Comment 5 for bug 698120

Revision history for this message
Minh Ha Duong (haduong) wrote : Re: Dual head corruption on 2.6.37 lucid backport

It looks like memory allocation trouble in DRM with the "intel" video driver.

I booted today with the 2.6.36-1-generic-pae #7~lucind1-Ubuntu kernel.
It does not set the external screen at the correct resolution (regression compared to .32)

Usually it works otherwise, but I had the bug occur today also with 2.6.36.

I switched between VT1 (console) and VT7 (the stuck X at kdm greeter stage) a few times for experiment.

In /var/log/kern.log I had about a dozen of the following:
Jan 7 09:41:46 wall-e kernel: [ 7.128517] [drm:i915_gem_madvise_ioctl] *ERROR* Attempted i915_gem_madvise_ioctl() on a pinned object

and a few pairs of:
Jan 7 09:42:37 wall-e kernel: [ 56.828954] [drm:drm_mode_getfb] *ERROR* invalid framebuffer id

These messages also appear in dmsg, where they are followed by
[ 56.856042] Skipping EDID probe due to cached edid

In xorg.log when I switch to the VT7 there is:
(II) Open ACPI successful (/var/run/acpid.socket)
(II) AIGLX: Resuming AIGLX clients after VT switch
(II) intel(0): No memory allocations
(EE) intel(0): Couldn't create pixmap for fbcon
(II) intel(0): EDID vendor "LEN", prod id 16500
(II) intel(0): Printing DDC gathered Modelines:
(II) intel(0): Modeline "1440x900"x0.0 102.00 1440 1488 1520 1836 900 903 909 926 -hsync -vsync (55.6 kHz)
(II) intel(0): Modeline "1440x900"x0.0 85.00 1440 1488 1520 1836 900 903 909 926 -hsync -vsync (46.3 kHz)