+2012-09-23 Paul Eggert <eggert@cs.ucla.edu>
+
+ Simplify and avoid signal-handling races (Bug#12471).
+ * alloc.c (die):
+ * sysdep.c (emacs_abort) [HAVE_NTGUI]:
+ Avoid recursive loop if there's a fatal error in the function itself.
+ * atimer.c (pending_atimers):
+ * blockinput.h: Don't include "atimer.h"; no longer needed.
+ (interrupt_input_pending): Remove. All uses removed.
+ pending_signals now counts both atimers and ordinary interrupts.
+ This is less racy than having three separate pending-signal flags.
+ (block_input, unblock_input, totally_unblock_input, unblock_input_to)
+ (input_blocked_p):
+ Rename from their upper-case counterparts BLOCK_INPUT,
+ UNBLOCK_INPUT, TOTALLY_UNBLOCK_INPUT, UNBLOCK_INPUT_TO,
+ INPUT_BLOCKED_P, and turn into functions. All uses changed.
+ This makes it easier to access volatile variables more accurately.
+ (BLOCK_INPUT_RESIGNAL): Remove. All uses replaced by unblock_input ().
+ (input_blocked_p): Prefer this to 'interrupt_input_blocked', as
+ that's more reliable if the code is buggy and sets
+ interrupt_input_blocked to a negative value. All uses changed.
+ * atimer.c (deliver_alarm_signal):
+ Remove. No need to deliver this to the parent; any thread can
+ handle this signal now. All uses replaced by underlying handler.
+ * atimer.c (turn_on_atimers):
+ * dispnew.c (handle_window_change_signal):
+ * emacs.c (handle_danger_signal):
+ * keyboard.c (kbd_buffer_get_event):
+ Don't reestablish signal handler; not needed with sigaction.
+ * blockinput.h (UNBLOCK_INPUT_TO, TOTALLY_UNBLOCK_INPUT)
+ (UNBLOCK_INPUT_TO):
+ Rework to avoid unnecessary accesses to volatile variables.
+ (UNBLOCK_INPUT_TO): Now a function.
+ (totally_unblock_input, unblock_input): New decls.
+ * data.c (handle_arith_signal, deliver_arith_signal): Move to sysdep.c
+ (init_data): Remove. Necessary stuff now done in init_signal.
+ * emacs.c, xdisp.c: Include "atimer.h", since we invoke atimer functions.
+ * emacs.c (handle_fatal_signal, deliver_fatal_signal): Move to sysdep.c.
+ (fatal_error_code): Remove; no longer needed.
+ (terminate_due_to_signal): Rename from fatal_error_backtrace, since
+ it doesn't always backtrace. All uses changed. No need to reset
+ signal to default, since sigaction and/or die does that for us now.
+ Use emacs_raise (FOO), not kill (getpid (), FOO).
+ (main): Check more-accurately whether we're dumping.
+ Move fatal-error setup to sysdep.c
+ * floatfns.c: Do not include "syssignal.h"; no longer needed.
+ * gtkutil.c (xg_get_file_name, xg_get_font):
+ Remove no-longer-needed signal-mask manipulation.
+ * keyboard.c, process.c (POLL_FOR_INPUT):
+ Don't depend on USE_ASYNC_EVENTS, a symbol that is never defined.
+ * keyboard.c (read_avail_input): Remove.
+ All uses replaced by gobble_input.
+ (Ftop_level): Use TOTALLY_UNBLOCK_INPUT rather than open code.
+ (kbd_buffer_store_event_hold, gobble_input):
+ (record_asynch_buffer_change) [USABLE_SIGIO]:
+ (store_user_signal_events):
+ No need to mess with signal mask.
+ (gobble_input): If blocking input and there are terminals, simply
+ set pending_signals to 1 and return. All hooks changed to not
+ worry about whether input is blocked.
+ (process_pending_signals): Clear pending_signals before processing
+ them, in case a signal comes in while we're processing.
+ By convention callers now test pending_signals before calling us.
+ (UNBLOCK_INPUT_TO, unblock_input, totally_unblock_input):
+ New functions, to support changes to blockinput.h.
+ (handle_input_available_signal): Now extern.
+ (reinvoke_input_signal): Remove. All uses replaced by
+ handle_async_input.
+ (quit_count): Now volatile, since a signal handler uses it.
+ (handle_interrupt): Now takes bool IN_SIGNAL_HANDLER as arg. All
+ callers changed. Block SIGINT only if not already blocked.
+ Clear sigmask reliably, even if Fsignal returns, which it can.
+ Omit unnecessary accesses to volatile var.
+ (quit_throw_to_read_char): No need to restore sigmask.
+ * keyboard.c (gobble_input, handle_user_signal):
+ * process.c (wait_reading_process_output):
+ Call signal-handling code rather than killing ourselves.
+ * lisp.h: Include <float.h>, for...
+ (IEEE_FLOATING_POINT): New macro, moved here to avoid duplication.
+ (pending_signals): Now volatile.
+ (syms_of_data): Now const if IEEE floating point.
