Bug 77023 - Scrolling produces black output
Summary: Scrolling produces black output
Status: RESOLVED DUPLICATE of bug 82090
Alias: None
Product: konqueror
Classification: Applications
Component: khtml (show other bugs)
Version: 3.2
Platform: Gentoo Packages Linux
: NOR normal
Target Milestone: ---
Assignee: Konqueror Developers
URL:
Keywords:
: 122208 (view as bug list)
Depends on:
Blocks:
 
Reported: 2004-03-08 15:55 UTC by Andrew Somerville
Modified: 2008-07-25 21:37 UTC (History)
2 users (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments
This is the scrolling bug as observed in kmail (176.78 KB, image/jpeg)
2004-03-08 15:57 UTC, Andrew Somerville
Details
Konqueror showing 4 separate parts of 2 separate web pages at once. Amazing, eh? (140.39 KB, image/jpeg)
2004-03-10 22:15 UTC, Andrew Somerville
Details
Yet another amazing reproduction of the scrolling bug. Maybe this will gain it it's much desired "NEW" status. (88.31 KB, image/jpeg)
2004-03-16 05:38 UTC, Andrew Somerville
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Andrew Somerville 2004-03-08 15:55:31 UTC
Version:           3.2.0 (using KDE KDE 3.2.0)
Installed from:    Gentoo Packages
Compiler:          gcc 3.3.3 
OS:          Linux

In konqueror, often scrolling up or down produces distinct lines where the top and the bottom of the scroll pane previously rested, with black above and below the previously visable content. 

I previously assumed this bug was limited to konqueror, but today I observed it with kmail. The screenshot is attached. I have no means or reproducing this, but hopefully as time goes on I will. (Or maybe someone else will.)
Comment 1 Andrew Somerville 2004-03-08 15:57:23 UTC
Created attachment 5081 [details]
This is the scrolling bug as observed in kmail
Comment 2 Andrew Somerville 2004-03-09 21:16:10 UTC
OK. This bug can be easily reproduced in Konq by scrolling while the page is loading. Hope that helps. 
Comment 3 Whitehawk Stormchaser 2004-03-10 01:42:24 UTC
Wrong software package
Comment 4 Andrew Somerville 2004-03-10 03:30:23 UTC
Sorry, but this problem is common to all KDE apps. As I said earlier, I've for sure observed it in konq (konqueror) and kmail, but it could be present in other apps that use this widget. 
Comment 5 Andrew Somerville 2004-03-10 22:15:09 UTC
Created attachment 5135 [details]
Konqueror showing 4 separate parts of 2 separate web pages at once. Amazing, eh?
Comment 6 Andrew Somerville 2004-03-16 05:38:08 UTC
Created attachment 5234 [details]
Yet another amazing reproduction of the scrolling bug. Maybe this will gain it it's much desired "NEW" status.
Comment 7 Stephan Kulow 2004-03-16 10:53:49 UTC
as long as only you see the problem, I'm not worried - I'm quite convinced, that a software update of your video driver will fix the problem for good ;)
Comment 8 Andrew Somerville 2004-03-16 16:51:59 UTC
I'm not using video drivers. I use VNC.
Comment 9 Andrew Somerville 2004-03-23 17:38:55 UTC
> I'm quite convinced, that a software update of your video driver will fix the problem for good ;) 

Well Stephen, I think that anyone who has RENDER/GLX (no sure which) turned off is likely to see it. I suspect that the KDE development team is assuming that at least the RENDER extention is present on the host display, which is not always the case. KDE should be tested with and wihout RENDER to avoid these problems.
Comment 10 Allan Sandfeld 2005-01-08 13:24:13 UTC
I also see it often on large pages. Only black boxes though. I think the multiple content bug is a separate and possibly hardware issue.
Comment 11 Florian Hackenberger 2005-04-18 20:25:39 UTC
I can observe this problem with KDE 3.4.0 with RENDER and GLX extensions enabled, although it only appears in Konqueror.
Comment 12 Tommi Tervo 2006-02-24 15:21:56 UTC
*** Bug 122208 has been marked as a duplicate of this bug. ***
Comment 13 Christian D. 2006-02-24 15:40:23 UTC
I never experienced this with a GTK+ app. This seems to be a toolkit (Qt) problem, triggering some bug in X/video driver. I first thought that this was for sure a video driver bug, but now this seems to be very unlikely, since we all use largely different hardware/drivers. Maybe someone with deep X.org knowledge could help out.
Comment 14 Martin Steigerwald 2006-02-26 15:14:59 UTC
Hello, I had this scrolling into blackness problem with Konqueror several times on my IBM ThinkPad T23 with X.org 6.9 savage video driver. I had it only with Konqueror. I never seen this with KMail or any other KDE appliction. (Maybe it happens in KMail when viewing HTML emails via KHTML?)

