- 97ea58e [LV] Ensure proper handling of multi-use case when collecting uniforms by Matthew Simpson · 9 years ago
- 92544c4 [InstCombine] regenerate checks by Sanjay Patel · 9 years ago
- 01b876d [RDF] Further improve handling of multiple phis reached from shadows by Krzysztof Parzyszek · 9 years ago
- 111f4fb [llvm-cov] Fix issues with segment highlighting in the html view by Vedant Kumar · 9 years ago
- 99a1cd6 [LV] Don't mark pointers used by scalarized memory accesses uniform by Matthew Simpson · 9 years ago
- 3d9234a [pdb] Make YamlTypeDumperCallbacks reuse *this. by Zachary Turner · 9 years ago
- 684da72 [YAMLIO] Add the ability to map with context. by Zachary Turner · 9 years ago
- 1bc1003 AMDGPU: Sign extend constants when splitting them by Matt Arsenault · 9 years ago
- 3358dc4 [Hexagon] Expand sext- and zextloads of vector types, not just extloads by Krzysztof Parzyszek · 9 years ago
- c0196eb AMDGPU: Try to commute when selecting s_addk_i32/s_mulk_i32 by Matt Arsenault · 9 years ago
- a50ca9f AArch64 .arch directive - Include default arch attributes with extensions. by Eric Christopher · 9 years ago
- d764af3 AMDGPU: Support commuting with immediate in src0 by Matt Arsenault · 9 years ago
- 86159cb Revert "[XRay] ARM 32-bit no-Thumb support in LLVM" by Renato Golin · 9 years ago
- b91e524 [LoopDataPrefetch] Use range based for loop; NFCI by Balaram Makam · 9 years ago
- e0563b1 [InstCombine] return a vector-safe true/false constant by Sanjay Patel · 9 years ago
- fcc57ae Add unittest for r280760 by Dehao Chen · 9 years ago
- f3c6524 [InstCombine][X86] Regenerate masked memory op combine tests by Simon Pilgrim · 9 years ago
- 2a7d3fa [InstCombine][X86] Regenerate vperm2f128/vperm2i128 combine tests by Simon Pilgrim · 9 years ago
- 0cd49c0 [InstCombine][X86] Regenerate insertps combine tests by Simon Pilgrim · 9 years ago
- c87eb16 [Coverage] Delete some dead code (NFC) by Vedant Kumar · 9 years ago
- e3aa0d9 [TableGen] AsmMatcher: Add AsmVariantName to Instruction class. by Sam Kolton · 9 years ago
- 88e07a9 Remove restriction that 'sret' must be on the first parameter by Reid Kleckner · 9 years ago
- 1a53e07 Give an x86 assembler test a triple by Reid Kleckner · 9 years ago
- 035ff49 revert r280427 by Dehao Chen · 9 years ago
- 08450b5 [ARM XRay] Try to fix Thumb-only failure by Renato Golin · 9 years ago
- 7442ca4 [SDAGBuilder] Don't create a binary tree for switches in minsize mode by James Molloy · 9 years ago
- ac84ac0 [Thumb1] AND with a constant operand can be converted into BIC by James Molloy · 9 years ago
- 210e77b [Thumb1] Fix cost calculation for complemented immediates by James Molloy · 9 years ago
- d88990b [SelectionDAG] Add BUILD_VECTOR support to computeKnownBits and SimplifyDemandedBits by Simon Pilgrim · 9 years ago
- 707719e [DAGCombiner] Enable AND combines of splatted constant vectors by Simon Pilgrim · 9 years ago
- ad84cf4 Revert "[ARM] Lower UDIV+UREM to UDIV+MLS (and the same for SREM)" by Pablo Barrio · 9 years ago
- 09e9256 [mips][microMIPS] Implement DBITSWAP, DLSA and LWUPC and add tests for AUI instructions by Hrvoje Varga · 9 years ago
- 52bc7ab [asan] Avoid lifetime analysis for allocas with can be in ambiguous state by Vitaly Buka · 9 years ago
- e53b49a Revert "[LoopUnroll] Properly update loop-info when cloning prologues and epilogues." by Michael Zolotukhin · 9 years ago
