[Intel-gfx] HD XvMC on i915

Steven J Newbury steve at snewbury.org.uk
Thu Oct 22 15:30:37 CEST 2009


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Nathan A. Stine wrote:
> On Tue, 2009-10-20 at 01:15 +0100, Sergio Monteiro Basto wrote:
>> On Mon, 2009-10-19 at 20:02 -0400, Nathan A. Stine wrote:
>>> On Mon, 2009-10-19 at 13:14 +0100, Sergio Monteiro Basto wrote:
>>>> Hi, 
>>>> On Sat, 2009-10-17 at 18:34 -0400, Nathan A. Stine wrote:
>>>>> Changing this to 1920x1080 gave a clean compile and would still
>>>>> properly accelerate and display 720x576 content, but not 1920x1080
>>>>> content.  The video was heavily distorted and had a green overlay.
>>>>>
>>>> I like to learn 
>>>> How you had test this ? 
>>>>
>>>> Thanks, 
>>> Sergio,
>>>
>>> All I did was apply the following patch:
>> Hi , thanks for replying , I mean how you test XvMC working .
>>  
>> >From what I understand I need to compile mplayer with Intel xvmc support
>> and run mplayer -vo xvmc -vc ffmpeg12mc , and this only works for
>> MPEG-1/2.
>> have you any other way to test XvMC ? with other play etc . 
>>
>> Thanks,
> 
> Sergio,
> 
> You may need to compile XvMC support into mplayer.  The package I use in
> Debian unstable has it compiled in already.  I simply ran:
> 
> stinerman at stine-1:~$ mplayer -vo xvmc -vc ffmpeg12mc file.mpeg
> 

Has anybody had any success getting Intel XvMC to work with the Xine
xxmc (or xvmc) plugin(s)?  MPlayer works as described above, but xine
crashes on startup.

> Well, I ran my 1080i test using a DVB card, but the theory is the same.
> I can use XvMC acceleration with standard definition DVDs.
> 
> That's the only way I know to test.
> 
You could always transcode existing media to a higher resolution for
testing, not that it would look good!
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkrgXn0ACgkQGcb56gMuC627EQCeJxj/Owkpv/DwZHdnDfdDupUu
tiEAoIww+GabswP/QPf3M5yMl4S/abmq
=zYI8
-----END PGP SIGNATURE-----



More information about the Intel-gfx mailing list