GitLab

POSIX does not specify timersub, and marks gettimeofday as
obsolescent.  Avoid porting problems by using timespec.h
functions instead.
* src/editfns.c: Include systime.h instead of sys/time.h.
(EXTRA_CONTEXT_FIELDS): Replace start and max_secs with
time_limit.  All uses changed.  This removes the need to call
gettimeofday or timersub.
* src/term.c (timeval_to_Time): Remove.  Replace all uses with ...
(current_Time): ... this new function, removing the need to
call gettimeofday.

* src/editfns.c (Freplace_buffer_contents): Add two optional arguments
  for mitigating performance issues.
* lisp/emacs-lisp/subr-x.el (replace-region-contents): Move from
  subr.el.  Add the same two arguments as for replace-buffer-contents.
* lisp/json.el (json-pretty-print-max-secs): New variable holding the
  default MAX-SECS value json-pretty-print passes to
  replace-buffer-contents.
  (json-pretty-print): Use it.
* doc/lispref/text.texi (Replacing): Add documentation for
  replace-buffer-contents two new optional arguments.  Document
  replace-region-contents.

* src/w32term.c (x_make_frame_visible): Call unblock_input
before returning early.  (Bug#34575)

* src/minibuf.c (Finternal_complete_buffer): Add
Qcycle_sort_function to completion table's metadata.
(syms_of_minibuf): New symbol Qcycle_sort_function.

* src/profiler.c (evict_lower_half, record_backtrace)
(setup_cpu_timer, cmpfn_profiler, hashfn_profiler):
Assume C99.  Use bool for boolean.
(timer_getoverrun): Remove; simplify use to not need it.
(Fprofiler_cpu_start): Any negative return from setup_cpu_timer fails.
(Fprofiler_cpu_stop): Simplify initialization.

* configure.ac: Add a check for the 'timer_getoverrun' function.

* src/profiler.c [CYGWIN] : Define 'timer_getoverrun' as a
macro only on versions of Cygwin where it is not already
defined as a function.

* src/w32.c (getpwnam): Support usernames provided through
LOGNAME or USERNAME environment variables, to mimic what
editfns.c:init_editfns does.

* src/keyboard.c (read_char, read_event_from_main_queue):
Ensure the global value of getcjmp is restored when the stack
is unwound by the likes of 'throw', by calling
record_unwind_protect_ptr instead of restoring the value
manually.  (Bug#34394)
(restore_getcjmp): Argument is now 'void *', to match the
signature of record_unwind_protect_ptr.

(cherry picked from commit 10527fca)

* src/keyboard.c (read_char, read_event_from_main_queue):
Ensure the global value of getcjmp is restored when the stack
is unwound by the likes of 'throw', by calling
record_unwind_protect_ptr instead of restoring the value
manually.  (Bug#34394)
(restore_getcjmp): Argument is now 'void *', to match the
signature of record_unwind_protect_ptr.

Revert part of ed962f2b.

* test/src/data-tests.el (data-tests-make-local-forwarded-var):
Add corresponding test.

* src/emacs.c (main) [HAVE_NTGUI]: Remove code that is only
for WINDOWSNT or condition it on WINDOWSNT.  (Bug#34409)

* src/editfns.c (Freplace_buffer_contents): Use lower value of
  too_expensive and enable heuristic.
* lisp/subr.el (replace-region-contents): New convenient wrapper
  function around replace-buffer-contents.

* src/callproc.c (Fcall_process, Fcall_process_region): Minor
fixes to doc strings.  Suggested by Nicholas Drozd
<nicholasdrozd@gmail.com>.  (Bug#34274)

* src/ftfont.h (struct font_info): New type, unifies similar types
from ftcrfont.c, ftfont.c and xftfont.c
* src/xftfont.c (struct xftfont_info): Remove, replace with struct
font_info.  Adjust all uses.
* src/ftcrfont.c (struct ftcrfont_info): Likewise.
* src/ftfont.c (struct ftfont_info): Likewise.

* src/process.c (Fprocess_thread): Correct docstring.

* src/emacs.c (main): Fix assertions and logic for pdump
loading in builds that can both unexec and pdump.

* src/emacs.c (main) [WINDOWSNT]: Fix logic of using dynamic
heap in unexec case.  (Bug#34277)

* src/emacs.c (main) [WINDOWSNT]: Allow to invoke temacs
interactively without the --temacs= option.

* src/coding.c (syms_of_coding): New symbol Qus_ascii.
(reset_coding_after_pdumper_load): Call
set-safe-terminal-system-internal to set up
safe_terminal_coding after restoring from pdump file.
Reported by Philippe Vaucher <philippe.vaucher@gmail.com>.

* src/xdisp.c (redisplay_internal): Set the
inhibit_clear_image_cache flag of a frame while its windows
are being redisplayed, and reset the flag after the call top
update_frame returns.
* src/image.c (clear_image_cache): Do nothing if the frame's
inhibit_clear_image_cache flag is set.  (Bug#34256)
* src/frame.h (struct frame): New flag inhibit_clear_image_cache.

* src/w32proc.c (LOCALE_IPAPERSIZE): Define if undefined.
(nl_langinfo): Support _NL_PAPER_WIDTH and _NL_PAPER_HEIGHT
like glibc does.
* src/fns.c (Flocale_info): Update the doc string.

* nt/inc/langinfo.h: Add _NL_PAPER_WIDTH and _NL_PAPER_HEIGHT
to the enumeration.
(_NL_PAPER_WIDTH, _NL_PAPER_HEIGHT): Define namesake macros.
* nt/mingw-cfg.site (emacs_cv_langinfo__nl_paper_width): Set
to 'yes'.

