Re: XFree86 5.0 TODO

Jens Owen <[email protected]>
Newsgroups gmane.comp.xfree86.forum
Message-ID <[email protected]>
Alan Cox wrote:
>>   This is only a legitimate usage when in DGA mode.  Many of the
>>players I've seen use DGA to get the framebuffer address and then
>>use that outside of the relatively safe full-screen mode to render
>>to the root window, which is not safe at all.  
> 
> 
> Its superior to the alternative of not doing it at all.

I'll argue that preventing a second pass at copying the data is not a 
legitimate reason to keep DGA if the mechanisms provided in a non-DGA 
solution allow two copies to be done faster than the single copy by DGA.

For example some Mesa drivers provides a fast AGP download path, where 
the host application writes data to AGP space, then the driver triggers 
the graphics hardware to read it from there.

> Incidentally
> part of the problem here is that we have no "fallback" Xv for such cases
> which does DMA direct to the display and locking.
> 
> Its also at least as safe as DRI on most x86 processors

It might be as secure, but the possibility of accidental hangs is much 
higher.  The DRI coordinates all hardware access at the driver level and 
doesn't require any cooperation from the application itself.  However, 
DGA requires the application to use the interface properly or risk 
hanging the graphics engine.

-- 
                                /\
          Jens Owen            /  \/\ _
   [email protected]  /    \ \ \   Steamboat Springs, Colorado
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.