caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
4ece50737d5385fb80cfa23f5297d1111f8eed39
clang-p2996/llvm/lib/Target/AMDGPU
History
Jeffrey Byrnes 1f08d3bc3a [AMDGPU] Further reduce attaching of implicit operands to spills 
Extension of https://reviews.llvm.org/D141101 to even further reduce the amount of implicit operands we attach. The main benefit is to improve cability of post-ra scheduler, and reduce unneeded dependency resolution (e.g. inserting snops).

Unfortunately, we run into regressions if we completely minimize the amount implicit operands (naively), we run into some regressions (e.g. dual_movs are replaced with multiple calls to v_mov). This is even more reason to switch to LiveRegUnits.

Nonetheless, this patch removes the operands which we can for free (more or less).

Change-Id: Ib4f409202b36bdbc59eed615bc2d19fa8bd8c057

Differential Revision: https://reviews.llvm.org/D141557

Change-Id: I8b039e3c0d39436b384083f8beb947ee1b1730b2
2023-01-19 14:31:07 -08:00
..
AsmParser
[AMDGPU][AsmParser][NFC] Refine defining single-bit custom operands.
2023-01-16 16:11:59 +00:00
Disassembler
[MC] llvm::Optional => std::optional
2022-12-04 21:36:08 +00:00
MCA
[Support] Move TargetParsers to new component
2022-12-20 11:05:50 +00:00
MCTargetDesc
[MC] Use MCRegister instead of unsigned in MCInstPrinter (NFC)
2023-01-17 22:39:39 +03:00
TargetInfo
…
Utils
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
AMDGPU.h
[llvm][PassSupport] don't require passes to be default constructible
2022-12-20 14:07:29 -08:00
AMDGPU.td
[AMDGPU] Add feature predicate for v_fmac_f64 instruction
2023-01-19 13:06:30 +01:00
AMDGPUAliasAnalysis.cpp
[AA] Allow for flow-sensitive analyses.
2022-12-15 21:04:38 -08:00
AMDGPUAliasAnalysis.h
[AA] Allow for flow-sensitive analyses.
2022-12-15 21:04:38 -08:00
AMDGPUAlwaysInlinePass.cpp
…
AMDGPUAnnotateKernelFeatures.cpp
…
AMDGPUAnnotateUniformValues.cpp
…
AMDGPUArgumentUsageInfo.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUArgumentUsageInfo.h
[amdgpu] Implement lds kernel id intrinsic
2022-07-19 17:46:19 +01:00
AMDGPUAsmPrinter.cpp
AMDGPU: Set scratch_en if there is dynamic stack but no fixed stack
2023-01-04 20:51:18 -05:00
AMDGPUAsmPrinter.h
[CodeGen] llvm::Optional => std::optional
2022-12-13 09:06:36 +00:00
AMDGPUAtomicOptimizer.cpp
AMDGPU: Create poison values instead of undef
2022-11-16 14:47:24 -08:00
AMDGPUAttributes.def
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
AMDGPUAttributor.cpp
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
AMDGPUCallingConv.td
AMDGPU/GlobalISel: Do not create readfirstlane with non-s32 type
2022-12-15 21:44:07 -05:00
AMDGPUCallLowering.cpp
Drop the ZeroBehavior parameter from countLeadingZeros and the like (NFC)
2023-01-18 19:58:44 -08:00
AMDGPUCallLowering.h
…
AMDGPUCodeGenPrepare.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUCombine.td
[GlobalISel] Add Predicates to GICombineRule
2022-10-26 07:13:40 +00:00
AMDGPUCombinerHelper.cpp
[CodeGen] llvm::Optional => std::optional
2022-12-13 09:06:36 +00:00
AMDGPUCombinerHelper.h
…
AMDGPUCtorDtorLowering.cpp
AMDGPU: Erase llvm.global_ctors/global_dtors after lowering
2022-12-09 14:25:32 -05:00
AMDGPUCtorDtorLowering.h
AMDGPU: Port AMDGPUCtorDtorLowering to new PM
2022-12-09 13:43:38 -05:00
AMDGPUExportClustering.cpp
…
AMDGPUExportClustering.h
…
AMDGPUFeatures.td
…
AMDGPUFrameLowering.cpp
[NFC][AMDGPU] Fix typo.
