Doctests using system brial crash (segfault) on debian-sid, -bullseye

[mkoeppe]

debian-sid (https://github.com/mkoeppe/sage/runs/985563463) and debian-bullseye:

sage -t --random-seed=0 src/doc/en/constructions/algebraic_geometry.rst  # 1 doctest failed
sage -t --random-seed=0 src/doc/en/developer/coding_in_other.rst  # 1 doctest failed
sage -t --random-seed=0 src/doc/en/reference/sat/index.rst  # Killed due to segmentation fault
sage -t --random-seed=0 src/sage/crypto/mq/sr.py  # Killed due to abort
sage -t --random-seed=0 src/sage/modules/fg_pid/fgp_module.py  # 3 doctests failed
sage -t --random-seed=0 src/sage/modules/free_module_morphism.py  # 3 doctests failed
sage -t --random-seed=0 src/sage/rings/complex_arb.pyx  # 6 doctests failed
sage -t --random-seed=0 src/sage/rings/polynomial/multi_polynomial_sequence.py  # Killed due to segmentation fault
sage -t --random-seed=0 src/sage/rings/polynomial/pbori.pyx  # Killed due to segmentation fault
sage -t --random-seed=0 src/sage/rings/polynomial/polynomial_ring_constructor.py  # Killed due to abort
sage -t --random-seed=0 src/sage/rings/real_arb.pyx  # 2 doctests failed
sage -t --random-seed=0 src/sage/sat/boolean_polynomials.py  # Killed due to segmentation fault
sage -t --random-seed=0 src/sage/sat/solvers/dimacs.py  # Killed due to segmentation fault
sage -t --random-seed=0 src/sage/schemes/hyperelliptic_curves/monsky_washnitzer.py  # Killed due to abort

Per previous tickets (#29490, #29792), we avoid brial 1.2.5 for the same reason.

Here it is 1.2.8 that crashes, illustrating once more why version number based tests are not robust enough.

CC: @kiwifb @dimpase @orlitzky @tobihan @SnarkBoojum

Component: packages: standard

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/30364

[dimpase]

comment:1

cc debian people

[dimpase]

comment:2

an explicit test on 1.2.5 was rather unclear how to accomplish, thus I opted for banning a version.

[vbraun]

comment:4

Maybe we should have positive lists instead of negative lists, so by default we don't use unknown system versions? Either way, sid/bullseye are not (and will not be in the near future) released versions -> not a blocker

[mkoeppe]

comment:5

The segfaults seem to have disappeared in more recent runs (https://groups.google.com/d/msg/sage-release/hobZzw74Rv0/0CkgcJYjCgAJ), so perhaps this was just a Debian bug.

[dimpase]

Reviewer: Dima Pasechnik