March 2021 Archives by author

• Messages sorted by: [ thread ] [ subject ] [ date ]
• More info on this list...

Starting: Mon Mar 1 05:21:22 PST 2021
Ending: Wed Mar 31 14:21:58 PDT 2021
Messages: 307

• [Mlir-commits] [mlir] 86c8a78 - Add C bindings for mlir::ExecutionEngine   Mehdi Amini
• [Mlir-commits] [mlir] 13cb431 - Add basic JIT Python Bindings   Mehdi Amini
• [Mlir-commits] [mlir] a7cac0d - Fix Dialect doc generation to special case for the Builtin dialect empty name   Mehdi Amini
• [Mlir-commits] [mlir] f8fe6d9 - Use gen-dialect-doc instead of gen-op-doc for the Builtin dialect   Mehdi Amini
• [Mlir-commits] [mlir] e94e557 - Forward the `LLVM_ENABLE_LIBCXX` CMake parameter to the mlir standalone test   Mehdi Amini
• [Mlir-commits] [mlir] 038f2a3 - Move LLVM::FMFAttr definition to TableGen (NFC)   Mehdi Amini
• [Mlir-commits] [mlir] 8205c1a - Rework LLVM Dialect LoopOptions attribute   Mehdi Amini
• [Mlir-commits] [mlir] 79f736c - Switch generatedTypeParser/generatedAttributeParser to return an OptionalParseResult   Mehdi Amini
• [Mlir-commits] [mlir] fe81e8f - Add default LoopOptionsAttrBuilder constructor and method to check if empty() (NFC)   Mehdi Amini
• [Mlir-commits] [mlir] cd9a692 - Fix LLVM Dialect LoopOptionsAttr round-tripping: the keywords were missing in the output   Mehdi Amini
• [Mlir-commits] [mlir] 890afad - Fix Flang build after MLIR API changes around `generatedTypeParser`   Mehdi Amini
• [Mlir-commits] [mlir] 75f3f77 - Fix MLIR test post 890afad954d   Mehdi Amini
• [Mlir-commits] [mlir] e1364f1 - Replace use of OperationState with builder::create in GPU Kernel Outlining (NFC)   Mehdi Amini
• [Mlir-commits] [mlir] c69550c - Revert "[mlir] Add linalg.fill bufferization conversion"   Mehdi Amini
• [Mlir-commits] [mlir] 270a336 - [mlir] Fix Python bindings tests failure in Debug mode after D98474   Mehdi Amini
• [Mlir-commits] [mlir] cdb6eb7 - Update syntax for amx.tile_muli to use two Unit attr to mark the zext case   Mehdi Amini
• [Mlir-commits] [mlir] a0c776f - Add a mechanism for Dialects to customize printing/parsing operations when they are unregistered   Mehdi Amini
• [Mlir-commits] [mlir] d905c10 - Add a mechanism for Dialects to provide a fallback for OpInterface   Mehdi Amini
• [Mlir-commits] [mlir] 973ddb7 - Define a `NoTerminator` traits that allows operations with a single block region to not provide a terminator   Mehdi Amini
• [Mlir-commits] [mlir] fcdf142 - Remove unused function, fix warning (NFC)   Mehdi Amini
• [Mlir-commits] [mlir] 86a2fa4 - Rename the 'concept' variable in SymbolInterfaces.td   Mehdi Amini
• [Mlir-commits] [mlir] 7a4d630 - Add a "register_runtime" method to the mlir.execution_engine and show calling back from MLIR into Python   Mehdi Amini
• [Mlir-commits] [mlir] a360a97 - Fix deletion of operations through the rewriter in a pattern matching a consumer operation   Mehdi Amini
• [Mlir-commits] [mlir] 2837991 - [mlir][linalg] Add support for memref inputs/outputs for `linalg.tiled_loop`.   Alexander Belyaev
• [Mlir-commits] [mlir] 628f5c9 - [mlir] Add a roundtrip test for 'linalg.tiled_loop' on buffers.   Alexander Belyaev
• [Mlir-commits] [mlir] 7f2236c - [mlir][linalg] Add output tensor args folding for linalg.tiled_loop.   Alexander Belyaev
• [Mlir-commits] [mlir] 883912a - Revert "[mlir] Introduce CloneOp and adapt test cases in BufferDeallocation."   Alexander Belyaev
• [Mlir-commits] [mlir] 465b9a4 - Revert "Revert "[mlir] Introduce CloneOp and adapt test cases in BufferDeallocation.""   Alexander Belyaev
• [Mlir-commits] [mlir] 5b333d3 - [mlir][sparse] do not ignore ordering for "dense" tensor linked with sparse type   Aart Bik
• [Mlir-commits] [mlir] 553cb6d - [mlir][sparse] fix bug in reduction chain   Aart Bik
• [Mlir-commits] [mlir] adc35b6 - [mlir][sparse] mask reduction update   Aart Bik
• [Mlir-commits] [mlir] e5c8fc7 - [mlir][vector] canonicalize unmasked gather/scatter/compress/expand directly into l/s   Aart Bik
• [Mlir-commits] [mlir] e7ee4ea - [mlir][sparse] disable nonunit stride dense vectorization   Aart Bik
• [Mlir-commits] [mlir] 6ad7b97 - [mlir][amx] Add Intel AMX dialect (architectural-specific vector dialect)   Aart Bik
• [Mlir-commits] [mlir] b388bbd - [mlir][amx] blocked tilezero integration test   Aart Bik
• [Mlir-commits] [mlir] b85d3e2 - [mlir][amx] reformatted examples   Aart Bik
