The buffer-local-versions of 'window-buffer-change-functions',
'window-size-change-functions',
'window-selection-change-functions' and
'window-state-change-functions' are now run with the respective
buffer temporarily current. Also, the local version of
'window-buffer-change-functions' is run for the buffer removed
from the window too.
* src/window.c (run_window_change_functions_locally)
(run_window_change_functions_globally): New functions replacing
'run_window_change_functions_1'.
(run_window_change_functions): Run the buffer local versions of
these hooks in their respective buffers. Run
'window-buffer-change-functions' for the buffer removed from the
window too.
(Vwindow_buffer_change_functions, Vwindow_size_change_functions)
(Vwindow_selection_change_functions)
(Vwindow_state_change_functions): Mention that the buffer-local
versions are run with their buffer temporarily current.
* doc/lispref/windows.texi (Window Hooks): Mention that
buffer-local-versions of window change functions are run with
their buffer temporarily current. Also say that
'window-buffer-change-functions' will be run for removed buffer
too.
* etc/NEWS: Advertise changes for the buffer-local versions of
window change functions.
In “Avoid duplicate calls to current_timespec” (2015-07-05)
we started caching current_timespec results in NOW.
However, this was buggy: we updated NOW only when the timeout was
nonzero, but the timeout can be set temporarily to zero in several
places in wait_reading_process_output (such as when checking for
process status changes), which would cause us to never update NOW
and therefore never detect that a timeout happened.
Also, this caching was wrong even in principle: since we call
Lisp code from wait_reading_process_output, substantial amounts
of time can pass, and we can be left using an outdated NOW and
incorrectly not time out. Also, nowadays we can use
monotonic_coarse_timespec which is fast, and which is better
anyway because it’s immune to manual clock changes.
Co-authored-by: Spencer Baugh <sbaugh@janestreet.com>
* src/process.c (wait_reading_process_output):
Stop caching the current realtime.
Instead, use the coarse monotonic clock without caching.
They are good enough for this purpose, and are cheaper to get.
* src/timefns.c (monotonic_coarse_timespec): New function.
* src/xterm.c [HAVE_XSYNC && !USE_GTK && HAVE_CLOCK_GETTIME]:
(x_sync_current_monotonic_time): Use it.
(CLOCK_MONOTONIC): Remove; no longer uneeded here.
* src/comp.c (ABI_VERSION): Bump new version
(comp_t): add 'eq'.
(helper_link_table): add 'slow_eq'.
(emit_slow_eq): New function.
(emit_EQ): Rework to reflect EQ implementation.
(declare_runtime_imported_funcs): Import 'slow_eq'.
(Fcomp__init_ctxt): Register emitter 'emit_eq' for op code Qeq.
This and the previous 2 commits increase elisp-benchmarks performance by
3.6% on my test machine. Also a (small) reduction in eln size comes with
it.
* doc/lispref/processes.texi (Accepting Output):
* src/process.c (Faccept_process_output): Document better the
meaning of the timeout of 'accept-process-output' a,d the fact
that it doesn't always return as soon as some output is available.
See https://lists.gnu.org/archive/html/emacs-devel/2025-08/msg00750.html
for more details.
* src/bytecode.c (exec_byte_code): Reset quitcounter to 1 after longjmp;
the exact value isn't important. This may reduce register pressure a
tiny bit or at least remove a useless stack slot.
This rids us of a bunch of unhygienic macros with free variables and
makes the marking code actually readable again. Even better, it is all
processed by the compiler even when the checks are disabled.
* src/alloc.c (CHECK_ALLOCATED, CHECK_LIVE, CHECK_ALLOCATED_AND_LIVE)
(CHECK_ALLOCATED_AND_LIVE_SYMBOL): Transform macros into...
(check_live, check_allocated_and_live, check_allocated_and_live_symbol)
(check_allocated_and_live_vectorlike): ...functions. Callers adapted.
Reloading a byte-compiled file when we fail to fetch some
docstring in it is not guaranteed to work and can cause
subtle bugs which are worse than simply not having a
docstring. Add 'documentation-dynamic-reload' variable to
allow disabling this behavior. See the discussion in
https://lists.gnu.org/archive/html/emacs-devel/2025-08/msg00304.html
for more details.
* src/doc.c (syms_of_doc): Add 'documentation-dynamic-reload'.
(Fdocumentation, Fdocumentation_property): Check value of
'documentation-dynamic-reload'.
Correct a few edge cases where we used the keymaps at point when
looking up keymaps for an event position which is outside the
current buffer. Namely:
- Clicking on a part of the mode line which is after the end of
mode-line-format produces an event with non-nil posn-area but
nil posn-string.
- Even if posn-string doesn't have a local keymap, we should
still ignore the keymaps at point if posn-string is non-nil.
* src/keymap.c (Fcurrent_active_maps): Ignore keymaps at point
for more positions outside the buffer. (bug#76620)
* lisp/emacs-lisp/shorthands.el (hack-read-symbol-shorthands):
Avoid inf-loops during bootstrap.
