LFortran can now compile LAPACK, single/double precision, real/complex, 32-bit/64-bit integers, all tests pass:
Thank you @ert for getting it over the finish line!
LFortran can now compile LAPACK, single/double precision, real/complex, 32-bit/64-bit integers, all tests pass:
Thank you @ert for getting it over the finish line!
That’s a significant milestone/achievement and great news! Thanks @certik, @ert, and all other contributors.
Very good. What about the modernization, fortran-lapack?
We should tackle the fortran-lapack soon — we currently don’t support quadruple precision, but probably a good time to implement it anyway.
@FedericoPerini should we move it under fortran-lang?
the modernized LAPACK work has been integrated into the Fortran Standard Library since v0.4.0 which was released a bit over two years ago but since then, development has continued directly within stdlib, with a growing set of higher-level APIs.
Congratulations on the big milestone! It’s nice to see interest from the compiler side as well.
Perfect, then we’ll just get the latest stdlib working again, we got some older version working, but not the very latest version.