- d9fa490 [LoopUnroll] Properly update loop-info when cloning prologues and epilogues. by Michael Zolotukhin · 9 years ago
- 837c30f [llvm-cov] Disable zlib compression in a test input, unbreaks bots by Vedant Kumar · 9 years ago
- d2e36ca [llvm-cov] Use less space to describe source names by Vedant Kumar · 9 years ago
- a57be10 [llvm-cov] Drop the longest common filename prefix from summaries by Vedant Kumar · 9 years ago
- 2a93bda [CGP] Be less conservative about tail-duplicating a ret to allow tail calls by Michael Kuperstein · 9 years ago
- 6d2157d [AA] Fix typo in comment (s/hase/has). by Justin Lebar · 9 years ago
- 0b559dd [XRay] Remove unused variable by Dean Michael Berris · 9 years ago
- 339ade7 [XRay] ARM 32-bit no-Thumb support in LLVM by Dean Michael Berris · 9 years ago
- 1d54212 Deleted right file by Piotr Padlewski · 9 years ago
- 6037ae7 Revert "[thinlto] Deleted unused test file" by Piotr Padlewski · 9 years ago
- 1c13d2b IR: Remove Value::intersectOptionalDataWith, replace all calls with calls to Instruction::andIRFlags. by Peter Collingbourne · 9 years ago
- f555e26 Revert "[asan] Avoid lifetime analysis for allocas with can be in ambiguous state" by Vitaly Buka · 9 years ago
- a7ad004 [thinlto] Deleted unused test file by Piotr Padlewski · 9 years ago
- f8002f1 [asan] Avoid lifetime analysis for allocas with can be in ambiguous state by Vitaly Buka · 9 years ago
- 774d87e [InstCombine] use m_APInt to allow icmp (and (sh X, Y), C2), C1 folds for splat constant vectors by Sanjay Patel · 9 years ago
- f0bb43f [CUDA] Rework "optimizations" and "publication" section in CompileCudaWithLLVM.rst. by Justin Lebar · 9 years ago
- 6676095 [CUDA] Clarify that -l and -L only need to be passed when linking, in CompileCudaWithLLVM.rst. by Justin Lebar · 9 years ago
- 70425af [CUDA] Further reformat "invoking clang" section of CompileCudaWithLLVM.rst. by Justin Lebar · 9 years ago
- 9e7800b [SimplifyCFG] Don't try to create metadata-valued PHIs by Hal Finkel · 9 years ago
- 16d3f88 [LoopUnroll] Correct a debug message. NFC. by Haicheng Wu · 9 years ago
- 22c45e1 Shift-left (ISD::SHL) operation crashes on "DAG Legalization" phase. by Elena Demikhovsky · 9 years ago
- 350a7a4 [InstCombine] allow icmp (and X, C2), C1 folds for splat constant vectors by Sanjay Patel · 9 years ago
- 5b033da [CUDA] Fix typo in link in CompileCudaWithLLVM. by Justin Lebar · 9 years ago
- e61b182 [CUDA] Move AXPY example into gist. by Justin Lebar · 9 years ago
- 4ecf8b9 [RDF] Fix liveness analysis for phi nodes with shadow uses by Krzysztof Parzyszek · 9 years ago
- 9c0826c Don't reuse a variable name in a nested scope. NFC. by Michael Kuperstein · 9 years ago
- 80708af [RDF] Introduce "undef" flag for ref nodes by Krzysztof Parzyszek · 9 years ago
- f1708bc [CUDA] Simplify build/install instructions in CompileCudaWithLLVM.rst. by Justin Lebar · 9 years ago
- 188c78b [CUDA] Call it "CUDA", not "CUDA C/C++" in our docs. by Justin Lebar · 9 years ago
- 78e95fa [CUDA] Expand upon --cuda-gpu-arch flag in CompileCudaWithLLVM doc. by Justin Lebar · 9 years ago
- a804c5a9 Rename test pr30298.ll to shrink_vmul_sse.ll, to make the name more meaningful, NFC. by Wei Mi · 9 years ago
- 145ae71 AMDGPU: Remove a useless variable which caused build failure for lld. by Yaxun Liu · 9 years ago
- ebe55c6 Don't reduce the width of vector mul if the target doesn't support SSE2. by Wei Mi · 9 years ago