+ (handle_input_available_signal) [USABLE_SIGIO]:
+ (terminate_due_to_signal, record_child_status_change): New decls.
+ * process.c (create_process): Avoid disaster if memory is exhausted
+ while we're processing a vfork, by tightening the critical section
+ around the vfork.
+ (send_process_frame, process_sent_to, handle_pipe_signal)
+ (deliver_pipe_signal): Remove. No longer needed, as Emacs now
+ ignores SIGPIPE.
+ (send_process): No need for setjmp/longjmp any more, since the
+ SIGPIPE stuff is now gone. Instead, report an error if errno
+ is EPIPE.
+ (record_child_status_change): Now extern. PID and W are now args.
+ Return void, not bool. All callers changed.
+ * sysdep.c (wait_debugging) [(BSD_SYSTEM || HPUX) && !defined (__GNU__)]:
+ Remove. All uses removed. This bug should be fixed now in a
+ different way.
+ (wait_for_termination_1): Use waitpid rather than sigsuspend,
+ and record the child status change directly. This avoids the
+ need to futz with the signal mask.
+ (process_fatal_action): Move here from emacs.c.
+ (emacs_sigaction_flags): New function, containing
+ much of what used to be in emacs_sigaction_init.
+ (emacs_sigaction_init): Use it. Block nonfatal system signals that are
+ caught by emacs, to make races less likely.
+ (deliver_process_signal): Rename from handle_on_main_thread.
+ All uses changed.
+ (BACKTRACE_LIMIT_MAX): Now at top level.
+ (thread_backtrace_buffer, threadback_backtrace_pointers):
+ New static vars.
+ (deliver_thread_signal, deliver_fatal_thread_signal):
+ New functions, for more-accurate delivery of thread-specific signals.
+ (handle_fatal_signal, deliver_fatal_signal): Move here from emacs.c.
+ (deliver_arith_signal): Handle in this thread, not
+ in the main thread, since it's triggered by this thread.
+ (maybe_fatal_sig): New function.
+ (init_signals): New arg DUMPING so that we can be more accurate
+ about whether we're dumping. Caller changed.
+ Treat thread-specific signals differently from process-general signals.
+ Block all signals while handling fatal error; that's safer.
+ xsignal from SIGFPE only on non-IEEE hosts, treating it as fatal
+ on IEEE hosts.
+ When batch, ignore SIGHUP, SIGINT, SIGTERM if they were already ignored.
+ Ignore SIGPIPE unless batch.
+ (emacs_backtrace): Output backtrace for the appropriate thread,
+ which is not necessarily the main thread.
+ * syssignal.h: Include <stdbool.h>.
+ (emacs_raise): New macro.
+ * xterm.c (x_connection_signal): Remove; no longer needed
+ now that we use sigaction.
+ (x_connection_closed): No need to mess with sigmask now.
+ (x_initialize): No need to reset SIGPIPE handler here, since
+ init_signals does this for us now.
+
+2012-09-23 Jan Djärv <jan.h.d@swipnet.se>
+
+ * nsterm.m (ns_dumpglyphs_image): dr is a new rect to draw image into,
+ background rect may be larger (Bug#12445).
+
+2012-09-23 Chong Yidong <cyd@gnu.org>
+
+ * keyboard.c (timer_check): Avoid quitting during Fcopy_sequence.
+
+2012-09-22 Paul Eggert <eggert@cs.ucla.edu>
+
+ * .gdbinit: Just stop at fatal_error_backtrace.
+ See Stefan Monnier's request in
+ <http://lists.gnu.org/archive/html/emacs-devel/2012-09/msg00549.html>.
+ Remove no-longer-used query of system type.
+
+2012-09-22 Chong Yidong <cyd@gnu.org>
+
+ * search.c (Freplace_match): Doc fix (Bug#12325).
+
+ * minibuf.c (Finternal_complete_buffer): Doc fix (Bug#12391).
+
+ * editfns.c (Fline_beginning_position): Doc fix (Bug#12416).
+ (Fline_end_position): Doc fix.
+
+ * cmds.c (Fforward_char, Fbackward_char): Doc fix (Bug#12414).
+
+2012-09-22 Chong Yidong <cyd@gnu.org>
+
+ * dispextern.h (struct image_type): Add new slot, storing a type
+ initialization function.
+
+ * image.c (define_image_type): Call the image initializer function
+ if it is defined. Arguments and return value changed.
+ (valid_image_p, make_image): Callers changed.
+ (xbm_type, xpm_type, pbm_type, png_type, jpeg_type, tiff_type)
+ (gif_type, imagemagick_type, svg_type, gs_type): Add
+ initialization functions.
+ (Finit_image_library): Call lookup_image_type.
+ (CHECK_LIB_AVAILABLE): Macro deleted.
+ (lookup_image_type): Call define_image_type here, rather than via
+ Finit_image_library, and without using CHECK_LIB_AVAILABLE.
+ (syms_of_image): Move define_image_type calls for xbm_type and
+ pbm_type to lookup_image_type.
