Re: stdckdint-h: Work around missing declarations with g++ 15.0.

On 2025-05-20 19:33, Collin Funk wrote: we have no way of portably setting the compiler flags if available. That's fine, but Gnulib does have a portable way to see whether the C++ compiler works properly with . Namely, compile a little test with the C++ compiler and if that compiles OK, do

Re: stdckdint-h: Work around missing declarations with g++ 15.0.

As the stdckdint.h bug in question is fixed in GCC 15.1, wouldn't it be better if the Gnulib workaround is used only if the bug is present (GCC 15.0)?

Re: re_node_set_alloc relies on malloc(0) returning non-NULL

ghtly, by avoiding a no-longer-needed runtime check inside pma_malloc.From de452358c2e518263093d423c402bf0aeb5f87b1 Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Mon, 19 May 2025 12:22:50 -0700 Subject: [PATCH] Change pma_malloc to behave more like GNU malloc This simplifies using Gnulib code

[PATCH] tests: simplify test_exit_status decl

index 09c5b008df..75792d4fe2 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,10 @@ +2025-05-18 Paul Eggert + + tests: simplify test_exit_status decl + * tests/macros.h (test_exit_status) [!NO_MAIN_HERE]: + Declare test_exit_status too, as this is simpler + and we might

[PATCH] Fix misspelling of ‘or1k’ in doc

--- doc/year2038.texi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/year2038.texi b/doc/year2038.texi index 47e26d85dd..43c63007e3 100644 --- a/doc/year2038.texi +++ b/doc/year2038.texi @@ -65,7 +65,7 @@ Avoiding the year 2038 problem @item Linux kernel 5.1 (2019) and l

Re: qcopy-acl: Fix copying of ACLs on CentOS 7

On 2025-05-13 06:26, Bruno Haible via Gnulib discussion list wrote: Paul Eggert wrote: - || strcmp (name, XATTR_NAME_NFSV4_ACL) == 0 - || attr_copy_action (name, ctx) == ATTR_ACTION_PERMISSIONS; + || strcmp (name, XATTR_NAME_NFSV4_ACL) == 0; } #endif

Re: qcopy-acl: Fix copying of ACLs on CentOS 7

On 2025-05-12 05:24, Bruno Haible wrote: Hi Pádraig, I checked one centos 7 system, and it didn't have a /etc/xattr.conf file which might explain the behavior noticed above. Then on centos 8 we have: system.nfs4_acl permissions system.nfs4acl permissions system.

Re: qcopy-acl: Fix copying of ACLs on CentOS 7

On 2025-05-12 05:43, Ondrej Valousek wrote: I think they intentionally removed the system.nfs4* in Fedora-42 because commands like "cp -a" complained when copying files. Yes, the removal in 2019 was due to a suggestion[1] by Pádraig, which Kamil Dudka was a bit dubious about[2] but went ahead

Re: [PATCH] acl-tests: link with $(FILE_HAS_ACL_LIB)

On 2025-05-10 05:46, Bruno Haible wrote: Here, after the qcopy-acl change yesterday, the following modules need to link with $(FILE_HAS_ACL_LIB), at least conditionally: * qcopy-acl * The dependents of qcopy-acl: $ ./gnulib-tool --extract-dependents qcopy-acl acl copy-f

[PATCH] acl-tests: link with $(FILE_HAS_ACL_LIB)

. --- ChangeLog | 9 + modules/acl-tests | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/ChangeLog b/ChangeLog index 39e2ae752d..bb1f55cce0 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,12 @@ +2025-05-09 Paul Eggert + + acl-tests: link with $(FILE_HAS_ACL_LIB

[PATCH] qcopy-acl: port better to NFSv4 on GNU/Linux

/qcopy-acl.c| 29 ++-- modules/qcopy-acl | 4 +- 7 files changed, 182 insertions(+), 53 deletions(-) diff --git a/ChangeLog b/ChangeLog index f2da0a6843..842a40a47f 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,26 @@ +2025-05-09 Paul Eggert + + qcopy-acl: port better to

Re: strftime-induced linker errors on Android API 35.

On 2025-05-08 17:50, Po Lu wrote: I will be unable to merge this update into Emacs for several days yet as my attention is currently concentrated elsewhere, but if anyone does it will be automatically tested within 24 hours. I merged it.

