AMD's UVD2-based XvBA Finally Does Something On Linux

[RealNC]

Instead of theorizing about possible bugs, it would be more helpful to actually try the patch. I can't see any flickering here, neither with Gtk+ nor with anything else.

[bridgman]

We need both, I guess. End user experience is what matters, but posts like md1032's help to identify where to test and what to look for.

[monraaf]

Richard just finished getting flow control instructions working on 6xx/7xx shader programs, and is cleaning up that code for release.

Any eta when Richard's work will hit the freedesktop repository? And IIRC internally someone was also working on adding interrupts to the oss driver, any news on that?

[bridgman]

I expect Richard's work will go out some time tomorrow. There are a couple of open issues with it (breaking out of subroutines doesn't work, for example) but there should be a lot of useful stuff there anyways.

The review for IRQs is still ongoing but I think we're getting pretty close to release. My guess is that it will go out within a week though. The review for Richard's flow control work was pretty fast because there's not much new IP - most of the flow control info was released in the r600isa.pdf doc in mid-2008. Alex's IRQ work, on the other hand, involves some new hardware information and an additional microcode image, so it takes longer and involves more people.

[kgonzales]

You're kidding, right ? We have three developers working full time on the open source drivers, although one of them (Cooper) works on supporting server customers with OSS drivers as top priority and you generally don't see that work.

Alex, Cooper and Richard got the 6xx/7xx 3D engine working in the first place, Richard wrote most of the initial 3D driver code, then all of them worked on that code to the point where community developers were comfortable jumping in and fixing bugs themselves. Look through the commit history for mesa/src/mesa/drivers/dri/r600 to get a better idea.

Alex has been a major contributor to the kernel modesetting code in drm. Check the recent commit messages there, including Linus's complaint about how many changes were being made

Cooper has been working on video acceleration over Gallium3D.

Richard just finished getting flow control instructions working on 6xx/7xx shader programs, and is cleaning up that code for release.

Alex's work is visible all over the driver stack - drm, radeon and radeonhd drivers, and mesa, and has been the primary developer adding support for new GPUs to both radeon and radeonhd. You just need to look at the commit logs. He also wrote most of the EXA and Textured Video code used in both radeon and radeonhd.

Sorry if this sounds like a rant, I was just a bit shocked that anyone could make a comment like that after all the work Alex, Richard and Cooper have done

And this is why I own two ATI Radeon 4890s for my gaming gigs/workstations. I'm going to purchase an iMac soon, and am glad the new cards are ATI Radeon 4850s. Don't worry, I'll dual boot it with Linux. As long as the stance of AMD/ATI contiues down this path, I will keep voting my money toward AMD/ATI video hardware solutions.

[herriejr]

I have finally been able to get it all working (Ubuntu 9.04 64bit, Radeon HD4870, Catalyst 9.10) with mplayer, but am I correct that - when using the vaapi - the OSD in mplayer is not working? (i.e. displaying subtitles, elapsed time/duration, etc...)

When some OSD stuff should be displayed I get the following error:

[vo_vaapi] vaAssociateSubpicture(): unknown libva error%  2.2% 13 0

Is it already possible to have OSD with vaapi, or is this still work in progress?

[Kano]

vaapi is not the problem, xvba output via opengl is. vaapi mplayer currently works with osd with intel g4x and via vdpau-video wrapper and nvidia card.

[h5n1]

Hi,
Can I use VAAPI with catalyst 9.11 ? I tested VAAPI from repo but I have some errors.
And second question. Who tested newest flashplugin or gnash with VAAPI support? For me it doesnt work (with catalyst 9.10).
best regards

[Melcar]

You need fglrx 8.66 and higher for XvBA if I'm not mistaken.

[h5n1]

9.10 worked ok. 9.11 not ;]