It didn't happen recently. I will monitor this.

martin@deepdance:~ -> xdpyinfo
name of display:    :0.0
version number:    11.0
vendor string:    The X.Org Foundation
vendor release number:    60900000
X.Org version: 6.9.0
maximum request size:  16777212 bytes
motion buffer size:  256
bitmap unit, bit order, padding:    32, LSBFirst, 32
image byte order:    LSBFirst
number of supported pixmap formats:    7
supported pixmap formats:
    depth 1, bits_per_pixel 1, scanline_pad 32
    depth 4, bits_per_pixel 8, scanline_pad 32
    depth 8, bits_per_pixel 8, scanline_pad 32
    depth 15, bits_per_pixel 16, scanline_pad 32
    depth 16, bits_per_pixel 16, scanline_pad 32
    depth 24, bits_per_pixel 32, scanline_pad 32
    depth 32, bits_per_pixel 32, scanline_pad 32
keycode range:    minimum 8, maximum 255
focus:  window 0x2800007, revert to PointerRoot
number of extensions:    33
    BIG-REQUESTS
    Composite
    DAMAGE
    DOUBLE-BUFFER
    DPMS
    Extended-Visual-Information
    FontCache
    GLX
    LBX
    MIT-SCREEN-SAVER
    MIT-SHM
    MIT-SUNDRY-NONSTANDARD
    RANDR
    RECORD
    RENDER
    SECURITY
    SGI-GLX
    SHAPE
    SYNC
    TOG-CUP
    X-Resource
    XC-APPGROUP
    XC-MISC
    XFIXES
    XFree86-Bigfont
    XFree86-DGA
    XFree86-DRI
    XFree86-Misc
    XFree86-VidModeExtension
    XInputExtension
    XKEYBOARD
    XTEST
    XVideo
[...]
screen #0:
  print screen:    no
  dimensions:    1024x768 pixels (289x217 millimeters)
  resolution:    90x90 dots per inch
  depths (7):    24, 1, 4, 8, 15, 16, 32
[...]

Regards, Martin
Comment 15 Christian D. 2006-02-26 15:33:45 UTC
~$ xdpyinfo
name of display:    :0.0
version number:    11.0
vendor string:    The X.Org Foundation
vendor release number:    70000000
X.Org version: 7.0.0
maximum request size:  16777212 bytes
motion buffer size:  256
bitmap unit, bit order, padding:    32, LSBFirst, 32
image byte order:    LSBFirst
number of supported pixmap formats:    7
supported pixmap formats:
    depth 1, bits_per_pixel 1, scanline_pad 32
    depth 4, bits_per_pixel 8, scanline_pad 32
    depth 8, bits_per_pixel 8, scanline_pad 32
    depth 15, bits_per_pixel 16, scanline_pad 32
    depth 16, bits_per_pixel 16, scanline_pad 32
    depth 24, bits_per_pixel 32, scanline_pad 32
    depth 32, bits_per_pixel 32, scanline_pad 32
keycode range:    minimum 8, maximum 255
focus:  window 0x2200007, revert to PointerRoot
number of extensions:    30
    BIG-REQUESTS
    DAMAGE
    DOUBLE-BUFFER
    DPMS
    Extended-Visual-Information
    GLX
    MIT-SCREEN-SAVER
    MIT-SHM
    MIT-SUNDRY-NONSTANDARD
    NV-CONTROL
    NV-GLX
    RANDR
    RECORD
    RENDER
    SECURITY
    SHAPE
    SYNC
    TOG-CUP
    X-Resource
    XC-APPGROUP
    XC-MISC
    XFIXES
    XFree86-Bigfont
    XFree86-DGA
    XFree86-Misc
    XFree86-VidModeExtension
    XInputExtension
    XKEYBOARD
    XTEST
    XVideo
default screen number:    0
number of screens:    1

