Video bugs when Virtualbox is active

The user reports that this is definitely a 3.0.0 regression. VBox 2.2.4 worked fine for him some minutes ago.

I have the very same bug on my Debian Squeeze 64 bits, and I've been working with Frank on it.

So here is a summary of our findings: 1. It's not related to the guest, as even the VirtualBox program (the main interface) has the same colormap issue 2. It's most likely not related to Qt, as the system was exactly the same in 2.2, and the 3.0 installation only de-installed 2.2 and installed 3.0. 3. It's not hardware-related, as it works on the very same hardware, but on Ubuntu intrepid 32bits instead of Debian Squeeze 64bits

It really seems to be a colormap issue, linked with gnome because: 4. When VBox is started before gnome, or without it (if you choose Failsafe Terminal as a login session, and start VBox from this terminal), there is no problem.

It brings me back in the 1990 when we had colormap issues when starting netscape after loading some other color-intensive tools like xv, on a Sun workstation. Maybe VBox is trying to allocate too many colors, and another application (started by gnome-session) uses maybe just one different too many.

Also, the main difference with the shared colormap is that, iirc (been over 10 years so), the application that had to use a private colormap was "reversed" when INACTIVE, here it's reversed when active.

Babar, do any of the following help?

• Disabling Compiz on the host
• Disabling 3D for the guest
• Starting the guest with the command line

  export XLIB_SKIP_ARGB_VISUALS=1 VirtualBox
  

assink: I presume that these are not relevant for you.

Replying to michael:

* Disabling Compiz on the host

I gave that as a hint on #vbox, but in fact, I'm no longer using compiz. Only tried it once last year, and didn't see any point, so I trashed it.

* Disabling 3D for the guest

Again, this isn't guest-related. Starting VirtualBox (the GUI) doesn't (shouldn't) require 3D, and it still fails. And anyway, 3D *is* disabled on the guest anyway.

* Starting the guest with the command line

export XLIB_SKIP_ARGB_VISUALS=1 VirtualBox

I assume you meant:

export XLIB_SKIP_ARGB_VISUALS=1
VirtualBox

or

XLIB_SKIP_ARGB_VISUALS=1 VirtualBox

Anyway, I've tried them both, and it doesn't change anything.

I'm starting to think it's one of the gnome applets I'm running, so I will try getting rid of them one by one, see if it changes anything. Will keep you posted (maybe on IRC).

Babar, could you try another thing to see if it helps. Remove VBoxTestOGL binary from VBox istallation folder. It's launched by vbox gui regardless of your guests etc.

Replying to leonid:

Babar, could you try another thing to see if it helps. Remove VBoxTestOGL binary from VBox istallation folder. It's launched by vbox gui regardless of your guests etc.

I renamed it:

# mv /usr/lib/virtualbox/VBoxTestOGL /usr/lib/virtualbox/VBoxTestOGL.disabled

But I still have exactly the same behavior :(

I've found the following:

• if I kill my window manager (no matter which one it is, I tried sawfish, metacity, wmaker, ion3, twm), the problem goes away
• but as soon as I start it, I get the "reversed private colormap" for VirtualBox

Of course, I can't live without a window manager, so that's of no use for me.

I couldn't figure out why Ubuntu Jaunty 32 bits as host works fine, when Debian Squeeze 64 bits has the issue. Might be something in xorg?

I tried to check all configuration files I could think of, and they're identical on both hosts.

Any hint where I could look?

Babar: could you check, using xwininfo, whether VirtualBox is using a TrueColor? visual and if not try starting it with:

VirtualBox -visual TrueColor

You might also check whether your window manager is using TrueColor? or DirectColor?.

Replying to michael:

Babar: could you check, using xwininfo, whether VirtualBox is using a TrueColor? visual

xwininfo: Window id: 0x4400159 "VirtualBox"
  Visual Class: DirectColor

Michael, you're a genius! It works perfectly fine with:

VirtualBox -visual TrueColor

I'll let assink close this bug as soon as he's back from holidays, but I'm pretty confident this will solve it for him too.

