Vector_modn_dense used for too large a modulus?

[mezzarobba]

sage.rings.finite_rings.integer_mod? says that IntegerMod_int64 "is used if the modulus is less than 2^{31}-1".

Consider however a vector of elements of GF(p) with p between 2³¹ and 2³²:

sage: K = GF(3037000493)
sage: v = vector(K, [42])
sage: type(v)
<type 'sage.modules.vector_modn_dense.Vector_modn_dense'>

Then:

sage: type(v[0])
<type 'sage.rings.finite_rings.integer_mod.IntegerMod_int64'>

which breaks some operations:

sage: ~v[0]
---------------------------------------------------------------------------
ZeroDivisionError                         Traceback (most recent call last)
<ipython-input-5-43d86000256e> in <module>()
----> 1 ~v[Integer(0)]

/home/marc/co/sage/local/lib/python2.7/site-packages/sage/rings/finite_rings/integer_mod.so in sage.rings.finite_rings.integer_mod.IntegerMod_int64.__invert__ (sage/rings/finite_rings/integer_mod.c:29992)()

/home/marc/co/sage/local/lib/python2.7/site-packages/sage/rings/finite_rings/integer_mod.so in sage.rings.finite_rings.integer_mod.mod_inverse_int64 (sage/rings/finite_rings/integer_mod.c:30920)()

ZeroDivisionError: Inverse does not exist.

This does not happens with modular integers themselves:

sage: type(K(42))
<type 'sage.rings.finite_rings.integer_mod.IntegerMod_gmp'>
sage: ~K(42)
2241595602

Matrices do not seem affected either.

Things work fine with sage-5.8 on the same machine. This makes me suspect that the issue might have been introduced in aa83d6e (#14627)—but I didn't check.

Component: linear algebra

Author: Volker Braun

Reviewer: Marc Mezzarobba

Merged: sage-5.12.rc0

Issue created by migration from https://trac.sagemath.org/ticket/15220

[vbraun]

Initial patch

[vbraun]

comment:1

Attachment: trac_15220_max_modulus.patch.gz

[vbraun]

Author: Volker Braun

[vbraun]

comment:2

Yes, this is from #14627. Would have been nice if there would have been doctests...

I think its pretty trivial fix so we should get that into the next Sage release.

[mezzarobba]

comment:3

Looks good to me--and solves my problem. Thanks!

[mezzarobba]

Reviewer: Marc Mezzarobba

[jdemeyer]

Merged: sage-5.12.rc0