This change was motivated by the desire to remove the weird
dependency of lib-src/profile.o on src/systime.h. profile.c
included systime.h only for current_timespec, and this
inclusion required systime.h to have #ifdef emacs in multiple
places and complicated further changes I have in mind.
The current_timespec decl belongs in timespec.h anyway,
and the main effect of this change is to move it there.
* lib-src/profile.c (INLINE): Remove.
Include timespec.h, not systime.h.
* lib/gettime.c (gettime): Prefer clock_gettime to nanotime,
and don’t worry about it failing on a CLOCK_REALTIME arg.
POSIX requires it to succeed and I don’t know of any
counterexamples where the fallbacks would work.
(current_timespec): Move here from src/systime.h.
Nowadays it seems to be better to not have this function
be inline.
* lib/timespec.h: Include arg-nonnull.h.
(current_timespec): New declaration.
(gettime, settime): Declare args to be nonnull.
* lib/gettime.c, lib/timespec.h: Copy from Gnulib.
* src/systime.h: Simplify by assuming ‘emacs’ is defined,
which it always is now.
(current_timespec): Move to lib/timespec.h.
This incorporates:
2018-05-21 crypto: omit stream ops Emacs doesn’t need
2018-05-13 truncate: Fix compilation error on Android
2018-05-13 imaxdiv: Fix compilation error on Android
2018-05-13 Support selective inclusion of recent mingw.org headers
2018-05-13 Add cross-compilation guesses for Linux systems sans glibc
2018-05-13 stdioext: Fix compilation errors with newer Android headers
2018-05-07 af_alg: Pacify --enable-gcc-warnings
2018-05-06 af_alg: Fix bug with streams that are not at position 0
2018-05-06 Followup to 'af_alg: New module'
2018-05-05 crypto/{md5,sha1,sha256,sha512}: simplify
2018-05-05 af_alg: New module
2018-05-05 af_alg: Improve function signature
2018-04-28 md5sum: Use AF_ALG when available
2018-04-28 sha512sum: Use AF_ALG when available
2018-04-28 sha256sum: Use AF_ALG when available
2018-04-28 sha1sum: Use AF_ALG when available
2018-05-05 all: Replace more http URLs by https URLs
2018-05-03 maint: port more modules to GCC 8
2018-05-03 Simplify code; drop support for Borland C++ on Windows
* admin/merge-gnulib (GNULIB_MODULES): Use crypto/md5-buffer
rather than crypto/md5, since Emacs doesn’t use the stream
operations that in recent Gnulib pull in other stuff Emacs doesn’t
need. Similarly for crypto/sha1-buffer, crypto/sha256-buffer,
crypto/sha512-buffer.
* build-aux/config.guess, build-aux/config.sub, lib/dosname.h:
* lib/dup2.c, lib/errno.in.h, lib/euidaccess.c, lib/fcntl.c:
* lib/fcntl.in.h, lib/fpending.c, lib/fsync.c, lib/getdtablesize.c:
* lib/getopt.c, lib/gettimeofday.c, lib/inttypes.in.h, lib/md5.c:
* lib/md5.h, lib/open.c, lib/pipe2.c, lib/putenv.c, lib/sha1.c:
* lib/sha1.h, lib/sha256.c, lib/sha256.h, lib/sha512.c:
* lib/sha512.h, lib/stat-time.h, lib/stdio-impl.h, lib/stdio.in.h:
* lib/stdlib.in.h, lib/sys_stat.in.h, lib/sys_types.in.h:
* lib/timespec.h, lib/unistd.in.h, lib/utimens.c, m4/c-strtod.m4:
* m4/gnulib-common.m4, m4/inttypes.m4, m4/lstat.m4, m4/nocrash.m4:
* m4/pselect.m4, m4/readlink.m4, m4/stdio_h.m4, m4/symlink.m4:
* m4/unistd_h.m4, m4/utimens.m4:
Copy from Gnulib.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
This incorporates:
2018-01-01 maint: Run 'make update-copyright'
2017-12-29 Add cross-compilation results for GNU/Hurd.
2017-12-12 explicit_bzero: port to macOS + Clang 9.0.0
This incorporates:
2017-10-29 timespec: prefer ‘assume’ to ‘assure’
2017-10-27 timespec.h: use "assure" to avoid a spurious warning
2017-10-09 getopt-posix: Fix build failure if ac_cv_header_getopt_h=no
* build-aux/config.guess, build-aux/config.sub:
* lib/timespec.h, lib/unistd.in.h:
Copy from Gnulib.
This gets Emacs working again with HP-UX Itanium cc.
