summaryrefslogtreecommitdiffstats
path: root/libclc
Commit message (Expand)AuthorAgeFilesLines
...
* vstore_half: Consolidate declarationsJan Vesely2018-02-061-39/+26
* vstore_half: Add support for custom rounding functionsJan Vesely2018-02-061-23/+39
* vstore_half: Make sure the helper function is always inlineJan Vesely2018-02-061-1/+1
* half_powr: Implement using powrJan Vesely2018-02-014-0/+15
* math.h: Use logical operations instead of bit operations for readabilityJan Vesely2018-01-311-1/+1
* math.h: Set HAVE_HW_FMA32 based on compiler provided macroJan Vesely2018-01-291-0/+5
* tanpi: Port from amd_builtinsJan Vesely2018-01-196-0/+168
* tan: Port from amd_builtinsJan Vesely2018-01-199-21/+164
* half_divide: Implement using x/yJan Vesely2018-01-185-0/+27
* half_tan: Implement using tanJan Vesely2018-01-184-0/+17
* half_sin: Implement using sinJan Vesely2018-01-184-0/+17
* half_recip: Implement using 1/xJan Vesely2018-01-184-0/+21
* half_log2: Implement using log2Jan Vesely2018-01-184-0/+17
* half_log10: Implement using log10Jan Vesely2018-01-184-0/+17
* half_log: Implement using logJan Vesely2018-01-184-0/+17
* half_exp10: Implement using exp10Jan Vesely2018-01-184-0/+17
* half_exp2: Implement using exp2Jan Vesely2018-01-184-0/+17
* half_exp: Implement using expJan Vesely2018-01-184-0/+17
* half_cos: Implement using cosJan Vesely2018-01-184-0/+17
* half_sqrt: Cleanup implementationJan Vesely2018-01-183-51/+2
* half_rsqrt: Cleanup implementationJan Vesely2018-01-184-51/+11
* rootn: Port from amd_builtinsJan Vesely2018-01-179-0/+399
* powr: Port from amd_builtinsJan Vesely2018-01-176-0/+411
* pown: Port from amd_builtinsJan Vesely2018-01-178-31/+395
* pow: Port from amd_builtinsJan Vesely2018-01-178-6/+1096
* configure.py: Add gfx900 (Vega, Raven)Vedran Miletic2017-11-271-2/+2
* math: Implement minmagJan Vesely2017-11-155-0/+17
* math: Implement maxmagJan Vesely2017-11-155-0/+17
* native_powr: Switch implementation to native_exp2 and native_log2Jan Vesely2017-11-144-1/+18
* native_divide: provide function implementation instead of macroJan Vesely2017-11-135-1/+17
* native_recip: provide function implementation instead of macroJan Vesely2017-11-134-1/+18
* native_rsqrt: Switch implementation to 1 / native_sqrtJan Vesely2017-11-134-1/+18
* native_tan: Switch implementation to use native_sin/native_cosJan Vesely2017-11-134-9/+17
* math: Use precomputed constant for log2(10.0)Jan Vesely2017-11-133-3/+11
* native_exp10: Switch implementation to llvm intrinsicJan Vesely2017-11-104-1/+18
* native_sqrt: Switch implementation to llvm intrinsicJan Vesely2017-11-103-1/+17
* native_sin: Switch implementation to llvm intrinsicJan Vesely2017-11-103-1/+17
* native_cos: Switch implementation to llvm intrinsicJan Vesely2017-11-103-1/+17
* native_exp2: Switch implementation to llvm intrinsicJan Vesely2017-11-103-1/+17
* native_exp: Switch implementation to llvm intrinsicJan Vesely2017-11-103-1/+17
* amdgpu: Add workaround for unimplemented llvm.exp intrinsicJan Vesely2017-11-103-0/+9
* native_log10: Switch to generic native intrinsic inc fileJan Vesely2017-11-102-8/+2
* native_log: Switch to generic native intrinsic inc fileJan Vesely2017-11-102-30/+2
* native_log2: Switch to generic native intrinsic inc fileJan Vesely2017-11-103-8/+29
* tgamma: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-2/+5
* tanh: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-24/+6
* tan: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-2/+6
* sqrt: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-2/+5
* sinpi: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-2/+5
* sinh: Use unary_decl instead of custom inc fileJan Vesely2017-11-022-24/+6
OpenPOWER on IntegriCloud