While working on the implementation of the matrix exponential expm, I ended up having to rewrite some of the handling functions for the lapack errors already written somewhere but not exposed.
Following the discussion here, I think grouping them all in one place (stdlib_linalg_lapack_aux seems a good choice) would make sense and avoid additional code duplication when re-using bits and pieces for other higher-level algorithms.
@perazz, @jvdp1, @jalvesz