Why care about indirect rendering ?
jg at freedesktop.org
Sun Sep 11 07:28:23 PDT 2005
At least on Linux these days, UNIX sockets are currently wicked fast. I
don't think this was true in the dim past.
So some actual benchmarking is in order.
On Sun, 2005-09-11 at 10:53 +1000, Dave Airlie wrote:
> > Rik Faith put quite a bit of time into a Shared Memory Transport
> > alternative to the network layer while at Precision Insight. The
> > results are committed to XFree86 on a branch (look for "smt" in the
> > branch name). As I understand it, results were not an improvement on
> > domain sockets, and I think this has been the case for most or all
> > attempts at this.
> I suspect in the GLX indirect rendering case (note not X) an SMT might
> be of use for large texture type transfers... but for the X case it
> doesn't look useful, there is a paper from Rik here
> xorg mailing list
> xorg at lists.freedesktop.org
More information about the xorg