- 2011-06-20 Paul Eggert <eggert@cs.ucla.edu>
++2011-06-21 Paul Eggert <eggert@cs.ucla.edu>
+
+ * termcap.c: Don't assume sizes fit in int and never overflow.
+ (struct termcap_buffer, tgetent): Use ptrdiff_t, not int, for sizes.
+ (gobble_line): Check for size-calculation overflow.
+
+ * minibuf.c (Fread_buffer):
+ * lread.c (intern, intern_c_string):
+ * image.c (xpm_scan) [HAVE_NS && !HAVE_XPM]:
+ Don't assume string length fits in int.
+
+ * keyboard.c (parse_tool_bar_item):
+ * gtkutil.c (style_changed_cb): Avoid need for strlen.
+
+ * font.c: Don't assume string length fits in int.
+ (font_parse_xlfd, font_parse_fcname, font_unparse_fcname):
+ Use ptrdiff_t, not int.
+ (font_intern_prop): Don't assume string length fits in int.
+ Don't assume integer property fits in fixnum.
+ * font.h (font_intern_prop): 2nd arg is now ptrdiff_t, not int.
+
+ * filelock.c: Fix some buffer overrun and integer overflow issues.
- (get_boot_time): Don't assume that gzip command string fits in 100 bytes.
++ (get_boot_time): Don't assume gzip command string fits in 100 bytes.
+ Reformulate so as not to need the command string.
+ Invoke gzip -cd rather than gunzip, as it's more portable.
+ (lock_info_type, lock_file_1, lock_file):
+ Don't assume pid_t and time_t fit in unsigned long.
+ (LOCK_PID_MAX): Remove; we now use more-reliable bounds.
+ (current_lock_owner): Prefer signed type for sizes.
+ Use memcpy, not strncpy, where memcpy is what is really wanted.
+ Don't assume (via atoi) that time_t and pid_t fit in int.
+ Check for time_t and/or pid_t out of range, e.g., via a network share.
+ Don't alloca where an auto var works fine.
+
- 2011-06-19 Paul Eggert <eggert@cs.ucla.edu>
-
+ * fileio.c: Fix some integer overflow issues.
+ (file_name_as_directory, Fexpand_file_name, Fsubstitute_in_file_name):
+ Don't assume string length fits in int.
+ (directory_file_name): Don't assume string length fits in long.
+ (make_temp_name): Don't assume pid fits in int, or that its print
+ length is less than 20.
+
+ * data.c (Fsubr_name): Rewrite to avoid a strlen call.
+
+ * coding.c (make_subsidiaries): Don't assume string length fits in int.
+
+ * callproc.c (child_setup): Rewrite to avoid two strlen calls.
+
+ * process.c (Fformat_network_address): Use EMACS_INT, not EMACS_UINT.
+ We prefer signed integers, even for size calculations.
+
+ * emacs.c: Don't assume string length fits in 'int'.
+ (DEFINE_DUMMY_FUNCTION, sort_args): Use ptrdiff_t, not int.
+ (main): Don't invoke strlen when not needed.
+
+ * dbusbind.c (XD_ERROR): Don't arbitrarily truncate string.
+ (XD_DEBUG_MESSAGE): Don't waste a byte.
+
+ * callproc.c (getenv_internal_1, getenv_internal)
+ (Fgetenv_internal):
+ * buffer.c (init_buffer): Don't assume string length fits in 'int'.
+
+ * lread.c (invalid_syntax): Omit length argument.
+ All uses changed. This doesn't fix a bug, but it simplifies the
+ code away from its former Hollerith-constant appearance, and it's
+ one less 'int' to worry about when looking at integer-overflow issues.
- (string_to_number): Simplify the 2011-04-26 change by invoking xsignal1.
++ (string_to_number): Simplify 2011-04-26 change by invoking xsignal1.
+
+ * lisp.h (DEFUN): Remove bogus use of sizeof (struct Lisp_Subr).
+ This didn't break anything, but it didn't help either.
+ It's confusing to put a bogus integer in a place where the actual
+ value does not matter.
+ (LIST_END_P): Remove unused macro and its bogus comment.
+ (make_fixnum_or_float): Remove unnecessary cast to EMACS_INT.
+
- 2011-06-18 Paul Eggert <eggert@cs.ucla.edu>
-
+ * lisp.h (union Lisp_Object.i): EMACS_INT, not EMACS_UINT.
+ This is for consistency with the ordinary, non-USE_LISP_UNION_TYPE,
+ implementation.
+ (struct Lisp_Bool_Vector.size): EMACS_INT, not EMACS_UINT.
