XEmacs hangs in gnus-score-edit-current-scores (`file-name-directory'/`make-directory' problem)
22 years, 9 months
Steven E. Harris
Steven E. Harris
Gnus v5.8.8
XEmacs 21.4 (patch 3) "Academic Rigor" [Lucid] (i686-pc-cygwin) of Thu May 17 2001 on SHALOM
configured using `configure --use-union-type --with-netinstall --with-x=no --with-site-lisp --bindir=/usr/local/bin/i686-pc-cygwin'
When trying to edit the score file in my Gnus mail group
"nnimap+speakeasy:INBOX", XEmacs hangs in an infinite loop trying to
create the directory "/home/sharr/doc/news/nnimap+speakeasy:" - which
should not and apparently cannot be created.
The trouble is in gnus-score.el within the function
`gnus-score-edit-current-scores':
,----
| (defun gnus-score-edit-current-scores (file)
| "Edit the current score alist."
| (interactive (list gnus-current-score-file))
| (if (not gnus-current-score-file)
| (error "No current score file")
| (let ((winconf (current-window-configuration)))
| (when (buffer-name gnus-summary-buffer)
| (gnus-score-save))
| (gnus-make-directory (file-name-directory file))
| (setq gnus-score-edit-buffer (find-file-noselect file))
| (gnus-configure-windows 'edit-score)
| (gnus-score-mode)
| (setq gnus-score-edit-exit-function 'gnus-score-edit-done)
| (make-local-variable 'gnus-prev-winconf)
| (setq gnus-prev-winconf winconf))
| (gnus-message
| 4 (substitute-command-keys
| "\\<gnus-score-mode-map>\\[gnus-score-edit-exit] to save edits"))))
`----
Note the call to `gnus-make-directory.' The argument "file" is
initially "/home/sharr/doc/news/nnimap+speakeasy:INBOX.SCORE". The
function `file-name-directory' applied to this file returns the path
"/home/sharr/doc/news/nnimap+speakeasy:" - which contains an illegal
colon character. When this path is eventually passed to the function
`make-directory', XEmacs hangs, consuming the full CPU and more and
more memory. Looking at the `make-directory' code, one can see a loop
that apparently gets stuck waiting for this impossible file name to
exist.
This is another artifact of the strange file name conversions used in
XEmacs on Windows. The function `nnheader-translate-file-chars' is
used occasionally - but not consistently - by Gnus to come up with
file names that can actually exist on Windows. Strangely, '+' is
considered an illegal character even though it's a valid Windows path
character. More dangerously, ':' is considered legal even though it's
obviously not:
,----
| (nnheader-translate-file-chars "a+b")
| "a-b"
|
| (nnheader-translate-file-chars "a:b")
| "a:b"
`----
This bug is a symptom of interaction among flaws in
`gnus-edit-current-scores', `nnheader-translate-file-chars',
`file-name-directory', and `make-directory' - all considered in the
context of Windows NT atop Cygwin.
--
Steven E. Harris :: seh(a)speakeasy.org
GnuPG :: 0x70248E67
- 2 participants
- 2 comments
extw-Xt.c patch resend
22 years, 9 months
Hollander, Isaac (IT)
Hollander, Isaac (IT)
The patch attached to my previous mail was incorrect -- it was based
on diffing against a work-in-progress file. Here's the real patch.
--- /ms/dev/fsf/xemacs/21.1.14/src/src/extw-Xt.c Wed Dec 18 17:44:05 1996
+++ extw-Xt.c Fri May 25 17:28:17 2001
＠＠ -23,6 +23,7 ＠＠
#ifdef emacs
#include <config.h>
+#include <stdio.h>
#ifndef EXTERNAL_WIDGET
ERROR! This ought not be getting compiled if EXTERNAL_WIDGET is undefined
＠＠ -34,47 +35,10 ＠＠
#endif
#include <X11/Intrinsic.h>
-#include <X11/IntrinsicP.h>
#include <stdlib.h>
#include <stdio.h>
#include "extw-Xt.h"
-/* Yeah, that's portable!
-
- Why the hell didn't the Xt people just export this function
- for real? */
-
-#if (XT_REVISION > 5)
-int _XtWaitForSomething(
- XtAppContext app,
- _XtBoolean ignoreEvents,
- _XtBoolean ignoreTimers,
- _XtBoolean ignoreInputs,
- _XtBoolean ignoreSignals,
- _XtBoolean block,
-#ifdef XTHREADS
- _XtBoolean drop_lock,
-#endif
- unsigned long *howlong);
-
-# ifndef XTHREADS
-# define _XtwaitForSomething(timers,inputs,events,block,howlong,appCtx) \
- _XtWaitForSomething(appCtx,events,timers,inputs,0,block,howlong)
-# else
-# define _XtwaitForSomething(timers,inputs,events,block,howlong,appCtx) \
- _XtWaitForSomething(appCtx,events,timers,inputs,0,block,1,howlong)
-# endif
-#else
-int _XtwaitForSomething(
- Boolean ignoreTimers,
- Boolean ignoreInputs,
- Boolean ignoreEvents,
- Boolean block,
- unsigned long *howlong,
- XtAppContext app
- );
-#endif
-
#ifdef DEBUG_WIDGET
static int geom_masks[] = {
＠＠ -195,49 +159,77 ＠＠
/* check if an event is of the sort we're looking for */
static Bool
-isMine(Display *dpy, XEvent *event, char *arg)
+isMine(XEvent *event, QueryStruct *q)
{
- QueryStruct *q = (QueryStruct *) arg;
- Widget w = q->w;
-
- if ( (dpy != XtDisplay(w)) || (event->xany.window != XtWindow(w)) ) {
- return FALSE;
- }
- if (event->xany.serial >= q->request_num) {
- if (event->type == ClientMessage &&
- event->xclient.message_type == a_EXTW_NOTIFY &&
- event->xclient.data.l[0] == 1 - extw_which_side &&
- event->xclient.data.l[1] == q->type)
- return TRUE;
- }
- return FALSE;
+ Widget w = q->w;
+
+ if ( (event->xany.display != XtDisplay(w)) || (event->xany.window != XtWindow(w)) )
+ {
+ return FALSE;
+ }
+ if (event->xany.serial >= q->request_num)
+ {
+ if (event->type == ClientMessage &&
+ event->xclient.message_type == a_EXTW_NOTIFY &&
+ event->xclient.data.l[0] == 1 - extw_which_side &&
+ event->xclient.data.l[1] == q->type)
+ {
+ return TRUE;
+ }
+ }
+ return FALSE;
+}
+
+void responseTimeOut(XtPointer clientData,XtIntervalId *id)
+{
+ Bool *expired=(Bool *)clientData;
+ *expired=TRUE;
}
/* wait for a ClientMessage of the specified type from the other widget, or
- time-out. isMine() determines whether an event matches. Culled from
- Shell.c. */
+ time-out. isMine() determines whether an event matches.
+ Took out the call to _XtWaitForSomething and replaced it with public
+ Xt api's.
+*/
Bool
extw_wait_for_response(Widget w, XEvent *event, unsigned long request_num,
en_extw_notify type, unsigned long timeout)
{
- XtAppContext app = XtWidgetToApplicationContext(w);
- QueryStruct q;
-
- XFlush(XtDisplay(w));
- q.w = w;
- q.request_num = request_num;
- q.type = type;
-
- for(;;) {
- /*
- * look for match event
- */
- if (XCheckIfEvent( XtDisplay(w), event, isMine, (char*)&q))
- return TRUE;
- if (_XtwaitForSomething(TRUE, TRUE, FALSE, TRUE, &timeout, app)
- != -1) continue;
- if (timeout == 0)
- return FALSE;
- }
+ XtAppContext app = XtWidgetToApplicationContext(w);
+ XtInputMask inputMask;
+ QueryStruct q;
+ Bool expired;
+ XtIntervalId id;
+
+ q.w = w;
+ q.request_num = request_num;
+ q.type = type;
+ expired=FALSE;
+
+ id=XtAppAddTimeOut(app,timeout,responseTimeOut,&expired);
+ while (!expired)
+ {
+ inputMask=XtAppPending(app);
+ if (inputMask & XtIMXEvent)
+ {
+ XtAppNextEvent(app,event);
+ if (isMine(event,&q))
+ {
+ if (!expired) XtRemoveTimeOut(id);
+ return True;
+ }
+ else
+ {
+ /* Do Nothing and go back to waiting */
+ }
+ }
+ if (inputMask & XtIMTimer)
+ {
+ /* Process the expired timer */
+ XtAppProcessEvent(app,XtIMTimer);
+ }
+ }
+ /* Must have expired */
+ return False;
}
--
Isaac Hollander Morgan Stanley Dean Witter
isaac.hollander(a)morganstanley.com 1585 Broadway, 2nd floor
(212) 761 3417 New York, NY
- 1 participants
- 0 comments
external widget and libXt
22 years, 9 months
Hollander, Isaac (IT)
Hollander, Isaac (IT)
This note documents why XEmacs external widget is important, and
provides a fix to some annoying behavior on Solaris 7 when using
external widget under XEmacs 21.1.14 on Solaris 7. This writeup and
the patch are both courtesy of Kenny Chien
(Kenny.Chien(a)morganstanley.com).
