I usually want to optimize the scalar code outside of the manually-vectorized body of the loops. A function call to that external compilation unit will be slower than inlining I have when everything is in the same unit. However, it could be the call overhead is small enough, obviously need to profile.
On GCC-like compilers, you could just use inline assembly. That definitely won’t get rewritten into some other instruction sequence, and it can handle things like register allocation and loads/stores for you. Downsides include that the compiler won’t be able to estimate instruction timings, the ease of screwing up the input/output notation, and that MSVC doesn’t support inline assembly on x64 at all.
Interesting idea, will try next time.
I usually want to optimize the scalar code outside of the manually-vectorized body of the loops. A function call to that external compilation unit will be slower than inlining I have when everything is in the same unit. However, it could be the call overhead is small enough, obviously need to profile.
On GCC-like compilers, you could just use inline assembly. That definitely won’t get rewritten into some other instruction sequence, and it can handle things like register allocation and loads/stores for you. Downsides include that the compiler won’t be able to estimate instruction timings, the ease of screwing up the input/output notation, and that MSVC doesn’t support inline assembly on x64 at all.