> Le 6 juil. 2022 à 07:53, Akim Demaille a écrit :
>
> Here's my updated proposal.
I had forgotten to s/test/tests/ in the commit message, sorry.
commit 813e5a1787ed156c70bd6d4bba39a8b2db4916db
Author: Akim Demaille
Date: Mon Jul 4 07:18:07 2022 +0200
gnulib-tool: add support for --
Hi Bruno,
Thanks for the feedback!
> Le 4 juil. 2022 à 23:31, Bruno Haible a écrit :
>
>> @@ -4518,7 +4526,9 @@ func_emit_initmacro_end ()
>> echo " sed_dirname4='s,\\(.\\)/[^/]*\$,\\1,'"
>> echo " sed_basename1='s,.*/,,'"
>> echo "changequote([, ])dnl"
>> - if $automake_subdir
On Tue, Jul 05, 2022 at 09:21:54PM -0500, Paul Eggert wrote:
> A couple of things. First, this doesn't seem to be glibc patch; it seems to
> be a reverse patch.
>
> Second, these days, why not just initialize to a null pointer all the time?
I think there's some confusion here. My patch is intended
On 7/5/22 21:01, Guilherme Janczak wrote:
I intend to send this file to glibc to make gnulib and glibc source the
same.
A couple of things. First, this doesn't seem to be glibc patch; it seems
to be a reverse patch.
Second, these days, why not just initialize to a null pointer all the
time?
On Wed, Jul 06, 2022 at 12:02:23AM +0200, Bruno Haible wrote:
> Since you are apparently speculating on a new compiler, please tell
> starting which version of GCC one can assume that (on ELF platforms)
>const char *x = NULL;
> produce code that is as good as
>const char *x;
My intention t
Guilherme Janczak wrote:
> I intend to send this file to glibc to make gnulib and glibc source the
> same.
Some comment / justification is missing. Ulrich Drepper, in his paper
"How to write shared libraries" [1], wrote
"The compiler will take care of avoid-
ing the initializers. To achieve t
I intend to send this file to glibc to make gnulib and glibc source the
same.
---
lib/argp-ba.c | 15 +--
1 file changed, 5 insertions(+), 10 deletions(-)
diff --git a/lib/argp-ba.c b/lib/argp-ba.c
index 5e43c75d3..2dedf2a94 100644
--- a/lib/argp-ba.c
+++ b/lib/argp-ba.c
@@ -20,15 +20
