XQuickChecck: 2007-09-10 MiguelMartinez

Bug #138572 reported by Miguel Martinez
2
Affects Status Importance Assigned to Milestone
xorg (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

Dell Inspirion 8600 (Pentium-M 1.70 GHz) w/ ATi Mobility Radeon 9600 128Mb RAM

ati 1:6.7.192-1ubuntu2
xserver 2:1.3.0.0.dfsg-12ubuntu4 (Sept 8 version)

3) X comes up normally. No artifacts
4) Resolution is OK
5) Xrandr mode list is OK, and the panel detected is "LVDS" (OK, too)
6) `xrandr -o 2' scrambled the screen and freezed the local machine (Ctrl+Alt+Backspace didn't work. I don't know hot to use SysReq). Didn't test further.
7) Videos play OK. A large 4:3 480p video uses about 16% CPU load at the lowest throttling (600 MHz). Scaling also works. Scaling on the fly also works, and CPU load stays low.
8) The CPU usage at asus.com is 33% with flashplugin-nonfree. Starcraft II's website is just awfull under scrolling, using all my CPU. Even staying just at the top of the site uses 66% of my CPU. The wiki pages, though, work OK, with the CPU load being nearly always under 20%. As Nicoló said, though, www.oakparkfoundation.com is really slow.
9) No errors shown by glxinfo, and the glxgears framerate is about 2140 FPS with 50% cpu usage at maximum throttling.

Under load:
7) Videos still play OK. Even the large divx one.
8) Scrolling in www.starcraft2.com is more or less as slow as before, although oakparkfoundation is worse. The wikis work OK.
9) glxgears still runs at 2140FPS (god bless the scheduler)

Compiz-fusion:
10) Starts OK. With XAANoOffScreenPixmaps, it is usually smooth except resizing gnome-terminal and gvim (both have columns/rows format), which seems slightly choppy. Pure CPU load doesn't change that (although heavy I/O does)
11) Videos play as before. Under CPU load, some frames seem to drop when resizing on the fly.
12) Scrolling in starcraft is slower, and the CPU without scrolling is higher than metacity, at 72%. The wiki pages are not as silk-smooth as before, but are still pretty smooth. (you need to drag the slider fast to se the difference).
13) glxinfo reports no errors, and glxgears runs at basically the same rate as before (2140 FPS).

Notes:
- EXA wasn't tested
- with this xserver (12ubuntu4 with less Fedora patches) fglrx works without problems, while the previous 12ubuntu4 would fail to start X.
- my device options are as follows (I should increase GARTSize to 128):

Section "Device"
        Identifier "ATI Mobility Radeon 9600"
        Driver "radeon"
        Option "AccelMethod" "XAA"
        Option "EnablePageFlip" "on"
        Option "XAANoOffscreenPixmaps" "on"
        Option "DynamicClocks" "on"
        Option "AGPMode" "4"
        Option "GARTSize" "64"
        BusID "PCI:1:0:0"
EndSection

- In Torcs sometimes the transparent info boxes shimmer.
- rotation via xrandr failed, and this is the backtrace:

Backtrace:
0: /usr/X11R6/bin/Xorg(xf86SigHandler+0x81) [0x80c8631]
1: [0xffffe420]
2: /usr/lib/xorg/modules/extensions//libGLcore.so(_mesa_free+0x1d) [0xaf7b237d]
3: /usr/lib/xorg/modules/extensions//libGLcore.so(XMesaDestroyVisual+0x1d) [0xaf8aa24d]
4: /usr/lib/xorg/modules/extensions//libGLcore.so [0xaf8a9a32]
5: /usr/lib/xorg/modules/extensions//libglx.so(__glXResetScreens+0x37) [0xb7be4657]
6: /usr/lib/xorg/modules/extensions//libglx.so [0xb7be37fc]
7: /usr/X11R6/bin/Xorg(CloseDownExtensions+0x46) [0x809bd36]
8: /usr/X11R6/bin/Xorg(main+0x4af) [0x8076e9f]
9: /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe0) [0xb7d64050]
10: /usr/X11R6/bin/Xorg(FontFileCompleteXLFD+0x1e5) [0x80761c1]

Fatal server error:
Caught signal 11. Server aborting

(II) RADEON(0): RADEONRestoreMemMapRegisters() :
(II) RADEON(0): MC_FB_LOCATION : 0x1fff0000
(II) RADEON(0): MC_AGP_LOCATION : 0x27ff2000

Revision history for this message
Miguel Martinez (el-quark) wrote :

The backtrace I posted is wrong. I just realised it's from another crash I had a week ago. However, /var/log/Xorg.0.log.old (the one that probably crashed, according to the date) has a lot of rubbish in the end. I'm attaching it just in case.

Revision history for this message
Miguel Martinez (el-quark) wrote :

And the lspci -vvnn output just in case.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Thanks for the check, you should try Hardy alph3 now to see if the issues are fixed, and if not, file bugs against the right components.

Changed in xorg:
status: New → Invalid
Revision history for this message
Miguel Martinez (el-quark) wrote : Re: [Bug 138572] Re: XQuickChecck: 2007-09-10 MiguelMartinez

Hi Timo,

It just happens that I distupgraded to hardy yesterday, and I'm
looking forward to doing lots of testing on the X subsystem. If you'd
rather have data from the live CD, don't hesitate to ask. Regards,

Miguel

2008/1/12, Timo Aaltonen <email address hidden>:
> Thanks for the check, you should try Hardy alph3 now to see if the
> issues are fixed, and if not, file bugs against the right components.
>
> ** Changed in: xorg (Ubuntu)
> Status: New => Invalid
>
> --
> XQuickChecck: 2007-09-10 MiguelMartinez
> https://bugs.launchpad.net/bugs/138572
> You received this bug notification because you are a direct subscriber
> of the bug.
>

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.