- bda6541 [Hexagon] Split large offsets into properly aligned addends by Krzysztof Parzyszek · 7 years ago
- 4eef9f6 [InstCombine] use SelectInst operand names to make code clearer; NFC by Sanjay Patel · 7 years ago
- a82b93e [x86] test codegen for unsigned saturated add; NFC by Sanjay Patel · 7 years ago
- 28b767e [ARC] Fix macro usage (DEBUG -> LLVM_DEBUG) by Tatyana Krasnukha · 7 years ago
- 6df7a517 [AMDGPU] Preliminary patch for divergence driven instruction selection. Inline immediate move to V_MADAK_F32. by Alexander Timofeev · 7 years ago
- 3e62673 [AST] Visit memtransfer arguments in order by Philip Reames · 7 years ago
- 8a7d229 [MIPS GlobalISel] Select icmp by Petar Jovanovic · 7 years ago
- e602052 HotColdSplitting: fix test failing because of last commit by Sebastian Pop · 7 years ago
- d126aab HotColdSplitting: check that target supports cold calling convention by Sebastian Pop · 7 years ago
- 09372b8 add flag instead of using a constant [NFC] by Sebastian Pop · 7 years ago
- 341b8a7 make flag name more specific to gvn [NFC] by Sebastian Pop · 7 years ago
- bd3beb4 [LSR] Add tests for small constants; NFC by Gil Rapaport · 7 years ago
- 0d018d1 InstCombine: move hasOneUse check to the top of foldICmpAddConstant by Tim Northover · 7 years ago
- a45f6b1 [Sparc] Move SparcTargetStreamer.h to the MC Desc, where the implementation is already by Benjamin Kramer · 7 years ago
- 6aabdc9 [LLD][COFF] Cleanup error messages / add more coverage tests by Alexandre Ganea · 7 years ago
- d94f77a [Target] Untangle disassemblers by Benjamin Kramer · 7 years ago
- 3ebb215 Don't create a temporary vector of loop blocks just to iterate over them. by Benjamin Kramer · 7 years ago
- 2c6fa12 [GVN] Invalidate cached info for values replaced by equality propagation by John Brawn · 7 years ago
- 32ca0aa AMDGPU: Remove function pointer type hack by Matt Arsenault · 7 years ago
- fa0980a AMDGPU: Stop reporting is-noop addrspacecast for constant 32-bit by Matt Arsenault · 7 years ago
- 2390ffb DAG: Handle odd vector sizes in calling conv splitting by Matt Arsenault · 7 years ago
- 01fb00c [llvm-exegesis] Ignore double spaced separators in asm strings by Simon Pilgrim · 7 years ago
- a5defd3 [AMDGPU] Prevent sequences of non-instructions disrupting GCNHazardRecognizer wait state counting by Carl Ritson · 7 years ago
- aa81d56 ReleaseNotes: update links to use https by Hans Wennborg · 7 years ago
- fa259d8 [IndVars] Set Changed if rewriteFirstIterationLoopExitValues changes IR. PR38863 by Max Kazantsev · 7 years ago
- 768427d [IndVars] Set Changed if sinkUnusedInvariants changes IR. PR38863 by Max Kazantsev · 7 years ago
- dfb80c8 Move a transformation routine from LoopUtils to LoopVectorize. by Vikram TV · 7 years ago
- 7fb521b [XRay] Fix buildbot failure by David Carlier · 7 years ago
- 5d1d358 Move createMinMaxOp() out of RecurrenceDescriptor. by Vikram TV · 7 years ago
- f32491f [Xray] tooling allow MachO format support by David Carlier · 7 years ago
- 69da26f [XRay] Remove unused reference by Dean Michael Berris · 7 years ago
- 7c299a7 AMDGPU: Fix tests using old number for constant address space by Matt Arsenault · 7 years ago
- 229ad4c [XRay] Add a BlockVerifier visitor for FDR Records by Dean Michael Berris · 7 years ago
- 814247c AMDGPU: Use GOT PSV since it has an address space now by Matt Arsenault · 7 years ago
