Log - e9987a1d2fad218f2da3c78a8bee60ecc87262d9 - toolchain/llvm-project

f74700a [CostModel][X86] Merged SK_PermuteSingleSrc/SK_PermuteTwoSrc into common shuffle cost LUTs. NFCI. by Simon Pilgrim · 9 years ago
bca02f9 [CostModel][X86] Add support for broadcast shuffle costs by Simon Pilgrim · 9 years ago
a62395a [CostModel][X86] Pulled out common type legalization code by Simon Pilgrim · 9 years ago
23599ba Currently isLikelyComplexAddressComputation tries to figure out if the given stride seems to be 'complex' and need some extra cost for address computation handling. by Mohammed Agabaria · 9 years ago
189e2d2 [Test Commit] fixing some format issue in X86TTI to match clang-format output. by Mohammed Agabaria · 9 years ago
bb895f3 [CostModel][X86] Updated vXi8 and vXi16 Reverse/Alternate shuffle costs by Simon Pilgrim · 9 years ago
939b8cd [X86] Merged Reverse/Alternate shuffle cost tables. NFCI. by Simon Pilgrim · 9 years ago
d96200d Fixed shuffle-reverse cost on AVX-512. by Elena Demikhovsky · 9 years ago
21706cb AVX-512 Loop Vectorizer: Cost calculation for interleave load/store patterns. by Elena Demikhovsky · 9 years ago
081abbb [X86][SSE] Improve lowering of vXi64 multiplies by Simon Pilgrim · 9 years ago
2f7f0e7 [CostModel][X86] Updated reverse shuffle costs by Simon Pilgrim · 9 years ago
841d7ca [X86][AVX512] Add support for v2i64 fptosi/fptoui/sitofp/uitofp on AVX512DQ-only targets by Simon Pilgrim · 9 years ago
4e9b9cb [X86][AVX512] Add support for v4i64 fptosi/fptoui/sitofp/uitofp on AVX512DQ-only targets by Simon Pilgrim · 9 years ago
03cd8f8 [CostModel][X86] Add missing AVX512DQ v8i64 fptosi/sitofp costs by Simon Pilgrim · 9 years ago
779da8e [CostModel][X86] Added mul costs for vXi8 vectors by Simon Pilgrim · 9 years ago
27fed8e [X86][AVX] Fixed v16i16/v32i8 ADD/SUB costs on AVX1 subtargets by Simon Pilgrim · 9 years ago
d02c552 [VectorLegalizer] Expansion of CTLZ using CTPOP when possible by Simon Pilgrim · 9 years ago
d07c731 Improved cost model for FDIV and FSQRT, by Andrew Tischenko by Alexey Bataev · 9 years ago
d23219b [X86][AVX512] Fix MUL v8i64 costs on non-AVX512DQ targets by Simon Pilgrim · 9 years ago
820e132 [X86][AVX512DQ] Improve lowering of MUL v2i64 and v4i64 by Simon Pilgrim · 9 years ago
6ac1e98 [X86][SSE] Add SSE41/AVX1 costs for vector shifts. by Simon Pilgrim · 9 years ago
b2443ed [X86] Enable interleaved memory access by default by Michael Kuperstein · 9 years ago
365be4f [CostModel][X86] Fixed AVX1/AVX512 sdiv/udiv uniformconst costs for 256/512 bit integer vectors by Simon Pilgrim · 9 years ago
025e26d [CostModel][X86] Fixed AVX1/AVX512 sdiv/udiv general costs for 256/512 bit integer vectors by Simon Pilgrim · 9 years ago
4ddc92b [X86][SSE] Add lowering to cvttpd2dq/cvttps2dq for sitofp v2f64/2f32 to 2i32 by Simon Pilgrim · 9 years ago
b271a58 NFC: The Cost Model specialization, by Andrey Tischenko by Alexey Bataev · 9 years ago
b03fd12 Replace "fallthrough" comments with LLVM_FALLTHROUGH by Justin Bogner · 9 years ago
e9c32c7 Revert "[X86] Support the "ms-hotpatch" attribute." by Charles Davis · 9 years ago
0822aa1 [X86] Support the "ms-hotpatch" attribute. by Charles Davis · 9 years ago
3ceac2b [LV, X86] Be more optimistic about vectorizing shifts. by Michael Kuperstein · 9 years ago
5d5ca9c [X86][SSE] Add initial costs for vector CTTZ/CTLZ by Simon Pilgrim · 9 years ago
f44b79d [AVX512] Don't use i128 masked gather/scatter/load/store. Do more accurately dataWidth check. by Igor Breger · 9 years ago