2022-08-20 08:30:42 +02:00
AMDGPUFrameLowering.h
…
AMDGPUGenRegisterBankInfo.def
[AMDGPU] Add support for new LLVM vector types
2022-11-29 17:02:04 +01:00
AMDGPUGISel.td
[AMDGPU] Handle nnan and fast on the call in fpmed3 patterns
2022-12-06 22:57:52 -08:00
AMDGPUGlobalISelUtils.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUGlobalISelUtils.h
[AMDGPU][NFC] Remove isLegalVOP3PShuffleMask
2022-11-22 14:31:59 +00:00
AMDGPUHSAMetadataStreamer.cpp
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
AMDGPUHSAMetadataStreamer.h
[AMDGPU] Add .uniform_work_group_size metadata to v5
2023-01-05 21:29:56 +00:00
AMDGPUIGroupLP.cpp
[AMDGPU] Treat WMMA the same as MFMA for sched_barrier
2023-01-19 10:52:31 -08:00
AMDGPUIGroupLP.h
[AMDGPU] Add iglp_opt builtin and MFMA GEMM Opt strategy
2022-08-19 15:38:36 -07:00
AMDGPUInsertDelayAlu.cpp
[AMDGPU] New AMDGPUInsertDelayAlu pass
2022-06-29 21:30:20 +01:00
AMDGPUInstCombineIntrinsic.cpp
[AMDGPU] Simplify simplifyAMDGCNMemoryIntrinsicDemanded. NFC.
2022-12-22 11:50:04 +00:00
AMDGPUInstrInfo.cpp
…
AMDGPUInstrInfo.h
…
AMDGPUInstrInfo.td
[AMDGPU] Add llvm.is.fpclass intrinsic to existing SelectionDAG fp
2022-11-28 16:00:36 -05:00
AMDGPUInstructions.td
[AMDGPU] Make use of !listremove. NFCI.
2022-12-12 17:01:04 +00:00
AMDGPUInstructionSelector.cpp
std::optional::value => operator*/operator->
2022-12-17 00:42:05 +00:00
AMDGPUInstructionSelector.h
[AMDGPU] Handle nnan and fast on the call in fpmed3 patterns
2022-12-06 22:57:52 -08:00
AMDGPUISelDAGToDAG.cpp
[llvm][PassSupport] don't require passes to be default constructible
2022-12-20 14:07:29 -08:00
AMDGPUISelDAGToDAG.h
[llvm][PassSupport] don't require passes to be default constructible
2022-12-20 14:07:29 -08:00
AMDGPUISelLowering.cpp
[AMDGPU] Cast sub-dword elements to i32 in concat_vectors
2023-01-09 15:35:49 -08:00
AMDGPUISelLowering.h
Enable roundeven.
2022-12-20 15:40:20 +00:00
AMDGPULateCodeGenPrepare.cpp
…
AMDGPULegalizerInfo.cpp
[llvm][ADT] Replace uses of makeMutableArrayRef with deduction guides
2023-01-16 14:49:37 -07:00
AMDGPULegalizerInfo.h
[AMDGPU] Always lower SHUFFLE_VECTOR
2022-10-04 14:23:17 +00:00
AMDGPULibCalls.cpp
try to fix build yet more after 16544cbe64
2022-09-28 15:40:52 -04:00
AMDGPULibFunc.cpp
[TableGen] Use MemoryEffects to represent intrinsic memory effects (NFCI)
2022-11-14 10:52:04 +01:00
AMDGPULibFunc.h
…
AMDGPULowerIntrinsics.cpp
…
AMDGPULowerKernelArguments.cpp
[NFC] Remove Function::getParamAlignment
2023-01-13 16:20:58 +00:00
AMDGPULowerKernelAttributes.cpp
[NFC][AMDGPULowerKernelAttributes] Factorize repeated code into function
2022-10-05 09:26:39 -05:00
AMDGPULowerModuleLDSPass.cpp
Revert D141134 "[NFC] Only expose getXXXSize functions in TypeSize"
2023-01-06 15:27:50 +00:00
AMDGPUMachineCFGStructurizer.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUMachineFunction.cpp
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
AMDGPUMachineFunction.h
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
AMDGPUMachineModuleInfo.cpp
…
AMDGPUMachineModuleInfo.h
std::optional::value => operator*/operator->
2022-12-17 00:42:05 +00:00
AMDGPUMacroFusion.cpp
…
AMDGPUMacroFusion.h
…
AMDGPUMCInstLower.cpp
[amdgpu] Reimplement LDS lowering
2022-12-07 22:02:54 +00:00
AMDGPUMCInstLower.h
…
AMDGPUMIRFormatter.cpp
AMDGPU: Remove ImagePSV and move images to addrspace 7
2022-11-30 11:32:34 +01:00
AMDGPUMIRFormatter.h
[llvm] Remove redundaunt virtual specifiers (NFC)
2022-07-24 21:50:35 -07:00
