--- Begin Message ---
- To: Jonathan Nieder <jrnieder@gmail.com>
- Subject: Re: xserver-xorg-video-radeon: 15x-slower performance regression in KMS mode for 2D operations
- From: Arno Schuring <aelschuring@hotmail.com>
- Date: Sun, 18 Mar 2012 00:46:18 +0100
- Message-id: <20120318004618.41010b7e@viper.intra.loos.site>
- In-reply-to: <[🔎] 20120316233425.GA32485@burratino>
- References: <20100529082132.3611.67275.reportbug@localhost.localdomain> <1275131136.29704.77.camel@thor.local> <20100529130253.GA2562@perso.beuc.net> <1275142906.29704.86.camel@thor.local> <20100529153352.GA2645@perso.beuc.net> <1275291813.29704.151.camel@thor.local> <1275292037.29704.153.camel@thor.local> <20100606203354.GA2641@perso.beuc.net> <20100613092917.GA3533@perso.beuc.net> <20111123130558.GA6497@elie.hsd1.il.comcast.net> <[🔎] 20120316233425.GA32485@burratino>
Hi Jonathan,
Jonathan Nieder (jrnieder@gmail.com on 2012-03-16 18:34 -0500):
> Jonathan Nieder wrote:
> >
> > Hm, how do 3.1.y kernels do?
>
> Ping? If you don't have access to this hardware or no longer are
> interested in pursuing this, that's fine, but please do let us know so
> we can plan accordingly.
The machine I used previously has died a few months ago. My current
machine has a (pcie) HD4350, which still has noticable input drag in 2D
redraw-heavy games. The profile however is completely different, so I
won't be of use for this particular bug.
For completeness sake, here are the top entries in the sysprof output
(picked a full-screen wine game for the test):
[/usr/bin/X] 0.00% 60.38%
r600_dri.so 28.85% 29.70%
-kernel- 0.00% 0.86%
ioctl 0.02% 11.95%
-kernel- 0.00% 11.93%
[Game.exe] 0.00% 24.68%
[wineserver] 0.00% 2.52%
Regards,
Arno
--- End Message ---