GAP fails its own test on SPARC Solaris 11

[dimpase]

running the usual "testinstall.tst" for GAP fails here:

gap> HashKeyBag([1,2,3],0,1,-1) <> HashKeyWholeBag([1,2,3],0);
Bus Error (core dumped)

See gap-system/gap#2191

Alignment problem, see
gap-system/gap#2196

Upstream: Fixed upstream, in a later stable release.

CC: @jdemeyer @dimpase

Component: porting: Solaris

Author: Dima Pasechnik

Branch/Commit: u/dimpase/sparc/gap_align @ b9179e9

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/24765

[dimpase]

Author: Dima Pasechnik

[dimpase]

Changed upstream from Reported upstream. No feedback yet. to Fixed upstream, in a later stable release.

[dimpase]

comment:1

see gap-system/gap#2195 for a fix

[dimpase]

comment:2

there are more issues:

gap-system/gap#2193
and
gap-system/gap#2194

[dimpase]

Commit: b9179e9

[dimpase]

New commits:

b9179e9

backport of alignment fix for (lib)gap

[dimpase]

Branch: u/dimpase/sparc/gap_align

[dimpase]

comment:4

With this patch, on SPARC Solaris tests in src/sage/interfaces/gap_workspace.py, src/sage/interfaces/gap.py and src/sage/libs/gap/ pass.

[dimpase]

comment:5

This does not solve all of GAP's problems on SPARC Solaris, but it at least gets rid of segfaults.

[mkoeppe]

comment:6

Outdated, should be closed

[dimpase]

Reviewer: Dima Pasechnik

[dimpase]

comment:9

we now have access to Solaris 11.4 (on Sparc), if anyone wants to give it a try...