[PATCH] Fix some ungrammatical uses of "allows to"

This buys back some comment changes from Emacs, and fixes other instances I noticed. --- ChangeLog| 6 +++--- build-aux/announce-gen | 4 ++-- build-aux/prefix-gnulib-mk | 4 ++-- build-aux/useless-if-before-free | 4 ++-- doc/posix-functions/utimes.texi

Re: [PATCH 1/2] obstack: check def before use in macros

0a6422c8f Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Wed, 7 May 2025 20:15:34 -0700 Subject: [PATCH] ChangeLog fix Problem reported by Bruno Haible in: https://lists.gnu.org/r/bug-gnulib/2025-05/msg00096.html --- ChangeLog | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff -

[PATCH] Omit a parenthesis pair again (fixing inadvertent revert)

--- lib/obstack.in.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/obstack.in.h b/lib/obstack.in.h index 4b133c32cc..698b6b37a7 100644 --- a/lib/obstack.in.h +++ b/lib/obstack.in.h @@ -593,7 +593,7 @@ extern int obstack_exit_failure; (((_OBSTACK_CPTR) (h)->chunk < (OB

[PATCH 1/2] obstack: check def before use in macros

changed, 17 insertions(+), 7 deletions(-) diff --git a/ChangeLog b/ChangeLog index efb647b3ab..dc81c8c726 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,11 @@ +2025-05-07 Paul Eggert + + obstack: check def before use in macros + * lib/obstack.in.h (__attribute_noreturn__

[PATCH 2/2] obstack: avoid undefined pointer comparison

--git a/ChangeLog b/ChangeLog index dc81c8c726..e63d92728a 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,10 @@ 2025-05-07 Paul Eggert + obstack: avoid undefined pointer comparison + * lib/obstack.in.h (_OBSTACK_CPTR): New macro. + (obstack_free): Use it instead of

[PATCH] obstack: omit a parenthesis pair

* lib/obstack.in.h (obstack_free) [!__GNUC__]: Omit an unnecessary pair of parentheses that are also omitted in glibc. --- lib/obstack.in.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/obstack.in.h b/lib/obstack.in.h index 174cab9d0a..e361dc67ac 100644 --- a/lib/obstack.

Re: [PATCH 12/19] obstack: _obstack_free → __obstack_free

t know anybody who still does that, but if they exist they'll likelier be happier this way.From acaa6c07d85fa5eeadae4e60b717fb25c919c23d Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Tue, 6 May 2025 15:09:02 -0700 Subject: [PATCH] obstack: let glibc user #define __obstack_

Re: [PATCH 02/19] obstack: stop depending on alignof

Thanks, I installed the attached.From 10f5798abc5e8ac8c4be0c8339e37bca14fd5957 Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Tue, 6 May 2025 15:14:50 -0700 Subject: [PATCH] obstack: fix comment * lib/obstack.c: Fix comment along the lines suggested by Bruno Haible in: https://lists.gnu.org/r

[PATCH] obstack: don’t assume INT_MAX < SIZE_MAX

.c | 67 + modules/obstack | 1 + 3 files changed, 47 insertions(+), 39 deletions(-) diff --git a/ChangeLog b/ChangeLog index 68b375e9b5..d60a9e5658 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,21 @@ +2025-05-06 Paul Eggert + + obstack: don’t assume INT_MAX < SIZE_MAX +

[PATCH 10/19] * lib/obstack.in.h: Move include.

This is merely to make the glibc reviewer’s job easier, since the include location doesn’t matter. --- lib/obstack.in.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/lib/obstack.in.h b/lib/obstack.in.h index 0f2677dba4..f53b78c15c 100644 --- a/lib/obstack.in.h +++ b/lib/

[PATCH 08/19] obstack: rename to _OBSTACK_CHUNK_SIZE_T

/ChangeLog +++ b/ChangeLog @@ -1,5 +1,9 @@ 2025-05-05 Paul Eggert + obstack: rename to _OBSTACK_CHUNK_SIZE_T + * lib/obstack.in.h (_OBSTACK_CHUNK_SIZE_T): + Rename from _CHUNK_SIZE_T, for namespace prefix consistency. + obstack: use int where glibc does Make it

