* admin/merge-gnulib (GNULIB_MODULES): Add memeq, streq and remove
stringeq, to adjust to a module splitup in Gnulib.
* lib/memeq.c, lib/streq.c, lib/strnul.c, m4/memeq.m4, m4/streq.m4:
New files.
* lib/string.c, m4/stringeq.m4: Remove.
* admin/merge-gnulib (GNULIB_MODULES): Add crypto/sha3-buffer.
* lib/sha3.c: New file, imported by running admin/merge-gnulib.
* lib/sha3.h: Likewise.
* m4/sha3.m4: Likewise.
* lib/gnulib.mk.in: Updated by admin/merge-gnulib.
* m4/gnulib-comp.m4: Likewise.
* src/fns.c: Include sha3.h
(Fsecure_hash_algorithms): Add Qsha3_224, Qsha3_256, Qsha3_384, and
Qsha3_512.
(secure_hash): Likewise.
(Fsecure_hash): List the SHA-3 algorithms in the docstring.
(syms_of_fns): Define Qsha3_224, Qsha3_256, Qsha3_384, and Qsha3_512.
* test/lisp/net/gnutls-tests.el (gnutls-tests-internal-macs-upcased):
Filter out the new SHA-3 algorithms since they are currently not
implemented in gnutls.
* test/src/fns-tests.el (test-secure-hash): Add test cases for the new
algorithms.
* doc/lispref/text.texi (Checksum/Hash): List the SHA-3 algorithms.
Mention that they are considered secure.
* etc/NEWS: Mention the new feature.
Uses of <vla.h>’s macro VLA_ELEMS were removed some time ago.
* admin/merge-gnulib (GNULIB_MODULES): Remove vla.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
* lib/vla.h: Remove.
* src/fns.c, src/lread.c: Do not include <vla.h>.
* admin/merge-gnulib (AVOIDED_MODULES): Add strncpy, as Emacs is not
likely to exercise the FreeBSD 15 strncpy bug that this module fixes.
* lib/issymlinkat.h: New file, taken from Gnulib.
This merges in changes that were made to Autoconf master
in April 2024. It causes ‘configure’ to prefer the C compiler to
enable C23 features if possible; otherwise, ‘configure’
behaves as before, using C11 if available and C99 if not.
* admin/merge-gnulib (GNULIB_MODULES): Add std-gnu23.
(AVOIDED_MODULES): Add (i.e., start avoiding) std-gnu11.
* m4/std-gnu11.m4: Remove.
* m4/std-gnu23.m4: Add.
* admin/merge-gnulib (GNULIB_MODULES): Add stringeq.
With current Gnulib it is already present as in indirect dependency;
listing it here because Emacs now depends on it directly.
* lib-src/ebrowse.c, lib-src/etags.c:
(streq): Remove, as Gnulib defines this now.
* lib/fseterr.c, lib/fseterr.h, lib/issymlink.c, lib/issymlink.h:
* lib/issymlinkat.c, lib/stdio-consolesafe.c, lib/string.c:
* m4/fseterr.m4, m4/gettext_h.m4, m4/stringeq.m4:
New files from Gnulib.
* src/conf_post.h (tzfree) [__ANDROID_API__ >= 35]: Remove.
Prepare for the next Gnulib merge by disabling the optional
support for non-Gregorian calendars recently added to Gnulib’s
nstrftime module. This would be tricky to integrate with Emacs,
as Emacs has its own (much fancier) calendar system, one that at
least for the Persian calendar sometimes disagrees with Gnulib’s,
and it’s likely not worth the trouble of reconciling the conflicts.
* src/conf_post.h (SUPPORT_NON_GREG_CALENDARS_IN_STRFTIME):
New macro, defined to false.
* admin/merge-gnulib (GNULIB_MODULES): nstrftime → nstrftime-limited.
(AVOIDED_MODULES): Add localcharset.
Near the end, also remove the files lib/calendar-ethiopian.h,
lib/calendar-persian.h, lib/calendar-thai.h, lib/calendars.h.
The following changes were made by hand,
so that admin/merge-gnulib could succeed
instead of failing because the diff didn’t match.
* admin/gnulib-patches/lib/getloadavg.c.diff:
Remove, as it is no longer needed now that recent
Gnulib has been merged.
* admin/merge-gnulib (GNULIB_TOOL_FLAGS):
Remove the --local-dir="$src"admin/gnulib-patches option,
as it is no longer needed either.
* admin/gnulib-patches/lib/getloadavg.c.diff: New file.
* admin/merge-gnulib (GNULIB_TOOL_FLAGS): Set --local-dir to
admin/gnulib-patches.