Morgan Stanley's in-house C++ widget toolkit is based solely on the
X11 libraries, with no dependency on Xt or Motif. This toolkit
implements a Motif look and feel.
The use of XEmacs's external widget feature was driven by a real
business need in a trading application. In this application, users
structure business logic using a high level business specific
programming language.
Instead of expending resources in implementing a full blown editor
widget, we decided to give XEmacs external widget a try. It has been
incorporated into our toolkit and used in production since 1997.
The editor widget is also supported via Java on Solaris and NT. At
least 3 production applications, two in C++ and one in Java, utilize
this editor widget. The response from the user community has been
excellent because there are simply no better editor widgets out there.
Now, for the technical content. The function extw_wait_for_response()
in extw-Xt.c behaves inconsistently between Solaris 2.5.1 and Solaris
7. More specifically, the underlying _XtWaitForSomething() function
invoked by extw_wait_for_response() behaves differently when a timeout
occurs. In Solaris 7, it doesn't seem to return -1 when it times out,
it returns 1 instead. extw_wait_for_response() therefore loops
indefinitely causing xemacs to hang.
The fix to this problem is to not use _XtWaitForSomething(). The
biggest reason is of course it is an undocumented internal Xt
function, there is technically no guarantee of the existence and
the behavior of this function in future releases.
This patch rewrites extw_wait_for_response() to use only public Xt
functions and at the same time preserve the semantics of this
function. It will now behave consistently and timeout properly.
Here is the patch, against XEmacs 21.1.14. extw_wait_for_response()
seems to have been rewritten in XEmacs 21.4, but it still uses the
undocumented _XtWaitForSomething function. This method is probably
more portable...
--- extw-Xt.c.save Fri May 25 17:28:57 2001
+++ extw-Xt.c Fri May 25 17:28:17 2001
＠＠ -35,47 +35,10 ＠＠
#endif
#include <X11/Intrinsic.h>
-#include <X11/IntrinsicP.h>
#include <stdlib.h>
#include <stdio.h>
#include "extw-Xt.h"
-/* Yeah, that's portable!
-
- Why the hell didn't the Xt people just export this function
- for real? */
-
-#if (XT_REVISION > 5)
-int _XtWaitForSomething(
- XtAppContext app,
- _XtBoolean ignoreEvents,
- _XtBoolean ignoreTimers,
- _XtBoolean ignoreInputs,
- _XtBoolean ignoreSignals,
- _XtBoolean block,
-#ifdef XTHREADS
- _XtBoolean drop_lock,
-#endif
- unsigned long *howlong);
-
-# ifndef XTHREADS
-# define _XtwaitForSomething(timers,inputs,events,block,howlong,appCtx) \
- _XtWaitForSomething(appCtx,events,timers,inputs,0,block,howlong)
-# else
-# define _XtwaitForSomething(timers,inputs,events,block,howlong,appCtx) \
- _XtWaitForSomething(appCtx,events,timers,inputs,0,block,1,howlong)
-# endif
-#else
-int _XtwaitForSomething(
- Boolean ignoreTimers,
- Boolean ignoreInputs,
- Boolean ignoreEvents,
- Boolean block,
- unsigned long *howlong,
- XtAppContext app
- );
-#endif
-
#ifdef DEBUG_WIDGET
static int geom_masks[] = {
＠＠ -195,35 +158,24 ＠＠
/* check if an event is of the sort we're looking for */
-static FILE *file;
static Bool
isMine(XEvent *event, QueryStruct *q)
{
Widget w = q->w;
- fprintf(file,"Event Type is %d\n",event->type);
- if (event->type == ClientMessage)
- {
- fprintf(file,"***** It's a ClientMessage.*****\n");
- }
-
+
if ( (event->xany.display != XtDisplay(w)) || (event->xany.window != XtWindow(w)) )
{
- if ( (event->xany.display != XtDisplay(w)))
- {
- fprintf(file,"Wrong Display\n");
- }
- else
- {
- fprintf(file,"Wrong Window ID\n");
- }
return FALSE;
- }
- if (event->xany.serial >= q->request_num) {
+ }
+ if (event->xany.serial >= q->request_num)
+ {
if (event->type == ClientMessage &&
event->xclient.message_type == a_EXTW_NOTIFY &&
event->xclient.data.l[0] == 1 - extw_which_side &&
event->xclient.data.l[1] == q->type)
- return TRUE;
+ {
+ return TRUE;
+ }
}
return FALSE;
}
＠＠ -235,8 +187,10 ＠＠
}
/* wait for a ClientMessage of the specified type from the other widget, or
- time-out. isMine() determines whether an event matches. Culled from
- Shell.c. */
+ time-out. isMine() determines whether an event matches.
+ Took out the call to _XtWaitForSomething and replaced it with public
+ Xt api's.
+*/
Bool
extw_wait_for_response(Widget w, XEvent *event, unsigned long request_num,
＠＠ -247,7 +201,6 ＠＠
QueryStruct q;
Bool expired;
XtIntervalId id;
- file=fopen("/tmp/debug","a");
q.w = w;
q.request_num = request_num;
＠＠ -257,46 +210,26 ＠＠
id=XtAppAddTimeOut(app,timeout,responseTimeOut,&expired);
while (!expired)
{
-/* fprintf(file,"Looking for Pending event.\n"); */
inputMask=XtAppPending(app);
if (inputMask & XtIMXEvent)
{
- fprintf(file,"Got an X Event.\n");
- fflush(file);
XtAppNextEvent(app,event);
- if (event->type==Expose)
- {
- fprintf(file,"Dispatching Expose.\n");
-/* XtDispatchEvent(event); */
- }
- else if (isMine(event,&q))
+ if (isMine(event,&q))
{
- fprintf(file,"===========> Got the Client Message Event\n");
- fflush(file);
if (!expired) XtRemoveTimeOut(id);
return True;
}
else
{
/* Do Nothing and go back to waiting */
- fprintf(file,"Don't want this X Event.\n");
- fflush(file);
}
}
if (inputMask & XtIMTimer)
{
/* Process the expired timer */
- fprintf(file,"Got a timer event.\n");
XtAppProcessEvent(app,XtIMTimer);
}
- if (inputMask & XtIMAlternateInput)
- {
-/* fprintf(file,"Got some Alternate Input Event.\n"); */
- }
}
/* Must have expired */
- fprintf(file,"Timer Expired.\n");
- fflush(file);
- fclose(file);
return False;
}
--
Isaac Hollander Morgan Stanley Dean Witter
isaac.hollander(a)morganstanley.com 1585 Broadway, 2nd floor
(212) 761 3417 New York, NY
- 1 participants
- 0 comments
RE: Dired Toolbar button broken
22 years, 9 months
Jim Horning
Jim Horning
This does not work as of 21.4.4 as well.
On Windows, the Open File dialog is popped up instead and will not allow you
to select a directory to display.