- 185f21f AMDGPU: Don't abort on unknown addrspace argument by Matt Arsenault · 7 years ago
- 27d4a67 [X86] Custom type legalize (v2i32 (fp_to_uint v2f64))) without avx512vl by widening to v4i32 and v4f64 instead of v8i32 and v8f64. Make it aware of x86-experimental-vector-widening-legalization by Craig Topper · 7 years ago
- 250785b [SelectionDAG] enhance vector demanded elements to look at a vector select condition operand by Sanjay Patel · 7 years ago
- 09d0743 [X86] Create paddus/psubus from narrower vectors with i8/i16 element types. by Craig Topper · 7 years ago
- d9ee98e [X86] Mark the ADCX and ADOX instruction as commutable. by Craig Topper · 7 years ago
- c4efbc0 [X86] Add test cases for commuting ADCX/ADOX instruction to avoid copies. by Craig Topper · 7 years ago
- ffff353 Revert "NFC: use bit_cast more in AArch64AddressingModes" by JF Bastien · 7 years ago
- d1fb1f5 NFC: use bit_cast more in AArch64AddressingModes by JF Bastien · 7 years ago
- 9bcfd97 [X86] Add commuted isel pattern for the load form of ADCX instructions. by Craig Topper · 7 years ago
- 1515971 [X86] Add load folding test cases for the addcarryx intrinsic. by Craig Topper · 7 years ago
- 4ed7b13 [X86] Add stack folding MIR test for ADCX/ADOX. by Craig Topper · 7 years ago
- 4bbbe05 Fix typo in previous commit by JF Bastien · 7 years ago
- 7dcaeaf ADT: add <bit> header, implement C++20 bit_cast, use by JF Bastien · 7 years ago
- 35bfd59 Fix typos. NFC by Fangrui Song · 7 years ago
- 90c166c Remove addBlockByrefAddress(), it is dead code as far as clang is concerned. by Adrian Prantl · 7 years ago
- 68a9873 [MemorySSA] Relax verification of clobbering accesses. by Alina Sbirlea · 7 years ago
- 1e06d13 Fix some of the PDB tests. by Zachary Turner · 7 years ago
- 5dc5797 Revert "ADT: add <bit> header, implement C++20 bit_cast, use" by JF Bastien · 7 years ago
- 25ccaa7 [PDB] Support pointer types in the native reader. by Zachary Turner · 7 years ago
- 265ce99 ADT: add <bit> header, implement C++20 bit_cast, use by JF Bastien · 7 years ago
- 443a8ce [COFF] Implement llvm.global_ctors priorities for MSVC COFF targets by Reid Kleckner · 7 years ago
- 72cb007 [SimplifyIndVar] Avoid generating truncate instructions with non-hoisted Laod operand. by Abderrazek Zaafrani · 7 years ago
- b9bd7dd Set cost of invariant group intrinsics to 0 by Piotr Padlewski · 7 years ago
- 16a4c18 [WebAssembly] v8x16.shuffle by Thomas Lively · 7 years ago
- a8ea812 [benchmark] Fix flags used to compile benchmark library with clang-cl by Reid Kleckner · 7 years ago
- cb88bb7 [benchmark] Re-enable benchmarks on all platforms including Windows by Reid Kleckner · 7 years ago
- a154fcb [InstCombine][x86] add tests for possible blendv transform (PR38814); NFC by Sanjay Patel · 7 years ago
- fe1f0fe [AST] Generalize argument specific aliasing by Philip Reames · 7 years ago
- 1de79cc [codeview] Add .cv_string directive for testing purposes by Reid Kleckner · 7 years ago
- dd2780f [X86] Add codegen tests for narrow PADDUS/PSUBUS patterns for PR38691. by Craig Topper · 7 years ago
- 5666935 [MemorySSA] Update MemoryPhi wiring for block splitting to consider if identical edges were merged. by Alina Sbirlea · 7 years ago
- a5f1bd0 [InstCombine] narrow vector select with padded condition and extracted result (PR38691) by Sanjay Patel · 7 years ago
- 4e9fc1e [WebAssembly] Change SIMD lane indices to vec_i8imm_op by Thomas Lively · 7 years ago