* lib/gnulib.mk.in: Update from Gnulib.
* admin/merge-gnulib (AVOIDED_MODULES):
Add locale-h, as Emacs should’t need this Gnulib module.
This change is needed for when we next run admin/merge-gnulib,
as some Gnulib dependencies changed recently.
Instead of worrying about whether malloc and realloc yield a
non-null pointer when given size 0 (the traditional behavior
recommended by Doug McIlroy, and the only behavior that makes
sense for Emacs), let Gnulib arrange for the traditional behavior.
* admin/merge-gnulib (GNULIB_MODULES): Add malloc-gnu (which we
were already depending on indirectly) and realloc-posix.
* lib/gnulib.mk.in, m4/gnulib-comp.m4:
Regenerate by running admin/merge-gnulib.
From suggestions by Eli Zaretski and Stefan Kangas in:
https://lists.gnu.org/r/emacs-devel/2025-01/msg00036.htmlhttps://lists.gnu.org/r/emacs-devel/2025-01/msg00037.html
* admin/merge-gnulib: For style consistency,
prefer ‘[ EXPR ]’ to ‘test EXPR’ and prefer -e to -f.
Fix bug that forgot to clean lib directory if it’s ./lib.
Check for lib/gnulib.mk.in, which we used directly, not ‘configure’.
(LC_ALL): Set to C.
(autogen): New function, to make it clearer to user when
commands are operating in some other directory.
(get_module_list): New function.
(old_modules, new_modules): New vars.
If they differ, issue a warning.
* admin/merge-gnulib (GNULIB_MODULES): Add builtin-expect,
since src/lisp.h now uses __builtin_expect directly,
which we should no longer rely on it being pulled in as
a dependency of some other Gnulib module that we use.
C23's <stdbit.h> in the long run should be better supported than
Gnulib's count-one-bits.h and similar headers, so switch to the
C23 primitives, with a Gnulib fallback for platforms lacking C23.
* admin/merge-gnulib (GNULIB_MODULES): Remove count-leading-zeros,
count-one-bits, count-trailing-zeros. Add stdc_bit_width,
stdc_count_ones, stdc_trailing_zeros.
* lib/count-leading-zeros.c, lib/count-leading-zeros.h:
* lib/count-one-bits.c, lib/count-one-bits.h:
* lib/count-trailing-zeros.c, lib/count-trailing-zeros.h: Remove.
* lib/stdbit.c, lib/stdbit.in.h, lib/stdc_bit_width.c:
* lib/stdc_count_ones.c, lib/stdc_leading_zeros.c:
* lib/stdc_trailing_zeros.c, m4/stdbit_h.m4:
New files, copied from Gnulib.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
* src/data.c: Do not include count-one-bits.h, count-trailing-zeros.h.
Instead, rely on lisp.h including stdbit.h.
(Flogcount, Fbool_vector_count_population)
(Fbool_vector_count_consecutive): Use stdbit.h macros instead of
count-one-bits.h and count-trailing-zeros.h macros.
(shift_right_ull, count_one_bits_word, pre_value)
(count_trailing_zero_bits): Remove; no longer needed.
* src/lisp.h: Include stdbit.h instead of count-leading-zeros.h.
(elogb): Use stdbit.h macro instead of count-leading-zeros.h macro.
* admin/merge-gnulib: Also remove the new Gnulib file
m4/off64_t.m4, as Emacs doesn’t need off64_t.
* configure.ac (gl_TYPE_OFF64_T): Define a nearly no-op.
Bruno Haible fixed Gnulib so that nstrftime no longer requires
locking code, which means we no longer need to avoid localename.
However, nstrftime now requires localename-unsafe-limited which
pulls in some Gnulib-specific locale code, and it’s likely this
needs to be replaced with Emacs-specific locale code. In the
meantime let’s continue to finess this by avoiding
localename-unsafe-limited.
* admin/merge-gnulib (AVOIDED_MODULES):
Avoid localename-unsafe-limited instead of localename.
Problem reported by Gerd Möllmann (Bug#68464).
* .gitignore: Ignore files in exec that are now copied from build-aux.
* admin/merge-gnulib (avoided_flags): Instead of clearing
autom4te.cache here ...
* autogen.sh (do_git): ... clear it here.
Use config.guess, config.sub and install-sh from
the Emacs repository, as they are more likely to be up to date.
This avoids unnecessary differences among different builders,
and avoids unnecessary ‘git diff’ output after autogen.sh.
Also, copy these files from build-aux to exec since there
should be no difference between the two copies.
* exec/config.guess, exec/config.sub, exec/install-sh:
Remove from repository, as autogen.sh now copies them from build-aux.
