aboutsummaryrefslogtreecommitdiff
path: root/libclc/amdgcn
AgeCommit message (Expand)AuthorFilesLines
2020-08-17libclc: Make all built-ins overloadableDaniel Stone9-70/+77
2018-11-27amdgcn: Consolidate atomic minmax helpersJan Vesely11-57/+4
2018-11-04amdgcn: Move __clc_amdgcn_s_waitcnt definition to clc fileJan Vesely4-15/+1
2018-11-04amdgcn: Convert get_num_groups to clcJan Vesely13-75/+16
2018-11-04amdgcn: Convert get_global_size to clcJan Vesely13-75/+16
2018-11-04amdgcn: Convert get_local_size to clcJan Vesely13-75/+16
2018-08-03amdgcn: Use __constant AS for amdgcn builtins.Jan Vesely2-2/+6
2018-05-17Add initial support for half precision builtinsJan Vesely2-0/+30
2018-04-17amdgcn/fmin: Fix typos that reduced precisionJan Vesely1-3/+3
2018-04-05amdgcn: Update datalayout after LLVM r328656Jan Vesely4-4/+4
2018-03-08amdgcn/fmax: fcanonicalize operandsJan Vesely2-0/+32
2018-03-08amdgcn/fmin: fcanonicalize operandsJan Vesely2-0/+32
2018-03-08amdgcn,popcount: Workaround broken llvm.ctpop intrinsic on some GCN ASICsJan Vesely3-0/+24
2018-02-23amdgcn: Fix build after GDS/const AS swap in r325030Jan Vesely6-10/+20
2018-02-23amdgcn: Fix datalayout after addition of 32bit const AS in r324747Jan Vesely4-4/+4
2018-02-23amdgcn: Fix datalayout after clang r324101Jan Vesely17-5/+150
2017-10-20amdgcn: Add missing datalayout info to .ll filesJan Vesely7-0/+14
2017-10-01Let get_work_dim take exactly 0 argumentsJeroen Ketema1-1/+1
2017-09-29Restore support for llvm-3.9Jan Vesely5-0/+60
2017-09-20Implement cl_khr_int64_extended_atomics builtinsJan Vesely2-0/+48
2017-09-04amdgcn,waitcnt: Add datalayout infoJan Vesely1-0/+2
2017-08-16amdgcn: rewrite barrier() using fence and clang __builtin_amdgcn_s_barrierJan Vesely3-33/+8
2017-08-16amdgcn: Implement {read_,write_,}mem_fence builtinJan Vesely3-0/+52
2016-08-25amdgcn: Fix return type of get_num_groupsMatt Arsenault2-0/+22
2016-08-24amdgcn: Fix return type for get_global_sizeMatt Arsenault2-0/+22
2016-08-20amdgpu: Fix default case value for get_local_sizeMatt Arsenault1-1/+1
2016-08-20amdgcn: Fix get_local_size IR return typeMatt Arsenault2-0/+22
2016-08-19amdgcn: Correct return types to be size_tMatt Arsenault3-3/+3
2016-07-22AMDGPU: Implement get_global_offset builtinJan Vesely2-0/+12
2016-07-22AMDGPU: Use clang intrinsics for workitem builtinsJan Vesely6-62/+34
2016-07-18Replace llvm.AMDGPU.ldexp with llvm.amdgcn.ldexpMatt Arsenault2-0/+48
2016-02-17amdgcn: Use new workitem intrinsicsMatt Arsenault3-0/+62
2016-02-13Split sources for amdgcn and r600Matt Arsenault3-0/+33
generated by cgit v1.1 at 2024-11-13 15:32:59 +0000