AMDGPUOpenCLEnqueuedBlockLowering.cpp
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
AMDGPUPerfHintAnalysis.cpp
[AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access
2022-07-19 15:16:28 +05:30
AMDGPUPerfHintAnalysis.h
[AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access
2022-07-19 15:16:28 +05:30
AMDGPUPostLegalizerCombiner.cpp
[GlobalISel] Add Predicates to GICombineRule
2022-10-26 07:13:40 +00:00
AMDGPUPreLegalizerCombiner.cpp
AMDGPU/GlobalISel: Fix combine crash because LI is not set in prelegalizer
2022-11-08 12:46:16 +01:00
AMDGPUPrintfRuntimeBinding.cpp
AMDGPU: Use getConstantStringInfo for printf format strings
2023-01-13 14:11:37 -05:00
AMDGPUPromoteAlloca.cpp
Deprecate MemIntrinsicBase::getDestAlignment() and MemTransferBase::getSourceAlignment()
2023-01-16 14:22:03 +00:00
AMDGPUPromoteKernelArguments.cpp
…
AMDGPUPropagateAttributes.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUPTNote.h
…
AMDGPURegBankCombiner.cpp
[CodeGen] llvm::Optional => std::optional
2022-12-13 09:06:36 +00:00
AMDGPURegisterBankInfo.cpp
MachineIRBuilder: Rename buildMerge. NFC
2023-01-13 09:32:58 +01:00
AMDGPURegisterBankInfo.h
…
AMDGPURegisterBanks.td
[AMDGPU] Add support for new LLVM vector types
2022-11-29 17:02:04 +01:00
AMDGPUReleaseVGPRs.cpp
[Target] Use std::nullopt instead of None (NFC)
2022-12-02 20:36:06 -08:00
AMDGPUReplaceLDSUseWithPointer.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUResourceUsageAnalysis.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
AMDGPUResourceUsageAnalysis.h
…
AMDGPURewriteOutArguments.cpp
[NFC] Cleanup: Replace Function::getBasicBlockList().splice() with Function::splice()
2022-12-14 15:34:19 -08:00
AMDGPURewriteUndefForPHI.cpp
Add override for runOnFunction()
2022-09-26 10:19:35 +08:00
AMDGPUSearchableTables.td
[AMDGPU] Mark wmma intrinsics as source of divergence
2023-01-11 14:16:18 -08:00
AMDGPUSetWavePriority.cpp
[AMDGPU][SetWavePriority] Fix dealing with MBBInfo records.
2022-09-30 14:27:50 +01:00
AMDGPUSubtarget.cpp
Attributes: Add function getter to parse integer string attributes
2022-12-14 13:12:35 -05:00
AMDGPUSubtarget.h
…
AMDGPUTargetMachine.cpp
AMDGPU: Use DenormalMode type in FP mode tracking
2022-12-21 20:35:48 -05:00
AMDGPUTargetMachine.h
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
AMDGPUTargetObjectFile.cpp
…
AMDGPUTargetObjectFile.h
…
AMDGPUTargetTransformInfo.cpp
[SLP]Fix cost of the broadcast buildvector/gather.
2023-01-06 09:25:05 -08:00
AMDGPUTargetTransformInfo.h
[SLP]Fix cost of the broadcast buildvector/gather.
2023-01-06 09:25:05 -08:00
AMDGPUUnifyDivergentExitNodes.cpp
[AMDGPU] Unify divergent nodes if the PostDom tree has one root
2023-01-04 10:45:03 -07:00
AMDGPUUnifyMetadata.cpp
[Target] Qualify auto in range-based for loops (NFC)
2022-08-28 10:41:50 -07:00
AMDKernelCodeT.h
…
BUFInstructions.td
AMDGPU/SIInsertWait: Skip dummy tied source
2023-01-11 09:59:35 +08:00
CaymanInstructions.td
…
CMakeLists.txt
[Support] Move TargetParsers to new component
2022-12-20 11:05:50 +00:00
DSInstructions.td
AMDGPU/SIInsertWait: Skip dummy tied source
2023-01-11 09:59:35 +08:00
EvergreenInstructions.td
R600: Remove broken atomicrmw patterns
2022-11-04 16:14:04 -07:00
EXPInstructions.td
…
FLATInstructions.td
AMDGPU/SIInsertWait: Skip dummy tied source
2023-01-11 09:59:35 +08:00
GCNCreateVOPD.cpp
[AMDGPU][GFX11][NFC] Refactor VOPD operands handling (part 2)
2022-11-18 14:15:05 +03:00
GCNDPPCombine.cpp
[AMDGPU] Use MCInstrDesc::getSize. NFC.
