[R] Compilation problems with R4.4.0

[Ivan Krylov]

Glad you got R working!

В Thu, 9 May 2024 12:19:49 +0000
Simon Andrews <simon.andrews using babraham.ac.uk> пишет:

> Does this need reporting upstream as a bug in R or lapack?

Maybe in AlmaLinux (and I think RHEL9 too, but that's up to Miguel
Esteva). Best case scenario, they somehow built the reference BLAS and
LAPACK packages with the wrong Fortran compiler version or flags, which
is why gfortran-11 gets a zero instead of the return value from DOUBLE
COMPLEX zdotu(...) and crashes when calling dgesv(...) from
lapack-3.9.0-10.el9.x86_64. Worst case scenario, the zdotu() problem
and the crash have two completely different causes elsewhere.

-- 
Best regards,
Ivan