- 9107e64 Add more triple to conditional-tailcall.ll test by Hans Wennborg · 9 years ago
- 371bac0 Typo. NFC. by Chad Rosier · 9 years ago
- c4c2318 CodeGen: ensure that libcalls are always AAPCS CC by Saleem Abdulrasool · 9 years ago
- 4a83266 X86: Fold tail calls into conditional branches where possible (PR26302) by Hans Wennborg · 9 years ago
- fab897d [lib/LTO] Add a way to run a custom pipeline by Davide Italiano · 9 years ago
- 6874fa8 AMDGPU: Add hidden kernel arguments to runtime metadata by Yaxun Liu · 9 years ago
- eadc14f [codeview] Add new directives to record inlined call site line info by Reid Kleckner · 9 years ago
- 452b6b2 [LoopInterchange] Improve debug output. NFC. by Chad Rosier · 9 years ago
- e760d89 [LoopInterchange] Improve debug output. NFC. by Chad Rosier · 9 years ago
- 600740b [LSV] Use the original loads' names for the extractelement instructions. by Justin Lebar · 9 years ago
- 05601b4 [x86] move combines of 'select of 2 constants' to its own function; NFC by Sanjay Patel · 9 years ago
- 0639c00 Fix typo in test - it should be masking bits0-15 not bit16 by Simon Pilgrim · 9 years ago
- 959abb8 Regenerate vector bitcast folding tests using update_test_checks.py. by Andrea Di Biagio · 9 years ago
- 56463dd [X86][SSE] Added or combine tests for known bits of vectors by Simon Pilgrim · 9 years ago
- b362b6e [X86][SSE] Added and+or+zext combine tests for known bits of vectors by Simon Pilgrim · 9 years ago
- 9313182 [X86][SSE] Added and+or combine tests currently failing with vectors by Simon Pilgrim · 9 years ago
- ba3ea4d [ARM] Lower UDIV+UREM to UDIV+MLS (and the same for SREM) by Pablo Barrio · 9 years ago
- 2fdf2bf [InstCombine][SSE4a] Fix assertion failure in the insertq/insertqi combining logic. by Andrea Di Biagio · 9 years ago
- fd11478 [InstCombine][SSE4a] Fix assertion failure caused by unsafe dyn_casts on the operands of extrq/extrqi intrinsic calls. by Andrea Di Biagio · 9 years ago
- 17c896a Revert "[EfficiencySanitizer] Adds shadow memory parameters for 40-bit virtual memory address." by Renato Golin · 9 years ago
- c7d7511 [mips] Disable the TImode shift libcalls for 32-bit targets. by Vasileios Kalintiris · 9 years ago
- 51cc6bb [EfficiencySanitizer] Adds shadow memory parameters for 40-bit virtual memory address. by Sagar Thakur · 9 years ago
- e50466a [SimplifyCFG] Followup fix to r280790 by James Molloy · 9 years ago
- 80fedcb [SimplifyCFG] Update workaround for PR30188 to also include loads by James Molloy · 9 years ago
- 1317eef [CMake] Use CMake's default RPATH for the unit tests by Diana Picus · 9 years ago
- 40fa86f [SimplifyCFG] Check PHI uses more accurately by James Molloy · 9 years ago
- 21c68fa [PowerPC] Fix address-offset folding for plain addi by Hal Finkel · 9 years ago
- 7840500 AVX512F: FMA intrinsic + FNEG - sequence optimization by Elena Demikhovsky · 9 years ago
- 1da68f6 AMDGPU: Make some scalar instructions commutable by Matt Arsenault · 9 years ago
- 774415b Remove unnecessary call to getAllocatableRegClass by Matt Arsenault · 9 years ago
- 4cf44f1 [X86] Add hasSideEffects=0 to some instructions. by Craig Topper · 9 years ago
- 83d6e11 [AVX-512] Add support for commuting masked instructions in findCommutedOpIndices. The default implementation doesn't skip the mask input or the preserved input. by Craig Topper · 9 years ago
- 4ccc33a Revert "CodeGen: ensure that libcalls are always AAPCS CC" by Saleem Abdulrasool · 9 years ago