[PATCH 02/19] obstack: stop depending on alignof

deletions(-) diff --git a/ChangeLog b/ChangeLog index 9ee3f6eea5..cdef5a8973 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,11 @@ 2025-05-05 Paul Eggert + obstack: stop depending on alignof + * lib/obstack.c (__alignof__) [!__GNUC__]: + Default to alignof, not to alignof_type

[PATCH 09/19] obstack: size_t → _OBSTACK_CHUNK_SIZE_T

+- 3 files changed, 30 insertions(+), 11 deletions(-) diff --git a/ChangeLog b/ChangeLog index 771309c182..403140cdf8 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,16 @@ 2025-05-05 Paul Eggert + obstack: size_t → _OBSTACK_CHUNK_SIZE_T + * lib/obstack.c

[PATCH 05/19] obstack: define PTR_INT_TYPE

, 19 insertions(+), 2 deletions(-) diff --git a/ChangeLog b/ChangeLog index 5c78f4dac4..7448c09ebc 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,11 @@ 2025-05-05 Paul Eggert + obstack: define PTR_INT_TYPE + * lib/obstack.in.h (PTR_INT_TYPE): Define, for compatibility

[PATCH 06/19] obstack: use presence of uintptr_t

+-- 3 files changed, 29 insertions(+), 29 deletions(-) diff --git a/ChangeLog b/ChangeLog index 7448c09ebc..cc7e7e3256 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,17 @@ 2025-05-05 Paul Eggert + obstack: use presence of uintptr_t + The code depends on the presence of

[PATCH 11/19] obstack: use glibc’s temp slot names

2 files changed, 32 insertions(+), 26 deletions(-) diff --git a/ChangeLog b/ChangeLog index 403140cdf8..3abdef82de 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,11 @@ 2025-05-05 Paul Eggert + obstack: use glibc’s temp slot names + * lib/obstack.in.h

[PATCH 12/19] obstack: _obstack_free → __obstack_free

insertions(+), 5 deletions(-) diff --git a/ChangeLog b/ChangeLog index 3abdef82de..c73702402d 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,10 @@ 2025-05-05 Paul Eggert + obstack: _obstack_free → __obstack_free + * lib/obstack.in.h (__obstack_free): Rename back from _obstack_free

[PATCH 15/19] libc-config, obstack: Oracle __extension__ support

..c0691ce564 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,9 @@ 2025-05-05 Paul Eggert + libc-config, obstack: Oracle __extension__ support + * lib/cdefs.h, lib/obstack.in.h (__extension__): + Oracle Developer Studio 12.6 also supports __extension__. + obstack: be more

[PATCH 17/19] obstack: use signed chunk sizes

100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,9 @@ 2025-05-05 Paul Eggert + obstack: use signed chunk sizes + * lib/obstack.in.h (_OBSTACK_CHUNK_SIZE_T) [!__GL_GNULIB_HEADER]: + Now long, not unsigned long, since that’s what glibc does. + obstack: fix undefined

[PATCH 18/19] obstack: simplify default size, alignment

/ChangeLog +++ b/ChangeLog @@ -1,5 +1,13 @@ 2025-05-05 Paul Eggert + obstack: simplify default size, alignment + * lib/obstack.c (MAX, DEFAULT_ALIGNMENT, DEFAULT_ROUNDING): Remove. + (_obstack_begin_worker): Use __alignof__ (max_align_t) + instead of calculating it by hand

[PATCH 13/19] obstack: be more like glibc in return values

-05-05 Paul Eggert + obstack: be more like glibc in return values + This partly reverts commit d15b2da0ac25e085ce30a9e2672624999ce910a6 + dated 2014-11-03. It keeps the part of that commit that fixed the + return type of non-gcc versions of obstack_make_room

[PATCH 14/19] obstack: be more like glibc for __extension__

@@ -1,5 +1,9 @@ 2025-05-05 Paul Eggert + obstack: be more like glibc for __extension__ + * lib/obstack.in.h (__extension__): Use glibc-like syntax + for testing gcc version. + obstack: be more like glibc in return values This partly reverts commit

[PATCH 19/19] obstack: match obstack.c with glibc

/obstack.c | 55 + modules/obstack | 1 + 3 files changed, 45 insertions(+), 18 deletions(-) diff --git a/ChangeLog b/ChangeLog index 20f9e7bc3c..f9da2b846b 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,12 @@ 2025-05-05 Paul Eggert