1b4f511 [X86][SSE] Add cost model values for CTPOP of vectors by Simon Pilgrim · 9 years ago
285d9e4 Strip trailing whitespace by Simon Pilgrim · 9 years ago
f0c5933 [X86] Make some cast costs more precise by Michael Kuperstein · 9 years ago
04b3496 [x86] fix cost of SINT_TO_FP for i32 --> float (PR21356, PR28434) by Sanjay Patel · 9 years ago
1b62e0e [X86] Sort cast cost tables. NFC. by Michael Kuperstein · 9 years ago
356e823 [X86][SSE] Add cost model for BSWAP of vectors by Simon Pilgrim · 9 years ago
3fc09f7 [CostModel][X86][SSE] Updated costs for vector BITREVERSE ops on SSSE3+ targets by Simon Pilgrim · 9 years ago
9a0542a [X86] Add costs for SSE zext/sext to v4i64 to TTI by Michael Kuperstein · 9 years ago
aedc347 [x86] avoid code explosion from LoopVectorizer for gather loop (PR27826) by Sanjay Patel · 10 years ago
14000b3 [CostModel][X86][XOP] Added XOP costmodel for BITREVERSE by Simon Pilgrim · 10 years ago
eec3a95 [X86][SSE] Improve cost model for i64 vector comparisons on pre-SSE42 targets by Simon Pilgrim · 10 years ago
468558a [X86]: Changing cost for “TRUNCATE v16i32 to v16i8” in SSE4.1 mode. by Ashutosh Nema · 10 years ago
867e914 Do not use getGlobalContext()... ever. by Mehdi Amini · 10 years ago
4c7d094 fix typo; NFC by Sanjay Patel · 10 years ago
9f6c4d5 [x86] fix cost model inaccuracy for vector memory ops by Sanjay Patel · 10 years ago
4d94d4d AVX512BW: Support llvm intrinsic masked vector load/store for i8/i16 element types on SKX by Igor Breger · 10 years ago
6d42141 AVX1 : Enable vector masked_load/store to AVX1. by Igor Breger · 10 years ago
5494698 Implemented cost model for masked gather and scatter operations by Elena Demikhovsky · 10 years ago
8df93ce [X86][SSE] Transform truncations between vectors of integers into X86ISD::PACKUS/PACKSS operations during DAG combine. by Cong Hou · 10 years ago
074e845 [X86] Prevent constant hoisting for a couple compare immediates that the selection DAG knows how to optimize into a shift. by Craig Topper · 10 years ago
59898d8 [X86][SSE] Update the cost table for integer-integer conversions on SSE2/SSE4.1. by Cong Hou · 10 years ago
a1a40cc AVX-512: Updated cost of FP/SINT/UINT conversion operations by Elena Demikhovsky · 10 years ago
1ca72e1 Pointers in Masked Load, Store, Gather, Scatter intrinsics by Elena Demikhovsky · 10 years ago
da4e8ae [X86] A small fix in X86/X86TargetTransformInfo.cpp: check a value type is simple before calling getSimpleVT(). by Cong Hou · 10 years ago
4b27576 Remove templates from CostTableLookup functions. All instantiations had the same type. by Craig Topper · 10 years ago
ee0c859 Convert cost table lookup functions to return a pointer to the entry or nullptr instead of the index. by Craig Topper · 10 years ago
0928585 Scalarizer for masked.gather and masked.scatter intrinsics. by Elena Demikhovsky · 10 years ago
eda02a9 Remove two unnecessary conversions from MVT to EVT. NFC by Craig Topper · 10 years ago
7ad0d56 Partially reverted changes from r250686 by Elena Demikhovsky · 10 years ago
20662e3 Removed parameter "Consecutive" from isLegalMaskedLoad() / isLegalMaskedStore(). by Elena Demikhovsky · 10 years ago
a18ae9b [CostModel] Fixed AVX integer shift costs by Simon Pilgrim · 10 years ago
083ca9b Fix Clang-tidy modernize-use-nullptr warnings in source directories and generated files; other minor cleanups. by Hans Wennborg · 10 years ago
79dd1bf [X86] Teach constant hoisting that ANDs with 64-bit immediates in the range 0x80000000-0xffffffff can be handled cheaply and don't need to be hoisted. by Craig Topper · 10 years ago