2023-01-11 13:22:38 +00:00
GCNHazardRecognizer.cpp
[AMDGPU] Replace getPhysRegClass with getPhysRegBaseClass
2022-12-20 16:22:14 +09:00
GCNHazardRecognizer.h
[AMDGPU][GFX11] Mitigate VALU mask write hazard
2022-10-01 16:21:24 +09:00
GCNILPSched.cpp
[llvm] Qualify auto in range-based for loops (NFC)
2022-08-13 12:55:42 -07:00
GCNIterativeScheduler.cpp
[AMDGPU] Refactor debug printing routines for GCNRPTracker
2022-10-28 04:22:46 +02:00
GCNIterativeScheduler.h
…
GCNMinRegStrategy.cpp
[llvm] Qualify auto in range-based for loops (NFC)
2022-08-13 12:55:42 -07:00
GCNNSAReassign.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
GCNPreRAOptimizations.cpp
…
GCNProcessors.td
[AMDGPU] user-sgpr-init16-bug does not apply to gfx1103
2022-07-29 14:21:13 +01:00
GCNRegPressure.cpp
Revert "[AMDGPU] Speedup GCNDownwardRPTracker::advanceBeforeNext"
2022-12-02 14:18:16 +00:00
GCNRegPressure.h
[AMDGPU] Fix GCNDownwardRPTracker::advanceBeforeNext at the end of MBB
2022-11-03 11:52:56 +01:00
GCNSchedStrategy.cpp
[AMDGPU] Introduce separate register limit bias in scheduler
2023-01-19 10:51:40 -08:00
GCNSchedStrategy.h
[AMDGPU] Introduce separate register limit bias in scheduler
2023-01-19 10:51:40 -08:00
GCNSubtarget.h
[AMDGPU] Add feature predicate for v_fmac_f64 instruction
2023-01-19 13:06:30 +01:00
GCNVOPDUtils.cpp
[NFC] Hide implementation details in anonymous namespaces
2023-01-08 17:37:02 +01:00
GCNVOPDUtils.h
[AMDGPU] gfx11 Generate VOPD Instructions
2022-07-05 09:18:19 -04:00
InstCombineTables.td
…
LDSDIRInstructions.td
…
MIMGInstructions.td
[AMDGPU][NFC] Rename GFX10A16 operands.
2023-01-09 17:18:46 +00:00
R600.h
[llvm][PassSupport] don't require passes to be default constructible
2022-12-20 14:07:29 -08:00
R600.td
[AMDGPU] Fix useDeprecatedPositionallyEncodedOperands errors in R600.
2022-09-25 17:55:09 -04:00
R600AsmPrinter.cpp
[llvm] LLVM_FALLTHROUGH => [[fallthrough]]. NFC
2022-08-08 11:24:15 -07:00
R600AsmPrinter.h
…
R600ClauseMergePass.cpp
…
R600ControlFlowFinalizer.cpp
[llvm] LLVM_FALLTHROUGH => [[fallthrough]]. NFC
2022-08-08 11:24:15 -07:00
R600Defines.h
…
R600EmitClauseMarkers.cpp
…
R600ExpandSpecialInstrs.cpp
…
R600FrameLowering.cpp
…
R600FrameLowering.h
…
R600InstrFormats.td
…
R600InstrInfo.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
R600InstrInfo.h
…
R600InstrInfo.td
…
R600Instructions.td
[AMDGPU] Fix useDeprecatedPositionallyEncodedOperands errors in R600.
