https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120674
--- Comment #3 from Robin Dapp <rdapp at gcc dot gnu.org> --- I believe the issue is that we can actually return 0 for factor in riscv_dwarf_poly_indeterminate_value. Using the second coefficient is reasonable but we cannot reasonably set a vector size when there is no vector support. So while [1 1] would prevent the issue here it doesn't feel right either. Maybe we should rather recognize the situation (BYTES_PER_RISC_VECTOR.coeffs[1] == 0) in riscv_dwarf_poly_indeterminate_value and set *factor = 1?
