Paul Eggert [Fri, 17 Jul 2015 18:54:24 +0000 (11:54 -0700)]
Fix hang with large yanks This should fix the bug fixed by Mike
Crowe's patch in:
https://lists.gnu.org/archive/html/emacs-devel/2015-07/msg00106.html
A problem in this area has been reported by several users; see
Bug#16737, Bug#17101, Bug#17026, Bug#17172, Bug#19320, Bug#20283.
This fix differs from Mike Crowe's patch in that it should avoid a
race condition that could lose SIGIO signals. ignore_sigio dates
back to the 1980s when some platforms couldn't block signals, and
could only ignore them, which led to races when signals arrived
while being ignored. We shouldn't have to worry about those old
platforms now.
* src/dispextern.h, src/sysdep.c (ignore_sigio): Remove.
* src/emacs.c (shut_down_emacs):
Don't call ignore_sigio; unrequest_sigio should suffice.
* src/keyboard.c (kbd_buffer_store_buffered_event):
Use unrequest_sigio, not ignore_sigio.
(kbd_buffer_get_event):
Call request_sigio when getting the ball rolling again.
Paul Eggert [Thu, 16 Jul 2015 07:48:40 +0000 (00:48 -0700)]
Better heuristic for C stack overflow
Improve the heuristic for distinguishing stack overflows from
other SIGSEGV causes (Bug#21004). Corinna Vinschen explained that
the getrlimit method wasn't portable to Cygwin; see:
https://www.cygwin.com/ml/cygwin/2015-07/msg00092.html
Corinna suggested pthread_getattr_np but this also has problems.
Instead, replace the low-level system stuff with a simple
heuristic based on known good stack addresses.
* src/eval.c, src/lisp.h (near_C_stack_top): New function.
* src/sysdep.c: Don't include <sys/resource.h>.
(stack_direction): Remove. All uses removed.
(stack_overflow): New function.
(handle_sigsegv): Use it instead of incorrect getrlimit heuristic.
Make SEGV fatal in non-main threads.
* epg-config.el (epg-gpgconf-program): New variable.
* epg.el (epg--start): Call `pinentry-start' if
allow-emacs-pinentry is set in ~/.gnupg/gpg-agent.conf.
Paul Eggert [Tue, 14 Jul 2015 13:59:26 +0000 (06:59 -0700)]
Clear gcprolist etc. after stack overflow
After stack overflow, command_loop calls init_eval, and this needs to
clear gcprolist and byte_stack_list (Bug#20996).
* src/alloc.c (init_alloc):
Move gcprolist and byte_stack_list initialization from here ...
* src/eval.c (init_eval): ... to here.
* src/macfont.m (macfont_family_cache): New variable.
(syms_of_macfont): Initialize it.
(macfont_available_families_cache): New variable.
(macfont_invalidate_family_cache, macfont_get_family_cache_if_present)
(macfont_set_family_cache, macfont_invalidate_available_families_cache)
(macfont_handle_font_change_notification)
(macfont_init_font_change_handler)
(macfont_copy_available_families_cache): New functions.
(macfont_create_family_with_symbol): Use font family caches.
(macfont_list, macfont_list_family): Use
macfont_copy_available_families_cache instead of
mac_font_create_available_families.
* lisp/progmodes/xref.el: Add `M-?' binding for
xref-find-references. Declare functions `grep-read-files' and
`grep-expand-template'.
(xref--read-identifier): Show the default value in the prompt.
When called with prefix argument, ask for file patterns to search as well
* lisp/progmodes/xref.el (xref-find-regexp): When called with
prefix argument, ask for file patterns to search as well. When
prompting for the directory, require an existing one.
(xref-collect-matches): Add a new argument, FILES. Use it in the
above function.
* lisp/progmodes/project.el (project-ignores): New generic
function, and an implementation for the VC project type.
* lisp/progmodes/xref.el (xref--rgrep-command): Split, as a
variant of rgrep-default-command that handles a generic list of
ignores.
