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

2025-04-27 Thread Paul Eggert
* m4/threadlib.m4: Refactor by prefering AS_CASE to case, " to ', $(...) to `, [[...]] to changequote, and #-to-newline to #-with-]. That way, Emacs commands recognize the sexps here. Omit some unnecessary quotes. --- ChangeLog | 6 ++ m4/threadlib.m4 | 183 ++--

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

2025-04-27 Thread Paul Eggert
* m4/threadlib.m4: When conditionally invoking a macro that uses or might use AC_REQUIRE, put it inside AS_CASE or AS_IF instead of using shell constructs directly. --- ChangeLog | 5 +++ m4/threadlib.m4 | 111 +++- 2 files changed, 58 insertions

[PATCH] manywarnings: omit -Wdisabled-optimization

2025-04-27 Thread Paul Eggert
* m4/manywarnings.m4 (gl_MANYWARN_ALL_GCC): Remove -Wdisabled-optimization, as this is a warning about the compiler not the program. I discovered the need for this when compiling bleeding-edge Emacs with gcc-15 (Ubuntu 15-20250404-0ubuntu1) 15.0.1 20250404 (experimental) [master r15-9193-g08e803aa

[PATCH] gendocs: do not depend on awk

2025-04-27 Thread Paul Eggert
* build-aux/gendocs.sh (calcsize): Use shell directly, rather than depending on awk. Allow spaces in outdir. Simplify. --- ChangeLog| 6 ++ build-aux/gendocs.sh | 6 +++--- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/ChangeLog b/ChangeLog index 930d181551..98ff

Update end-of-life.txt.

2025-04-27 Thread Collin Funk
OpenBSD 7.7 was released 2025-04-28 [1]. This means that 7.5 is EOL the same day [2]. Applied the attached patch to the maint-tools repository end-of-life.txt. I created a VM with OpenBSD 7.7 and ran a testdir of all modules and saw only one failure for test-nonblocking-socket.sh. But I assume tha

Make use of gcc's __nonnull_if_nonzero__ attribute

2025-04-27 Thread Bruno Haible via Gnulib discussion list
ISO C decided to standardize the long-standing behaviour of functions like memcpy(): that it's OK to call them with a NULL pointer argument if the corresponding size parameter is 0. [1] (They still refuse to talk about arrays of length 0, though. This is not only a matter of terminology. This forbi