As after each function call GCC clobbers the pointer to the function
relocation table. This commit modify the code generation to create a
local copy of it for each function. This reduces the average number
of loads for each function call into C from two to one.
* src/comp.c (comp_t): Add 'func_relocs_ptr_type' and
'func_relocs_local' fields.
(emit_call): Use the local func_relocs pointer when possible.
(emit_ctxt_code): Fill 'comp.func_relocs_ptr_type'.
(compile_function): Declare 'func_relocs_ptr_local'.
(compile_function): Assign 'func_relocs_ptr_local' from the global
value in each function prologue.
We used to store module runtime and environment pointers in the static
lists Vmodule_runtimes and Vmodule_environments. However, this is
incorrect because these objects have to be kept per-thread. With this
naive approach, interleaving module function calls in separate threads
leads to environments being removed in the wrong order, which in turn
can cause local module values to be incorrectly garbage-collected.
The fix isn't completely trivial: specbinding the lists wouldn't work
either, because then the garbage collector wouldn't find the
environments in other threads than the current ones, again leading to
objects being garbage-collected incorrectly. While introducing custom
pseudovector types would fix this, it's simpler to put the runtime and
environment pointers into the specbinding list as new specbinding
kinds. This works since we need to unwind them anyway, and we only
ever treat the lists as a stack. The thread switching machinery
ensures that the specbinding lists are thread-local, and that all
elements of the specbinding lists in all threads are marked during
garbage collection.
Module assertions now have to walk the specbinding list for the
current thread, which is more correct since they now only find
environments for the current thread. As a result, we can now remove
the faulty Vmodule_runtimes and Vmodule_environments variables
entirely.
Also add a unit test that exemplifies the problem. It interleaves two
module calls in two threads so that the first call ends while the
second one is still active. Without this change, this test triggers
an assertion failure.
* src/lisp.h (enum specbind_tag): Add new tags for module runtimes and
environments.
* src/eval.c (record_unwind_protect_module): New function to record a
module object in the specpdl list.
(do_one_unbind): Unwind module objects.
(backtrace_eval_unrewind, default_toplevel_binding, lexbound_p)
(Fbacktrace__locals): Deal with new specbinding types.
(mark_specpdl): Mark module environments as needed.
* src/alloc.c (garbage_collect): Remove call to 'mark-modules'.
Garbage collection of module values is now handled as part of marking
the specpdl of each thread.
* src/emacs-module.c (Fmodule_load, funcall_module): Use specpdl to
record module runtimes and environments.
(module_assert_runtime, module_assert_env, value_to_lisp): Walk
through specpdl list instead of list variables.
(mark_module_environment): Rename from 'mark_modules'. Don't attempt
to walk though current thread's environments only, since that would
miss other threads.
(initialize_environment, finalize_environment): Don't change
Vmodule_environments variable; environments are now in the specpdl
list.
(finalize_environment_unwind, finalize_runtime_unwind): Make 'extern'
since do_one_unbind now calls them.
(finalize_runtime_unwind): Don't change Vmodule_runtimes variable;
runtimes are now in the specpdl list.
(syms_of_module): Remove Vmodule_runtimes and Vmodule_environments.
* test/data/emacs-module/mod-test.c (Fmod_test_funcall): New test
function.
(emacs_module_init): Bind it.
* test/src/emacs-module-tests.el (emacs-module-tests--variable): New
helper type to guard access to state in a thread-safe way.
(emacs-module-tests--wait-for-variable)
(emacs-module-tests--change-variable): New helper functions.
(emacs-module-tests/interleaved-threads): New unit test.
For certain patterns, re-matching the same regexp on the matched
substring does not produce correctly translated match data
(bug#15107 and bug#44861).
Using a new builtin function also improves performance since the
number of calls to string-match is halved.
Reported by Kevin Ryde and Shigeru Fukaya.
* lisp/subr.el (replace-regexp-in-string): Translate the match data
using match-data--translate instead of trusting a call to string-match
on the matched string to do the job.
* test/lisp/subr-tests.el (subr-replace-regexp-in-string):
Add test cases.
* src/search.c (Fmatch_data__translate): New internal function.
(syms_of_search): Register it as a subroutine.
6442cdc0e4 Revert extra focus redirection in do_switch_frame (Bug#24803)
fc4379f1ae Minor cleanup of tramp-tests.el on MS Windows
dea3d6aa18 Fix handling of defcustom :local tag
* src/frame.c (do_switch_frame): Do not also redirect frame
focus when FRAME has its minibuffer window on the selected
frame which was intended to fix Bug#24500. It may cause
Bug#24803 and lead to a nasty state where no active cursor is
shown on any frame, see
https://lists.gnu.org/archive/html/emacs-devel/2020-11/msg01137.html.
* src/Makefile.in (temacs$(EXEEXT)): Codesign the executable on
recent (ARM) MacOS systems (bug#43878). Without this, building
Emacs fails.
Copyright-paperwork-exempt: yes
86cbc9d216 Make ignoring modifiers on IME input optional
32b97bb9e0 Ignore modifiers when processing WM_IME_CHAR messages
f641ef1a07 Improve documentation of 'font-spec'
# Conflicts:
# etc/NEWS
* src/minibuf.c (read_minibuf): In the record_unwind_protect for the second
restore_window_configuration (separate minibuffer frame case) arrange for the
future switching back to the original frame by Fset_window_configuration.