• [Mlir-commits] [mlir] 9705caf - [mlir][amx] regression test for tile-muli (all zero/sign-extension combinations)   Aart Bik
• [Mlir-commits] [mlir] f2557cf - [mlir][cpu-runner] register all llvm ir dialects   Aart Bik
• [Mlir-commits] [mlir] 4fda0dc - Fix use of deprecated API   David Blaikie
• [Mlir-commits] [mlir] 8e3075c - [MLIR] Fix lowering of Affine IfOp in the presence of yield values.   Uday Bondhugula
• [Mlir-commits] [mlir] 0b20413 - Revert "[Canonicalizer] Process regions top-down instead of bottom up & reuse existing constants."   Uday Bondhugula
• [Mlir-commits] [mlir] 849f818 - [mlir] Fix ConstantOp verifier   Marius Brehler
• [Mlir-commits] [mlir] b635492 - [mlir][Affine][NFC] Return BlockArgument in AffineForOp::getInductionVar   Diego Caballero
• [Mlir-commits] [mlir] 71a8624 - [mlir] Extend Operation visitor with pre-order traversal   Diego Caballero
• [Mlir-commits] [mlir] 2de6dbd - [mlir] Add 'Skip' result to Operation visitor   Diego Caballero
• [Mlir-commits] [mlir] 95db7b4 - [mlir][Vector][Affine] Improve affine vectorizer algorithm   Diego Caballero
• [Mlir-commits] [mlir] 77a9d15 - [mlir][Affine][Vector] Add initial support for 'iter_args' to Affine vectorizer.   Diego Caballero
• [Mlir-commits] [mlir] ed193bc - [mlir][Vector][Affine] Fix heap-use-after-free in vectorizer   Diego Caballero
• [Mlir-commits] [mlir] 96891f0 - Reland: [mlir][Vector][Affine] Improve affine vectorizer algorithm   Diego Caballero
• [Mlir-commits] [mlir] 0fd0fb5 - Reland: [mlir][Affine][Vector] Add initial support for 'iter_args' to Affine vectorizer.   Diego Caballero
• [Mlir-commits] [mlir] de155f4 - [MLIR][OpenMP] Pretty printer and parser for omp.wsloop   Kiran Chandramohan
• [Mlir-commits] [mlir] bcc9b37 - Split `ElementwiseMappable` trait into four more precise traits.   Frederik Gossen
• [Mlir-commits] [mlir] 3b9667a - Clarify documentation for `Elementwise`, `Scalarizable`, `Vectorizable`, and   Frederik Gossen
• [Mlir-commits] [mlir] b975e3b - [MLIR] Add canoncalization for `shape.is_broadcastable`   Frederik Gossen
• [Mlir-commits] [mlir] 2a71f95 - [MLIR] Allow compatible shapes in `Elementwise` operations   Frederik Gossen
• [Mlir-commits] [mlir] b55f424 - [MLIR] Add canonicalization for `shape.broadcast`   Frederik Gossen
• [Mlir-commits] [mlir] 1ce70c1 - [MLIR] Canonicalize broadcast operations on single shapes   Frederik Gossen
• [Mlir-commits] [mlir] d78374b - [MLIR] Add callback builder for `shape.assuming` op   Frederik Gossen
• [Mlir-commits] [mlir] f368b3a - [MLIR][Shape] Canonicalize duplicate operands in `shape.cstr_broadcastable`   Frederik Gossen
• [Mlir-commits] [mlir] 5f8acd4 - [MLIR] Canonicalize `shape.assuming` op to yield only inner values   Frederik Gossen
• [Mlir-commits] [mlir] 94ef248 - Revert "[MLIR] Canonicalize `shape.assuming` op to yield only inner values"   Frederik Gossen
• [Mlir-commits] [mlir] 630afc6 - [MLIR][Shape] Canonicalize casted dynamic extent tensor   Frederik Gossen
• [Mlir-commits] [mlir] 46ef6ff - [NFC] Test commit. Add empty lines.   Sergei Grechanik
• [Mlir-commits] [mlir] fd2b089 - [mlir][Vector] Lowering of transfer_read/write to vector.load/store   Sergei Grechanik
• [Mlir-commits] [mlir] 2aef202 - [mlir] Fix invalid hoisting of dependent allocs in buffer hoisting pass.   Julian Gross
• [Mlir-commits] [mlir] e231070 - [MLIR] Create memref dialect and move dialect-specific ops from std.   Julian Gross
• [Mlir-commits] [mlir] ea51e7d - Added documentation for SSA like property in Bufferization.   Julian Gross
• [Mlir-commits] [mlir] 06b0380 - [mlir] Introduce CloneOp and adapt test cases in BufferDeallocation.   Julian Gross
• [Mlir-commits] [mlir] c1a4cd5 - [mlir][linalg] refactor the result handling during vectorization.   Tobias Gysi
• [Mlir-commits] [mlir] 8808222 - [mlir][linalg] Do not call region builder during vectorization.   Tobias Gysi
• [Mlir-commits] [mlir] 5837fdc - [mlir][llvm] Pass struct results as parameter in c wrapper   Stephan Herhut
• [Mlir-commits] [mlir] c714b44 - [mlir][Shape] Make cstr_eq more like cstr_broadcastable   Benjamin Kramer
• [Mlir-commits] [mlir] 24acade - [mlir][Shape] Make shape_eq nary   Benjamin Kramer
• [Mlir-commits] [mlir] 73cb58d - [mlir][Shape] Lower cstr_eq to shape_eq + assert   Benjamin Kramer
• [Mlir-commits] [mlir] 42c195f - [mlir][Shape] Allow shape.split_at to return extent tensors and lower it to std.subtensor   Benjamin Kramer
