Possible new display bug

Possible new display bug

[Lars Magne Ingebrigtsen]

I'm not filing a bug report, because I don't have a test case.

But recently -- the last day or so -- I'm sometimes seeing windows not
being updated, especially if they have images in them.  The symptoms are
that if you have a buffer that is a window-and-a-half long, and you're
looking at the first window-full, and then `down' to scroll down, the
portion of the window that's after the end of the buffer doesn't get
updated at all.  The pixels just remain the same.

I'm not able to reproduce it reliably, but it's happened like ten
times. 

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Possible new display bug

[Eli Zaretskii]

> From: Lars Magne Ingebrigtsen <larsi@gnus.org>
> Date: Mon, 19 Sep 2011 20:36:09 +0200
> 
> But recently -- the last day or so -- I'm sometimes seeing windows not
> being updated, especially if they have images in them.  The symptoms are
> that if you have a buffer that is a window-and-a-half long, and you're
> looking at the first window-full, and then `down' to scroll down, the
> portion of the window that's after the end of the buffer doesn't get
> updated at all.  The pixels just remain the same.

Did you ever try to "M-x redraw-display RET" in that situation?  If
not, try it next time.

Re: Possible new display bug

[Lars Magne Ingebrigtsen]

joakim@verona.se writes:

> I had some very similar issues. I had juggled some settings on the Intel
> DRI graphics driver and later forgot about it. Emacs showed the symptoms
> you describe but oddly enough no other application. Have you recently
> begun using GNOME3 or something?

I don't think I've installed anything in particular other than some
fonts lately...

I haven't seen the display glitch again the past day.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Possible new display bug

[Lars Magne Ingebrigtsen]

Eli Zaretskii <eliz@gnu.org> writes:

> Did you ever try to "M-x redraw-display RET" in that situation?

Nope.

> If not, try it next time.

Will do.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Possible new display bug

[joakim]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> I'm not filing a bug report, because I don't have a test case.
>
> But recently -- the last day or so -- I'm sometimes seeing windows not
> being updated, especially if they have images in them.  The symptoms are
> that if you have a buffer that is a window-and-a-half long, and you're
> looking at the first window-full, and then `down' to scroll down, the
> portion of the window that's after the end of the buffer doesn't get
> updated at all.  The pixels just remain the same.
>
> I'm not able to reproduce it reliably, but it's happened like ten
> times. 

I had some very similar issues. I had juggled some settings on the Intel
DRI graphics driver and later forgot about it. Emacs showed the symptoms
you describe but oddly enough no other application. Have you recently
begun using GNOME3 or something?

-- 
Joakim Verona

Re: Possible new display bug

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

>> Did you ever try to "M-x redraw-display RET" in that situation?
>
> Nope.
>
>> If not, try it next time.
>
> Will do.

It finally happened again.  `M-x redraw-display RET' fixed the issue.

So Emacs is not redrawing certain parts of the display?  It seems to be
reproducible for me right now, but it involves using Gnus to view blog
items that have images in them.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Possible new display bug

[Eli Zaretskii]

> From: Lars Magne Ingebrigtsen <larsi@gnus.org>
> Cc: emacs-devel@gnu.org
> Date: Sat, 01 Oct 2011 18:36:12 +0200
> MailScanner-NULL-Check: 1318091775.61063@wbRskF8kn+B91ur5w1lehw
> 
> It finally happened again.  `M-x redraw-display RET' fixed the issue.
> 
> So Emacs is not redrawing certain parts of the display?

It probably means Emacs is employing some redisplay optimization when
it shouldn't.

> It seems to be reproducible for me right now, but it involves using
> Gnus to view blog items that have images in them.

If you can make a reproducible test case starting from "emacs -Q", I
could try looking into it.

Re: Possible new display bug

[Lars Magne Ingebrigtsen]

Eli Zaretskii <eliz@gnu.org> writes:

> If you can make a reproducible test case starting from "emacs -Q", I
> could try looking into it.

It involves using Gnus and stuff, and seems to be dependent on
... things.  I haven't been able to create an "emacs -Q" test case,
unfortunately.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

Re: Possible new display bug

[Eli Zaretskii]

> From: Lars Magne Ingebrigtsen <larsi@gnus.org>
> Cc: emacs-devel@gnu.org
> Date: Thu, 06 Oct 2011 21:23:38 +0200
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> > If you can make a reproducible test case starting from "emacs -Q", I
> > could try looking into it.
> 
> It involves using Gnus and stuff, and seems to be dependent on
> ... things.  I haven't been able to create an "emacs -Q" test case,
> unfortunately.

Can you describe a few instances of that, preferably with screenshots?

Re: Possible new display bug

[Lars Magne Ingebrigtsen]

Eli Zaretskii <eliz@gnu.org> writes:

> Can you describe a few instances of that, preferably with screenshots?

I thought I had noted what article triggered the bug, but now I can't
find it.  Grr.

The next time it happens, I'll grab some screenshots.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/