[PATCH 16/19] obstack: fix undefined behavior if big alignment

+++- lib/obstack.in.h | 6 -- 3 files changed, 58 insertions(+), 16 deletions(-) diff --git a/ChangeLog b/ChangeLog index c0691ce564..8967f3d54f 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,17 @@ 2025-05-05 Paul Eggert + obstack: fix undefined behavior if big

[PATCH 01/19] obstack: fix unlikely buffer overrun in glibc

erflow. --- ChangeLog| 9 + lib/obstack.c| 19 --- lib/obstack.in.h | 4 ++-- 3 files changed, 27 insertions(+), 5 deletions(-) diff --git a/ChangeLog b/ChangeLog index adf06a2645..9ee3f6eea5 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,12 @@ +2025-05-05 Paul

[PATCH 04/19] obstack: change license wording

Change license wording to match glibc, to ease future merging. --- lib/obstack.c| 19 ++- lib/obstack.in.h | 19 ++- 2 files changed, 20 insertions(+), 18 deletions(-) diff --git a/lib/obstack.c b/lib/obstack.c index 7d19f6bd3a..92e2a1206f 100644 --- a/lib/obst

[PATCH 03/19] obstack: merge better with glibc

. --- ChangeLog| 8 lib/obstack.in.h | 13 - modules/obstack | 1 + 3 files changed, 13 insertions(+), 9 deletions(-) diff --git a/ChangeLog b/ChangeLog index cdef5a8973..5c78f4dac4 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,13 @@ 2025-05-05 Paul Eggert

[PATCH 07/19] obstack: use int where glibc does

deletions(-) diff --git a/ChangeLog b/ChangeLog index cc7e7e3256..ed1c57575b 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,16 @@ 2025-05-05 Paul Eggert + obstack: use int where glibc does + Make it easier to merge with glibc, by using int where + the glibc API uses int, when

[PATCH 00/19] prepare to merge Gnulib obstack back into glibc

e.org/bugzilla/show_bug.cgi?id=32393 Paul Eggert (19): obstack: fix unlikely buffer overrun in glibc obstack: stop depending on alignof obstack: merge better with glibc obstack: change license wording obstack: define PTR_INT_TYPE obstack: use presence of uintptr_t obstack: use int where

Re: obstack changes: relicense under LGPLv2+

On 2025-05-04 03:10, Bruno Haible wrote: Would you please relicense your changes under LGPLv2+? That is, give us the ability to relicense the 'obstack' module under LGPLv2+ and then be able to flow back the changes into glibc? Yes, of course. And I don't think things were unclear at all, as th

Re: [PATCH] regex: Add \A and \z synonyms to \` and \'

On 2025-05-01 08:15, Eric Blake wrote: If POSIX were to try to standardize \` and \' instead of \A and \z, it would have the benefit of existing practice (glibc), but the drawback of less familiarity (lots more languages would have to adapt, whereas now \A is fairly common and \` is rare). I do

Re: obstack.in.h license

On 4/30/25 18:33, Collin Funk wrote: (I would not bet that code like this: #if 0 # if @FOOBAR@ # endif #endif works for all C and C++ compilers. Better avoid that.) I always wondered if a compiler would fail with that. The C standard has required it to work ever since C89 (see its

Re: [PATCH] manywarnings: omit -Wdisabled-optimization

On 2025-04-28 02:43, Bruno Haible wrote: Paul Eggert wrote: Remove -Wdisabled-optimization, as this is a warning about the compiler not the program. Does gcc's warning message give a hint how the compiler options could be adjusted, so as to enable the optimization? Unfortunately no

Re: obstack.in.h license

On 4/29/25 18:30, Collin Funk wrote: Thanks! I submitted the patch for the clang undefined behavior sanitizer issue and Paul fixed it to account for your later Oracle cc fix [1]. That glibc patch needs more work, as I think Gnulib is misusing _LIBC. _LIBC means we're using the header while com

Re: obstack.in.h license

On 4/29/25 18:50, Bruno Haible wrote: Paul, what do you think? Should we change 'obstack' from LGPLv3+ to LGPLv2+ ? That would be less confusing, yes.

Re: [PATCH 2/2] threadlib: follow Autoconf AC_REQUIRE rules better

On 2025-04-28 02:53, Bruno Haible wrote: This change was not needed, because the use of AS_CASE or AS_IF w.r.t. AC_REQUIRE is not needed inside AC_DEFUNs [1]. Ah, I didn't know that. Thanks for letting me know. No harm done by the patch, I hope.

