71 / asmdot Goto Github PK

Can it used for Arm64 or X86-64?

Right now, integers larger than 8 bits are either encoded in little endian, or in the endianness of the running machine. It would be nice to have the ability to switch it to big endian.

In Rust and Haskell, it's only a matter of changing the functions we use. In other languages though, it either requires a manual shift or another library.

I feel like the names I gave to the root directories aren't really helpful to someone who doesn't know how the project is structured.

If anyone has ideas that could improve these names, please submit them below.

It would be nice to use a stream instead of raw pointers like we do in C.

This comparison of generated binaries shows that using the immutable buffer option (which is the default option in the CLI, and can be toggled off via --update-buffer) leads to fewer generated instructions in most cases.

Since the intended goal of --update-buffer was to make things faster when emitting x86 code (at the cost of making things slower for ARM), it can be argued that it does not serve this purpose, and that both the --update-buffer and the --return void options can be removed, making the library more consistent and significantly easier to maintain.

If anyone sees a problem with this removal, please comment below. Otherwise, these features will be removed shortly.