Eyal Rozenberg

Would you believe PTX doesn't have rotate left and rotate right? It's true... and neither do we. Let's add them as non-builtins. Note that one can use `funnel_shift` to neatly...

Need to make a pass over the code to ensure I'm not missing any `volatile` qualifiers anywhere, particularly on inline PTX instructions and their wrappers.

Much of the code is lacking Doxygen comments - especially file-level comments. Write those in.

We currently have the `collaborative/` directory and the `sequence_ops/` directory. This is less than ideal, for three reasons: 1. `sequence_ops/` are also a kind of collaborative primitives 2. Some primitives...

We currently have the builtin `load_global_with_non_coherent_cache()`, which translates to the CUDA intrinsic `__ldg()`, which translates into one of the PTX instruction starting with `ld.global.nc.` (depending on the size). But -...

In `on_device/builtins.cuh` we don't have multi-PTX-instruction implementations - although some of the functions, for some of the types, need them (e.g. maximum, minimum). It is probably a good idea, therefore,...

Some of our math functions have two implementations: A more efficient one for runtime (e.g. using PTX instructions), and a less-efficient one which is `constexpr`. I keep them separate both...

While it's rarely a great idea, for the sake of completeness, we may want to have implementations of the Add abstract `` and `` algorithms which could be run by...

"Built-ins" in cuda-kat means those functions which translate into single PTX instructions (not necessarily single SASS instructions though!) We have `on_device/builtins.cuh`, and `on_device/non-builtins.cuh` which contains functions which are builtin-like, or...

We've already added support for some of the `` methods for accessing ranges, like `std::begin()` and `std::end()`. But - we haven't added any of their "reverse" variant,s e.g.` std::rbegin()` and...