Mailinglist Archive: radeonhd (271 mails)
| < Previous | Next > |
Re: [radeonhd] radeon vs radeonhd
- From: Matthias Hopf <mhopf@xxxxxxx>
- Date: Thu, 12 Nov 2009 19:45:53 +0100
- Message-id: <20091112184553.GB21943@xxxxxxx>
On Nov 12, 09 10:00:21 -0800, Yang Zhao wrote:
I doubt the optimizations are the reason with glxgears - which produces
only a very small load regarding the command stream.
Matthias
--
Matthias Hopf <mhopf@xxxxxxx> __ __ __
Maxfeldstr. 5 / 90409 Nuernberg (_ | | (_ |__ mat@xxxxxxxxx
Phone +49-911-74053-715 __) |_| __) |__ R & D www.mshopf.de
--
To unsubscribe, e-mail: radeonhd+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: radeonhd+help@xxxxxxxxxxxx
For r5xx, both drivers have to do some context switching when
interleaving 2D and 3D operations. There's certain amount of waiting
that needs to be done when this switch happens. radeonhd's 2D accel
code is a very old fork of radeon's, and there are optimizations that
are just missing in the former.
I doubt the optimizations are the reason with glxgears - which produces
only a very small load regarding the command stream.
Matthias
--
Matthias Hopf <mhopf@xxxxxxx> __ __ __
Maxfeldstr. 5 / 90409 Nuernberg (_ | | (_ |__ mat@xxxxxxxxx
Phone +49-911-74053-715 __) |_| __) |__ R & D www.mshopf.de
--
To unsubscribe, e-mail: radeonhd+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: radeonhd+help@xxxxxxxxxxxx
| < Previous | Next > |