl::desc("Lower kernel argument loads in IR pass"), cl::init(true), cl::Hidden)

static

Referenced by llvm::AMDGPUPassConfig::addCodeGenPrepare().

EnableLowerModuleLDS

cl::opt<bool, true> EnableLowerModuleLDS("amdgpu-enable-lower-module-lds", cl::desc("Enable lower module lds pass"), cl::location(AMDGPUTargetMachine::EnableLowerModuleLDS), cl::init(true), cl::Hidden)

static

Referenced by llvm::AMDGPUPassConfig::addIRPasses().

EnablePreRAOptimizations

cl::opt<bool> EnablePreRAOptimizations("amdgpu-enable-pre-ra-optimizations", cl::desc("Enable Pre-RA optimizations pass"), cl::init(true), cl::Hidden)

static

EnablePromoteKernelArguments

cl::opt<bool> EnablePromoteKernelArguments("amdgpu-enable-promote-kernel-arguments", cl::desc("Enable promotion of flat kernel pointer arguments to global"), cl::Hidden, cl::init(true))

static

Referenced by llvm::AMDGPUTargetMachine::adjustPassManager(), and llvm::AMDGPUTargetMachine::registerPassBuilderCallbacks().

EnableRegReassign

cl::opt<bool> EnableRegReassign("amdgpu-reassign-regs", cl::desc("Enable register reassign optimizations on gfx10+"), cl::init(true), cl::Hidden)

static

EnableScalarIRPasses

cl::opt<bool> EnableScalarIRPasses("amdgpu-scalar-ir-passes", cl::desc("Enable scalar IR passes"), cl::init(true), cl::Hidden)

static

Referenced by llvm::AMDGPUPassConfig::addIRPasses().

EnableSDWAPeephole

cl::opt<bool> EnableSDWAPeephole("amdgpu-sdwa-peephole", cl::desc("Enable SDWA peepholer"), cl::init(true))

static

EnableSIModeRegisterPass

cl::opt<bool> EnableSIModeRegisterPass("amdgpu-mode-register", cl::desc("Enable mode register pass"), cl::init(true), cl::Hidden)

static

EnableSROA

cl::opt<bool> EnableSROA("amdgpu-sroa", cl::desc("Run SROA after promote alloca pass"), cl::ReallyHidden, cl::init(true))

static

Referenced by llvm::AMDGPUPassConfig::addIRPasses().

EnableStructurizerWorkarounds

cl::opt<bool> EnableStructurizerWorkarounds("amdgpu-enable-structurizer-workarounds", cl::desc("Enable workarounds for the StructurizeCFG pass"), cl::init(true), cl::Hidden)

static

GCNILPSchedRegistry

MachineSchedRegistry GCNILPSchedRegistry("gcn-ilp", "Run GCN iterative scheduler for ILP scheduling (experimental)", createIterativeILPMachineScheduler)

static

GCNMaxOccupancySchedRegistry

MachineSchedRegistry GCNMaxOccupancySchedRegistry("gcn-max-occupancy", "Run GCN scheduler to maximize occupancy", createGCNMaxOccupancyMachineScheduler)

static

GCNMinRegSchedRegistry

MachineSchedRegistry GCNMinRegSchedRegistry("gcn-minreg", "Run GCN iterative scheduler for minimal register usage (experimental)", createMinRegScheduler)

static

InternalizeSymbols

cl::opt<bool> InternalizeSymbols("amdgpu-internalize-symbols", cl::desc("Enable elimination of non-kernel functions and unused globals"), cl::init(false), cl::Hidden)

static

Referenced by llvm::AMDGPUTargetMachine::adjustPassManager(), and llvm::AMDGPUTargetMachine::registerPassBuilderCallbacks().

IterativeGCNMaxOccupancySchedRegistry

MachineSchedRegistry IterativeGCNMaxOccupancySchedRegistry("gcn-max-occupancy-experimental", "Run GCN scheduler to maximize occupancy (experimental)", createIterativeGCNMaxOccupancyMachineScheduler)

static

LateCFGStructurize

cl::opt<bool, true> LateCFGStructurize("amdgpu-late-structurize", cl::desc("Enable late CFG structurization"), cl::location(AMDGPUTargetMachine::EnableLateStructurizeCFG), cl::Hidden)

static

OptExecMaskPreRA

cl::opt<bool> OptExecMaskPreRA("amdgpu-opt-exec-mask-pre-ra", cl::Hidden, cl::desc("Run pre-RA exec mask optimizations"), cl::init(true))

static

OptVGPRLiveRange

cl::opt<bool> OptVGPRLiveRange("amdgpu-opt-vgpr-liverange", cl::desc("Enable VGPR liverange optimizations for if-else structure"), cl::init(true), cl::Hidden)

static

RegAllocOptNotSupportedMessage

const char RegAllocOptNotSupportedMessage[]

static

Initial value:

=
  "-regalloc not supported with amdgcn. Use -sgpr-regalloc and -vgpr-regalloc"

Definition at line 1304 of file AMDGPUTargetMachine.cpp.

ScalarizeGlobal

cl::opt<bool> ScalarizeGlobal("amdgpu-scalarize-global-loads", cl::desc("Enable global load scalarization"), cl::init(true), cl::Hidden)

static

Referenced by llvm::GCNTargetMachine::getSubtargetImpl().

SISchedRegistry

MachineSchedRegistry SISchedRegistry("si", "Run SI's custom scheduler", createSIMachineScheduler)

static