bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[Chong Yidong]

Thanks for the backtrace.

Could you go to the third breakpoint, i.e.

  Fset_frame_size (frame=147078268, cols=896, rows=384)

and check the following variables?

n
n
n
n (until f is assigned)
p f->new_text_lines
p f->new_text_cols

Also, could you send me the minimal .emacs and .Xresources required to
reproduce this bug for you?

Thanks.

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[Tim Van Holder]

[-- Attachment #1: Type: text/plain, Size: 1587 bytes --]

On Wed, Oct 1, 2008 at 2:01 AM, Chong Yidong <cyd@stupidchicken.com> wrote:
> Thanks for the backtrace.
>
> Could you go to the third breakpoint, i.e.
>
>  Fset_frame_size (frame=147078268, cols=896, rows=384)
>
> and check the following variables?
>
> n
> n
> n
> n (until f is assigned)
> p f->new_text_lines
> p f->new_text_cols
>
> Also, could you send me the minimal .emacs and .Xresources required to
> reproduce this bug for you?
>
> Thanks.

gdb log attached.
Note: 'f' was not available inside Fset_frame_size (due to
optimization perhaps),
so I got the info inside x_set_window_size() instead.

My recipe:
- set aside .emacs (I have no .Xresources)
- run emacs (-> gets me an 80x37 frame)
- run emacs --geometry 132x60 ( gets me an 132x57 frame)
  => these are explained by menu/toolbars, if I disable those, X reports the
       window as 132x60; while this bothers me a little, if it just means I have
       to ask for 132x63, that would be fine I guess
- customize-face default, set font to terminus (foundry xos4), and height to 135
- run emacs --geometry 132x60
  -> erratic results

Checking now, it looks like it's the terminus font causing the current troubles,
a few other fonts I've checked (e.g. adobe courier) seem to end up at 132x57
reliably.
This puzzles me, as I am sure I originally reproduced it by changing only the
font height and keeping the font itself unchanged.
I'm also getting max-specpdl-size exceeded on C-x 5 2, so I cannot check the
previously observed behaviour that even when the initial frame is erratically
sized, subsequent frames are fine.

[-- Attachment #2: emacs598-gdb2.txt --]
[-- Type: text/plain, Size: 3852 bytes --]

tim@leeloo:~/gnu/build/linux/emacs/src$ gdb emacs
GNU gdb 6.4.90-debian
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i486-linux-gnu"...Using host libthread_db library "/lib/libthread_db.so.1".

DISPLAY = leeloo.anubex.internal:5.0
TERM = xterm
Breakpoint 1 at 0x811cc16: file /home/tim/gnu/src/emacs/src/emacs.c, line 420.
Breakpoint 2 at 0x8137f49: file /home/tim/gnu/src/emacs/src/sysdep.c, line 1133.
(gdb) break Fset_frame_size
Breakpoint 3 at 0x8061896: file /home/tim/gnu/src/emacs/src/frame.c, line 2742.
(gdb) break x_set_window_size
Breakpoint 4 at 0x80df4f6: file /home/tim/gnu/src/emacs/src/xterm.c, line 8847.
(gdb) disable 4
(gdb) commands 3
Type commands for when breakpoint 3 is hit, one per line.
End with a line saying just "end".
>enable 4
>cont
>end
(gdb) commands 4
Type commands for when breakpoint 4 is hit, one per line.
End with a line saying just "end".
>disable 4
>p f->new_text_lines
>p f->new_text_cols
>cont
>end
(gdb) r
The program being debugged has been started already.
Start it from the beginning? (y or n) y
Starting program: /home/tim/gnu/build/linux/emacs/src/emacs
[Thread debugging using libthread_db enabled]
[New Thread 16384 (LWP 31757)]
[Switching to Thread 16384 (LWP 31757)]

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$3 = 0
$4 = 0

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$5 = 60
$6 = 132

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$7 = 0
$8 = 0

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$9 = 0
$10 = 0

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$11 = 0
$12 = 0

Breakpoint 3, Fset_frame_size (frame=147028492, cols=1056, rows=480) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=132, rows=60) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$13 = 60
$14 = 132

