Clément Pernet

first naive versionn done in #96. Letting this issue open as a reminder to consider using precomputed lists of primes

We should, but I see very little advantage of having both prime integers and irreducible polynomials sharing the same class. There would be too many features specific to each one:...

I could not find any usage of it. If your think there is no reason to keep it (e.g. it slows down a 2 args compose) then drop it!

It may depend on the dimension/magnitude: although the overall arithmetic count is always in favor of the PLUQ variant, solving with a PLUQ at each iteration may be expensive: poorer...

I think the last 2 points are strong enough to motivate this change. I kind of remember discussing this with @jgdumas who had a good argument for the existence of...

Not sure we want to fix this problem on our side, unless someone shows up who actually needs it to be solved.

Suggestion: - decouple detection of int128_t from the capacity to do `std::make_unsigned` - when `std::make_unsigned` is not defined, couldn't we force define it in Givaro?

For the record, I have already decoupled FFLAS-FFPACK detection of int128 from that of Givaro, since fflas-ffpack does not require the ability to do `std:make_unsigned`

Can't reproduce, neither on my Linux Mint 20 with g++-10, nor on a Ubuntu Xenial with g++-8. I tried using the exact same compilation arguments passed to g++ and running...

Although I could reproduce your bug on the same machine linking against your install of OpenBLAS, I then built a fresh OpenBLAS from upstream `develop` branch. I suspect that your...