On 18:45 Fri 11 Jan , Dirk Gassen wrote: > I currently have a performance problem on a Optiplex 745/755 and the > onboard graphics card. We would like to run tvtime as a TV application. > Unfortunately, when tvtime starts X consumes >30% CPU, which results in a > somewhat sluggish system (responses to user actions are slow, closed > captioning looses characters, etc.). > > This system did already work well on an Optiplex 620GX (X CPU usage <15%). > > Here's some more info about the systems:
To simplify things, I'm just going to point out what I think are the important facts. > Optiplex 620GX (works): > > 00:02.1 Display controller: Intel Corporation 82945G/GZ Integrated Graphics > Controller (rev 02) > sys-kernel/gentoo-sources-2.6.19 > x11-base/xorg-server-1.3.0.0 > x11-drivers/xf86-video-i810-1.7.4 945G with the 1.x intel driver > Optiplex 745 (performance issues): > 00:02.1 Display controller: Intel Corporation 82Q963/Q965 Integrated > Graphics Controller (rev 02) > sys-kernel/gentoo-sources-2.6.19 > x11-base/xorg-server-1.3.0.0 > x11-drivers/xf86-video-i810-2.0.0 (with 1.7.4 there were two bars with > noise visible on the sides) 965Q with 2.0 intel driver > Optiplex 755 (also performance issues): > 00:02.1 Display controller: Intel Corporation 82Q35 Express Integrated > Graphics Controller (rev 02) > sys-kernel/gentoo-sources-2.6.23 > x11-base/xorg-server-1.3.0.0 > x11-drivers/xf86-video-i810-2.0.0 Q35 with 2.0 intel driver You've tested the new hardware with the old driver, and it didn't work for unrelated reasons. Could you test the old hardware with the new driver, to test whether it's the driver version change creating a slowdown? Also, could you try 2.2 instead of 2.0? If that doesn't work, try 2.1. It's also possible that the driver's just not as good on 965 and Q35 as on 945. > I suspect there's something wrong with the MTRR setup on the 745/755 (wild > guess): > localhost ~ # cat /proc/mtrr > reg00: base=0x00000000 ( 0MB), size=65536MB: write-back, count=1 > reg01: base=0x3d600000 ( 982MB), size= 2MB: uncachable, count=1 > reg02: base=0x3d800000 ( 984MB), size= 8MB: uncachable, count=1 > reg03: base=0x3e000000 ( 992MB), size= 32MB: uncachable, count=1 > reg04: base=0x3d500000 ( 981MB), size= 1MB: uncachable, count=1 > reg05: base=0x40000000 (1024MB), size=1024MB: uncachable, count=1 > reg06: base=0x80000000 (2048MB), size=2048MB: uncachable, count=1 > > ... and X reports in its log: > (WW) intel(0): Failed to set up write-combining range > (0xd0000000,0x10000000) Yeah, this could cause problems with speed. If it worked in the old versions with otherwise identical software configurations (e.g. kernel features), this could be a driver bug. Thanks, Donnie -- [email protected] mailing list
