llvm-project/Conversion at fa6cc7eadedd61347456f088ef40f6dfdcd751e2 - llvm-project - shylie's gitea

shylie/llvm-project

History

Matthias Springer 6c640b86e6

[mlir][LLVM] Fix unsupported FP lowering in VectorConvertToLLVMPattern (#166513 )

Fixes a bug in `VectorConvertToLLVMPattern`, which converted operations
with unsupported FP types. E.g., `arith.addf ... : f4E2M1FN` was lowered
to `llvm.fadd ... : i4`, which does not verify. There are a few more
patterns that have the same bug. Those will be fixed in follow-up PRs.

This commit is in preparation of adding an `APFloat`-based lowering for
`arith` operations with unsupported floating-point types.

2025-11-05 12:24:59 +00:00

..

AffineToStandard

[mlir][Affine] Add nsw to lowering of AffineMulExpr. (#121535 )

2025-01-06 14:57:24 -08:00

[mlir][amdgpu][rocdl] Allow for graceful wmma conversion failures (#165616 )

2025-10-29 16:09:59 -04:00

[mlir][ArithToAMDGPU] limit scaling truncf/extf support to gfx950 (#155431 )

2025-09-17 14:26:52 -05:00

…

[mlir][emitc] Isolate expressions from above (#155641 )

2025-09-01 15:10:17 +03:00

[mlir][LLVM] Fix unsupported FP lowering in VectorConvertToLLVMPattern (#166513 )

2025-11-05 12:24:59 +00:00

[mlir][spirv] Add pattern matching for arith.index_cast i1 to index for ArithToSPIRV (#155729 )

2025-09-03 14:04:35 -04:00

[AArch64][SME] Improve codegen for aarch64.sme.cnts* when not in streaming mode (#154761 )

2025-09-12 10:23:57 +01:00

[mlir][sme] Use signed comparison in ArmSMEToSCF (#156834 )

2025-09-05 08:58:52 +01:00

…

BufferizationToMemRef

…

…

[mlir][Transforms] Dialect Conversion Driver without Rollback (#151865 )

2025-08-13 17:40:55 +02:00

ComplexToROCDLLibraryCalls

[MLIR] Add new complex.powi op (#158722 )

2025-09-19 01:36:13 +00:00

…

ComplexToStandard

[mlir][complex] Fix exp accuracy (#164952 )

2025-10-28 15:01:15 +01:00

ControlFlowToLLVM

[mlir][Transforms] Dialect Conversion Driver without Rollback (#151865 )

2025-08-13 17:40:55 +02:00

ControlFlowToSCF

…

ControlFlowToSPIRV

…

[mlir][bufferize] Make buffer-results-to-out-params support only functions that are neither public nor extern (#162441 )

2025-10-08 16:54:02 +08:00

[MLIR][Affine] Extend getVectorReductionOp to support xor/maxnumf/minnumf (#163310 )

2025-11-04 12:52:51 +08:00

[mlir][EmitC] Add pass that combines all available emitc conversions (#117549 )

2025-05-01 07:24:01 -07:00

[mlir][ToLLVM] Fix the index bitwidth handling for the dynamic case of convert-to-llvm (#156557 )

2025-09-03 08:07:18 -04:00

[mlir][spirv] Add 8-bit float type emulation (#148811 )

2025-07-30 17:39:49 -05:00

[mlir][Vector] Move vector.mask canonicalization to folder (#140324 )

2025-05-21 17:25:01 -07:00

[MLIR][GPU] Generalize gpu.printf op lowering to LLVM call pattern. (#164297 )

2025-10-23 08:32:53 -07:00

[MLIR] Add sincos op to math dialect (#160772 )

2025-09-30 07:36:13 -07:00

[mlir][GPU] Generalize gpu.printf to not need gpu.module (#161266 )

2025-09-29 18:21:24 -07:00

[mlir][gpu][spirv] Add conversion for gpu.subgroup_mma_elementwise mulf (#158832 )

2025-09-16 16:01:48 +01:00

[mlir][Transforms] Dialect Conversion Driver without Rollback (#151865 )

2025-08-13 17:40:55 +02:00

…

…

[mlir][EmitC] Add MathToEmitC pass for math function lowering to EmitC (#113799 )

2025-01-20 09:26:41 +01:00

…

[mlir][Vector] add vector.insert canonicalization pattern to convert a chain of insertions to vector.from_elements (#142944 )

2025-08-19 13:43:31 +01:00

[MLIR] Add sincos op to math dialect (#160772 )

2025-09-30 07:36:13 -07:00

[MLIR][ROCDL] Add math.clampf -> rocdl.fmed3 conversion (#163520 )

2025-10-17 08:14:58 -04:00

[MLIR][SPIRV] Add spirv.IsFinite and lower math.{isfinite,isinf,isnan} to spirv. (#151552 )

2025-07-31 13:54:14 -04:00

[mlir] Use dyn_cast instead of cast in MathToVCIX conversion (#134047 )