Jim
-----Original Message-----
From: David Masterson [mailto:dmaster＠synopsys.com]
Sent: Wednesday, July 11, 2001 3:23 PM
To: xemacs-beta(a)xemacs.org
Subject: Dired Toolbar button broken
In XEmacs 21.4 (patch 3) "Academic Rigor" [Lucid] (i586-pc-win32) of Thu May
17 2001 on SHALOM
configured using `configure UNKNOWN'
Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:
If you click on the Dired toolbar button, an MS-Windows style "Open
File" dialog box is popped up. The problem is that this popup does
not allow you to pick a *directory* which is what Dired wants.
Recent keystrokes:
misc-user
Recent messages (most recent first):
Loading mail-abbrevs...
Loading emacsbug...done
Loading emacsbug...
Loading c:\MyDocuments\.emacs-dir\my-xemacs.el...done
Auto Fill enabled in Text modes
Loading smtpmail...done
Loading smtpmail...
Loading clearcase...done
Loading clearcase...
Loading c:\MyDocuments\.emacs-dir\my-xemacs.el...
- 5 participants
- 4 comments
Large TAG files not working in XEmacs
22 years, 9 months
Sanjeev Shashikant Deshmukh
Sanjeev Shashikant Deshmukh
This bug report will be sent to the XEmacs Development Team,
not to your local site managers!!
Please write in English, because the XEmacs maintainers do not have
translators to read other languages for them.
In XEmacs 21.5 (beta1) "anise" [Lucid] (sparc-sun-solaris2.6) of Fri Jul 6 2001 on urvashi
configured using `configure --site-includes=/usr/local/include:/users/sdeshmuk/gnu/include --site-libraries=/usr/local/lib:/users/sdeshmuk/gnu/lib:/usr/openwin/lib/:/net/urvashi/usr2/sdeshmuk/kde/slib '--prefix=/users/sdeshmuk/gnu/XEmacs/beta/xemacs-21.5.1 ' --with-xpm=yes --with-workshop=yes --with-mule=no'
Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:
**********************************************************************
TAGS files big in size are not working. I have tags files which are
> 10 Mb. They work just fine with "vi" and "GNU Emacs" both. As you
load this TAGS file in XEmacs, it waits forever showing the garbage
collection cursor.
**********************************************************************
Recent keystrokes:
down down down down down down down down down down down
down down down down down down down down down down down
down down down down down up C-e C-a down button1 button1up
button2 button2up M-x v s BS BS s e t - v a r i TAB
RET t a g s - a l TAB RET t RET M-. down down down
C-e up C-a C-e down down down down down down down down
prior button1 button1up prior C-x 1 up C-up up up up
up up up up up up up C-e button1 button1up C-g mis
c-user
Recent messages (most recent first):
Parsing BBDB... (frobnicating...)
Loading byte-optimize...done
Loading byte-optimize...
Upgrade BBDB to version 6? (y or n) Yes
Upgrade BBDB to version 6? (y or n)
Loading bbdb-migrate...done
Loading bbdb-migrate...
Parsing BBDB... (reading...)
Parsing BBDB...
Loading bbdb-com...done
- 3 participants
- 3 comments
Release candidate for XEmacs 21.4.4 / my vacation
22 years, 9 months
Stephen J. Turnbull
Stephen J. Turnbull
Finally I've had time to get to the backlog of patches aimed at the
next release of the 21.4 series.
Once I've finished with 21.4.4, I'll try to roll up a 21.5 beta. This
will likely exhaust my available time before my vacation (7/30--8/6
upstate NY, 8/6--8/12 Bay Area). I expect to be in email contact, but
not CVS-able.
I'd like to avoid the kind of fiasco that happened with 21.4.2, and
I've also had too many context switches since early May to really be
on top of things. So I'd appreciate it if people would scan the list
of changes below for any glaring errors of omission or commission.
I've still got some checking to do myself (especially Windows builds),
and a few outstanding patches for Japanese I haven't decided what to
do with, so I'm targeting next Monday (7/23) for the release. All
problems are my responsibility, of course, but your help would be very
welcome!
Ben, do you remember what the story is with set_file_times()? Adrian
thinks it's essential, but it seems from your discussion that his
patch shouldn't actually be any help. Is the mouse-track-insert stuff
safe/recommended for 21.4? (The set_file_times patches are IN, the
mouse-track-insert patches are NOT, yet.)
The release candidate is on a temporary CVS branch "candidate-21-4".
Get it with "cvs update -r candidate-21-4". Build reports (especially
from Master Builder Perry ;-) would be very welcome.
You can also look at the changes applied to the code base with "cvs
diff -r release-21-4 -r candidate-21-4 ..." (either globally or for a
directory, subset of files, or individual file).
The following list of changes is also available as 21.4.4-Changes in
the root directory of the candidate-21-4 branch.
============================== 21.4.4-Changes ==============================
The changes in this release candidate are grouped in the following sections.
Some patches overlap a couple of sections. I (somewhat arbitrarily) assigned
them to the section I thought was most important.
UI and API CHANGES
GENERAL BUG FIXES
MS WINDOWS - printing
MS WINDOWS - set_file_times
MS WINDOWS - C-g handling
MS WINDOWS - miscellaneous
GTK
MISCELLANEOUS
DOCUMENTATION
Please review especially the UI and GENERAL sections, and others as you
have interest/expertise. The MISCELLANEOUS and DOCUMENTATION sections are
entirely uncontroversial and may safely be ignored.
I'm not committed to anything here irrevocably; this is just my first pass
at the next release candidate. Nor are things I've not included here
necessarily ruled out. Let me know if anything in needs out, or anything
out needs in, in your opinion.
UI and API CHANGES
<y9lwv966j49.fsf(a)sams.informatik.uni-tuebingen.de>
./src/ChangeLog
2001-03-31 Mike Sperber <mike(a)xemacs.org>
* search.c (Freplace_match): Support FSF-style specification of
match subexpression when applied to a buffer.
<15160.50052.863645.73187(a)europe.nortel.com>
./lisp/ChangeLog
2001-06-26 Robert Pluim <rpluim(a)bigfoot.com>
* packages.el (locate-library): Use read-library-name for completion.
<15160.50790.460079.416332(a)europe.nortel.com>
./lisp/ChangeLog
2001-06-26 Robert Pluim <rpluim(a)bigfoot.com>
* package-ui.el (pui-install-selected-packages): reverse the lists
of packages so that they get handled in the same order as they
were selected.
<microsoft-free.x4pub3lxg4.fsf(a)slackware.mynet.pc>
./src/ChangeLog
2001-07-14 Sean MacLennan <seanm(a)storm.ca>
* package-admin.el (package-install-hook): New.
(package-delete-hook): New.
(package-admin-add-single-file-package): Use package-delete-hook.
* package-get.el (package-get): Use package-install-hook.
<sxsoftm4b29.fsf(a)florida.arsdigita.de>
./lisp/ChangeLog
2001-04-24 Hrvoje Niksic <hniksic(a)arsdigita.com>
* about.el (about-mailto-link): Use compose-mail for sending mail.
<muxy9sz343m.fsf(a)uzeb.lrde.epita.fr>
./lisp/ChangeLog
2001-04-17 Didier Verna <didier(a)xemacs.org>
* startup.el (load-user-init-file): define `custom-file' before
loading the user's init file.
<mux1yqqdqbq.fsf(a)uzeb.lrde.epita.fr>
./lisp/ChangeLog
2001-04-18 Didier Verna <didier(a)xemacs.org>
* cus-edit.el (Custom-reset-standard): reset to standard settings
not only when the buffer's :custom-state is 'modified, but also
when it is 'set or 'saved.
GENERAL BUG FIXES
<uofrqvxzn.fsf(a)ispras.ru>
./lisp/ChangeLog
2001-06-15 Nick Pakoulin <npak(a)ispras.ru>
* help.el (frob-help-extents): Require 'hyper-apropos for faces.
<mqqoftzthg6.fsf(a)eng.cam.ac.uk>
./ChangeLog
2001-04-14 Gordon Sadler <gbsadler1(a)lcisp.com>
* configure.in: Add GTK_CFLAGS to CPPFLAGS for glade.h check.
<15167.31192.971882.809294(a)tyranny.hsys.msk.ru>
./ChangeLog
2001-07-01 Alexey Mahotkin <alexm(a)hsys.msk.ru>
* configure.in (option processing): with_widgets=m is ambiguous;
don't allow it.
