Re: bug#77800: grep-3.12: write-error-msg test failure on fedora rawhide (f43)

On Thu, Apr 17, 2025 at 11:15 PM Grisha Levit wrote: > > On Fri, Apr 18, 2025 at 1:51 AM Jim Meyering wrote: > > > > Surprised to find that coreutils-9.5 (fedora 41 stock) works fine: > > > > $ { /bin/printf %4095s; /bin/printf %4096s; } > /dev/full > > /bin/printf: write error: No space left

Re: bug#77800: grep-3.12: write-error-msg test failure on fedora rawhide (f43)

On Tue, Apr 15, 2025 at 9:17 PM Jim Meyering wrote: ... > We're going to have to revise that code. > The difference I see is that before rawhide, that fclose would fail. > It's perfectly fine for fclose to succeed in this case, as now happens > on rawhide (because with 4k BUFSIZ, the fclose wrote

Re: [PATCH 2/5] Pacify GCC 15’s new -Wtrailing-whitespace option

Hi Paul, Paul Eggert writes: > @@ -82,7 +82,7 @@ AC_DEFUN([AM_ICONV_LINK] > LIBS="$LIBS $LIBICONV" >fi >am_cv_func_iconv_works=no > - for ac_iconv_const in '' 'const'; do > + for ac_iconv_const in '/*empty*/' 'const'; do > AC_RUN_IFELSE( >[

Re: bug#77800: grep-3.12: write-error-msg test failure on fedora rawhide (f43)

On Thu, Apr 17, 2025 at 10:35 PM Grisha Levit wrote: > On Fri, Apr 18, 2025, 01:18 Jim Meyering wrote: > > > > On Tue, Apr 15, 2025 at 9:17 PM Jim Meyering wrote: > > ... > > > We're going to have to revise that code. > > > The difference I see is that before rawhide, that fclose would fail. > >

[PATCH] Update some POSIX.1-2004 citations

--- doc/pastposix-functions/usleep.texi | 2 +- doc/posix-headers/sys_timeb.texi| 3 ++- doc/posix-headers/ucontext.texi | 3 ++- lib/unistd.in.h | 2 +- 4 files changed, 6 insertions(+), 4 deletions(-) diff --git a/doc/pastposix-functions/usleep.texi b/doc/pastposix-

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On 2025-04-17 09:46, Simon Josefsson wrote: I don't think it's worth making a distinction here between Autoconf and "Autoconf + Automake", since most packages that use Autoconf also use Automake. I would appreciate distinguishing this, it helps to make it more clear where dependencies are coming

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On 2025-04-17 04:31, Simon Josefsson wrote: Paul Eggert writes: [ awk cat cmp cp diff echo expr false grep ls mkdir mv printf pwd rm rmdir sed sort test touch tr true Of those, I'm guessing awk, grep, sed and ls are the most complex. The 'diff' tool also stands out, and most of my packages b

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On 4/17/25 10:14 AM, Eric Blake wrote: > On Thu, Apr 17, 2025 at 02:28:36PM +0200, Bruno Haible via Bug reports for > autoconf wrote: >> Jeffrey Walton wrote: >>> Awk is a standard Posix utility: >>> . >> >> The newest POSIX is P

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Simon Josefsson wrote: > the docker container images (which are typically > used for CI) or the cloud images (which are typically used in Kubernetes > environments). I'm not sure there is a trend that container images for > OS's are growing, I think most distributions are actively trying to make >

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Bruno Haible via Bug reports for autoconf writes: > Simon Josefsson wrote: >> 0) Modify autoconf to continue to work in this situation without awk, >> replacing it with more POSIX shell or something else? > > Many years ago, Autoconf produced configure scripts that used 'sed' > for the job of

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On Thu, Apr 17, 2025 at 02:28:36PM +0200, Bruno Haible via Bug reports for autoconf wrote: > Jeffrey Walton wrote: > > Awk is a standard Posix utility: > > . > > The newest POSIX is POSIX:2024. Update your URLs: >

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Simon Josefsson wrote: > It seems awk is going away as a standard tool, and it is possible to > either fight that or just accept it. I don't think that the set of "standard tools" is shrinking. Rather, it's expanding. On OpenBSD for example, perl is installed by default. Similarly, Python is consi