+
+2012-09-22 Eli Zaretskii <eliz@gnu.org>
+
+ * keyboard.c (timer_check_2): Move calculation of 'timers' and
+ 'idle_timers' from here ...
+ (timer_check): ... to here. Use Fcopy_sequence to copy the timer
+ lists, to avoid infloops when the timer does something stupid,
+ like reinvoke itself with the same or smaller time-out.
+ (Bug#12447)
+
+2012-09-22 Martin Rudalics <rudalics@gmx.at>
+
+ * window.c (Fsplit_window_internal): Handle only Qt value of
+ Vwindow_combination_limit separately.
+ (Qtemp_buffer_resize): New symbol.
+ (Vwindow_combination_limit): New default value. Rewrite
+ doc-string.
+
+2012-09-22 Eli Zaretskii <eliz@gnu.org>
+
+ * xdisp.c (next_overlay_string): Initialize it->end_charpos for
+ the new overlay string. (Bug#10159)
+
+2012-09-22 Paul Eggert <eggert@cs.ucla.edu>
+
+ * emacs.c (shut_down_emacs): Don't assume stderr is buffered,
+ or that fprintf is async-signal-safe. POSIX doesn't require
+ either assumption.
+
+2012-09-22 Chong Yidong <cyd@gnu.org>
+
+ * buffer.c (Fset_buffer_modified_p): Handle indirect buffers
+ (Bug#8207).
+
+2012-09-22 Kenichi Handa <handa@gnu.org>
+
+ * composite.c (composition_reseat_it): Handle the case that a
+ grapheme cluster is not covered by a single font (Bug#12352).
+
+2012-09-21 Chong Yidong <cyd@gnu.org>
+
+ * image.c (define_image_type): Avoid adding duplicate types to
+ image_types (Bug#12463). Suggested by Jörg Walter.
+
+2012-09-21 YAMAMOTO Mitsuharu <mituharu@math.s.chiba-u.ac.jp>
+
+ * unexmacosx.c: Define LC_DATA_IN_CODE if not defined.
+ (print_load_command_name): Add case LC_DATA_IN_CODE.
+ (dump_it) [LC_DATA_IN_CODE]: Call copy_linkedit_data.
+
+2012-09-21 Glenn Morris <rgm@gnu.org>
+
+ * eval.c (Frun_hook_with_args_until_success)
+ (Frun_hook_with_args_until_failure): Doc fixes. (Bug#12393)
+
+2012-09-21 Andreas Schwab <schwab@linux-m68k.org>
+
+ * fileio.c (Ffile_selinux_context): Only call freecon when
+ lgetfilecon succeeded.
+ (Fset_file_selinux_context): Likewise. (Bug#12444)
+
+2012-09-21 Eli Zaretskii <eliz@gnu.org>
+
+ * xdisp.c (try_window_reusing_current_matrix): Under bidi
+ reordering, locate the cursor by calling set_cursor_from_row; if
+ that fails, clear the desired glyph matrix before returning a
+ failure indication to the caller. Fixes leaving garbled display
+ when fast scrolling with a down-key. (Bug#12403)
+ (compute_stop_pos_backwards): Fix a typo that caused crashes while
+ scrolling through multibyte text.
+
+2012-09-20 Stefan Monnier <monnier@iro.umontreal.ca>
+
+ * alloc.c (mark_object) <PVEC_WINDOW>: Mark prev/next_buffers *after*
+ calling mark_vectorlike since that's the one that marks the window.
+ (mark_discard_killed_buffers): Mark the final cdr.
+ * window.h (struct window): Move prev/next_buffers to the
+ non-standard fields.
+ * window.c (make_window): Initialize prev/next_buffers manually.
+
+2012-09-20 Paul Eggert <eggert@cs.ucla.edu>
+
+ Omit unused arg EXPECTED from socket hooks.
+ * keyboard.c (gobble_input, read_avail_input, tty_read_avail_input):
+ * nsterm.m (ns_term_init):
+ * termhooks.h (struct terminal.read_socket_hook):
+ * w32inevt.c (w32_console_read_socket):
+ * w32term.c (w32_read_socket):
+ * xterm.c (XTread_socket):
+ Omit unused arg EXPECTED. All callers changed.
+ (store_user_signal_events): Return void, not int, since callers no
+ longer care about the return value. All uses changed.
+
+2012-09-20 Juanma Barranquero <lekktu@gmail.com>
+
+ * w32gui.h (XParseGeometry): Do not declare.
+
+2012-09-19 Paul Eggert <eggert@cs.ucla.edu>
+
+ * w32inevt.c (w32_console_read_socket): Return -1 on failure, not 0.
+ Ignore 'expected'. See Eli Zaretskii in
+ <http://bugs.gnu.org/12471#8> (last line).
+
+ * frame.c (read_integer): Remove. All uses replaced by strtol/strtoul.
+ (XParseGeometry): Now static. Substitute extremal values for
+ values that are out of range.
+
+2012-09-19 Jan Djärv <jan.h.d@swipnet.se>
+
+ * w32xfns.c (read_integer, XParseGeometry): Move to frame.c.