Re: [PATCH 1/2] threadlib: make m4 work better with Emacs

On 2025-04-28 03:46, Bruno Haible wrote: But the indentation in your patch is terrible: I did that to minimize the size of the diff output for that patch. It's of course fine to fix the indentation, as you did later.

[PATCH 1/2] threadlib: make m4 work better with Emacs

b.m4 | 183 ++-- 2 files changed, 89 insertions(+), 100 deletions(-) diff --git a/ChangeLog b/ChangeLog index 84c5fecf1b..6a432b7950 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,11 @@ 2025-04-27 Paul Eggert + threadlib: make m4 work better with Emacs + * m4/th

[PATCH 2/2] threadlib: follow Autoconf AC_REQUIRE rules better

insertions(+), 58 deletions(-) diff --git a/ChangeLog b/ChangeLog index 6a432b7950..6a6a174be3 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,10 @@ 2025-04-27 Paul Eggert + threadlib: follow Autoconf AC_REQUIRE rules better + * m4/threadlib.m4: When conditionally invoking a macro that

[PATCH] manywarnings: omit -Wdisabled-optimization

b/ChangeLog index 98ffe323cc..84c5fecf1b 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,14 @@ 2025-04-27 Paul Eggert + manywarnings: omit -Wdisabled-optimization + * m4/manywarnings.m4 (gl_MANYWARN_ALL_GCC): + Remove -Wdisabled-optimization, as this is a warning about the

[PATCH] gendocs: do not depend on awk

..98ffe323cc 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,9 @@ +2025-04-27 Paul Eggert + + gendocs: do not depend on awk + * build-aux/gendocs.sh (calcsize): Use shell directly, rather than + depending on awk. Allow spaces in outdir. Simplify. + 2025-04-27 Bruno Haible

Re: [PATCH] regex: Add \A and \z synonyms to \` and \'

On 2025-04-24 09:13, Eric Blake wrote: case '\'': +case 'z': Python uses \Z for this and Python's \A ... \Z is cleaner and easier to remember than Perl's \A ... \z, so how about if we follow Python's lead instead? Users could get the effect of Perl's \Z, which is less commonl

[PATCH 2/2] doc: update some GCC version comments

GCC 15.1 is released, and we can update some commentary now. --- doc/gnulib.texi| 2 +- lib/diffseq.h | 2 +- m4/gnulib-common.m4| 7 +-- tests/test-sprintf-posix.h | 3 ++- 4 files changed, 9 insertions(+), 5 deletions(-) diff --git a/doc/gnulib.texi b/doc/gn

[PATCH 1/2] stddef-h: port better to future GCC 13.4, 14.3

changed, 13 insertions(+), 5 deletions(-) diff --git a/ChangeLog b/ChangeLog index 99c2c9c893..37bfcc814a 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,9 @@ +2025-04-25 Paul Eggert + + stddef-h: port better to future GCC 13.4, 14.3 + * m4/stddef_h.m4 (gl_STDDEF_H): The idempotency

Re: Android API 8 compilation errors

It should work the same as Po Lu's patch, while not unnecessarily repeating symbols which was my concern, and also while not assuming __ANDROID__ implies __linux__ which was your concern. From cd9859b6f5908f933d2db7ee92bcd1e0db9aac1f Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Mon, 21 Ap

Re: Android API 8 compilation errors

On 2025-04-21 05:15, Po Lu wrote: -# if !defined (LDAV_DONE) && (defined __linux__ || defined __ANDROID__) - /* Linux without glibc, Android, Cygwin */ +# if !defined LDAV_DONE && (defined __linux__ || 13 <= __ANDROID_API__) + /*

Re: Android API 8 compilation errors

On 2025-04-20 21:54, Po Lu wrote: Probably. I don't know whether any non-Android systems exist where __ANDROID_API__ is defined. I think it unlikely, but if we run into any we can complicate Gnulib later. Could you please review and try the attached patch? Thanks. diff --git a/lib/getloadav

Re: Android API 8 compilation errors

On 2025-04-20 18:45, Po Lu wrote: -# if !defined (LDAV_DONE) && (defined __linux__ || defined __ANDROID__) +# if !defined (LDAV_DONE) && (defined __linux__ || defined __ANDROID__) \ + && (!defined __ANDROID__ || __ANDROID_API__ >= 13) Wouldn't the following be simpler, and work as well? #if