(AC_CHECKING for the Athena widgets): Don't check for Athena when
with_widgets=no.
<5020000.994949898＠flashingchance>
./src/ChangeLog
2001-07-11 Matt Tucker <tuck(a)whistlingfish.net>
* syntax.c (find_start_of_comment):
Fix `unbalanced parentheses' bug when dealing with mixed comment
styles in c++/java/etc. mode
<15111.24585.586600.484683(a)gargle.gargle.HOWL>
./lisp/ChangeLog
2001-05-20 Martin Buchholz <martin(a)xemacs.org>
* bytecomp.el (byte-compile-arithcompare):
Only optimize (= x) ==> t if byte-compile-delete-errors is nil.
<15112.31246.180077.975624(a)gargle.gargle.HOWL>
./lisp/ChangeLog
2001-05-21 Martin Buchholz <martin(a)xemacs.org>
* byte-optimize.el (=): `=' is not a binary predicate!
Fix incorrect optimization: (= 0 0 (random 2)) ==> 0 or 1.
Use byte-optimize-predicate, not byte-optimize-binary-predicate,
<15123.9022.247285.42002(a)gargle.gargle.HOWL>
./ChangeLog
2001-05-29 Martin Buchholz <martin(a)xemacs.org>
* configure.in: Remove stray backslash.
<15127.8885.609982.512779(a)gargle.gargle.HOWL>
./tests/ChangeLog
2001-05-29 Martin Buchholz <martin(a)xemacs.org>
* automated/mule-tests.el:
Avoid test failure if (temp-directory) is a symlink.
<B72977D8.45533%abegel(a)eecs.berkeley.edu>
./src/ChangeLog
2001-05-17 Andrew Begel <abegel(a)eecs.berkeley.edu>
* lrecord.h (INIT_EXTERNAL_LRECORD_IMPLEMENTATION): Add needed cast.
<B729780C.45534%abegel(a)eecs.berkeley.edu>
./ChangeLog
2001-05-17 Andrew Begel <abegel(a)eecs.berkeley.edu>
* aclocal.m4 (LTLD): Don't use "gcc", use the ld reported by gcc.
<87u22s1ffj.fsf(a)u.sanpo.t.u-tokyo.ac.jp>
./src/ChangeLog
2001-05-11 Yoshiki Hayashi <yoshiki(a)xemacs.org>
* lisp.h (EMACS_INT_MAX): This should be one bit less since
EMACS_INT is signed.
<200104180223.WAA13090(a)mclean.mail.mindspring.net>
2001-04-17 Isaac Hollander <ysh(a)mindspring.com>
* gnuserv.c (setup_table): fix 2 signed/unsigned mismatches
* gnuslib.c (connect_to_internet_server): fix signed/unsigned mismatch
<200106220151.VAA31665(a)gwyn.tux.org>
./src/ChangeLog
2001-06-21 Ben Wing <ben(a)xemacs.org>
* dired-msw.c (Fmswindows_insert_directory):
Fix problem noted by Michael Sperber with directories containing
[] and code that destructively modifies an existing string.
<200105121126.HAA26657(a)gwyn.tux.org>
./lisp/ChangeLog
2001-05-12 Ben Wing <ben(a)xemacs.org>
* find-paths.el (paths-find-recursive-path):
fix error with null EXCLUDE-REGEXP.
* font-lock.el (font-lock-mode):
fix problem reported by hrvoje with buffers starting with a space.
<200106220151.VAA31665(a)gwyn.tux.org>
./lisp/ChangeLog
2001-06-21 Ben Wing <ben(a)xemacs.org>
* help.el (frob-help-extents): Fix problem with undefined face.
* update-elc.el (update-elc-files-to-compile): Compile in proper order.
* keydefs.el (global-map): Add a defn for M-?, previously
undefined, to access help -- in case the terminal is not set up
right, or f1 gets redefined.
MS WINDOWS - printing
<200105060838.EAA10366(a)gwyn.tux.org>
./src/ChangeLog
2001-05-06 Ben Wing <ben(a)xemacs.org>
* console-msw.h:
* device-msw.c:
* device-msw.c (print_dialog_worker):
* device-msw.c (mswindows_handle_print_dialog_box):
* device-msw.c (syms_of_device_mswindows):
* dialog-msw.c (mswindows_make_dialog_box_internal):
* general-slots.h:
implement printing the selection when it's selected.
<200105060838.EAA10366(a)gwyn.tux.org>
./lisp/ChangeLog
2001-05-06 Ben Wing <ben(a)xemacs.org>
* dialog.el (make-dialog-box):
* menubar-items.el (default-menubar):
* printer.el (generic-print-buffer):
* printer.el (generic-print-region):
implement printing the selection when it's selected.
MS WINDOWS - set_file_times
<1yp8gt63.fsf(a)rapier.ecf.teradyne.com>
./src/ChangeLog
2001-05-29 Adrian Aichner <adrian(a)xemacs.org>
* fileio.c: Include nt.h. Remove lisp_string_set_file_times()
because set_file_times() now takes Lisp_Object path, instead of
char*.
* nt.c: Include buffer.h.
* nt.c (convert_from_time_t): New.
* nt.c (mswindows_utime): New. Use utime, since SetFileTime does
not set mtime correctly.
* nt.h: Declare mswindows_utime().
* sysdep.c (set_file_times): set_file_times() now takes Lisp_Object
path, instead of char*.
* systime.h: Include <sys/utime.h> on WIN32_NATIVE.
<200106020618.CAA12498(a)gwyn.tux.org>
./src/ChangeLog
2001-06-01 Ben Wing <ben(a)xemacs.org>
* nt.c (mswindows_utime):
* nt.c (close_file_data):
fix off-by-one-indirection error.
MS WINDOWS - C-g handling
<200105311247.IAA28237(a)gwyn.tux.org>
./src/ChangeLog addition:
2001-05-30 Ben Wing <ben(a)xemacs.org>
For 21.4:
(Stephen, just take all event-msw.c patches. This includes
the "iconify" fix below.)
* event-msw.c:
* event-msw.c (mswindows_dequeue_dispatch_event):
* event-msw.c (assert):
* event-msw.c (emacs_mswindows_quit_p):
* event-msw.c (debug_mswin_messages):
* event-msw.c (debug_output_mswin_message):
* event-msw.c (vars_of_event_mswindows):
Fix yet more problems with C-g handling.
Implement debug-mswindows-events.
MS WINDOWS - miscellaneous
<15119.11444.70000.574489(a)gargle.gargle.HOWL>
./src/ChangeLog
2001-05-25 Craig Lanning <CraigL(a)Knology.net>
* s\mingw32.h:
Properly find MinGW's <process.h> inside Cygwin's restructured
include directories. Don't try to include <cygwin/version.h>
since we are dropping support for MinGW in versions of Cygwin
earlier than b21.
* nt.c:
Drop support for MinGW in versions of Cygwin before b21.
* sysdep.c:
Properly find MinGW's <process.h> inside Cygwin's restructured
include directories.
* syswindows.h:
Drop support for MinGW in versions of Cygwin before b21.
* unexcw.c:
Even though a.out.h is no longer detected by configure, allow
MinGW to use it until we figure out how to do the job with Win32.
<FENWAYUfKRfcjcMzfoV00000f4e(a)fenway.corp.soundbite.com>
./nt/ChangeLog
2001-05-28 Paul Krause <pkrause(a)soundbite.com>
* xemacs.mak (DEPEND): defined(_) isn't a reliable method for
detecting if perl is from cygwin. Check $OSNAME as a backup.
<4.3.2.7.2.20010517121830.00b0a920(a)san-francisco.beasys.com>
./netinstall/ChangeLog
2001-05-17 Andy Piper <andy(a)xemacs.org>
* sysfile.h: don't assume that file attributes are boolean
2001-05-17 Andy Piper <andy(a)xemacs.org>
* win32.h:
* win32.h (NOCOMATTRIBUTE): sync with latest cygwin version.
