Performance issue reported by Eli Zaretskii (Bug#41321#149).
* src/alloc.c (GC_OBJECT_ALIGNMENT_MINIMUM): New constant.
(maybe_lisp_pointer): Use it instead of GCALIGNMENT.
* src/lisp.h (struct Lisp_Float): Declare via
GCALIGNED_UNION_MEMBER, not via GCALIGNED_STRUCT, since alloc.c
creates these in arrays and GCALIGNED_STRUCT does not necessarily
suffice to align struct Lisp_Float when it’s used in an array.
This avoids undefined behavior on oddball machines where
sizeof (struct Lisp_Float) is not a multiple of 8 and the compiler
does not support __attribute__ ((aligned 8)).
* src/alloc.c (union emacs_align_type): Move to here ...
* src/lisp.h: ... from here, and uncomment out some of the
types that alloc.c can see but lisp.h cannot.
* src/alloc.c (MALLOC_ALIGNMENT_BOUND): Remove.
(LISP_ALIGNMENT): Go back to yesterday’s version, except use
union emacs_align_type instead of max_align_t.
(MALLOC_IS_LISP_ALIGNED): Go back to yesterday’s version.
(maybe_lisp_pointer): Check against GCALIGNMENT, not LISP_ALIGNMENT.
* src/lisp.h (union emacs_align_type): Bring back.
This is simpler, and fixes a bug in the previous fix.
* src/alloc.c (MALLOC_ALIGNMENT_BOUND): Simplify by
using max_align_t, since the buggy implementations won’t
break this simpler implementation.
(LISP_ALIGNMENT): Simplify by just using GCALIGNMENT, since the
fancier implementation wasn’t correct anyway, and fixing it
isn’t worth the trouble on practical platforms.
* src/lisp.h (union emacs_align_type): Remove.
Problem reported by Eli Zaretskii (Bug#41321).
* src/alloc.c (MALLOC_ALIGNMENT_BOUND): New constant.
(LISP_ALIGNMENT): Lower it to avoid crashes on MinGW and similarly
buggy platforms where malloc returns pointers not aligned to
alignof (max_align_t). But keep it higher on platforms where this
is known to work, as it helps GC performance.
(MALLOC_IS_LISP_ALIGNED): Define in terms of the other two.
* src/alloc.c (stacktop_sentry):
* src/thread.c (run_thread):
Don’t overalign or oversize stack sentries; they need to be
aligned only for pointers and Lisp_Object, not for arbitrary
pseudovector contents.
* src/lisp.h (union emacs_align_type): New type, used for
LISP_ALIGNMENT.
d6a0b66a0c (origin/emacs-27) * lisp/subr.el (save-match-data): Clarif...
1a6d59eeba Improve the documentation of setting up fontsets
c7737d40f2 ; * etc/TODO (Ligatures): Update the entry based on recent...
fb2e34cd21 ; * etc/TODO (Ligatures): Update the entry based on recent...
13b6dfd4f7 * doc/emacs/killing.texi (Rectangles): Improve indexing.
a10254dd46 Fix accessing files on networked drives on MS-Windows
* lisp/version.el: Don’t put an RCS Id style string into the
executable via purecopy, as this does not work with the pdumper.
* src/emacs.c (RCS_Id): New constant, for 'ident'.
(cherry picked from commit 3d1bcfba5e)
Check Lisp_Compiled objects better as they’re created,
so that the byte-code interpreter needn’t do the checks
each time it executes them. This improved performance
of ‘make compile-always’ by 1.5% on my platform. Also,
improve the quality of the (still-incomplete) checks, as
this is more practical now that they’re done less often.
* src/alloc.c (make_byte_code): Remove. All uses removed.
(Fmake_byte_code): Put a better (though still incomplete)
check here instead. Simplify by using Fvector instead
of make_uninit_vector followed by memcpy, and by using
XSETPVECTYPE instead of make_byte_code followed by XSETCOMPILED.
* src/bytecode.c (Fbyte_code): Do sanity check and conditional
translation to unibyte here instead of each time the function is
executed.
(exec_byte_code): Omit no-longer-necessary sanity and
unibyte checking. Use SCHARS instead of SBYTES where
either will do, as SCHARS is faster.
* src/eval.c (fetch_and_exec_byte_code): New function.
(funcall_lambda): Use it.
(funcall_lambda, lambda_arity, Ffetch_bytecode):
Omit no-longer-necessary sanity checks.
(Ffetch_bytecode): Add sanity check if actually fetching.
* src/lisp.h (XSETCOMPILED): Remove. All uses removed.
* src/lread.c (read1): Check byte-code objects more thoroughly,
albeit still incompletely, and do translation to unibyte here
instead of each time the function is executed.
(read1): Use XSETPVECYPE instead of make_byte_code.
(read_vector): Omit no-longer-necessary sanity check.
* src/fns.c (Ffillarray, Fclear_string):
Add CHECK_IMPURE here, to be consistent with Faset etc.
(Ffillarray): Prefer memset when the fill is a single byte.
* lisp/version.el: Don’t put an RCS Id style string into the
executable via purecopy, as this does not work with the pdumper.
* src/emacs.c (RCS_Id): New constant, for 'ident'.
It has been missing since 2012-07-03 (Emacs 24.3)
"Cleanup basic buffer management", when undo_list was moved to
the end of struct buffer. (Bug#33492)
* src/buffer.c (buffer_local_variables_1): New function.
(Fbuffer_local_variables): Explicitly add buffer-undo-list.
* src/alloc.c (SET_STACK_TOP_ADDRESS): Do not call
__builtin_unwind_init.
(flush_stack_call_func1): Rename from 'flush_stack_call_func'.
(flush_stack_call_func): New function to spill all registers
before calling 'flush_stack_call_func1'. This to make sure the
top of the stack identified includes those registers.
b4937f64cd (origin/emacs-27) Improve documentation of manually instal...
efd4e973a4 Reflect the emacs-devel ELPA/MELPA dispute in FAQ
28541674cd Consider face inheritance when checking region face backgr...
e75f6be6cc Fix dired default file operation (bug#41261)
406fb0746c Fix documentation related to 'command-switch-alist'.
747e0a2523 Improve ediff readability in misterioso theme (Bug#41221)
48830c73e7 Fix a crash in handle_display_spec
a37290a6f9 In x_hide_tip reset tip_last_frame for GTK+ tooltips only ...
3d81995692 Fix docstring of flymake-make-diagnostic (bug#40351)
632aa9d57a Go back to “Bahá’í”
e2406ff60f * lisp/dired.el (dired-toggle-marks): Doc fix. (Bug#41097)
# Conflicts:
# doc/emacs/building.texi
Pacify GCC 10.1.0 so that it does not issue false alarms
when Emacs is configured with --enable-gcc-warnings.
* src/dispnew.c (clear_glyph_row):
* src/fns.c (hash_clear):
* src/keyboard.c (append_tab_bar_item):
* src/lisp.h (vcopy):
* src/xfaces.c (get_lface_attributes_no_remap)
(Finternal_copy_lisp_face, realize_default_face):
* src/xmenu.c (set_frame_menubar):
Work around -Warray-bounds false alarm in GCC 10.1.0.
* src/intervals.c (copy_properties):
Avoid -Wnull-dereference false alarm in GCC 10.1.0.
* src/lisp.h (xvector_contents_addr, xvector_contents):
New functions, useful for working around GCC bug 95072.
be0d1cac83 (origin/emacs-27) Small fix for type of 'display-fill-colu...
c5e5839776 Fix customization of 'display-fill-column-indicator-charac...
d5c184aa3e Refer to fill column indicator Info node in some places.
e13300ae50 Merge branch 'emacs-27' of git.sv.gnu.org:/srv/git/emacs i...
0bae57033f Fix GTK's Tool Bar menu radio buttons
4c98aa7ea5 Minor clarifications in NEWS
a1cbd05f38 Improve documentation of 'with-suppressed-warnings'.
4a895c1b26 Fix a typo in a comment
2caf3e997e Improve documentation of Hi Lock mode
7081c1d66f Fix typos in the Emacs user manual
0385771e2f Fix references to Speedbar in VHDL mode
a76cafea0d Fix handling of FROM = t and TO = t by 'window-text-pixel-...
# Conflicts:
# etc/NEWS
# src/xdisp.c
* src/xdisp.c (syms_of_xdisp): Add reference to manual in
documentation strings for variables related to fill column indicators.
* lisp/display-fill-column-indicator.el (display-fill-column-indicator)
(display-fill-column-indicator-mode): Add reference to manual.
* src/textprop.c (Fnext_single_char_property_change): Clarify in
the doc string the behavior when LIMIT is past the end of OBJECT.
Stop the search when position gets to end of buffer, for when LIMIT
is beyond that. (Bug#40000)
Problem reported by Andreas Schwab in:
https://lists.gnu.org/r/emacs-devel/2020-05/msg00650.html
* src/conf_post.h (HAS_ATTR_no_sanitize): Define to false in case
cpp is picky, fixing a longstanding glitch here.
(ATTRIBUTE_NO_SANITIZE_ADDRESS, ATTRIBUTE_NO_SANITIZE_UNDEFINED):
Use HAS_ATTRIBUTE, not __has_attribute.
attribute.h is partly designed for C2X forward compatibility,
since C2X will add some standard attributes. Using its macros
should help insulate Emacs from C2X teething problems.
* src/conf_post.h: Include attribute.h.
(HAS_ATTRIBUTE, HAS_FEATURE): Rename from __has_attribute and
__has_feature, to avoid polluting the builtin namespace.
All uses changed.
(ATTRIBUTE_COLD, ATTRIBUTE_FORMAT, FALLTHROUGH, ATTRIBUTE_CONST)
(ATTRIBUTE_PURE, ATTRIBUTE_UNUSED, ATTRIBUTE_MAY_ALIAS)
(ATTRIBUTE_MALLOC, ATTRIBUTE_ALLOC_SIZE)
(ATTRIBUTE_RETURNS_NONNULL): Remove, as attribute.h does this now.
(NO_INLINE, EXTERNALLY_VISIBLE, ARG_NONNULL, ATTRIBUTE_UNUSED):
Simplify by defining in terms of attribute.h macros.
* src/systhread.h (ATTRIBUTE_WARN_UNUSED_RESULT): Remove.
All uses replaced by attribute.h’s NODISCARD.
1f17193e00 Expand file name for remote dirs as well
7a12ab5ea2 Fix project.el commands in "transient" projects
274ec97e3c Make sure alist-related functions say so in their doc
It is conceptually wrong for apropos commands that search doc
strings to look for matches of several words only on the same
line, because division of doc strings between lines is
ephemeral.
* lisp/apropos.el (apropos-parse-pattern): Accept an optional
argument MULTILINE-P, and if that is non-nil, produce regexps that
match words in the list even if they are separated by line
boundaries.
(apropos-value, apropos-local-value, apropos-documentation): Use
the new optional argument in apropos commands that search
multiline text, such as doc strings.
* src/search.c (Fposix_looking_at, Fposix_string_match)
(Fposix_search_backward, Fposix_search_forward): Make sure Posix
appears in the doc strings near REGEXP, for better matches.
* src/term.c (encode_terminal_code): Each character from an
automatic composition is a multibyte character, so its multibyte
representation can take up to MAX_MULTIBYTE_LENGTH bytes.
Account for that when allocating storage for characters to be
encoded. (Bug#40913)
* src/character.h (multibyte_length):
Merge tests so that there are fewer conditional branches.
This improved CPU speed of ‘make compile-always’
by about 1.5% on my platform.
* src/editfns.c (Fstring_to_char): Simplify.
* src/editfns.c (Fstring_to_char): Simplify.
This tweak improved the CPU time performance of
‘make compile-always’ by about 1.8% on my platform.
This tweak improved the CPU time performance of
‘make compile-always’ by about 1.7% on my platform.
* src/character.c (string_char): Remove; no longer used.
* src/character.h (string_char_and_length): Redo so that it
needn’t call string_char. This helps the caller, which can now
become a leaf function.
This reclaims a bit of performance when compiling with gcc -Og.
These functions were macros until I changed them in
2020-04-17T14:57:25Z!eggert@cs.ucla.edu.
* src/casefiddle.c (make_char_unibyte):
* src/ccl.c (GET_TRANSLATION_TABLE): Now inline.
45a64c97c7 (origin/emacs-27) Clarify semantics of trace-function CONT...
821760fdc4 Don't let a code literal get modified in mml parsing (Bug#...
74a92be16d * lisp/simple.el (kill-ring-save): Doc fix. (Bug#40797)
3d0e859692 Minor doc clarification regarding fringe bitmaps
4d86c7f822 Fix documentation of fringe bitmaps
a76af88dd8 Tweak mutability doc a bit more
f7e488d206 Calc: fix autoload errors (bug#40800)
369761b36d ; * src/xdisp.c: Improve the introductory commentary.
a92ca1f177 Improve indexing of ELisp manual
5a25d17760 * lisp/image-mode.el (image-transform-resize): Remove FIXM...
37ebec3a95 Improve the default value of 'doc-view-ghostscript-program'.
ba6104d1e8 Change doc-view-mode-map prefix key 's' to 'c'.
400ff5cd19 Improve wording about constants
d2836fe71b Improve the default value of 'doc-view-ghostscript-program'.
fc55f65305 Minor improvements in documentation of the last change
a64da75961 Add image-auto-resize defcustoms to image-mode.el
692ad40539 Improve the documentation of tab-bar and tab-line
# Conflicts:
# etc/NEWS
Without a call to image_find_image, we can get a file name that
is relative to data-directory/images/, or a file name that
starts with "~/", in which case w32_load_image would fail.
* src/image.c (native_image_load): Call image_find_image_file to
resolve and encode the image file name.
* src/w32image.c (w32_load_image): No need to encode the file
name, as it's already encoded by native_image_load.