2022-09-25 17:55:09 -04:00
R600ISelDAGToDAG.cpp
[NFC] Hide implementation details in anonymous namespaces
2023-01-08 17:37:02 +01:00
R600ISelLowering.cpp
Move from llvm::makeArrayRef to ArrayRef deduction guides - llvm/ part
2023-01-05 14:11:08 +01:00
R600ISelLowering.h
AMDGPU/R600: Special case addrspacecast lowering for null
2022-12-18 08:02:45 -05:00
R600MachineCFGStructurizer.cpp
…
R600MachineFunctionInfo.cpp
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
R600MachineFunctionInfo.h
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
R600MachineScheduler.cpp
…
R600MachineScheduler.h
…
R600MCInstLower.cpp
[NFC] Hide implementation details in anonymous namespaces
2023-01-08 17:37:02 +01:00
R600OpenCLImageTypeLoweringPass.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
R600OptimizeVectorRegisters.cpp
…
R600Packetizer.cpp
…
R600Processors.td
…
R600RegisterInfo.cpp
PEI should be able to use backward walk in replaceFrameIndicesBackward.
2022-11-18 15:57:34 +01:00
R600RegisterInfo.h
PEI should be able to use backward walk in replaceFrameIndicesBackward.
2022-11-18 15:57:34 +01:00
R600RegisterInfo.td
…
R600Schedule.td
…
R600Subtarget.cpp
…
R600Subtarget.h
…
R600TargetMachine.cpp
[NFC] Hide implementation details in anonymous namespaces
2023-01-08 17:37:02 +01:00
R600TargetMachine.h
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
R600TargetTransformInfo.cpp
[SLP]Fix cost of the broadcast buildvector/gather.
2023-01-06 09:25:05 -08:00
R600TargetTransformInfo.h
[SLP]Fix cost of the broadcast buildvector/gather.
2023-01-06 09:25:05 -08:00
R700Instructions.td
…
SIAnnotateControlFlow.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
SIDefines.h
Partially reapply "AMDGPU: Invert handling of enqueued block detection"
2023-01-12 15:02:16 -05:00
SIFixSGPRCopies.cpp
[AMDGPU] Replace getPhysRegClass with getPhysRegBaseClass
2022-12-20 16:22:14 +09:00
SIFixVGPRCopies.cpp
…
SIFoldOperands.cpp
AMDGPU: Use DenormalMode type in FP mode tracking
2022-12-21 20:35:48 -05:00
SIFormMemoryClauses.cpp
Attributes: Add function getter to parse integer string attributes
2022-12-14 13:12:35 -05:00
SIFrameLowering.cpp
Revert "[AMDGPU][SILowerSGPRSpills] Spill SGPRs to virtual VGPRs"
2022-12-21 16:17:42 +05:30
SIFrameLowering.h
Revert "[AMDGPU][SILowerSGPRSpills] Spill SGPRs to virtual VGPRs"
2022-12-21 16:17:42 +05:30
SIInsertHardClauses.cpp
…
SIInsertWaitcnts.cpp
AMDGPU/SIInsertWait: Skip dummy tied source
2023-01-11 09:59:35 +08:00
SIInstrFormats.td
AMDGPU/SIInsertWait: Skip dummy tied source
2023-01-11 09:59:35 +08:00
SIInstrInfo.cpp
[AMDGPU] Further reduce attaching of implicit operands to spills
2023-01-19 14:31:07 -08:00
SIInstrInfo.h
[AMDGPU] Treat WMMA the same as MFMA for sched_barrier
2023-01-19 10:52:31 -08:00
SIInstrInfo.td
[AMDGPU][AsmParser][NFC] Refine defining single-bit custom operands.
2023-01-16 16:11:59 +00:00
SIInstructions.td
[AMDGPU] Add feature predicate for v_fmac_f64 instruction
2023-01-19 13:06:30 +01:00
SIISelLowering.cpp
Drop the ZeroBehavior parameter from countLeadingZeros and the like (NFC)
2023-01-18 19:58:44 -08:00
SIISelLowering.h
[AMDGPU] Allow finer grain control of an unaligned access speed
2022-11-17 09:23:53 -08:00
SILateBranchLowering.cpp
[Target] Qualify auto in range-based for loops (NFC)
2022-08-28 10:41:50 -07:00
SILoadStoreOptimizer.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
SILowerControlFlow.cpp
[AMDGPU] SILowerControlFlow uses LiveIntervals
2022-07-12 16:53:53 +01:00
SILowerI1Copies.cpp
AMDGPU: Factor out hasDivergentBranch(). NFC
2022-09-14 13:27:21 +08:00
SILowerSGPRSpills.cpp
Revert "[AMDGPU][SILowerSGPRSpills] Spill SGPRs to virtual VGPRs"
2022-12-21 16:17:42 +05:30
SIMachineFunctionInfo.cpp
CodeGen: Don't lazily construct MachineFunctionInfo
2022-12-21 10:49:32 -05:00
SIMachineFunctionInfo.h
Move from llvm::makeArrayRef to ArrayRef deduction guides - llvm/ part
2023-01-05 14:11:08 +01:00
SIMachineScheduler.cpp
[CodeGen][Target] Remove uses of Register::isPhysicalRegister/isVirtualRegister. NFC
2023-01-13 23:12:48 -08:00
SIMachineScheduler.h
…
SIMemoryLegalizer.cpp
std::optional::value => operator*/operator->
2022-12-17 00:42:05 +00:00
SIModeRegister.cpp
[AMDGPU][GFX11] Use VGPR_32_Lo128 for VOP1,2,C
2022-09-20 09:56:28 -04:00
SIOptimizeExecMasking.cpp
[NFC][AMDGPU] Some cleanups in the SIOptimizeExecMasking pass.