(xref-collect-matches): Use it, and pass through to it the value
of the newly added argument.
(xref-find-regexp): Handle ignored paths within the project.
Remove outdated comment.
* lisp/vc/vc.el (vc-default-ignore-completion-table):
Skip the comments and the empty lines.
* lisp/cedet/ede.el: (project-try-ede): New function.
(project-root): New implementation.
* lisp/progmodes/elisp-mode.el (emacs-lisp-mode):
Set project-search-path-function.
(elisp--xref-find-references): Delegate some logic to
project-search-path.
(elisp-search-path): New function.
(elisp-xref-find): Don't implement `matches' anymore.
* lisp/progmodes/etags.el: Don't implement `matches'.
Delegate some logic to project-search-path.
(etags-search-path): New function.
* lisp/progmodes/xref.el (xref-find-function):
Remove `matches' from the API.
(xref-find-regexp): Move whatever common logic was in elisp and
etags implementations, and search the directories returned by
project-directories and project-search-path.
Nicolas Petton [Thu, 9 Jul 2015 17:43:41 +0000 (19:43 +0200)]
Add support for gv.el in map.el
* lisp/emacs-lisp/map.el (map-elt, map-delete): Declare a gv-expander.
* lisp/emacs-lisp/map.el (map-put): Refactor using `setf' and `map-elt'.
* test/automated/map-tests.el: Update tests to work with the new
implementations of map-elt and map-put.
* lisp/gnus/nnimap.el (nnimap-open-connection-1): Always query
capabilities, so that a 'plain value for the `nnimap-stream' server
variable is handled correctly.
* doc/misc/gnus.texi (Customizing the IMAP Connection):
Document the 'plain option.
Eric Abrahamsen [Wed, 8 Jul 2015 22:28:50 +0000 (22:28 +0000)]
gnus-group.el: Check if group names are already strings
* lisp/gnus/gnus-group.el (gnus-group-group-name):
The group name may already be a string.
Specifically, in the group list reached from the *Server* buffer,
the 'gnus-group text property returns a string. Everywhere else
it returns a symbol.
Stefan Monnier [Tue, 7 Jul 2015 15:37:04 +0000 (11:37 -0400)]
(gv-setter, gv-synthetic-place, gv-delay-error): New funs/macros
* lisp/emacs-lisp/gv.el (gv-setter): New function.
(gv-invalid-place): New error.
(gv-get): Use them.
(gv-synthetic-place, gv-delay-error): New places.
* lisp/emacs-lisp/cl-generic.el (cl--generic-setf-rewrite): Remove.
(cl-defgeneric, cl-defmethod): Use gv-setter.
Ken Brown [Tue, 7 Jul 2015 14:24:09 +0000 (10:24 -0400)]
Improve recent change to emacsclient on Cygwin
* lisp/server.el (server-process-filter): Remove redundant check
that 'cygwin-convert-file-name-from-windows' is defined as a
function on Cygwin. Don't call that function unless its argument
starts with a drive letter.
Stefan Monnier [Tue, 7 Jul 2015 06:14:16 +0000 (02:14 -0400)]
Add online-help support to describe types
* lisp/help-fns.el (describe-symbol-backends): Move to help-mode.el.
(describe-symbol): Improve the selection of default.
* lisp/help-mode.el: Require cl-lib.
(describe-symbol-backends): Move from help-fns.el.
(help-make-xrefs): Use it.
* lisp/emacs-lisp/cl-extra.el (describe-symbol-backends): Add entry
for types.
(cl--typedef-regexp): New const.
(find-function-regexp-alist): Add entry for types.
(cl-help-type, cl-type-definition): New buttons.
(cl-find-class): New function.
(cl-describe-type): New command.
(cl--describe-class, cl--describe-class-slot)
(cl--describe-class-slots): New functions, moved from eieio-opt.el.
* lisp/emacs-lisp/cl-generic.el (cl--generic-method-documentation)
(cl--generic-all-functions, cl--generic-specializers-apply-to-type-p):
New functions. Moved from eieio-opt.el.
