ChangeLog 322 KB
Newer Older
1 2 3 4 5
2011-10-29  Glenn Morris  <rgm@gnu.org>

	* process.c (wait_reading_process_output): Revert 2009-08-30 change.
	Not clear what it was for, and it causes various bugs.  (Bug#9839)

6 7 8 9 10 11
2011-10-28  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (note_mouse_highlight): Initialize `part', to avoid a
	possible random value that matches one of those tested as
	condition to clear the mouse face.

12 13 14 15
2011-10-28  Chong Yidong  <cyd@gnu.org>

	* xdisp.c (note_mouse_highlight): Fix use of uninitialized var.

16 17 18 19
2011-10-28  Dan Nicolaescu  <dann@ics.uci.edu>

	* window.c (make_window): Initialize phys_cursor_on_p.

20 21 22 23
2011-10-28  Stefan Monnier  <monnier@iro.umontreal.ca>

	* lisp.h (struct Lisp_Symbol): Update comments.

24 25 26 27
2011-10-28  Juanma Barranquero  <lekktu@gmail.com>

	* w32font.c (w32_load_unicows_or_gdi32): Add missing return.

28 29 30 31 32 33 34 35 36 37 38
2011-10-28  Eli Zaretskii  <eliz@gnu.org>

	Fix Emacs on Windows 9X (bug#8562).  Thanks to oslsachem
	<oslsachem@gmail.com> for helping to debug this.

	* w32font.c (g_b_init_is_w9x, g_b_init_get_outline_metrics_w)
	(g_b_init_get_text_metrics_w, g_b_init_get_glyph_outline_w)
	(g_b_init_get_glyph_outline_w): New static variables.
	(GetOutlineTextMetricsW_Proc, GetTextMetricsW_Proc)
	(GetGlyphOutlineW_Proc): New typedefs.
	(w32_load_unicows_or_gdi32, get_outline_metrics_w)
39 40 41 42
	(get_text_metrics_w, get_glyph_outline_w, globals_of_w32font):
	New functions.
	(w32font_open_internal, compute_metrics):
	Call get_outline_metrics_w, get_text_metrics_w, and get_glyph_outline_w
43 44 45 46 47 48
	instead of calling the "wide" APIs directly.

	* emacs.c (main) [HAVE_NTGUI]: Call globals_of_w32font.

	* w32.h (syms_of_w32font): Add prototype.

49 50 51 52 53 54 55
2011-10-27  Juanma Barranquero  <lekktu@gmail.com>

	* window.c (Fframe_root_window, Fframe_first_window, Fwindow_end)
	(Fframe_selected_window, Ftemp_output_buffer_show, Fnext_window)
	(Fdelete_window_internal, Fwindow_parameters): Fix typos in docstrings.
	(Fmove_to_window_line): Doc fix.

Chong Yidong's avatar
Chong Yidong committed
56 57 58 59 60 61 62 63 64 65 66
2011-10-27  Chong Yidong  <cyd@gnu.org>

	* process.c (make_process): Set gnutls_state to NULL.

	* gnutls.c (emacs_gnutls_deinit): Deinit the gnutls_state if it is
	non-NULL, regardless of GNUTLS_INITSTAGE.
	(Fgnutls_boot): Cleanups.  Call emacs_gnutls_deinit if we signal
	an error.  Set process slots as soon as we allocate them.

	* gnutls.h (GNUTLS_LOG, GNUTLS_LOG2): Fix macros.

67 68
2011-10-27  Chong Yidong  <cyd@gnu.org>

69 70
	* gnutls.c (emacs_gnutls_deinit): New function.
	Deallocate credentials structures as well as calling gnutls_deinit.
71 72 73 74 75
	(Fgnutls_deinit, Fgnutls_boot): Use it.

	* process.c (make_process): Initialize GnuTLS credentials to NULL.
	(deactivate_process): Call emacs_gnutls_deinit.

76 77 78 79 80 81 82
2011-10-27  Juanma Barranquero  <lekktu@gmail.com>

	* image.c (x_create_x_image_and_pixmap):
	* w32.c (sys_rename, w32_delayed_load):
	* w32font.c (fill_in_logfont):
	* w32reg.c (x_get_string_resource): Silence compiler warnings.

83 84 85 86
2011-10-26  Juanma Barranquero  <lekktu@gmail.com>

	* w32fns.c (w32_default_color_map): New function,
	extracted from Fw32_default_color_map.
Juanma Barranquero's avatar
Juanma Barranquero committed
87
	(Fw32_default_color_map, Fx_open_connection): Use it.  (Bug#9785)
88

89 90 91 92
2011-10-25  Paul Eggert  <eggert@cs.ucla.edu>

	* dispextern.h (Fcontrolling_tty_p): New decl (Bug#6649 part 2).

93 94 95 96 97
2011-10-25  Stefan Monnier  <monnier@iro.umontreal.ca>

	* keyboard.c (test_undefined): New function (bug#9751).
	(read_key_sequence): Use it to detect when a key is bound to `undefined'.

98 99 100 101 102
2011-10-25  Enami Tsugutomo  <tsugutomo.enami@jp.sony.com>

	* sysdep.c (init_sys_modes): Fix the check for the controlling
	terminal (Bug#6649).

103 104 105 106 107 108 109 110 111 112 113 114 115
2011-10-20  Eli Zaretskii  <eliz@gnu.org>

	* dispextern.h (struct bidi_it): New member next_en_type.

	* bidi.c (bidi_line_init): Initialize the next_en_type member.
	(bidi_resolve_explicit_1): When next_en_pos is valid for the
	current character, check also for next_en_type being WEAK_EN.
	(bidi_resolve_weak): Don't enter the expensive loop if the current
	position is before next_en_pos.  Record the bidi type of the first
	non-ET, non-BN character we find, in addition to its position.
	(bidi_level_of_next_char): Invalidate next_en_type when
	next_en_pos is over-stepped.

116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136
2011-10-20  Paul Eggert  <eggert@cs.ucla.edu>

	Time zone name fixes for non-ASCII locales (Bug#641, Bug#9794)
	* editfns.c: Rewrite current-time-zone so that it invokes
	the equivalent of (format-time-string "%Z") to get the time zone name.
	This fixes a bug when the time zone name contains characters that
	need converting from the system time locale to Emacs internal format.
	This fixes a shortcoming that I introduced in my 1999-10-19 patch:
	that patch fixed format-time-string to do the conversion, but
	I forgot to fix current-time-zone.
	(format_time_string): New function, containing most of
	what Fformat_time_string used to contain.
	(Fformat_time_string): Rewrite in terms of format_time_string.
	This doesn't change this function's behavior.
	(current-time-zone): Rewrite to use format_time_string.
	This fixes the bug reported by Michael Schierl in
	<http://lists.gnu.org/archive/html/emacs-devel/2007-06/msg00334.html>.
	Jason Rumney's 2007-06-07 change worked around this bug, but
	didn't fix it.
	* systime.h (tzname, timezone): Remove no-longer-used declarations.

137 138 139 140
2011-10-19  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (start_display): If the character at POS is displayed
	via a display vector, reset IT->current.dpvec_index to zero.
141 142 143 144
	(try_window_reusing_current_matrix): If a line ends in a display
	vector or the next line starts in a display vector, continue
	redrawing the window even though the character position of
	start_row was reached.
145 146
	(Bug#9771, part 2)

147 148 149 150 151
2011-10-18  Chong Yidong  <cyd@gnu.org>

	* xdisp.c (get_next_display_element): Handle U+2010 and U+2011
	with nobreak-char-display too.

152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
2011-10-18  Eli Zaretskii  <eliz@gnu.org>

	Fix part 3 of bug#9771.
	* bidi.c (bidi_line_init): Initialize next_en_pos to zero, not -1.
	(bidi_resolve_neutral): Don't enter the expensive loop looking for
	non-neutral characters if the current character is a paragraph
	separator (a.k.a. Newline).  This avoids running the same
	expensive loop twice, once when we consume the preceding newline
	and the other time when the line actually needs to be displayed.
	Avoid the loop when we see neutrals on the base embedding level
	following a character whose directionality is the same as the
	paragraph's.  This avoids running the expensive loop when a line
	ends in a long sequence of neutrals, like control characters.
	Add assertion against STRONG_AL type.  Slightly rearrange code
	that determines the type of a neutral given the first non-neutral
	that follows it.
	(bidi_level_of_next_char): Set next_en_pos to zero when
	invalidating its info.

171 172 173 174 175
2011-10-17  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (push_display_prop): Determine whether to record string
	or buffer position by IT->string, not by IT->method.  Allow
	GET_FROM_DISPLAY_VECTOR as IT->method on entry.  (Bug#9771, part 4)
176 177 178
	(move_it_vertically_backward): Don't look for character position
	immediately after the newline when in a continuation line.
	(Bug#9771, part 1)
179

180 181 182 183 184
2011-10-15  Martin Rudalics  <rudalics@gmx.at>

	* window.c (coordinates_in_window): Rewrite and delabelize
	vertical border check.  (Bug#5357) (Bug#9618)

185 186 187 188 189
2011-10-14  Stefan Monnier  <monnier@iro.umontreal.ca>

	* xterm.c (frame_highlight, frame_unhighlight): Ignore unexplained
	errors in XSetWindowBorder (bug#9310).