I knew it had something to do with colormaps, didn't think of their new "name"!

Oups, I was too quick to say it solved it.

You're right about the cause, but I can't seem to pass -visual TrueColor? to a VM. So now I can start the configuration interface (VirtualBox) in the proper visual class, but I can't start the VM with this class :( If I start the VM from the interface, it starts it in DirectColor? :(

Any idea how I can pass this along to VBoxManage startvm or something?

Thanks!

Sorry, michael helped me on IRC, but my post here was lost.

Anyway, I can launch the VM using:

VirtualBox -visual DirectColor -startvm VM

And I can put DirectColor? or TrueColor? in there, it doesn't make any difference.

Also, I just noticed today that another tool (kdiff3) has the very same issue, so it's not related to vbox! As it's fixed on Ubuntu, maybe there is a launchpad bug item about that...

I think that we have fixed this locally. I will link to test builds here (the links will work for a week or so), so that anyone interested can give the fix a try (Babar has already confirmed on IRC that it works for him). The usual warnings about test builds apply - please only use official releases on production systems, and please only use testbuilds if they fix a specific problem that you are experiencing :) That said, here they are:

http://www.virtualbox.org/download/testcase/VirtualBox-2009-07-23-18-19-36-lin32-3.0-3.0.3-r50373.run

http://www.virtualbox.org/download/testcase/VirtualBox-2009-07-23-18-09-37-lin64-3.0-3.0.3-r50373.run

And for those interested, this is the fix that was applied:

http://www.virtualbox.org/changeset/21760 and http://www.virtualbox.org/changeset/21761

I don't think this is completely resolved.

It looks like VirtualBox itself works correctly now, but if I startup up another OpenGL client (glxgears, any KDE4 application) *after* I start up VirtualBox, the same thing happens (when that application is active, the screen goes mostly black). If I pass the --visual TrueColor? argument to any KDE applications, however, they work correctly.

Note that I can start any number of OpenGL clients before starting VirtualBox, but VirtualBox seems to be doing something that results in future applications doing something incorrectly.

BTW, this is with the radeon driver on a Radeon Mobility X1400. xdpyinfo and glxinfo -t output enclosed.

name of display:    :0.0
version number:    11.0
vendor string:    The X.Org Foundation
vendor release number:    10603000
X.Org version: 1.6.3
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 0x3400022, revert to PointerRoot
number of extensions:    28
    BIG-REQUESTS
    Composite
    DAMAGE
    DOUBLE-BUFFER
    DPMS
    DRI2
    GLX
    Generic Event Extension
    MIT-SCREEN-SAVER
    MIT-SHM
    RANDR
    RECORD
    RENDER
    SECURITY
    SGI-GLX
    SHAPE
    SYNC
    X-Resource
    XC-MISC
    XFIXES
    XFree86-DGA
    XFree86-DRI
    XFree86-VidModeExtension
    XINERAMA
    XInputExtension
    XKEYBOARD
    XTEST
    XVideo
default screen number:    0
number of screens:    1

screen #0:
  print screen:    no
  dimensions:    1920x1200 pixels (677x423 millimeters)
  resolution:    72x72 dots per inch
  depths (7):    24, 1, 4, 8, 15, 16, 32
  root window id:    0x7f
  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 NO, save-unders NO
  largest cursor:    64x64
  current input event mask:    0xfa4031
    KeyPressMask             EnterWindowMask          LeaveWindowMask          
    KeymapStateMask          StructureNotifyMask      SubstructureNotifyMask   
    SubstructureRedirectMask FocusChangeMask          PropertyChangeMask       
    ColormapChangeMask       
  number of visuals:    16
  default visual id:  0x21
  visual:
    visual id:    0x21
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x22
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x71
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x72
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x73
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x74
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x75
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x76
    class:    TrueColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x77
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x78
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x79
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x7a
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x7b
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x7c
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x7d
    class:    DirectColor
    depth:    24 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits
  visual:
    visual id:    0x60
    class:    TrueColor
    depth:    32 planes
    available colormap entries:    256 per subfield
    red, green, blue masks:    0xff0000, 0xff00, 0xff
    significant bits in color specification:    8 bits