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

-n "\$AWK" || { + awk '' >"$CONFIG_STATUS" <<\_ACEOF || ac_write_fail=1 -- 2.48.1 From 3d79cd9ff455d85e788398bfe4fdc4544fabceb5 Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Fri, 18 Apr 2025 14:01:38 -0700 Subject: [PATCH 2/2] Avoid grep in mini

[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?

g with 'cmp', 'gcc', 'make', and 'touch'.From 95b849dd837210dc3109adfbd52a0929fbec141e Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Thu, 17 Apr 2025 12:14:41 -0700 Subject: [PATCH] Improve list of 'configure' prereqs * doc/install.texi (Installation Pre

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

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

. Maybe Autoconf should be changed so that 'configure' fails quickly if there's no awk. That shouldn't be much work.From 56860215d1af490eddb821d73154a190137aa8ec Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Thu, 17 Apr 2025 01:08:52 -0700 Subject: [PATCH] Document t

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

+++ 3 files changed, 10 insertions(+), 6 deletions(-) diff --git a/ChangeLog b/ChangeLog index 6da050ff2d..b143d1311b 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,11 @@ 2025-04-17 Paul Eggert + getlogin-tests: remove xalloc dependency + * modules/getlogin-tests (Depends-on

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

@@ -1,3 +1,13 @@ +2025-04-17 Paul Eggert + + 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

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

/ChangeLog +++ b/ChangeLog @@ -1,5 +1,9 @@ 2025-04-17 Paul Eggert + memset_explicit-tests: pacify -Wuse-after-free + * tests/test-memset_explicit.c: Ignore -Wuse-after-free, + since the test deliberately uses after free. + Pacify -Wunterminated-string-initialization

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

4 deletions(-) diff --git a/ChangeLog b/ChangeLog index b143d1311b..7a5586a8d7 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,10 @@ 2025-04-17 Paul Eggert + Pacify -Wunterminated-string-initialization + * tests/test-fwrite.c (main): Mark nonstrings with + _GL_ATTRIBUTE_NONSTR

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

0644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,14 @@ 2025-04-17 Paul Eggert + 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/

Re: regex: Fix undefined behaviour

Thanks for fixing that. I installed the attached followup patch, which cleans up some of the mess I made when I put in that macro in the first place.From 7fddad0294390f85855c93175bfc513107ebd7ee Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Mon, 14 Apr 2025 15:42:40 -0700 Subject: [PATCH

Re: [PATCH] Update RE_SYNTAX_EMACS to include features used by GNU Emacs

2001 From: Paul Eggert Date: Mon, 14 Apr 2025 14:43:02 -0700 Subject: [PATCH] =?UTF-8?q?regex:=20don=E2=80=99t=20check=20RE=5FSYNTAX=5F?= =?UTF-8?q?EMACS?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * m4/regex.m4 (gl_REGEX): Do not check RE_SYNTAX_EMA

Re: [PATCH] Update RE_SYNTAX_EMACS to include features used by GNU Emacs

On 2025-04-13 20:39, Paul Eggert wrote:    - Will they change RE_BACKSLASH_ESCAPE_IN_LISTS from    ((unsigned long int) 1) to    (1UL) in order to make the RE_* values usable in preprocessor expressions? No; I followed up with a later patch to remove that assumption. Oops

Re: [PATCH] Update RE_SYNTAX_EMACS to include features used by GNU Emacs

On 2025-04-13 19:28, Bruno Haible wrote: Paul Eggert wrote: Thanks; I installed the attached somewhat fancier patch into Gnulib. ... and then installed the attach further patch to fix a thinko in the previously-mentioned one. I don't understand this patch. What changes are you expecti

Re: [PATCH] Update RE_SYNTAX_EMACS to include features used by GNU Emacs

On 2025-04-13 18:04, Paul Eggert wrote: Thanks; I installed the attached somewhat fancier patch into Gnulib. ... and then installed the attach further patch to fix a thinko in the previously-mentioned one.From f6d648a883676256894f5687cbceffb1f4209e3d Mon Sep 17 00:00:00 2001 From: Paul Eggert