E.g. this can occur while loading `uni-special-lowercase.el` where
`hack-local-variables--find-variables` uses `downcase` which triggers
loading `uni-special-lowercase.el`, ...
* lisp/international/mule-cmds.el (ucs-names): Explicitly require
`charprop`.
* src/Makefile.in ($(lispsource)/loaddefs.el): Depend on `charprop`.
The doc strings for re-search-forward and re-search-backward have been
improved (bug#25193, bug#31584) but the corresponding posix- and
non-regexp search functions still have the original text, despite
describing the exact same mechanism.
There is no reason for them to differ, so we make them all use the
updated version which also is more readable. Any future doc string
improvements should be made to all of them in the same way for
consistency.
* src/search.c (Fsearch_backward, Fsearch_forward)
(Fposix_search_backward, Fposix_search_forward):
Use text from Fre_search_forward and Fre_search_backward.
* src/process.c (server_accept_connection): Make the client
process be locked to the same thread as the parent server process,
or unlocked if the server process was unlocked. (Bug#79367)
* src/indent.c (vertical-motion): If iterator is inside an overlay,
reset it to the beginning of line before trying to reach goal
column. This prevents point from being stuck at the beginning
of overlay strings during upward motions.
Copyright-paperwork-exempt: yes
* src/xdisp.c (next_element_from_display_vector): Only switch the
iterator from unibyte to multibyte, never in the other direction,
and not if the original character came from a unibyte buffer.
(Bug#79317)
The following changes make the changes for bug#79275 less
radical, closer to their previous shape, while still fixing
that bug.
* src/xdisp.c (push_prefix_prop, get_it_property): Restore
original code that determined the object and position on it.
(get_line_prefix_it_property): Take FROM_BUFFER from the actual
object of the prefix property.
Only allow string mutation that is certain not to require string data to
be resized and reallocated: writing bytes into a unibyte string, and
changing ASCII to ASCII in a multibyte string.
This ensures that mutation will never transform a unibyte string to
multibyte, that the size of a string in bytes never changes, and that
the byte offsets of characters remain the same. Most importantly, it
removes a long-standing obstacle to reform of string representation and
allow for future performance improvements.
* src/data.c (Faset): Disallow resizing string mutation.
* src/fns.c (clear_string_char_byte_cache):
* src/alloc.c (resize_string_data): Remove.
* test/src/data-tests.el (data-aset-string): New test.
* test/lisp/subr-tests.el (subr--subst-char-in-string):
Skip error cases.
* test/src/alloc-tests.el (aset-nbytes-change):
Remove test that is no longer relevant.
* doc/lispref/strings.texi (Modifying Strings):
* doc/lispref/sequences.texi (Array Functions):
* doc/lispref/text.texi (Substitution): Update manual.
* etc/NEWS: Announce.
* src/xdisp.c (push_it): Reset the 'string_from_prefix_prop_p' flag.
(try_window_id): Disable this optimization if the last unchanged
at-beg row begins with a display or overlay string and there;s a
line/wrap-prefix property on the row.
(push_prefix_prop): Accept an additional argument FROM_BUFFER to
indicate that the prefix property was found on buffer text
underlying a display or overlay property, and set up the position
to pop to accordingly. Reset the 'string_from_display_prop_p' flag
of the iterator after pushing IT to set up for iterating the
prefix string.
(get_it_property): Use it->string, not it->object, as indication
that prefix property is on a string.
(get_line_prefix_it_property): Accept an additional argument:
pointer to a flag indicating that the prefix property was found on
buffer text underlying a display or overlay property. Callers
adjusted.
(handle_line_prefix): Use the FROM_BUFFER flag to correctly handle
prefix properties on buffer text at the same position as a display
string. (Bug#79275)
We use the current parsing position instead. The line and column in the
error weren't used (nor very accurate to begin with) and the user can
easily compute them when needed. The line number calculation is kept
just in case but deprecated, for removal in Emacs 32.
* src/json.c (struct json_parser, json_parser_init): Update parser state.
(json_signal_error): New position computation.
(json_skip_whitespace_internal): Remove.
(is_json_whitespace): New.
(json_skip_whitespace, json_skip_whitespace_if_possible)
(json_parse_unicode, json_parse_string, json_parse_number)
(json_parse_value): Simplify and rewrite for efficiency.
(count_chars, count_newlines)
(string_byte_to_pos, string_byte_to_line)
(buffer_byte_to_pos, buffer_byte_to_line): New.
(Fjson_parse_string, Fjson_parse_buffer): Adapt to new parser state.
* test/src/json-tests.el (json-tests--parse-string-error-pos)
(json-tests--parse-buffer-error-pos, json-parse-error-position): New.
* etc/NEWS: Note deprecation of line and column.
end_of_file_error previously always signaled end-of-file with
load-true-file-name if that was non-nil (and a string).
However, this might be the wrong thing to do; for example, if a
file being loaded calls read on a buffer.
