Commits · d44cb1838f848370fc8894ad461f3935e4b22a06 · Roger Ferrer / llvm-epi

Browse files

Jun 02, 2017
- generic: add missing get_work_dim include · 2f2a3bc0
  Jan Vesely authored Jun 02, 2017
  2f2a3bc0
Feb 25, 2017
- math: Implement sinh function · 9f717296
  Jan Vesely authored Feb 25, 2017
  9f717296
Feb 24, 2017
- .gitignore: Ignore amdgcn-mesa object directory · c3868c8f
  Jan Vesely authored Feb 24, 2017
  c3868c8f
Feb 23, 2017
- math: Add native_tan as wrapper to tan · dfec3c8e
  Aaron Watry authored Feb 23, 2017
  dfec3c8e
Feb 12, 2017
- Move BufferPtr into the block where it it being used · 80d2e8ff
  Jeroen Ketema authored Feb 12, 2017
  80d2e8ff
- Add the correct prefixes to the cl_khr_fp64 pragma · ed98e8d0
  Jeroen Ketema authored Feb 12, 2017
  ed98e8d0
Feb 09, 2017
- math: Add native_rsqrt builtin function · 9df2b978
  Matt Arsenault authored Feb 09, 2017
  9df2b978
Jan 18, 2017
- math: Add logb builtin · c606efab
  Aaron Watry authored Jan 18, 2017
  c606efab
- math: Add expm1 builtin function · 900bd7eb
  Aaron Watry authored Jan 18, 2017
  900bd7eb
Nov 14, 2016
- Fix build since r286752. · d83eb34e
  Tom Stellard authored Nov 14, 2016
  d83eb34e
Nov 11, 2016
- Fix build since llvm r286566 and require at least llvm 4.0 · 088faab4
  Tom Stellard authored Nov 11, 2016
  088faab4
Sep 21, 2016
- Provide vstore_half helper to workaround clc restrictions · 0a5aac3f
  Jan Vesely authored Sep 21, 2016
  0a5aac3f
Sep 17, 2016
- configure: Add amdgcn-mesa-mesa3d target · 6b195ece
  Tom Stellard authored Sep 16, 2016
  6b195ece
- amdgcn-amdhsa: Add get_num_groups implementation · f19cf403
  Tom Stellard authored Sep 16, 2016
  f19cf403
- amdgcn-amdhsa: Add get_global_size() implementation · e7ad23ba
  Tom Stellard authored Sep 16, 2016
  e7ad23ba
Sep 15, 2016
- math: Implement tgamma · af569547
  Aaron Watry authored Sep 15, 2016
  af569547
- math: Implement lgamma · e9009cdd
  Aaron Watry authored Sep 15, 2016
  e9009cdd
- math: Implement lgamma_r · 0ab07e1b
  Aaron Watry authored Sep 15, 2016
  0ab07e1b
- Add ADDR_SPACE parameter to _CLC_V_V_VP_VECTORIZE · f969413a
  Aaron Watry authored Sep 15, 2016
  f969413a
Sep 08, 2016
- Replace nextafter implementation · fbfd828d
  Matt Arsenault authored Sep 08, 2016
  fbfd828d
- Avoid ambiguity in calling atom_add functions. · eade1727
  Jan Vesely authored Sep 07, 2016
  eade1727
Aug 30, 2016
- configure.py: Add polaris10 and polaris11 · 63f71057
  Niels Ole Salscheider authored Aug 30, 2016
  63f71057
Aug 25, 2016
- amdgcn: Fix return type of get_num_groups · 958fce31
  Matt Arsenault authored Aug 25, 2016
  958fce31
- Strip opencl.ocl.version metadata · 7ef7e6aa
  Matt Arsenault authored Aug 25, 2016
  7ef7e6aa
Aug 24, 2016
- amdgcn: Also correct get_local_size type for HSA · d0a27522
  Matt Arsenault authored Aug 24, 2016
  d0a27522
- amdgcn: Fix return type for get_global_size · 26d9c41f
  Matt Arsenault authored Aug 24, 2016
  26d9c41f
Aug 20, 2016
- amdgpu: Fix default case value for get_local_size · 314364cb
  Matt Arsenault authored Aug 20, 2016
  314364cb
- amdgcn: Fix get_local_size IR return type · 220268d1
  Matt Arsenault authored Aug 20, 2016
  220268d1
- amdgcn: Correct return types to be size_t · 2ce3d94a
  Matt Arsenault authored Aug 19, 2016
  2ce3d94a
Aug 17, 2016
- Implement vstore_half{,n} · ad867272
  Jan Vesely authored Aug 17, 2016
  ad867272
Jul 26, 2016
- Make min follow the OCL 1.0 specs · 4c59714a
  Jan Vesely authored Jul 25, 2016
  4c59714a
Jul 23, 2016
- Implement cbrt builtin · d835b3f1
  Tom Stellard authored Jul 22, 2016
  d835b3f1
- Implement cosh builtin · 9cb070f9
  Tom Stellard authored Jul 22, 2016
  9cb070f9
- geometric/floatn.inc: Add vec8 and vec16 types · ff13926a
  Tom Stellard authored Jul 22, 2016
  ff13926a
Jul 22, 2016
- AMDGPU: Implement get_global_offset builtin · a82e080b
  Jan Vesely authored Jul 22, 2016
  a82e080b
- AMDGPU: Use clang intrinsics for workitem builtins · 74f02db9
  Jan Vesely authored Jul 22, 2016
  74f02db9
- ptx: Fix builtin names after clang r274770 · 7846c9b8
  Jan Vesely authored Jul 22, 2016
  7846c9b8
Jul 19, 2016
- amdgpu: Use right builtn for rsq · 633d749d
  Matt Arsenault authored Jul 19, 2016
  633d749d
Jul 18, 2016
- R600: Use new barrier intrinsic · 1ab0d9c1
  Matt Arsenault authored Jul 18, 2016
  1ab0d9c1
- Replace llvm.AMDGPU.ldexp with llvm.amdgcn.ldexp · b456c6dd
  Matt Arsenault authored Jul 18, 2016
  b456c6dd