ChangeLog 160 KB
Newer Older
1 2 3 4 5 6 7 8
2011-06-07  Martin Rudalics  <rudalics@gmx.at>

	* window.c (Fwindow_total_size, Fwindow_left_column)
	(Fwindow_top_line, window_body_lines, Fwindow_body_size)
	(Fwindow_list_1): New functions.
	(window_box_text_cols): Replace with window_body_cols.
	(Fwindow_width, Fscroll_left, Fscroll_right): Use
	window_body_cols instead of window_box_text_cols.
9 10 11 12
	(delete_window, Fset_window_configuration): Call
	delete_all_subwindows with window as argument.
	(delete_all_subwindows): Take a window as argument and not a
	structure.  Rewrite.
13 14

	* window.h: Extern window_body_cols instead of
15 16
	window_box_text_cols.  delete_all_subwindows now takes a
	Lisp_Object as argument.
17 18 19 20

	* indent.c (compute_motion, Fcompute_motion): Use
	window_body_cols instead of window_box_text_cols.

21 22 23
	* frame.c (delete_frame): Call delete_all_subwindows with root
	window as argument.

24 25 26 27
2011-06-07  Daniel Colascione  <dan.colascione@gmail.com>

	* fns.c (Fputhash): Document return value.

28 29 30 31
2011-06-06  Chong Yidong  <cyd@stupidchicken.com>

	* image.c (gif_load): Implement gif89a spec "no disposal" method.

32
2011-06-06  Paul Eggert  <eggert@cs.ucla.edu>
33

Paul Eggert's avatar
Paul Eggert committed
34
	Cons<->int and similar integer overflow fixes (Bug#8794).
Paul Eggert's avatar
Paul Eggert committed
35

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
	Check for overflow when converting integer to cons and back.
	* charset.c (Fdefine_charset_internal, Fdecode_char):
	Use cons_to_unsigned to catch overflow.
	(Fencode_char): Use INTEGER_TO_CONS.
	* composite.h (LGLYPH_CODE): Use cons_to_unsigned.
	(LGLYPH_SET_CODE): Use INTEGER_TO_CONS.
	* data.c (long_to_cons, cons_to_long): Remove.
	(cons_to_unsigned, cons_to_signed): New functions.
	These signal an error for invalid or out-of-range values.
	* dired.c (Ffile_attributes): Use INTEGER_TO_CONS.
	* fileio.c (Fset_visited_file_modtime): Use CONS_TO_INTEGER.
	* font.c (Ffont_variation_glyphs):
	* fontset.c (Finternal_char_font): Use INTEGER_TO_CONS.
	* lisp.h: Include <intprops.h>.
	(INTEGER_TO_CONS, CONS_TO_INTEGER): New macros.
	(cons_to_signed, cons_to_unsigned): New decls.
	(long_to_cons, cons_to_long): Remove decls.
	* undo.c (record_first_change): Use INTEGER_TO_CONS.
	(Fprimitive_undo): Use CONS_TO_INTEGER.
	* xfns.c (Fx_window_property): Likewise.
	* xselect.c: Include <limits.h>.
	(x_own_selection, selection_data_to_lisp_data):
	Use INTEGER_TO_CONS.
	(x_handle_selection_request, x_handle_selection_clear)
	(x_get_foreign_selection, Fx_disown_selection_internal)
	(Fx_get_atom_name, x_send_client_event): Use CONS_TO_INTEGER.
	(lisp_data_to_selection_data): Use cons_to_unsigned.
	(x_fill_property_data): Use cons_to_signed.
	Report values out of range.

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
	Check for buffer and string overflow more precisely.
	* buffer.h (BUF_BYTES_MAX): New macro.
	* lisp.h (STRING_BYTES_MAX): New macro.
	* alloc.c (Fmake_string):
	* character.c (string_escape_byte8):
	* coding.c (coding_alloc_by_realloc):
	* doprnt.c (doprnt):
	* editfns.c (Fformat):
	* eval.c (verror):
	Use STRING_BYTES_MAX, not MOST_POSITIVE_FIXNUM,
	since they may not be the same number.
	* editfns.c (Finsert_char):
	* fileio.c (Finsert_file_contents):
	Likewise for BUF_BYTES_MAX.

81 82 83 84 85 86
	* image.c: Use ptrdiff_t, not int, for sizes.
	(slurp_file): Switch from int to ptrdiff_t.
	All uses changed.
	(slurp_file): Check that file size fits in both size_t (for
	malloc) and ptrdiff_t (for sanity and safety).

87 88 89
	* fileio.c (Fverify_visited_file_modtime): Avoid time overflow
	if b->modtime has its maximal value.

90 91
	* dired.c (Ffile_attributes): Don't assume EMACS_INT has >32 bits.

92 93 94 95 96
	Don't assume time_t can fit into int.
	* buffer.h (struct buffer.modtime): Now time_t, not int.
	* fileio.c (Fvisited_file_modtime): No need for time_t cast now.
	* undo.c (Fprimitive_undo): Use time_t, not int, for time_t value.

97 98 99 100 101 102 103 104 105 106 107 108 109 110
	Minor fixes for signed vs unsigned integers.
	* character.h (MAYBE_UNIFY_CHAR):
	* charset.c (maybe_unify_char):
	* keyboard.c (read_char, reorder_modifiers):
	XINT -> XFASTINT, since the integer must be nonnegative.
	* ftfont.c (ftfont_spec_pattern):
	* keymap.c (access_keymap, silly_event_symbol_error):
	XUINT -> XFASTINT, since the integer must be nonnegative.
	(Fsingle_key_description, preferred_sequence_p): XUINT -> XINT,
	since it makes no difference and we prefer signed.
	* keyboard.c (record_char): Use XUINT when all the neighbors do.
	(access_keymap): NATNUMP -> INTEGERP, since the integer must be
	nonnegative.

111 112 113 114
2011-06-06  Stefan Monnier  <monnier@iro.umontreal.ca>

	* window.h (Fwindow_frame): Declare.

115 116 117 118 119 120 121
2011-06-06  Paul Eggert  <eggert@cs.ucla.edu>

	* alloc.c: Simplify handling of large-request failures (Bug#8800).
	(SPARE_MEMORY): Always define.
	(LARGE_REQUEST): Remove.
	(memory_full): Use SPARE_MEMORY rather than LARGE_REQUEST.

122 123
2011-06-06  Martin Rudalics  <rudalics@gmx.at>

124 125 126 127 128 129 130 131 132 133 134 135 136 137 138
	* lisp.h: Move EXFUNS for Fframe_root_window,
	Fframe_first_window and Fset_frame_selected_window to window.h.

	* window.h: Move EXFUNS for Fframe_root_window,
	Fframe_first_window and Fset_frame_selected_window here from
	lisp.h.

	* frame.c (Fwindow_frame, Fframe_first_window)
	(Fframe_root_window, Fframe_selected_window)
	(Fset_frame_selected_window): Move to window.c.
	(Factive_minibuffer_window): Move to minibuf.c.
	(Fother_visible_frames_p): New function.

	* minibuf.c (Factive_minibuffer_window): Move here from frame.c.

139 140 141
	* window.c (decode_window, decode_any_window): Move up in code.
	(Fwindowp, Fwindow_live_p): Rewrite doc-strings.
	(inhibit_frame_unsplittable): Remove unused variable.
142 143 144
	(Fwindow_buffer): Move up and rewrite doc-string.
	(Fwindow_parent, Fwindow_vchild, Fwindow_hchild, Fwindow_next)
	(Fwindow_prev): New functions.
145 146 147 148 149 150 151
	(Fwindow_frame): Move here from frame.c.  Accept any window as
	argument.
	(Fframe_root_window, Fframe_first_window)
	(Fframe_selected_window): Move here from frame.c.  Accept frame
	or arbitrary window as argument.  Update doc-strings.
	(Fminibuffer_window): Move up in code.
	(Fwindow_minibuffer_p): Move up in code and simplify.
152 153
	(Fset_frame_selected_window): Move here from frame.c.
	Marginal rewrite.
154 155
	(Fselected_window, select_window, Fselect_window): Move up in
	code.  Minor doc-string fixes.
156

157 158 159 160 161 162 163 164
2011-06-06  Paul Eggert  <eggert@cs.ucla.edu>

	* alloc.c (memory_full) [SYSTEM_MALLOC]: Port to MacOS (Bug#8800).
	Do not assume that spare memory exists; that assumption is valid
	only if SYSTEM_MALLOC.
	(LARGE_REQUEST): New macro, so that the issue of large requests
	is separated from the issue of spare memory.

165 166
2011-06-05  Andreas Schwab  <schwab@linux-m68k.org>

Andreas Schwab's avatar
Andreas Schwab committed
167 168 169
	* editfns.c (Fformat): Correctly handle zero flag with hexadecimal
	format.  (Bug#8806)

170 171
	* gtkutil.c (xg_get_default_scrollbar_width): Avoid warning.

172 173 174
	* xfns.c (x_set_scroll_bar_default_width): Move declarations
	before statements.

175 176 177 178 179 180 181 182 183
2011-06-05  Jan Djärv  <jan.h.d@swipnet.se>

	* gtkutil.c (xg_get_default_scrollbar_width): New function.

	* gtkutil.h: Declare xg_get_default_scrollbar_width.