* src/keymap.c (describe_vector): Say which command shadows this
binding. (Bug#9293)
* test/src/keymap-tests.el
(help--describe-vector/bug-9293-one-shadowed-in-range): Adapt
test.
* src/keymap.c (describe_vector): Do not say binding is shadowed if
the other key binding points to the same command. (Bug#9293)
* test/src/keymap-tests.el
(help--describe-vector/bug-9293-same-command-does-not-shadow): New
test.
* src/keymap.c (describe_vector): Make sure found consecutive keys
are either not shadowed or, if they are, that they are shadowed by
the same command. (Bug#9293)
* test/src/keymap-tests.el
(help--describe-vector/bug-9293-one-shadowed-in-range): New test.
By default, ignore modifier keys on IME input, but add
a variable to get back old behavior.
* src/w32fns.c (syms_of_w32fns): New variable
w32-ignore-modifiers-on-IME-input.
(w32_wnd_proc): Use it to ignore modifier keys when IME input is
used. (Bug#44641)
* etc/NEWS: Announce the change and the new variable.
* lisp/emacs-lisp/comp.el (native--compile-async): New defun extracted
from native-compile-async.
(native-compile-async): Remove load argument and use above new defun.
* src/comp.c (maybe_defer_native_compilation): Use above new
defun. (Bug#44676)
* src/image.c (rsvg_handle_get_dimensions, init_svg_functions):
Make 'rsvg_handle_get_dimensions' available and defined for all
versions of librsvg. (Bug#44655)
Set the default value when `set` encounters a PER_BUFFER variable
which has been let-bound globally, to match the behavior seen with
`make-variable-buffer-local`.
* test/src/data-tests.el (binding-test--let-buffer-local):
Add corresponding test.
(data-tests--set-default-per-buffer): Add tentative test for the
performance problem encountered in bug#41029.
All-ASCII strings cannot have substrings with non-ASCII characters in
them; use this fact to avoid searching entirely.
* src/fns.c (Fstring_search): For multibyte non-ASCII needle and
unibyte haystack, don't check if the haystack is all-ASCII; it's a
waste of time. For multibyte non-ASCII needle and multibyte
all-ASCII haystack, fail immediately.
* test/src/fns-tests.el (string-search): Add more test cases.
This was problematic when minibuffer-follows-selected-frame was non-nil.
Introduce a new parameter DONT-SET-FRAME to set-window-configuration.
* doc/lispref/windows.texi (Window Configurations): Describe the new &optional
parameter to set-window-configuration.
* etc/NEWS (Lisp Changes): Note the new parameter to set-window-configuration.
* src/keyboard.c (read_char_help_form_unwind): Add a new Qnil argument to the
call of Fset_window_configuration.
* src/minibuf.c (read_minibuf): Cons up a Qt with the window configuration in
the argument to record_unwind_protect for the window configuration (twice).
* src/window.c (Fset_window_configuration): Add the new &optional parameter
and document it in the doc string. At the final do_switch_frame operation,
restore the original frame when DONT-SET-FRAME is non-nil.
(restore_window_configuration): Handle the new parameter when the supplied
argument is a cons.
19da602991 Fix input method translation near read-only text
5aabf2cc7f Fix display of truncated R2L lines on TTY frames
daff3bda10 Avoid crashes when a reversed glyph row starts with a comp...
b697bb91a1 ; * .gitignore: src/fingerprint.c not generated since 2019...
* src/xterm.c (x_term_init):
* src/w32term.c (w32_initialize_display_info): Sync initial frame
title with new value of Vframe_title_format.
Problem reported by Angelo Graziosi <angelo.g0@libero.it>.
* src/comp.c (Fcomp__compile_ctxt_to_file): Now that we do not
rely anymore on globlal variables move logic in from
'Fcomp__init_ctxt' so comp.debug is already set correctly.
* src/process.c (init_process_emacs): force glib's g_unix_signal
handler into lib_child_handler where it should belong.
Copyright-paperwork-exempt: yes
We capitalize all other messages during the dump, so capitalize the
"dump mode" and "dumping fingerprint" ones as well for consistency.
* src/pdumper.c (Fdump_emacs_portable): Capitalize fingerprint message
prefix.
* lisp/loadup.el: Capitalize "dump mode" message.
* src/dispnew.c (update_mouse_position): New function for mouse
movement logic in 'handle_one_term_event' that can be shared across
different mouse backends.
(display--update-for-mouse-movement): New lisp function, call it.
* lisp/xt-mouse.el (xterm-mouse--handle-mouse-movement): New function
that calls 'display--update-for-mouse-movement'.
(xterm-mouse-translate-1): Call it.
* src/term.c (handle_one_term_event): Inline logic from
'term_mouse_movement' and call 'update_mouse_position'.
(term_mouse_movement): Delete.
* src/xdisp.c (extend_face_to_end_of_line): Use a while-loop, not
a do-while loop, to avoid appending an extra glyph at the end of a
line that is one character shorter than the window-width. This is
needed to fix display of reversed glyph rows that are almost as
wide as the window, because append_space_for_newline already added
one space glyph.
* src/dispnew.c (build_frame_matrix_from_leaf_window): Add an
assertion to prevent us from overwriting non-char glyphs with the
vertical border glyph.
* src/xdisp.c (extend_face_to_end_of_line): Account for one glyph
possibly inserted by append_space_for_newline. (Bug#44506)
Remove a kludgey correction for an off-by-one error in column
counting, which is no longer needed.