(cl--generic-class-parents): New function, extracted from
cl--generic-struct-specializers.
(cl--generic-struct-specializers): Use it.
* lisp/emacs-lisp/cl-macs.el (cl-defstruct): Use pcase-dolist.
Improve constructor's docstrings.
(cl-struct-unknown-slot): New error.
(cl-struct-slot-offset): Use it.
* lisp/emacs-lisp/cl-preloaded.el (cl-struct-define): Record the type
definition in current-load-list.
* lisp/emacs-lisp/eieio-core.el (eieio--known-slot-names): New var.
(eieio--add-new-slot): Set it.
(eieio-defclass-internal): Use new name for current-load-list.
(eieio-oref): Add compiler-macro to warn about unknown slots.
* lisp/emacs-lisp/eieio.el (defclass): Update eieio--known-slot-names
as compile-time as well. Improve constructor docstrings.
* lisp/emacs-lisp/eieio-opt.el (eieio-help-class)
(eieio--help-print-slot, eieio-help-class-slots): Move to cl-extra.el.
(eieio-class-def): Remove button.
(eieio-help-constructor): Use new name for load-history element.
(eieio--specializers-apply-to-class-p, eieio-all-generic-functions)
(eieio-method-documentation): Move to cl-generic.el.
(eieio-display-method-list): Use new names.
* lisp/help-fns.el (describe-symbol-backends): New var.
(help-xref-stack-item): Declare.
(describe-symbol): Rename from describe-function-or-variable.
Rewrite using describe-symbol-backends instead of help-xref-interned.
* lisp/help.el (help-map): Use it.
* lisp/help-mode.el (help-symbol, help-follow-symbol): Use it.
(help-xref-interned): Make it into an obsolete alias.
Wolfgang Jenkner [Mon, 6 Jul 2015 13:10:03 +0000 (15:10 +0200)]
Fix parsing glitches in dired-mark-sexp (bug#13575)
* lisp/dired-x.el (dired-x--string-to-number): New function.
(dired-mark-sexp): Use it. Tweak dired-re-inode-size. Fix usage
of directory-listing-before-filename-regexp. Consider
forward-word harmful and replace it. Add more verbiage in
comments and doc string.
python.el: Respect process environment for remote shells
* lisp/progmodes/python.el
(python-shell-calculate-process-environment): Calculate
process-environment or tramp-remote-process-environment depending
whether current file is remote.
(python-shell-calculate-exec-path): Calculate exec-path or
tramp-remote-path depending whether current file is remote.
(python-shell-with-environment): New macro.
(python-shell-prompt-detect, python-shell-calculate-command)
(python-shell-make-comint, python-check): Use it.
python.el: Avoid making let-bound defvars buffer local (Bug#18244)
* lisp/progmodes/python.el (python-shell--interpreter)
(python-shell--interpreter-args): New vars.
(inferior-python-mode, python-shell-make-comint): Use them.
Ian Kelling [Mon, 6 Jul 2015 01:14:25 +0000 (18:14 -0700)]
Avoid returning early reading process output due to SIGIO
* src/process.c (wait_reading_process_output): Extend the behavior of
not breaking due to not finding output when a timer has lowered the
timeout to include when SIGIO lowers the timeout.
Ian Kelling [Mon, 6 Jul 2015 00:00:26 +0000 (17:00 -0700)]
Don't return as fast reading any process output
* src/process.c (wait_reading_process_output):
The patch for Bug#17647 returns too fast sometimes when reading
from any processes. Revert part of it, and limit the timeout more
sensibly (Bug#20978).
Ian Kelling [Sun, 5 Jul 2015 22:38:29 +0000 (15:38 -0700)]
; Rename local var nsecs to adaptive_nsecs
* src/process.c (wait_reading_process_output): Rename inner nsecs to
adaptive_nsecs. There is already an nsecs, and this function is
confusing enough (Bug#20978).
