Commits · b95b867dd83703cc52ace3ed5c14eeba568005f4 · Roger Ferrer / llvm-epi-0.8 · GitLab

This project is archived. Its data is read-only.

Oct 15, 2017
- [globalisel][tablegen] Import ComplexPattern when used as an operator · b95b867d
  Daniel Sanders authored Oct 15, 2017
  b95b867d
- [X86] Remove the SlowBTMem feature flag entirely · 27381173
  Craig Topper authored Oct 15, 2017
  27381173
- [unittests] Only build llvm-cfi-verify if X86 is in LLVM_TARGETS_TO_BUILD · c2e2e0a6
  Krzysztof Parzyszek authored Oct 15, 2017
  c2e2e0a6
- [AVX512] Don't mark EXTLOAD as legal with AVX512. Continue using custom lowering. · a5af4a64
  Craig Topper authored Oct 15, 2017
  a5af4a64
- [X86] Add FeatureSlowBTMem to Haswell, Broadwell, Skylake, Cannonlake, and Knights Landing CPUs. · a1f9c9dd
  Craig Topper authored Oct 15, 2017
  a1f9c9dd
- [TableGen] Remove error checks incorrectly failing on non-error conditions · 4b3876f7
  Krzysztof Parzyszek authored Oct 15, 2017
  4b3876f7
- revert r314984: revert r314698 - [InstCombine] remove one-use restriction for... · 934738a3
  Sanjay Patel authored Oct 15, 2017
  934738a3
- [SimplifyCFG] use range-for-loops, tidy; NFCI · 30f30d37
  Sanjay Patel authored Oct 15, 2017
  30f30d37
- Reverting r315590; it did not include changes for llvm-tblgen, which is... · 615eb470
  Aaron Ballman authored Oct 15, 2017
  615eb470
- [MergeFunctions] Merge small functions if possible without a thunk. · ae12efab
  whitequark authored Oct 15, 2017
  ae12efab
- [MergeFunctions] Replace all uses of unnamed_addr functions. · b2ce9ffe
  whitequark authored Oct 15, 2017
  b2ce9ffe
- [X86] Ignore DBG instructions in X86CmovConversion optimization to resolve PR34565 · c8d67979
  Amjad Aboud authored Oct 15, 2017
  c8d67979
- [LoopInfo][Refactor] Make SetLoopAlreadyUnrolled a member function of the Loop Pass, NFC. · 73f65043
  Hongbin Zheng authored Oct 15, 2017
  73f65043
- [X86] Lower vselect with constant condition to vector_shuffle even with AVX512 instructions. · a9cd59fb
  Craig Topper authored Oct 15, 2017
  a9cd59fb
- [X86] Don't use constant condition for select instruction when testing masking ops. · f02e9785
  Craig Topper authored Oct 15, 2017
  f02e9785
- Remove unused variables · 7450398e
  Vitaly Buka authored Oct 15, 2017
  7450398e
- [globalisel][tablegen] Map ld and st to G_LOAD and G_STORE. NFC · 39690bdf
  Daniel Sanders authored Oct 15, 2017
  39690bdf
- [tablegen] Handle common load/store predicates inside tablegen. NFC. · 3f267bf7
  Daniel Sanders authored Oct 15, 2017
  3f267bf7
- [Hexagon] Mark RangeTree::dump() with LLVM_DUMP_METHOD. · 76067588
  Davide Italiano authored Oct 14, 2017
  76067588
- AMDGPU: Temporary disable pal metadata check line in llvm-readobj test · 263f7f66
  Konstantin Zhuravlyov authored Oct 14, 2017
  263f7f66
- AMDGPU: Don't use TargetStreamer if it has not been initialized · 8c18f5b3
  Konstantin Zhuravlyov authored Oct 14, 2017
  8c18f5b3
Oct 14, 2017
- [X86] Remove a bunch of dead FileCheck lines with the wrong prefix. · dfb443e8
  Craig Topper authored Oct 14, 2017
  dfb443e8
- [TableGen] Avoid unnecessary std::string creations · 6ecae9fc
  Simon Pilgrim authored Oct 14, 2017
  6ecae9fc
- [X86][SSE] Don't attempt to reduce the imul vector width of odd sized vectors (PR34947) · 36fe00ee
  Simon Pilgrim authored Oct 14, 2017
  36fe00ee
- [X86][SSE] Test vector imul reduction on 32 and 64-bit targets · 3f49b988
  Simon Pilgrim authored Oct 14, 2017
  3f49b988
- Revert "[AArch64][RegisterBankInfo] Use the statically computed mappings for COPY" · caac2fbd
  Bruno Cardoso Lopes authored Oct 14, 2017
  caac2fbd
- AMDGPU: Add AMDGPU HSA Kernel Descriptor · 13376a4b
  Konstantin Zhuravlyov authored Oct 14, 2017
  13376a4b
- AMDGPU: Bring HSA metadata on par with the specification · a01d8b0b
  Konstantin Zhuravlyov authored Oct 14, 2017
  a01d8b0b
- llvm-readobj: Print AMDGPU note contents · b3c605d6
  Konstantin Zhuravlyov authored Oct 14, 2017
  b3c605d6
- Pull out repeated calls to VT.getVectorNumElements(). NFCI. · f5b9f353
  Simon Pilgrim authored Oct 14, 2017
  f5b9f353
- Cleanup update_llc_test_checks.py notes. · 5bd4431a
  Simon Pilgrim authored Oct 14, 2017
  5bd4431a
- AMDGPU: Cleanup elf-notes.ll test · 7b4be1ed
  Konstantin Zhuravlyov authored Oct 14, 2017
  7b4be1ed
- Use DAG::getBitcast() helper. NFCI. · cded8283
  Simon Pilgrim authored Oct 14, 2017
  cded8283
- llvm-readobj: Print AMDGPU note type names · 716af741
  Konstantin Zhuravlyov authored Oct 14, 2017
  716af741
- AMDGPU: Improve note directive verification in assembler · 219066ba
  Konstantin Zhuravlyov authored Oct 14, 2017
  219066ba
- AMDGPU: Do not emit deprecated notes for code object v3 · eda425ed
  Konstantin Zhuravlyov authored Oct 14, 2017
  eda425ed
- AMDGPU: Add support for isa version note · 9c05b2bc
  Konstantin Zhuravlyov authored Oct 14, 2017
  9c05b2bc
- [X86][SSE] Support combining AND(EXTRACT(SHUF(X)), C) -> EXTRACT(SHUF(X)) · f367c27d
  Simon Pilgrim authored Oct 14, 2017
  f367c27d
- [X86] Add patterns for vzmovl+cvtpd2dq/cvttpd2dq with a load. · f7e77776
  Craig Topper authored Oct 14, 2017
  f7e77776
- [X86] Add AVX512 versions of VCVTPD2PS to load folding tables. · 61010a85
  Craig Topper authored Oct 14, 2017
  61010a85