190 191 192 193 194
2011-10-13  Dmitry Antipov  <dmantipov@yandex.ru>

	* editfns.c (Fset_time_zone_rule): Replace free with xfree to
	avoid crash when xmalloc overrun checking is enabled.

195 196 197 198 199 200 201 202 203
2011-10-13  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (Fcurrent_bidi_paragraph_direction): Initialize
	itb.paragraph_dir to NEUTRAL_DIR.  Fixes an occasional incorrect
	cursor motion with <left> and <right> arrow keys.

	* bidi.c (bidi_init_it): Don't initialize paragraph_dir here, as
	some callers set that themselves.

204 205 206 207 208 209
2011-10-12  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (find_row_edges): Handle the case where ROW comes from a
	display string and the previous row comes from the same string and
	is empty.  (Bug#9739)  (Bug#9738)

210 211 212 213
2011-10-12  Stefan Monnier  <monnier@iro.umontreal.ca>

	* doc.c (get_doc_string): Encode file name (bug#9735).

214 215
2011-10-12  Eli Zaretskii  <eliz@gnu.org>

216 217 218 219
	* bidi.c (bidi_level_of_next_char):
	* xdisp.c (get_visually_first_element): Remove old incorrect
	comments regarding the Unicode Line Separator character.

220 221
	* bidi.c (bidi_init_it): Initialize paragraph_dir to NEUTRAL_DIR.

222 223 224 225 226 227
2011-10-12  Dmitry Antipov  <dmantipov@yandex.ru>

	* alloc.c (Fgc_status): Do not access beyond zombies array
	boundary if nzombies > MAX_ZOMBIES.
	* alloc.c (dump_zombies): Add missing format specifier.

228 229
2011-10-12  Paul Eggert  <eggert@cs.ucla.edu>

230 231 232
	* xdisp.c (set_cursor_from_row): Simplify conditionals,
	to pacify GCC 4.6.1 x86-64 with -O2 -Wstrict-overflow.

233 234 235
	* lread.c (read_escape): Allow hex escapes as large as ?\xfffffff.
	Some packages use them to denote characters with modifiers.

236 237 238 239 240 241
2011-10-11  Andreas Schwab  <schwab@linux-m68k.org>

	* lisp.h (GCPRO1_VAR, GCPRO2_VAR, GCPRO3_VAR, GCPRO4_VAR)
	(GCPRO5_VAR, GCPRO6_VAR, UNGCPRO_VAR): Add whitespace to avoid
	matching a pp-number.  Rename parameter var to var1.

242 243 244 245
2011-10-11  Stefan Monnier  <monnier@iro.umontreal.ca>

	* minibuf.c (Finternal_complete_buffer): Fix last change (bug#9709).

Glenn Morris's avatar
Glenn Morris committed
246 247 248 249 250
2011-10-08  Glenn Morris  <rgm@gnu.org>

	* callint.c (Fcall_interactively): Give a more explicit error for the
	'c' case with a non-character input.  (Bug#8479)

251 252
2011-10-08  Eli Zaretskii  <eliz@gnu.org>

253 254
	* xdisp.c (hscroll_window_tree): Support hscroll in right-to-left
	lines.
255 256
	(set_cursor_from_row): Fix cursor positioning in mixed L2R+R2L
	lines that are hscrolled on the left.
257

258 259 260
	* dispnew.c (buffer_posn_from_coords): Account for a possible
	presence of header-line.  (Bug#4426)

261 262
2011-10-07  Stefan Monnier  <monnier@iro.umontreal.ca>

263 264
	* buffer.c (syms_of_buffer) <enable-multibyte-characters>:
	Don't advertise functionality which we discourage or doesn't work.
265

266 267 268 269 270 271 272 273 274 275 276 277
2011-10-07  Paul Eggert  <eggert@cs.ucla.edu>

	* alloc.c (GC_LISP_OBJECT_ALIGNMENT): Use offsetof, not __alignof__
	or sizeof.  __alignof__ gives the wrong answer on Fedora x86-64
	with GCC 4.6.1 when configured with CC='gcc -m32' --with-wide-int;
	this makes Emacs dump core during garbage collection on rare
	occasions.  sizeof is obviously inferior to offsetof here, so
	stick with offsetof.
	(GC_POINTER_ALIGNMENT): New macro.
	(mark_memory): Omit 3rd (offset) arg; caller changed.
	Don't assume EMACS_INT alignment is the same as pointer alignment.

278 279 280 281 282 283
2011-10-03  Stefan Monnier  <monnier@iro.umontreal.ca>

	* keyboard.c (read_key_sequence_remapped): New var.
	(read_key_sequence): Compute remapping in the right buffer.
	(command_loop_1): Use read_key_sequence's remapping directly.

284 285
2011-10-02  Stefan Monnier  <monnier@iro.umontreal.ca>

286 287 288 289
	* dired.c (file_name_completion): Don't expand file name.
	(Ffile_name_completion, Ffile_name_all_completions): Expand file name
	before checking file name handler.

290 291 292
	* minibuf.c (Finternal_complete_buffer): Only show internal buffers if
	they've been requested explicitly (bug#9591).

Chong Yidong's avatar
Chong Yidong committed
293
2011-10-01  Andreas Schwab  <schwab@linux-m68k.org>
294 295 296 297 298

	* keymap.c (Fsingle_key_description): Use make_specified_string
	instead of build_string to build string from push_key_description.
	(Bug#5193)

299 300
2011-09-30  Paul Eggert  <eggert@cs.ucla.edu>

301 302 303 304 305 306 307 308
	* buffer.h (struct buffer): Use time_t, not int, for a time stamp.
	This fixes a Y2038 bug on 64-bit hosts.
	* buffer.c (reset_buffer):
	* fileio.c (Fdo_auto_save, Fset_buffer_auto_saved)
	(Fclear_buffer_auto_save_failure):
	Use 0, not -1, to represent an unset failure time, since time_t
	might not be signed.

309 310 311 312 313 314 315 316 317 318 319 320 321 322
	Remove dependency on glibc malloc internals.
	* alloc.c (XMALLOC_OVERRUN_CHECK_OVERHEAD, XMALLOC_OVERRUN_CHECK_SIZE):
	Move back here from lisp.h, but with their new implementations.
	(XMALLOC_BASE_ALIGNMENT, COMMON_MULTIPLE, XMALLOC_HEADER_ALIGNMENT)
	(XMALLOC_OVERRUN_SIZE_SIZE): Move these new lisp.h macros here.
	* charset.c (charset_table_init): New static var.
	(syms_of_charset): Use it instead of xmalloc.  This removes a
	dependency on glibc malloc internals.  See Eli Zaretskii's comment in
	<http://lists.gnu.org/archive/html/emacs-devel/2011-09/msg00815.html>.
	* lisp.h (XMALLOC_OVERRUN_CHECK_OVERHEAD, XMALLOC_OVERRUN_CHECK_SIZE):
	Move back to alloc.c.
	(XMALLOC_BASE_ALIGNMENT, COMMON_MULTIPLE, XMALLOC_HEADER_ALIGNMENT)
	(XMALLOC_OVERRUN_SIZE_SIZE): Move to alloc.c.

323 324 325 326 327 328
2011-09-30  Jan Djärv  <jan.h.d@swipnet.se>

	* nsterm.m (windowDidResize): Call x_set_window_size only when
	ns_in_resize is true.  Otherwise set pixelwidth/height and
	call change_frame_size (Bug#9628).

329 330
2011-09-30  Paul Eggert  <eggert@cs.ucla.edu>

331 332 333 334
	Port --enable-checking=all to Fedora 14 x86-64.
	* charset.c (syms_of_charset): Also account for glibc malloc's
	internal overhead when calculating the initial malloc maximum.

335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
	Port --enable-checking=all to Fedora 14 x86.
	* alloc.c (XMALLOC_OVERRUN_CHECK_OVERHEAD, XMALLOC_OVERRUN_CHECK_SIZE):
	Move to lisp.h.
	(xmalloc_put_size, xmalloc_get_size, overrun_check_malloc)
	(overrun_check_realloc, overrun_check_free):
	Use XMALLOC_OVERRUN_SIZE_SIZE, not sizeof (size_t).
	That way, xmalloc returns a properly-aligned pointer even if
	XMALLOC_OVERRUN_CHECK is defined.  The old debugging code happened
	to align OK on typical 64-bit hosts, but not on Fedora 14 x86.
	* charset.c (syms_of_charset): Take XMALLOC_OVERRUN_CHECK_OVERHEAD
	into account when calculating the initial malloc maximum.
	* lisp.h (XMALLOC_OVERRUN_CHECK_OVERHEAD, XMALLOC_OVERRUN_CHECK_SIZE):
	Move here from alloc.c, so that charset.c can use it too.
	Properly align; the old code wasn't right for common 32-bit hosts
	when configured with --enable-checking=all.
	(XMALLOC_BASE_ALIGNMENT, COMMON_MULTIPLE, XMALLOC_HEADER_ALIGNMENT)
	(XMALLOC_OVERRUN_SIZE_SIZE): New macros.