screen #0:
  dimensions:    1152x864 pixels (361x271 millimeters)
  resolution:    81x81 dots per inch
  depths (7):    24, 1, 4, 8, 15, 16, 32
  root window id:    0x61
  depth of root window:    24 planes
  number of colormaps:    minimum 1, maximum 1
  default colormap:    0x20
  default number of colormap cells:    256
  preallocated pixels:    black 0, white 16777215
  options:    backing-store YES, save-unders YES
  largest cursor:    32x32
  current input event mask:    0xfa4035
    KeyPressMask             ButtonPressMask          EnterWindowMask
    LeaveWindowMask          KeymapStateMask          StructureNotifyMask
    SubstructureNotifyMask   SubstructureRedirectMask FocusChangeMask
    PropertyChangeMask       ColormapChangeMask
  number of visuals:    16
  default visual id:  0x21
Comment 16 Christian D. 2006-02-26 15:37:24 UTC
Hi martin, could you please post your settings of backing-store and save-unders? For me they are:

options:    backing-store YES, save-unders YES

I suspect that it's a memory management bug.
Comment 17 Maksim Orlovich 2006-02-26 16:13:03 UTC
Backing store: on? I know it's supposed to be buggy, though I don't know the details. I suggest you turn it off, and try to see if the bug goes away. (And I'll talk to a more X-aware of our guys to ask about details)

Comment 18 Christian D. 2006-02-26 16:23:54 UTC
I still believe that's an X bug triggered by Qt somehow. Maybe backing-store off could help, but I doubt it. At least it can't harm to turn it off for the people who encounter this bug everyday, since you can't really work with that kind of a grave bug. It didn't happen here for a long time (even with bs on), so 3.5.1 really improved on this for me very much!
Comment 19 Martin Steigerwald 2006-02-26 21:40:52 UTC
Hello Christian,

I have backing store set to on in order to run AROS (http://www.aros.org) without refresh problems ;-).

martin@deepdance:~ -> xdpyinfo | grep backing
  options:    backing-store YES, save-unders YES

What's save-unders being used for?

Regards,
Martin Steigerwald
Comment 20 Christian D. 2006-02-27 10:55:19 UTC
Save-unders is nearly the same as backing-store. The difference is, that it only stores the part of a windows which gets overdrawn by popups, instead of storing a whole window with child-windows in offscreen memory.
Comment 21 Florian Hackenberger 2006-05-18 22:58:13 UTC
Never happened to me on KDE 3.5.2

Anyone still experiencing this problem? I think we should close the bug if not.
Comment 22 Damien Gombault 2006-05-18 23:33:27 UTC
I have this problem with KDE 3.5.2.
Comment 23 Martin Steigerwald 2006-05-18 23:41:11 UTC
It still happens from time to time here... I am using KDE 3.5.2 on Debian Etch/Sid/Experimental (mostly Etch tough, not much experimental). I am using modularized X.org xserver-xorg 7.0.18 together with this video driver:

xserver-xorg-video-savage 2.0.2.3-4

My driver settings are:

Section "Device"
        Identifier      "Savage"
        Driver          "savage"
        BusID           "PCI:1:0:0"
        Option          "Accel" "True"
        Option          "HWCursor" "True"
        # AGP: The chip should be able to do 4x according to IBM specs for TP23
        Option          "AGPMode" "4"
        # Powermanagement: Dont know whether it works with Savage at all
        Option          "DynamicClocks" "True"
        # For AROS ;)
        Option          "BackingStore" "True"
EndSection

I have an IBM ThinkPad T23 with this videocard:

0000:01:00.0 VGA compatible controller: S3 Inc. SuperSavage IX/C SDR (rev 05)
0000:01:00.0 0300: 5333:8c2e (rev 05)

When it happens it looks similar to attachment

Yet another amazing reproduction of the scrolling bug. Maybe this will gain it it's much desired "NEW" status.

Can provide a screenshot if required.
Comment 24 peppelorum 2006-06-16 10:51:42 UTC
http://bugs.kde.org/show_bug.cgi?id=82090 seems to be about the same thing.
Comment 25 Germain Garand 2007-01-27 20:26:24 UTC

*** This bug has been marked as a duplicate of 82090 ***