Checking my config.log gives: configure:20626: checking for posix_spawn configure:20626: clang -o conftest -O3 -march=native -mtune=native -flto=thin -flto-jobs=5 -std=gnu\ 17 -Wno-unused-command-line-argument -Wno-unknown-pragmas conftest.c -lpng -ljpeg >&5 configure:20626: $? = 0 configure:20626: result: yes configure:20632: checking for posix_spawn_file_actions_addchdir configure:20632: clang -o conftest -O3 -march=native -mtune=native -flto=thin -flto-jobs=5 -std=gnu\ 17 -Wno-unused-command-line-argument -Wno-unknown-pragmas conftest.c -lpng -ljpeg >&5 undef: _posix_spawn_file_actions_addchdir Undefined symbols for architecture x86_64: "_posix_spawn_file_actions_addchdir", referenced from: _main in 0.x86_64.thinlto.o ld: symbol(s) not found for architecture x86_64 clang: error: linker command failed with exit code 1 (use -v to see invocation) And the final set of defines for configure includes the following: ... #define HAVE_POSIX_SPAWN 1 #define HAVE_POSIX_SPAWN_FILE_ACTIONS_ADDCHDIR_NP 1 #define HAVE_POSIX_SPAWNATTR_SETFLAGS 1 #define HAVE_DECL_POSIX_SPAWN_SETSID 1 ... configure: exit 0 Best regards, Jacob Faibussowitsch (Jacob Fai - booss - oh - vitch) > On Nov 17, 2021, at 11:58, Eli Zaretskii wrote: > >> From: Jacob Faibussowitsch >> Date: Wed, 17 Nov 2021 11:50:46 -0600 >> >> This is by no means breaking or disrupting anything on my end (at least that I can tell), but I did notice in >> some commit messages a few weeks that someone was attempting to replace vfork with posix_spawn >> specifically on macOS, so maybe my system is some unhandled corner-case? > > That change is already on master. If your build of a recent master > branch doesn't use the posix_spawn code, the first place to look is in > config.log, where you should try to find the reason(s) why the > configure script decided your system doesn't have posix_spawn. > > Thanks.