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Jeffrey Walton wrote: > Awk is a standard Posix utility: > . The newest POSIX is POSIX:2024. Update your URLs: . Bruno

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Simon Josefsson wrote: > 0) Modify autoconf to continue to work in this situation without awk, > replacing it with more POSIX shell or something else? Many years ago, Autoconf produced configure scripts that used 'sed' for the job of replacing the various @FOO@ occurrences in *.in files. When

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On Thu, Apr 17, 2025 at 2:33 AM Simon Josefsson via Gnulib discussion list wrote: > > I got a CI/CD build failure [1] for libidn on the new release fedora 42. > > checking that generated files are newer than configure... done > configure: creating ./config.status > config.status: creating csharpco

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Simon Josefsson wrote: > I got a CI/CD build failure [1] for libidn on the new release fedora 42. > > checking that generated files are newer than configure... done > configure: creating ./config.status > config.status: creating csharpcomp.sh > ./config.status: line 2711: awk: command not found A

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

Paul Eggert writes: > On 2025-04-16 23:31, Simon Josefsson via Bug reports for autoconf wrote: > >> I tried reading through the autoconf manual to see if 'awk' is a >> run-time dependency for running generated ./configure scripts > > It is a dependency, and the documentation should mention this

Re: fedora 42 doesn't have awk: how to deal with autoconf subst?

On 2025-04-16 23:31, Simon Josefsson via Bug reports for autoconf wrote: I tried reading through the autoconf manual to see if 'awk' is a run-time dependency for running generated ./configure scripts It is a dependency, and the documentation should mention this. I installed the attached patch

[PATCH 3/5] getlogin-tests: remove xalloc dependency

* modules/getlogin-tests (Depends-on): Remove xalloc. (test_getlogin_LDADD): Remove @LIBINTL@. * tests/test-getlogin.c: Do not include xalloc.h. (main): Use a static rather than heap storage. --- ChangeLog | 6 ++ modules/getlogin-tests | 3 +-- tests/test-getlogin.c | 7 +++

[PATCH 1/5] manywarnings: update C warnings for GCC 15

Adjust for C programs compiled by GCC 15. This is merely changes to commentary; no changes to the flags passed to GCC. * build-aux/gcc-warning.spec: Add warnings introduced in GCC 15. Sort for LC_ALL=C, so that the update procedure is reproducible. * m4/manywarnings.m4: Specify LC_ALL=C in comment.

[PATCH 5/5] memset_explicit-tests: pacify -Wuse-after-free

* tests/test-memset_explicit.c: Ignore -Wuse-after-free, since the test deliberately uses after free. --- ChangeLog| 4 tests/test-memset_explicit.c | 4 2 files changed, 8 insertions(+) diff --git a/ChangeLog b/ChangeLog index 7a5586a8d7..6df7b09a1e 100644 --- a/Cha

[PATCH 4/5] Pacify -Wunterminated-string-initialization

* tests/test-fwrite.c (main): Mark nonstrings with _GL_ATTRIBUTE_NONSTRING. * tests/test-mbsinit.c (main): "\303" → {'\303'} in initializer. --- ChangeLog| 5 + tests/test-fwrite.c | 6 +++--- tests/test-mbsinit.c | 2 +- 3 files changed, 9 insertions(+), 4 deletions(-) diff --gi

[PATCH 2/5] Pacify GCC 15’s new -Wtrailing-whitespace option

* m4/builtin-expect.m4 (gl___BUILTIN_EXPECT): * m4/iconv.m4 (AM_ICONV_LINK, AM_ICONV): Avoid whitespace at line end in C code. * modules/unitypes-h (unitypes_restrict): Likewise, bearing in mind that configure.ac lines will be indented so such lines containing only '])' should be avoided when they

fedora 42 doesn't have awk: how to deal with autoconf subst?

Hi I got a CI/CD build failure [1] for libidn on the new release fedora 42. checking that generated files are newer than configure... done configure: creating ./config.status config.status: creating csharpcomp.sh ./config.status: line 2711: awk: command not found config.status: error: could not c