Paul Eggert wrote:
> Come to think of it, that new use of AC_REQUIRE([AC_CANONICAL_HOST]) at
> gzip/configure.ac's is problematic.
Yes, I agree.
> I guess it works because Gnulib has
> already called AC_CANONICAL_HOST by then, but we shouldn't assume that
> without checking.
Maybe a better way would have been to wrap this code in an AC_DEFUN
and invoke it just immediately afterwards. This way, one can still use
AC_REQUIRE.
> I worked around the issue by instead checking that
> Gnulib has called AC_CANONICAL_HOST, and by using AS_CASE rather than
> 'case' while I was in the neighborhood.
Unfortunately, there's a typo here: The output in the configure file
looks like this. Note the "Avoid a crash" line:
case $DEFS in #(
*NO_ASM*) :
;; #(
*) :
# 'host_os?' ensures Gnulib sets host_os as usual.
case ${host_os?} in #(
*-musl* | midipix*) :
Avoid a crash in musl libc's startup code,
# because musl doesn't support TEXTRELs.
# https://www.openwall.com/lists/musl/2020/09/25/1
;; #(
*) :
if cp $srcdir/lib/match.c _match.S &&
Here's a fix.
>From 7927b8446cf5cb5a628dc32e045da38d979715e5 Mon Sep 17 00:00:00 2001
From: Bruno Haible <[email protected]>
Date: Sun, 19 Apr 2026 09:16:18 +0200
Subject: [PATCH] build: Fix syntax error in configure (regression yesterday)
* configure.ac: Fix typo in yesterday's commit.
---
configure.ac | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 4f69098..ca3126e 100644
--- a/configure.ac
+++ b/configure.ac
@@ -253,7 +253,7 @@ AC_CACHE_CHECK([for an assembler syntax supported by this package],
[# 'host_os?' ensures Gnulib sets host_os as usual.
AS_CASE([${host_os?}],
[*-musl* | midipix*],
- [Avoid a crash in musl libc's startup code,
+ [# Avoid a crash in musl libc's startup code,
# because musl doesn't support TEXTRELs.
# https://www.openwall.com/lists/musl/2020/09/25/1
],
--
2.52.0