	* xfns.c (x_set_scroll_bar_default_width): If USE_GTK, get
	min width by calling x_set_scroll_bar_default_width (Bug#8505).

184 185 186 187
2011-06-05  Juanma Barranquero  <lekktu@gmail.com>

	* xdisp.c (single_display_spec_intangible_p): Remove declaration.

188 189 190 191
2011-06-04  Chong Yidong  <cyd@stupidchicken.com>

	* xselect.c (x_clipboard_manager_save): Remove redundant arg.
	(x_clipboard_manager_save): Add return value.
192 193
	(x_clipboard_manager_error_1, x_clipboard_manager_error_2):
	New error handlers.
194 195 196 197
	(x_clipboard_manager_save_frame, x_clipboard_manager_save_all):
	Obey Vx_select_enable_clipboard_manager.  Catch errors in
	x_clipboard_manager_save (Bug#8779).
	(Vx_select_enable_clipboard_manager): New variable.
198
	(x_get_foreign_selection): Reduce scope of x_catch_errors (Bug#8790).
199

200
2011-06-04  Dan Nicolaescu  <dann@ics.uci.edu>
201 202 203

	* emacs.c (main): Warn when starting a GTK emacs in daemon mode.

204
2011-06-04  YAMAMOTO Mitsuharu  <mituharu@math.s.chiba-u.ac.jp>
205 206 207 208

	* fringe.c (update_window_fringes): Don't update overlay arrow bitmap
	in the current matrix if keep_current_p is non-zero.

209 210 211 212
2011-06-04  Eli Zaretskii  <eliz@gnu.org>

	* bidi.c (bidi_level_of_next_char): Fix last change.

213 214
2011-06-03  Eli Zaretskii  <eliz@gnu.org>

215
	Support bidi reordering of text covered by display properties.
216

217 218 219 220
	* bidi.c (bidi_copy_it): Use offsetof instead of emulating it.
	(bidi_fetch_char, bidi_fetch_char_advance): New functions.
	(bidi_cache_search, bidi_cache_iterator_state)
	(bidi_paragraph_init, bidi_resolve_explicit, bidi_resolve_weak)
221 222
	(bidi_level_of_next_char, bidi_move_to_visually_next):
	Support character positions inside a run of characters covered by a
223 224 225 226 227 228 229 230 231 232
	display string.
	(bidi_paragraph_init, bidi_resolve_explicit_1)
	(bidi_level_of_next_char): Call bidi_fetch_char and
	bidi_fetch_char_advance instead of FETCH_CHAR and
	FETCH_CHAR_ADVANCE.
	(bidi_init_it): Initialize new members.
	(LRE_CHAR, RLE_CHAR, PDF_CHAR, LRO_CHAR, RLO_CHAR): Remove macro
	definitions.
	(bidi_explicit_dir_char): Lookup character type in bidi_type_table,
	instead of using explicit *_CHAR codes.
233 234
	(bidi_resolve_explicit, bidi_resolve_weak):
	Use FETCH_MULTIBYTE_CHAR instead of FETCH_CHAR, as reordering of
235 236 237 238 239 240 241 242 243 244
	bidirectional text is supported only in multibyte buffers.
	(bidi_init_it): Accept additional argument FRAME_WINDOW_P and use
	it to initialize the frame_window_p member of struct bidi_it.
	(bidi_cache_iterator_state, bidi_resolve_explicit_1)
	(bidi_resolve_explicit, bidi_resolve_weak)
	(bidi_level_of_next_char, bidi_move_to_visually_next): Abort if
	bidi_it->nchars is non-positive.
	(bidi_level_of_next_char): Don't try to lookup the cache for the
	next/previous character if nothing is cached there yet, or if we
	were just reseat()'ed to a new position.
245

246 247 248 249 250 251
	* xdisp.c (set_cursor_from_row): Set start and stop points
	according to the row's direction when priming the loop that looks
	for the glyph on which to display cursor.
	(single_display_spec_intangible_p): Function deleted.
	(display_prop_intangible_p): Reimplement to call
	handle_display_spec instead of single_display_spec_intangible_p.
252 253
	Accept 3 additional arguments needed by handle_display_spec.
	This fixes incorrect cursor motion across display property with complex
254 255 256 257 258
	values: lists, `(when COND...)' forms, etc.
	(single_display_spec_string_p): Support property values that are
	lists with the argument STRING its top-level element.
	(display_prop_string_p): Fix the condition for processing a
	property that is a list to be consistent with handle_display_spec.
259
	(handle_display_spec): New function, refactored from the
260 261 262 263 264 265 266 267 268 269
	last portion of handle_display_prop.
	(compute_display_string_pos): Accept additional argument
	FRAME_WINDOW_P.  Call handle_display_spec to determine whether the
	value of a `display' property is a "replacing spec".
	(handle_single_display_spec): Accept 2 additional arguments BUFPOS
	and FRAME_WINDOW_P.  If IT is NULL, don't set up the iterator from
	the display property, but just return a value indicating whether
	the display property will replace the characters it covers.
	(Fcurrent_bidi_paragraph_direction): Initialize the nchars and
	frame_window_p members of struct bidi_it.
270 271
	(compute_display_string_pos, compute_display_string_end):
	New functions.
272 273 274
	(push_it): Accept second argument POSITION, where pop_it should
	jump to continue iteration.
	(reseat_1): Initialize bidi_it.disp_pos.
275

276 277
	* keyboard.c (adjust_point_for_property): Adjust the call to
	display_prop_intangible_p to its new signature.
278 279

	* dispextern.h (struct bidi_it): New member frame_window_p.
280 281
	(bidi_init_it): Update prototypes.
	(display_prop_intangible_p): Update prototype.
282 283
	(compute_display_string_pos, compute_display_string_end):
	Declare prototypes.
284 285
	(struct bidi_it): New members nchars and disp_pos.  ch_len is now
	EMACS_INT.
286

287
2011-06-02  Paul Eggert  <eggert@cs.ucla.edu>
288

289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
	Malloc failure behavior now depends on size of allocation.
	* alloc.c (buffer_memory_full, memory_full): New arg NBYTES.
	* lisp.h: Change signatures accordingly.
	* alloc.c, buffer.c, editfns.c, menu.c, minibuf.c, xterm.c:
	All callers changed.  (Bug#8762)

	* gnutls.c: Use Emacs's memory allocators.
	Without this change, the gnutls library would invoke malloc etc.
	directly, which causes problems on non-SYNC_INPUT hosts, and which
	runs afoul of improving memory_full behavior.  (Bug#8761)
	(fn_gnutls_global_set_mem_functions): New macro or function pointer.
	(emacs_gnutls_global_init): Use it to specify xmalloc, xrealloc,
	xfree instead of the default malloc, realloc, free.
	(Fgnutls_boot): No need to check for memory allocation failure,
	since xmalloc does that for us.

Paul Eggert's avatar
Paul Eggert committed
305
	Remove arbitrary limit of 2**31 entries in hash tables.  (Bug#8771)
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
	* category.c (hash_get_category_set):
	* ccl.c (ccl_driver):
	* charset.c (Fdefine_charset_internal):
	* charset.h (struct charset.hash_index):
	* composite.c (get_composition_id, gstring_lookup_cache)
	(composition_gstring_put_cache):
	* composite.h (struct composition.hash_index):
	* dispextern.h (struct image.hash):
	* fns.c (next_almost_prime, larger_vector, cmpfn_eql)
	(cmpfn_equal, cmpfn_user_defined, hashfn_eq, hashfn_eql)
	(hashfn_equal, hashfn_user_defined, make_hash_table)
	(maybe_resize_hash_table, hash_lookup, hash_put)
	(hash_remove_from_table, hash_clear, sweep_weak_table, SXHASH_COMBINE)
	(sxhash_string, sxhash_list, sxhash_vector, sxhash_bool_vector)
	(Fsxhash, Fgethash, Fputhash, Fmaphash):
	* image.c (make_image, search_image_cache, lookup_image)
	(xpm_put_color_table_h):
	* lisp.h (struct Lisp_Hash_Table):
324
	* minibuf.c (Ftry_completion, Fall_completions, Ftest_completion):
325 326
	* print.c (print):  Use 'EMACS_UINT' and 'EMACS_INT'
	for hashes and hash indexes, instead of 'unsigned' and 'int'.
327 328 329 330
	* alloc.c (allocate_vectorlike):
	Check for overflow in vector size calculations.
	* ccl.c (ccl_driver):
	Check for overflow when converting EMACS_INT to int.
331 332
	* fns.c, image.c: Remove unnecessary static decls that would otherwise
	need to be updated by these changes.
333 334
	* fns.c (make_hash_table, maybe_resize_hash_table):
	Check for integer overflow with large hash tables.
335 336 337 338 339 340 341 342
	(make_hash_table, maybe_resize_hash_table, Fmake_hash_table):
	Prefer the faster XFLOAT_DATA to XFLOATINT where either will do.
	(SXHASH_REDUCE): New macro.
	(sxhash_string, sxhash_list, sxhash_vector, sxhash_bool_vector):
	Use it instead of discarding useful hash info with large hash values.
	(sxhash_float): New function.
	(sxhash): Use it.  No more need for "& INTMASK" due to above changes.
	* lisp.h (FIXNUM_BITS): New macro, useful for SXHASH_REDUCE etc.
343 344
	(MOST_NEGATIVE_FIXNUM, MOST_POSITIVE_FIXNUM, INTMASK):
	Rewrite to use FIXNUM_BITS, as this simplifies things.
345 346 347 348 349
	(next_almost_prime, larger_vector, sxhash, hash_lookup, hash_put):
	Adjust signatures to match updated version of code.
	(consing_since_gc): Now EMACS_INT, since a single hash table can
	use more than INT_MAX bytes.

