public inbox for openembedded-core@lists.openembedded.org
 help / color / mirror / Atom feed
From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: martin.jansa@gmail.com, openembedded-core@lists.openembedded.org
Subject: Re: [OE-core] [PATCH] apr: fix build with autoconf-2.73
Date: Sun, 12 Apr 2026 07:01:06 +0100	[thread overview]
Message-ID: <8ffabe3414bdb206a716aa9ee5db2fc964d07e79.camel@linuxfoundation.org> (raw)
In-Reply-To: <20260410134915.3936107-1-martin.jansa@gmail.com>

On Fri, 2026-04-10 at 15:49 +0200, Martin Jansa via lists.openembedded.org wrote:
> From: Martin Jansa <martin.jansa@gmail.com>
> 
> On hosts with gcc-13 it was trying to use -std=gnu23 and failing
> with unrecognized command-line option (gnu23 needs gcc-14 and newer)
> 
> Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
> ---
>  ...rog_cc_for_build.m4-update-to-latest.patch | 189 ++++++++++++++++++
>  meta/recipes-support/apr/apr_1.7.6.bb         |   1 +
>  2 files changed, 190 insertions(+)
>  create mode 100644 meta/recipes-support/apr/apr/0001-ax_prog_cc_for_build.m4-update-to-latest.patch
> 
> diff --git a/meta/recipes-support/apr/apr/0001-ax_prog_cc_for_build.m4-update-to-latest.patch b/meta/recipes-support/apr/apr/0001-ax_prog_cc_for_build.m4-update-to-latest.patch
> new file mode 100644
> index 0000000000..2607f6f860
> --- /dev/null
> +++ b/meta/recipes-support/apr/apr/0001-ax_prog_cc_for_build.m4-update-to-latest.patch
> @@ -0,0 +1,189 @@
> +From 24eb9fd3469e610b6887517a80b8e038c200e807 Mon Sep 17 00:00:00 2001
> +From: Martin Jansa <martin.jansa@gmail.com>
> +Date: Fri, 10 Apr 2026 15:04:43 +0200
> +Subject: [PATCH] ax_prog_cc_for_build.m4: update to latest
> +MIME-Version: 1.0
> +Content-Type: text/plain; charset=UTF-8
> +Content-Transfer-Encoding: 8bit
> +
> +After the autoconf update to 2.73 this was using -std=gnu23
> +even on hosts with gcc-13 which doesn't support it:
> +
> +ac_cv_prog_ac_ct_CC_FOR_BUILD='gcc '
> +CC_FOR_BUILD='gcc  -std=gnu23'
> +ac_ct_CC_FOR_BUILD='gcc '
> +
> +Leading to:
> +gcc: error: unrecognized command-line option ‘-std=gnu23’; did you mean ‘-std=gnu2x’?
> +
> +Needs the updated ac_cv_prog_cc_c23 from:
> +https://github.com/autoconf-archive/autoconf-archive/commit/8a970ce96721f516fef4226e5eca8da341159765
> +
> +Upstream-Status: Submitted [https://github.com/apache/apr/pull/72]
> +Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
> +---
> + build/ax_prog_cc_for_build.m4 | 99 ++++++++++++++++++++++++++---------
> + 1 file changed, 74 insertions(+), 25 deletions(-)
> +
> +diff --git a/build/ax_prog_cc_for_build.m4 b/build/ax_prog_cc_for_build.m4
> +index 26daf36..4d1de99 100644
> +--- a/build/ax_prog_cc_for_build.m4
> ++++ b/build/ax_prog_cc_for_build.m4
> +@@ -32,28 +32,32 @@
> + #   and this notice are preserved. This file is offered as-is, without any
> + #   warranty.
> + 
> +-#serial 9
> ++#serial 26
> + 
> + AU_ALIAS([AC_PROG_CC_FOR_BUILD], [AX_PROG_CC_FOR_BUILD])
> + AC_DEFUN([AX_PROG_CC_FOR_BUILD], [dnl
> + AC_REQUIRE([AC_PROG_CC])dnl
> + AC_REQUIRE([AC_PROG_CPP])dnl
> +-AC_REQUIRE([AC_EXEEXT])dnl
> +-AC_REQUIRE([AC_CANONICAL_HOST])dnl
> ++AC_REQUIRE([AC_CANONICAL_BUILD])dnl
> + 
> + dnl Use the standard macros, but make them use other variable names
> + dnl
> + pushdef([ac_cv_prog_CPP], ac_cv_build_prog_CPP)dnl