3d11c99 [X86][XOP] Added support for the lowering of 128-bit vector shifts to XOP shift instructions by Simon Pilgrim · 10 years ago
93205eb [TTI] Make the cost APIs in TargetTransformInfo consistently use 'int' by Chandler Carruth · 10 years ago
d566fb1 Rename hasCompatibleFunctionAttributes->areInlineCompatible based by Eric Christopher · 10 years ago
86478c6 [X86][SSE] Vectorize i64 ASHR operations by Simon Pilgrim · 10 years ago
e2c244f [X86][SSE] Reordered cast vectorization costs. NFCI. by Simon Pilgrim · 10 years ago
59764dc [X86][SSE] Updated SHL/LSHR i64 vectorization costs. by Simon Pilgrim · 10 years ago
0b305db Prune trailing whitespaces and CRs. by NAKAMURA Takumi · 10 years ago
64cc4ad [X86][SSE] Vectorized v4i32 non-uniform shifts. by Simon Pilgrim · 10 years ago
44ede33 Make TargetLowering::getPointerTy() taking DataLayout as an argument by Mehdi Amini · 10 years ago
8fbf1c1 [X86][SSE] Vectorized i64 uniform constant SRA shifts by Simon Pilgrim · 10 years ago
e100226 Implement TargetTransformInfo::hasCompatibleFunctionAttributes for X86. by Eric Christopher · 10 years ago
5965680 [X86][SSE] Vectorized i8 and i16 shift operators by Simon Pilgrim · 10 years ago
0be4fa7 [X86][AVX2] Vectorized i16 shift operators by Simon Pilgrim · 11 years ago
062c744 [X86] Disable loop unrolling in loop vectorization pass when VF is 1. by Wei Mi · 11 years ago
93dcdc4 [PM] Switch the TargetMachine interface from accepting a pass manager by Chandler Carruth · 11 years ago
705b185 [PM] Change the core design of the TTI analysis to use a polymorphic by Chandler Carruth · 11 years ago
a3232f7 Implemented cost model for masked load/store operations. by Elena Demikhovsky · 11 years ago
fb81b93 Masked Load/Store - Changed the order of parameters in intrinsics. by Elena Demikhovsky · 11 years ago
3fcafa2 Loop Vectorizer minor changes in the code - by Elena Demikhovsky · 11 years ago
f1de34b Masked Load / Store Intrinsics - the CodeGen part. by Elena Demikhovsky · 11 years ago
5bf9578 [X86] Clean up whitespace as well as minor coding style by Michael Liao · 11 years ago
9bc81fb Revert "Masked Vector Load and Store Intrinsics." by Duncan P. N. Exon Smith · 11 years ago
8c5128b Add missing override keywords. by Craig Topper · 11 years ago
9e5089a Masked Vector Load and Store Intrinsics. by Elena Demikhovsky · 11 years ago
d5e95b5 AVX-512: SINT_TO_FP cost model and some bugfixes by Elena Demikhovsky · 11 years ago
360460b [X86] Custom lower UINT_TO_FP from v4f32 to v4i32, and for v8f32 to v8i32 if by Quentin Colombet · 11 years ago
2701247 AVX-512: added cost for some AVX-512 instructions by Elena Demikhovsky · 11 years ago
b653de1 Rename getMaximumUnrollFactor -> getMaxInterleaveFactor; also rename option names controlling this variable. by Sanjay Patel · 11 years ago
7f33ff7 Allow vectorization of division by uniform power of 2. by Karthik Bhat · 11 years ago
d913448 Remove the TargetMachine forwards for TargetSubtargetInfo based by Eric Christopher · 11 years ago
2820a5b [X86] AVX512: Enable it in the Loop Vectorizer by Adam Nemet · 11 years ago
c8e8bda [CostModel][x86] Improved cost model for alternate shuffles. by Andrea Di Biagio · 11 years ago
e03a25d Add Support to Recognize and Vectorize NON SIMD instructions in SLPVectorizer. by Karthik Bhat · 11 years ago
3c890c4 X86: stifle GCC warning by Saleem Abdulrasool · 11 years ago
b2e4edb [ConstantHoisting][X86] Improve the cost model for small constants with large types (i64 and above). by Juergen Ributzka · 11 years ago