350 351 352 353 354 355
2011-06-01  Dan Nicolaescu  <dann@ics.uci.edu>

	Make it possible to build with GCC-4.6+ -O2 -flto.

	* emacs.c (__malloc_initialize_hook): Mark as EXTERNALLY_VISIBLE.

356 357 358 359 360
2011-06-01  Stefan Monnier  <monnier@iro.umontreal.ca>

	* minibuf.c (get_minibuffer, read_minibuf_unwind):
	Call minibuffer-inactive-mode.

361 362 363 364 365
2011-05-31  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in ($(BLD)/data.$(O), $(BLD)/editfns.$(O)):
	Update dependencies.

366 367 368 369 370
2011-05-31  Dan Nicolaescu  <dann@ics.uci.edu>

	* data.c (init_data): Remove code for UTS, this system is not
	supported anymore.

371 372 373 374 375 376 377 378
2011-05-31  Dan Nicolaescu  <dann@ics.uci.edu>

	Don't force ./temacs to start in terminal mode.

	* frame.c (make_initial_frame): Initialize faces in all cases, not
	only when CANNOT_DUMP is defined.
	* dispnew.c (init_display): Remove CANNOT_DUMP condition.

379 380 381 382 383
2011-05-31  Dan Nicolaescu  <dann@ics.uci.edu>

	* dispnew.c (add_window_display_history): Use const for the string
	pointer.  Remove declaration, not needed.

Paul Eggert's avatar
Paul Eggert committed
384
2011-05-31  Paul Eggert  <eggert@cs.ucla.edu>
385

Paul Eggert's avatar
Paul Eggert committed
386
	Use 'inline', not 'INLINE'.
Paul Eggert's avatar
Paul Eggert committed
387
	<http://lists.gnu.org/archive/html/emacs-devel/2011-05/msg00914.html>
Paul Eggert's avatar
Paul Eggert committed
388 389 390 391 392 393 394
	* alloc.c, fontset.c (INLINE): Remove.
	* alloc.c, bidi.c, charset.c, coding.c, dispnew.c, fns.c, image.c:
	* intervals.c, keyboard.c, process.c, syntax.c, textprop.c, w32term.c:
	* xdisp.c, xfaces.c, xterm.c: Replace all uses of INLINE with inline.
	* gmalloc.c (register_heapinfo): Use inline unconditionally.
	* lisp.h (LISP_MAKE_RVALUE): Use inline, not __inline__.

395 396 397 398 399 400 401 402 403
2011-05-31  Dan Nicolaescu  <dann@ics.uci.edu>

	Make it possible to run ./temacs.

	* callproc.c (set_initial_environment): Remove CANNOT_DUMP code,
	syms_of_callproc does the same thing.  Remove test for
	"initialized", do it in the caller.
	* emacs.c (main): Avoid calling set_initial_environment when dumping.

404 405 406 407 408 409 410 411 412
2011-05-31  Stefan Monnier  <monnier@iro.umontreal.ca>

	* minibuf.c (Finternal_complete_buffer): Return `category' metadata.
	(read_minibuf): Use get_minibuffer.
	(syms_of_minibuf): Use DEFSYM.
	(Qmetadata): New var.
	* data.c (Qbuffer): Don't make it static.
	(syms_of_data): Use DEFSYM.

413 414 415 416 417
2011-05-31  Paul Eggert  <eggert@cs.ucla.edu>

	* ccl.c (CCL_CODE_RANGE): Allow negative numbers.  (Bug#8751)
	(CCL_CODE_MIN): New macro.

Paul Eggert's avatar
Paul Eggert committed
418 419
2011-05-30  Paul Eggert  <eggert@cs.ucla.edu>

420 421
	* alloc.c (lisp_align_malloc): Omit unnecessary val==NULL tests.

Paul Eggert's avatar
Paul Eggert committed
422 423 424 425 426
	* eval.c (Qdebug): Now static.
	* lisp.h (Qdebug): Remove decl.  This reverts a part of the
	2011-04-26T11:26:05Z!dan.colascione@gmail.com that inadvertently undid part of
	2011-04-14T06:48:41Z!eggert@cs.ucla.edu.

427 428 429 430 431
2011-05-29  Chong Yidong  <cyd@stupidchicken.com>

	* image.c: Various fixes to ImageMagick code comments.
	(Fimagemagick_types): Doc fix.

432 433
2011-05-29  Paul Eggert  <eggert@cs.ucla.edu>

434 435 436 437
	Minor fixes prompted by GCC 4.6.0 warnings.

	* xselect.c (converted_selections, conversion_fail_tag): Now static.

438 439 440 441
	* emacs.c [HAVE_X_WINDOWS]: Include "xterm.h".
	(x_clipboard_manager_save_all): Move extern decl to ...
	* xterm.h: ... here, so that it can be checked for consistency.

442 443 444 445 446 447 448 449 450 451 452
2011-05-29  Chong Yidong  <cyd@stupidchicken.com>

	* xselect.c (x_clipboard_manager_save_frame)
	(x_clipboard_manager_save_all): New functions.
	(Fx_clipboard_manager_save): Lisp function deleted.

	* emacs.c (Fkill_emacs): Call x_clipboard_manager_save_all.
	* frame.c (delete_frame): Call x_clipboard_manager_save_frame.

	* xterm.h: Update prototype.

453 454 455 456 457
2011-05-28  William Xu  <william.xwl@gmail.com>

	* nsterm.m (ns_term_shutdown): Synchronize user defaults before
	exiting (Bug#8239).

458 459
2011-05-28  Jim Meyering  <meyering@redhat.com>

Eli Zaretskii's avatar
Eli Zaretskii committed
460
	Avoid a sign-extension bug in crypto_hash_function.
461 462 463 464 465 466
	* fns.c (to_uchar): Define.
	(crypto_hash_function): Use it to convert some newly-signed
	variables to unsigned, to avoid sign-extension bugs.  For example,
	without this change, (md5 "truc") would evaluate to
	45723a2aff78ff4fff7fff1114760e62 rather than the expected
	45723a2af3788c4ff17f8d1114760e62.  Reported by Antoine Levitt in
Eli Zaretskii's avatar
Eli Zaretskii committed
467
	https://lists.gnu.org/archive/html/emacs-devel/2011-05/msg00883.html.
468

Paul Eggert's avatar
Paul Eggert committed
469 470 471
2011-05-27  Paul Eggert  <eggert@cs.ucla.edu>

	Integer overflow fixes.
472

473 474 475 476 477 478
	* dbusbind.c: Serial number integer overflow fixes.
	(CHECK_DBUS_SERIAL_GET_SERIAL): New macro.
	(Fdbus_call_method_asynchronously, xd_read_message_1): Use a float
	to hold a serial number that is too large for a fixnum.
	(Fdbus_method_return_internal, Fdbus_method_error_internal):
	Check for serial numbers out of range.  Decode any serial number
Paul Eggert's avatar
Paul Eggert committed
479
	that was so large that it became a float.  (Bug#8722)
480

481 482 483 484 485 486
	* dbusbind.c: Use XFASTINT rather than XUINT, and check for nonneg.
	(Fdbus_call_method, Fdbus_call_method_asynchronously):
	Use XFASTINT rather than XUINT when numbers are nonnegative.
	(xd_append_arg, Fdbus_method_return_internal):
	(Fdbus_method_error_internal): Likewise.  Also, for unsigned
	arguments, check that Lisp number is nonnegative, rather than
Paul Eggert's avatar
Paul Eggert committed
487
	silently wrapping negative numbers around.  (Bug#8722)
488
	(xd_read_message_1): Don't assume dbus_uint32_t can fit in int.
Paul Eggert's avatar
Paul Eggert committed
489
	(Bug#8722)
490

491 492
	* data.c (arith_driver, Flsh): Avoid unnecessary casts to EMACS_UINT.

493 494
	* ccl.c (ccl_driver): Redo slightly to avoid the need for 'unsigned'.

Paul Eggert's avatar
Paul Eggert committed
495 496 497 498 499 500 501
	ccl: add integer overflow checks
	* ccl.c (CCL_CODE_MAX, GET_CCL_RANGE, GET_CCL_CODE, GET_CCL_INT):
	(IN_INT_RANGE): New macros.
	(ccl_driver): Use them to check for integer overflow when
	decoding a CCL program.  Many of the new checks are whether XINT (x)
	fits in int; it doesn't always, on 64-bit hosts.  The new version
	doesn't catch all possible integer overflows, but it's an
Paul Eggert's avatar
Paul Eggert committed
502
	improvement.  (Bug#8719)
Paul Eggert's avatar
Paul Eggert committed
503

504 505 506
	* alloc.c (make_event_array): Use XINT, not XUINT.
	There's no need for unsigned here.