* src/lread.c (end_of_file_error): <source>: New argument; check
it to determine what data to signal with. (bug#68546)
(read_char_escape, read_char_literal, read_string_literal)
(skip_space_and_comments, read0): Pass source to
end_of_file_error.
.waiting_thread and .thread could be left set to non-NULL values
in a deleted fd_callback_info entry. These would never be
cleared by e.g. clear_waiting_thread_info since that only clears
fd_callback_info entries up to max_desc. Clear fd_callback_info
entirely when deleting an entry.
* src/process.c (clear_fd_callback_data): Add.
(delete_write_fd, delete_keyboard_wait_descriptor): Call
clear_fd_callback_data. (bug#79201)
(delete_read_fd): Remove duplicated clearing code.
(deactivate_process): Remove duplicate recompute_max_desc.
* lisp/treesit.el (treesit-enabled-modes): New user option.
* src/treesit.c (treesit-major-mode-remap-alist): New variable.
* lisp/progmodes/c-ts-mode.el:
* lisp/progmodes/csharp-mode.el:
* lisp/progmodes/java-ts-mode.el:
* lisp/progmodes/js.el:
* lisp/progmodes/json-ts-mode.el:
* lisp/progmodes/ruby-ts-mode.el:
* lisp/progmodes/sh-script.el:
* lisp/textmodes/css-mode.el:
* lisp/textmodes/mhtml-ts-mode.el:
* lisp/textmodes/toml-ts-mode.el:
Add ts-mode mapping to 'treesit-major-mode-remap-alist'
for ts-modes that already have the corresponding non-ts mode
association in 'auto-mode-alist'.
* lisp/progmodes/cmake-ts-mode.el (cmake-ts-mode-maybe):
* lisp/progmodes/dockerfile-ts-mode.el (dockerfile-ts-mode-maybe):
* lisp/progmodes/elixir-ts-mode.el (elixir-ts-mode-maybe):
* lisp/progmodes/go-ts-mode.el (go-ts-mode-maybe)
(go-mod-ts-mode-maybe, go-work-ts-mode-maybe):
* lisp/progmodes/heex-ts-mode.el (heex-ts-mode-maybe):
* lisp/progmodes/lua-ts-mode.el (lua-ts-mode-maybe):
* lisp/progmodes/php-ts-mode.el (php-ts-mode-maybe):
* lisp/progmodes/rust-ts-mode.el (rust-ts-mode-maybe):
* lisp/progmodes/typescript-ts-mode.el (typescript-ts-mode-maybe)
(tsx-ts-mode-maybe):
* lisp/textmodes/markdown-ts-mode.el (markdown-ts-mode-maybe):
* lisp/textmodes/yaml-ts-mode.el (yaml-ts-mode-maybe):
Add a wrapper function to 'auto-mode-alist'
for ts-modes that have no corresponding non-ts mode.
Also add a mapping to 'treesit-major-mode-remap-alist'
for the case when a non-ts mode is installed from an external
source to be able to customize it with 'treesit-enabled-modes'.
* src/inotify.c (inotify_callback): Set event.frame_or_window to Qnil.
* src/keyboard.c (make_lispy_event): Make a better check for the
HAVE_W32NOTIFY case when creating a Qfile_notify event.
Fixes bug#74074, bug#79164.
* src/nsterm.m ([EmacsView windowDidEndLiveResize:]):
([EmacsView updateFramePosition]): New functions.
([EmacsView windowDidMove:]): Move contents of this function into
updateFramePosition and call it.
Copyright-paperwork-exempt: yes
* src/thread.c (Fmake_thread): Add new argument (bug#76969).
(thread_set_error): New function, extracted from thread-signal.
(Fthread_buffer_disposition): Add getter.
(Fthread_set_buffer_disposition): And setter.
(thread_check_current_buffer): Check the values of threads'
buffer_disposition.
(thread_all_before_buffer_killed): New function.
(init_threads): Set buffer_disposition to nil for the main thread.
(syms_of_threads): Add new symbols and define the error.
* src/thread.h (thread_state): New field buffer_disposition.
(thread_all_before_buffer_killed): Declare.
* src/buffer.c (Fkill_buffer): Call thread_check_current_buffer
one more time after all hooks and after that call
thread_all_before_buffer_killed.
* src/comp.c (ABI_VERSION): Increase the value.
* test/src/thread-tests.el (thread-buffer-disposition-t)
(thread-buffer-disposition-nil)
(thread-buffer-disposition-silently)
(thread-set-buffer-disposition)
(thread-set-buffer-disposition-main-thread): New tests.
* doc/lispref/threads.texi (Basic Thread Functions): Document
buffer-disposition in make-thread and its getter and setter.
* etc/NEWS: Add entry.
* src/print.c (print_create_variable_mapping): Fix a typo,
`unreadeable-function' to `unreadable-function'. Add `symbols-bare'
as an override for `print-symbols-bare'. (Bug#79161)