Breakpoint 3, Fset_frame_size (frame=147028492, cols=896, rows=384) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=112, rows=48) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$15 = 0
$16 = 0

Breakpoint 3, Fset_frame_size (frame=147028492, cols=896, rows=384) at /home/tim/gnu/src/emacs/src/frame.c:2742
2742    {

Breakpoint 4, x_set_window_size (f=0x8c37a08, change_gravity=1, cols=112, rows=48) at /home/tim/gnu/src/emacs/src/xterm.c:8847
8847      BLOCK_INPUT;
$17 = 0
$18 = 0

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[martin rudalics]

 > I'm also getting max-specpdl-size exceeded on C-x 5 2, so I cannot check the
 > previously observed behaviour that even when the initial frame is erratically
 > sized, subsequent frames are fine.

That's most likely bug#911 (aka nine-eleven) see

http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=911

Could you try the recipe Chong gave there?

martin

bug#911: bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[Tim Van Holder]

On Wed, Oct 1, 2008 at 10:47 AM, martin rudalics <rudalics@gmx.at> wrote:
>> I'm also getting max-specpdl-size exceeded on C-x 5 2, so I cannot check
>> the
>> previously observed behaviour that even when the initial frame is
>> erratically
>> sized, subsequent frames are fine.
>
> That's most likely bug#911 (aka nine-eleven) see
>
> http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=911
>
> Could you try the recipe Chong gave there?

I tried a slightly different recipe (below) and that made the error go away.
I assume Chong's will work as well - I just wanted to set the inhibit
flag as early as possible (and I wasn't 100% sure whether or not the let
would use the defvar'd flag instead of its own scoped var).

--- faces.el.~1.425.~	2008-09-25 10:49:34.000000000 +0200
+++ faces.el	2008-10-01 11:48:01.000000000 +0200
@@ -1839,10 +1839,13 @@ variable with `setq'; this won't have th
 (declare-function x-get-resource "frame.c"
 		  (attribute class &optional component subclass))

+(defvar inhibit-frame-set-background-mode nil)
 (defun frame-set-background-mode (frame)
   "Set up display-dependent faces on FRAME.
 Display-dependent faces are those which have different definitions
 according to the `background-mode' and `display-type' frame parameters."
+  (unless inhibit-frame-set-background-mode
+    (setq inhibit-frame-set-background-mode t)
   (let* ((bg-resource
 	  (and (window-system frame)
 	       (x-get-resource "backgroundMode" "BackgroundMode")))
@@ -1914,7 +1917,8 @@ according to the `background-mode' and `
 	;; parameters, unless they have been locally modified.
 	(dolist (face (face-list))
 	  (unless (memq face locally-modified-faces)
-	    (face-spec-recalc face frame)))))))
+	    (face-spec-recalc face frame))))))
+      (setq inhibit-frame-set-background-mode nil)))

 \f
 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[martin rudalics]

> I tried a slightly different recipe (below) and that made the error go away.

I suppose the original problem (that of erratic size changes) persists?

martin

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[Tim Van Holder]

(removed bug #911 from cc: list, as it's no longer about that issue)

On Wed, Oct 1, 2008 at 3:05 PM, martin rudalics <rudalics@gmx.at> wrote:
>> I tried a slightly different recipe (below) and that made the error go
>> away.
>
> I suppose the original problem (that of erratic size changes) persists?

Yes, it does. It seems to be font-specific though - one font in particular shows
the problem for me, others don't.

bug#598: 23.0.60; frame size issue (--geometry or default-frame-alist) when using customized default face

[Tim Van Holder]

On Wed, Oct 1, 2008 at 10:12 AM, Tim Van Holder <tim.vanholder@gmail.com> wrote:
> Checking now, it looks like it's the terminus font causing the current troubles,
> a few other fonts I've checked (e.g. adobe courier) seem to end up at 132x57
> reliably.

Turns out this is not the case - emacs started up in a 168x61 frame
this morning, with
adobe-courier set as font. So it just seems to show up much less often.
Given that I'm on a system that I can't update (all new packages
require kernel 2.6,
which doesn't support the system's RAID controller), could it be that
an outdated
gtk and/or X is causing the problem?