1 |
On 18:45 Fri 11 Jan , Dirk Gassen wrote: |
2 |
> I currently have a performance problem on a Optiplex 745/755 and the |
3 |
> onboard graphics card. We would like to run tvtime as a TV application. |
4 |
> Unfortunately, when tvtime starts X consumes >30% CPU, which results in a |
5 |
> somewhat sluggish system (responses to user actions are slow, closed |
6 |
> captioning looses characters, etc.). |
7 |
> |
8 |
> This system did already work well on an Optiplex 620GX (X CPU usage <15%). |
9 |
> |
10 |
> Here's some more info about the systems: |
11 |
|
12 |
To simplify things, I'm just going to point out what I think are the |
13 |
important facts. |
14 |
|
15 |
> Optiplex 620GX (works): |
16 |
> |
17 |
> 00:02.1 Display controller: Intel Corporation 82945G/GZ Integrated Graphics |
18 |
> Controller (rev 02) |
19 |
> sys-kernel/gentoo-sources-2.6.19 |
20 |
> x11-base/xorg-server-1.3.0.0 |
21 |
> x11-drivers/xf86-video-i810-1.7.4 |
22 |
|
23 |
945G with the 1.x intel driver |
24 |
|
25 |
> Optiplex 745 (performance issues): |
26 |
> 00:02.1 Display controller: Intel Corporation 82Q963/Q965 Integrated |
27 |
> Graphics Controller (rev 02) |
28 |
> sys-kernel/gentoo-sources-2.6.19 |
29 |
> x11-base/xorg-server-1.3.0.0 |
30 |
> x11-drivers/xf86-video-i810-2.0.0 (with 1.7.4 there were two bars with |
31 |
> noise visible on the sides) |
32 |
|
33 |
965Q with 2.0 intel driver |
34 |
|
35 |
> Optiplex 755 (also performance issues): |
36 |
> 00:02.1 Display controller: Intel Corporation 82Q35 Express Integrated |
37 |
> Graphics Controller (rev 02) |
38 |
> sys-kernel/gentoo-sources-2.6.23 |
39 |
> x11-base/xorg-server-1.3.0.0 |
40 |
> x11-drivers/xf86-video-i810-2.0.0 |
41 |
|
42 |
Q35 with 2.0 intel driver |
43 |
|
44 |
You've tested the new hardware with the old driver, and it didn't work |
45 |
for unrelated reasons. Could you test the old hardware with the new |
46 |
driver, to test whether it's the driver version change creating a |
47 |
slowdown? Also, could you try 2.2 instead of 2.0? If that doesn't work, |
48 |
try 2.1. |
49 |
|
50 |
It's also possible that the driver's just not as good on 965 and Q35 as |
51 |
on 945. |
52 |
|
53 |
> I suspect there's something wrong with the MTRR setup on the 745/755 (wild |
54 |
> guess): |
55 |
> localhost ~ # cat /proc/mtrr |
56 |
> reg00: base=0x00000000 ( 0MB), size=65536MB: write-back, count=1 |
57 |
> reg01: base=0x3d600000 ( 982MB), size= 2MB: uncachable, count=1 |
58 |
> reg02: base=0x3d800000 ( 984MB), size= 8MB: uncachable, count=1 |
59 |
> reg03: base=0x3e000000 ( 992MB), size= 32MB: uncachable, count=1 |
60 |
> reg04: base=0x3d500000 ( 981MB), size= 1MB: uncachable, count=1 |
61 |
> reg05: base=0x40000000 (1024MB), size=1024MB: uncachable, count=1 |
62 |
> reg06: base=0x80000000 (2048MB), size=2048MB: uncachable, count=1 |
63 |
> |
64 |
> ... and X reports in its log: |
65 |
> (WW) intel(0): Failed to set up write-combining range |
66 |
> (0xd0000000,0x10000000) |
67 |
|
68 |
Yeah, this could cause problems with speed. If it worked in the old |
69 |
versions with otherwise identical software configurations (e.g. kernel |
70 |
features), this could be a driver bug. |
71 |
|
72 |
Thanks, |
73 |
Donnie |
74 |
-- |
75 |
gentoo-desktop@l.g.o mailing list |