projects / oota-llvm.git / history
? search:
summary | shortlog | log | commit | commitdiff | tree
first ⋅ prev ⋅ next
[AVX512] Remove alternate data type versions of VALIGND, VALIGNQ, VMOVSHDUP and VMOVS...
[oota-llvm.git] / lib / Target / AMDGPU / SIShrinkInstructions.cpp
2015-10-20 Matt ArsenaultAMDGPU: Add MachineInstr overloads for instruction... blob | commitdiff | raw
2015-09-10 Matt ArsenaultAMDGPU: Simplify debug printing blob | commitdiff | raw | diff to current
2015-08-08 Matt ArsenaultAMDGPU/SI: Remove VCCReg blob | commitdiff | raw | diff to current
2015-08-08 Matt ArsenaultAMDGPU/SI: Remove source uses of VCCReg blob | commitdiff | raw | diff to current
2015-07-14 Tom StellardAMDGPU/SI: Add support for shrinking v_cndmask_b32_e32... blob | commitdiff | raw | diff to current
2015-07-13 Tom StellardAMDGPU/SI: Select mad patterns to v_mac_f32 blob | commitdiff | raw | diff to current
2015-07-09 Tom StellardAMDGPU/SI: The SIShrinkInstructions pass should only... blob | commitdiff | raw | diff to current
2015-06-13 Tom StellardR600 -> AMDGPU rename blob | commitdiff | raw | diff to current
C/C++ LLVM-based compilers that forbids OOTA behaviors