Eli Zaretskii's avatar
Eli Zaretskii committed
353 354
2011-09-29  Eli Zaretskii  <eliz@gnu.org>

355
	* sysdep.c (snprintf) [!EOVERFLOW]: If EOVERFLOW is not defined,
Eli Zaretskii's avatar
Eli Zaretskii committed
356 357
	use EDOM.

358 359 360 361 362 363 364 365 366
2011-09-28  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (compute_display_string_end): If there's no display
	string at CHARPOS, return -1.

	* bidi.c (bidi_fetch_char): When compute_display_string_end
	returns a negative value, treat the character as a normal
	character not covered by a display string.  (Bug#9624)

Juanma Barranquero's avatar
Juanma Barranquero committed
367 368 369 370
2011-09-28  Juanma Barranquero  <lekktu@gmail.com>

	* lread.c (Fread_from_string): Fix typo in docstring.

371 372 373 374 375
2011-09-27  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (handle_invisible_prop): If invisible text ends on a
	newline, reseat the iterator instead of bidi-iterating there one
	character at a time.  (Bug#9610)
376 377
	(BUFFER_POS_REACHED_P, move_it_in_display_line_to): Bail when past
	TO_CHARPOS if the bidi iterator is at base embedding level.
378

379 380 381 382 383
2011-09-27  Andreas Schwab  <schwab@linux-m68k.org>

	* lread.c (readevalloop): Use correct code for NBSP.
	(read1): Likewise.  (Bug#9608)

384 385 386 387 388
2011-09-25  Michael Albinus  <michael.albinus@gmx.de>

	* dbusbind.c (Fdbus_register_signal): When service is not
	registered, use nil in Vdbus_registered_objects_table.  (Bug#9581)

389 390 391 392
2011-09-25  Glenn Morris  <rgm@gnu.org>

	* buffer.c (truncate-lines): Doc fix.

393 394 395 396 397
2011-09-24  Chong Yidong  <cyd@stupidchicken.com>

	* window.c (Fwindow_prev_buffers, Fset_window_prev_buffers)
	(Fset_window_next_buffers): Doc fix.

398 399 400 401
2011-09-24  Glenn Morris  <rgm@gnu.org>

	* minibuf.c (read_minibuf): Disable line truncation.  (Bug#5715)

402 403
2011-09-24  Paul Eggert  <eggert@cs.ucla.edu>

404 405
	Fix minor problems found by static checking.
	* xdisp.c (string_from_display_spec): Don't assume vecsize fits in int.
406 407
	* indent.c (Fvertical_motion): Fix == vs = typo.

408 409
2011-09-24  Eli Zaretskii  <eliz@gnu.org>

410 411
	* dispnew.c (syms_of_display) <redisplay-dont-pause>:
	Default value is now t.  Doc fix.
412

413
	* indent.c (Fvertical_motion): Compute and apply the overshoot
414
	logic when moving up, not only when moving down.  Fix the
415
	confusing name and values of the it_overshoot_expected variable;
416
	logic changes accordingly.  (Bug#9254) (Bug#9549)
417 418 419 420 421 422

	* xdisp.c (pos_visible_p): Produce correct pixel coordinates when
	CHARPOS is covered by a display string which includes newlines.
	(move_it_vertically_backward): Avoid inflooping when START_CHARPOS
	is covered by a display string with embedded newlines.

423 424 425 426
2011-09-24  Michael Albinus  <michael.albinus@gmx.de>

	* dbusbind.c (Fdbus_register_signal): Add match rule to
	Vdbus_registered_objects_table.  (Bug#9581)
427 428
	(Fdbus_register_method, Vdbus_registered_objects_table):
	Fix docstring.
429

430 431
2011-09-24  Jim Meyering  <meyering@redhat.com>

432
	do not ignore write error for any output size
433 434 435 436 437 438 439 440 441 442 443 444
	The previous change was incomplete.
	While it makes emacs --batch detect the vast majority of stdout
	write failures, errors were still ignored whenever the output size is
	k * (BUFSIZ+1) - 4.  E.g., on a system with BUFSIZ of 4096,
	  $ emacs --batch --eval '(print (format "%4093d" 0))' > /dev/full \
	      && echo FAIL: ignored write error
	  FAIL: ignored write error
	  $ emacs --batch --eval '(print (format "%20481d" 0))' > /dev/full \
	      && echo FAIL: ignored write error
	  FAIL: ignored write error
	* emacs.c (Fkill_emacs): Also test ferror.  (Bug#9574)

445 446 447 448 449
2011-09-23  Andreas Schwab  <schwab@linux-m68k.org>

	* emacs.c (Fkill_emacs): In noninteractive mode exit
	non-successfully if a write error occurred on stdout.  (Bug#9574)

450 451 452 453 454 455 456 457
2011-09-21  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (pop_it): Allow it->object that is a cons cell to pass
	the xassert test.

	* dispextern.h (struct it): Update the comment documenting what
	can it->OBJECT be.

458 459 460 461 462 463 464
2011-09-20  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (set_cursor_from_row): If the row ends in a newline from
	a display string, extend search for cursor position to end of row.
	(find_row_edges): If the row ends in a newline from a display
	string, increment its MATRIX_ROW_END_CHARPOS by one.  (Bug#9549)
	Handle the case of a display string with multiple newlines.
465 466 467
	(Fcurrent_bidi_paragraph_direction): Fix search for previous
	non-empty line.  Fixes confusing cursor motion with arrow keys at
	the beginning of a line that starts with whitespace.
468

469 470 471 472 473
2011-09-19  Lars Magne Ingebrigtsen  <larsi@gnus.org>

	* lread.c (Fread_from_string): Document what FINAL-STRING-INDEX is
	(bug#9493).

474 475 476 477 478
2011-09-18  Chong Yidong  <cyd@stupidchicken.com>

	* xfns.c (Fx_create_frame): Handle the bitmapIcon resource as
	boolean (Bug#9154).

Eli Zaretskii's avatar
Eli Zaretskii committed
479 480 481 482 483 484 485
2011-09-18  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (display_line): Record maximum and minimum buffer
	positions even if no glyphs were produced (e.g., by a zero-width
	stretch).  Fixes bug#9530 on a TTY.  Under word-wrap, don't record
	buffer positions that will be removed from the glyph row because
	they don't fit.
486 487 488
	(produce_stretch_glyph): Fix a bug in :align-to on a TTY when the
	column is beyond frame width: don't subtract 1 "pixel" when
	computing width of the stretch.
489 490 491
	(reseat_at_next_visible_line_start): Undo the change made on
	2011-09-17 that saved paragraph information and restored it after
	the call to `reseat'.  (Bug#9545)
Eli Zaretskii's avatar
Eli Zaretskii committed
492

493
2011-09-18  YAMAMOTO Mitsuharu  <mituharu@math.s.chiba-u.ac.jp>
494 495 496 497

	* xdisp.c (expose_window): Save original value of phys_cursor_on_p
	and turn window cursor on if cleared (Bug#9415).

498
2011-09-18  Andreas Schwab  <schwab@linux-m68k.org>
499 500 501 502

	* search.c (boyer_moore): Take unibyte characters from pattern
	literally.  (Bug#9458)

503 504 505 506
2011-09-18  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (reseat_at_next_visible_line_start): Fix last change.

507 508
2011-09-18  Paul Eggert  <eggert@cs.ucla.edu>

509 510 511 512
	Fix minor problem found by static checking.
	* xdisp.c (reseat_at_next_visible_line_start): Mark locals as
	initialized, to pacify gcc -Wuninitialized.

513 514 515 516 517
	* fileio.c: Report proper errno when syscall falls.
	(Finsert_file_contents): Save and restore errno,
	so that report_file_error outputs the correct diagnostic.
	(Fwrite_region) [CLASH_DETECTION]: Likewise.

518 519 520 521
2011-09-18  Eli Zaretskii  <eliz@gnu.org>

	* .gdbinit (pgx): Fix references to fields of `struct glyph'.

522 523 524 525 526
2011-09-17  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (produce_stretch_glyph): Another fix for changes made on
	2011-08-30T17:32:44Z!eliz@gnu.org.  (Bug#9530)

527 528
2011-09-17  Eli Zaretskii  <eliz@gnu.org>

529
	* xdisp.c (reseat_at_next_visible_line_start): Keep information
530
	about the current paragraph and restore it after the call to reseat.
531 532 533 534 535 536 537 538

	* bidi.c (MAX_PARAGRAPH_SEARCH): New macro.
	(bidi_find_paragraph_start): Search back for paragraph beginning
	at most MAX_PARAGRAPH_SEARCH lines; if not found, return BEGV_BYTE.
	(bidi_move_to_visually_next): Only trigger paragraph-related
	computations when the last character is a newline or at EOB, not
	just any NEUTRAL_B.  (Bug#9470)

539 540 541
	* xdisp.c (set_cursor_from_row): Don't invoke special treatment of
	truncated lines if point is covered by a display string.  (Bug#9524)

542 543 544 545 546 547 548 549
2011-09-16  Paul Eggert  <eggert@cs.ucla.edu>

	* xselect.c: Relax test for outgoing X longs (Bug#9498).
	(cons_to_x_long): New function.
	(lisp_data_to_selection_data): Use it.  Correct the test for
	short-versus-long data; it was negated.  Break out of vector
	loop, for efficiency, when a long datum is discovered.

550 551 552 553
2011-09-16  Stefan Monnier  <monnier@iro.umontreal.ca>

	* eval.c (Fquote): Document its non-consing behavior (bug#9482).

554 555 556 557 558
2011-09-16  Eli Zaretskii  <eliz@gnu.org>

	* image.c (tiff_handler): Work around a bug in MinGW GCC 3.x (see
	GCC PR/17406) by declaring this function with external scope.

559 560 561 562 563
2011-09-15  Paul Eggert  <eggert@cs.ucla.edu>

	* editfns.c (Fformat): Fix bug in text-property fix (Bug#9514).
	Don't mishandle (length (format "%%")) and (format "%4000s%%" "").

564 565 566 567
2011-09-15  Andreas Schwab  <schwab@linux-m68k.org>

	* editfns.c (Fformat): Correctly handle text properties on "%%".

568 569 570 571 572 573 574 575 576 577
2011-09-15  Eli Zaretskii  <eliz@gnu.org>

	* xterm.c (x_draw_composite_glyph_string_foreground):
	* w32term.c (x_draw_composite_glyph_string_foreground):
	* term.c (encode_terminal_code):
	* composite.c (composition_update_it, get_composition_id):
	* xdisp.c (get_next_display_element)
	(fill_composite_glyph_string): Add comments about special meaning
	of TAB characters in a composition.

578 579 580
2011-09-15  Paul Eggert  <eggert@cs.ucla.edu>

	* editfns.c (Fformat): Fix off-by-1 bug for "%%b" (Bug#9514).
Paul Eggert's avatar
Paul Eggert committed
581 582
	This occurs when processing a multibyte format.
	Problem reported by Wolfgang Jenker.
583

584 585 586 587 588
2011-09-15  Johan Bockgård  <bojohan@gnu.org>

	* xdisp.c (try_cursor_movement): Only check for exact match if
	cursor hpos found by set_cursor_from_row is valid.  (Bug#9495)

Paul Eggert's avatar
Paul Eggert committed
589 590 591 592 593 594 595 596 597 598
2011-09-14  Paul Eggert  <eggert@cs.ucla.edu>

	Remove unused external symbols.
	* dispextern.h (calc_pixel_width_or_height): Remove decl.
	* xdisp.c (calc_pixel_width_or_height): Now static.
	* doprnt.c (exprintf) [! (HAVE_X_WINDOWS && USE_X_TOOLKIT)]: Remove.
	* indent.c (check_display_width):
	* w32term.c: Fix comment to match code.
	* xterm.c, xterm.h (x_catching_errors): Remove.

599 600 601 602 603 604 605 606 607 608
2011-09-14  Paul Eggert  <eggert@cs.ucla.edu>

	* xselect.c: Use signed conversions more consistently (Bug#9498).
	(selection_data_to_lisp_data): Assume incoming selection data are
	signed integers, not unsigned.  This is to be consistent with
	outgoing selection data, which was modified to use signed integers
	in as part of the fix to Bug#9196 in response to Jan D.'s comment
	in <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9196#32> that X11
	expects long, not unsigned long.

609 610 611 612 613 614
2011-09-14  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (try_window_reusing_current_matrix): Fix incorrect
	computation of loop end.  Reported by Johan Bockgård
	<bojohan@gnu.org>.

615 616 617 618
2011-09-13  Chong Yidong  <cyd@stupidchicken.com>

	* frame.c (Fother_visible_frames_p): Function deleted.

619 620 621 622 623
2011-09-12  Eli Zaretskii  <eliz@gnu.org>

	* indent.c (compute_motion): Process display vector front to back
	rather than the other way around.  (Bug#2496)

624 625 626 627
2011-09-12  Stefan Monnier  <monnier@iro.umontreal.ca>

	* fileio.c (Finsert_file_contents): Don't assume beg_offset is 0.

628 629 630 631
2011-09-11  Chong Yidong  <cyd@stupidchicken.com>

	* minibuf.c (Fread_from_minibuffer): Doc fix.

632 633 634 635 636
2011-09-11  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (produce_stretch_glyph): Fix a typo made in changes from
	2011-08-30T17:32:44Z!eliz@gnu.org.  (Bug#9475)

637 638 639 640 641
2011-09-11  Lars Magne Ingebrigtsen  <larsi@gnus.org>

	* fileio.c (Fvisited_file_modtime): Document `(-1 65535)' as a
	value for non-existent files.

642 643 644 645 646 647 648 649
2011-09-11  Eli Zaretskii  <eliz@gnu.org>

	* fileio.c (Finsert_file_contents): If the file cannot be opened,
	set its "size" to -1.  This will set the modtime_size field of
	the corresponding buffer to -1, which is what
	verify-visited-file-modtime expects for files that do not exist.
	(Bug#9139)

650 651 652 653 654 655 656 657 658
2011-09-11  Paul Eggert  <eggert@cs.ucla.edu>

	* keymap.h (KEY_DESCRIPTION_SIZE, push_key_description): Move decls
	here ...
	* lisp.h: ... from here.  push_key_description is no longer
	defined in keyboard.c, so its declaration should not be in
	lisp.h's "Defined in keyboard.c" section, and KEY_DESCRIPTION_SIZE
	logically belongs with push_key_description.

659 660 661 662 663 664
2011-09-10  Paul Eggert  <eggert@cs.ucla.edu>

	* buffer.h: Include <sys/types.h> instead of <time.h>.
	Otherwise, off_t wasn't defined on FreeBSD 9.0-BETA2 i386.
	Problem reported by Herbert J. Skuhra.

665 666 667 668 669
2011-09-10  Lars Magne Ingebrigtsen  <larsi@gnus.org>

	* xml.c (parse_region): Make the parsing work for
	non-comment-starting XML files again (bug#9144).

670 671 672 673 674
2011-09-10  Andreas Schwab  <schwab@linux-m68k.org>

	* image.c (gif_load): Fix calculation of bottom and right corner.
	(Bug#9468)

675 676 677 678 679
2011-09-10  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (MAX_DISP_SCAN): Decrease to 250.  Prevents sluggish
	redisplay in small windows.

680 681 682 683
2011-09-09  Eli Zaretskii  <eliz@gnu.org>

	* frame.c (x_report_frame_params): Cast to avoid compiler warnings.

684 685 686 687 688
2011-09-08  Martin Rudalics  <rudalics@gmx.at>

	* window.c (Fset_window_prev_buffers, Fset_window_next_buffers):
	Operate on live windows only.

689 690 691 692
2011-09-08  Juanma Barranquero  <lekktu@gmail.com>

	* emacs.c (my_heap_start): #ifdef to avoid warnings when unused.

693 694 695 696 697
2011-09-07  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (move_it_in_display_line_to): Call RESTORE_IT on ppos_it
	only under bidi iteration.

698 699 700 701
2011-09-07  Jan Djärv  <jan.h.d@swipnet.se>

	* gtkutil.c (xg_make_tool_item): Insert comment about eventbox.

702 703 704 705 706 707 708 709 710 711 712 713 714 715
2011-09-06  Paul Eggert  <eggert@cs.ucla.edu>

	isnan: Fix porting problem to Solaris 10 with bundled gcc.
	Without this fix, the command to link temacs failed due to an
	undefined symbol __builtin_isnan.  This is because
	/usr/include/iso/math_c99.h #defines isnan(x) to
	__builtin_isnan(x), but the bundled gcc, which identifies itself
	as gcc 3.4.3 (csl-sol210-3_4-branch+sol_rpath), does not have
	a __builtin_isnan.
	* floatfns.c (isnan): #undef, and then #define to a clone of
	what's in data.c.
	(Fisnan): Always define, since it's always available now.
	(syms_of_floatfns): Always define isnan at the Lisp level.

Paul Eggert's avatar
Paul Eggert committed
716
2011-09-06  Paul Eggert  <eggert@cs.ucla.edu>
Paul Eggert's avatar
Paul Eggert committed
717 718 719

	* Makefile.in (gl-stamp): move-if-change now in build-aux (Bug#9169).

Paul Eggert's avatar
Paul Eggert committed
720
2011-09-06  Paul Eggert  <eggert@cs.ucla.edu>
721

Paul Eggert's avatar
Paul Eggert committed
722
	* fileio.c: Fix bugs with large file offsets (Bug#9428).
723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738
	The previous code assumed that file offsets (off_t values) fit in
	EMACS_INT variables, which is not true on typical 32-bit hosts.
	The code messed up by falsely reporting buffer overflow in cases
	such as (insert-file-contents "big" nil 1 2) into an empty buffer
	when "big" contains more than 2**29 bytes, even though this
	inserts just one byte and does not overflow the buffer.
	(Finsert_file_contents): Store file offsets as off_t
	values, not as EMACS_INT values.  Check for overflow when
	converting between EMACS_INT and off_t.  When checking for
	buffer overflow or for overlap, take the offsets into account.
	Don't use EMACS_INT for small values where int suffices.
	When checking for overlap, fix a typo: ZV was used where
	ZV_BYTE was intended.
	(Fwrite_region): Don't assume off_t fits into 'long'.
	* buffer.h (struct buffer.modtime_size): Now off_t, not EMACS_INT.

739 740 741 742
2011-09-05  Michael Albinus  <michael.albinus@gmx.de>

	* dbusbind.c (xd_signature_cat): Rename from signature_cat.

Paul Eggert's avatar
Paul Eggert committed
743
2011-09-04  Paul Eggert  <eggert@cs.ucla.edu>
Paul Eggert's avatar
Paul Eggert committed
744

Paul Eggert's avatar
Paul Eggert committed
745
	sprintf-related integer and memory overflow issues (Bug#9412).
746 747

	* doprnt.c (doprnt): Support printing ptrdiff_t and intmax_t values.
748
	(esprintf, exprintf, evxprintf): New functions.
749
	* keyboard.c (command_loop_level): Now EMACS_INT, not int.
750
	(cmd_error): Kbd macro iterations count is now EMACS_INT, not int.
751 752 753 754 755 756 757 758 759 760 761
	(modify_event_symbol): Do not assume that the length of
	name_alist_or_stem is safe to alloca and fits in int.
	(Fexecute_extended_command): Likewise for function name and binding.
	(Frecursion_depth): Wrap around reliably on integer overflow.
	* keymap.c (push_key_description): First arg is now EMACS_INT, not int,
	since some callers pass EMACS_INT values.
	(Fsingle_key_description): Don't crash if symbol name contains more
	than MAX_ALLOCA bytes.
	* minibuf.c (minibuf_level): Now EMACS_INT, not int.
	(get_minibuffer): Arg is now EMACS_INT, not int.
	* lisp.h (get_minibuffer, push_key_description): Reflect API changes.
762
	(esprintf, exprintf, evxprintf): New decls.
763 764
	* window.h (command_loop_level, minibuf_level): Reflect API changes.

765 766
	* dbusbind.c (signature_cat): New function.
	(xd_signature, Fdbus_register_signal):
767 768
	Do not overrun buffer; instead, report string overflow.

769 770 771
	* dispnew.c (add_window_display_history): Don't overrun buffer.
	Truncate instead; this is OK since it's just a log.

772 773 774 775
	* editfns.c (Fcurrent_time_zone): Don't overrun buffer
	even if the time zone offset is outlandishly large.
	Don't mishandle offset == INT_MIN.

776 777 778
	* emacs.c (main) [NS_IMPL_COCOA]: Don't overrun buffer
	when creating daemon; the previous buffer-overflow check was incorrect.

779 780 781
	* eval.c (verror): Simplify by rewriting in terms of evxprintf,
	which has the guts of the old verror function.

782 783 784
	* filelock.c (lock_file_1, lock_file): Don't blindly alloca long name;
	use SAFE_ALLOCA instead.  Use esprintf to avoid int-overflow issues.

785 786
	* font.c: Include <float.h>, for DBL_MAX_10_EXP.
	(font_unparse_xlfd): Don't blindly alloca long strings.
787
	Don't assume XINT result fits in int, or that XFLOAT_DATA * 10
788
	fits in int, when using sprintf.  Use single snprintf to count
789 790 791 792 793 794
	length of string rather than counting it via multiple sprintfs;
	that's simpler and more reliable.
	(font_unparse_fcname): Use it to avoid sprintf buffer overrun.
	(generate_otf_features) [0 && HAVE_LIBOTF]: Use esprintf, not
	sprintf, in case result does not fit in int.

795 796 797
	* fontset.c (num_auto_fontsets): Now printmax_t, not int.
	(fontset_from_font): Print it.

798 799 800 801 802 803 804 805 806 807
	* frame.c (tty_frame_count): Now printmax_t, not int.
	(make_terminal_frame, set_term_frame_name): Print it.
	(x_report_frame_params): In X, window IDs are unsigned long,
	not signed long, so print them as unsigned.
	(validate_x_resource_name): Check for implausibly long names,
	and don't assume name length fits in 'int'.
	(x_get_resource_string): Don't blindly alloca invocation name;
	use SAFE_ALLOCA.  Use esprintf, not sprintf, in case result does
	not fit in int.

808 809
	* gtkutil.c: Include <float.h>, for DBL_MAX_10_EXP.
	(xg_check_special_colors, xg_set_geometry):
810 811
	Make sprintf buffers a bit bigger, to avoid potential buffer overrun.

812 813 814
	* lread.c (dir_warning): Don't blindly alloca buffer; use SAFE_ALLOCA.
	Use esprintf, not sprintf, in case result does not fit in int.

815 816 817 818 819 820
	* macros.c (executing_kbd_macro_iterations): Now EMACS_INT, not int.
	(Fend_kbd_macro): Don't mishandle MOST_NEGATIVE_FIXNUM by treating
	it as a large positive number.
	(Fexecute_kbd_macro): Don't assume repeat count fits in int.
	* macros.h (executing_kbd_macro_iterations): Now EMACS_INT, not int.

821 822 823
	* nsterm.m ((NSSize)windowWillResize): Use esprintf, not sprintf,
	in case result does not fit in int.

824 825 826 827 828 829 830 831
	* print.c (float_to_string): Detect width overflow more reliably.
	(print_object): Make sprintf buffer a bit bigger, to avoid potential
	buffer overrun.  Don't assume list length fits in 'int'.  Treat
	print length of 0 as 0, not as infinity; to be consistent with other
	uses of print length in this function.  Don't overflow print length
	index.  Don't assume hash table size fits in 'long', or that
	vectorlike size fits in 'unsigned long'.

832 833 834
	* process.c (make_process): Use printmax_t, not int, to format
	process-name gensyms.

835 836
	* sysdep.c (snprintf) [! HAVE_SNPRINTF]: New function.

837 838 839
	* term.c (produce_glyphless_glyph): Make sprintf buffer a bit bigger
	to avoid potential buffer overrun.

840 841 842
	* xfaces.c (x_update_menu_appearance): Don't overrun buffer
	if X resource line is longer than 512 bytes.

843 844 845
	* xfns.c (x_window): Make sprintf buffer a bit bigger
	to avoid potential buffer overrun.

846 847
	* xterm.c (x_io_error_quitter): Don't overrun sprintf buffer.

848 849
	* xterm.h (x_check_errors): Add ATTRIBUTE_FORMAT_PRINTF.

850 851
2011-09-04  Paul Eggert  <eggert@cs.ucla.edu>

852
	Integer overflow fixes for scrolling, etc.
Paul Eggert's avatar
Paul Eggert committed
853 854
	Without these, Emacs silently mishandles large integers sometimes.
	For example, "C-u 4294967297 M-x recenter" was treated as if
855 856
	it were "C-u 1 M-x recenter" on a typical 64-bit host.

Paul Eggert's avatar
Paul Eggert committed
857 858
	* xdisp.c (try_window_id): Check Emacs fixnum range before
	converting to 'int'.
Paul Eggert's avatar
Paul Eggert committed
859

Paul Eggert's avatar
Paul Eggert committed
860
	* window.c (window_scroll_line_based, Frecenter):
861 862 863 864 865 866 867 868
	Check that an Emacs fixnum is in range before assigning it to 'int'.
	(Frecenter, Fmove_to_window_line): Use EMACS_INT, not int, for
	values converted from Emacs fixnums.
	(Frecenter): Don't wrap around a line count if it is out of 'int'
	range; instead, treat it as an extreme value.
	(Fset_window_configuration, compare_window_configurations):
	Use ptrdiff_t, not int, for index that might exceed 2 GiB.

Paul Eggert's avatar
Paul Eggert committed
869 870 871
	* search.c (Freplace_match): Use ptrdiff_t, not int, for indexes
	that can exceed INT_MAX.  Check that EMACS_INT value is in range
	before assigning it to the (possibly-narrower) index.
872 873
	(match_limit): Don't assume that a fixnum can fit in 'int'.

Paul Eggert's avatar
Paul Eggert committed
874
	* print.c (print_object): Use ptrdiff_t, not int, for index that can
Paul Eggert's avatar
Paul Eggert committed
875 876
	exceed INT_MAX.

Paul Eggert's avatar
Paul Eggert committed
877
	* indent.c (position_indentation): Now takes ptrdiff_t, not int.
878 879 880 881
	(Fvertical_motion): Don't wrap around LINES values that don't fit
	in 'int'.  Instead, treat them as extreme values.  This is good
	enough for windows, which can't have more than INT_MAX lines anyway.

882 883
2011-09-03  Lars Magne Ingebrigtsen  <larsi@gnus.org>

884 885
	* Require libxml/parser.h to avoid compilation warning.

886 887 888 889 890
	* emacs.c (shut_down_emacs): Call xmlCleanupParser on shutdown.

	* xml.c (parse_region): Don't call xmlCleanupParser after parsing,
	since this reportedly can destroy thread storage.