+
+ * nsfns.m (XParseGeometry): Remove.
+ (Fx_create_frame): Call x_set_offset to correctly interpret
+ top_pos in geometry.
+
+ * frame.c (read_integer, XParseGeometry): Moved from w32xfns.c.
+ (Fx_parse_geometry): If there is a space in string, call
+ Qns_parse_geometry, otherwise do as on other terms (Bug#12368).
+
+2012-09-17 Eli Zaretskii <eliz@gnu.org>
+
+ * search.c (scan_buffer): Use character positions in calls to
+ region_cache_forward and region_cache_backward, not byte
+ positions. (Bug#12196)
+
+ * w32term.c (w32_read_socket): Set pending_signals to 1, like
+ xterm.c does. Reported by Daniel Colascione <dancol@dancol.org>.
+
+ * ralloc.c (r_alloc_init) [!SYSTEM_MALLOC]: Initialize
+ __malloc_extra_blocks to 32 instead of 64, like alloc.c did in
+ emacs_blocked_malloc, now deleted.
+
+2012-09-17 Paul Eggert <eggert@cs.ucla.edu>
+
+ Remove no-longer-needed Solaris 2.4 vfork bug workaround.
+ The workaround was for improving performance on Solaris 2.4, but
+ is getting in the way now. Emacs will still work if someone is
+ still running Solaris 2.4 in a museum somewhere; Sun dropped
+ support for Solaris 2.4 in 2003.
+ * callproc.c (Fcall_process) [HAVE_WORKING_VFORK]:
+ * process.c (create_process) [HAVE_WORKING_VFORK]:
+ Omit now-unnecessary workaround for the Solaris 2.4 vfork bug,
+ since Emacs no longer uses vfork on that platform.
+
+2012-09-17 Glenn Morris <rgm@gnu.org>
+
+ * emacs.c: Use COPYRIGHT.
+
+2012-09-16 Paul Eggert <eggert@cs.ucla.edu>
+
+ Remove configure's --without-sync-input option (Bug#12450).
+ When auditing signal-handling in preparation for cleaning it up,
+ I found that SYNC_INPUT has race conditions and would be a real
+ pain to fix. Since it's an undocumented and deprecated
+ configure-time option, now seems like a good time to remove it.
+ Also see <http://bugs.gnu.org/11080#16>.
+ * alloc.c (_bytes_used, __malloc_extra_blocks, _malloc_internal)
+ (_free_internal) [!DOUG_LEA_MALLOC]: Remove decls.
+ (alloc_mutex) [!SYSTEM_MALLOC && !SYNC_INPUT && HAVE_PTHREAD]:
+ (malloc_hysteresis):
+ (check_depth) [XMALLOC_OVERRUN_CHECK]:
+ (MALLOC_BLOCK_INPUT, MALLOC_UNBLOCK_INPUT):
+ (__malloc_hook, __realloc_hook, __free_hook, BYTES_USED)
+ (dont_register_blocks, bytes_used_when_reconsidered)
+ (bytes_used_when_full, emacs_blocked_free, emacs_blocked_malloc)
+ (emacs_blocked_realloc, reset_malloc_hooks, uninterrupt_malloc):
+ [!SYSTEM_MALLOC && !SYNC_INPUT]:
+ Remove. All uses removed.
+ (MALLOC_BLOCK_INPUT, MALLOC_UNBLOCK_INPUT): Use a different
+ implementation, one that depends on whether the new macro
+ XMALLOC_BLOCK_INPUT_CHECK is defined, not on whether SYNC_INPUT
+ is defined.
+ * atimer.c (run_timers, handle_alarm_signal):
+ * keyboard.c (pending_signal, poll_for_input_1, poll_for_input)
+ (handle_async_input, process_pending_signals)
+ (handle_input_available_signal, init_keyboard):
+ * nsterm.m (ns_read_socket):
+ * process.c (wait_reading_process_output):
+ * regex.c (immediate_quit, IMMEDIATE_QUIT_CHECK):
+ * sysdep.c (emacs_sigaction_init) [SA_RESTART]:
+ (emacs_write):
+ * xterm.c (XTread_socket):
+ Assume SYNC_INPUT.
+ * conf_post.h (SA_RESTART) [IRIX6_5]: Do not #undef.
+ * eval.c (handling_signal): Remove. All uses removed.
+ * lisp.h (ELSE_PENDING_SIGNALS): Remove.
+ All uses replaced with the SYNC_INPUT version.
+ (reset_malloc_hooks, uninterrupt_malloc, handling_signal):
+ Remove decls.
+ * sysdep.c, syssignal.h (main_thread) [FORWARD_SIGNAL_TO_MAIN_THREAD]:
+ Now static.
+
+ * font.c (Ffont_shape_gstring): Remove unused local.
+
+2012-09-16 Glenn Morris <rgm@gnu.org>
+
+ * Makefile.in (clean): No longer run nextstep's clean.
+
+ * Makefile.in (ns_appdir, ns_appbindir, ns_appsrc): Remove variables.