• [Mlir-commits] [mlir] 6327a7c - [mlir][Linalg] Make LLVM_DEBUG region bigger to avoid warnings in Release builds   Benjamin Kramer
• [Mlir-commits] [mlir] 2995e16 - [mlir]: Add canonicalization for dim of 1D alloc of size rank.   Adrian Kuegel
• [Mlir-commits] [mlir] 4a8c01a - Move BaseOpWithOffsetSizesAndStrides to OpBase.td   Adrian Kuegel
• [Mlir-commits] [mlir] c691b96 - [mlir] Add an option to still use bottom-up traversal   Adrian Kuegel
• [Mlir-commits] [mlir] b5d9a3c - [Canonicalizer] Process regions top-down instead of bottom up & reuse existing constants.   Chris Lattner
• [Mlir-commits] [mlir] 91a6ad5 - [m_Constant] Check #operands/results before hasTrait()   Chris Lattner
• [Mlir-commits] [mlir] b2f232b - [testsuite] Make testsuite more stable vs canonicalization change. NFC.   Chris Lattner
• [Mlir-commits] [mlir] 361b7d1 - [Canonicalizer] Process regions top-down instead of bottom up & reuse existing constants.   Chris Lattner
• [Mlir-commits] [mlir] 3a506b3 - Change OwningRewritePatternList to carry an MLIRContext with it.   Chris Lattner
• [Mlir-commits] [mlir] ffde3ac - [ShapeDialect] Silence a build warning, NFC   Chris Lattner
• [Mlir-commits] [mlir] 1d909c9 - Remove the extraneous MLIRContext argument from populateWithGenerated. NFC.   Chris Lattner
• [Mlir-commits] [mlir] 6874726 - [PatternMatching] Add convenience insert method to OwningRewritePatternList. NFC.   Chris Lattner
• [Mlir-commits] [mlir] 549e190 - [PatternRewriter] Rename OwningRewritePatternList -> RewritePatternSet and insert -> add   Chris Lattner
• [Mlir-commits] [mlir] dc4e913 - [PatternMatch] Big mechanical rename OwningRewritePatternList -> RewritePatternSet and insert -> add. NFC   Chris Lattner
• [Mlir-commits] [mlir] 289eccc - Tidy up some docs.   Chris Lattner
• [Mlir-commits] [mlir] 79d7f61 - Rename FrozenRewritePatternList -> FrozenRewritePatternSet; NFC.   Chris Lattner
• [Mlir-commits] [mlir] 782c534 - [ODS] Implement a new 'hasCanonicalizeMethod' bit for cann patterns.   Chris Lattner
• [Mlir-commits] IMPORTANT NOTICE - Subscription to Mailman lists disabled immediately   Tanya Lattner
• [Mlir-commits] [mlir] d36a15d - [mlir][linalg] Memoize indexing map generation.   Stella Laurenzo
• [Mlir-commits] [mlir] 6d2fd3d - [mlir][linalg] Replace monomorphic contration ops with polymorphic variants.   Stella Laurenzo
• [Mlir-commits] [mlir] a9ccdfb - NFC: Glob all python sources in the MLIR Python bindings.   Stella Laurenzo
• [Mlir-commits] [mlir] 0b5f1b8 - [mlir][linalg] Add linalg_opdsl tool first draft.   Stella Laurenzo
• [Mlir-commits] [mlir] e31c77b - [mlir][python] Reorganize MLIR python into namespace packages.   Stella Laurenzo
• [Mlir-commits] [mlir] d4cba4a - [mlir][linalg] Add structured op builders from python opdsl.   Stella Laurenzo
• [Mlir-commits] [mlir] 436c6c9 - NFC: Break up the mlir python bindings into individual sources.   Stella Laurenzo
• [Mlir-commits] [mlir] d9343e6 - [mlir][python] Function decorator for capturing a FuncOp from a python function.   Stella Laurenzo
• [Mlir-commits] [mlir] 8d05a28 - [mlir][python] Adapt to `segment_sizes` attribute type change.   Stella Laurenzo
• [Mlir-commits] [mlir] bdf4e93 - Fix extraneous context parameter in templated helper function.   Stella Laurenzo
• [Mlir-commits] [mlir] ec294eb - [mlir][linalg] Add an InitTensorOp python builder.   Stella Laurenzo
• [Mlir-commits] [mlir] 594e0ba - [mlir][python] Add docs for op class extension mechanism.   Stella Laurenzo
• [Mlir-commits] [mlir] 4ca39da - NFC: Update MLIR python bindings docs to install deps via requirements.txt.   Stella Laurenzo
• [Mlir-commits] [mlir] 9a9214f - [mlir] Add C and python API for is_registered_operation.   Stella Laurenzo
• [Mlir-commits] [mlir] ffa455d - [mlir] Translate global initializers after creating all LLVM IR globals   Jean Perier
• [Mlir-commits] [mlir] 2f0b4db - [mlir] Add convenience grouping for tensor type inference   Jacques Pienaar
• [Mlir-commits] [mlir] 08f0764 - Remove use of tuple for multiresult type storage   Jacques Pienaar
• [Mlir-commits] [mlir] 87e05eb - Revert "Remove use of tuple for multiresult type storage"   Jacques Pienaar
• [Mlir-commits] [mlir] 113baa2 - Update examples post OwningRewritePatternList change   Jacques Pienaar
