Created attachment 34988 [details]
screenshot of window manager doesn't work after gnom-session startup

Bug description update:
After gnome-session startup with compiz, window manager doesn't work(see attached srceenshot).

This bug is the same as bug #26162, not bug #27258. As bug #26162 described, gnome-session with compiz startup fails on all platform before bug#26814 fixed. With bug#26814 fixed, it works well on i965 platform but still fails on 945gm, aspire1 and 855gm. The behaviour is window manager doesn't work after gnome-session startup under compiz. Finally We find it is pageflip caused the problem, and with fix, it works ok on 945gm and aspire1(we didn't test 855gm that time). Now the problem is that this issue still happens on 855gm whether pagefilip enabled or disabled.

Created attachment 34989 [details]
xorg log

*** Bug 27722 has been marked as a duplicate of this bug. ***

Can you try starting compiz by hand?  How does it fail?  If it crashes, please try to collect a backtrace...

Created attachment 36058 [details]
Sreenshot of start compiz by hand

Start compiz by hand, window can't be moved, border seems disappear(see attached screeshot). X doesn't crash.

Chris like 8xx bugs. :)

Chris would like a 855gm to work with as well!

Daniel Vetter added i8xx paths for page-flipping which have been pushed to anholt/drm-intel-next:

commit 31f568a5a9ed5b2179a4c5ccb47a63c4c7be5ef7
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Aug 4 21:22:10 2010 +0200

    drm/i915: i8xx also doesn't like multiple oustanding pageflips
    
    My i855GM suffers from a 80k/s interrupt storm without this.
    So add 2nd gen to the list of things that don't like more than
    one outstanding pageflip request.
    
    Furthermore I've changed the busy loop into a ringbuffer wait.
    Busy-loops that don't check whether the chip died are simply evil.
    And performance should actually improve, because there's usually
    a decent amount of rendering queued on the gpu, hopefully rendering
    that MI_WAIT into a noop by the time it's executed.
    
    The current code holds dev->struct_mutex while executing this loop,
    hence stalling all other gem activity anyway.
    
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
    Cc: stable@kernel.org
    Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
    [anholt: resolved against conflict]
    Signed-off-by: Eric Anholt <eric@anholt.net>

commit f8485322f05aaf1a2037e40ccd5e6f111bc720c6
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Aug 4 21:22:09 2010 +0200

    drm/i915: fixup pageflip ringbuffer commands for i8xx
    
    Add a new path for 2nd gen chips that uses the commands for i81x
    chips (where public docs do exist) augmented with the plane bits
    from i915. It seems to work and doesn't result in a black screen
    like before.
    
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
    Cc: stable@kernel.org
    [anholt: resolved against conflict]
    Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
    Signed-off-by: Eric Anholt <eric@anholt.net>

I believe this is now fixed in http://cgit.freedesktop.org/~ickle/drm-intel/