name of display: :0.0
display: :0  screen: 0
direct rendering: Yes
server glx vendor string: SGI
server glx version string: 1.2
server glx extensions:
    GLX_ARB_multisample, GLX_EXT_import_context, GLX_EXT_texture_from_pixmap, 
    GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_MESA_copy_sub_buffer, 
    GLX_OML_swap_method, GLX_SGI_make_current_read, GLX_SGI_swap_control, 
    GLX_SGIS_multisample, GLX_SGIX_fbconfig, GLX_SGIX_visual_select_group
client glx vendor string: SGI
client glx version string: 1.4
client glx extensions:
    GLX_ARB_get_proc_address, GLX_ARB_multisample, GLX_EXT_import_context, 
    GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_MESA_allocate_memory, 
    GLX_MESA_copy_sub_buffer, GLX_MESA_swap_control, 
    GLX_MESA_swap_frame_usage, GLX_OML_swap_method, GLX_OML_sync_control, 
    GLX_SGI_make_current_read, GLX_SGI_swap_control, GLX_SGI_video_sync, 
    GLX_SGIS_multisample, GLX_SGIX_fbconfig, GLX_SGIX_pbuffer, 
    GLX_SGIX_visual_select_group, GLX_EXT_texture_from_pixmap
GLX extensions:
    GLX_ARB_get_proc_address, GLX_ARB_multisample, GLX_EXT_import_context, 
    GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_MESA_copy_sub_buffer, 
    GLX_MESA_swap_control, GLX_MESA_swap_frame_usage, GLX_OML_swap_method, 
    GLX_SGI_make_current_read, GLX_SGI_swap_control, GLX_SGI_video_sync, 
    GLX_SGIS_multisample, GLX_SGIX_fbconfig, GLX_SGIX_visual_select_group
OpenGL vendor string: DRI R300 Project
OpenGL renderer string: Mesa DRI R300 20060815  TCL
OpenGL version string: 1.4 Mesa 7.5
OpenGL extensions:
    GL_ARB_depth_texture, GL_ARB_draw_buffers, GL_ARB_fragment_program, 
    GL_ARB_imaging, GL_ARB_multisample, GL_ARB_multitexture, 
    GL_ARB_point_parameters, GL_ARB_shadow, GL_ARB_shadow_ambient, 
    GL_ARB_texture_border_clamp, GL_ARB_texture_compression, 
    GL_ARB_texture_cube_map, GL_ARB_texture_env_add, 
    GL_ARB_texture_env_combine, GL_ARB_texture_env_crossbar, 
    GL_ARB_texture_env_dot3, GL_MESAX_texture_float, 
    GL_ARB_texture_mirrored_repeat, GL_ARB_texture_rectangle, 
    GL_ARB_transpose_matrix, GL_ARB_vertex_buffer_object, 
    GL_ARB_vertex_program, GL_ARB_window_pos, GL_EXT_abgr, GL_EXT_bgra, 
    GL_EXT_blend_color, GL_EXT_blend_equation_separate, 
    GL_EXT_blend_func_separate, GL_EXT_blend_logic_op, GL_EXT_blend_minmax, 
    GL_EXT_blend_subtract, GL_EXT_compiled_vertex_array, GL_EXT_convolution, 
    GL_EXT_copy_texture, GL_EXT_draw_range_elements, GL_EXT_fog_coord, 
    GL_EXT_gpu_program_parameters, GL_EXT_histogram, GL_EXT_multi_draw_arrays, 
    GL_EXT_packed_pixels, GL_EXT_point_parameters, GL_EXT_polygon_offset, 
    GL_EXT_rescale_normal, GL_EXT_secondary_color, 
    GL_EXT_separate_specular_color, GL_EXT_shadow_funcs, 
    GL_EXT_stencil_two_side, GL_EXT_stencil_wrap, GL_EXT_subtexture, 
    GL_EXT_texture, GL_EXT_texture3D, GL_EXT_texture_edge_clamp, 
    GL_EXT_texture_env_add, GL_EXT_texture_env_combine, 
    GL_EXT_texture_env_dot3, GL_EXT_texture_filter_anisotropic, 
    GL_EXT_texture_lod_bias, GL_EXT_texture_mirror_clamp, 
    GL_EXT_texture_object, GL_EXT_texture_rectangle, GL_EXT_vertex_array, 
    GL_APPLE_packed_pixels, GL_ATI_blend_equation_separate, 
    GL_ATI_texture_env_combine3, GL_ATI_texture_mirror_once, 
    GL_ATI_separate_stencil, GL_IBM_multimode_draw_arrays, 
    GL_IBM_rasterpos_clip, GL_IBM_texture_mirrored_repeat, 
    GL_INGR_blend_func_separate, GL_MESA_pack_invert, GL_MESA_ycbcr_texture, 
    GL_MESA_window_pos, GL_NV_blend_square, GL_NV_light_max_exponent, 
    GL_NV_texture_rectangle, GL_NV_texgen_reflection, GL_NV_vertex_program, 
    GL_OES_read_format, GL_SGI_color_matrix, GL_SGI_color_table, 
    GL_SGIS_generate_mipmap, GL_SGIS_texture_border_clamp, 
    GL_SGIS_texture_edge_clamp, GL_SGIS_texture_lod, GL_SUN_multi_draw_arrays
glu version: 1.3
glu extensions:
    GLU_EXT_nurbs_tessellator, GLU_EXT_object_space_tess