* admin/merge-gnulib (AVOIDED_MODULES): Add ‘access’,
since I don’t think Emacs cares about the Mac OS X 10.5
bug where faccessat ignores a trailing slash on a symlink
to a directory. This will matter the next time we run
admin/merge-gnulib.
Simplify Emacs proper by using Gnulib’s boot-time module
instead of doing it all by hand. This should port Emacs
better to obscurish hosts, as Bruno Haible has merged the
best of Emacs’s and Gnulib’s boot-time gathering.
* lib/boot-time-aux.h, lib/boot-time.c, lib/boot-time.h:
* lib/readutmp.h, m4/readutmp.m4: New files, copied from Gnulib.
* admin/merge-gnulib (GNULIB_MODULES): Add boot-time.
* configure.ac: Do not check for utmp.h;
the boot-time module now does this.
(BOOT_TIME_FILE): Remove; no longer used.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
* src/filelock.c [__FreeBSD__]: Do not include <sys/sysctl.h>.
[HAVE_UTMP_H]: Do not include utmp.h.
Include boot-time instead: boot-time does the work now.
(BOOT_TIME) [HAVE_ANDROID && !ANDROID_STUBIFY]: Don’t undef.
(WTMP_FILE): Don’t define.
(boot_time, boot_time_initialized, get_boot_time_1, get_boot_time):
Remove.
(get_boot_sec): New function that simply calls Gnulib get_boot_time.
(lock_file_1, current_lock_owner): Use get_boot_sec instead
of get_boot_time.
It’s not worth the porting hassle, and as the glibc manual says,
“this function is generally a poor choice for processing strings”.
* admin/merge-gnulib (GNULIB_MODULES): Remove stpncpy.
* exec/configure.ac: Do not check for stpncpy.
* exec/exec.c (rpl_stpncpy, stpncpy): Remove this replacement.
(exec_0): Properly clear buffer1. Use memcpy instead of
stpncpy to add the trailing name. This code is clearly
still suboptimal but efficiency is not that important here
and I tried to minimize the change.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
The Gnulib regex module now pulls in some wide-char stuff
that I think Emacs doesn’t need.
* admin/merge-gnulib (AVOIDED_MODULES): Add iswblank, iswctype,
iswdigit, iswxdigit, wctype.
* m4, lib: Update from Gnulib.
* msdos/sedlibmk.inp: Remove variables deleted as part of previous
change.
* admin/merge-gnulib (GNULIB_MODULES): Remove vasprintf and
printf-posix.
* admin/merge-gnulib: Remove m4/locale-fr.m4.
* configure.ac (gt_LOCALE_FR_UTF8): Replace Gnulib’s
definition with a near no-op, as Emacs doesn’t care
about the Solaris 10 MB_CUR_MAX bug.
* lib/gnulib.mk.in: Regenerate.
* m4/locale-fr.m4: Remove; no longer needed due
to above changes.
This adds support for 64-bit time_t on 32-bit GNU/Linux ARM and
x86, where time_t was historically 32-bit. The resulting Emacs
should continue to work for timestamps after mid-January 2038.
* admin/merge-gnulib (GNULIB_MODULES): Add year2038.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
* admin/merge-gnulib: Copy m4/codeset.m4 too. This is mostly for
documentation, as it’s now automatically copied by other modules.
* configure.ac (EMACS_PAPER_WIDTH): New macro. This relies on
Gnulib, to avoid duplicate tests for HAVE_LANGINFO_CODESET.
Run admin/merge-gnulib along with the following change:
* admin/merge-gnulib (GNULIB_MODULES): Replace ‘time’ with
‘time-h’, adjusting to a recent renaming in Gnulib modules.
Emacs doesn’t appear to need the new Gnulib ‘time’ module.
* .gitignore: Simplify cross/lib rule.
* admin/merge-gnulib (avoided_flags): Stop copying to cross/lib.
* configure.ac: Link gnulib source and header files to
cross/lib.
* cross/Makefile.in (LIB_SRCDIR): Make relative to builddir.
(maintainer-clean): Merge with distclean. Remove links created
by configure.
These are neccessary because Android's printf is missing basic format
modifiers such as t.
* admin/merge-gnulib (GNULIB_MODULES): Add printf-posix and
vasprintf-posix. Update from gnulib.
* configure.ac (CFLAGS): Add -DHAVE_CONFIG_H.
* src/android.c (android_run_select_thread, android_init_events)
(android_select): Add alternative android_select implementation
for API 16 and lower.
* src/androidterm.c (handle_one_android_event): Fix
use-after-frees.
