[Openchrome-users] help getting hardware accelleration going on vx900

[Kevin Brace]

Hi Kevin,

> Date: Mon, 16 May 2016 21:19:51 -0700
> From: kevinbanjo <kevinbanjo at gmail.com>
> Cc: openchrome-users at lists.freedesktop.org
> Subject: Re: [Openchrome-users] help getting hardware accelleration
> 	going on	vx900
> Message-ID:
> 	<CAOHcZqPw1dC9D=t_zK8TkPc1rr8CU=8xdZmd+VttKvG6K53MOA at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
> 
> Okay, I've had to resort to using grep against the X modules and I found
> some of the unresolved symbols and added them to my xorg.conf
> 
> (here:  https://pastebin.sabayon.org/pastie/17763 )
> 
> Now I'm getting this message:
> 
> https://pastebin.sabayon.org/pastie/17762
> 
> Specifically:
> 
> 
>    1. [170611.011] (II) AIGLX: Screen 0 is not DRI2 capable
>    2. [170611.011] (EE) AIGLX: reverting to software rendering
> 
> 
> IIRC this was were I was at with 0.3.3
> 
> The screen scrolling within web pages isn't bad, but if you grab a
> window and drag it around you can tell its not really accelerated, and
> video (e.g. hulu, youtube, etc...) basically doesn't work at all (but
> you can hear the audio).

OpenChrome currently does not have DRI2 support nor it has 3D acceleration support, and as I have explained in the previous e-mail, my time resources are completely tied up with fixing problems in the UMS (User Mode Setting) code paths.
Since I have released OpenChrome Version 0.4.0 back in the end of March 2016, I have done many more commits to the code, but there are number of issues that I do not know how to fix it at this point.
Those issues are really holding up the release of the next version of OpenChrome (Version 0.5.0) at this point.
    Regarding VX900 chipset, I know that 2D acceleration is somewhat sluggish with VX800 chipset, and this can possibly related to your issue about performance since VX800 and VX900 chipsets are similar in terms of the hardware.
I have not touched the rendering code paths so far, and do not plan to do so for a while due to the reason I have already given in this message (i.e., fixing UMS code is the highest priority).
However, down the road, I will take a look at the issue since I have noticed that VX800 chipset's rendering performance appears sub-par, and this is noticeable when you move a window around with a mouse.