• [Mlir-commits] [mlir] cec2443 - Fix the order of directives and the target string   Jacques Pienaar
• [Mlir-commits] [mlir] 5d6b4aa - [mlir] Compare elements directly rather than creating pair first   Jacques Pienaar
• [Mlir-commits] [mlir] 7ce07c6 - [mlir] Remove unneeded ShapeFunctionLibraryTerminatorOp   Jacques Pienaar
• [Mlir-commits] [mlir] f3f0c6c - [mlir] Remove uses of type-less CreateLoad() APIs (NFC)   Nikita Popov
• [Mlir-commits] [mlir] 42eb658 - [OpaquePtrs] Remove some uses of type-less CreateGEP() (NFC)   Nikita Popov
• [Mlir-commits] [mlir] 25a20b8 - [mlir] Correct verifyCompatibleShapes   Tres Popp
• [Mlir-commits] [mlir] bc888a0 - [MLIR, test] Fix variable def in Dialect/Linalg/tile-and-distribute.mlir   Thomas Preud'homme
• [Mlir-commits] [mlir] 3dfa861 - [mlir][IR] Refactor the internal implementation of Value   River Riddle
• [Mlir-commits] [mlir] 55f878b - [mlir][pdl] Add a new !pdl.range<> type   River Riddle
• [Mlir-commits] [mlir] e07c968 - [mlir][pdl][NFC] Rename InputOp to OperandOp   River Riddle
• [Mlir-commits] [mlir] 83ef862 - [mlir] Add support for generating Attribute classes for ODS   River Riddle
• [Mlir-commits] [mlir] 6bc767c - [mlir] Add a DialectAsmParser::getChecked method   River Riddle
• [Mlir-commits] [mlir] 1447ec5 - [mlir][AttrDefGen] Add support for specifying the value type of an attribute   River Riddle
• [Mlir-commits] [mlir] 2f37cdd - [mlir][IR][NFC] Move a majority of the builtin attributes to ODS   River Riddle
• [Mlir-commits] [mlir] f175ba4 - [mlir][AsmPrinter] Don't use string comparison when filtering list attributes   River Riddle
• [Mlir-commits] [mlir] a4bb667 - [mlir][IR][NFC] Define the Location classes in ODS instead of C++   River Riddle
• [Mlir-commits] [mlir] 0d01dfb - [mlir][IR][NFC] Move the remaining builtin types to ODS   River Riddle
• [Mlir-commits] [mlir] 4a7aed4 - [mlir][IR] Add a new SymbolUserMap class   River Riddle
• [Mlir-commits] [mlir] a776ecb - [mlir][IR] Add an Operation::eraseOperands that supports batch erasure   River Riddle
• [Mlir-commits] [mlir] 4e02eb8 - [mlir] Optimize the implementation of RegionDCE   River Riddle
• [Mlir-commits] [mlir] 31bb8ef - [mlir][StorageUniquer] Properly call the destructor on non-trivially destructible storage instances   River Riddle
• [Mlir-commits] [mlir] 242762c - [mlir][pdl] Restructure how results are represented.   River Riddle
• [Mlir-commits] [mlir] 02c4c0d - [mlir][pdl] Remove CreateNativeOp in favor of a more general ApplyNativeRewriteOp.   River Riddle
• [Mlir-commits] [mlir] 1eb6994 - [mlir][PDL] Add support for variadic operands and results in PDL   River Riddle
• [Mlir-commits] [mlir] 3a833a0 - [mlir][PDL] Add support for variadic operands and results in the PDL Interpreter   River Riddle
• [Mlir-commits] [mlir] 85ab413 - [mlir][PDL] Add support for variadic operands and results in the PDL byte code   River Riddle
• [Mlir-commits] [mlir] 1f13963 - [mlir][pdl] Cast the OperationPosition to Position to fix MSVC miscompile   River Riddle
• [Mlir-commits] [mlir] 425e11e - [mlir][AttrTypeDefGen] Add support for custom parameter comparators   River Riddle
• [Mlir-commits] [mlir] caa7038 - [mlir][IR] Move the remaining builtin attributes to ODS.   River Riddle
• [Mlir-commits] [mlir] e60d574 - [mlir][Python] Fix test broken after D98474   River Riddle
• [Mlir-commits] [mlir] d70185e - [mlir][IR] Support parsing hex float values in the DialectSymbolParser   River Riddle
• [Mlir-commits] [mlir] ee74860 - [mlir][Toy] Update the tutorial to use tablegen for dialect declarations   River Riddle
• [Mlir-commits] [mlir] 5a8d5a2 - [mlir][Toy] Tidy up the first half of Chapter 2.   River Riddle
• [Mlir-commits] [mlir] cde203e - [mlir][Pass] Coalesce dynamic pass pipelines before running   River Riddle
• [Mlir-commits] [mlir] d75a611 - [mlir] Update `simplifyRegions` to use RewriterBase for erasure notifications   River Riddle
• [Mlir-commits] [mlir] 1a75be0 - [mlir][NFC] Use the native range instead of APInt when computing operand ranges   River Riddle
• [Mlir-commits] [mlir] caddfbd - [mlir][docs] Remove the BuiltinDialect documentation from langref and generate it from ODS   River Riddle
• [Mlir-commits] [mlir] 6d6fe9c - [mlir][OpAsmFormat] Add support for an "else" group on optional elements   River Riddle
• [Mlir-commits] [mlir] 76f3c2f - [mlir][Pattern] Add better support for using interfaces/traits to match root operations in rewrite patterns   River Riddle