- 906fd9a [AArch64] Support reserving x1-7 registers. by Nick Desaulniers · 7 years ago
- 183b21b [X86] Don't create ZERO_EXTEND_INREG/SIGN_EXTEND_INREG for v1iX vectors. by Craig Topper · 7 years ago
- 7df3d72 [X86] Don't create X86ISD::AVG nodes from v1iX vectors. by Craig Topper · 7 years ago
- 33d40b3 [PGO] Fix some style issue of ControlHeightReduction by Fangrui Song · 7 years ago
- a2fde74 [CMake] Fix LLVM_ENABLE_LTO option on Windows by Alexandre Ganea · 7 years ago
- 84c1ade [X86] Modify the the rdtscp intrinsic to return values instead of taking a pointer argument by Craig Topper · 7 years ago
- d71790c [codeview] Improve readobj FPO dumper and pdbutil register names by Reid Kleckner · 7 years ago
- 96a04e6 [PGO][CHR] Build/warning fix by Hiroshi Yamauchi · 7 years ago
- f119734 [RISCV] Fix crash in decoding instruction with unknown floating point rounding mode by Ana Pazos · 7 years ago
- 1b609e0 [Error] Reintroduce type validation in createFileError() by Alexandre Ganea · 7 years ago
- 4115a63 [llvm-dwp] Clean up tests X86/*.test by Fangrui Song · 7 years ago
- e853ba4 [RISCV] Fix AddressSanitizer heap-buffer-overflow in disassembling by Ana Pazos · 7 years ago
- 537da42 NFC: remove magic bool in LoopIdiomRecognize by JF Bastien · 7 years ago
- 66f1f16 [PGO][CHR] Small cleanup. by Hiroshi Yamauchi · 7 years ago
- 35cb717 [Bindings][Go] Fixed go.test failure due to C-API argument type mismatch. by Kristina Brooks · 7 years ago
- 0c1ea5a utils/abtest: Refactor and add bisection method by Matthias Braun · 7 years ago
- 3071b63 [X86] Change the addcarry and subborrow intrinsics to return 2 results and remove the pointer argument. by Craig Topper · 7 years ago
- d25c18e [X86] Use regular expressions to make test immune to register allocation changes. by Craig Topper · 7 years ago
- 43fc2d6 [X86] Teach X86DAGToDAGISel::foldLoadStoreIntoMemOperand to handle loads in operand 1 of commutable operations. by Craig Topper · 7 years ago
- 189c2c8 [InstCombine] Fold (min/max ~X, Y) -> ~(max/min X, ~Y) when Y is freely invertible by Craig Topper · 7 years ago
- 78b2edd [LV] Fix code gen for conditionally executed loads and stores by Anna Thomas · 7 years ago
- 55883c9 Hot cold splitting pass by Aditya Kumar · 7 years ago
- 1b57b3d [InstCombine] Do not fold scalar ops over select with vector condition. by Florian Hahn · 7 years ago
- 66586cc [DebugInfo] Handle stack slot offsets for spilled sub-registers in LDV by David Stenberg · 7 years ago
- a067621 Add support for getRegisterByName. by Sid Manning · 7 years ago
- 4aea3b7 [X86][SSE] Add additional fadd/fsub(x, bitcast_fneg(y)) tests with different integer bitwidths by Simon Pilgrim · 7 years ago
- e14a443 [DAGCombiner] foldBitcastedFPLogic - Add basic vector support by Simon Pilgrim · 7 years ago
- 49e8ec0 [NewGVN] Mark function as changed if we erase instructions. by Florian Hahn · 7 years ago
- fe3b800 [X86][SSE] Add fadd/fsub(x, bitcast_fneg(y)) tests by Simon Pilgrim · 7 years ago
- 60c6009 [dsymutil] Prevent non-determinism due to threading. by Jonas Devlieghere · 7 years ago
- 0e364e1 [MSan] don't access MsanCtorFunction when using KMSAN by Alexander Potapenko · 7 years ago
- 361c297 ARM: fix Thumb2 CodeGen for ldrex with folded frame-index. by Tim Northover · 7 years ago
- f774071 [MSan] Add KMSAN instrumentation to MSan pass by Alexander Potapenko · 7 years ago