2025-04-02 18:30:42 +08:00

[MLIR][MathToXeVM] Remove requirement for ModuleOp op type for MathToXeVM (#163619 )

2025-10-17 12:54:30 -05:00

[mlir][EmitC] Expand the MemRefToEmitC pass - Lowering CopyOp (#151206 )

2025-08-14 05:25:55 -07:00

[mlir][memref] Introduce memref.distinct_objects op (#156913 )

2025-10-01 15:01:37 +03:00

[mlir][memref-to-spirv]: Remap Image Load Coordinates (#160495 )

2025-09-30 09:44:06 +01:00

[mlir][mpi] fixing in-place and 0d mpi.all_reduce (#134225 )

2025-04-03 15:53:40 +02:00

[MLIR][NVVM] Update mbarrier Ops to use AnyTypeOf[] (2/n) (#165993 )

2025-11-05 15:38:24 +05:30

[MLIR][NVVM] Update mbarrier Ops to use AnyTypeOf[] (2/n) (#165993 )

2025-11-05 15:38:24 +05:30

[flang][acc] Update assembly formats to include asyncOnly, async, and wait (#140122 )

2025-05-15 12:56:15 -07:00

[Flang][OpenMP][Dialect] Swap to using MLIR dialect enum to encode map flags (#164043 )

2025-10-21 21:54:25 +02:00

…

[mlir][ptr] Extend ptr_add operation to support shaped operands (#156374 )

2025-09-03 15:50:21 +00:00

ReconcileUnrealizedCasts

[mlir][Transforms] Fix crash in reconcile-unrealized-casts (#158298 )

2025-09-12 14:28:02 +00:00

SCFToControlFlow

[mlir][SCF] scf.for: Add support for unsigned integer comparison (#153379 )

2025-08-15 10:59:14 +02:00

[mlir][emitc] Add emitc.do op to the dialect (#143008 )

2025-10-08 16:02:53 +03:00

[mlir][gpu] Loose the condition to convert scf.parallel to gpu.launch (#164978 )

2025-10-30 13:43:49 +00:00

[flang][OpenMP] Enable tiling (#143715 )

2025-09-10 09:25:40 -04:00

[mlir][spirv] Rework type capability queries (#160113 )

2025-09-22 15:28:41 +00:00

ShapeToStandard

…

[NFC][mlir][mesh,shard] Fixing misnomers in mesh dialect, renaming 'mesh' dialect to 'shard' (#150177 )

2025-07-25 16:53:08 +02:00

[mlir][LLVM] Switch undef for poison for uninitialized values (#125629 )

2025-02-06 12:49:30 -06:00

[mlir][tensor] Fix createFillOrGenerateOp (#121205 )

2025-04-15 09:29:44 +08:00

…

[mlir][tosa] Avoid requirement to specify enum name for enum attributes (#155695 )

2025-08-29 16:32:49 +01:00

[mlir][tosa] Support RescaleOp with dynamic extension in TosaToLinalg (#155967 )

2025-10-14 10:10:18 +08:00

TosaToMLProgram

[tosa] Change VariableOp to align with spec (#142240 )

2025-06-03 17:41:33 +01:00

[MLIR][TOSA] Update IfOp print/parse to support ranked condition tens… (#149791 )

2025-07-28 10:44:21 +01:00

[mlir][tosa] Fix indexing in TosaToTensor (#140906 )

2025-05-22 17:08:05 +03:00

[mlir][Transforms] Dialect Conversion Driver without Rollback (#151865 )

2025-08-13 17:40:55 +02:00

[mlir][spirv] Fix some issues related to converting ub.poison to SPIR-V (#125905 )

2025-02-06 19:07:34 +01:00

[mlir][amx] Direct AMX data transfers (#154114 )

2025-08-27 08:54:54 +02:00

[MLIR][Vector] Remove vector.splat (#162167 )

2025-10-10 09:58:18 -07:00

[mlir][gpu][vector] Lower Vector dialect to GPU for element-wise ops only (#159091 )

2025-09-17 09:10:52 +01:00

[MLIR][Conversion] Vector to LLVM: Remove unneeded vector shuffle (#162946 )

2025-10-15 13:21:26 -07:00

[mlir][vector] Folder: shape_cast(extract) -> extract (#146368 )

2025-07-21 11:12:50 -07:00

[mlir][spirv][vector] Use adaptor.getElements() in FromElements lowering. (#156972 )

2025-09-04 16:06:33 -07:00

[MLIR][XeGPU][VectorToXeGPU] Lower vector.load/store/transfer_read/transfer_write to new offsets syntax (#162095 )

2025-11-04 13:52:23 +01:00

[MLIR][Conversion] XeGPU to XeVM: Lower ranked dynamic base memory for create_nd_tdesc. (#164283 )

2025-10-29 09:55:50 -07:00

[MLIR][XeVM] XeVM to LLVM: Add conversion patterns for id ops (#162536 )

2025-10-10 09:20:44 -07:00