Jim Lin 18f5ada244 [DAGCombiner] Don't reduce BUILD_VECTOR to BITCAST before LegalizeTypes if VT is legal. ...

Targets may lose some optimization opportunities for certain vector operation
if we reduce BUILD_VECTOR to BITCAST early.

And if VT is not legal, reduce BUILD_VECTOR to BITCAST before LegailizeTypes
can get benefit. Because type-legalizer often scalarizes illegal type of vectors.

Reviewed By: sebastian-ne

Differential Revision: https://reviews.llvm.org/D156645

2023-08-19 12:53:50 +08:00

..

CMakeLists.txt

Split out CodeGenTypes from CodeGen for LLT/MVT

2023-05-03 00:13:20 +09:00

DAGCombiner.cpp

[DAGCombiner] Don't reduce BUILD_VECTOR to BITCAST before LegalizeTypes if VT is legal.

2023-08-19 12:53:50 +08:00

FastISel.cpp

[XRay][AArch64] Suppport __xray_customevent/__xray_typedevent

2023-06-23 09:24:18 -07:00

FunctionLoweringInfo.cpp

[1/11][IR] Permit load/store/alloca for struct of the same scalable vector type

2023-05-19 09:39:36 -07:00

InstrEmitter.cpp

[SDAG] Preserve unpredictable metadata, teach X86CmovConversion to respect this metadata

2023-06-01 20:56:44 +02:00

InstrEmitter.h

[DebugInfo] Store instr-ref mode of MachineFunction in member

2023-01-20 14:47:11 +00:00

LegalizeDAG.cpp

[LegalizeDAG] Prevent LegalizeLoadOps from creating extloads that mix int and fp types.

2023-07-12 08:03:35 -07:00

LegalizeFloatTypes.cpp

[LegalizeTypes] Use report_fatal_error instead of llvm_unreachable in the default case of some type legalization handlers.

2023-07-22 11:05:24 -07:00

LegalizeIntegerTypes.cpp

[LegalizeTypes][NFC] Combine ExpandIntOp_{S,U}INT_TO_FP to ExpandIntOp_XINT_TO_FP

2023-08-14 10:38:29 +01:00

LegalizeTypes.cpp

[DAG] Use legal shift amount type in DAGTypeLegalizer::JoinIntegers

2023-07-12 08:12:09 +01:00

LegalizeTypes.h

[LegalizeTypes][NFC] Combine ExpandIntOp_{S,U}INT_TO_FP to ExpandIntOp_XINT_TO_FP

2023-08-14 10:38:29 +01:00

LegalizeTypesGeneric.cpp

[LegalizeType][X86] Support WidenVecRes_AssertZext and SplitVecRes_AssertZext for ISD::AssertZext during LegalizeType procedure

2023-05-24 10:17:08 +08:00

LegalizeVectorOps.cpp

IR: Add llvm.frexp intrinsic

2023-06-28 14:50:16 -04:00

LegalizeVectorTypes.cpp

[LegalizeTypes] Factor in vscale_range when widening insert_subvector

2023-08-14 09:58:15 +01:00

ResourcePriorityQueue.cpp

Remove unneeded cl::ZeroOrMore for cl::opt/cl::list options

2022-06-05 00:31:44 -07:00

ScheduleDAGFast.cpp

Fix uninitialized scalar members in CodeGen

2023-04-21 12:22:34 +08:00

ScheduleDAGRRList.cpp

Restore CodeGen/MachineValueType.h from Support

2023-05-03 00:13:20 +09:00

ScheduleDAGSDNodes.cpp

Clean come dead code

2023-04-23 20:46:34 -04:00

ScheduleDAGSDNodes.h

Restore CodeGen/MachineValueType.h from Support

2023-05-03 00:13:20 +09:00

ScheduleDAGVLIW.cpp

…

SDNodeDbgValue.h

[DebugInfo] Remove dbg.addr from CodeGen

2023-03-02 09:29:43 +00:00

SelectionDAG.cpp

[DAG] NFC: Add getBitcastedExtOrTrunc

2023-08-17 14:29:17 -07:00

SelectionDAGAddressAnalysis.cpp

[X86] Fix aliasing check between TargetFrameIndex and FrameIndex

2023-08-17 10:48:08 +01:00

SelectionDAGBuilder.cpp

[SelectionDAG] Use TypeSize variant of ComputeValueVTs to compute correct offsets for scalable aggregate types.

2023-08-16 11:56:31 +00:00

SelectionDAGBuilder.h

Restore CodeGen/MachineValueType.h from Support

2023-05-03 00:13:20 +09:00

SelectionDAGDumper.cpp

IR: Add llvm.frexp intrinsic

2023-06-28 14:50:16 -04:00

SelectionDAGISel.cpp

[SelectionDAG] Add Opc_CheckPatternPredicate2 to support targets with more than 256 predicates.

2023-08-05 16:59:45 -07:00

SelectionDAGPrinter.cpp

…

SelectionDAGTargetInfo.cpp

…

StatepointLowering.cpp

Fix "this this" duplicate typo in comment. NFC.

2023-06-27 11:46:02 +01:00

StatepointLowering.h

…

TargetLowering.cpp

[DAG] SimplifyDemandedBits - Use DemandedBits intead of OriginalDemandedBits to when simplifying UMIN/UMAX to AND/OR.

2023-08-18 11:59:18 -07:00