summaryrefslogtreecommitdiffstats
path: root/libclc
Commit message (Expand)AuthorAgeFilesLines
* math: Add logb builtinAaron Watry2017-01-185-0/+36
* math: Add expm1 builtin functionAaron Watry2017-01-186-0/+293
* Fix build since r286752.Tom Stellard2016-11-141-1/+2
* Fix build since llvm r286566 and require at least llvm 4.0Tom Stellard2016-11-112-3/+4
* Provide vstore_half helper to workaround clc restrictionsJan Vesely2016-09-214-26/+75
* configure: Add amdgcn-mesa-mesa3d targetTom Stellard2016-09-161-1/+5
* amdgcn-amdhsa: Add get_num_groups implementationTom Stellard2016-09-163-0/+14
* amdgcn-amdhsa: Add get_global_size() implementationTom Stellard2016-09-162-0/+40
* math: Implement tgammaAaron Watry2016-09-155-0/+77
* math: Implement lgammaAaron Watry2016-09-155-0/+49
* math: Implement lgamma_rAaron Watry2016-09-156-0/+518
* Add ADDR_SPACE parameter to _CLC_V_V_VP_VECTORIZEAaron Watry2016-09-151-12/+27
* Replace nextafter implementationMatt Arsenault2016-09-082-28/+29
* Avoid ambiguity in calling atom_add functions.Jan Vesely2016-09-074-4/+4
* configure.py: Add polaris10 and polaris11Niels Ole Salscheider2016-08-301-2/+2
* amdgcn: Fix return type of get_num_groupsMatt Arsenault2016-08-255-2/+24
* Strip opencl.ocl.version metadataMatt Arsenault2016-08-251-0/+7
* amdgcn: Also correct get_local_size type for HSAMatt Arsenault2016-08-241-5/+8
* amdgcn: Fix return type for get_global_sizeMatt Arsenault2016-08-245-2/+24
* amdgpu: Fix default case value for get_local_sizeMatt Arsenault2016-08-202-2/+2
* amdgcn: Fix get_local_size IR return typeMatt Arsenault2016-08-205-5/+27
* amdgcn: Correct return types to be size_tMatt Arsenault2016-08-193-3/+3
* Implement vstore_half{,n}Jan Vesely2016-08-173-19/+68
* Make min follow the OCL 1.0 specsJan Vesely2016-07-251-2/+2
* Implement cbrt builtinTom Stellard2016-07-227-0/+869
* Implement cosh builtinTom Stellard2016-07-227-0/+370
* geometric/floatn.inc: Add vec8 and vec16 typesTom Stellard2016-07-221-0/+16
* AMDGPU: Implement get_global_offset builtinJan Vesely2016-07-229-1/+33
* AMDGPU: Use clang intrinsics for workitem builtinsJan Vesely2016-07-2214-136/+71
* ptx: Fix builtin names after clang r274770Jan Vesely2016-07-225-13/+13
* amdgpu: Use right builtn for rsqMatt Arsenault2016-07-191-1/+6
* R600: Use new barrier intrinsicMatt Arsenault2016-07-181-4/+3
* Replace llvm.AMDGPU.ldexp with llvm.amdgcn.ldexpMatt Arsenault2016-07-183-3/+3
* configure: Remove device specific definesJan Vesely2016-06-171-25/+11
* nvptx: Drop feature defines.Jan Vesely2016-06-171-6/+4
* 64 bit integers are legal in full profile without an extensionJan Vesely2016-06-172-6/+12
* math: Use single precision fmax in sp pathJan Vesely2016-05-171-1/+1
* math: Add erf ported from amd-builtinsJan Vesely2016-05-064-0/+413
* math: Add fdim implementationAaron Watry2016-05-066-0/+86
* prepare-builtins: Remove call to getGlobalContext()Tom Stellard2016-04-151-1/+1
* [AMDGPU] Implement get_local_size for amdgcn--amdhsa tripleKonstantin Zhuravlyov2016-04-075-1/+41
* Update copyright year to 2016.Paul Robinson2016-03-301-1/+1
* math: Fix ilogb(double) return typeAaron Watry2016-02-241-1/+1
* math: Add ilogb ported from amd-builtinsAaron Watry2016-02-236-0/+68
* Add .gitignore for build directoriesMatt Arsenault2016-02-171-0/+13
* amdgcn: Use new workitem intrinsicsMatt Arsenault2016-02-179-38/+124
* Update page to list supported targetsMatt Arsenault2016-02-131-2/+2
* Split sources for amdgcn and r600Matt Arsenault2016-02-1334-38/+75
* configure: Remove llvm 3.6 definesJan Vesely2016-02-091-3/+3
* configure: Remove cl_khr_fp64 for device that don't support doublesJan Vesely2016-02-091-5/+5
OpenPOWER on IntegriCloud