mirror of https://github.com/intel/intel-qs.git
709 B
709 B
List of TODO
Instead of printing TODO messages during compilation, the suggestions are collected here. The messages are left in the source code as comments (search for keyword TODO).
highperfkernels.cpp:
- Not using ICC vectorization: need to rewrite in intrincics
- Allow for case where state is not aligned: need SIMD ISA for un-aligned access
- Add nthreads check to clamp to smaller number if too little work
- Generalize for AVX3 cases so we check for pos <=1 etc
qureg_apply1qubitgate.cpp
- Fix problem when coming here from controlled gate
- Add diagonal special case
qureg_applyctrl1qubitgate.cpp
- Way to fix problem with X and Y specialization
- Insert Loop_SN specialization for this case