> + pushdef([ac_cv_prog_gcc], ac_cv_build_prog_gcc)dnl
> ++pushdef([ac_cv_prog_cc_c89], ac_cv_build_prog_cc_c89)dnl
> ++pushdef([ac_cv_prog_cc_c99], ac_cv_build_prog_cc_c99)dnl
> ++pushdef([ac_cv_prog_cc_c11], ac_cv_build_prog_cc_c11)dnl
> ++pushdef([ac_cv_prog_cc_c23], ac_cv_build_prog_cc_c23)dnl
> ++pushdef([ac_cv_prog_cc_stdc], ac_cv_build_prog_cc_stdc)dnl
> + pushdef([ac_cv_prog_cc_works], ac_cv_build_prog_cc_works)dnl
> + pushdef([ac_cv_prog_cc_cross], ac_cv_build_prog_cc_cross)dnl
> + pushdef([ac_cv_prog_cc_g], ac_cv_build_prog_cc_g)dnl
> +-pushdef([ac_cv_exeext], ac_cv_build_exeext)dnl
> +-pushdef([ac_cv_objext], ac_cv_build_objext)dnl
> ++pushdef([ac_prog_cc_stdc], ac_build_prog_cc_stdc)dnl
> + pushdef([ac_exeext], ac_build_exeext)dnl
> + pushdef([ac_objext], ac_build_objext)dnl
> + pushdef([CC], CC_FOR_BUILD)dnl
> + pushdef([CPP], CPP_FOR_BUILD)dnl
> ++pushdef([GCC], GCC_FOR_BUILD)dnl
> + pushdef([CFLAGS], CFLAGS_FOR_BUILD)dnl
> + pushdef([CPPFLAGS], CPPFLAGS_FOR_BUILD)dnl
> + pushdef([LDFLAGS], LDFLAGS_FOR_BUILD)dnl
> +@@ -67,27 +71,58 @@ pushdef([ac_cv_host_alias], ac_cv_build_alias)dnl
> + pushdef([ac_cv_host_cpu], ac_cv_build_cpu)dnl
> + pushdef([ac_cv_host_vendor], ac_cv_build_vendor)dnl
> + pushdef([ac_cv_host_os], ac_cv_build_os)dnl
> +-pushdef([ac_cpp], ac_build_cpp)dnl
> +-pushdef([ac_compile], ac_build_compile)dnl
> +-pushdef([ac_link], ac_build_link)dnl
> ++pushdef([ac_tool_prefix], ac_build_tool_prefix)dnl
> ++pushdef([am_cv_CC_dependencies_compiler_type], am_cv_build_CC_dependencies_compiler_type)dnl
> ++pushdef([am_cv_prog_cc_c_o], am_cv_build_prog_cc_c_o)dnl
> ++pushdef([cross_compiling], cross_compiling_build)dnl
> ++dnl
> ++dnl These variables are problematic to rename by M4 macros, so we save
> ++dnl their values in alternative names, and restore the values later.
> ++dnl
> ++dnl _AC_COMPILER_EXEEXT and _AC_COMPILER_OBJEXT internally call
> ++dnl AC_SUBST which prevents the renaming of EXEEXT and OBJEXT
> ++dnl variables. It's not a good idea to rename ac_cv_exeext and
> ++dnl ac_cv_objext either as they're related.
> ++dnl Renaming ac_exeext and ac_objext is safe though.
> ++dnl
> ++ac_cv_host_exeext=$ac_cv_exeext
> ++AS_VAR_SET_IF([ac_cv_build_exeext],
> ++  [ac_cv_exeext=$ac_cv_build_exeext],
> ++  [AS_UNSET([ac_cv_exeext])])
> ++ac_cv_host_objext=$ac_cv_objext
> ++AS_VAR_SET_IF([ac_cv_build_objext],
> ++  [ac_cv_objext=$ac_cv_build_objext],
> ++  [AS_UNSET([ac_cv_objext])])
> ++dnl
> ++dnl ac_cv_c_compiler_gnu is used in _AC_LANG_COMPILER_GNU (called by
> ++dnl AC_PROG_CC) indirectly.
> ++dnl
> ++ac_cv_host_c_compiler_gnu=$ac_cv_c_compiler_gnu
> ++AS_VAR_SET_IF([ac_cv_build_c_compiler_gnu],
> ++  [ac_cv_c_compiler_gnu=$ac_cv_build_c_compiler_gnu],
> ++  [AS_UNSET([ac_cv_c_compiler_gnu])])
> + 
> +-save_cross_compiling=$cross_compiling
> +-save_ac_tool_prefix=$ac_tool_prefix
> +-cross_compiling=no
> +-ac_tool_prefix=
> ++cross_compiling_build=no
> + 
> ++ac_build_tool_prefix=
> ++AS_IF([test -n "$build"],      [ac_build_tool_prefix="$build-"],
> ++      [test -n "$build_alias"],[ac_build_tool_prefix="$build_alias-"])
> ++
> ++AC_LANG_PUSH([C])
> + AC_PROG_CC
> ++_AC_COMPILER_EXEEXT
> ++_AC_COMPILER_OBJEXT
> + AC_PROG_CPP
> +-AC_EXEEXT
> + 
> +-ac_tool_prefix=$save_ac_tool_prefix
> +-cross_compiling=$save_cross_compiling
> ++BUILD_EXEEXT=$ac_cv_exeext