<62532263.3200087799(a)[148.59.233.133]>
./src/ChangeLog
2001-05-29 Mike Alexander <mta(a)arbortext.com>
* console-msw.h (FRAME_MSPRINTER_CHARHEIGHT): Unswap
FRAME_MSPRINTER_CHARWIDTH and FRAME_MSPRINTER_CHARHEIGHT
<200105121126.HAA26657(a)gwyn.tux.org>
./src/ChangeLog
2001-05-12 Ben Wing <ben(a)xemacs.org>
* event-msw.c (mswindows_dde_callback):
* event-msw.c (mswindows_wnd_proc):
eliminate cygwin warnings.
<200105231035.GAA13007(a)gwyn.tux.org>
./src/ChangeLog
2001-05-23 Ben Wing <ben(a)xemacs.org>
* device-msw.c (msprinter_init_device):
signal an error rather than crash with an unavailable network
printer (from Mike Alexander).
* event-msw.c:
* event-msw.c (struct winsock_stream):
* event-msw.c (winsock_writer):
* event-msw.c (winsock_closer):
* event-msw.c (make_winsock_stream_1):
cleanup headers. fix (hopefully) an error with data corruption
when sending to a network connection.
* fileio.c (Fexpand_file_name): Fix evil code that attempts
to handle the ~user prefix by (a) always assuming we're referencing
ourselves and not even verifying the user -- hence any file with
a tilde as its first char is invalid! (b) if there wasn't a slash
following the filename, the pointer was set *past* the end of
file and we started reading from uninitialized memory. Now we
simply treat these as files, always.
* buffer.c (common_init_complex_vars_of_buffer): comment change.
<200105231147.HAA15653(a)gwyn.tux.org>
./src/ChangeLog
2001-05-23 Ben Wing <ben(a)xemacs.org>
* event-msw.c (winsock_writer):
* event-msw.c (winsock_closer):
prior kludgy code was clobbering the buffer, thinking it was
"unused" space to store an unneeded return value; instead, use
the variable we've already got staring us in the face.
<4.3.2.7.2.20010524160839.02dfa610＠pop>
./src/ChangeLog
2001-05-24 Kirill 'Big K' Katsnelson <kkm(a)dtmx.com>
* event-msw.c (WM_DROPFILES): Use correct type for OLE characters.
GTK
<86snhme1pz.fsf(a)hel.bp.aventail.com>
./src/ChangeLog
2001-05-30 William M. Perry <wmperry(a)gnu.org>
* event-gtk.c: Check for buttons as modifier keys.
<8666eh375k.fsf(a)hel.bp.aventail.com>
./lisp/ChangeLog
2001-05-31 William M. Perry <wmperry(a)gnu.org>
* font-menu.el: Avoid arithmetic with 'nil' when using the font menu
in XEmacs/GTK.
<863d9hf0rd.fsf(a)hel.bp.aventail.com>
./lisp/ChangeLog
2001-06-03 William M. Perry <wmperry(a)gnu.org>
* dialog-gtk.el (popup-builtin-open-dialog): Don't forget to
return the filename selected by the user!
<987338379.7732.0.camel(a)loki.lausch.at>
./src/ChangeLog
2001-04-15 Michael Lausch <mla(a)1012surf.net>
* scrollbar-gtk.c (gtk_create_scrollbar_instance): Call
gtk_widget_request_size() to give scrollbar >0 width for
GTK >1.2.8.
<86ae3t378p.fsf(a)hel.bp.aventail.com>
./src/ChangeLog
2001-05-20 Drazen Kacar <dave(a)arsdigita.com>
* glyphs-gtk.c (gtk_resource_instantiate): C++ comment -> #if 0
<15189.33914.633203.591427(a)turnbull.sk.tsukuba.ac.jp>
./lisp/ChangeLog
2001-07-18 Stephen J. Turnbull <stephen(a)xemacs.org>
* specifier.el: define-specifier-tag 'gtk.
MISCELLANEOUS
<200105230148.VAA13292(a)mclean.mail.mindspring.net>
./src/ChangeLog
2001-05-22 Isaac Hollander <ysh(a)mindspring.com>
* Makefile.in.in: Use TAR macro instead of hardcoding
tar. Allows use of GNU tar instead of system tar
DOCUMENTATION
A bunch of typo fixes from the late Beta Release Engineer with no
ChangeLogs; leave them that way in memorium.
<iluk82vvtx8.fsf(a)barbar.josefsson.org>
./man/ChangeLog
2001-06-02 Simon Josefsson <simon(a)josefsson.org>
* internals/internals.texi (Techniques for XEmacs Developers): Fix
typo in ./configure line.
<87bsof9zbk.fsf(a)n-pl.org>
./ChangeLog
2001-05-27 Kazuo Oishi <oishi(a)n-pl.org>
* etc/xemacs-ja.1: Replace with new translation.
<20010529193619.2664.qmail(a)hsys.msk.ru>
./man/ChangeLog
2001-05-29 Alexey Mahotkin <alexm(a)hsys.msk.ru>
* xemacs/custom.texi: Documented keyboard shortcut.
* xemacs/mule.texi: Updated to match reality; tiny fixes.
<20010529193619.2664.qmail(a)hsys.msk.ru>
./lisp/ChangeLog
2001-05-29 Alexey Mahotkin <alexm(a)hsys.msk.ru>
* coding.el: Tiny typo fixed.
<200106081225.IAA31075(a)gwyn.tux.org>
./man/ChangeLog
2001-06-08 Ben Wing <ben(a)xemacs.org>
* xemacs-faq.texi (Top):
* xemacs-faq.texi (MS Windows):
* xemacs-faq.texi (Q6.2.1):
* xemacs-faq.texi (Q6.2.2):
* xemacs-faq.texi (Q6.2.3):
* xemacs-faq.texi (Q6.2.4):
* xemacs-faq.texi (Q6.2.5):
* xemacs-faq.texi (Q6.3.1):
* xemacs-faq.texi (Q6.3.2):
* xemacs-faq.texi (Q6.3.3):
* xemacs-faq.texi (Q6.3.4):
* xemacs-faq.texi (Q6.4.1):
* xemacs-faq.texi (Q6.4.2):
* xemacs-faq.texi (Current Events):
* xemacs-faq.texi (Q7.0.1):
* xemacs-faq.texi (Q7.0.2):
* xemacs-faq.texi (Q7.0.3):
* xemacs-faq.texi (Q7.0.4):
* xemacs-faq.texi (Q7.0.5):
* xemacs-faq.texi (Q7.0.6):
Merge in the rest of Hrvoje's Windows FAQ. Redo section 7
to update current reality and add condensed versions of
new changes for 21.1 and 21.4. (Not quite done for 21.4.)
Lots more Windows updates.
<200106101053.GAA12271(a)gwyn.tux.org>
./man/ChangeLog
2001-06-10 Ben Wing <ben(a)xemacs.org>
* xemacs-faq.texi (Q1.0.10):
* xemacs-faq.texi (Q1.0.11):
* xemacs-faq.texi (Q1.0.12):
Update sections on Windows and MacOS availability.
<8zi7hxlo.fsf(a)ecf.teradyne.com>
./man/ChangeLog
2001-07-02 Adrian Aichner <adrian(a)xemacs.org>
* lispref\windows.texi (Window Configurations): Fix typos.
* lispref\x-windows.texi (Resources): Ditto.
* internals\internals.texi (XEmacs From the Inside): Ditto.
* emodules.texi (Initialization Mode): Ditto.
* xemacs-faq.texi (Q6.3.2): Ditto.
<1ynzhwt3.fsf(a)ecf.teradyne.com>
./man/ChangeLog
2001-07-02 Adrian Aichner <adrian(a)xemacs.org>
* xemacs-faq.texi (Q1.0.14): Update information following a
request by Tom Mostyn.
./man/ChangeLog
2001-06-24 Stephen J. Turnbull <stephen(a)xemacs.org>
* xemacs/programs.texi: Add node name (Top) to Ebrowse xref.
[Long since committed; added ChangeLog]
<scm66jlgfkm.fsf(a)hookah.kazbek.ispras.ru>
./lisp/ChangeLog
2001-01-12 Nick V. Pakoulin <npak(a)ispras.ru>
* cl-macs.el (cl-upcase-arg): New function.
(cl-function-arglist): New function.
(cl-transform-lambda): Automatically add CL-style lambda list to
documentation string using functions above.
