[Openchrome-devel] K8M800 artifacts, trac #145, trac #164

Thomas Hellström thomas
Sun Mar 16 13:03:15 PDT 2008


Xavier Bachelot wrote:
> Thomas Hellstr?m wrote:
>> Xavier Bachelot wrote:
>>> Benno Schulenberg wrote:
>>>> Thomas Hellstr?m wrote:
>>>>> Benno Schulenberg wrote:
>>>>>> I've already asked Andrey Melentyev to try the 
>>>>>> 7d3d15e67de27f7c47859f36bb55002f0c9d52d6 commit.
>>>>
>>>> Andrey has confirmed:
>>>>
>>>> Andrey Melentyev wrote:
>>>>> I've patched 2.6.24-gentoo-r4 kernel with your patch, now I 
>>>>> believe the bug is gone. I didn't see any artifacts for a
>>>>> while, with "EnableAGPDMA" enabled
>>>>
>>>> Xavier, did you mail <stefanscheffler at gmx.net>?
>>>>
>>> No I did not had time to mail anyone and not even to test myself. 
>>> Actually, I can just test if latest drm tree works, not if it fixes
>>>  anything because I never saw the bogus behavior neither on my
>>> CLE266 nor on my KM400A. I'll let you know how it goes for me.
>>>
>>> Xavier
>>>
>> Forgive me guys, but I pushed one more drm AGPDMA change. This one 
>> should hopefully not affect stability in any way, just efficiency.
>>
>> /Thomas
>>
> drm tree from a few minutes ago doesn't break the KM400A and drm tree
> from few hours ago doesn't break the CLE266. I'll test again on the 
> CLE266 with the latest, but I don't expect it to break.
>
> Just a question, how do one enable verbose mode on the XvMC libs ?
>
> Xavier
>
Hmm,
IIRC, you need to hack the function "flushXvMCLowLevel" to print out an 
error if there is any, and make sure the video player doesn't filter the 
message out.

/Thomas






More information about the Openchrome-devel mailing list