507 508 509 510
	* mem-limits.h (EXCEEDS_LISP_PTR) [!USE_LSB_TAG]: EMACS_UINT -> uintptr_t
	This follows up to the 2011-05-06 change that substituted uintptr_t
	for EMACS_INT.  This case wasn't caught back then.

511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530
	Rework Fformat to avoid integer overflow issues.
	* editfns.c: Include <float.h> unconditionally, as it's everywhere
	now (part of C89).  Include <verify.h>.
	(MAX_10_EXP, CONVERTED_BYTE_SIZE): Remove; no longer needed.
	(pWIDE, pWIDElen, signed_wide, unsigned_wide): New defns.
	(Fformat): Avoid the prepass trying to compute sizes; it was only
	approximate and thus did not catch overflow reliably.  Instead, walk
	through the format just once, formatting and computing sizes as we go,
	checking for integer overflow at every step, and allocating a larger
	buffer as needed.  Keep track separately whether the format is
	multibyte.  Keep only the most-recently calculated precision, rather
	than them all.  Record whether each argument has been converted to
	string.  Use EMACS_INT, not int, for byte and char and arg counts.
	Support field widths and precisions larger than INT_MAX.  Avoid
	sprintf's undefined behavior with conversion specifications such as %#d
	and %.0c.  Fix bug with strchr succeeding on '\0' when looking for
	flags.  Fix bug with (format "%c" 256.0).  Avoid integer overflow when
	formatting out-of-range floating point numbers with int
	formats. (Bug#8668)

531 532
	* lisp.h (FIXNUM_OVERFLOW_P): Work even if arg is a NaN.

533 534 535 536 537 538 539 540 541
	* data.c: Avoid integer truncation in expressions involving floats.
	* data.c: Include <intprops.h>.
	(arith_driver): When there's an integer overflow in an expression
	involving floating point, convert the integers to floating point
	so that the resulting value does not suffer from catastrophic
	integer truncation.  For example, on a 64-bit host (* 4
	most-negative-fixnum 0.5) should yield about -4.6e+18, not zero.
	Do not rely on undefined behavior after integer overflow.

542 543
	merge count_size_as_multibyte, parse_str_to_multibyte
	* character.c, character.h (count_size_as_multibyte):
544
	Rename from parse_str_to_multibyte; all uses changed.
545 546 547 548 549 550
	Check for integer overflow.
	* insdel.c, lisp.h (count_size_as_multibyte): Remove,
	since it's now a duplicate of the other.  This is more of
	a character than a buffer op, so better that it's in character.c.
	* fns.c, print.c: Adjust to above changes.

551 552 553 554
2011-05-27  Stefan Monnier  <monnier@iro.umontreal.ca>

	* xselect.c (x_convert_selection): Yet another int/Lisp_Object mixup.

555 556
2011-05-27  Paul Eggert  <eggert@cs.ucla.edu>

557 558 559 560 561
	* xselect.c: Fix minor problems prompted by GCC 4.6.0 warnings.
	(x_handle_selection_request, frame_for_x_selection): Remove unused vars.
	(x_clipboard_manager_save): Now static.
	(Fx_clipboard_manager_save): Rename local to avoid shadowing.

562 563 564 565
	* fns.c: Fix minor problems prompted by GCC 4.6.0 warnings.
	(crypto_hash_function): Now static.
	Fix pointer signedness problems.  Avoid unnecessary initializations.

Chong Yidong's avatar
Chong Yidong committed
566 567 568 569 570 571 572 573 574 575 576 577 578
2011-05-27  Chong Yidong  <cyd@stupidchicken.com>

	* termhooks.h (Vselection_alist): Make it terminal-local.

	* terminal.c (create_terminal): Initialize it.

	* xselect.c: Support for clipboard managers.
	(Vselection_alist): Move to termhooks.h as terminal-local var.
	(LOCAL_SELECTION): New macro.
	(x_atom_to_symbol): Handle x_display_info_for_display fail case.
	(symbol_to_x_atom): Remove gratuitous arg.
	(x_handle_selection_request, lisp_data_to_selection_data)
	(x_get_foreign_selection, Fx_register_dnd_atom): Callers changed.
579 580
	(x_own_selection, x_get_local_selection, x_convert_selection):
	New arg, specifying work frame.  Use terminal-local Vselection_alist.
Chong Yidong's avatar
Chong Yidong committed
581 582 583 584 585 586 587 588 589 590 591 592 593
	(some_frame_on_display): Delete unused function.
	(Fx_own_selection_internal, Fx_get_selection_internal)
	(Fx_disown_selection_internal, Fx_selection_owner_p)
	(Fx_selection_exists_p): New optional frame arg.
	(frame_for_x_selection, Fx_clipboard_manager_save): New functions.
	(x_handle_selection_clear): Don't treat other terminals with the
	same keyboard specially.  Use the terminal-local Vselection_alist.
	(x_clear_frame_selections): Use Frun_hook_with_args.

	* xterm.c (x_term_init): Intern ATOM and CLIPBOARD_MANAGER atoms.

	* xterm.h: Add support for those atoms.

594 595 596 597 598 599 600
2011-05-26  Chong Yidong  <cyd@stupidchicken.com>

	* xselect.c: ICCCM-compliant handling of MULTIPLE targets.
	(converted_selections, conversion_fail_tag): New global variables.
	(x_selection_request_lisp_error): Free the above.
	(x_get_local_selection): Remove unnecessary code.
	(x_reply_selection_request): Args changed; handle arbitrary array
601 602
	of converted selections stored in converted_selections.
	Separate the XChangeProperty and SelectionNotify steps.
603 604 605 606 607 608 609 610 611
	(x_handle_selection_request): Rewrite to handle MULTIPLE target.
	(x_convert_selection): New function.
	(x_handle_selection_event): Simplify.
	(x_get_foreign_selection): Don't ignore incoming requests while
	waiting for an answer; this will fail when we implement
	SAVE_TARGETS, and seems unnecessary anyway.
	(selection_data_to_lisp_data): Recognize ATOM_PAIR type.
	(Vx_sent_selection_functions): Doc fix.

612 613 614 615
2011-05-26  Leo Liu  <sdl.web@gmail.com>

	* editfns.c (Ftranspose_regions): Allow empty regions.  (Bug#8699)

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632
2011-05-25  YAMAMOTO Mitsuharu  <mituharu@math.s.chiba-u.ac.jp>

	* dispextern.h (struct glyph_row): New member fringe_bitmap_periodic_p.

	* dispnew.c (shift_glyph_matrix, scrolling_window): Mark scrolled row
	for fringe update if it has periodic bitmap.
	(row_equal_p): Also compare left_fringe_offset,	right_fringe_offset,
	and fringe_bitmap_periodic_p.

	* fringe.c (get_fringe_bitmap_data): New function.
	(draw_fringe_bitmap_1, update_window_fringes): Use it.
	(update_window_fringes): Record periodicity of fringe bitmap in glyph
	row.  Mark glyph row for fringe update if periodicity changed.

	* xdisp.c (try_window_reusing_current_matrix): Don't mark scrolled row
	for fringe update unless it has periodic bitmap.

633 634 635 636 637
2011-05-25  Kenichi Handa  <handa@m17n.org>

	* xdisp.c (get_next_display_element): Set correct it->face_id for
	a static composition.

638 639 640 641 642 643 644 645 646
2011-05-24  Leo Liu  <sdl.web@gmail.com>

	* deps.mk (fns.o):
	* makefile.w32-in ($(BLD)/fns.$(O)): Include sha1.h.

	* fns.c (crypto_hash_function, Fsha1): New function.
	(Fmd5): Use crypto_hash_function.
	(syms_of_fns): Add Ssha1.

647 648 649 650 651 652 653 654
2011-05-22  Paul Eggert  <eggert@cs.ucla.edu>

	* gnutls.c: Remove unused macros.
	(fn_gnutls_transport_set_lowat, fn_gnutls_transport_set_pull_function):
	(fn_gnutls_transport_set_push_function) [!WINDOWSNT]:
	Remove macros that are defined and never used.
	Caught by gcc -Wunused-macros (GCC 4.6.0, Fedora 14).

655 656 657 658 659 660 661
2011-05-22  Chong Yidong  <cyd@stupidchicken.com>

	* xselect.c (syms_of_xselect): Remove unused symbol SAVE_TARGETS.
	(Fx_get_selection_internal): Minor cleanup.
	(Fx_own_selection_internal): Rename arguments for consistency with
	select.el.

662 663 664 665
2011-05-22  Paul Eggert  <eggert@cs.ucla.edu>

	* xselect.c (QSAVE_TARGETS): New static var, to fix build failure.

666 667 668 669
2011-05-22  Chong Yidong  <cyd@stupidchicken.com>

	* xselect.c (syms_of_xselect): Include character.h; use DEFSYM.

670 671 672 673 674
2011-05-21  YAMAMOTO Mitsuharu  <mituharu@math.s.chiba-u.ac.jp>

	* dispnew.c (scrolling_window): Don't exclude the case that the
	last enabled row in the desired matrix touches the bottom boundary.