+ (ns_frag): Remove.
+ (ns-app): Move here from ns.mk, and simplify.
+ (clean): Simplify nextstep entry.
+ * ns.mk: Remove file.
+
+2012-09-17 Kenichi Handa <handa@gnu.org>
+
+ * font.c (Ffont_shape_gstring): Fix previous change; GLYPHs may
+ not covert the last few charactes.
+
+2012-09-16 Kenichi Handa <handa@gnu.org>
+
+ * font.c (Ffont_shape_gstring): Don't adjust grapheme cluster
+ here, but just check the validity of glyphs in the glyph-string.
+
+2012-09-16 Martin Rudalics <rudalics@gmx.at>
+
+ * window.c (Fwindow_parameter, Fset_window_parameter): Accept
+ any window as argument (Bug#12452).
+
+2012-09-16 Jan Djärv <jan.h.d@swipnet.se>
+
+ * nsfns.m (Fx_open_connection): Move initialization of ns_*_types
+ to ns_term_init to avoid memory leak.
+
+ * nsterm.m (ns_update_begin): Initialize bp after lcokFocus, use
+ explicit retain/release.
+ (ns_term_init): Only allow one display. Initialize outerpool and
+ ns_*_types.
+
+2012-09-15 Paul Eggert <eggert@cs.ucla.edu>
+
+ Port _setjmp fix to POSIXish hosts as well as Microsoft.
+ * image.c (_setjmp) [!HAVE__SETJMP]: Restore definition, as
+ it's needed on POSIXish hosts that lack _setjmp. Attempt to solve
+ the Microsoft problem in a different way, by altering ../nt/config.nt.
+
+2012-09-15 Eli Zaretskii <eliz@gnu.org>
+
+ * w32xfns.c:
+ * w32uniscribe.c:
+ * w32term.c:
+ * w32select.c:
+ * w32reg.c:
+ * w32proc.c:
+ * w32menu.c:
+ * w32inevt.c:
+ * w32heap.c:
+ * w32font.c:
+ * w32fns.c:
+ * w32console.c:
+ * w32.c:
+ * w16select.c: Remove inclusion of setjmp.h, as it is now included
+ by lisp.h. This completes removal of setjmp.h inclusion
+ erroneously announced in the previous commit. (Bug#12446)
+
+ * lisp.h [!HAVE__SETJMP, !HAVE_SIGSETJMP]: Make the commentary
+ more accurate.
+
+ * image.c (_setjmp) [!HAVE__SETJMP]: Define only if 'setjmp' is
+ not defined as a macro. The latter happens on MS-Windows.
+ (Bug#12446)
+
+2012-09-15 Paul Eggert <eggert@cs.ucla.edu>
+
+ Port better to POSIX hosts lacking _setjmp (Bug#12446).
+ * lisp.h: Include <setjmp.h> here, since we use its symbols here.
+ Some instances of '#include <setjmp.h>' removed, if the
+ only reason for the instance was because "lisp.h" was included.
+ (sys_jmp_buf, sys_setjmp, sys_longjmp): New symbols.
+ Unless otherwise specified, replace all uses of jmp_buf, _setjmp,
+ and _longjmp with the new symbols. Emacs already uses _setjmp if
+ available, so this change affects only POSIXish hosts that have
+ sigsetjmp but not _setjmp, such as some versions of Solaris and
+ Unixware. (Also, POSIX-2008 marks _setjmp as obsolescent.)
+ * image.c (_setjmp, _longjmp) [HAVE_PNG && !HAVE__SETJMP]: New macros.
+ (png_load_body) [HAVE_PNG]:
+ (PNG_LONGJMP) [HAVE_PNG && PNG_LIBPNG_VER < 10500]:
+ (PNG_JMPBUF) [HAVE_PNG && PNG_LIBPNG_VER >= 10500]:
+ Use _setjmp and _longjmp rather than sys_setjmp and sys_longjmp,
+ since PNG requires jmp_buf. This is the only exception to the
+ general rule that we now use sys_setjmp and sys_longjmp.
+ This exception is OK since this code does not change the signal
+ mask or longjmp out of a signal handler.
+
+2012-09-14 Paul Eggert <eggert@cs.ucla.edu>
+
+ * alloc.c [!SYSTEM_MALLOC && !SYNC_INPUT && HAVE_PTHREAD]:
+ Include "syssignal.h", for 'main_thread'.
+
+2012-09-14 Dmitry Antipov <dmantipov@yandex.ru>
+
+ Avoid out-of-range marker position (Bug#12426).
+ * insdel.c (replace_range, replace_range_2): Adjust
+ markers before overlays, as suggested by comments.
+ (insert_1_both, insert_from_buffer_1, adjust_after_replace):
+ Remove redundant check before calling offset_intervals.
+
+2012-09-14 Martin Rudalics <rudalics@gmx.at>
+
+ * xdisp.c (Fformat_mode_line): Unconditionally save/restore
+ current buffer (Bug#12387).