• [Mlir-commits] [mlir] caa5144 - [mlir] Use mlir::OpState::operator->() to get to Operation::getAttrs().   Christian Sigg
• [Mlir-commits] [mlir] 9d7be77 - [mlir] Move cuda tests   Christian Sigg
• [Mlir-commits] [mlir] b6ac26f - [mlir] Silence -Wglobal-constructors error in CudaRuntimeWrapper.cpp   Christian Sigg
• [Mlir-commits] [mlir] f69d5a7 - [mlir] Initialize CUDA context lazily.   Christian Sigg
• [Mlir-commits] [mlir] 5fedf30 - [mlir] Make cuInit() call thread-safe.   Christian Sigg
• [Mlir-commits] [mlir] 7cdcb4a - [mlir] NFC: Add #endif comment.   Christian Sigg
• [Mlir-commits] [mlir] 840ff84 - [mlir] Default for gpu-binary-annotation option.   Christian Sigg
• [Mlir-commits] [mlir] 4d295cf - [mlir] Add base class for GpuKernelToBlobPass   Christian Sigg
• [Mlir-commits] [mlir] 6a291ed - [mlir] Remove unnecessary copying of pass options   Christian Sigg
• [Mlir-commits] [mlir] 2224221 - [mlir] Add NVVM to CUBIN conversion to mlir-opt   Christian Sigg
• [Mlir-commits] [mlir] bafe418 - [mlir] Change test-gpu-to-cubin to derive from SerializeToBlobPass   Christian Sigg
• [Mlir-commits] [mlir] 1ef544d - [mlir] Remove mlir-cuda-runner   Christian Sigg
• [Mlir-commits] [mlir] a825fb2 - [mlir] Remove mlir-rocm-runner   Christian Sigg
• [Mlir-commits] [mlir] 74ffe8d - [mlir] Remove ConvertKernelFuncToBlob   Christian Sigg
• [Mlir-commits] [mlir] a5f9cda - [mlir] Rename gpu-to-llvm pass implementation file   Christian Sigg
• [Mlir-commits] [mlir] ddae61d - [mlir] Remove deprecated methods from mlir::OpState   Christian Sigg
• [Mlir-commits] [mlir] 0524a09 - [mlir] Tune error message for assertion.   Sean Silva
• [Mlir-commits] [mlir] acce0ea - [mlir][AVX512] Add mask.compress to AVX512 dialect.   Matthias Springer
• [Mlir-commits] [mlir] c40e0d7 - [mlir][AVX512] Implement sparse vector dot product integration test.   Matthias Springer
• [Mlir-commits] [mlir] 581672b - [mlir][AVX512] Add while loop-based sparse vector-vector dot product variants.   Matthias Springer
• [Mlir-commits] [mlir] 95f8135 - [mlir] Change vector.transfer_read/write "masked" attribute to "in_bounds".   Matthias Springer
• [Mlir-commits] [mlir] 801067f - [mlir][lldb] Fix several gcc warnings in mlir and lldb   Stella Stamenova
• [Mlir-commits] [mlir] f6e0fc2 - [mlir] Fix tile-and-distribute.mlir   Stella Stamenova
• [Mlir-commits] [mlir] 087bc20 - [MLIR][TOSA] Lower tosa.transpose to linalg.generic   Rob Suderman
• [Mlir-commits] [mlir] cb3542e - [MLIR][TOSA] Added lowerings for Reduce operations to Linalg   Rob Suderman
• [Mlir-commits] [mlir] f4bb076 - [mlir][tosa] Add tosa.slice to std.subtensor lowering   Rob Suderman
• [Mlir-commits] [mlir] 5627564 - [mlir][tosa] Add tosa.concat to subtensor inserts lowering   Rob Suderman
• [Mlir-commits] [mlir] 286a9d4 - [mlir][tosa] Add lowering for tosa.rescale to linalg.generic   Rob Suderman
• [Mlir-commits] [mlir] 1b74981 - [mlir][tosa] Add tosa.logical_* to linalg lowerings   Rob Suderman
• [Mlir-commits] [mlir] 47286fc - [mlir][tosa] Add tosa.cast to linalg lowering   Rob Suderman
• [Mlir-commits] [mlir] e990fa2 - [mlir][tosa] Add tosa.reverse lowering to linalg.generic   Rob Suderman
• [Mlir-commits] [mlir] d7c44a5 - [mlir][tosa] Fix tosa.mul to use tosa.apply_scale   Rob Suderman
• [Mlir-commits] [mlir] e20911b - [mlir][tosa] Add tosa.matmul and tosa.fully_connected lowering   Rob Suderman
• [Mlir-commits] [mlir] 2d72b67 - [mlir][tosa] Add tosa.tile to linalg.generic lowering   Rob Suderman
• [Mlir-commits] [mlir] 4157a07 - [mlir][tosa] Add tosa.pad to linalg.pad operation   Rob Suderman
• [Mlir-commits] [mlir] 28e6420 - [mlir][tosa] Add tosa.argmax to linalg lowering   Rob Suderman
• [Mlir-commits] [mlir] f5ba3ee - [mlir][tosa] Add tosa.bitwise_not lowering to constant and xor   Rob Suderman
• [Mlir-commits] [mlir] ec46e03 - [mlir][tosa] TOSA MLIR dialect update to v0.22, part 1   Rob Suderman
• [Mlir-commits] [mlir] 888c506 - Move non-spec TOSA operators into TosaUtilOps.td   Rob Suderman
• [Mlir-commits] [mlir] a879a1b - [mlir][tosa] Add tosa.reciprocal and tosa.sigmoid lowerings   Rob Suderman
• [Mlir-commits] [mlir] ea7f211 - [mlir] Add polynomial approximation for math::ExpOp   Ahmed Taei
• [Mlir-commits] [mlir] da1e37a - Fold full-size subview of static shapes.   Ahmed Taei
