<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Jan 1, 2015 at 10:31 PM, Keith Packard <span dir="ltr"><<a href="mailto:keithp@keithp.com" target="_blank">keithp@keithp.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="">Rob Clark <<a href="mailto:robdclark@gmail.com">robdclark@gmail.com</a>> writes:<br>
<br>
> What about simply not using GL_QUADS for the normal GL paths? Is<br>
> using quads, vs tri's and a few extra vertices really going to be a<br>
> win on some other hw? If not, avoiding quads would be a big help for<br>
> freedreno too..<br>
<br>
</span>Without quads, you end up replicating a lot of vertices, which means a<br>
lot more data motion from CPU to GPU.<span class=""><br></span></blockquote><div><br></div><div>On a UMA system, that's effectively free. With a triangle fan and glMultiDrawElements, you can get it down to four, same as a quad. Though I'm not sure if mesa fully supports glMultiDrawElements.<br><br>I believe mesa supports primitive reset, so with a triangle fan and primitive reset, you can get it down to five indices per quad.<br></div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="">
>> The big performance win, though, is fixing copyarea to give the GL<br>
>> information about the area that might be damaged by the operation,<br>
>> using the scissor. Given that it's negative lines of code and not<br>
>> significant on i965, I think this is a pretty good idea. I do wonder<br>
>> if we don't want to just always leave scissoring on, like I did to<br>
>> logic op.<br>
><br>
> I'd implemented a similar thing in XA once upon a time... pretty much<br>
> must-have for tilers..<br>
<br>
</span>Do you actually want a closer bounding box on the operation? Given that<br>
we're likely to have generated a very tight bounds when computing<br>
damage, I wonder if rejiggering the API to somehow remember that for the<br>
rendering code could be done?<span class=""><font color="#888888"><br></font></span></blockquote><div><br></div><div>Yeah, a tight scissor box for rendering would really help.<br></div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class=""><font color="#888888">
--<br>
-keith<br>
</font></span><br>_______________________________________________<br>
<a href="mailto:xorg-devel@lists.x.org">xorg-devel@lists.x.org</a>: X.Org development<br>
Archives: <a href="http://lists.x.org/archives/xorg-devel" target="_blank">http://lists.x.org/archives/xorg-devel</a><br>
Info: <a href="http://lists.x.org/mailman/listinfo/xorg-devel" target="_blank">http://lists.x.org/mailman/listinfo/xorg-devel</a><br></blockquote></div><br><br clear="all"><br>-- <br><div class="gmail_signature"> Jasper<br></div>
</div></div>