[PATCH] regex: pacify gcc -Wcalloc-transposed-args

++-- lib/regex_internal.c | 4 ++-- lib/regexec.c| 10 +- 4 files changed, 20 insertions(+), 13 deletions(-) diff --git a/ChangeLog b/ChangeLog index c20c151757..e202b14530 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,12 @@ 2025-04-13 Paul Eggert + regex: pacify

Re: [PATCH] Update RE_SYNTAX_EMACS to include features used by GNU Emacs

Thanks; I installed the attached somewhat fancier patch into Gnulib.From efd5c380ff8062541d5fd98b050ecd3cb295917c Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Sun, 13 Apr 2025 18:01:08 -0700 Subject: [PATCH] regex: match current Emacs behavior * config/srclist.txt: Comment out regex.h

Re: localeconv tests: Add multithread-safety test

On 2025-04-12 14:01, Bruno Haible wrote: Paul Eggert wrote: To me it means that if two threads simultanously call localeconv and any other function (including setlocale and uselocale!), localeconv might have a data race. You mean, "localeconv will *not* have a data race" ?

Re: localeconv tests: Add multithread-safety test

On 4/12/25 00:39, Bruno Haible via Gnulib discussion list wrote: POSIX [1] says about localeconv(): "The localeconv() function need not be thread-safe; however, localeconv() shall avoid data races with all other functions." [1]https://pubs.opengroup.org/onlinepubs/9799919799/functions/loc

Re: symbolic link curiousity in 3.6.0

On 2025-04-10 15:47, Bruno Haible wrote: But what I could do is to change Gnulib's file-has-acl module so that if the environment variable CYGWINLY_PEDANTIC is set, Oh, let's not do that. We have too many environment variables already. Thanks for explaining.

Re: symbolic link curiousity in 3.6.0

On 4/9/25 04:09, Corinna Vinschen wrote: I would like to ask you to reconsider and to remove the Cygwin-special code and let Cygwin decide by itself, what a trivial vs. non-trivial ACL is by using the POSIXy functions Cygwin provides. Just use acl_extended_file(), please. Sorry, I've lost conte

Re: Build failure on IBM z/OS

On 4/10/25 07:02, Sachin T wrote: I'm encountering a build failure while trying to compile Coreutils v9.7 on IBM z/OS version 2.4. The error message is: “Please port gnulib getlocalename_l-unsafe.c to your platform! Report this to bug-gnulib”. How did you configure and run coreutils, and what

Re: symbolic link curiousity in 3.6.0

On 2025-04-04 10:40, Pádraig Brady wrote: Is this patch is OK to be applied to gnulib, with the tweaks Corinna mentioned? Yes, I did that just now, and propagated it into coreutils.

Re: symbolic link curiousity in 3.6.0

On 2025-03-30 05:50, Corinna Vinschen wrote: The definition of O_PATH requires an additional #include Thanks, I added that and installed the patch into Gnulib. However, assuming not only Cygwin is affected, shouldn't the patch rather use O_PATH if it's available, O_RDONLY if not? I h

Re: parse-duration: Work around an strtoul() misfeature

On 3/21/25 08:33, Bruno Haible wrote: Regarding speed, I'm not sure: the byte describing the properties of *str is in the L1 cache after the execution of isspace ((unsigned char) *str) Oh, and I just checked with gcc -O2 on x86-64, which optimizes the isdigit version into a fewer bytes' wor

Re: symbolic link curiousity in 3.6.0

patch work for you? I haven't tested or installed it (I don't use Cygwin).From e245ab6ac865c7ff723837645886eb717c53a754 Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Sat, 29 Mar 2025 10:27:01 -0600 Subject: [PATCH] file-has-acl: port symlink code to Cygwin Problem reported by Cori

Re: symbolic link curiousity in 3.6.0

On 3/31/25 12:26, Corinna Vinschen wrote: ls(1) always potentially shows a past state anyway. Sure, but traditionally (and I'm talking about 7th edition Unix) a single output line of 'ls' corresponded to a state obtained atomically from the file system. I realize we can't always do that nowad

Re: symbolic link curiousity in 3.6.0

On 3/31/25 11:27, Pádraig Brady wrote: The file could be deleted at any time. We're just suppressing errors in the edge case it's deleted More generally, though, the file could be renamed and another put in its place, which means that an attacker could cause 'ls' to generate a line that does