+
+2012-09-14 Juanma Barranquero <lekktu@gmail.com>
+
+ * makefile.w32-in ($(BLD)/alloc.$(O)): Update dependencies.
+
+2012-09-13 Paul Eggert <eggert@cs.ucla.edu>
+
+ Use a more backwards-compatible timer format (Bug#12430).
+ * keyboard.c (decode_timer): Get PSECS from the 8th (origin-0)
+ vector element, not from the 4th, since PSECS is now at the end.
+ (Fcurrent_idle_time): Doc fix.
+
+2012-09-13 Dmitry Antipov <dmantipov@yandex.ru>
+
+ Function to mark objects and remove killed buffers at once.
+ * alloc.c (discard_killed_buffers): Rename to ...
+ (mark_discard_killed buffers) ... new name. Add marking
+ of remaining objects. Fix comment. Adjust users.
+ (mark_object): Do not touch frame buffer lists here.
+ * frame.c (delete_frame): Reset frame buffer lists here.
+
+2012-09-13 Paul Eggert <eggert@cs.ucla.edu>
+
+ Better workaround for GNOME bug when --enable-gcc-warnings.
+ * emacsgtkfixed.c (G_STATIC_ASSERT): Remove, undoing last change.
+ Instead, disable -Wunused-local-typedefs. See Dmitry Antipov in
+ <http://lists.gnu.org/archive/html/emacs-devel/2012-09/msg00335.html>.
+
+ Simplify SIGIO usage (Bug#12408).
+ The code that dealt with SIGIO was crufty and confusing, e.g., it
+ played tricks like "#undef SIGIO" but these tricks were not used
+ consistently. Simplify mostly by not #undeffing standard symbols,
+ e.g., use "defined USABLE_SIGIO" (our symbol, which we can define
+ or not as we please) rather than "defined SIGIO" (standard symbol
+ that we probably shouldn't #undef).
+ * conf_post.h [USG5_4]: Do not include <sys/wait.h> here.
+ Modules that need it can include it.
+ [USG5_4 && emacs]: Likewise, do not include the streams stuff here.
+ * dispextern.h (ignore_sigio): New decl.
+ * emacs.c (shut_down_emacs): Invoke unrequest_sigio
+ unconditionally, since it's now a no-op if !USABLE_SIGIO.
+ * emacs.c (shut_down_emacs):
+ * keyboard.c (kbd_buffer_store_event_hold):
+ Use ignore_sigio rather than invoking 'signal' directly.
+ * keyboard.c (USABLE_FIONREAD && USG5_4): Include <sys/filio.h>,
+ for FIONREAD.
+ (FIONREAD, SIGIO): Do not #undef.
+ (tty_read_avail_input): Use #error rather than a syntax error.
+ * process.c [USG5_4]: Include <sys/stream.h> and <sys/stropts.h>,
+ for I_PIPE, used by SETUP_SLAVE_PTY.
+ (DATAGRAM_SOCKETS): Simplify defn, based on USABLE_FIONREAD.
+ * sysdep.c (croak): Remove; no longer needed. This bit of
+ temporary code, with Fred N. Fish's comment that it's temporary,
+ has been in Emacs since at least 1992!
+ (init_sigio, reset_sigio, request_sigio, unrequest_sigio):
+ Arrange for them to be no-ops in all cases when ! USABLE_SIGIO.
+ * syssignal.h (croak): Remove decl.
+ (SIGIO, SIGPOO, SIGAIO, SIGPTY): Do not #undef; that's too fragile.
+ * systty.h [!NO_TERMIO]: Do not include <termio.h>; no longer needed
+ now that we're termios-only.
+ (FIONREAD, ASYNC) [BROKEN_FIONREAD]: Do not #undef.
+ * term.c (dissociate_if_controlling_tty): Use #error rather than
+ a run-time error.
+
+ Work around GCC and GNOME bugs when --enable-gcc-warnings.
+ * emacsgtkfixed.c (G_STATIC_ASSERT): Redefine to use 'verify',
+ to work around GNOME bug 683906.
+ * image.c (jpeg_load_body) [HAVE_JPEG && lint]: Pacify gcc -Wclobber.
+ (struct my_jpeg_error_mgr) [HAVE_JPEG && lint]: New member fp.
+ This works around GCC bug 54561.
+
+2012-09-12 Paul Eggert <eggert@cs.ucla.edu>
+
+ More fixes for 'volatile' and setjmp/longjmp.
+ * eval.c (Fdefvar, Fcondition_case): Remove unnecessary 'volatile's.
+ * image.c (struct png_load_context) [HAVE_PNG]: New type.
+ (png_load_body) [HAVE_PNG]:
+ (jpeg_load_body) [HAVE_JPEG]:
+ New function, with most of the old parent function's body.
+ (png_load) [HAVE_PNG]:
+ (jpeg_load) [HAVE_JPEG]:
+ Invoke the new function, to avoid longjmp munging our locals.
+ (struct my_jpeg_error_mgr) [HAVE_JPEG]: New members cinfo, failure_code.
