[Openchrome-devel] kms_test results on VX800-based netbook

Xavier Bachelot xavier
Tue May 15 12:58:28 PDT 2012


Hi Sebastian.

First sorry for the late answer and thanks for testing.

On 04/30/2012 09:37 PM, Svenska wrote:
> Hi,
> 
> I have a netbook "One A110", same as Quanta IL1 and I flashed it with a
> BIOS from the "AIRIS KIRA". It is based on the VX800 chipset and runs
> Debian Wheezy, i.e. a linux kernel 3.2.0 and 0.2.904 UMS Openchrome,
> which works. The internal LCD has 800x480
> 
James owns a Quanta IL1, so this particular hardware has been tested. I
used to have one too, but unfortunately it died some time ago...

> I tested both git xf86-video-openchrome and drm-openchrome from 2012-04-28.
> 
> Debian Wheezy Kernel (3.2.0)
> ============================
> UMS xf86-video-openchrome 0.3.0 on debian wheezy kernel needs 'Option
> "SWCursor" "true"' ('enabled' is not recognized), otherwise the cursor
> is random pixels.
> Dual-Screen is somewhat broken, it works if LVDS+VGA are both at
> position 0x0, otherwise the LVDS shows garbage. The LVDS screen area is
> not accessible at all.
> "xrandr --panning" does not work. If I set up a size of 1024x480, the
> screen does not move, if I select 2048x2048, the complete screen is
> garbled.
> "xrandr --scale" doesn't work either, I only get "xrandr: Configure crtc
> 1 failed".
> 
> Kernel log (dmesg.32) and Xorg log (xorg.32) are attached.
> 
Dual screen is known not to work at the moment, and I've seen the same
behaviour as you describe on both a VX800 and a VX900 based laptops.
You should however be able to switch from the LVDS to the VGA port and
back without issue.

> Kernel from kms_branch (3.4rc2)
> ===============================
> The kernel does not do modesetting at all, I only get these errors while
> booting:
> 
> [drm:via_detect_agp] *ERROR* Failed acquiring AGP device.
> [drm:via_driver_load] *ERROR* Failed to allocate AGP
>
That's expected, you have PCIe ;-)

> UMS does not come up either, the Xorg log shows
> 
> [drm] KMS not enabled
> 
> and
> 
> (II) CHROME(0): VIAScreenInit
> 1536000 bytes of DRI2 memory allocated at 40000, handle 1
> mmap failed with error -22
> 
> Fatal server error:
> AddScreen/ScreenInit failed for driver 0
> 
> Kernel log (dmesg.34rc2) and Xorg log (xorg.34rc2) are attached.
> 
The DRM driver has a number of known bugs and is not functionnal at the
moment, even with UMS. Also, you need to enable KMS, UMS is still the
default. See http://www.openchrome.org/trac/wiki/TtmGemKms. The page
could use some update though.

Anyway, there's not much work going into the kernel at the moment, the
focus is on killing the last X driver bugs from the kms_branch. Most
notably, there is an issue with memory management that shows up very
proheminently with EXA.

> I hope this helps somewhat. If I can do more to help, just tell me. I
> don't know much about X programming, but have some C knowledge.
>
Feel free to dive in the code and work on a part that suits you. You're
very welcome to ask more specific questions on the mailing list, we'll
do our best to help.

Again, thanks for testing.

Regards,
Xavier





More information about the Openchrome-devel mailing list