1 | initial version |
This is definitely a bug, thanks for reporting, it is now trac ticket 30702.
As for tracking a minimal example, it seems that when a finite field uses givaro
as a backend, it is possible to deal with polynomial rings in a single variable (generic Sage implementation) :
sage: F = GF(3, impl='givaro')
sage: F
Finite Field of size 3
sage: type(F)
<class 'sage.rings.finite_rings.finite_field_givaro.FiniteField_givaro_with_category'>
sage: R.<a0> = PolynomialRing(F)
sage: type(R)
<class 'sage.rings.polynomial.polynomial_ring.PolynomialRing_dense_finite_field_with_category'>
sage: F(R(2))
2
But dealing with multivariate polynomial ring (whose implementation relies on libsingular
) leads to segfaults:
sage: R.<a0,a1> = PolynomialRing(F)
sage: type(R)
<class 'sage.rings.polynomial.multi_polynomial_libsingular.MPolynomialRing_libsingular'>
sage: F(R(2))
....
Segmentation fault