Prepare for autoconf-2.71 #30668
Comments
|
comment:1
Well, fortunately end users do not run autoconf. |
|
comment:2
Lots of people clone the git repo and run |
|
comment:3
I think it's a good idea to check whether we can point out some issued, before the release is final. |
|
comment:4
Bootstrap output (cleaned up): |
|
comment:5
AC_TRY_LINK comes from ppl.m4, and has an easy replacement. Reported upstream at https://www.cs.unipr.it/mantis/view.php?id=2751 |
|
comment:6
AC_TRY_RUN, too: https://www.cs.unipr.it/mantis/view.php?id=2752 |
|
comment:7
AC_FOREACH: http://savannah.gnu.org/patch/index.php?9983 |
|
comment:9
Dropping the backslashes that it complains about in diff --git a/m4/sage_spkg_collect.m4 b/m4/sage_spkg_collect.m4
index 7cba7252c4..3867bc16c9 100644
--- a/m4/sage_spkg_collect.m4
+++ b/m4/sage_spkg_collect.m4
@@ -183,7 +183,7 @@ for DIR in $SAGE_ROOT/build/pkgs/*; do
multiple installation records for $SPKG_NAME:
m4_newline($(ls -l "$SAGE_SPKG_INST/$SPKG_NAME"-*))
m4_newline([only one should exist, so please delete some or all
- of these files and re-run \"$srcdir/configure\"])
+ of these files and re-run "$srcdir/configure"])
]))
])
stampfile=yes |
|
comment:10
Setting new milestone based on a cursory review of ticket status, priority, and last modification date. |
|
comment:11
on Gentoo, autoconf is now version 2.71. Running |
|
Commit: |
|
New commits:
|
|
Branch: u/dimpase/configure/autoconf271 |
|
comment:13
I think that leaves only the obsolete with since we already call But that would skip the C99 check for people using older autoconf. Should we...
|
|
comment:14
Replying to @orlitzky:
No, that would be too disruptive |
|
Author: Dima Pasechnik, Michael Orlitzky |
|
comment:15
This looks sufficient for both autoconf-2.69 and autoconf-2.71. New commits:
|
|
Changed branch from u/dimpase/configure/autoconf271 to u/mjo/ticket/30668 |
|
comment:16
As you could check, can be removed, as nothing sets |
|
comment:17
Replying to @dimpase:
Of course I haven't tested on a system with a brand-new autoconf and a twenty-year-old gcc, but the documentation for >=autoconf-2.70 says that https://www.gnu.org/software/autoconf/manual/autoconf-2.70/html_node/C-Compiler.html |
|
comment:42
You can unmask autoconf-2.71 on Gentoo by adding the line I am sceptical of the unreported race condition, but have too many other things to worry about at the moment. |
|
Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:
|
|
comment:44
I've removed all the changes to ax_* files - except the one that was updated upstream, so I included this update here. |
|
comment:46
weirdly, feels like wrong quoting somewhere? As well: |
|
comment:47
the latter is fixed by I'll open another ticket for this. |
|
comment:48
Replying to @dimpase:
fixed by #31871 |
|
comment:49
please merge this soon, a number of distros use autoconf 2.70+ now, and it breaks CI stuff not to support it properly. |
|
comment:50
Made it a blocker, as I am tired of seeing every time I run also, |
This comment has been minimized.
This comment has been minimized.
|
Changed author from Dima Pasechnik, Michael Orlitzky to Dima Pasechnik, Michael Orlitzky |
|
Changed branch from u/dimpase/ticket/30668 to |
|
comment:54
Thanks, Volker. Were "typos" (well, unprintable chars - thanks Michael for spotting these) in the list of authors tripping up your scripts? |
|
Changed commit from |
Autoconf is finally getting an update after 8 years, and there are some breaking changes. A beta version has been released for testing:
https://lists.gnu.org/archive/html/autoconf/2020-09/msg00006.html
We should try to bootstrap sage with the new version to make sure everything works before any end-users get a chance to.
This ticket also unbreaks make targets such as
distclean, which rely on normal return from./bootstrap, for these on autoconf 2.71Upstream: Not yet reported upstream; Will do shortly.
CC: @mkoeppe @kiwifb @dimpase @vbraun
Component: build
Author: Dima Pasechnik, Michael Orlitzky
Branch:
9bfbc41Reviewer: Dima Pasechnik, Matthias Koeppe
Issue created by migration from https://trac.sagemath.org/ticket/30668
The text was updated successfully, but these errors were encountered: