* src/nsterm.m (ns_draw_window_cursor): Test glyph width vs cursor width
before setting final size.

* src/insdel.c (adjust_markers_bytepos): New function.
* src/lisp.h (adjust_markers_bytepos): Add prototype.
* src/insdel.c (replace_range, replace_range_2):
* src/editfns.c (Ftranspose_regions): Call
adjust_markers_bytepos.  (Bug#5131)

* test/src/editfns-tests.el (transpose-test-reverse-word)
(transpose-test-get-byte-positions): New functions.
(transpose-ascii-regions-test)
(transpose-nonascii-regions-test-1)
(transpose-nonascii-regions-test-2): New tests.

* lisp/image.el (image-animate-timeout): Fix the logic that tests if
an animation is too big (bug#24004).

* lisp/emulation/viper-cmd.el (viper-adjust-undo): Use it.
(viper-set-complex-command-for-undo): Save current state with
prepare-change-group.
* lisp/emulation/viper-init.el (viper-undo-needs-adjustment)
(viper-buffer-undo-list-mark): Remove.

* src/emacs.c (main) [WINDOWSNT]: Move init_environment calls after the
set_initial_environment call.  This prevents Emacs' modifications to the
environment from contaminating Vprocess_environment and
Vinitial_environment (Bug #10980).
* src/callproc.c (getenv_internal) [WINDOWSNT]: Consult Emacs' internal
environment in as a fallback to Vprocess_environment.
* test/src/callproc-tests.el (initial-environment-preserved): New Test.

* lisp/emacs-lisp/cl-seq.el (cl-reduce): Explain what reducing means.

This reverts commit 5811404f.

It doesn't have the consensus, as evidenced by
http://lists.gnu.org/archive/html/emacs-devel/2016-06/msg00138.html

This reverts commit 001d88b6.

It doesn't have the consensus, as evidenced by
http://lists.gnu.org/archive/html/emacs-devel/2016-06/msg00138.html

* src/nsmenu.m (pop_down_menu, ns_popup_dialog): Remove references to
autorelease pool and Popdown_data struct.

* lisp/hexl.el (hexl-mode):
* lisp/ielm.el (inferior-emacs-lisp-mode):
* lisp/progmodes/cfengine.el (cfengine3-mode):
* lisp/progmodes/elisp-mode.el (emacs-lisp-mode):
* lisp/progmodes/octave.el (octave-mode, inferior-octave-mode):
* lisp/progmodes/python.el (python-mode):
* lisp/simple.el (read--expression): Add buffer-locally to hook
eldoc-documentation-functions.

* lisp/international/ucs-normalize.el (ucs-normalize-combining-chars-regexp):
(quick-check-list-to-regexp): Use regexp-opt-charset instead of
regexp-opt.
* lisp/international/ucs-normalize.el (quick-check-list): Reuse a single
temp buffer for the whole loop.

Some tests are marked as expected to fail.

* test/lisp/international/ucs-normalize-tests.el: New tests.
* admin/unidata/NormalizationTest.txt: Add data for tests.
* admin/unidata/README: Add URL for NormalizationTest.txt.
* admin/notes/unicode: Add note about running (and updating the data
for) the new tests.  Remove note about normalization being unsupported.

* src/xdisp.c (get_phys_cursor_geometry): Fix invalid C operator.

* src/bytecode.c (PUSH): Remove workaround for long-obsolete compiler.

It is no longer needed now that we assume GC_MARK_STACK ==
GC_MAKE_GCPROS_NOOPS.
* src/bytecode.c (BYTE_MAINTAIN_TOP): Remove.
All uses removed, and code simplified accordingly.
(BEFORE_POTENTIAL_GC, AFTER_POTENTIAL_GC): Remove, since they
are always no-ops now.  All uses removed.
(MAYBE_GC): Remove.  All uses replaced by maybe_gc,
since it is now equivalent.

* src/bytecode.c: Remove comments that are no longer accurate.
Most of these are actually old ChangeLog entries.

Inspired by a suggestion by Florian Weimer in:
https://sourceware.org/ml/libc-alpha/2016-07/msg00425.html
* configure.ac (HAVE_PERSONALITY_ADDR_NO_RANDOMIZE):
Rename from HAVE_PERSONALITY_LINUX32, and check for
ADDR_NO_RANDOMIZE (the crucial thing) instead of for LINUX32.
All uses changed.
* src/emacs.c (main) [HAVE_PERSONALITY_ADDR_NO_RANDOMIZE]:
Use ADDR_NO_RANDOMIZE from personality.h rather than inventing the
flag ourselves.  Just set that flag, rather than also setting the
persona.  When doing it, avoid functions like putenv that may
allocate memory.

This reverts commit bdda4855.

Fixes bug #23944.

* lisp/progmodes/cc-engine.el (c-literal-limits): make the sixth argument of
an invocation of parse-partial-sexp 'syntax-table, not the fourth.

* configure.ac ("$hybrid_malloc" = yes): Set GNU_MALLOC to no,
parenthesize value of GNU_MALLOC_reason.

* lisp/sort.el (delete-duplicate-lines): Delete duplicate
first line when operating backward (bug#23863).

* Makefile.in (base_obj): Remove $(SHEAP_OBJ).

Use shr-image-map instead of shr-map (bug#23964).

* lisp/cus-edit.el (mouse): Remove both group definitions.

* lisp/emacs-lisp/autoload.el (update-directory-autoloads): Do not
exclude preloaded libraries or remove entries generated for them.
(autoload-generate-file-autoloads): Do not generate autoload
statements for preloaded libraries.

* src/xwidget.c: Remove #include directives that are not needed.
Also, don’t conditionalize #includes on HAVE_X_WINDOWS, since this
file is compiled only if the X interface is available.

* src/xwidget.c (x_draw_xwidget_glyph_string, xwidget_end_redisplay):
Adjust to pacify GCC.  Add a couple of FIXME comments, suggesting
possible bugs found by GCC.

* lisp/progmodes/grep.el (grep-compute-defaults): Quote braces
in all commands to be passed to a shell (Bug#23959).

This reverts commit 20038f8a.

I am reverting this change because it was applied without prior discussion
on emacs-devel, and has been found to break the NS port.  It needs more
testing and review before it should be applied here.

(cl-prog, cl-prog*): New macros.

* lisp/emacs-lisp/cl-generic.el (cl-defmethod): Make docstring dynamic.
(cl--generic-make-defmethod-docstring): New function for that.
(cl-defmethod, cl-generic-generalizers): Tweak docstrings accordingly.
(cl-generic-define-method, cl--generic-describe): Change `load-history'
format of cl-defmethods, so as not to confused methods with equal
specializers but different qualifiers.
* lisp/emacs-lisp/eieio-core.el (cl-generic-generalizers): Provide docstrings.

* src/w32fns.c (setup_w32_kbdhook): Don't install the keyboard
hook if we are being debugged.  This avoids hosing the debugger,
because the hook is global, and is called in the context of the
thread which installed it, and that thread is stopped when GDB has
control.  Reported by Fabrice Popineau <fabrice.popineau@gmail.com>.

Problems were caused by such a string spanning jit-lock chunks, and by a flaw
in the +-500 bytes boundaries imposed for macros.

* lisp/progmodes/cc-mode.el (c-extend-region-for-CPP): Check the +-500 byte
macro boundaries here.
(c-extend-font-lock-region-for-macros): Remove the check on the +-500 byte
lower boundary.  Fix the check on the upper boundary.

* lisp/progmodes/cc-fonts.el (c-font-lock-raw-strings): Handle the starting
point already being within a raw string.

* lisp/progmodes/cc-engine.el (c-raw-string-pos)
(c-depropertize-raw-strings-in-region, c-after-change-re-mark-raw-strings):
Modify regexp element "\\{,16\\}" to "\\{0,16\\}" for greater compatibility
with other Emacsen.

* src/dispextern.h (toplevel): Remove 'tip_frame' and 'tip_window' decls.
* src/frame.h (struct frame): New bitfield to indicate tooltip frame.
(FRAME_TOOLTIP_P): New macro.
* src/frame.c (make_frame): Mark new frame as regular frame by default.
(Fframe_list, delete_frame): Redesign to use FRAME_TOOLTIP_P.
(syms_of_frame): Don't DEFSYM 'Qtooltip' but use 'Qtooltip_timer'
and 'Qtooltip_parameters' instead.
* src/gtkutil.h (toplevel): Add 'xg_hide_tip' decl.
* src/gtkutil.c (xg_hide_tip): New function.
(xg_hide_tooltip): Adjust to cancel GTK event loop timeout if needed.
* src/menu.c (Fx_popup_menu): Adjust call to Fx_hide_tip.
* src/nsfns.c (toplevel): Remove 'tip_frame' leftover.
* src/w32fns.c (unwind_create_tip_frame): Remove.
(w32_display_monitor_attributes_list)
(w32_display_monitor_attributes_list_fallback): Use FRAME_TOOLTIP_P.
(toplevel): Remove 'tip_frame', 'tip_window' and 'last_show_tip_args'.
(x_create_tip_frame): Use do_unwind_create_frame.  Mark new frame
as a tooltip frame and record it using appropriate display info.
(x_hide_tooltip): Add frame arg.
(Fx_show_tip): Adjust to avoid globals, store tooltip parameters among
base frame parameters, store tooltip hide timer among tooltip frame
parameters.
(Fx_hide_tip): Add frame arg, hide tooltips on all displays by default.
(syms_of_w32fns): Don't DEFSYM 'Qtip_frame', don't initialize
and GC-protect 'tip_timer', 'tip_frame' and 'last_show_tip_args'.
* src/w32term.c (w32_read_socket): Extract tooltip window id from
per-display data.  Use FRAME_TOOLTIP_P where appropriate.
(x_new_font): Use FRAME_TOOLTIP_P.
(x_free_frame_resources): Reset pointer to tooltip frame.
* src/w32term.h (struct w32_display_info): New member 'w32_tooltip_frame'.
* src/xdisp.c (init_iterator, x_consider_frame_title, prepare_menu_bars):
Use FRAME_TOOLTIP_P.
* src/xfns.c (x_make_monitor_attribute_list)
(Fx_display_monitor_attributes_list): Likewise.
(unwind_create_tip_frame): Remove.
(toplevel): Remove 'tip_frame', 'tip_window' and 'last_show_tip_args'.
(x_create_tip_frame): Use do_unwind_create_frame.  Mark new frame
as a tooltip frame and record it using appropriate display info.
(x_hide_tooltip): Add frame arg.
(Fx_show_tip): Adjust to avoid globals, store tooltip parameters among
base frame parameters, store tooltip hide timer among tooltip frame
parameters.  To hide GTK system tooltip, use timeout hooked into GTK
event loop.
(Fx_hide_tip): Add frame arg, hide tooltips on all displays by default.
(syms_of_xfns): Don't DEFSYM 'Qtip_frame', don't initialize
and GC-protect 'tip_timer', 'tip_frame' and 'last_show_tip_args'.
* src/xterm.c (handle_one_xevent): Extract tooltip window id from
per-display data.  Use FRAME_TOOLTIP_P where appropriate.
(x_new_font, x_set_window_size): Use FRAME_TOOLTIP_P.
(x_free_frame_resources): Reset pointer to tooltip frame.
* src/xterm.h (struct x_display_info): New member 'x_tooltip_frame'.
(struct x_output) [USE_GTK_TOOLTIP]: New member 'ttip_timeout'.

* test/src/fns-tests.el (fns-tests-mapcan): New test.