675 676 677
2011-05-21  Glenn Morris  <rgm@gnu.org>

	* Makefile.in ($(etc)/DOC): Make second command line even shorter.
678 679
	(SOME_MACHINE_OBJECTS): Replace FONT_OBJ by its maximal expansion,
	and add some more files.
680

681 682 683 684 685
2011-05-20  Eli Zaretskii  <eliz@gnu.org>

	* callproc.c (Fcall_process) [MSDOS]: Fix arguments to
	report_file_error introduced by the change from 2011-05-07.

686 687 688 689 690 691 692
2011-05-20  Paul Eggert  <eggert@cs.ucla.edu>

	* systime.h (Time): Define only if emacs is defined.
	This is to allow ../lib-src/profile.c to be compiled on FreeBSD,
	where the include path doesn't have X11/X.h by default.  See
	<http://lists.gnu.org/archive/html/emacs-devel/2011-05/msg00561.html>.

Kenichi Handa's avatar
Kenichi Handa committed
693
2011-05-20 Kenichi Handa  <handa@m17n.org>
694 695 696

	* composite.c (find_automatic_composition): Fix previous change.

697 698 699 700 701 702 703
2011-05-20  Glenn Morris  <rgm@gnu.org>

	* lisp.mk: New file, split from Makefile.in.
	* Makefile.in (lisp): Move to separate file, inserted by @lisp_frag@.
	(shortlisp): Remove.
	($(etc)/DOC): Edit lisp.mk rather than using $shortlisp.

704 705 706 707 708 709 710 711 712 713 714
2011-05-19  Glenn Morris  <rgm@gnu.org>

	* Makefile.in (MSDOS_SUPPORT_REAL, MSDOS_SUPPORT, NS_SUPPORT)
	(REAL_MOUSE_SUPPORT, GPM_MOUSE_SUPPORT, MOUSE_SUPPORT, TOOLTIP_SUPPORT)
	(BASE_WINDOW_SUPPORT, X_WINDOW_SUPPORT, WINDOW_SUPPORT): Remove.
	(lisp): Set the order to that of loadup.el.
	(shortlisp): Make it a copy of $lisp.
	(SOME_MACHINE_LISP): Remove.
	($(etc)/DOC): Depend just on $lisp, not $SOME_MACHINE_LISP too.
	Use just $shortlisp, not $SOME_MACHINE_LISP too.

715 716 717 718 719 720
2011-05-18  Kenichi Handa  <handa@m17n.org>

	* composite.c (CHAR_COMPOSABLE_P): Add more check for efficiency.
	(BACKWARD_CHAR): Wrap the arg STOP by parenthesis.
	(find_automatic_composition): Mostly rewrite for efficiency.

721 722 723 724
2011-05-18  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in: Update dependencies.

725 726 727 728 729
2011-05-18  Christoph Scholtes  <cschol2112@googlemail.com>

	* menu.c: Include limits.h (fixes the MS-Windows build broken by
	revision 104625).

Paul Eggert's avatar
Paul Eggert committed
730
2011-05-18  Paul Eggert  <eggert@cs.ucla.edu>
731

732 733
	Fix some integer overflow issues, such as string length overflow.

734 735
	* insdel.c (count_size_as_multibyte): Check for string overflow.

736 737 738 739 740 741 742 743 744
	* character.c (lisp_string_width): Check for string overflow.
	Use EMACS_INT, not int, for string indexes and lengths; in
	particular, 2nd arg is now EMACS_INT, not int.  Do not crash if
	the resulting string length overflows an EMACS_INT; instead,
	report a string overflow if no precision given.  When checking for
	precision exhaustion, use a check that cannot possibly have
	integer overflow.  (Bug#8675)
	* character.h (lisp_string_width): Adjust to new signature.

745 746 747 748 749 750
	* alloc.c (string_overflow): New function.
	(Fmake_string): Use it.  This doesn't change behavior, but saves
	a few bytes and will simplify future changes.
	* character.c (string_escape_byte8): Likewise.
	* lisp.h (string_overflow): New decl.

751 752 753
	Fixups, following up to the user-interface timestamp change.
	* nsterm.m (last_mouse_movement_time, ns_mouse_position): Use Time
	for UI timestamps, instead of unsigned long.
754 755
	* msdos.c (mouse_get_pos): Likewise.
	* w32inevt.c (movement_time, w32_console_mouse_position): Likewise.
756 757 758
	* w32gui.h (Time): Define by including "systime.h" rather than by
	declaring it ourselves.  (Bug#8664)

759 760 761
	* dispextern.h (struct image): Don't assume time_t <= unsigned long.
	* image.c (clear_image_cache): Likewise.

762 763
	* term.c (term_mouse_position): Don't assume time_t wraparound.

764 765
	Be more systematic about user-interface timestamps.
	Before, the code sometimes used 'Time', sometimes 'unsigned long',
766 767
	and sometimes 'EMACS_UINT', to represent these timestamps.
	This change causes it to use 'Time' uniformly, as that's what X uses.
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786
	This makes the code easier to follow, and makes it easier to catch
	integer overflow bugs such as Bug#8664.
	* frame.c (Fmouse_position, Fmouse_pixel_position):
	Use Time, not unsigned long, for user-interface timestamps.
	* keyboard.c (last_event_timestamp, kbd_buffer_get_event): Likewise.
	(button_down_time, make_lispy_position, make_lispy_movement): Likewise.
	* keyboard.h (last_event_timestamp): Likewise.
	* menu.c (Fx_popup_menu) [!HAVE_X_WINDOWS]: Likewise.
	* menu.h (xmenu_show): Likewise.
	* term.c (term_mouse_position): Likewise.
	* termhooks.h (struct input_event.timestamp): Likewise.
	(struct terminal.mouse_position_hook): Likewise.
	* xmenu.c (create_and_show_popup_menu, xmenu_show): Likewise.
	* xterm.c (XTmouse_position, x_scroll_bar_report_motion): Likewise.
	* systime.h (Time): New decl.  Pull it in from <X11/X.h> if
	HAVE_X_WINDOWS, otherwise define it as unsigned long, which is
	what it was before.
	* menu.h, termhooks.h: Include "systime.h", for Time.

787 788 789 790 791 792 793
	* keyboard.c (make_lispy_event): Fix problem in integer overflow.
	Don't assume that the difference between two unsigned long values
	can fit into an integer.  At this point, we know button_down_time
	<= event->timestamp, so the difference must be nonnegative, so
	there's no need to cast the result if double-click-time is
	nonnegative, as it should be; check that it's nonnegative, just in
	case.  This bug is triggered when events are more than 2**31 ms
Paul Eggert's avatar
Paul Eggert committed
794
	apart (about 25 days).  (Bug#8664)
795

796
	* xselect.c (last_event_timestamp): Remove duplicate decl.
797
	(x_own_selection): Remove needless cast to unsigned long.
798

799 800 801
	* xmenu.c (set_frame_menubar): Use int, not EMACS_UINT, for indexes
	that always fit in int.  Use a sentinel instead of a counter, to
	avoid a temp and to allay GCC's concerns about possible int overflow.
802 803
	* frame.h (struct frame): Use int for menu_bar_items_used
	instead of EMACS_INT, since it always fits in int.
804

805 806
	* menu.c (grow_menu_items): Check for int overflow.

807 808
	* xmenu.c (set_frame_menubar): Don't mishandle vectors with no nils.

809 810 811 812 813 814 815 816 817
	* xterm.c: Use EMACS_INT for Emacs modifiers, and int for X modifiers.
	Before, the code was not consistent.  These values cannot exceed
	2**31 - 1 so there's no need to make them unsigned.
	(x_x_to_emacs_modifiers): Accept int and return EMACS_INT.
	(x_emacs_to_x_modifiers): Accept EMACS_INT and return int.
	(x_x_to_emacs_modifiers, x_emacs_to_x_modifiers): Reject non-integers
	as modifiers.
	* xterm.h (x_x_to_emacs_modifiers): Adjust to signature change.

818 819 820 821 822
	* lisp.h (XINT) [USE_LISP_UNION_TYPE]: Cast to EMACS_INT.
	(XUINT) [USE_LISP_UNION_TYPE]: Cast to EMACS_UINT.
	Otherwise, GCC 4.6.0 warns about printf (pI, XINT (...)),
	presumably because the widths might not match.

823 824
	* window.c (size_window): Avoid needless test at loop start.

825 826 827 828
2011-05-18  Courtney Bane  <emacs-bugs-7626@cbane.org>  (tiny change)

	* term.c (Fresume_tty): Restore hooks before reinitializing (bug#8687).

829
2011-05-12  Drew Adams  <drew.adams@oracle.com>
830 831 832

	* textprop.c (Fprevious_single_char_property_change): Doc fix (bug#8655).

833
2011-05-12  YAMAMOTO Mitsuharu  <mituharu@math.s.chiba-u.ac.jp>
834 835 836 837 838

	* w32term.c (w32_draw_fringe_bitmap): Rename local vars `left' and
	`width' to `bar_area_x' and `bar_area_width', respectively.
	(x_scroll_run): Take account of fringe background extension.

839 840
	* xterm.c (x_draw_fringe_bitmap) [USE_TOOLKIT_SCROLL_BARS]:
	Rename local vars `left' and `width' to `bar_area_x' and
841 842 843 844
	`bar_area_width', respectively.
	(x_scroll_run) [USE_TOOLKIT_SCROLL_BARS]: Take account of fringe
	background extension.