Re: symbolic link curiousity in 3.6.0

On 2025-03-30 07:26, Pádraig Brady wrote: On 30/03/2025 13:50, Corinna Vinschen wrote: In terms of coreutils, I think either ls(1) gobble_file() or file_has_aclinfo_cache() should still handle ENOENT from file_has_aclinfo() and not print any error message.  After all, due to the preconditions fo

[PATCH 2/2] same-inode: don't assume memory objects have ino

/same-inode.h | 9 +++-- 2 files changed, 14 insertions(+), 2 deletions(-) diff --git a/ChangeLog b/ChangeLog index d6377489ce..d60af23955 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,12 @@ 2025-03-24 Paul Eggert + same-inode: don't assume memory objects have ino +

[PATCH 1/2] same-inode: update now-wrong dependency

+1,9 @@ 2025-03-24 Paul Eggert + same-inode: update now-wrong dependency + * modules/same-inode (Depends-on): Depend on sys_stat-h, + not sys_types.h. + libtextstyle-optional: pacify gcc -Wunused-parameter * lib/textstyle.in.h (html_ostream_create): Mark arg as

[PATCH] libtextstyle-optional: pacify gcc -Wunused-parameter

@@ +2025-03-24 Paul Eggert + + libtextstyle-optional: pacify gcc -Wunused-parameter + * lib/textstyle.in.h (html_ostream_create): Mark arg as unused. + 2025-03-24 Bruno Haible getlocalename_l-unsafe: Work around Cygwin 3.6.0 bug. diff --git a/lib/textstyle.in.h b/lib

[PATCH] nproc, openmp-init: omit unnecessary strtol code

| 10 ++ lib/nproc.c| 22 +- lib/omp-init.c | 22 +- 3 files changed, 28 insertions(+), 26 deletions(-) diff --git a/ChangeLog b/ChangeLog index dae23affdb..8a788702e6 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,13 @@ +2025-03-19 Paul

Re: parse-duration: Work around an strtoul() misfeature

On 3/21/25 05:16, Bruno Haible via Gnulib discussion list wrote: + while (isspace ((unsigned char) *str)) +str++; + if (isdigit ((unsigned char) *str)) +{ + unsigned long ret = strtoul (str, (char **)ppz, base); Perhaps "*str != '+' && *str != '-'" instead of "isdigit ((unsigned

Re: bug#76876: logname output is often wrong when linked with glibc

On 2025-03-09 11:49, Bruno Haible wrote: Nicolas Boos wrote: This page says that the result of the logname command and the LOGNAME variable must be the same: https://www.ibm.com/docs/en/aix/7.3?topic=l-logname-command An AIX man page is not a specification for what we run on GNU systems. Plus

Re: bug#76876: logname output is often wrong when linked with glibc

which also occurs with getlogin_r.From 2a60c939bec320abe18737360bb9841595f5d8d1 Mon Sep 17 00:00:00 2001 From: Paul Eggert Date: Sat, 8 Mar 2025 15:36:33 -0800 Subject: [PATCH] getlogin: document musl bug Problem reported by Nicolas Boos <https://bugs.gnu.org/76876>. --- ChangeLog

Re: tar --mtime

On 2025-03-03 03:54, Bruno Haible via Gnulib discussion list wrote: I don't think we have any users on HP-UX any more, for several years already. still seems to be active. It has a 64-bit Itanium 2 copy of Coreutils 9.6, released in January. Insta

Re: new module 'vc-mtime'

On 2025-02-24 17:04, Collin Funk wrote: Isn't unsigned long 32 bits on x86_64 Windows? It's also narrower than time_t on some other platforms, e.g., GNU/Linux x86 when compiled with -D_TIME_BITS=64 (the default if Gnulib's year2038 module is used). && endptr == line + strlen

Re: new module 'vc-mtime'

On 2/24/25 11:10, Bruno Haible via Gnulib discussion list wrote: - If F is under version control and not modified locally: the time of the last change of F in the version control system. It's worth mentioning that in Git, this is the commit timestamp and not the author timestamp.

Re: hard_locale in a single-threaded app

On 2025-02-22 10:02, Bruno Haible wrote: The latest release already contains some multithreading support already: That was inadvertent. When you intersect the resulting list with the gnulib_modules defined in diffutils/bootstrap.conf, the result is: exclude nstrftime regex In detail, we

  1   2   3   4   5   6   7   8   9   10   >