Commits · e7d8ed32f9f98d02801c2a0a2dd094689e830a21 · Roger Ferrer / llvm-epi · GitLab

Dec 22, 2016
- AMDGPU: Swap order of operands in fadd/fsub combine · e7d8ed32
  Matt Arsenault authored Dec 22, 2016
  e7d8ed32
- AMDGPU: Check fast math flags in fadd/fsub combines · 46e6b7ad
  Matt Arsenault authored Dec 22, 2016
  46e6b7ad
- AMDGPU: Form more FMAs if fusion is allowed · 770ec868
  Matt Arsenault authored Dec 22, 2016
  770ec868
- Only substitute into type of non-type template parameter once, rather than · e27c6dfd
  Richard Smith authored Dec 22, 2016
  e27c6dfd
- AMDGPU: Move combines into separate functions · d8b73d53
  Matt Arsenault authored Dec 22, 2016
  d8b73d53
- AMDGPU: Enable some f32 fadd/fsub combines for f16 · ef82ad94
  Matt Arsenault authored Dec 22, 2016
  ef82ad94
- AMDGPU: Implement isFMAFasterThanFMulAndFAdd for f16 · 9e22bc2c
  Matt Arsenault authored Dec 22, 2016
  9e22bc2c
- AMDGPU: setcc test cleanup · 2920f624
  Matt Arsenault authored Dec 22, 2016
  2920f624
- Driver: use the triple to query the arch, not the toolchain · 2511d604
  Saleem Abdulrasool authored Dec 22, 2016
  2511d604
- Driver: remove unnecessary parameter · 491d1d4b
  Saleem Abdulrasool authored Dec 22, 2016
  491d1d4b
- Driver: rename parameter to reduce confusion · 416852fe
  Saleem Abdulrasool authored Dec 22, 2016
  416852fe
- AMDGPU: Allow rcp and rsq usage with f16 · cdff21b1
  Matt Arsenault authored Dec 22, 2016
  cdff21b1
- AMDGPU: Custom lower f16 fdiv · 4052a576
  Matt Arsenault authored Dec 22, 2016
  4052a576
- AMDGPU: Implement f16 fcanonicalize · ce84130f
  Matt Arsenault authored Dec 22, 2016
  ce84130f
- AMDGPU: Update isFPImmLegal for f16 · 4e55c1ec
  Matt Arsenault authored Dec 22, 2016
  4e55c1ec
- Clear the PendingTypeTests vector after moving from it. · 704f814a
  Peter Collingbourne authored Dec 22, 2016
  704f814a
- Add the alloc_size attribute to clang, attempt 2. · e3763371
  George Burgess IV authored Dec 22, 2016
  e3763371
- [AArch64] Correct the check of signed 9-bit imm in getIndexedAddressParts(). · 9ac20a1e
  Haicheng Wu authored Dec 22, 2016
  9ac20a1e
- Pass GetAssumptionCache to InlineFunctionInfo constructor · 180bd9f6
  Easwaran Raman authored Dec 22, 2016
  180bd9f6
- [NVVMIntrRange] Only set range metadata if none is already present · 5fa7d48b
  David Majnemer authored Dec 22, 2016
  5fa7d48b
- Renumber testcase metadata nodes after r290153. · 1eadba1c
  Adrian Prantl authored Dec 22, 2016
  1eadba1c
- [LLParser] Make the line field of DIMacro(File) optional. · 58c19106
  Adrian Prantl authored Dec 22, 2016
  58c19106
- [clang-tidy] Ignore `size() == 0` in the container implementation. · b5ca17f8
  Alexander Kornienko authored Dec 21, 2016
  b5ca17f8
- Legalize metadata in legacy testcases · ec9ebba7
  Adrian Prantl authored Dec 21, 2016
  ec9ebba7
- Legalize metadata in legacy testcases · 762e4b72
  Adrian Prantl authored Dec 21, 2016
  762e4b72
- Legalize metadata in legacy testcases · aad5df48
  Adrian Prantl authored Dec 21, 2016
  aad5df48
- Legalize metadata in legacy testcases · b767f312
  Adrian Prantl authored Dec 21, 2016
  b767f312
- [GlobalISel] Add basic Selector-emitter tblgen backend. · 36f7035b
  Ahmed Bougacha authored Dec 21, 2016
  36f7035b
- [AsmWriter] Remove redundant cast<>s. NFC. · aa9fe532
  Ahmed Bougacha authored Dec 21, 2016
  aa9fe532
- specify -DNDEBUG for BNI builds of all targets in the Xcode build · 756cb33b
  Sean Callanan authored Dec 21, 2016
  756cb33b
- [WebAssembly] Fix the opcode value for i64.rotr. · a2b9b349
  Dan Gohman authored Dec 21, 2016
  a2b9b349
- IR: Function summary representation for type tests. · 1b4137a7
  Peter Collingbourne authored Dec 21, 2016
  1b4137a7
Dec 21, 2016
- Increase the treshold in unit test to accomodate for qurantine size increase. · 76a60a8c
  Evgeniy Stepanov authored Dec 21, 2016
  76a60a8c
- [sancov] skip duplicated points · bfe5045b
  Mike Aizatsky authored Dec 21, 2016
  bfe5045b
- [sancov] hash prefix results in huge merge files, use shorter prefix · 987f6420
  Mike Aizatsky authored Dec 21, 2016
  987f6420
- Perform type-checking for a converted constant expression in a template · 52e624f3
  Richard Smith authored Dec 21, 2016
  52e624f3
- [AArch64] Remove a redundant check. NFC. · 6bb0e393
  Haicheng Wu authored Dec 21, 2016
  6bb0e393
- Add the ability for DWARFDie objects to get the parent DWARFDie. · 78a07bfa
  Greg Clayton authored Dec 21, 2016
  78a07bfa
- [CMake] Support distribution install for LLDB.framework · e9ce09b8
  Chris Bieneman authored Dec 21, 2016
  e9ce09b8
- Fix for the __kmpc_global_num_threads function to return the value of the __kmp_all_nth global var. · 76d42854
  Andrey Churbanov authored Dec 21, 2016
  76d42854