+ (my_error_exit) [HAVE_JPEG]: Don't trust 'setjmp' to return 2 when
+ longjmp is passed 2, as the C standard doesn't guarantee this.
+ Instead, store the failure code into mgr->failure_code.
+
+2012-09-12 Stefan Monnier <monnier@iro.umontreal.ca>
+
+ * keyboard.c (read_char, requeued_events_pending_p, Finput_pending_p)
+ (Fdiscard_input, quit_throw_to_read_char, init_keyboard)
+ (syms_of_keyboard): Remove support for unread-command-char.
+
+2012-09-12 Eli Zaretskii <eliz@gnu.org>
+
+ * w32proc.c (sys_kill): If PID is our process ID and the signal is
+ SIGABRT, call emacs_abort. Avoids silently exiting upon assertion
+ violation. (Bug#12426)
+
+2012-09-12 Paul Eggert <eggert@cs.ucla.edu>
+
+ * image.c (jpeg_memory_src): Don't assume string len fits in unsigned.
+
+2012-09-12 Stefan Monnier <monnier@iro.umontreal.ca>
+
+ * eval.c: Add `inhibit-debugger'.
+ (Qinhibit_debugger): New symbol.
+ (call_debugger): Bind it instead of Qdebug_on_error.
+ (maybe_call_debugger): Test Vinhibit_debugger.
+ (syms_of_eval): Define inhibit-debugger.
+ * xdisp.c (set_message): Don't bind Qinhibit_debug_on_message.
+ (syms_of_xdisp): Remove inhibit-debug-on-message.
+
+2012-09-11 Paul Eggert <eggert@cs.ucla.edu>
+
+ Avoid _setjmp/_longjmp problems with local nonvolatile variables.
+ If a nonvolatile local variable is written before a _longjmp to
+ the frame containing the variable, and is read after the _longjmp,
+ the value read is indeterminate. Some local variables of type
+ 'struct handler' and 'struct catchtag' are used in this way, so
+ mark each of their slots as volatile if the slot can be set before
+ _longjmp and read afterwards.
+ * lisp.h (struct handler): var and chosen_clause are now volatile.
+ (struct catchtag): val, next, and pdlcount are now volatile.
+
+ * bidi.c (bidi_push_it, bidi_pop_it):
+ * fns.c (copy_hash_table):
+ * image.c (define_image_type):
+ * keyboard.c (kbd_buffer_store_event_hold):
+ * process.c (Fprocess_send_eof):
+ * xfaces.c (x_create_gc) [HAVE_NS]:
+ * xgselect.c (xg_select):
+ Prefer assignment to memcpy when either will do.
+
+ * alloc.c (discard_killed_buffers): Tune and simplify a bit.
+ Use pointer-to-a-pointer to simplify and avoid a NILP check each
+ time an item is removed. No need to mark this function 'inline';
+ the compiler knows better than we do.
+
+2012-09-11 Jan Djärv <jan.h.d@swipnet.se>
+
+ * nsterm.m (ns_judge_scroll_bars): Pass NO to updateFrameSize.
+ (updateFrameSize:): Add delay parameter to updateFrameSize, send it
+ to change_frame_size (Bug#12388).
+ (windowDidResize:): Pass YES to updateFrameSize.
+
+ * nsterm.h: Add delay parameter to updateFrameSize.
+
+2012-09-11 Dmitry Antipov <dmantipov@yandex.ru>
+
+ Discard killed buffers from deleted window and frame objects.
+ This reduces an amount of references to killed buffers and
+ helps GC to reclaim them faster.
+ * alloc.c (discard_killed_buffers): New function.
+ (mark_object): Use it for deleted windows and frames.
+ (mark_object): If symbol's value is set up for a killed buffer
+ or deleted frame, restore it's global binding.
+ * data.c (swap_in_global_binding): Add GC notice.
+ (swap_in_symval_forwarding): Use convenient set_blv_where.
+ * window.c (wset_next_buffers, wset_prev_buffers): Move ...
+ * window.h: ... to here.
+
+2012-09-11 Dmitry Antipov <dmantipov@yandex.ru>
+
+ Convenient macro to check whether the buffer is live.
+ * buffer.h (BUFFER_LIVE_P): New macro.
+ * alloc.c, buffer.c, editfns.c, insdel.c, lread.c, marker.c:
+ * minibuf.c, print.c, process.c, window.c, xdisp.c: Use it.
+
2012-09-11 YAMAMOTO Mitsuharu <mituharu@math.s.chiba-u.ac.jp>
* xdisp.c (right_overwritten, right_overwriting): Also handle gstring
(ns_maybe_dumpglyphs_background): Remove fringe/internal border
adjustment.
(ns_dumpglyphs_image): Ditto.
- (ns_dumpglyphs_stretch): Fix coding style. Remove fringe/internal
+ (ns_dumpglyphs_stretch): Fix coding style. Remove fringe/internal
border adjustment.
(ns_set_vertical_scroll_bar): Remove variables barOnVeryLeft/Right and
their usage. Add fringe_extended_p and its use as in other terms.