• [Mlir-commits] [mlir] f596394 - Add arm_neon.sdot operation   Ahmed Taei
• [Mlir-commits] [mlir] 0588215 - [mlir][Linalg] NFC - Add isOutputTensor to LinalgInterfaces.td   Nicolas Vasilache
• [Mlir-commits] [mlir] 4f4f3f1 - [mlir] NFC - Add runner util functions to only print MemRef metadata.   Nicolas Vasilache
• [Mlir-commits] [mlir] a756f12 - [mlir][Linalg] Add folding of linalg.copy that are in fact identities.   Nicolas Vasilache
• [Mlir-commits] [mlir] f3cc854 - [mlir][Vector] Add folding of vector transfers from/into tensor producing ops.   Nicolas Vasilache
• [Mlir-commits] [mlir] f21d786 - [mlir] Tighten the rules around folding TensorLoadOp   Nicolas Vasilache
• [Mlir-commits] [mlir] 3590840 - [mlir][scf] Canonicalize scf.for last tensor iteration result.   Nicolas Vasilache
• [Mlir-commits] [mlir] c86d3c1 - [mlir][Linalg] Fix order of dimensions in hoistPaddingOnTensors.   Nicolas Vasilache
• [Mlir-commits] [mlir] fd7eee6 - scf::ForOp: Fold away iterator arguments with no use and for which the corresponding input is yielded   Nicolas Vasilache
• [Mlir-commits] [mlir] b661788 - [mlir] NFC - Expose GlobalCreator so it can be reused.   Nicolas Vasilache
• [Mlir-commits] [mlir] 0a74a71 - [mlir] scf::ForOp: Drop iter arguments (and corresponding result) with no use   Nicolas Vasilache
• [Mlir-commits] [mlir] 5b2d850 - [mlir][Linalg] NFC - Expose helper function `substituteMin`.   Nicolas Vasilache
• [Mlir-commits] [mlir] bcd6424 - [mlir][Linalg] Fix linalg on tensor fusion   Nicolas Vasilache
• [Mlir-commits] [mlir] 2240568 - [MLIR][Linalg] Hoist padding across multiple levels of tiling   Nicolas Vasilache
• [Mlir-commits] [mlir] 7716e55 - [mlir] Fixes to hoist padding   Nicolas Vasilache
• [Mlir-commits] [mlir] 69d01e0 - [mlir][python] NFC - Fix stale path in doc   Nicolas Vasilache
• [Mlir-commits] [mlir] 2f367f3 - [mlir][Linalg] Allow calling named ops when available and make it the default.   Nicolas Vasilache
• [Mlir-commits] [mlir] 43b9fa3 - [mlir][Linalg][Python] Create the body of builtin named Linalg ops   Nicolas Vasilache
• [Mlir-commits] [mlir] 37eca08 - [mlir][NFC] Rename `MemRefType::getMemorySpace` to `getMemorySpaceAsInt`   Vladislav Vinogradov
• [Mlir-commits] [mlir] e3b350c - [mlir][NFC] Use `BaseMemRefType::isValidElementType` in Parser   Vladislav Vinogradov
• [Mlir-commits] [mlir] 5d613e4 - [mlir][ODS] Use StringLiteral instead of StringRef when applicable   Vladislav Vinogradov
• [Mlir-commits] [mlir] 2241b39 - [mlir][CMAKE] Fix cross-compilation build   Vladislav Vinogradov
• [Mlir-commits] [mlir] f3bf5c0 - [mlir] Model MemRef memory space as Attribute   Vladislav Vinogradov
• [Mlir-commits] [mlir] fee9054 - [mlir][ODS] Support specialized Attribute class for Enums   Vladislav Vinogradov
• [Mlir-commits] [mlir] 02834e1 - [mlir][ODS] Get rid of limitations in rewriters generator   Vladislav Vinogradov
• [Mlir-commits] [mlir] 18a2f47 - [mlir][NFC] Replace `getMemorySpaceAsInt` with `getMemorySpace` where possible   Vladislav Vinogradov
• [Mlir-commits] [mlir] 70b6f16 - [mlir] Support MemRefType with multiple AffineMaps in getStridesAndOffset   Vladislav Vinogradov
• [Mlir-commits] [mlir] 15b76e6 - [mlir][ODS] Fix `VariadicRegion` code generation for `NoTerminator` Ops   Vladislav Vinogradov
• [Mlir-commits] [mlir] c0f8115 - [mlir][linalg] Only generate one var for an attrUse.   Hanhan Wang
• [Mlir-commits] [mlir] 497b7b8 - [mlir][linalg] Delete unused vars if there are shaped-only operands.   Hanhan Wang
• [Mlir-commits] [mlir] 83c56aa - [mlir][linalg] Add depthwise_conv_2d_input_nhwc_filter_hwcf to Linalg TC ops.   Hanhan Wang
• [Mlir-commits] [mlir] b47c6c6 - [mlir][linalg] Add suffix "Op" to pooling TC ops.   Hanhan Wang
• [Mlir-commits] [mlir] d5d4fb6 - [mlir][linalg] Add support for using scalar attributes in TC ops.   Hanhan Wang
• [Mlir-commits] [mlir] 2ce4caf - Moved getStaticLoopRanges and getStaticShape methods to LinalgInterfaces.td to add static shape verification   Hanhan Wang
• [Mlir-commits] [mlir] f584633 - Added static verification for Linalg Ops.   Hanhan Wang
• [Mlir-commits] [mlir] f178c13 - [mlir] Support use-def cycles in graph regions during regionDCE   Andrew Young
• [Mlir-commits] [mlir] 9c61c76 - [mlir][cse] do not replace operands in previously simplified operations   Andrew Young
• [Mlir-commits] [mlir] 1d297f9 - [lit] Sort test start times based on prior test timing data   David Zarzycki