It incorporates:
2017-03-19 stdalign: tweak version# and test for HP-UX IA64
2017-03-18 stdalign: restore previous behavior for HP-UX IA64
2017-03-17 stat-time, timespec: Support header files in C++ mode
2017-03-17 stdalign: Make it work with HP-UX cc
2017-03-17 flexmember: try to detect HP-UX 11.31 cc bug
2017-03-16 stdint: Fix test compilation failure with HP-UX 11 cc.
2017-03-14 gnulib-tool: don't produce tests with only snippets
2017-03-14 limits-h: Make it work with HP-UX cc.
* etc/PROBLEMS: Remove now-obsolete entry for HP-UX 11.31.
* lib/gnulib.mk.in, m4/gnulib-comp.m4: Regenerate.
* lib/limits.in.h, lib/stat-time.h, lib/stdalign.in.h:
* lib/stdint.in.h, lib/timespec.h, m4/flexmember.m4, m4/stdalign.m4:
Copy from gnulib.
2014-12-12 stddef: support C11's max_align_t
2014-12-08 apply _GL_ATTRIBUTE_PURE to some inline functions
2014-12-02 support GNU format printf and scanf on mingw
* doc/misc/texinfo.tex, lib/stat-time.h, lib/stddef.in.h:
* lib/timespec.h, m4/extensions.m4, m4/stddef_h.m4, m4/stdio_h.m4:
Update from gnulib.
* lib/gnulib.mk: Regenerate.
This incorporates the following changes:
2013-09-19 stdio: OS X port of putc_unlocked + extern inline
2013-09-19 signal: OS X port of sigaddset etc. + extern inline
2013-09-19 extern-inline: do not always suppress extern inline on OS X
2013-09-17 getgroups: statement without effect
2013-08-28 headers: check that _GL_INLINE_HEADER_BEGIN is defined
This portability layer is no longer needed, since Emacs has been
using struct timespec as a portability layer for some time.
Merge from gnulib, incorporating:
2013-08-27 timespec: new convenience constants and function
* src/atimer.h, src/buffer.h, src/dispextern.h, src/xgselect.h:
Include <time.h> rather than "systime.h"; that's all that's needed now.
* src/dispnew.c: Include <timespec.h> rather than "systime.h";
that's all that's needed now.
* src/systime.h (EMACS_TIME): Remove. All uses changed to struct timespec.
(EMACS_TIME_RESOLUTION): Remove. All uses changed to
TIMESPEC_RESOLUTION.
(LOG10_EMACS_TIME_RESOLUTION): Remove. All uses changed to
LOG10_TIMESPEC_RESOLUTION.
(EMACS_SECS, emacs_secs_addr): Remove. All uses changed to tv_sec.
(EMACS_NSECS): Remove. All uses changed to tv_nsec.
(make_emacs_time): Remove. All used changed to make_timespec.
(invalid_timespec): Rename from invalid_emacs_time. All uses changed.
(current_timespec): Rename from current_emacs_time. All uses changed.
(add_emacs_time): Remove. All uses changed to timespec_add.
(sub_emacs_time): Remove. All uses change dot timespec_sub.
(EMACS_TIME_SIGN): Remove. All uses changed to timespec_sign.
(timespec_valid_p): Rename from EMACS_TIME_VALID_P. All uses changed.
(EMACS_TIME_FROM_DOUBLE): Remove. All uses changed to dtotimespec.
(EMACS_TIME_TO_DOUBLE): Remove. All uses changed to timespectod.
(current_timespec): Rename from current_emacs_time. All uses changed.
(EMACS_TIME_EQ, EMACS_TIME_LT, EMACS_TIME_LE): Remove. All uses
changed to timespec_cmp.
* src/xgselect.c: Include <timespec.h>, since our .h files don't.
2012-08-01 extern-inline: new module
2012-08-01 stat-time, timespec, u64, utimens: use extern-inline
* lib/stat-time.c, lib/utimespec.c, lib/u64.c, m4/extern-inline.m4:
New files. The new .c files are for instantiating extern inline
functions.
Fixes: debbugs:12116
lib/makefile.w32-in ($(BLD)/dtotimespec.$(O)):
($(BLD)/timespec-add.$(O)):
($(BLD)/timespec-sub.$(O)): Don't depend on
$(EMACS_ROOT)/nt/inc/sys/time.h.
lib/stat-time.h:
lib/timespec.h:
lib/utimens.h: Revert last change.
src/makefile.w32-in (TIMESPEC_H): Remove nt/inc/sys/time.h.
(SYSTIME_H): Add nt/inc/sys/time.h.
src/systime.h [WINDOWSNT]: Include sys/time.h.
src/s/ms-w32.h (struct timespec): Definition moved from
nt/inc/sys/time.h. Suggested by Paul Eggert <eggert@cs.ucla.edu>.
nt/inc/sys/time.h (struct timespec): Don't define it here, it is
now defined in src/s/ms-w32.h.
Fixes: debbugs:9000