Ian Kelling [Sun, 5 Jul 2015 22:35:23 +0000 (15:35 -0700)]
; Rename local var to match function name
* src/process.c (wait_reading_process_output, status_notify):
Previously the function wait_reading_process_input was renamed to the
more logical wait_reading_process_output. Make its local variables
consistent with that change (Bug#20978).
Ian Kelling [Sun, 5 Jul 2015 22:30:27 +0000 (15:30 -0700)]
Remove ADAPTIVE_READ_BUFFERING ifdef
* src/process.c (make-process, make-pipe-process, deactivate_process)
(wait_reading_process_output, read_process_output, send_process)
(init_process_emacs): ifdef ADAPTIVE_READ_BUFFERING was originally
added in case there was an operating system in which it was not
useful. That was 11 years ago and it hasn't happened. Make
development easier by not considering the effect of changes on a
theoretical OS where this is disabled (Bug#20978).
Christoph Wedler [Fri, 19 Jun 2015 13:38:24 +0000 (13:38 +0000)]
Respect `prog-indentation-context' in python.el
* lisp/progmodes/python.el (python-indent-guess-indent-offset)
(python-indent-context, python-indent--calculate-indentation)
(python-info-current-defun)
(python-info-dedenter-opening-block-message)
(python-info-line-ends-backslash-p)
(python-info-beginning-of-backslash)
(python-info-continuation-line-p): Use `prog-widen'.
(python-indent--calculate-indentation)
(python-indent--calculate-levels)
(python-indent-calculate-indentation): Use `prog-first-column'.
(python-indent--calculate-levels): Simplify.
Ignore also initial empty lines for syntax calculation.
* lisp/progmodes/python.el (python-indent-context): Return
:no-indent for first non-empty line, not just in line 1.
* test/automated/python-tests.el (python-indent-base-case)
(python-indent-inside-paren-1, python-indent-inside-paren-2)
(python-indent-inside-paren-3, python-indent-inside-paren-4)
(python-indent-inside-paren-5, python-indent-inside-paren-6)
(python-indent-after-backslash-1)
(python-indent-after-backslash-2)
(python-indent-after-backslash-3)
(python-indent-after-backslash-4, python-indent-inside-string-1):
Expect :no-indent for first non-empty line.
Eli Zaretskii [Sat, 4 Jul 2015 11:23:27 +0000 (14:23 +0300)]
Fix mouse pointer on w32 when a menu is active
* src/w32fns.c (w32_wnd_proc): Don't change the mouse pointer
shape while a menu is in use. This started happening since we now
send WM_EMACS_SHOWCURSOR messages when the mouse moves.
Martin Rudalics [Fri, 3 Jul 2015 13:15:51 +0000 (15:15 +0200)]
Fix some issues with `window-divider-mode'
* lisp/frame.el (window-divider-default-places): New option.
(window-divider-mode): Remove option.
(window-divider-mode): Make it a "regular" minor mode.
(window-divider-width-valid-p): Drop frame- prefix.
(window-divider-mode-apply): New argument ENABLE. Drop frame-
prefix. Handle `window-divider-default-places'.
(frame--window-divider-mode-set-and-apply): Remove.
(window-divider-default-bottom-width)
(window-divider-default-right-width): Drop :group entries.
* lisp/menu-bar.el (menu-bar-bottom-and-right-window-divider)
(menu-bar-right-window-divider, menu-bar-bottom-window-divider)
(menu-bar-no-window-divider): Set `window-divider-default-places'
and call `window-divider-mode'.
* doc/emacs/frames.texi (Window Dividers): Document
`window-divider-default-places'.
Martin Rudalics [Thu, 2 Jul 2015 12:58:31 +0000 (14:58 +0200)]
Some further fixes in Change Window node. (Bug#20183)
* doc/emacs/windows.texi (Change Window): Replace "rearranging"
by "resizing" in section title. Add some concept indices.
Suggested by N. Jackson (Bug#20183).