Vis  Vis   Visual Trans  buff lev render DB ste  r   g   b   a  aux dep ste  accum buffers  MS   MS
 ID Depth   Type  parent size el   type     reo sz  sz  sz  sz  buf th  ncl  r   g   b   a  num bufs
----------------------------------------------------------------------------------------------------
0x21 24 TrueColor    0     32  0  rgba   1   0   8   8   8   8   0   24  8   0   0   0   0   0   0
0x22 24 DirectColor  0     32  0  rgba   1   0   8   8   8   8   0   24  8   0   0   0   0   0   0
0x71 24 TrueColor    0     32  0  rgba   0   0   8   8   8   8   0   24  0   0   0   0   0   0   0
0x72 24 TrueColor    0     32  0  rgba   0   0   8   8   8   8   0   24  0  16  16  16  16   0   0
0x73 24 TrueColor    0     32  0  rgba   1   0   8   8   8   8   0   24  0  16  16  16  16   0   0
0x74 24 TrueColor    0     32  0  rgba   0   0   8   8   8   8   0   24  8   0   0   0   0   0   0
0x75 24 TrueColor    0     32  0  rgba   0   0   8   8   8   8   0   24  8  16  16  16  16   0   0
0x76 24 TrueColor    0     32  0  rgba   1   0   8   8   8   8   0   24  8  16  16  16  16   0   0
0x77 24 DirectColor  0     32  0  rgba   0   0   8   8   8   8   0   24  0   0   0   0   0   0   0
0x78 24 DirectColor  0     32  0  rgba   0   0   8   8   8   8   0   24  0  16  16  16  16   0   0
0x79 24 DirectColor  0     32  0  rgba   1   0   8   8   8   8   0   24  0   0   0   0   0   0   0
0x7a 24 DirectColor  0     32  0  rgba   1   0   8   8   8   8   0   24  0  16  16  16  16   0   0
0x7b 24 DirectColor  0     32  0  rgba   0   0   8   8   8   8   0   24  8   0   0   0   0   0   0
0x7c 24 DirectColor  0     32  0  rgba   0   0   8   8   8   8   0   24  8  16  16  16  16   0   0
0x7d 24 DirectColor  0     32  0  rgba   1   0   8   8   8   8   0   24  8  16  16  16  16   0   0
0x60 32 TrueColor    0     32  0  rgba   1   0   8   8   8   8   0   24  0   0   0   0   0   0   0

Is it enough to start the VirtualBox machine selector without starting any VMs to trigger this?