• [Mlir-commits] [mlir] 4d90e46 - [MLIR][SPIRV] Rename `spv.spcConstant...` to `spv.SpcConstant...`.   Lei Zhang
• [Mlir-commits] [mlir] 5abdca4 - [MLIR][SPIRV] Rename `spv.constant` to `spv.Constant`.   Lei Zhang
• [Mlir-commits] [mlir] c74eb46 - [MLIR][SPIRV] Rename `spv.globalVariable` to `spv.GlobalVariable`.   Lei Zhang
• [Mlir-commits] [mlir] 29812a6 - [MLIR][SPIRV] Rename `spv.loop` to `spv.mlir.loop`.   Lei Zhang
• [Mlir-commits] [mlir] d48ceb4 - [MLIR][SPIRV] Rename `spv.undef` to `spv.Undef`.   Lei Zhang
• [Mlir-commits] [mlir] bb6f5c8 - [mlir][spirv] Convert tensor.extract for very small tensors   Lei Zhang
• [Mlir-commits] [mlir] 50000ab - [mlir] Use affine.apply when distributing to processors   Lei Zhang
• [Mlir-commits] [mlir] 619c150 - [mlir][spirv] Define spv.Image Operation   Lei Zhang
• [Mlir-commits] [mlir] fcc1ce0 - Revert "Revert "[mlir] Add linalg.fill bufferization conversion""   Lei Zhang
• [Mlir-commits] [mlir] ddf93ab - [mlir][linalg] NFC: Move makeTiledShapes into Utils.{h|cpp}   Lei Zhang
• [Mlir-commits] [mlir] e58597e - [mlir][linalg] Fuse producers with non-permutation indexing maps   Lei Zhang
• [Mlir-commits] [mlir] 23fd266 - [mlir][affine] Deduplicate affine min/max op expressions   Lei Zhang
• [Mlir-commits] [mlir] f66120a - [mlir][affine] Add canonicalization to merge affine min/max ops   Lei Zhang
• [Mlir-commits] [mlir] 7f28d27 - [mlir][linalg] Allow controlling folding unit dim reshapes   Lei Zhang
• [Mlir-commits] [mlir] c241e1c - [mlir][linalg] Support dropping unit dimensions for init tensors   Lei Zhang
• [Mlir-commits] [mlir] 19435d3 - [mlir][linalg] Fold fill -> tensor_reshape chain   Lei Zhang
• [Mlir-commits] [mlir] c089170 - [mlir] Add polynomial approximation for math::Log2   Eugene Zhulenev
• [Mlir-commits] [mlir] 39b2cd4 - [mlir] Annotate functions used only in debug mode with LLVM_ATTRIBUTE_UNUSED   Eugene Zhulenev
• [Mlir-commits] [mlir] 74f6138 - [mlir] Add lowering from math::Log1p to LLVM   Eugene Zhulenev
• [Mlir-commits] [mlir] 32a744a - [mlir] Add linalg.fill bufferization conversion   Eugene Zhulenev
• [Mlir-commits] [mlir] 6c9fbcf - [mlir] Linalg: add tile interchange flag to test-linalg-codegen-strategy pass   Eugene Zhulenev
• [Mlir-commits] [mlir] 4a2930f - [mlir] Add loop codegen options to some LLVM dialect ops.   Alex Zinenko
• [Mlir-commits] [mlir] 19db802 - [mlir] make implementations of translation to LLVM IR interfaces private   Alex Zinenko
• [Mlir-commits] [mlir] 32c49c7 - [mlir] ODS: change OpBuilderDAG to OpBuilder   Alex Zinenko
• [Mlir-commits] [mlir] 4e39335 - [mlir] Add an AccessGroup attribute to load/store LLVM dialect ops and generate the access_group LLVM metadata.   Alex Zinenko
• [Mlir-commits] [mlir] 6410ee0 - [mlir] Squash LLVM_ArmNeon dialect into ArmNeon   Alex Zinenko
• [Mlir-commits] [mlir] 90fec5e - [mlir] make MLIRPresburger depend on MLIRIR   Alex Zinenko
• [Mlir-commits] [mlir] 8184247 - [mlir] move LLVM target import header and tests   Alex Zinenko
• [Mlir-commits] [mlir] 0af53de - [mlir] simplify type constraints in AVX512 dialect   Alex Zinenko
• [Mlir-commits] [mlir] a776942 - [mlir] squash LLVM_AVX512 dialect into AVX512   Alex Zinenko
• [Mlir-commits] [mlir] 78f3fb4 - [mlir] Update comments in ArmNeon dialect. NFC   Alex Zinenko
• [Mlir-commits] [mlir] e02dd79 - [mlir] fix typo in OpDefinitions.md   Alex Zinenko
• [Mlir-commits] [mlir] b599f46 - [mlir][CMAKE] Fix build with BUILD_SHARED_LIBS=ON   Alex Zinenko
• [Mlir-commits] [mlir] ed71553 - Revert "[mlir][Affine][Vector] Add initial support for 'iter_args' to Affine vectorizer."   Alex Zinenko
• [Mlir-commits] [mlir] 79da91c - Revert "[mlir][Vector][Affine] Improve affine vectorizer algorithm"   Alex Zinenko
• [Mlir-commits] [mlir] b4a516c - [mlir] Add LLVM loop codegen options to control software pipelining   Alex Zinenko
• [Mlir-commits] [mlir] 3ba14fa - [mlir] Introduce data layout modeling subsystem   Alex Zinenko
• [Mlir-commits] [mlir] 2710439 - [mlir] fix cmake build   Alex Zinenko
• [Mlir-commits] [mlir] be5b844 - [mlir] fix memory leak on failure path in parser   Alex Zinenko
• [Mlir-commits] [mlir] 4affd0c - [mlir] fix a memory leak in NestedPattern   Alex Zinenko
• [Mlir-commits] [mlir] 40d8e4d - Revert "[Canonicalizer] Process regions top-down instead of bottom up & reuse existing constants."   Alex Zinenko
