givaro spkg-configure: Reject too new versions

[mkoeppe]

The linbox team has recently released new versions of their packages: givaro-4.2.0, fflas-ffpack-2.5.0, linbox-1.7.0, which some cutting-edge distributions are already shipping: https://repology.org/project/givaro/versions

Unfortunately our spkg-configure for givaro only checks a lower version bound, not an upper bound. We do not have an spkg-configure for linbox - that's #29631; so we try to build an old linbox with a new givaro, which fails (upstream does not support this).
Hence these new, incompatible releases retroactively break released Sage versions.

This leads to build errors in our linbox spkg, such as

• https://groups.google.com/g/sage-devel/c/seTJ79k33Ac/m/wr4kbNAhDQAJ
• https://groups.google.com/g/sage-release/c/3b_sh9htMg0/m/ft4DZsmpCwAJ

A workaround for users of released versions of Sage is to use ./configure --without-system-givaro.

Depends on #32947

CC: @dimpase @orlitzky @ClementPernet @kiwifb @EmmanuelCharpentier @antonio-rojas

Component: build: configure

Author: Matthias Koeppe

Branch/Commit: 1da5d31

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/33042

[mkoeppe]

Author: Matthias Koeppe

[mkoeppe]

Description changed:

--- 
+++ 
@@ -3,7 +3,8 @@
 Unfortunately our spkg-configure for `givaro` only checks a lower version bound, not an upper bound. So these new, incompatible releases retroactively break released Sage versions.
 
 This leads to build errors in our `linbox` spkg, such as  
-https://groups.google.com/g/sage-devel/c/seTJ79k33Ac/m/wr4kbNAhDQAJ (likely).
+- https://groups.google.com/g/sage-devel/c/seTJ79k33Ac/m/wr4kbNAhDQAJ
+- https://groups.google.com/g/sage-release/c/3b_sh9htMg0/m/ft4DZsmpCwAJ
 
 
 (Note that we do not have an `spkg-configure` for `linbox` - that's #29631).

[mkoeppe]

Branch: u/mkoeppe/givaro_spkg_configure__reject_too_new_versions

[mkoeppe]

Dependencies: #32947

[mkoeppe]

New commits:

d2d73dc

build/pkgs/givaro/spkg-configure.m4: Reject givaro >= 4.2.0

[mkoeppe]

Commit: d2d73dc

[sagetrac-git]

Changed commit from d2d73dc to f8d5fa2

[sagetrac-git]

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

ab4fb99	tox.ini: Add ubuntu-jammy, debian-bookworm, linuxmint-20.3, fedora-36
e51225a	.github/workflows/tox*.yml: Remove debian-jessie, add ubuntu-jammy, debian-bookworm, linuxmint-20.3, fedora-36
ac47e7d	.github/workflows/tox-gcc_spkg.yml: Remove
f8d5fa2	Merge #32947

[mkoeppe]

Description changed:

--- 
+++ 
@@ -1,11 +1,9 @@
 The linbox team has recently released new versions of their packages: givaro-4.2.0, fflas-ffpack-2.5.0, linbox-1.7.0, which some cutting-edge distributions are already shipping: https://repology.org/project/givaro/versions
 
-Unfortunately our spkg-configure for `givaro` only checks a lower version bound, not an upper bound. So these new, incompatible releases retroactively break released Sage versions.
+Unfortunately our spkg-configure for `givaro` only checks a lower version bound, not an upper bound. We do not have an `spkg-configure` for `linbox` - that's #29631; so we try to build an old `linbox` with a new `givaro`, which fails (upstream does not support this).
+Hence these new, incompatible releases retroactively break released Sage versions.
 
 This leads to build errors in our `linbox` spkg, such as  
 - https://groups.google.com/g/sage-devel/c/seTJ79k33Ac/m/wr4kbNAhDQAJ
 - https://groups.google.com/g/sage-release/c/3b_sh9htMg0/m/ft4DZsmpCwAJ
 
-
-(Note that we do not have an `spkg-configure` for `linbox` - that's #29631).
-

[mkoeppe]

comment:7

tox -e docker-ubuntu-focal-standard -- config.status

correctly accept givaro 4.1.1

tox -e docker-debian-bookworm-standard -- config.status

correctly rejects givaro 4.2.0

[mkoeppe]

Description changed:

--- 
+++ 
@@ -7,3 +7,5 @@
 - https://groups.google.com/g/sage-devel/c/seTJ79k33Ac/m/wr4kbNAhDQAJ
 - https://groups.google.com/g/sage-release/c/3b_sh9htMg0/m/ft4DZsmpCwAJ
 
+A workaround for users of released versions of Sage is to use `./configure --without-system-givaro`.
+

[mkoeppe]

comment:10

Let's get this in please.

[dimpase]

comment:11

+    m4_pushdef([SAGE_GIVARO_LTVER])
     m4_popdef([SAGE_GIVARO_MINVER])

That should be popdef,
obviously.