845 846 847 848
2011-05-10  Jim Meyering  <meyering@redhat.com>

	* xdisp.c (x_intersect_rectangles): Fix typo "the the -> the".

849 850 851 852 853
2011-05-10  Juanma Barranquero  <lekktu@gmail.com>

	* image.c (Finit_image_library): Return t for built-in image types,
	like pbm and xbm.  (Bug#8640)

854 855 856 857
2011-05-09  Andreas Schwab  <schwab@linux-m68k.org>

	* w32menu.c (set_frame_menubar): Fix submenu allocation.

858 859
2011-05-07  Eli Zaretskii  <eliz@gnu.org>

860 861 862 863
	* w32console.c (Fset_screen_color): Doc fix.
	(Fget_screen_color): New function.
	(syms_of_ntterm): Defsubr it.

864 865 866 867 868
	* callproc.c (call_process_cleanup) [MSDOS]: Don't close and
	unlink the temporary file if Fcall_process didn't create it in the
	first place.
	(Fcall_process) [MSDOS]: Don't create tempfile if stdout of the
	child process will be redirected to a file specified with `:file'.
869 870 871
	Don't try to re-open tempfile in that case, and set fd[0] to -1 as
	cue to call_process_cleanup not to close that handle.

872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888
2011-05-07  Ben Key  <bkey76@gmail.com>

	* makefile.w32-in: The bootstrap-temacs rule now makes use of
	one of two shell specific rules, either bootstrap-temacs-CMD or
	bootstrap-temacs-SH.  The bootstrap-temacs-SH rule is identical
	to the previous implementation of the bootstrap-temacs rule.
	The bootstrap-temacs-CMD rule is similar to the previous
	implementation of the bootstrap-temacs rule except that it
	makes use of the ESC_CFLAGS variable instead of the CFLAGS
	variable.

	These changes, along with some changes to nt/configure.bat,
	nt/gmake.defs, and nt/nmake.defs, are required to extend my
	earlier fix to add support for --cflags and --ldflags options
	that include quotes so that it works whether make uses cmd or
	sh as the shell.

889 890 891 892 893 894 895 896 897 898
2011-05-06  Michael Albinus  <michael.albinus@gmx.de>

	* dbusbind.c (QCdbus_type_unix_fd): Declare static.
	(xd_remove_watch): Don't check QCdbus_type_unix_fd for SYMBOLP, it
	is a constant.
	(Fdbus_init_bus, xd_read_queued_messages): Bus can be a symbol or
	a string.  Handle both cases.
	(Fdbus_call_method_asynchronously, Fdbus_register_signal)
	(Fdbus_register_method): Use Qinvalid_function.

899 900 901 902 903 904
2011-05-06  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in: Update dependencies.
	(LISP_H): Add inttypes.h and stdin.h.
	(PROCESS_H): Add unistd.h.

905 906 907 908 909
2011-05-06  Eli Zaretskii  <eliz@gnu.org>

	* lread.c: Include limits.h (fixes the MS-Windows build broken by
	2011-05-06T07:13:19Z!eggert@cs.ucla.edu).

Paul Eggert's avatar
Paul Eggert committed
910
2011-05-06  Paul Eggert  <eggert@cs.ucla.edu>
911

912 913
	* image.c (Finit_image_library) [!HAVE_NTGUI]: Omit unused local.

914 915 916
	* term.c (vfatal): Remove stray call to va_end.
	It's not needed and the C Standard doesn't allow it here anyway.

917 918 919
	Use C99's va_copy to avoid undefined behavior on x86-64 GNU/Linux.
	* eval.c (verror): doprnt a copy of ap, not the original.  (Bug#8545)

920 921 922
	* eval.c (verror): OK to create a string of up to MOST_POSITIVE_FIXNUM
	bytes.

923 924
	* term.c: Don't include <stdarg.h>, as <lisp.h> does that.

925 926
	* callproc.c (Fcall_process): Use 'volatile' to avoid vfork clobbering.

927 928
	* process.c (Fformat_network_address): Fix typo: args2 -> *args2.

929 930
	* xmenu.c (set_frame_menubar): Fix typo: int * -> int (3 times).

931 932
	* coding.c (detect_coding_charset): Fix typo: * 2 -> *4 (Bug#8601).

933 934 935
	* charset.h (struct charset.code_space): Now has 15 elements, not 16.
	* charset.c (Fdefine_charset_internal): Don't initialize
	charset.code_space[15].  The value was garbage, on hosts with
Paul Eggert's avatar
Paul Eggert committed
936
	32-bit int (Bug#8600).
937 938 939 940 941 942 943

	* lread.c (read_integer): Be more consistent with string-to-number.
	Use string_to_number to do the actual conversion; this avoids
	rounding errors and fixes some other screwups.  Without this fix,
	for example, #x1fffffffffffffff was misread as -2305843009213693952.
	(digit_to_number): Move earlier, for benefit of read_integer.
	Return -1 if the digit is out of range for the base, -2 if it is
Paul Eggert's avatar
Paul Eggert committed
944
	not a digit in any supported base.  (Bug#8602)
945

946 947
	* doprnt.c (doprnt): Support arbitrary pI values, such as "I64".

948 949 950 951
	* dispnew.c (scrolling_window): Return 1 if we scrolled,
	to match comment at start of function.  This also removes a
	GCC warning about overflow in a 32+64-bit port.

952 953
	* lisp.h (EMACS_INT, EMACS_UINT, BITS_PER_EMACS_INT, pI): Simplify.

954 955 956
	* dbusbind.c: Do not use XPNTR on a value that may be an integer.
	Reported by Stefan Monnier in
	<http://lists.gnu.org/archive/html/emacs-devel/2011-04/msg00919.html>.
957 958
	(xd_remove_watch, Fdbus_init_bus, xd_read_queued_messages):
	Use SYMBOLP-guarded XSYMBOL, not XPNTR.
959

960 961 962
	* lisp.h (EMACS_INTPTR): Remove.  All uses changed to intptr_t.
	(EMACS_UINTPTR): Likewise, with uintptr_t.

963 964 965 966
	* lisp.h: Prefer 64-bit EMACS_INT if available.
	(EMACS_INT, EMACS_UINT, BITS_PER_EMACS_INT, pI): Define to 64-bit
	on 32-bit hosts that have 64-bit int, so that they can access
	large files.
967 968
	However, temporarily disable this change unless the temporary
	symbol WIDE_EMACS_INT is defined.
969

970 971
	* lread.c, process.c: Do not include <inttypes.h>; lisp.h does it now.

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 1001
	Prefer intptr_t/uintptr_t for integers the same widths as pointers.
	This removes an assumption that EMACS_INT and long are the same
	width as pointers.  The assumption is true for Emacs porting targets
	now, but we want to make other targets possible.
	* lisp.h: Include <inttypes.h>, for INTPTR_MAX, UINTPTR_MAX.
	(EMACS_INTPTR, EMACS_UINTPTR): New macros.
	In the rest of the code, change types of integers that hold casted
	pointers to EMACS_INTPTR and EMACS_UINTPTR, systematically
	replacing EMACS_INT, long, EMACS_UINT, and unsigned long.
	(XTYPE): Don't cast arg to EMACS_UINT; normally is not needed.
	(XSET): Cast type of XTYPE arg to EMACS_INTPTR; it is needed here.
	No need to cast type when ORing.
	(XPNTR): Return a value of type EMACS_INTPTR or EMACS_UINTPTR.
	* alloc.c (lisp_align_malloc): Remove a no-longer-needed cast.
	* doc.c (store_function_docstring): Use EMACS_INTPTR, so as not to
	assume EMACS_INT is the same width as char *.
	* gtkutil.c (xg_gtk_scroll_destroy, xg_tool_bar_button_cb):
	(xg_tool_bar_callback, xg_tool_bar_help_callback, xg_make_tool_item):
	Remove no-longer-needed casts.
	(xg_create_scroll_bar, xg_tool_bar_button_cb, xg_tool_bar_callback):
	(xg_tool_bar_help_callback, xg_make_tool_item):
	Use EMACS_INTPTR to hold an integer
	that will be cast to void *; this can avoid a GCC warning
	if EMACS_INT is not the same width as void *.
	* menu.c (find_and_call_menu_selection): Remove no-longer-needed cast.
	* xdisp.c (display_echo_area_1, resize_mini_window_1):
	(current_message_1, set_message_1):
	Use a local to convert to proper width without a cast.
	* xmenu.c (dialog_selection_callback): Likewise.

1002 1003 1004 1005 1006 1007
	* sysdep.c (get_random): Don't assume EMACS_INT is no wider than long.
	Also, don't assume VALBITS / RAND_BITS is less than 5,
	and don't rely on undefined behavior when shifting a 1 left into
	the sign bit.
	* lisp.h (get_random): Change signature to match.

1008 1009 1010 1011 1012 1013 1014 1015
	* lread.c (hash_string): Use size_t, not int, for hash computation.
	Normally we prefer signed values; but hashing is special, because
	it's better to use unsigned division on hash table sizes so that
	the remainder is nonnegative.  Also, size_t is the natural width
	for hashing into memory.  The previous code used 'int', which doesn't
	retain enough info to hash well into very large tables.
	(oblookup, oblookup_last_bucket_number, Funintern): Likewise.