And if you start up VirtualBox using --visual TrueColor?, does that make a difference?

1) Starting up VirtualBox without starting any VM is sufficient to trigger this.

2) The --visual TrueColor? option still triggers this.

Oh, and if I kill off my window manager, it doesn't happen. However, any window manager I use (twm, even) does cause this problem (even if I restart it after the fact).

The window manager is responsible for swapping colourmaps, so that is probably the reason for what you mention in your last comment. Does changing the Qt style (qtconfig here, not sure about systems running KDE) affect this problem? A bit of googling suggests to me that this problem may be common to (nearly?) all pure Qt applications running at the same time as KDE applications.

Running qtconfig (which I believe is a pure Qt app) followed by kmines does not trigger the problem.

I meant using qtconfig to change the Qt theme in use...

Very interesting.

I haven't restarted my X server yet, but if I change styles with qtconfig, the next time I run qtconfig the qtconfig window itself (with the new style) is fine. However, the decorations and the rest of the screen are *not* correct.

If I then switch back to the original style, the qtconfig window itself is bad, along with everything else.

rlk, which Linux distribution is that? And you said you have a Radeon X1400. I wonder why this graphics card is handled by the R300 DRI driver.

This is OpenSUSE 11.1, but with newer packages from their repositories. I had also tried radeonhd, although maybe not with VB 3.0.4, but the radeon driver seems to give better performance. These are the specific X11 packages I'm using:

xorg-x11-7.4-44.1 xorg-x11-fonts-7.4-4.13 xorg-x11-driver-video-unichrome-20081229-8.8 xorg-x11-libXrender-7.4-3.17 xorg-x11-libX11-32bit-7.4-17.8 xorg-x11-libxcb-7.4-16.5 xorg-x11-libXau-7.4-4.12 xorg-x11-libXprintUtil-7.4-3.20 xorg-x11-libXfixes-devel-7.4-3.16 libqt4-x11-4.5.2-57.1 xorg-x11-libXmu-devel-7.4-4.17 xorg-x11-libxcb-32bit-7.4-16.5 xorg-x11-libxkbfile-32bit-7.4-3.17 xorg-x11-Xvnc-7.4-67.4 xorg-x11-libXt-devel-7.4-5.10 xorg-x11-libXv-7.4-3.17 xorg-x11-libxcb-devel-7.4-16.5 xorg-x11-libSM-devel-7.4-6.2 xorg-x11-fonts-core-7.4-4.13 xorg-x11-util-devel-7.4-1.31 xorg-x11-xtrans-devel-7.4-6.1 xorg-x11-libICE-7.4-6.13 xorg-x11-libxkbfile-7.4-3.17 xorg-x11-libXrender-devel-7.4-3.17 xorg-x11-fonts-devel-7.4-4.2 xorg-x11-libXau-32bit-7.4-4.12 xorg-x11-libs-32bit-7.4-41.8 xorg-x11-driver-input-7.4-48.1 xorg-x11-libX11-ccache-7.4-1.133 xorg-x11-libXt-7.4-5.10 xorg-x11-libXpm-devel-7.4-3.18 x11-input-wacom-tools-0.8.3-5.10 xorg-x11-libXfixes-7.4-3.16 xorg-x11-libXp-devel-7.4-3.16 xorg-x11-libXext-32bit-7.4-8.3 x11-tools-0.1-167.7 xorg-x11-libXdmcp-7.4-3.12 xorg-x11-libX11-devel-7.4-17.8 xorg-x11-libSM-32bit-7.4-6.2 xorg-x11-libXv-32bit-7.4-3.17 xorg-x11-xauth-7.4-44.1 xorg-x11-libXext-7.4-8.3 xorg-x11-libfontenc-7.4-3.12 xorg-x11-libfontenc-devel-7.4-3.12 xorg-x11-doc-7.4-1.44 x11-input-wacom-0.8.3-5.10 xorg-x11-libSM-7.4-6.2 xorg-x11-libICE-devel-7.4-6.13 xorg-x11-libs-7.4-41.8 xorg-x11-libxkbfile-devel-7.4-3.17 xorg-x11-server-7.4-67.4 xorg-x11-libXpm-7.4-3.18 xorg-x11-libXv-devel-7.4-3.17 xorg-x11-devel-7.4-41.8 xorg-x11-libICE-32bit-7.4-6.13 xorg-x11-libXmu-32bit-7.4-4.17 xorg-x11-libXau-devel-7.4-4.12 xorg-x11-libXp-7.4-3.16 xorg-x11-libXprintUtil-devel-7.4-3.20 xorg-x11-libfontenc-32bit-7.4-3.12 xorg-x11-server-sdk-7.4-67.4 xorg-x11-libXext-devel-7.4-8.3 xorg-x11-libXfixes-32bit-7.4-3.16 xorg-x11-libXmu-7.4-4.17 xorg-x11-libXpm-32bit-7.4-3.18 xorg-x11-libXt-32bit-7.4-5.10 xorg-x11-driver-video-radeonhd-1.2.5_20090506_4be5f71-7.52 xorg-x11-libXdmcp-devel-7.4-3.12 xorg-x11-libXp-32bit-7.4-3.16 xorg-x11-libXdmcp-32bit-7.4-3.12 xorg-x11-proto-devel-7.4-28.1 xorg-x11-libXprintUtil-32bit-7.4-3.20 xorg-x11-driver-video-7.4-120.1 xorg-x11-libX11-7.4-17.8 xorg-x11-libXrender-32bit-7.4-3.17 xorg-x11-server-extra-7.4-67.4

