Re: Poor Xorg performance in Etch
Andrei Popescu wrote:
On Sat, Aug 18, 2007 at 06:01:10PM -0400, Marty wrote:
I recently installed Etch on my new system and found X performance to be
sluggish, particularly when using Iceape. It appears as if there is little
no hardware acceleration, remeniscent of a K7 with a slow graphics card.
scrolling a web page down (but never up), or moving windows the display
sometimes freezes for about one second or more, with up to 100% CPU
Did you check with top what is using the CPU?
Xorg. To see the instantaneous usage I have to increase top's update frequency.
Incidently, I noticed that I was running a lot of plugins, so I removed flash
My hardware is a K8M800 chipset motherboard with an Athlon64 X2 5000+ and
7500. I have tested all recent kernels up to and including linux
You call that a slow card?
For 2D apps it should be very fast. Its responsiveness seems normal under
moderate graphics load, only a few percent of the CPU, but jumps to 100% under
heavy load, such as scolling up a large web page (i.e. a blog), opening a web
page, starting or stopping an X app. When I enabled on the onboard VIA graphics
I saw the same effect, but more pronounced. On a much older, slower system
which I recently upgraded to Etch, I see a similar but less pronounced effect.
grep enabled /var/log/Xorg.0.log
(==) AIGLX enabled
(II) RADEON(0): Color tiling enabled by default
(**) RADEON(0): Backing store enabled
(II) RADEON(0): Direct rendering enabled
(II) RADEON(0): Render acceleration enabled
(II) RADEON(0): Acceleration enabled
(**) RADEON(0): DPMS enabled
(==) RADEON(0): Silken mouse enabled
(==) RandR enabled
Any errors in the log?
If I enable the frame buffer (which I did since my last post):
(EE) RADEON(0): MergedFB does not work with Option UseFBDev, MergedFB mode is
I also get the following warnings:
(WW) RADEON(0): DRI init changed memory map, adjusting ...
(WW) RADEON(0): MC_FB_LOCATION was: 0x1fff0000 is: 0x1fff0000
(WW) RADEON(0): MC_AGP_LOCATION was: 0xffffffc0 is: 0xe07fe000
(WW) AIGLX: 3D driver claims to not support visual 0x23
The last one is followed by several similar lines. The framebuffer does not
seem to affect the freeze issue.