<200105170631.CAA32586(a)gwyn.tux.org>
./man/ChangeLog
2001-05-16 Ben Wing <ben(a)xemacs.org>
* xemacs-faq.texi (Top):
* xemacs-faq.texi (Installation):
* xemacs-faq.texi (Q2.1.15):
* xemacs-faq.texi (Q2.1.18):
* xemacs-faq.texi (Q2.1.19):
document how to debug X errors
<200105200118.VAA32224(a)gwyn.tux.org>
./src/ChangeLog
2001-05-19 Ben Wing <ben(a)xemacs.org>
* eval.c: doc comment about gcpro'ing in record_unwind_protect.
* process-nt.c:
* process-nt.c (vars_of_process_nt):
remove unused mswindows-quote-process-args. rec for 21.4.
<200105231035.GAA13007(a)gwyn.tux.org>
./nt/ChangeLog
2001-05-23 Ben Wing <ben(a)xemacs.org>
* xemacs.mak (installation):
call `ver' to get the exact os version and put it in the
installation; suggestion from adrian.
<200105231035.GAA13007(a)gwyn.tux.org>
./ChangeLog addition
2001-05-23 Ben Wing <ben(a)xemacs.org>
* etc\sample.init.el:
* etc\sample.init.el (Init-safe-require): New.
* etc\sample.init.el ((fboundp 'pending-delete-mode)):
* etc\sample.init.el ((eq system-type 'windows-nt)):
* etc\sample.init.el (dired):
* etc\sample.init.el ((Init-safe-require 'efs-auto)):
* etc\sample.init.el (completer):
* etc\sample.init.el (crypt):
* etc\sample.init.el (filladapt):
* etc\sample.init.el ((fboundp 'turn-on-lazy-lock)):
* etc\sample.init.el ((fboundp 'resize-minibuffer-mode)):
* etc\sample.init.el ((Init-safe-require 'scroll-in-place)):
Rewrite to be much more careful about loading features -- now
it decays gracefully even in the complete absence of packages.
Also avoid doing obnoxious things when loading efs.
<200105311247.IAA28237(a)gwyn.tux.org>
./man/ChangeLog
2001-05-30 Ben Wing <ben(a)xemacs.org>
* xemacs-faq.texi (Top):
* xemacs-faq.texi (MS Windows):
* xemacs-faq.texi (Q6.0.1):
* xemacs-faq.texi (Q6.0.2):
* xemacs-faq.texi (Q6.0.3):
* xemacs-faq.texi (Q6.0.4):
* xemacs-faq.texi (Q6.1.1):
* xemacs-faq.texi (Q6.1.4):
* xemacs-faq.texi (Q6.1.5):
* xemacs-faq.texi (Q6.1.6):
* xemacs-faq.texi (Q6.2.1):
* xemacs-faq.texi (Q6.2.2):
* xemacs-faq.texi (Q6.3.1):
* xemacs-faq.texi (Q6.3.2):
* xemacs-faq.texi (Q6.3.3):
* xemacs-faq.texi (Q6.4.1):
* xemacs-faq.texi (Current Events):
Major rewrite.
Update all MS Windows info to current.
Redo section 6.1 almost completely.
Incorporate sections 1 and 2 of Hrvoje's FAQ.
<8zhqhahu.fsf(a)ecf.teradyne.com>
./lisp/ChangeLog
2001-07-15 Adrian Aichner <adrian(a)xemacs.org>
* about.el (about-personal-info): Fix word duplication typos.
* dialog.el (make-dialog-box): Ditto.
* lisp-mode.el (lisp-indent-function): Ditto.
* package-admin.el (package-admin-delete-binary-package): Ditto.
* package-get.el (package-get-staging-dir): Ditto.
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091
_________________ _________________ _________________ _________________
What are those straight lines for? "XEmacs rules."
- 10 participants
- 16 comments
execute_rare_opcode(long * 0x0082da38, const unsigned char * 0x035beb23, int 0)
22 years, 9 months
Adrian Aichner
Adrian Aichner
This bug report will be sent to the XEmacs Development Team,
not to your local site managers!!
Please write in English, because the XEmacs maintainers do not have
translators to read other languages for them.
In XEmacs 21.4 (patch 4) "Artificial Intelligence (candidate 3)" [Lucid] (i586-pc-win32) of Mon Jul 23 2001 on D5DC120J
configured using `configure UNKNOWN'
Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:
Crash during visit of .content file.
Local variables in the file are trying to enter xml-mode.
Boom!
Adrian
Recent keystrokes:
C-x m M-p M-p RET M-p M-< M-> M-P M-< C-x C-b C-x o
C-n RET C-x 1 C-p C-p C-p C-p C-p C-p C-p C-p C-p C-p
C-p C-p C-p C-p C-p C-p C-p C-x k RET M-x r e p o SPC
r SPC e SPC SPC RET
Recent messages (most recent first):
Loading emacsbug...done
Loading emacsbug...
Making completion list...
Loading cl-macs...done
Loading cl-macs...
Loading iso-acc...done
Loading iso-acc...
Loading lazy-lock...done
Loading lazy-lock...
Loading gnus-ems...done
Call Stack:
NTDLL! 77fa018c()
execute_rare_opcode(long * 0x0082da38, const unsigned char * 0x035beb23, int 0) line 1477 + 20 bytes
execute_optimized_program(const unsigned char * 0x035beb20, int 5, long * 0x01dcedd0) line 656 + 17 bytes
funcall_compiled_function(long 31004524, int 0, long * 0x0082dd18) line 518 + 53 bytes
Ffuncall(int 1, long * 0x0082dd14) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x035bea88, int 4, long * 0x01477638) line 746 + 16 bytes
funcall_compiled_function(long 21608788, int 2, long * 0x0082dff0) line 518 + 53 bytes
Ffuncall(int 3, long * 0x0082dfec) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x0241cae8, int 6, long * 0x01475aa0) line 746 + 16 bytes
funcall_compiled_function(long 21606324, int 1, long * 0x0082e2d0) line 518 + 53 bytes
Ffuncall(int 2, long * 0x0082e2cc) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x0246dfa0, int 4, long * 0x01475b4c) line 746 + 16 bytes
funcall_compiled_function(long 21606352, int 1, long * 0x0082e5a8) line 518 + 53 bytes
Ffuncall(int 2, long * 0x0082e5a4) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x0082e780, int 2, long * 0x0146ebac) line 746 + 16 bytes
Fbyte_code(long 20652816, long 21425052, long 5) line 2405 + 38 bytes
Feval(long 21223200) line 3331 + 187 bytes
condition_case_1(long 21223152, long (long)* 0x01071b4c Feval(long), long 21223200, long (long, long)* 0x0106ee25 run_condition_case_handlers(long, long), long 20470344) line 1651 + 7 bytes
condition_case_3(long 21223200, long 20470344, long 21223152) line 1732 + 27 bytes
execute_rare_opcode(long * 0x0082ea6c, const unsigned char * 0x02462233, int 143) line 1271 + 19 bytes
execute_optimized_program(const unsigned char * 0x02462220, int 3, long * 0x014777f0) line 656 + 17 bytes
funcall_compiled_function(long 21608900, int 1, long * 0x0082ed44) line 518 + 53 bytes
Ffuncall(int 2, long * 0x0082ed40) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x024620c0, int 5, long * 0x01477c84) line 746 + 16 bytes
funcall_compiled_function(long 21609320, int 2, long * 0x0082f020) line 518 + 53 bytes
Ffuncall(int 3, long * 0x0082f01c) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x0082f1fc, int 3, long * 0x0146eae0) line 746 + 16 bytes
Fbyte_code(long 20652416, long 21424848, long 7) line 2405 + 38 bytes
Feval(long 21222732) line 3331 + 187 bytes
condition_case_1(long 21222360, long (long)* 0x01071b4c Feval(long), long 21222732, long (long, long)* 0x0106ee25 run_condition_case_handlers(long, long), long 20474472) line 1651 + 7 bytes
condition_case_3(long 21222732, long 20474472, long 21222360) line 1732 + 27 bytes
execute_rare_opcode(long * 0x0082f560, const unsigned char * 0x02455b1c, int 143) line 1271 + 19 bytes
execute_optimized_program(const unsigned char * 0x02455a40, int 6, long * 0x014780ac) line 656 + 17 bytes