Out of interest, could you give the RadeonHD driver a go, to see if this issue is present with that too? The R300 driver is still considered unstable according to the DRI wiki, so this might be a driver problem.

Exactly the same problem with radeonhd.

I'm running a Slackware 13 host (qt-4.5.1), VirtualBox 3.0.4 and I have the same problem. If I run "dolphin" my screen gets colormap reversed. But if I run "dolphin --cmap" everything looks ok (this installs a private color map).

Is --cmap a generic Qt parameter? Does it work on VirtualBox too?

it's a qt parameter, it's not working on VirtualBox.
Try: dolphin --help-qt (or konqueror, kspread... any qt app you choose), you'll see this parameter.

Hallo! I have this too!

I am also using the radeon driver and SuSE 11.1 ... (32 bit). VirtualBox 3.0.4 r50677. And Xorg repository. X.Org X Server 1.6.3.901 (1.6.4 RC 1) ?. In the previous release before 0.4 of Virtualbox there has been that issue where all VM's showed up like their color was inverted to negative. Then in the even more prior version of... there was another color problem ... Restarting SuSE Desktop helped to get to normal. About in the time of the last mentioned version there was a patch for Xorg and it went better after this one. (Maybe there was a similar problem with another tool?) Let me show you: (Description of the fix): .. in packet: xorg-x11-driver-video

Di 02 Jun 2009 14:00:00 CEST sndirsch@suse.de - xf86-video-vmware 10.16.6

• various fixes

This packet contains the driver "radeon" as far as i know.?

Then there was the second mentioned Virtualbox Version. This worked nearly the same way. ... some rebooting after usage i remember in another patch but i cant find it anymore... And with the latest version. There was some fixing for Video in Virtualbox as it looks`?? This version shows the Virtual machine normally but when closing VM's and Virtualbox. For example opening a new Konsole messes up the whole desktop (Color like inverted) until the Konsole gets closed or minized again even leaving that window with a mouse would work. But the messing seems to happen for all new windows after Virtualbox was used. You differed that in more detail already. In this case restarting (KDE in my case) brings back normal Desktop colors. I did not try your suggestions. But maybe the information about this VMWARE thing might be interesting for your search??

"radeon" is quite ok vor X1xxx series cards. Radeon HD is in my point more interesting for cards which have a additional Audio onboard. You can have a look at this: http://wiki.x.org/wiki/RadeonFeature

Is there anything that can be done about the Color thing? I got a X1950 XTX. I am on KDE 4.3 with the Xorg Repo.