1016 1017 1018 1019 1020
	* dbusbind.c: Don't possibly lose pointer info when converting.
	(xd_remove_watch, Fdbus_init_bus, xd_read_queued_messages):
	Use XPNTR rather than XHASH, so that the high-order bits of
	the pointer aren't lost when converting through void *.

1021 1022
	* eval.c (Fautoload): Don't double-shift a pointer.

1023 1024
	* fns.c (Frandom): Let EMACS_UINT be wider than unsigned long.

1025 1026 1027 1028 1029
2011-05-06  Juanma Barranquero  <lekktu@gmail.com>

	* gnutls.c (DEF_GNUTLS_FN):
	* image.c (DEF_IMGLIB_FN): Make function pointers static.

1030 1031 1032 1033 1034
2011-05-05  Andreas Schwab  <schwab@linux-m68k.org>

	* lread.c (lisp_file_lexically_bound_p): Stop scanning at end
	marker.  (Bug#8610)

1035
2011-05-05 Eli Zaretskii  <eliz@gnu.org>
1036 1037 1038 1039

	* w32heap.c (allocate_heap) [USE_LISP_UNION_TYPE || USE_LSB_TAG]:
	New version that can reserve upto 2GB of heap space.

1040
2011-05-05  Chong Yidong  <cyd@stupidchicken.com>
1041 1042 1043

	* nsfns.m (Fns_read_file_name): Doc fix (Bug#8534).

1044 1045 1046 1047 1048
2011-05-05  Teodor Zlatanov  <tzz@lifelogs.com>

	* gnutls.c (fn_gnutls_certificate_set_x509_key_file): Add alias to
	`gnutls_certificate_set_x509_key_file'.

1049 1050 1051 1052 1053
2011-05-05  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in ($(BLD)/image.$(O), $(BLD)/process.$(O)):
	Update dependencies.

1054 1055 1056 1057 1058 1059 1060
2011-05-04  Juanma Barranquero  <lekktu@gmail.com>

	* gnutls.h (emacs_gnutls_write, emacs_gnutls_read):
	* gnutls.c (emacs_gnutls_write, emacs_gnutls_read):
	Remove unused parameter `fildes'.
	* process.c (read_process_output, send_process): Don't pass it.

1061 1062 1063 1064 1065 1066
2011-05-04  Juanma Barranquero  <lekktu@gmail.com>

	Fix previous change: the library cache is defined in w32.c.
	* image.c (CACHE_IMAGE_TYPE) [!HAVE_NTGUI]: Define to noop.
	(Finit_image_library): Wrap Vlibrary_cache on "#ifdef HAVE_NTGUI".

1067 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 1107 1108 1109 1110 1111
2011-05-04  Juanma Barranquero  <lekktu@gmail.com>

	Implement dynamic loading of GnuTLS on Windows.

	* gnutls.h (GNUTLS_EMACS_ERROR_NOT_LOADED): New macro.
	(emacs_gnutls_write, emacs_gnutls_read): Mark as extern.
	(emacs_gnutls_record_check_pending, emacs_gnutls_transport_set_errno):
	Declare.

	* gnutls.c (Qgnutls_dll): Define.
	(DEF_GNUTLS_FN, LOAD_GNUTLS_FN): New macros.
	(gnutls_*): Declare function pointers.
	(init_gnutls_functions): New function to initialize function pointers.
	(emacs_gnutls_handshake, Fgnutls_error_string, Fgnutls_deinit)
	(emacs_gnutls_global_init, Fgnutls_bye): Use function pointers.
	(emacs_gnutls_record_check_pending, emacs_gnutls_transport_set_errno):
	Wrappers for gnutls_record_check_pending and gnutls_transport_set_errno.
	(emacs_gnutls_write, emacs_gnutls_read)
	(emacs_gnutls_handle_error, Fgnutls_error_fatalp)
	(Fgnutls_available_p): New function.
	(Fgnutls_boot): Call Fgnutls_available_p.  Use function pointers.
	(syms_of_gnutls) <Qgnutls_dll>: Initialize and staticpro it.
	(syms_of_gnutls) <Sgnutls_available_p>: defsubr it.

	* image.c: Include w32.h.
	(Vimage_type_cache): Delete.
	(syms_of_image) <Vimage_type_cache>: Don't initialize and staticpro it.
	(CACHE_IMAGE_TYPE, Finit_image_library): Use Vlibrary_cache instead.
	(w32_delayed_load): Move to w32.c.

	* w32.h (VlibraryCache, QCloaded_from, w32_delayed_load): Declare.

	* w32.c (QCloaded_from, Vlibrary_cache): Define.
	(w32_delayed_load): Move from image.c.  When loading a library, record
	its filename in the :loaded-from property of the library id.
	(globals_of_w32) <QCloaded_from, Vlibrary_cache>:
	Initialize and staticpro them.
	(emacs_gnutls_pull, emacs_gnutls_push): Call emacs_gnutls_* functions.

	* process.c: Include lisp.h before w32.h, not after.
	(wait_reading_process_output): Call emacs_gnutls_record_check_pending
	instead of gnutls_record_check_pending.

	* callproc.c, emacs.c: Include lisp.h before w32.h, not after.

1112 1113 1114 1115 1116 1117
2011-05-04  Teodor Zlatanov  <tzz@lifelogs.com>

	* gnutls.c (Fgnutls_boot): Support :keylist and :crlfiles options
	instead of :keyfiles.  Give GnuTLS the keylist and the CRL lists
	as passed in.

1118 1119 1120 1121 1122
2011-05-03  Jan Djärv  <jan.h.d@swipnet.se>

	* xterm.c (x_set_frame_alpha): Do not set property on anything
	else than FRAME_X_OUTER_WINDOW (Bug#8608).

1123 1124 1125 1126
2011-05-02  Juanma Barranquero  <lekktu@gmail.com>

	* sysdep.c (get_tty_size) [WINDOWSNT]: Implement.  (Bug#8596)

1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138
2011-05-02  Juanma Barranquero  <lekktu@gmail.com>

	* gnutls.c (Qgnutls_log_level, Qgnutls_code, Qgnutls_anon)
	(Qgnutls_x509pki, Qgnutls_e_interrupted, Qgnutls_e_again)
	(Qgnutls_e_invalid_session, Qgnutls_e_not_ready_for_handshake)
	(gnutls_global_initialized, Qgnutls_bootprop_priority)
	(Qgnutls_bootprop_trustfiles, Qgnutls_bootprop_keyfiles)
	(Qgnutls_bootprop_callbacks, Qgnutls_bootprop_loglevel)
	(Qgnutls_bootprop_hostname, Qgnutls_bootprop_verify_flags)
	(Qgnutls_bootprop_verify_error, Qgnutls_bootprop_verify_hostname_error)
	(Qgnutls_bootprop_callbacks_verify): Make static.

1139 1140
2011-05-01  Andreas Schwab  <schwab@linux-m68k.org>

1141 1142
	* callproc.c: Indentation fixup.

1143
	* sysdep.c (wait_for_termination_1): Make static.
1144 1145
	(wait_for_termination, interruptible_wait_for_termination):
	Move after wait_for_termination_1.
1146

1147 1148 1149 1150 1151 1152 1153 1154 1155 1156
2011-05-01  Lars Magne Ingebrigtsen  <larsi@gnus.org>

	* sysdep.c (interruptible_wait_for_termination): New function
	which is like wait_for_termination, but allows keyboard
	interruptions.

	* callproc.c (Fcall_process): Add (:file "file") as an option for
	the STDOUT buffer.
	(Fcall_process_region): Ditto.

1157 1158
2011-04-30  Eli Zaretskii  <eliz@gnu.org>

1159 1160 1161
	* dosfns.c (Fint86, Fdos_memget, Fdos_memput): Use `ASIZE (FOO)'
	rather than `XVECTOR (FOO)->size'.

1162 1163 1164 1165
	* process.c: Remove HAVE_INTTYPES_H condition from inclusion of
	inttypes.h, as a gnulib replacement is used if it not available in
	system headers.

1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180
2011-04-21  Eli Zaretskii  <eliz@gnu.org>

	Lift the MOST_POSITIVE_FIXNUM/4 limitation on visited files.
	* fileio.c (Finsert_file_contents): Don't limit file size to 1/4
	of MOST_POSITIVE_FIXNUM.  (Bug#8528)

	* coding.c (coding_alloc_by_realloc): Error out if destination
	will grow beyond MOST_POSITIVE_FIXNUM.
	(decode_coding_emacs_mule): Abort if there isn't enough place in
	charbuf for the composition carryover bytes.  Reserve an extra
	space for up to 2 characters produced in a loop.
	(decode_coding_iso_2022): Abort if there isn't enough place in
	charbuf for the composition carryover bytes.

2011-04-21  Eli Zaretskii  <eliz@gnu.org>
1181

1182 1183 1184 1185 1186
	* doprnt.c (doprnt) [!HAVE_LONG_LONG_INT]: Error out instead of
	aborting when %lld or %lll format is passed.
	[!HAVE_UNSIGNED_LONG_LONG_INT]: Error out instead of aborting when
	%llo or %llx format is passed.  (Bug#8545)

1187 1188 1189
	* window.c (window_scroll_line_based): Use a marker instead of
	simple variables to record original value of point.  (Bug#7952)

1190 1191 1192
	* doprnt.c (doprnt): Fix the case where a multibyte sequence
	produced by %s or %c overflows available buffer space.  (Bug#8545)

1193 1194 1195
2011-04-28  Paul Eggert  <eggert@cs.ucla.edu>

	* doprnt.c (doprnt): Omit useless test; int overflow check (Bug#8545).
1196
	(SIZE_MAX): Move defn after all includes, as they might #define it.
1197

1198 1199 1200 1201
2011-04-28  Juanma Barranquero  <lekktu@gmail.com>

	* w32.c (init_environment): Warn about defaulting HOME to C:\.