> ++BUILD_OBJEXT=$ac_cv_objext
> + 
> + dnl Restore the old definitions
> + dnl
> +-popdef([ac_link])dnl
> +-popdef([ac_compile])dnl
> +-popdef([ac_cpp])dnl
> ++popdef([cross_compiling])dnl
> ++popdef([am_cv_prog_cc_c_o])dnl
> ++popdef([am_cv_CC_dependencies_compiler_type])dnl
> ++popdef([ac_tool_prefix])dnl
> + popdef([ac_cv_host_os])dnl
> + popdef([ac_cv_host_vendor])dnl
> + popdef([ac_cv_host_cpu])dnl
> +@@ -101,26 +136,40 @@ popdef([host])dnl
> + popdef([LDFLAGS])dnl
> + popdef([CPPFLAGS])dnl
> + popdef([CFLAGS])dnl
> ++popdef([GCC])dnl
> + popdef([CPP])dnl
> + popdef([CC])dnl
> + popdef([ac_objext])dnl
> + popdef([ac_exeext])dnl
> +-popdef([ac_cv_objext])dnl
> +-popdef([ac_cv_exeext])dnl
> ++popdef([ac_prog_cc_stdc])dnl
> + popdef([ac_cv_prog_cc_g])dnl
> + popdef([ac_cv_prog_cc_cross])dnl
> + popdef([ac_cv_prog_cc_works])dnl
> ++popdef([ac_cv_prog_cc_stdc])dnl
> ++popdef([ac_cv_prog_cc_c23])dnl
> ++popdef([ac_cv_prog_cc_c11])dnl
> ++popdef([ac_cv_prog_cc_c99])dnl
> ++popdef([ac_cv_prog_cc_c89])dnl
> + popdef([ac_cv_prog_gcc])dnl
> + popdef([ac_cv_prog_CPP])dnl
> ++dnl
> ++ac_cv_exeext=$ac_cv_host_exeext
> ++EXEEXT=$ac_cv_host_exeext
> ++ac_cv_objext=$ac_cv_host_objext
> ++OBJEXT=$ac_cv_host_objext
> ++ac_cv_c_compiler_gnu=$ac_cv_host_c_compiler_gnu
> ++ac_compiler_gnu=$ac_cv_host_c_compiler_gnu
> ++
> ++dnl restore global variables ac_ext, ac_cpp, ac_compile,
> ++dnl ac_link, ac_compiler_gnu (dependent on the current
> ++dnl language after popping):
> ++AC_LANG_POP([C])
> + 
> + dnl Finally, set Makefile variables
> + dnl
> +-BUILD_EXEEXT=$ac_build_exeext
> +-BUILD_OBJEXT=$ac_build_objext
> +-AC_SUBST(BUILD_EXEEXT)dnl
> +-AC_SUBST(BUILD_OBJEXT)dnl
> ++AC_SUBST([BUILD_EXEEXT])dnl
> ++AC_SUBST([BUILD_OBJEXT])dnl
> + AC_SUBST([CFLAGS_FOR_BUILD])dnl
> + AC_SUBST([CPPFLAGS_FOR_BUILD])dnl
> + AC_SUBST([LDFLAGS_FOR_BUILD])dnl
> + ])
> +-
> diff --git a/meta/recipes-support/apr/apr_1.7.6.bb b/meta/recipes-support/apr/apr_1.7.6.bb
> index 13fa5daa3c..86fac1e4c8 100644
> --- a/meta/recipes-support/apr/apr_1.7.6.bb
> +++ b/meta/recipes-support/apr/apr_1.7.6.bb
> @@ -22,6 +22,7 @@ SRC_URI = "${APACHE_MIRROR}/apr/${BPN}-${PV}.tar.bz2 \
>             file://0001-configure-Remove-runtime-test-for-mmap-that-can-map-.patch \
>             file://autoconf-2.73.patch \
>             file://0001-dso-Check-for-NULL-handle-in-apr_dso_sym.patch \
> +           file://0001-ax_prog_cc_for_build.m4-update-to-latest.patch \
>             "
>  
>  SRC_URI[sha256sum] = "49030d92d2575da735791b496dc322f3ce5cff9494779ba8cc28c7f46c5deb32"

I think this caused a reproducibility issue:

https://valkyrie.yocto.io/pub/repro-fail/oe-reproducible-20260412-a5w4mcrk/packages/diff-html/

Cheers,

Richard


  reply	other threads:[~2026-04-12  6:01 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-10 13:49 [PATCH] apr: fix build with autoconf-2.73 martin.jansa
2026-04-12  6:01 ` Richard Purdie [this message]
2026-04-12  7:21   ` [OE-core] " Khem Raj
2026-04-12 11:17     ` Martin Jansa
2026-04-12 17:55       ` Khem Raj

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8ffabe3414bdb206a716aa9ee5db2fc964d07e79.camel@linuxfoundation.org \
    --to=richard.purdie@linuxfoundation.org \
    --cc=martin.jansa@gmail.com \
    --cc=openembedded-core@lists.openembedded.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox