Bug 639062 - glxgears framerate is not constant on Intel 945 (8086:27a2)
Summary: glxgears framerate is not constant on Intel 945 (8086:27a2)
Keywords:
Status: CLOSED DUPLICATE of bug 582861
Alias: None
Product: Fedora
Classification: Fedora
Component: xorg-x11-drv-intel
Version: 14
Hardware: All
OS: Linux
low
medium
Target Milestone: ---
Assignee: Adam Jackson
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2010-09-30 18:07 UTC by John Reiser
Modified: 2018-04-11 08:21 UTC (History)
5 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2010-10-07 03:14:49 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
/var/log/Xorg.0.log (192.13 KB, text/plain)
2010-10-01 15:37 UTC, John Reiser
no flags Details
output of dmesg (122.51 KB, text/plain)
2010-10-01 15:38 UTC, John Reiser
no flags Details
/var/log/messages (146.60 KB, text/plain)
2010-10-01 15:43 UTC, John Reiser
no flags Details

Description John Reiser 2010-09-30 18:07:37 UTC
Description of problem: The observed framerate while running 'glxgears' is not constant.  It looks a little jerky, sometimes stutters, and sometimes appears to run backwards, even on an "idle" system.


Version-Release number of selected component (if applicable):
xorg-x11-drv-intel-2.12.0-6.fc14.i686


How reproducible: every time


Steps to Reproduce:
1.https://fedoraproject.org/wiki/QA:Testcase_intelvideo_glx using i686 LiveCD http://adamwill.fedorapeople.org/graphics_test_week_201009/graphics_test_20100928-i686.iso
2.
3.
  
Actual results:
$ glxgears
Running synchronized to the vertical refresh.  The framerate should be
approximately the same as the monitor refresh rate.
208 frames in 5.0 seconds = 41.599 FPS
269 frames in 5.0 seconds = 53.661 FPS
169 frames in 5.0 seconds = 33.638 FPS
166 frames in 5.0 seconds = 33.144 FPS
191 frames in 5.0 seconds = 37.900 FPS
212 frames in 5.0 seconds = 42.395 FPS
162 frames in 5.0 seconds = 32.391 FPS
216 frames in 5.0 seconds = 43.147 FPS
292 frames in 5.0 seconds = 58.292 FPS
287 frames in 5.0 seconds = 57.312 FPS
196 frames in 5.0 seconds = 39.130 FPS
159 frames in 5.0 seconds = 31.666 FPS
186 frames in 5.0 seconds = 36.978 FPS
161 frames in 5.0 seconds = 32.144 FPS
121 frames in 5.0 seconds = 24.161 FPS
141 frames in 5.0 seconds = 27.945 FPS
212 frames in 5.0 seconds = 42.338 FPS
236 frames in 5.0 seconds = 47.067 FPS
245 frames in 5.0 seconds = 48.815 FPS
276 frames in 5.0 seconds = 55.044 FPS
159 frames in 5.0 seconds = 31.493 FPS
148 frames in 5.0 seconds = 29.521 FPS
148 frames in 5.1 seconds = 29.289 FPS
151 frames in 5.0 seconds = 30.090 FPS
150 frames in 5.0 seconds = 29.962 FPS
142 frames in 5.1 seconds = 28.020 FPS
145 frames in 5.0 seconds = 28.914 FPS
242 frames in 5.0 seconds = 48.136 FPS
208 frames in 5.0 seconds = 41.579 FPS
203 frames in 5.0 seconds = 40.409 FPS
228 frames in 5.0 seconds = 45.427 FPS


Expected results: constant frame rate with no jerking and no stuttering.


Additional info:  Apple Macintosh mini Core Duo i686 only.  Applications > System tools > System Monitor shows less than 30% CPU usage.  'xrandr' says:
DVI1 connected 1280x1024+0+0 (normal left inverted right x axis y axis) 338mm x 270mm
   1280x1024      60.0*+   75.0  

lspci says:
00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)
00:02.0 0300: 8086:27a2 (rev 03)
00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])
	Subsystem: Intel Corporation Device 7270
	Flags: bus master, fast devsel, latency 0, IRQ 16
	Memory at 90380000 (32-bit, non-prefetchable) [size=512K]
	I/O ports at 20f0 [size=8]
	Memory at 80000000 (32-bit, prefetchable) [size=256M]
	Memory at 90400000 (32-bit, non-prefetchable) [size=256K]
	Expansion ROM at <unassigned> [disabled]
	Capabilities: <access denied>
	Kernel driver in use: i915
	Kernel modules: i915

Comment 1 Matěj Cepl 2010-10-01 11:07:47 UTC
Thanks for the bug report.  We have reviewed the information you have provided above, and there is some additional information we require that will be helpful in our diagnosis of this issue.

Please add drm.debug=0x04 to the kernel command line, restart computer, and attach

* your X server config file (/etc/X11/xorg.conf, if available),
* X server log file (/var/log/Xorg.*.log)
* output of the dmesg command, and
* system log (/var/log/messages)

to the bug report as individual uncompressed file attachments using the bugzilla file attachment link above.

We will review this issue again once you've had a chance to attach this information.

Thanks in advance.

Comment 2 John Reiser 2010-10-01 15:37:35 UTC
Created attachment 451038 [details]
/var/log/Xorg.0.log

This is the Xorg log.

There is no .conf file:
   /etc/X11/xorg.conf: No such file or directory

$ glxgears
Running synchronized to the vertical refresh.  The framerate should be
approximately the same as the monitor refresh rate.
205 frames in 5.0 seconds = 40.856 FPS
248 frames in 5.0 seconds = 49.597 FPS
140 frames in 5.0 seconds = 27.958 FPS
136 frames in 5.0 seconds = 27.028 FPS
110 frames in 5.0 seconds = 21.933 FPS
125 frames in 5.0 seconds = 24.880 FPS
110 frames in 5.0 seconds = 21.891 FPS
121 frames in 5.0 seconds = 24.056 FPS
120 frames in 5.0 seconds = 23.980 FPS
154 frames in 5.1 seconds = 30.471 FPS
114 frames in 5.0 seconds = 22.754 FPS
206 frames in 5.0 seconds = 41.078 FPS
221 frames in 5.0 seconds = 44.119 FPS

Comment 3 John Reiser 2010-10-01 15:38:10 UTC
Created attachment 451039 [details]
output of dmesg

Comment 4 John Reiser 2010-10-01 15:43:27 UTC
Created attachment 451044 [details]
/var/log/messages

As root, /var/log/messages was copied to /tmp, then "chmod a+rw" in order to get around 0600 protection on /var/log/messages.

Jerkiness of glxgears did not go away when USB2.0 webcam was unplugged.  (The cam was never used, but the USB plug is the only ON/OFF switch.)

Comment 5 Adam Williamson 2010-10-06 23:09:01 UTC
is it smoother if you keep the CPU occupied? There's a bug several reporters seem to have where GL performance is bad if the CPU is idle...



-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 6 Adam Williamson 2010-10-06 23:09:15 UTC

-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 7 John Reiser 2010-10-07 00:24:02 UTC
When the CPU is busy, then the framerate of glxgears is much closer to the rate of video refresh, but I still see some stuttering and "backwards" motion with Expose events.  (Two non-overlapping Gnome-terminal shell windows, glxgears graphics window overlapping both windows.  Click in sequence on title bars to hide and expose the windows.)

Busy:  $ while true; do date; done

$ glxgears   ## a different Gnome terminal window
Running synchronized to the vertical refresh.  The framerate should be
approximately the same as the monitor refresh rate.
294 frames in 5.0 seconds = 58.706 FPS
301 frames in 5.0 seconds = 60.017 FPS
293 frames in 5.0 seconds = 58.425 FPS
301 frames in 5.0 seconds = 60.007 FPS
292 frames in 5.0 seconds = 58.228 FPS
299 frames in 5.0 seconds = 59.626 FPS
293 frames in 5.0 seconds = 58.390 FPS
301 frames in 5.0 seconds = 60.044 FPS
283 frames in 5.0 seconds = 56.569 FPS
292 frames in 5.0 seconds = 58.274 FPS
278 frames in 5.0 seconds = 55.444 FPS
289 frames in 5.0 seconds = 57.796 FPS
278 frames in 5.0 seconds = 55.376 FPS
285 frames in 5.0 seconds = 56.878 FPS
282 frames in 5.0 seconds = 56.108 FPS
297 frames in 5.0 seconds = 59.360 FPS
292 frames in 5.0 seconds = 58.224 FPS
301 frames in 5.0 seconds = 59.975 FPS
292 frames in 5.0 seconds = 58.263 FPS
300 frames in 5.0 seconds = 59.813 FPS
292 frames in 5.0 seconds = 58.222 FPS
300 frames in 5.0 seconds = 59.804 FPS
^C
$

Comment 8 Adam Williamson 2010-10-07 02:52:38 UTC
from the numbers, it sure looks like this is: https://bugzilla.redhat.com/show_bug.cgi?id=582861

Comment 9 John Reiser 2010-10-07 03:14:49 UTC
It certainly does look the same as 582861 (which has been around for 5 months.)

*** This bug has been marked as a duplicate of bug 582861 ***


Note You need to log in before you can comment on or make changes to this bug.