1202 1203 1204 1205 1206 1207 1208 1209
2011-04-28  Juanma Barranquero  <lekktu@gmail.com>

	* keyboard.c (Qdelayed_warnings_hook): Define.
	(command_loop_1): Run `delayed-warnings-hook'
	if Vdelayed_warnings_list is non-nil.
	(syms_of_keyboard) <delayed-warnings-hook>: DEFSYM it.
	(syms_of_keyboard) <delayed-warnings-list>: DEFVAR_LISP it.

1210 1211 1212 1213 1214
2011-04-28  Eli Zaretskii  <eliz@gnu.org>

	* doprnt.c (doprnt): Don't return value smaller than the buffer
	size if the message was truncated.  (Bug#8545).

1215 1216 1217 1218 1219
2011-04-28  Juanma Barranquero  <lekktu@gmail.com>

	* w32fns.c (Fx_change_window_property, Fx_delete_window_property)
	(Fx_window_property): #if-0 the whole functions, not just the bodies.

1220 1221 1222 1223
2011-04-27  Paul Eggert  <eggert@cs.ucla.edu>

	* doprnt.c (doprnt): Support "ll" length modifier, for long long.

1224 1225 1226 1227
2011-04-27  Juanma Barranquero  <lekktu@gmail.com>

	* makefile.w32-in: Update dependencies.

1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
2011-04-27  Eli Zaretskii  <eliz@gnu.org>

	Improve `doprnt' and its usage.  (Bug#8545)
	* doprnt.c (doprnt): Make sure `format' is never accessed beyond
	`format_end'.  Remove support for %l as a conversion specifier.
	Don't use xrealloc.  Improve diagnostics when the %l size modifier
	is used.  Update the commentary.

	* eval.c (verror): Simplify calculation of size_t.

	* coding.c (Ffind_operation_coding_system): Fix diagnostic error
	messages.

1241 1242 1243 1244 1245
2011-04-27  Yoshiaki Kasahara  <kasahara@nc.kyushu-u.ac.jp>  (tiny change)

	* buffer.c (init_buffer) [USE_MMAP_FOR_BUFFERS]: Adjust to aliasing
	change.

1246 1247 1248 1249 1250
2011-04-27  Paul Eggert  <eggert@cs.ucla.edu>

	* nsmenu.m: Replace all uses of XVECTOR with ASIZE and AREF.
	This makes this file independent of the recent pseudovector change.

Paul Eggert's avatar
Paul Eggert committed
1251
2011-04-26  Paul Eggert  <eggert@cs.ucla.edu>
1252

1253 1254
	* keyboard.c (handle_user_signal): Fix pointer signedness problem.

Paul Eggert's avatar
Paul Eggert committed
1255
	* gnutls.c (emacs_gnutls_handle_error): Remove unused local.
1256
	(Fgnutls_boot): gnutls_certificate_verify_peers2 wants unsigned *.
Paul Eggert's avatar
Paul Eggert committed
1257
	Remove unused local.
1258
	(emacs_gnutls_write): Don't use uninitialized rtnval if nbyte <= 0.
1259

Juanma Barranquero's avatar
Juanma Barranquero committed
1260
	* lisp.h: Fix a problem with aliasing and vector headers.  (Bug#8546)
1261 1262
	GCC 4.6.0 optimizes based on type-based alias analysis.
	For example, if b is of type struct buffer * and v of type struct
1263 1264 1265 1266 1267
	Lisp_Vector *, then gcc -O2 was incorrectly assuming that &b->size
	!= &v->size, and therefore "v->size = 1; b->size = 2; return
	v->size;" must therefore return 1.  This assumption is incorrect
	for Emacs, since it type-puns struct Lisp_Vector * with many other
	types.  To fix this problem, this patch adds a new type struct
Paul Eggert's avatar
Paul Eggert committed
1268
	vectorlike_header that documents the constraints on layout of vectors
1269 1270 1271
	and pseudovectors, and helps optimizing compilers not get fooled
	by Emacs's type punning.  It also adds the macros XSETTYPED_PVECTYPE
	XSETTYPED_PSEUDOVECTOR, TYPED_PSEUDOVECTORP, for similar reasons.
Paul Eggert's avatar
Paul Eggert committed
1272 1273
	* lisp.h (XSETTYPED_PVECTYPE): New macro, specifying the name of
	the size member.
1274 1275 1276 1277 1278
	(XSETPVECTYPE): Rewrite in terms of new macro.
	(XSETPVECTYPESIZE): New macro, specifying both type and size.
	This is a bit clearer, and further avoids the possibility of
	undesirable aliasing.
	(XSETTYPED_PSEUDOVECTOR): New macro, specifying the size.
Paul Eggert's avatar
Paul Eggert committed
1279
	(XSETPSEUDOVECTOR): Rewrite in terms of XSETTYPED_PSEUDOVECTOR.
1280 1281
	(XSETSUBR): Rewrite in terms of XSETTYPED_PSEUDOVECTOR and XSIZE,
	since Lisp_Subr is a special case (no "next" field).
1282 1283
	(ASIZE): Now uses header.size rather than size.
	All previous uses of XVECTOR (foo)->size replaced to use this macro,
Paul Eggert's avatar
Paul Eggert committed
1284 1285
	to avoid the hassle of writing XVECTOR (foo)->header.size.
	(struct vectorlike_header): New type.
1286 1287 1288 1289 1290 1291 1292
	(TYPED_PSEUDOVECTORP): New macro, also specifying the C type of the
	object, to help avoid aliasing.
	(PSEUDOVECTORP): Rewrite in terms of TYPED_PSEUDOVECTORP.
	(SUBRP): Likewise, since Lisp_Subr is a special case.
	* lisp.h (struct Lisp_Vector, struct Lisp_Char_Table):
	(struct Lisp_Sub_Char_Table, struct Lisp_Bool_Vector):
	(struct Lisp_Hash_Table): Combine first two members into a single
Paul Eggert's avatar
Paul Eggert committed
1293
	struct vectorlike_header member.  All uses of "size" and "next" members
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
	changed to be "header.size" and "header.next".
	* buffer.h (struct buffer): Likewise.
	* font.h (struct font_spec, struct font_entity, struct font): Likewise.
	* frame.h (struct frame): Likewise.
	* process.h (struct Lisp_Process): Likewise.
	* termhooks.h (struct terminal): Likewise.
	* window.c (struct save_window_data, struct saved_window): Likewise.
	* window.h (struct window): Likewise.
	* alloc.c (allocate_buffer, Fmake_bool_vector, allocate_pseudovector):
	Use XSETPVECTYPESIZE, not XSETPVECTYPE, to avoid aliasing problems.
	* buffer.c (init_buffer_once): Likewise.
	* lread.c (defsubr): Use XSETTYPED_PVECTYPE, since Lisp_Subr is a
	special case.
	* process.c (Fformat_network_address): Use local var for size,
	for brevity.

1310 1311
	* bytecode.c (exec_byte_code): Don't use XVECTOR before CHECK_VECTOR.

Paul Eggert's avatar
Paul Eggert committed
1312
	Make the Lisp reader and string-to-float more consistent (Bug#8525)
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322
	* data.c (atof): Remove decl; no longer used or needed.
	(digit_to_number): Move to lread.c.
	(Fstring_to_number): Use new string_to_number function, to be
	consistent with how the Lisp reader treats infinities and NaNs.
	Do not assume that floating-point numbers represent EMACS_INT
	without losing information; this is not true on most 64-bit hosts.
	Avoid double-rounding errors, by insisting on integers when
	parsing non-base-10 numbers, as the documentation specifies.
	* lisp.h (string_to_number): New decl, replacing ...
	(isfloat_string): Remove.
1323
	* lread.c: Include <inttypes.h>, for uintmax_t and strtoumax.
1324
	(read1): Do not accept +. and -. as integers; this
1325 1326 1327 1328 1329 1330 1331
	appears to have been a coding error.  Similarly, do not accept
	strings like +-1e0 as floating point numbers.  Do not report
	overflow for integer overflows unless the base is not 10 which
	means we have no simple and reliable way to continue.
	Break out the floating-point parsing into a new
	function string_to_number, so that Fstring_to_number parses
	floating point numbers consistently with the Lisp reader.
1332
	(digit_to_number): Move here from data.c.  Make it static inline.
1333 1334 1335 1336 1337
	(E_CHAR, EXP_INT): Remove, replacing with ...
	(E_EXP): New macro, to solve the "1.0e+" problem mentioned below.
	(string_to_number): New function, replacing isfloat_string.
	This function checks for valid syntax and produces the resulting
	Lisp float number too.  Rework it so that string-to-number
1338
	no longer mishandles examples like "1.0e+".  Use strtoumax,
Paul Eggert's avatar