+ We prefer signed types, and the value cannot exceed the EMACS_INT
+ range anyway (because otherwise the length would not be representable).
+ (XSET) [USE_LISP_UNION_TYPE]: Use uintptr_t and intptr_t,
+ not EMACS_UINT and EMACS_INT, when converting pointer to integer.
+ This avoids a GCC warning when WIDE_EMACS_INT.
+
+ * indent.c (sane_tab_width): New function.
+ (current_column, scan_for_column, Findent_to, position_indentation)
+ (compute_motion): Use it. This is just for clarity.
+ (Fcompute_motion): Don't assume hscroll and tab offset fit in int.
+
- * image.c (xbm_image_p): Don't assume stated width and height fit in int.
++ * image.c (xbm_image_p): Don't assume stated width, height fit in int.
+
+ * lisp.h (lint_assume): New macro.
+ * composite.c (composition_gstring_put_cache):
+ * ftfont.c (ftfont_shape_by_flt): Use it to pacify GCC 4.6.0.
+
+ * editfns.c, insdel.c:
+ Omit unnecessary forward decls, to simplify future changes.
+
+ * ftfont.c (ftfont_shape_by_flt): Use signed integers for lengths.
+
+ * font.c (Ffont_shape_gstring): Don't assume glyph len fits in 'int'.
+
+ * fns.c (Ffillarray): Don't assume bool vector size fits in 'int'.
+ Use much-faster test for byte-length change.
+ Don't assume string byte-length fits in 'int'.
+ Check that character arg fits in 'int'.
+ (mapcar1): Declare byte as byte, for clarity.
+
+ * alloc.c (Fmake_bool_vector): Avoid unnecessary multiplication.
+
+ * fns.c (concat): Catch string overflow earlier.
+ Do not rely on integer wraparound.
+
- * dispextern.h (struct it.overlay_strings_charpos): EMACS_INT, not int.
-
- 2011-06-17 Paul Eggert <eggert@cs.ucla.edu>
-
- * dispextern.h (struct it.selective): Now EMACS_INT, not int.
++ * dispextern.h (struct it.overlay_strings_charpos)
++ (struct it.selective): Now EMACS_INT, not int.
+ * xdisp.c (forward_to_next_line_start)
+ (back_to_previous_visible_line_start)
+ (reseat_at_next_visible_line_start, next_element_from_buffer):
+ Don't arbitrarily truncate the value of 'selective' to int.
+
+ * xdisp.c (init_iterator): Use XINT, not XFASTINT; it might be < 0.
+
+ * composite.c: Don't truncate sizes to 'int'.
+ (composition_gstring_p, composition_reseat_it)
+ (composition_adjust_point): Use EMACS_INT, not int.
+ (get_composition_id, composition_gstring_put_cache): Use EMACS_INT,
+ not EMACS_UINT, for indexes.
+
+ * category.h (CATEGORY_SET_P): Remove unnecessary cast to EMACS_INT.
+
+ * buffer.c: Include <verify.h>.
+ (struct sortvec.priority, struct sortstr.priority):
+ Now EMACS_INT, not int.
+ (compare_overlays, cmp_for_strings): Avoid subtraction overflow.
+ (struct sortstr.size, record_overlay_string)
+ (struct sortstrlist.size, struct sortlist.used):
+ Don't truncate size to int.
+ (record_overlay_string): Check for size-calculation overflow.
+ (init_buffer_once): Check at compile-time, not run-time.
+
+ 2011-06-20 Deniz Dogan <deniz@dogan.se>
+
+ * process.c (Fset_process_buffer): Clarify return value in
+ docstring.
+
+ 2011-06-18 Chong Yidong <cyd@stupidchicken.com>
+
+ * dispnew.c (add_window_display_history): Use BVAR.
+
+ * xdisp.c (debug_method_add): Use BVAR.
+ (check_window_end, dump_glyph_matrix, dump_glyph)
+ (dump_glyph_row, dump_glyph_string): Convert arglist to ANSI C.
+
+ * xfaces.c (check_lface_attrs, check_lface, dump_realized_face):
+ Likewise.
+
+ * xfns.c (Fx_create_frame, x_create_tip_frame): Delay image cache
+ check till after the cache is created in init_frame_faces.
+
+ 2011-06-17 Stefan Monnier <monnier@iro.umontreal.ca>
+
+ * fns.c (Fsafe_length): Yet another int/Lisp_Object mixup.
+
2011-06-16 Paul Eggert <eggert@cs.ucla.edu>
* lisp.h: Include <limits.h>, for INT_MAX, LONG_MAX, LLONG_MAX.