* doc/lispref/nonascii.texi (Locales): Update the
documentation of 'locale-info' for the argument of 'paper'.

* etc/NEWS: Update the locale-info entry.

This fixes Bug#34193

* src/process.c (Fmake_network_process): Explicitly use ::1 when
using IPv6 with 'local.  Update docstring.

* test/lisp/net/network-stream-tests.el
(make-ipv6-tcp-server-with-unspecified-port):
(make-ipv6-tcp-server-with-specified-port): Test creating ipv6
local server.
(make-server): Add optional family argument, default ipv4
(echo-server-with-local-ipv4): Test connecting to 'local ipv4
(echo-server-with-local-ipv6): Test connecting to 'local ipv6

* doc/lispref/processes.texi (Network Processes): Describe
behavior when using 'local.

* etc/NEWS: Document new 'make-network-process' behavior when
connecting to 'local with ipv6.

* configure.ac (HAVE_LANGINFO__NL_PAPER_WIDTH): New macro.
* src/fns.c (Flocale_info) [HAVE_LANGINFO__NL_PAPER_WIDTH]:
Get paper width and height from locale.

* src/editfns.c (Fuser_full_name):
Don’t assume uid fits into fixnum.

* src/emacs.c (main) [WINDOWSNT]: Fix logic of determining by
argv[0] whether to use the static heap.  (Bug#34277)

* src/xdisp.c (redisplay_internal): Run window change
functions before updating the display so changes induced by
these functions can get caught by redisplay (Bug#34138).
* src/window.c (run_window_change_functions): Bind
Qinhibit_redisplay to avoid that the minibuffer window gets
resized while running window change functions (Bug#34179,
Bug#34260).

* configure.ac (CANNOT_DUMP): Remove.  All uses removed,
or changed to the negative of with_unexec.
(HAVE_UNEXEC): New macro.  All uses of CANNOT_DUMP changed
to the negative of this macro.

* src/charset.c (charset_table_size):
* src/pdumper.c (dump_private):
* src/sysdep.c (init_sigsegv):
* src/window.c (old_selected_window):
Now static.
* src/charset.c (charset_table_size):
Now int, since the value always fits in int.
* src/gtkutil.c (xg_gtk_initialized): Now present only if
HAVE_XWIDGETS, to make it clearer that this is an xwidgets
hack.  All uses changed.
* src/lread.c (ndefsubr): Remove; unused.
* src/pdumper.h: Use usual GNU indenting style for functions,
since my static-vs-extern checking hack relies on it.
(dump_public): Always declare; simpler and doesn’t hurt.
(pdumper_handle_page_fault): Remove unused decl.

Widen modification counts to at least 64 bits, to make
wraparound practically impossible.
* doc/lispref/buffers.texi (Buffer Modification):
Don’t say the modification-count can wrap around.
* src/buffer.c (Frestore_buffer_modified_p, Fbuffer_swap_text)
(modify_overlay):
* src/insdel.c (insert_1_both, insert_from_string_1)
(insert_from_gap, insert_from_buffer_1)
(adjust_after_replace, replace_range, replace_range_2)
(del_range_2, modify_text):
* src/textprop.c (modify_text_properties):
Use modiff_incr instead of incrementing	directly.
(Fbuffer_modified_tick, Fbuffer_chars_modified_tick):
Don’t assume modification counts fit into fixnums.
* src/buffer.h (struct buffer_text, struct buffer):
* src/cmds.c (internal_self_insert):
* src/fileio.c (Finsert_file_contents):
* src/indent.c (last_known_column_modified):
* src/keyboard.c (command_loop_1):
* src/marker.c (cached_modiff):
* src/syntax.c (find_start_modiff, parse_sexp_propertize)
(find_defun_start):
* src/window.h (struct window):
Use modiff_count for modification counts.
* src/editfns.c (Fsubst_char_in_region):
Copy instead of incrementing modification counts,
since integer overflow checking is not needed here.
* src/lisp.h (modiff_count): New type.
(modiff_incr, modiff_to_integer): New inline functions.
* src/pdumper.c (dump_buffer): Update hash.

* src/pdumper.c (dump_roots, pdumper_load): Simplify initialization.
(dump_bitset_init, pdumper_load): Omit unnecessary assignments.
(dump_bitset_destroy): Remove; never called.  All callers removed.
(dump_do_dump_relocation, pdumper_load): Add FIXME comment.
(pdumper_load): Simplify by assuming C99.  Remove unused local.

This fixes a regression from Emacs 25.3 (Bug#34134).
* src/process.c (server_accept_connection):
Set host correctly, fixing a bug introduced in
2017-09-16T21:29:18Z!eggert@cs.ucla.edu
when working around a GCC bug.

Else build number increments without limit even in clean builds.

* src/xfaces.c (init_xfaces): Don't rely of 'face'
property of a face to be a natural number.

* src/xfaces.c (init_xfaces): New function.
* src/emacs.c (main) [HAVE_PDUMPER]: If dumped with pdumper,
call init_xfaces.  (Bug#34226)
* src/lisp.h (init_xfaces) [HAVE_PDUMPER]: Add prototype.

* test/lisp/faces-tests.el (faces--test-face-id): New test for
bug#34226.

* src/cmds.c (Fforward_line): Support bignum arg.
(scan_newline): Return void since no caller was using the
return value.
* src/search.c (find_newline, scan_newline_from_point)
(find_newline1): Return the number of newlines counted, not
the count shortage, so that the return value always fits in
ptrdiff_t even if the original count was a bignum.  All
callers changed.
* test/src/cmds-tests.el (forward-line-with-bignum): New test.