funcall_compiled_function(long 21609600, int 1, long * 0x0082f848) line 518 + 53 bytes
Ffuncall(int 2, long * 0x0082f844) line 3563 + 17 bytes
execute_optimized_program(const unsigned char * 0x035b08e0, int 3, long * 0x01478178) line 746 + 16 bytes
funcall_compiled_function(long 21609628, int 2, long * 0x0082fb14) line 518 + 53 bytes
Ffuncall(int 3, long * 0x0082fb10) line 3563 + 17 bytes
Fcall_interactively(long 20441904, long 20494704, long 20494704) line 940 + 22 bytes
Fcommand_execute(long 20441904, long 20494704, long 20494704) line 2970 + 17 bytes
execute_command_event(command_builder * 0x01e2b0b8, long 36525840) line 3915 + 25 bytes
Fdispatch_event(long 36525840) line 4207 + 70 bytes
Fcommand_loop_1() line 583 + 9 bytes
command_loop_1(long 20494704) line 495
condition_case_1(long 20494344, long (long)* 0x01051306 command_loop_1(long), long 20494704, long (long, long)* 0x01050d20 cmd_error(long, long), long 20494704) line 1651 + 7 bytes
command_loop_3() line 256 + 35 bytes
command_loop_2(long 20494704) line 269
internal_catch(long 20322480, long (long)* 0x01050e70 command_loop_2(long), long 20494704, int * volatile 0x00000000) line 1317 + 7 bytes
initial_command_loop(long 20494704) line 305 + 25 bytes
STACK_TRACE_EYE_CATCHER(int 1, char * * 0x00e64138, char * * 0x00e62d60, int 0) line 2346
main(int 1, char * * 0x00e64138, char * * 0x00e62d60) line 2714
mainCRTStartup() line 338 + 17 bytes
KERNEL32! 77e97d08()
execute_rare_opcode(long * 0x0082da38, const unsigned char * 0x035beb23, int 0) line 1477 + 20 bytes
- stack_ptr 0x0082da38
20494704
- program_ptr 0x035beb23 ""
0 ''
opcode 0
execute_optimized_program(const unsigned char * 0x035beb20, int 5, long * 0x01dcedd0) line 656 + 17 bytes
n 0
opcode 0
- program 0x035beb20 ""
0 ''
stack_depth 5
- constants_data 0x01dcedd0
3
- program_ptr 0x035beb23 ""
0 ''
speccount 34
- stack_ptr 0x0082da38
20494704
- gcpro1 {...}
- next 0x0082ded8
- next 0x0082e1b8
- next 0x0082e490
- next 0x0082e760
- next 0x0082e84c
- next 0x0082e920
- next 0x0082ec2c
- next 0x0082ef08
- next 0x0082f1dc
- next 0x0082f340
- next 0x0082f414
- next 0x0082f72c
- next 0x0082fa00
- next 0x0082fbc4
- next 0x0082fbd0
- next 0x0082fd0c
- next 0x0082fd9c
- next 0x0082fda8
- next 0x0082fdd0
- next 0x00000000
next CXX0030: Error: expression cannot be evaluated
var CXX0030: Error: expression cannot be evaluated
nvars CXX0030: Error: expression cannot be evaluated
- var 0x0082fe50
20494704
nvars 1
- var 0x0082fd94
36525840
nvars 1
- var 0x0082fd98
20494704
nvars 1
- var 0x0082fd28
36525840
nvars 1
- var 0x0082fc90
20494704
nvars 1
- var 0x0082fb10
20441904
nvars 7
- var 0x0082f840
20464152
nvars 3
- var 0x0082f560
20474472
nvars 6
- var 0x0082f494
20474472
nvars 1
- var 0x0082f34c
20652416
nvars 3
- var 0x0082f01c
20440296
nvars 3
- var 0x0082ed40
20439312
nvars 5
- var 0x0082ea6c
20470344
nvars 3
- var 0x0082e9a0
20470344
nvars 1
- var 0x0082e858
20652816
nvars 3
- var 0x0082e5a4
20430600
nvars 2
- var 0x0082e2cc
20430504
nvars 4
- var 0x0082dfec
20438952
nvars 6
- var 0x0082dd14
20399112
nvars 4
- var 0x0082da38
20494704
nvars 5
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
- 1 participants
- 0 comments
Re: [PATCH] line_number_cache corruption with buffer_replace_char()
22 years, 9 months
Andy Piper
Andy Piper
Looks ok to me, but Ben should comment
andy
At 12:17 AM 7/26/01 +0100, Nix wrote:
>Synopsis: buffer_replace_char() forgets to check for \n insertions and
>deletions, potentially leaving the line-number cache out of sync with
>reality.
>
>To see it:
>
>- Visit a big buffer, like redisplay.c, with hs-minor-mode from
> hideshow.el turned on; `selective-display' will be `t'.
>
>- Invoke `hs-hide-all'.
>
>- Go to the bottom of the buffer; sensibly, the line number is low,
> because most of the `\n's in the buffer are now `^M's.
>
>- Invoke `hs-show-all'.
>
>- Watch the line number not change, because the cache is not
> invalidated, despite the transformation of all those `^M's back into
> `\n's again. If the buffer's big enough, moving backwards in units of
> less than LINE_NUMBER_FAR chars can give you a negative line number.
>
>This definitely affects subst-char-in-region (which hideshow uses),
>and might well affect translate-region and replace-match calls that
>don't change lengths too.
>
>
>I've tried to MULEize this, but this is my first exposure to MULE inside
>of XEmacs, so the patch, although tiny, may be wrong or wildly
>inefficient or not the idiomatic way to do things. If so, please fix and
>castigate :)
>
>diff -durN xemacs-orig/src/insdel.c xemacs/src/insdel.c
>--- xemacs-orig/src/insdel.c Mon Apr 16 23:33:02 2001
>+++ xemacs/src/insdel.c Wed Jul 25 23:48:44 2001
>＠＠ -2922,8 +2922,19 ＠＠
> MAP_INDIRECT_BUFFERS (buf, mbuf, bufcons)
> {
> mbuf->changes->newline_was_deleted = 1;
>+ delete_invalidate_line_number_cache (mbuf, pos, pos+1);
> }
> }
>+
>+ if (ch == '\n')
>+ {
>+ Bufbyte inserted_char[MAX_EMCHAR_LEN];
>+ Bytecount len;
>+
>+ len = set_charptr_emchar (inserted_char, ch);
>+ insert_invalidate_line_number_cache (buf, pos, inserted_char, len);
>+ }
>+
> MARK_BUFFERS_CHANGED;
> if (!not_real_change)
> {
>
>
>--
>`It's all about bossing computers around. Users have to say "please".
>Programmers get to say "do what I want NOW or the hard disk gets it".'
> -- Richard Heathfield on the nature of programming
- 1 participants
- 0 comments
hexl.el changes
22 years, 9 months
Edwin Steiner
Edwin Steiner
Hello!
I've implemented some things mentioned in the thread
"Changes to avoid (inappropriate) font-locking during hexl-find-file".
hexl-mode now behaves as I like it. You can avoid unneccessary
font-locking, revert-buffer works (without unnecc. prompting),
font-lock is switched off when entering hexl-mode and restored by
hexl-mode-exit. :-)
I would appreciate feedback, whether this is also how /you/ would like
hexl-mode to behave, or if I missed something and this changes could
break existing code.
The changes also need more testing, of course.
There are two independent (orthogonal) changes to hexl.el:
CHANGE 1:
There is a new custom. variable *hexl-find-file-raw*, which determines
if hexl-find-file uses find-file (default) or find-file-noselect with
RAWFILE non-nil. See the doc string and the comments in the patch
for further explanation.
find-file is called as (find-file filename), not as
(find-file filename 'binary) as before. See the comments in the patch
for my justification.
CHANGE 2:
A new custom. variable *hexl-switch-font-lock*, which determines if hexl
should leave font-lock alone (default) or if it should try to turn off
font-lock on entering hexl-mode and try to restore it on leaving
hexl-mode.
hexl-after-revert-hook was removed. hexl-mode now sets its own
revert-buffer-function, which calls the default revert-buffer, but
sets PRESERVE-MODES to non-nil, so the buffer stays in hexl-mode.
