| | 2 | |
| | 3 | CLRadeonExtender 0.1.4: |
| | 4 | |
| | 5 | * add AMD RX VEGA support (GCN 1.4/VEGA) |
| | 6 | * add symbol scopes |
| | 7 | * add support for 32-bit AMD OpenCL 2.0 binaries |
| | 8 | * update GPU device ids to latest drivers |
| | 9 | * add Ellesmere and Baffin support for AMD OpenCL 1.2 binaries |
| | 10 | * add support for LLVM 3.9, LLVM 4.0 and Mesa3D 17.0 |
| | 11 | * add new options to clrxasm (--llvmVersion) |
| | 12 | * add GCN 1.2 instruction set documentation |
| | 13 | * add new SMEM instruction (s_buffer_atomics) |
| | 14 | * add GDS segment size to AMD OpenCL 2.0 binaries |
| | 15 | * add code of samples for GCN 1.2 |
| | 16 | * add option to use old AMD OpenCL 1.2 binary format into samples |
| | 17 | * add editor's syntax (NotePad++, Kate, Gedit, VIM) |
| | 18 | * minor fixes in GCN assembler |
| | 19 | * add modifier's parametrization |
| | 20 | * add options to control case-sensitiviness in macro names |
| | 21 | * fixed handling AMDOCL names for 32-bit Windows environment |
| | 22 | * add installation rules for AMDGPU-PRO drivers (OpenSUSE and Ubuntu) |
| | 23 | * add new pseudo-ops '.get_64bit', '.get_arch', '.get_format', '.get_gpu' |
| | 24 | * add autodetection for LLVM and Mesa3D version |
| | 25 | * find correct AMDOCL, MesaOCL and llvm-config at runtime |