Many Thanks! And i hope this text was not too confusing. And of help. Basically i want to pop in and tell you about that VMWARE case and say that i have the same Problem and hope that you could fix this. And sorry for the load of text but my english is not good enough for a more precise description.

Greetings.

R

revealed: not quite sure whether I understood you correctly. Am I correct in thinking that with all updates to everything, you still have the issue with colour being messed up in some windows while VirtualBox is running? And were you saying that VMWare had similar problems?

Replying to michael: Hello! No i wanted to say that ... i think that this VMWare update might have fixed a part of that issue in past, because it went better at this time in combination with a new Virtualbox version at this (past) time. The color within Virtualbox is not being messed up with the latest version. But the desktop around Virtualbox (new opened Windows like e.g "Konsole")after Vritualbox VM usage. (After closing all Virtualbox Windows and Machines). Already opened windows before usage are not affected... only new ones. I wish i could help... but how?

Thank you.

Replying to michael: Sorry i have to admit... i havent tried since the latest few updates.... With "todays version (KDE 4.3 factory Xorg Repository and all that stuff). It did not happen. But i dont know if it is fixed definitiveley? Will have a look at it and tell you :) I have a Debian guest (no 3d due to kernel) and a XP guest with 3D enabled.

I have the same issue with an update to the latest KDE 4.3.0. It started when I updated to the latest Qt last week. Prior to this update, everything was fine.

Fedora 11 qt-4.5.2-13.fc11.i586

Oh yeah. Another comment. This does not happen until I start MS Outlook 2007 under Windows XP. It does not matter if 3D is enabled or not.

Replying to [comment:36 ]:

Hello! I said that i would report back... so thats it. It seems that the issue got lost with: 3.0.6r52128 for me. Thanks!

Greetings and good luck!

I upgraded VB to 3.0.6r52128 version, but I get the reverse video effect on every window after opening any other qt app. :-(

This appears to have cleared up for me with 3.0.6.

rlk: that is rather strange, as I'm not aware of any changes between 3.0.4 and 3.0.6 that might have fixed this - since we can't actually reproduce this ourselves, and aren't quite sure what causes it, it is hard for us to work on.

It's also possible that I've upgraded my Qt4/KDE4 installation, which I do fairly often. It appears that I last upgraded that on September 3 (to KDE 4.3.1 and Qt 4.5.2). I've also installed newer versions of the X server (I tend to stay somewhat on the bleeding edge with some of the OpenSUSE stuff). So unfortunately I don't have a clean trail here.

I guess this one's going to keep floating along until someone has a really clear test case for it.

OK folks, I've downgraded from VB 3.0.6 to 2.2.4 and all is working fine to me. Yes, I know this is the easy way, but may be this is a clue for someone.

As sollericos I've downgraded from VB 3.0.6 to 2.2.4.

My procedures were:

-Run VB 3.0.6
-Install VB 2.2.4
-Run VB 2.2.4
Result: VB was in reverse mode, as all other programs used to be.

Then:

-Restart Computer
-VB 2.2.4 (OK)
-Start winxp with 3D Enabled
-Dolphin, etc (All Reverse mode)

Then:
-Restart Computer
-VB 2.2.4 (OK)
-Start winxp with 3D Disabled
-Dolphin, etc (All OK)

I Hope this information can contribute tosolve the problem.

I can also confirm this bug using VirtualBox 3.0.6.

I am also using the radeon x.org driver.

With a ATI Technologies Inc RV280 [Radeon 9200 PRO] card.

One way to stop this happening for me is to change the DefaultDepth? option in my xorg.conf file from 24 to 16.

Hope this helps someone.

I have this bug on Debian with libqt4 v. 4.5.2 and VirtualBox v. 3.0.8 and now 3.0.10.

I've found this workaround: http://forums.virtualbox.org/viewtopic.php?f=11&t=9534

Basically the idea is: if you can't prevent vbox to load a private color map, remove it after with the command:

xstdcmap -delete all