Restoring point after revert-buffer now works. (This change is not
entirely independent from the new font-lock handling, because the new
revert-buffer logic avoids unneccessary switching of major modes,
which also effects font-locking.)
END OF CHANGES
Here comes the patch, if you want to try it:
(Finally got rid of tab-ignorant Outlook Express :-))
-Edwin
Index: text-modes/hexl.el
===================================================================
RCS file: /usr/CVSroot/XEmacs/packages/xemacs-packages/text-modes/hexl.el,v
retrieving revision 1.5
diff -u -r1.5 hexl.el
--- text-modes/hexl.el 2001/07/20 14:09:16 1.5
+++ text-modes/hexl.el 2001/07/25 13:09:37
＠＠ -102,6 +102,30 ＠＠
:group 'hexl
:version "20.3")
+(defcustom hexl-find-file-raw nil
+ "*If non-nil `hexl-find-file' reads the specified file literally.
+Per default hexl-find-file uses `find-file' to read the specified file,
+so that `hexl-mode-exit' gets you back to the appropriate major mode and
+coding-system for the file.
+Setting this variable to non-nil speeds up hexl-find-file, because no
+major mode (apart from hexl-mode) is loaded and no font-locking is done.
+Using hexl-mode-exit in this case will get you back to a buffer in
+fundamental-mode and the binary coding-system.
+NOTE: Regardless of hexl-find-file-raw's value hexl-mode itself always
+displays the literal binary contents of a file (in hex dump format)."
+ :type 'boolean
+ :group 'hexl)
+
+(defcustom hexl-switch-font-lock nil
+ "*If non-nil, turn off font-lock-mode when entering hexl-mode.
+When hexl-mode exits, it tries to restore font-lock to its previous settings.
+If this variable is non-nil, you can set customized font-locking for
+hexl-mode in your `hexl-mode-hook'.
+Set this variable to nil if you experience problems with font-lock after
+leaving hexl-mode."
+ :type 'boolean
+ :group 'hexl)
+
(defvar hexl-max-address 0
"Maximum offset into hexl buffer.")
＠＠ -113,6 +137,9 ＠＠
(defvar hexl-mode-old-write-contents-hooks)
(defvar hexl-mode-old-require-final-newline)
(defvar hexl-mode-old-syntax-table)
+(defvar hexl-mode-old-revert-buffer-function nil)
+(defvar hexl-mode-old-font-lock-mode nil)
+(defvar hexl-mode-old-font-lock-defaults nil)
(defvar hexl-ascii-extent nil
"Extent used to highlight ASCII element corresponding to current point.")
＠＠ -243,8 +270,9 ＠＠
(setq require-final-newline nil)
;; Add hooks to rehexlify or dehexlify on various events.
- (make-local-hook 'after-revert-hook)
- (add-hook 'after-revert-hook 'hexl-after-revert-hook nil t)
+ (setq hexl-mode-old-revert-buffer-function revert-buffer-function)
+ (make-local-variable 'revert-buffer-function)
+ (setq revert-buffer-function 'hexl-revert-buffer)
(make-local-hook 'change-major-mode-hook)
(add-hook 'change-major-mode-hook 'hexl-maybe-dehexlify-buffer nil t)
＠＠ -252,10 +280,17 ＠＠
(if hexl-follow-ascii (hexl-follow-ascii 1)))
(run-hooks 'hexl-mode-hook))
-(defun hexl-after-revert-hook ()
- (setq hexl-max-address (1- (buffer-size)))
- (hexlify-buffer)
- (set-buffer-modified-p nil))
+(defun hexl-revert-buffer (ignore-auto noconfirm)
+ "hexl-mode installs this function as `revert-buffer-function'.
+This function makes sure that the major mode is not changed during
+`revert-buffer' and that the buffer is rehexlified after reverting."
+ (let ((revert-buffer-function hexl-mode-old-revert-buffer-function)
+ (opoint (point)))
+ (revert-buffer ignore-auto noconfirm t)
+ (setq hexl-max-address (1- (buffer-size)))
+ (hexlify-buffer)
+ (goto-char (min opoint (point-max)))
+ (set-buffer-modified-p nil)))
(defvar hexl-in-save-buffer nil)
＠＠ -293,7 +328,13 ＠＠
"Edit file FILENAME in hexl-mode.
Switch to a buffer visiting file FILENAME, creating one if none exists."
(interactive "fFilename: ")
- (find-file filename 'binary)
+ (if hexl-find-file-raw
+ (let ((find-file-run-dired nil))
+ (switch-to-buffer (find-file-noselect filename nil t)))
+ ;; (find-file filename 'binary) makes no sense here IMO, since it mixes
+ ;; "automatic major mode" with "forced binary coding-system".
+ ;; See hexl-find-file-raw for further info. --Edwin Steiner
+ (find-file filename))
(if (not (eq major-mode 'hexl-mode))
(hexl-mode)))
＠＠ -310,17 +351,26 ＠＠
(set-buffer-modified-p modified)
(goto-char original-point)))
- (remove-hook 'after-revert-hook 'hexl-after-revert-hook t)
(remove-hook 'change-major-mode-hook 'hexl-maybe-dehexlify-buffer t)
(remove-hook 'post-command-hook 'hexl-follow-ascii-find t)
(setq hexl-ascii-extent nil)
+ (setq revert-buffer-function hexl-mode-old-revert-buffer-function)
(setq write-contents-hooks hexl-mode-old-write-contents-hooks)
(setq require-final-newline hexl-mode-old-require-final-newline)
(setq mode-name hexl-mode-old-mode-name)
(use-local-map hexl-mode-old-local-map)
(set-syntax-table hexl-mode-old-syntax-table)
(setq major-mode hexl-mode-old-major-mode)
+
+ (if (and hexl-switch-font-lock (featurep 'font-lock))
+ ;; Restore pre hexl-mode font-lock settings
+ (progn
+ (font-lock-mode 0)
+ (setq font-lock-defaults hexl-mode-old-font-lock-defaults)
+ (if hexl-mode-old-font-lock-mode
+ (font-lock-set-defaults t))))
+
(force-mode-line-update)
(run-hooks 'hexl-mode-exit-hook))
＠＠ -613,6 +663,17 ＠＠
(or (y-or-n-p "Converting to hexl format discards undo info; ok? ")
(error "Aborted")))
(setq buffer-undo-list nil)
+ ;; Switch font-lock-mode off if we are entering hexl-mode.
+ ;; This cannot be in the function hexl-mode, because it must be done after
+ ;; the y-or-no-p above and before the call to shell-command-on-region below.
+ (if (and hexl-switch-font-lock (featurep 'font-lock))
+ (if (eq major-mode 'hexl-mode)
+ nil
+ (make-local-variable 'hexl-mode-old-font-lock-mode)
+ (setq hexl-mode-old-font-lock-mode font-lock-mode)
+ (make-local-variable 'hexl-mode-old-font-lock-defaults)
+ (setq hexl-mode-old-font-lock-defaults font-lock-defaults)
+ (font-lock-mode 0)))
;; Don't decode text in the ASCII part of `hexl' program output.
(let ((coding-system-for-read 'raw-text)
(coding-system-for-write
- 2 participants
- 3 comments
Re: Am I the only one?
22 years, 9 months
Dan Colascione
Dan Colascione
On Tue, Jul 24, 2001 at 08:50:56AM -0700, Andy Piper wrote:
> Is anyone elses user experience this bad?
>
> XEmacs 21.4.3 is totally unusable for me under RH 7.1. It locks up almost
> instantly in the selection code and worse locks up the window manager as
> well. I have tried configuring without ESD to no avail. Gnome or KDE makes
> no difference.
>
> Maybe its OpenMotif, does anyone else use this?
I use the true Motif code with XEmacs (but only for dialogs).
Worksforme.
Very stable for me, although the popup menus are broken without a
patch (which I sent before, but was never accepted for some reason)
--
The first book written on cryptology in English, by John Wilkins in
1641, remarked that `If all those useful Inventions that are liable to
abuse, should therefore be concealed, there is not any Art or Science
which might be lawfully profest'.
- 4 participants
- 6 comments