Add helpers for modular arithmetic #634
Conversation
The prime-testing techniques we will use (Miller-Rabin, Strong Lucas) all make heavy usage of modular arithmetic. Therefore, we lay those foundations here, adding utilities to perform the basic arithmetic operations robustly. Since these are internal-only helper functions, we don't bother checking the preconditions, although we state them clearly in the contract comment for each utility. After C++26, we could add contracts for these. Helps mpusz#509.
Apparently some freestanding builds need this?
| @@ -32,6 +32,7 @@ | |||
| import std; | |||
| #else | |||
| #include <array> | |||
| #include <bits/types.h> | |||
There was a problem hiding this comment.
This is normally not the correct header. A tool like clangd will suggest this include due to the internals of the clangd compiler, which matches in most cases the gcc compiler.
For most compilers a higher level include is normally need, as this one is an implementation detail of the specific compiler.
Unfortunately I am not sure what the correct C++ header is that must be included
There was a problem hiding this comment.
I quickly looked at the compile error, it appears that <cstdint> is the correct high level include since it complains that 'uint64_t' must be declared before it is used and that should be found in the mentioned header according to https://en.cppreference.com/w/cpp/types/integer
There was a problem hiding this comment.
Thanks! <cstdint> is already included, so I guess the error is just spurious.
Apparently, that's not the right approach.
| // Precondition: (a < n). | ||
| // Precondition: (b < n). | ||
| // Precondition: (n > 0). |
There was a problem hiding this comment.
You can use MP_UNITS_EXPECTS_DEBUG if needed.
There was a problem hiding this comment.
I'd love to express these in code, but here is what I see:
This is with #include <mp-units/compat_macros.h>.
Any advice for a quick fix?
There was a problem hiding this comment.
Sure 😉
mp-units/src/core/include/mp-units/framework/quantity.h
Lines 39 to 40 in dc6fb93
| // Precondition: (a < n). | ||
| // Precondition: (b < n). | ||
| // Precondition: (n > 0). | ||
| [[nodiscard]] consteval uint64_t add_mod(uint64_t a, uint64_t b, uint64_t n) |
There was a problem hiding this comment.
Usage of cstdint without the std:: namespace prefix will probably fail to compile in import std; build.
The prime-testing techniques we will use (Miller-Rabin, Strong Lucas) all make heavy usage of modular arithmetic. Therefore, we lay those foundations here, adding utilities to perform the basic arithmetic operations robustly. Getting these foundations right is extremely important for the work that will follow.
Since these are internal-only helper functions, we don't bother checking the preconditions, although we state them clearly in the contract comment for each utility. After C++26, we could add contracts for these.
Helps #509.