Matt Arsenault 460ffcddd9 AMDGPU: Make bf16/v2bf16 legal types (#76215) ...

There are some intrinsics are using i16 vectors in place of bfloat
vectors.
Move towards making bf16 vectors legal so these can migrate. Leave the
larger vectors for a later change.

Depends #76213 #76214

2024-01-04 22:31:18 +07:00

..

CMakeLists.txt

…

DAGCombiner.cpp

[DAGCombiner][RISCV] Preserve disjoint flag in folding (shl (or x, c1), c2) -> (or (shl x, c2), c1 << c2) (#76860)

2024-01-03 13:14:13 -08:00

FastISel.cpp

[IR] Fix GEP offset computations for vector GEPs (#75448)

2024-01-04 10:08:21 +01:00

FunctionLoweringInfo.cpp

[CodeGen] Use range-based for loops (NFC)

2023-12-24 22:45:50 -08:00

InstrEmitter.cpp

[llvm][NFC] Use SDValue::getConstantOperandVal(i) where possible (#76708)

2024-01-02 13:14:28 +00:00

InstrEmitter.h

…

LegalizeDAG.cpp

AMDGPU: Make bf16/v2bf16 legal types (#76215)

2024-01-04 22:31:18 +07:00

LegalizeFloatTypes.cpp

DAG: Implement promotion for strict_fpextend (#74310)

2023-12-22 17:15:52 +07:00

LegalizeIntegerTypes.cpp

[RISCV] Add codegen support for experimental.vp.splice (#74688)

2023-12-21 08:38:32 -08:00

LegalizeTypes.cpp

[SelectionDAG] Adjust placement of blank lines in debug output. NFC (#70969)

2023-11-02 09:35:04 -07:00

LegalizeTypes.h

DAG: Implement promotion for strict_fpextend (#74310)

2023-12-22 17:15:52 +07:00

LegalizeTypesGeneric.cpp

[llvm][TypeSize] Fix addition/subtraction in TypeSize. (#72979)

2023-11-22 08:52:53 +00:00

LegalizeVectorOps.cpp

[RISCV] Support Strict FP arithmetic Op when only have Zvfhmin (#68867)

2023-11-09 09:55:48 +08:00

LegalizeVectorTypes.cpp

[SelectionDAG] Fix assertion for widening of mask operand in MSTORE. (#73295)

2023-11-29 12:56:30 +01:00

ResourcePriorityQueue.cpp

…

ScheduleDAGFast.cpp

[llvm][NFC] Use SDValue::getConstantOperandVal(i) where possible (#76708)

2024-01-02 13:14:28 +00:00

ScheduleDAGRRList.cpp

[llvm][NFC] Use SDValue::getConstantOperandVal(i) where possible (#76708)

2024-01-02 13:14:28 +00:00

ScheduleDAGSDNodes.cpp

Fix MSVC signed/unsigned mismatch warning. NFC.

2023-12-01 18:34:01 +00:00

ScheduleDAGSDNodes.h

…

ScheduleDAGVLIW.cpp

[NFC][CodeGen] Change CodeGenOpt::Level/CodeGenFileType into enum classes (#66295)

2023-09-14 14:10:14 -07:00

SDNodeDbgValue.h

…

SelectionDAG.cpp

[SelectionDAG][X86] Use disjoint flag in SelectionDAG::isADDLike. (#76847)

2024-01-03 11:54:29 -08:00

SelectionDAGAddressAnalysis.cpp

[DAG] BaseIndexOffset::equalBaseIndex - early out on failed matches. NFCI.

2024-01-04 10:50:43 +00:00

SelectionDAGBuilder.cpp

[IR] Fix GEP offset computations for vector GEPs (#75448)

2024-01-04 10:08:21 +01:00

SelectionDAGBuilder.h

[DebugInfo][RemoveDIs] Handle dbg.declares in SelectionDAGISel (#73496)

2023-12-12 11:32:19 +00:00

SelectionDAGDumper.cpp

[SelectionDAG] Add initial plumbing for the disjoint flag. (#76751)

2024-01-02 21:58:00 -08:00

SelectionDAGISel.cpp

[llvm][NFC] Use SDValue::getConstantOperandVal(i) where possible (#76708)

2024-01-02 13:14:28 +00:00

SelectionDAGPrinter.cpp

…

SelectionDAGTargetInfo.cpp

…

StatepointLowering.cpp

[StatepointLowering] Take return attributes of gc.result into account (#68439)

2023-10-14 18:38:18 +02:00

StatepointLowering.h

…

TargetLowering.cpp

[DAG] SimplifyDemandedBits - call demanded elts variant directly for SELECT/SELECT_CC nodes.

2024-01-04 10:53:45 +00:00