2022-08-23 18:16:47 +02:00
SIOptimizeExecMaskingPreRA.cpp
AMDGPU: Fix assert when trying to overextend liverange
2022-11-04 15:14:43 -07:00
SIOptimizeVGPRLiveRange.cpp
[AMDGPU] Fix duplicated words in comments
2022-11-03 15:33:30 +00:00
SIPeepholeSDWA.cpp
[AMDGPU] Stop using make_pair and make_tuple. NFC.
2022-12-14 13:22:26 +00:00
SIPostRABundler.cpp
[AMDGPU] Use llvm::any_of (NFC)
2022-10-16 09:19:09 -07:00
SIPreAllocateWWMRegs.cpp
[AMDGPU] Add pseudo wavemode to optimize strict_wqm
2022-10-28 09:45:17 +09:00
SIPreEmitPeephole.cpp
[Target] Qualify auto in range-based for loops (NFC)
2022-08-28 10:41:50 -07:00
SIProgramInfo.cpp
…
SIProgramInfo.h
[AMDGPU] Add remarks to output some resource usage
2022-07-15 11:01:53 -07:00
SIRegisterInfo.cpp
[AMDGPU] Further reduce attaching of implicit operands to spills
2023-01-19 14:31:07 -08:00
SIRegisterInfo.h
[AMDGPU] Replace getPhysRegClass with getPhysRegBaseClass
2022-12-20 16:22:14 +09:00
SIRegisterInfo.td
[AMDGPU] Replace getPhysRegClass with getPhysRegBaseClass
2022-12-20 16:22:14 +09:00
SISchedule.td
…
SIShrinkInstructions.cpp
[AMDGPU] Add support for new LLVM vector types
2022-11-29 17:02:04 +01:00
SIWholeQuadMode.cpp
[AMDGPU] Replace getPhysRegClass with getPhysRegBaseClass
2022-12-20 16:22:14 +09:00
SMInstructions.td
[AMDGPU][CodeGen] Support (soffset + offset) s_buffer_load's.
2022-09-05 12:53:05 +01:00
SOPInstructions.td
[AMDGPU] S_MULK_I32 does not define SCC. NFCI.
2023-01-09 15:44:56 +00:00
VIInstrFormats.td
…
VINTERPInstructions.td
[AMDGPU][MC][GFX11][NFC] Correct VINTERP src operands
2022-11-07 15:52:55 +03:00
VOP1Instructions.td
[AMDGPU] Combine redundant Asm64 and AsmVOP3DPPBase. NFC
2023-01-06 14:09:42 -05:00
VOP2Instructions.td
[AMDGPU] Add feature predicate for v_fmac_f64 instruction
2023-01-19 13:06:30 +01:00
VOP3Instructions.td
[AMDGPU] Remove permlane discard vdst_in optimization from isel
2022-12-22 15:49:26 +00:00
VOP3PInstructions.td
[AMDGPU] Combine redundant Asm64 and AsmVOP3DPPBase. NFC
2023-01-06 14:09:42 -05:00
VOPCInstructions.td
[AMDGPU] Combine redundant Asm64 and AsmVOP3DPPBase. NFC
2023-01-06 14:09:42 -05:00
VOPDInstructions.td
[AMDGPU] Fix delay alu for VOPD with src2acc
2022-10-25 13:11:17 -04:00
VOPInstructions.td
[AMDGPU] Add llvm.is.fpclass intrinsic to existing SelectionDAG fp
2022-11-28 16:00:36 -05:00