gr_poly_scalar_mul ?

[rburing]

This may be a bit of a corner case, but for a polynomial ring over a matrix ring one might want to multiply by a "constant" (i.e. a matrix) on the left instead of on the right (as gr_poly_mul_scalar does), see e.g. https://gitlab.inria.fr/ricardo-thomas.buring/d-finite-fun/-/blob/463d254ded0f4fc78fcae16ead856fa7e5e72be3/newton_iteration_flint.c#L258-L267

[fredrik-johansson]

There should be separate gr_poly_scalar_mul (scalar on the left) and gr_poly_mul_scalar (scalar on the right).

Likewise gr_poly_mul_other and gr_poly_other_mul.

There are some other places too where commutativity is assumed that ought to be fixed.