891 892 893 894 895
2011-08-30  Chong Yidong  <cyd@stupidchicken.com>

	* syntax.c (find_defun_start): Update all cache variables if
	exiting early (Bug#9401).

896 897
2011-08-30  Eli Zaretskii  <eliz@gnu.org>

898 899
	* image.c (x_bitmap_pixmap): Cast to int to avoid compiler warnings.

900 901 902 903 904 905 906
	* xdisp.c (produce_stretch_glyph): No longer static, compiled also
	when HAVE_WINDOW_SYSTEM is not defined.  Support both GUI and TTY
	frames.  Call tty_append_glyph in the TTY case.  (Bug#9402)

	* term.c (tty_append_glyph): New function.
	(produce_stretch_glyph): Static function and its prototype deleted.

907 908
	* dispextern.h (produce_stretch_glyph, tty_append_glyph):
	Add prototypes.
909

910 911 912 913 914
2011-08-29  Paul Eggert  <eggert@cs.ucla.edu>

	* image.c (parse_image_spec): Check for nonnegative, not for positive,
	when checking :margin (Bug#9390).
	(IMAGE_NON_NEGATIVE_INTEGER_VALUE_OR_PAIR):
915
	Rename from IMAGE_POSITIVE_INTEGER_VALUE_OR_PAIR,
916 917
	so that the name doesn't mislead.  All uses changed.

918 919 920 921 922
2011-08-28  Johan Bockgård  <bojohan@gnu.org>

	* term.c (init_tty) [HAVE_GPM]: Move mouse settings after
	set_tty_hooks.

923 924 925 926 927 928 929
2011-08-27  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (move_it_to): Don't bail out early when reaching
	position beyond to_charpos, if we are scanning backwards.
	(move_it_vertically_backward): When DY == 0, make sure we get to
	the first character in the line after the newline.

930 931 932 933 934 935 936 937 938 939
2011-08-27  Paul Eggert  <eggert@cs.ucla.edu>

	* ccl.c: Improve and simplify overflow checking (Bug#9196).
	(ccl_driver): Do not generate an out-of-range pointer.
	(Fccl_execute_on_string): Remove unnecessary check for
	integer overflow, noted by Stefan Monnier in
	<http://lists.gnu.org/archive/html/emacs-devel/2011-08/msg00979.html>.
	Remove a FIXME that didn't need fixing.
	Simplify the newly-introduced buffer reallocation code.

940 941 942 943
2011-08-27  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in ($(BLD)/alloc.$(O)): Depend on lib/verify.h.

Paul Eggert's avatar
Paul Eggert committed
944
2011-08-26  Paul Eggert  <eggert@cs.ucla.edu>
945

Paul Eggert's avatar
Paul Eggert committed
946
	Integer and memory overflow issues (Bug#9196).
947

948 949 950 951 952
	* doc.c (get_doc_string): Rework so that
	get_doc_string_buffer_size is the actual buffer size, rather than
	being 1 less than the actual buffer size; this makes xpalloc more
	convenient.

Paul Eggert's avatar
Paul Eggert committed
953 954 955 956
	* image.c (x_allocate_bitmap_record, cache_image):
	* xselect.c (Fx_register_dnd_atom):
	Simplify previous changes by using xpalloc.

957 958 959
	* buffer.c (overlay_str_len): Now ptrdiff_t, not EMACS_INT,
	since either will do and ptrdiff_t is convenient with xpalloc.

960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
	* charset.c (charset_table_size)
	(struct charset_sort_data.priority): Now ptrdiff_t.
	(charset_compare): Don't overflow if priorities differ greatly.
	(Fsort_charsets): Don't assume list length fits in int.
	Check for size-calculation overflow when allocating sort data.
	(syms_of_charset): Allocate an initial charset table that is
	just under 64 KiB, to avoid problems with glibc malloc and mmap.

	* cmds.c (internal_self_insert): Check for size-calculation overflow.

	* composite.h (struct composition.glyph_len): Now int, not unsigned.
	The actual value is always <= INT_MAX, and leaving it unsigned made
	overflow checking harder.

	* dispextern.h (struct glyph_matrix.rows_allocated)
	(struct face_cache.size): Now ptrdiff_t, for convenience in use
	with xpalloc.  The values are still always <= INT_MAX.

	* indent.c (compute_motion): Adjust to region_cache_forward sig change.

	* lisp.h (xnmalloc, xnrealloc, xpalloc): New decls.
	(SAFE_NALLOCA): New macro.

	* region-cache.c (struct boundary.pos, find_cache_boundary)
	(move_cache_gap, insert_cache_boundary, delete_cache_boundaries)
	(set_cache_region, invalidate_region_cache)
	(revalidate_region_cache, know_region_cache, region_cache_forward)
	(region_cache_backward, pp_cache):
	Use ptrdiff_t, not EMACS_INT, since either will do.  This is needed
	so that ptrdiff_t * can be passed to xpalloc.
	(struct region_cache): Similarly, for gap_start, gap_len, cache_len,
	beg_unchanged, end_unchanged, buffer_beg, buffer_end members.
	(pp_cache): Don't assume cache_len fits in int.
	* region-cache.h: Adjust extern decls to match.

	* search.c (scan_buffer, Freplace_match): Use ptrdiff_t, not
	EMACS_INT, since either will do, for xpalloc.

	* alloc.c: Include verify.h, and check that int fits in ptrdiff_t.
	(xnmalloc, xnrealloc, xpalloc): New functions.

1001 1002 1003 1004
	* bidi.c (bidi_shelve_header_size): New constant.
	(bidi_cache_ensure_space, bidi_shelve_cache): Use it.
	(bidi_cache_ensure_space): Avoid integer overflow when allocating.

1005
	* bidi.c (bidi_cache_shrink):
1006 1007 1008 1009
	* buffer.c (overlays_at, overlays_in, record_overlay_string)
	(overlay_strings):
	Don't update size of array until after memory allocation succeeds,
	because xmalloc/xrealloc may not return.
1010 1011 1012 1013
	(struct sortstrlist.bytes): Now ptrdiff_t, as EMACS_INT doesn't help
	now that we have proper integer overflow checking.
	(record_overlay_string, overlay_strings): Catch overflows when
	calculating size of overlay_str_buf.
1014

1015 1016 1017 1018 1019
	* callproc.c (Fcall_process): Check for size overflow when
	calculating size of args2.
	(child_setup): Avoid overflow by using size_t rather than ptrdiff_t.
	Normally we prefer signed values, but sticking with ptrdiff_t would
	require adding more-complicated checks.
1020 1021 1022 1023

	* ccl.c (Fccl_execute_on_string): Check for memory overflow.
	Use ptrdiff_t rather than EMACS_INT where ptrdiff_t will do.
	Redo buffer-overflow calculations to avoid integer overflow.
1024
	Add a FIXME comment where memory seems to be over-allocated.
1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063

	* character.c (Fstring): Check for size-calculation overflow.

	* coding.c (produce_chars): Redo buffer-overflow calculations to avoid
	unnecessary integer overflow.  Check for size overflow.
	(encode_coding_object): Don't update size until xmalloc succeeds.

	* composite.c (get_composition_id): Check for overflow in glyph
	length calculations.

	Integer and memory overflow fixes for display code.
	* dispextern.h (struct glyph_pool.nglyphs): Now ptrdiff_t, not int.
	* dispnew.c (adjust_glyph_matrix, realloc_glyph_pool)
	(scrolling_window): Check for overflow in size calculations.
	(line_draw_cost, realloc_glyph_pool, add_row_entry):
	Don't assume glyph table len fits in int.
	(struct row_entry.bucket, row_entry_pool_size, row_entry_idx)
	(row_table_size): Now ptrdiff_t, not int.
	(scrolling_window): Avoid overflow in size calculations.
	Don't update size until allocation succeeds.
	* fns.c (concat): Check for overflow in size calculations.
	(next_almost_prime): Verify NEXT_ALMOST_PRIME_LIMIT.
	* lisp.h (RANGED_INTEGERP, TYPE_RANGED_INTEGERP): New macros.
	(NEXT_ALMOST_PRIME_LIMIT): New constant.

	* doc.c (get_doc_string_buffer_size): Now ptrdiff_t, not int.
	(get_doc_string): Check for size calculation overflow.
	Don't update size until allocation succeeds.
	(get_doc_string, Fsubstitute_command_keys): Use ptrdiff_t, not
	EMACS_INT, where ptrdiff_t will do.
	(Fsubstitute_command_keys): Check for string overflow.

	* editfns.c (set_time_zone_rule): Don't assume environment length
	fits in int.
	(message_length): Now ptrdiff_t, not int.
	(Fmessage_box): Don't update size until allocation succeeds.
	Don't assume message length fits in int.
	(Fformat): Use ptrdiff_t, not EMACS_INT, where ptrdiff_t will do.

1064 1065 1066 1067
	* emacs.c (main): Do not reallocate argv, since there is a null at
	the end that can be overwritten, and this way there's no need to
	worry about size-calculation overflow.
	(sort_args): Check for size-calculation overflow.
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106

	* eval.c (init_eval_once, grow_specpdl): Don't update size until
	alloc succeeds.
	(call_debugger, grow_specpdl): Redo calculations to avoid overflow.

	* frame.c (set_menu_bar_lines, x_set_frame_parameters)
	(x_set_scroll_bar_width, x_figure_window_size):
	Check for integer overflow.
	(x_set_alpha): Do not assume XINT fits in int.

	* frame.h (struct frame): Use int, not EMACS_INT, where int works.
	This is for the members text_lines, text_cols, total_lines, total_cols,
	where the system imposes an 'int' limit.

	* fringe.c (Fdefine_fringe_bitmap):
	Don't update size until alloc works.

	* ftfont.c (ftfont_get_open_type_spec, setup_otf_gstring)
	(ftfont_shape_by_flt): Check for integer overflow in size calculations.

	* gtkutil.c (get_utf8_string, xg_store_widget_in_map):
	Check for size-calculation overflow.
	(get_utf8_string): Use ptrdiff_t, not size_t, where either will
	do, as we prefer signed integers.
	(id_to_widget.max_size, id_to_widget.used)
	(xg_store_widget_in_map, xg_remove_widget_from_map)
	(xg_get_widget_from_map, xg_get_scroll_id_for_window)
	(xg_remove_scroll_bar, xg_update_scrollbar_pos):
	Use and return ptrdiff_t, not int.
	(xg_gtk_scroll_destroy): Don't assume ptrdiff_t fits in int.
	* gtkutil.h: Change prototypes to match the above.

	* image.c (RANGED_INTEGERP, TYPE_RANGED_INTEGERP): Remove; these
	are duplicate now that they've been promoted to lisp.h.
	(x_allocate_bitmap_record, x_alloc_image_color)
	(make_image_cache, cache_image, xpm_load):
	Don't update size until alloc is done.
	(xpm_load, lookup_rgb_color, lookup_pixel_color, x_to_xcolors)
	(x_detect_edges):
1107
	Check for size calculation overflow.
1108 1109 1110
	(ct_colors_allocated_max): New constant.
	(x_to_xcolors, x_detect_edges): Reorder multiplicands to avoid
	overflow.
1111

1112 1113 1114 1115 1116
	* keyboard.c (read_char, menu_bar_items, tool_bar_items)
	(read_char_x_menu_prompt, read_char_minibuf_menu_width)
	(read_char_minibuf_menu_prompt, follow_key, read_key_sequence):
	Use ptrdiff_t, not int, to count maps.
	(read_char_minibuf_menu_prompt): Check for overflow in size
1117 1118
	calculations.  Don't update size until allocation succeeds.
	Redo calculations to avoid overflow.
1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
	* keyboard.h: Change prototypes to match the above.

	* keymap.c (cmm_size, current_minor_maps): Use ptrdiff_t, not int,
	to count maps.
	(current_minor_maps): Check for size calculation overflow.
	* keymap.h: Change prototypes to match the above.

	* lread.c (read1, init_obarray): Don't update size until alloc done.

	* macros.c (Fstart_kbd_macro): Don't update size until alloc done.
	(store_kbd_macro_char): Reorder multiplicands to avoid overflow.

	* nsterm.h (struct ns_color_table.size, struct ns_color_table.avail):
	Now ptrdiff_t, not int.
	* nsterm.m (ns_index_color): Use ptrdiff_t, not int, for table indexes.
	(ns_draw_fringe_bitmap): Rewrite to avoid overflow.

	* process.c (Fnetwork_interface_list): Check for overflow
	in size calculation.

	* region-cache.c (move_cache_gap): Check for size calculation overflow.

	* scroll.c (do_line_insertion_deletion_costs): Check for size calc
	overflow.  Don't bother calling xmalloc when xrealloc will do.

	* search.c (Freplace_match): Check for size calculation overflow.
	(Fset_match_data): Don't assume list lengths fit in 'int'.

	* sysdep.c (system_process_attributes): Use ptrdiff_t, not int,
	for command line length.  Do not attempt to address one before the
	beginning of an array, as that's not portable.

	* term.c (max_frame_lines): Remove; unused.
	(encode_terminal_src_size, encode_terminal_dst_size): Now ptrdiff_t,
	not int.
	(encode_terminal_code, calculate_costs): Check for size
	calculation overflow.
	(encode_terminal_code): Use ptrdiff_t, not int, to record glyph
	table lengths and related sizes.  Don't update size until alloc
	done.  Redo calculations to avoid overflow.
	(calculate_costs): Don't bother calling xmalloc when xrealloc will do.

	* termcap.c (tgetent): Use ptrdiff_t, not int, to record results of
	subtracting pointers.
	(gobble_line): Check for overflow more carefully.  Don't update size
	until alloc done.

	* tparam.c (tparam1): Use ptrdiff_t, not int, for sizes.
	Don't update size until alloc done.
	Redo size calculations to avoid overflow.
	Check for size calculation overflow.
1170
	(main) [DEBUG]: Fix typo in invoking tparam1.
1171 1172 1173 1174 1175

	* xdisp.c (store_mode_line_noprop_char, x_consider_frame_title):
	Use ptrdiff_t, not int, for sizes.
	(store_mode_line_noprop_char): Don't update size until alloc done.

1176 1177 1178 1179 1180 1181
	* xfaces.c (lface_id_to_name_size, Finternal_make_lisp_face):
	Use ptrdiff_t, not int, for sizes.
	(Finternal_make_lisp_face, cache_face):
	Check for size calculation overflow.
	(cache_face): Treat size calculation overflows as if they were
	memory exhaustion (the usual treatment), rather than aborting.
1182 1183 1184 1185 1186 1187

	* xfns.c (x_encode_text, x_set_name_internal)
	(Fx_change_window_property): Use ptrdiff_t, not int, to count
	sizes, since they can exceed INT_MAX in size.  Check for size
	calculation overflow.

1188 1189
	* xgselect.c (gfds_size): Now ptrdiff_t, for convenience with xpalloc.
	(xg_select): Check for size calculation overflow.
1190 1191
	Don't update size until alloc done.

1192
	* xrdb.c (get_environ_db): Don't assume path length fits in int,
1193
	as sprintf is limited to int lengths.
1194

1195 1196
	* xselect.c (X_LONG_SIZE, X_SHRT_MAX, X_SHRT_MIN, X_LONG_MAX)
	(X_LONG_MIN): New macros.
1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
	Use them to make the following changes clearer.
	(MAX_SELECTION_QUANTUM): Make the other bounds on this value clearer.
	This change doesn't affect the value now, but it may help remind
	future maintainers not to raise the value too much later.
	(SELECTION_QUANTUM): Remove, replacing with ...
	(selection_quantum): ... new function, which avoids overflow.
	All uses changed.
	(struct selection_data.size): Now ptrdiff_t, not int, to avoid
	assumption that selection length fits in 'int'.
	(x_reply_selection_request, x_handle_selection_request)
	(x_get_window_property, receive_incremental_selection)
	(x_get_window_property_as_lisp_data, selection_data_to_lisp_data)
	(lisp_data_to_selection_data, clean_local_selection_data):
	Use ptrdiff_t, not int, to record length of selection.
	(x_reply_selection_request, x_get_window_property)
	(receive_incremental_selection, x_property_data_to_lisp):
	Redo calculations to avoid overflow.
	(x_reply_selection_request): When sending hint, ceiling it at
1215
	X_LONG_MAX rather than relying on wraparound overflow to send
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
	something.
	(x_get_window_property, receive_incremental_selection)
	(lisp_data_to_selection_data, x_property_data_to_lisp):
	Check for size-calculation overflow.
	(x_get_window_property, receive_incremental_selection)
	(lisp_data_to_selection_data, Fx_register_dnd_atom):
	Don't store size until memory allocation succeeds.
	(x_get_window_property): Plug memory leak on memory exhaustion.
	Don't double-block input; malloc is safe here.  Don't assume 2**34
	- 4 fits in unsigned long.  Add an xassert to check
	XGetWindowProperty overflow.  Be more careful about overflow
	calculations, and distinguish size from memory overflow better.
	(receive_incremental_selection): When tracing, don't assume
	unsigned int is less than INT_MAX.
	(x_selection_data_to_lisp_data): Remove unnecessary (and in theory
	harmful) conversions of unsigned short to int.
	(lisp_data_to_selection_data): Don't assume that integers
	in the range -65535 through -1 fit in an X unsigned short.
	Don't assume that ULONG_MAX == X_ULONG_MAX.  Don't store into
	result parameters unless successful.  Rely on cons_to_unsigned
	to report problems with elements; the old code wasn't right anyway.
	(x_check_property_data): Check for int overflow; we cannot use
	a wider type due to X limits.
	(x_handle_dnd_message): Use unsigned int, to avoid int overflow.

1241
	* xsmfns.c (smc_save_yourself_CB): Check for size calc overflow.
1242

1243 1244
	* xterm.c (x_color_cells, x_send_scrollbar_event, handle_one_xevent)
	(x_term_init): Check for size calculation overflow.
1245 1246
	(x_color_cells): Don't store size until memory allocation succeeds.
	(handle_one_xevent): Use ptrdiff_t, not int, for byte counts.
1247
	Don't assume alloca size is less than MAX_ALLOCA.
1248 1249
	(x_term_init): Don't assume length fits in int (sprintf is limited
	to int size).
Paul Eggert's avatar
Paul Eggert committed
1250

1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261
	Use ptrdiff_t for composition IDs.
	* character.c (lisp_string_width):
	* composite.c (composition_table_size, n_compositions)
	(get_composition_id, composition_gstring_from_id):
	* dispextern.h (struct glyph_string.cmp_id, struct composition_it.id):
	* xdisp.c (BUILD_COMPOSITE_GLYPH_STRING):
	* window.c (Frecenter):
	Use ptrdiff_t, not int, for composition IDs.
	* composite.c (get_composition_id): Check for integer overflow.
	* composite.h: Adjust prototypes to match the above changes.

1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276
	Use ptrdiff_t for hash table indexes.
	* category.c (hash_get_category_set):
	* ccl.c (ccl_driver):
	* charset.h (struct charset.hash_index, CHECK_CHARSET_GET_ID):
	* coding.c (coding_system_charset_list, detect_coding_system):
	* coding.h (struct coding_system.id):
	* composite.c (get_composition_id, gstring_lookup_cache):
	* fns.c (hash_lookup, hash_put, Fgethash, Fputhash):
	* image.c (xpm_get_color_table_h):
	* lisp.h (hash_lookup, hash_put):
	* minibuf.c (Ftest_completion):
	Use ptrdiff_t for hash table indexes, not int (which is too
	narrow, on 64-bit hosts) or EMACS_INT (which is too wide, on
	32-bit --with-wide-int hosts).

1277 1278 1279 1280
	* charset.c (Fdefine_charset_internal): Check for integer overflow.
	Add a FIXME comment about memory leaks.
	(syms_of_charset): Don't assume xmalloc returns.

1281 1282 1283 1284 1285 1286
	Don't assume that stated character widths fit in int.
	* character.c (Fchar_width, c_string_width, lisp_string_width):
	* character.h (CHAR_WIDTH):
	* indent.c (MULTIBYTE_BYTES_WIDTH):
	Use sanitize_char_width to avoid undefined and/or bad behavior
	with outlandish widths.
1287
	* character.h (sanitize_tab_width): Rename from sanitize_width,
1288 1289 1290
	now that we have two such functions.  All uses changed.
	(sanitize_char_width): New inline function.

1291 1292 1293 1294 1295 1296 1297 1298
	Don't assume that tab-width fits in int.
	* character.h (sanitize_width): New inline function.
	(SANE_TAB_WIDTH): New macro.
	(ASCII_CHAR_WIDTH): Use it.
	* indent.c (sane_tab_width): Remove.  All uses replaced by
	SANE_TAB_WIDTH (current_buffer).
	* xdisp.c (init_iterator): Use SANE_TAB_WIDTH.

1299 1300 1301
	* fileio.c: Integer overflow issues with file modes.
	(Fset_file_modes, auto_save_1): Don't assume EMACS_INT fits in int.

1302 1303 1304 1305
	* charset.c (read_hex): New arg OVERFLOW.  All uses changed.
	Remove unreachable code.
	(read_hex, load_charset_map_from_file): Check for integer overflow.

1306 1307 1308 1309 1310 1311
	* xterm.c: don't go over XClientMessageEvent limit
	(scroll_bar_windows_size): Now ptrdiff_t, as we prefer signed.
	(x_send_scroll_bar_event): Likewise.  Check that the size does not
	exceed limits imposed by XClientMessageEvent, as well as the usual
	ptrdiff_t and size_t limits.

1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
	* keyboard.c: Overflow, signedness and related fixes.
	(make_lispy_movement): Use same integer type in forward decl
	that is used in the definition.
	(read_key_sequence, keyremap_step):
	Change bufsize argument back to int, undoing my 2011-03-30 change.
	We prefer signed types, and int is wide enough here.
	(parse_tool_bar_item): Don't assume tool_bar_max_label_size is less
	than TYPE_MAXIMUM (EMACS_INT) / 2.  Don't let the label size grow
	larger than STRING_BYTES_BOUND.  Use ptrdiff_t for Emacs string
	length, not size_t.  Use ptrdiff_t for index, not int.
	(keyremap_step, read_key_sequence): Redo bufsize check to avoid
	possibility of integer overflow.

1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352
	Overflow, signedness and related fixes for images.

	* dispextern.h (struct it.stack[0].u.image.image_id)
	(struct_it.image_id, struct image.id, struct image_cache.size)
	(struct image_cache.used, struct image_cache.ref_count):
	* gtkutil.c (update_frame_tool_bar):
	* image.c (x_reference_bitmap, Fimage_size, Fimage_mask_p)
	(Fimage_metadata, free_image_cache, clear_image_cache, lookup_image)
	(cache_image, mark_image_cache, x_kill_gs_process, Flookup_image):
	* nsmenu.m (update_frame_tool_bar):
	* xdisp.c (calc_pixel_width_or_height):
	* xfns.c (image_cache_refcount):
	Image IDs are now ptrdiff_t, not int, to avoid arbitrary limits
	on typical 64-bit hosts.

	* image.c (RANGED_INTEGERP, TYPE_RANGED_INTEGERP): New macros.
	(x_bitmap_pixmap, x_create_x_image_and_pixmap):
	Omit unnecessary casts to int.
	(parse_image_spec): Check that integers fall into 'int' range
	when the callers expect that.
	(image_ascent): Redo ascent calculation to avoid int overflow.
	(clear_image_cache): Avoid overflow when sqrt (INT_MAX) < nimages.
	(lookup_image): Remove unnecessary tests.
	(xbm_image_p): Locals are now of int, not EMACS_INT,
	since parse_image_check makes sure they fit into int.
	(png_load, gif_load, svg_load_image):
	Prefer int to unsigned where either will do.
	(tiff_handler): New function, combining the cores of the
1353 1354
	old tiff_error_handler and tiff_warning_handler.
	This function is rewritten to use vsnprintf and thereby avoid
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
	stack buffer overflows.  It uses only the features of vsnprintf
	that are common to both POSIX and native Microsoft.
	(tiff_error_handler, tiff_warning_handler): Use it.
	(tiff_load, gif_load, imagemagick_load_image):
	Don't assume :index value fits in 'int'.
	(gif_load): Omit unnecessary cast to double, and avoid double-rounding.
	(imagemagick_load_image): Check that crop parameters fit into
	the integer types that MagickCropImage accepts.  Don't assume
	Vimagemagick_render_type has a nonnegative value.  Don't assume
	size_t fits in 'long'.
	(gs_load): Use printmax_t to print the widest integers possible.
	Check for integer overflow when computing image height and width.

1368 1369 1370 1371 1372
2011-08-26  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (redisplay_window): Don't force window start if point
	will be invisible in the resulting window.  (Bug#9324)

1373 1374 1375 1376 1377 1378 1379 1380 1381
2011-08-25  Eli Zaretskii  <eliz@gnu.org>

	* xdisp.c (compute_display_string_pos): Return 2 in DISP_PROP when
	the display spec is of the form `(space ...)'.
	(handle_display_spec): Return the value returned by
	handle_single_display_spec, not just 1 or zero.
	(handle_single_display_spec): If the display spec is of the form
	`(space ...)', and specifies display in the text area, return 2
	rather than 1.
1382
	(try_cursor_movement): Check for the need to scroll more
1383 1384
	accurately, and prefer exact match for point under bidi.
	Don't advance `row' beyond the last row of the window.
1385 1386 1387 1388 1389 1390 1391 1392

	* dispextern.h (struct bidi_it): Rename the disp_prop_p member
	into disp_prop; all users changed.

	* bidi.c (bidi_fetch_char): If compute_display_string_pos returns
	DISP_PROP = 2, substitute the u+2029 PARAGRAPH SEPARATOR character
	for the text covered by the display property.

1393 1394 1395 1396 1397 1398
2011-08-25  Chong Yidong  <cyd@stupidchicken.com>

	* buffer.c (Fbury_buffer_internal): Rename from Funrecord_buffer.
	Change return value to nil.
	(Frecord_buffer): Delete unused function.

Eli Zaretskii's avatar
Eli Zaretskii committed
1399 1400
2011-08-24  Eli Zaretskii  <eliz@gnu.org>

1401 1402
	* xdisp.c (Fcurrent_bidi_paragraph_direction): For unibyte
	buffers, return left-to-right.
1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415
	(set_cursor_from_row): Consider candidate row a win if its glyph
	represents a newline and point is on that newline.  Fixes cursor
	positioning on the newline at EOL of R2L text within L2R
	paragraph, and vice versa.
	(try_cursor_movement): Check continued rows, in addition to
	continuation rows.  Fixes unwarranted scroll when point enters a
	continued line of R2L text within an L2R paragraph, or vice versa.
	(cursor_row_p): Consider the case of point being equal to
	MATRIX_ROW_END_CHARPOS.  Prevents cursor being stuck when moving
	from the end of a short line to the beginning of a continued line
	of R2L text within L2R paragraph.
	(RECORD_MAX_MIN_POS): For max_pos, use IT_CHARPOS even for
	composed characters.
1416