Disallow positive characteristic in the symbolic ring

[bgrenet]

Elements of positive characteristic in the symbolic ring are complete nonsense

sage: x = SR.var('x')
sage: f(x) = exp(GF(3).one() * x)
sage: f
x |--> e^x
sage: bool(f(x) == exp(x))
True
sage: f(3*x)
1

or

sage: solve(x^2 == 2, [x])
[x == -sqrt(2), x == sqrt(2)]
sage: solve(x^2 == GF(3)(2), [x])
[x == -sqrt(2), x == sqrt(2)]

More dramatically, it leads to segmentation faults

sage: x = polygen(GF(3))
sage: a = SR.var('a')
sage: (2*x + 1) * a
segmentation fault

Even going through conversion

sage: p = SR(2*x + 1)
sage: p * a
segmentation fault

We simply disallow wrapping of element of positive characteristic in SR.

Solve #18787 and #21391

Original reports for the segfault:

• Ask
• sage-devel

Component: commutative algebra

Keywords: polynomial, symbolics

Author: Vincent Delecroix

Branch/Commit: da8b056

Reviewer: Ralf Stephan, Travis Scrimshaw

Issue created by migration from https://trac.sagemath.org/ticket/24072

[dimpase]

comment:1

The real bug is that SR(sigma) succeeds (and returns something broken).
It should just fail, and then * will fail too, as it should.

[videlec]

Changed keywords from polynomial to polynomial, symbolics

[videlec]

Author: Vincent Delecroix

[videlec]

Commit: 148340c

[videlec]

comment:4

A branch where SR does not accept anymore positive characteristic... let see what the patchbot has to say.

---

New commits:

148340c

24072: disallow positive characteristic in SR

[videlec]

Branch: u/vdelecroix/24072

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

402865e	24072: fix doctest in matrix_space.py
8f956e0	24072: fix doctests

[sagetrac-git]

Changed commit from 148340c to 8f956e0

[videlec]

comment:7

I think I fixed all doctests!

[rwst]

Reviewer: Ralf Stephan

[rwst]

comment:10

What a relief, thanks for your work.

[sagetrac-git]

Changed commit from 8f956e0 to de4e760

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1 and set ticket back to needs_review. New commits:

de4e760

24072: py3 syntax

[videlec]

comment:12

A tiny change for py3 syntax (patchbot complaint). Setting back to positive review...

[dimpase]

comment:13

Does this mean that #21391 can now be closed as duplicate?

[vbraun]

comment:14

Merge conflict

[videlec]

comment:15

Volker, can you provide more information?

[dimpase]

comment:17

Hmm, I guess, look at https://github.com/vbraun/sage/tree/develop ?

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

32ffd59	24072: disallow positive characteristic in SR
3878ed6	24072: fix doctests

[sagetrac-git]

Changed commit from de4e760 to 3878ed6

[videlec]

comment:19

rebased on 8.1.beta9

[sagetrac-git]

Changed commit from 3878ed6 to da8b056

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

da8b056

24072: fix doctests

[videlec]

comment:21

ping?

[tscrim]

Changed reviewer from Ralf Stephan to Ralf Stephan, Travis Scrimshaw

[vbraun]

Changed branch from u/vdelecroix/24072 to da8b056