* conf_post.h [SIGNAL_H_AHB]: Do not include <signal.h>;
no longer needed here.
* emacs.c (main): Inspect existing signal handler with sigaction,
- so that there's no need to block and unblock SIGHUP.
+ so that there's no need to block and unblock SIGHUP.
* sysdep.c (struct save_signal): New member 'action', replacing
old member 'handler'.
(save_signal_handlers, restore_signal_handlers):
2012-08-28 Jan Djärv <jan.h.d@swipnet.se>
* nsmenu.m (initWithContentRect:styleMask:backing:defer:): Initialize
- button_values to NULL. Call setStykeMask so dialogs get a close button.
+ button_values to NULL. Call setStykeMask so dialogs get a close button.
(windowShouldClose:): Set window_closed.
(dealloc): New member, free button_values.
- (process_dialog:): Make member function. Remove window argument,
- replace window with self. Count buttons and allocate and store values
+ (process_dialog:): Make member function. Remove window argument,
+ replace window with self. Count buttons and allocate and store values
in button_values.
(addButton:value:row:): value is int with the name tag. Call setTag
- with tag. Remove return self, declare return value as void.
+ with tag. Remove return self, declare return value as void.
(addString:row:): Remove return self, declare return value as void.
(addSplit): Remove return self, declare return value as void.
(clicked:): Remove return self, declare return value as void.
- Set dialog_return to button_values[seltag]. Code formatting change.
+ Set dialog_return to button_values[seltag]. Code formatting change.
(initFromContents:isQuestion:): Adjust call to process_dialog.
Code formatting change.
(timeout_handler:): Set timer_fired to YES.
* fontset.c (FONTSET_ADD): Return void, not Lisp_Object.
Otherwise, the compiler complains about (A?B:C) where B is void
- and C is Lisp_Object. This fixes an incompatibility with Sun C 5.12.
+ and C is Lisp_Object. This fixes an incompatibility with Sun C 5.12.
(fontset_add): Return void, for FONTSET_ADD.
2012-08-21 Paul Eggert <eggert@cs.ucla.edu>
Start main loop and wait for application defined event.
Inform select thread to stop selecting after main loop is exited.
(ns_term_init): Create selfds pipe and set non-blocking.
- Initialize select_mutex. Start the select thread (fd_handler).
+ Initialize select_mutex. Start the select thread (fd_handler).
(fd_handler:): Loop forever, wait for info from the main thread
to either start or stop selecting. When select returns, send
and appdefined event.
(xg_create_tool_bar): Allocate and initialize a xg_frame_tb_info
if not present.
(update_frame_tool_bar): Return early if data in xg_frame_tb_info
- is up to date. Otherwise store new data.
+ is up to date. Otherwise store new data.
(free_frame_tool_bar): Free xg_frame_tb_info if present.
2012-08-13 Dmitry Antipov <dmantipov@yandex.ru>
2012-08-11 Jan Djärv <jan.h.d@swipnet.se>
* nsterm.m (not_in_argv): New function.
- (application:openFile, application:openTempFile:):
+ (application:openFile, application:openTempFile:):
(application:openFileWithoutUI:, application:openFiles:): Open file
if not_in_argv returns non-zero (bug#12171).
2012-07-21 Jan Djärv <jan.h.d@swipnet.se>
- * nsterm.m (accessibilityAttributeValue): New function. (Bug#11134).
+ * nsterm.m (accessibilityAttributeValue): New function. (Bug#11134).
(conversationIdentifier): Return value is NSInteger.
* nsterm.m (accessibilityAttributeValue): Surround with NS_IMPL_COCOA.
* lisp.h (struct vectorlike_header): New field `nbytes',
adjust comment accordingly.
* alloc.c (enum mem_type): New type `MEM_TYPE_VECTOR_BLOCK'
- to denote vector blocks. Adjust users (live_vector_p,
+ to denote vector blocks. Adjust users (live_vector_p,
mark_maybe_pointer, valid_lisp_object_p) accordingly.
(COMMON_MULTIPLE): Move outside #if USE_LSB_TAG.
(VECTOR_BLOCK_SIZE, vroundup, VECTOR_BLOCK_BYTES),
(allocate_vector_from_block, init_vectors, allocate_vector_from_block)
(sweep_vectors): New functions.
(allocate_vectorlike): Return `zero_vector' as the only vector of
- 0 items. Allocate new vector from block if vector size is less than
+ 0 items. Allocate new vector from block if vector size is less than
or equal to VBLOCK_BYTES_MAX.
(Fgarbage_collect): Move all vector sweeping code to sweep_vectors.
(init_alloc_once): Add call to init_vectors.
change it's type from Lisp_Object to bitfield.
Change type of 'force_start', 'optional_new_start',
'last_had_star', 'update_mode_line' and 'start_at_line_beg'
- fields from Lisp_Object to bitfield. Adjust users accordingly.
+ fields from Lisp_Object to bitfield. Adjust users accordingly.
2012-05-31 Paul Eggert <eggert@cs.ucla.edu>
HCoop / bpt / emacs.git / blobdiff