[Openchrome-users] PAL TV-output with a lot of flickering after 0.2.904 with VIA EPIA SP8000E (chipset CN400/VT1622)

[Vicente Navarro]

On 27/05/2013 14:26, Xavier Bachelot wrote:
> On 05/27/2013 02:02 PM, Xavier Bachelot wrote:
>> On 05/27/2013 11:22 AM, Vicente Navarro wrote:
>>> Hello
>>>
>>> I have a VIA EPIA SP8000E (chipset CN400/VT1622) connected to an old 
>>> PAL
>>> TV using the S-Video connector.
>>>
>> ...
>>
>>> But recently I upgraded again to Debian 7 
>>> (xserver-xorg-video-openchrome
>>> (1:0.2.906-2) ):
>>>
>>> http://packages.debian.org/wheezy/xserver-xorg-video-openchrome
>>>
>>> just to find out that the TV-out flickering problem is still there! To
>>> make things worse, the DRI driver is no longer available (as it is a
>>> DRI1-only driver) and I cannot compile 0.2.904 for Debian 7. So I 
>>> had to
>>> reinstall the system with Debian 6.
>>>
>>> So, some change between 0.2.904 and 0.2.904+svn842-2 broke the PAL
>>> TV-Out for VT1622 and it is still broken.
>>>
>>> Please find attached the Xorg logs for 0.2.904 (works file) and for the
>>> Debian 6 version (flickering).
>>>
>> Can you try with latest release 0.3.3 ?
>> If the flickering still happens with 0.3.3, I'll dig out the patch, I
>> still have it somewhere.
>>
> And here's the patch : 
> http://pkgs.fedoraproject.org/cgit/xorg-x11-drv-openchrome.git/plain/openchrome-0.2.904-fix_tvout_flickering.patch
>
> X.
>

Thanks for the information and the assistance.

I tried yesterday with 0.3.3. It compiled fine on top of Debian 6. But 
unfortunately, it still shows a slightly flickering. It is the same with 
the version distributed with Debian 7: a slightly flickering. The 
flickering with 0.2.904+svn842-2 (Debian 6) is stronger.

I tried that "openchrome-0.2.904-fix_tvout_flickering.patch" patch with 
0.3.3, but there was no improvement. The patch just seems to remove a 
few lines for VT1625, but my chip is VT1622, so I wouldn't expect any 
change for my issue.

Furthermore, I have compared the file via_vt162x.c between 0.2.904 and 
0.2.904+svn842-2 and they are identical. If I compare that file between 
0.2.904 and 0.3.3, I see minimal changes. Therefore, I wouldn't expect 
the issue to be in via_vt162x.c.

Vicente