• [Mlir-commits] [mlir] 0308515 - [mlir] fix cmake for generating data layout documentation   Alex Zinenko
• [Mlir-commits] [mlir] a883714 - [mlir] better formatting in interface docs   Alex Zinenko
• [Mlir-commits] [mlir] 0fb4a20 - [mlir] fix shared-lib build fallout of e2310704d890ad252aeb1ca28b4b84d29514b1d1   Alex Zinenko
• [Mlir-commits] [mlir] e82a30b - [mlir] enable Python bindings for the MemRef dialect   Alex Zinenko
• [Mlir-commits] [mlir] 7aa6f3a - [mlir] fix integration tests post e2310704d890ad252aeb1ca28b4b84d29514b1d1   Alex Zinenko
• [Mlir-commits] [mlir] 0aceb61 - [mlir] make memref.cast implement ViewLikeOpInterface   Alex Zinenko
• [Mlir-commits] [mlir] b868a3e - [mlir] fix SPIR-V CPU and Vulkan runners after e2310704d890ad252aeb1ca28b4b84d29514b1d1   Alex Zinenko
• [Mlir-commits] [mlir] 5fac87d - [mlir] verify that operand/result_segment_sizes attributes have i32 element   Alex Zinenko
• [Mlir-commits] [mlir] 20c68d9 - [mlir] silence -Wunused-variable in release mode in Linalg transforms   Alex Zinenko
• [Mlir-commits] [mlir] 1916b0e - [mlir] support data layout specs on ModuleOp   Alex Zinenko
• [Mlir-commits] [mlir] f9cdc61 - [mlir] provide a version of data layout size hooks in bits   Alex Zinenko
• [Mlir-commits] [mlir] 842d243 - [mlir] forward data layout query to scoping op in absence of specification   Alex Zinenko
• [Mlir-commits] [mlir] b3386a7 - [mlir] introduce data layout entry for index type   Alex Zinenko
• [Mlir-commits] [mlir] db694c5 - [mlir] fix -Wsign-compare in memref unit tests   Alex Zinenko
• [Mlir-commits] [mlir] d68ba1f - [mlir] Register Linalg passes in C API and Python Bindings   Alex Zinenko
• [Mlir-commits] [mlir] 4c782a2 - [mlir] Fix typo in SCF.cpp (NFC)   lorenzo chelini
• [Mlir-commits] [mlir] a9e68db - [mlir] Add canonicaliations for subtensor_insert operation.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 3b021fb - [MLIR][LinAlg] Detensorize interal function control flow.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 5d7e0a2 - [mlir] Add LinalgInterface method to clone with a given BlockAndValueMapping.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] c118fdc - [mlir] Remove incorrect folding for SubTensorInsertOp   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 3fb384d - [MLIR][SPIRV] Rename `spv.selection` to `spv.mlir.selection`.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 1694765 - [mlir][linalg] Extend linalg vectorization to support non-identity input maps   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 1a572f4 - [mlir] Add vector op support to cuda-runner including vector.print   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 44f24f3 - [mlir] Fix build failure due to 1a572f4   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 3587728 - [mlir] Fix cuda integration test failure   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] a531bbd - [MLIR] Test pattern benefit sorting between operation specific and operation agnostic patterns.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 5657f93 - [mlir] Canonicalize IfOp with trivial `then` and `else` bodies to list of SelectOp's   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 7219b31 - [mlir] Additional folding for SelectOp   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] e5f2898 - [MLIR][STD] Fold trunci (zexti).   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] c52a5f2 - MLIR][STD] Fold trunci (sexti).   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 5288c25 - [mlir][vector] Add lowering of Transfer_read with broadcast and permutation map   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 7d8b478 - [mlir][Linalg] Drop spurious error message   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] f0a2fe7 - [mlir][Linalg] Rewrite SubTensors that take a slice out of a unit-extend dimension.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] 9b05170 - [mlir] Enhance InferShapedTypeOpInterface and move LinalgOps to use them.   llvmlistbot at llvm.org
• [Mlir-commits] [mlir] c4d5b95 - Fix broken build for commit 9b0517035faee275ce1feabb03d0c7606ea7f819   llvmlistbot at llvm.org

Last message date: Wed Mar 31 14:21:58 PDT 2021
Archived on: Wed Mar 31 14:27:29 PDT 2021

• Messages sorted by: [ thread ] [ subject ] [ date ]
• More info on this list...