Bugzilla – Bug 575603
X.org second desktop uses 100% CPU
Last modified: 2010-06-25 01:37:54 UTC
Created attachment 339846 [details] Xorg.0.log User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.6) Gecko/20091201 SUSE/3.5.6-1.1.1 Firefox/3.5.6 Using kdm and KDE with OpenSuse 11.2 upgraded from 11.1, upgraded from 11.0 etc... When two users are logged in KDE via the switch user method, the desktop that is in the background will eventually eat up all the cycles of one CPU. It doesn't happen right away but after a few minutes delay. I think it is related to when the screen saver on the non-visible desktop would have started. Or it could be when the power saving mode is "entered" on the non-visible desktop, I am not sure. In any case it's a big problem for a laptop. Visiting the stuck desktop with CTRL-ALT-F8 and back with F7 cures the problem momentarily but it comes back again after a delay. The screensaver on all accounts is "blank", so the problem is not with openGL or similar. Here is a few lines from top: top - 10:19:32 up 1:24, 5 users, load average: 1.16, 1.25, 1.21 Tasks: 169 total, 2 running, 166 sleeping, 1 stopped, 0 zombie Cpu(s): 8.1%us, 44.7%sy, 0.0%ni, 47.2%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 3060628k total, 1580744k used, 1479884k free, 61624k buffers Swap: 4072468k total, 0k used, 4072468k free, 1010784k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 6726 root 20 0 192m 51m 5232 R 98 1.7 30:53.89 Xorg 1471 root 20 0 387m 58m 7940 S 2 2.0 3:26.13 Xorg Reproducible: Always Steps to Reproduce: 1. Log in a KDE session 2. Switch user and log into another KDE session 3. Go back to first session (not sure if this is necessary) 4. Wait 20 minutes 5. Voila! Top will show 100% CPU usage by Xorg Actual Results: Xorg uses 100% CPU Expected Results: Xorg should use a few percent CPU at best for blank screen saver...
Created attachment 339847 [details] Xorg.1.log
It may be a regression from Bug 472046 "X.Org grabs almost 100% of CPU load on an idle system" https://bugzilla.novell.com/show_bug.cgi?id=472046. This bug seems to be the same problem in 11.1 and is marked as fixed for that distribution.
Created attachment 339896 [details] xorg.conf
You're using the radeon driver for Radeon X1400. Does this issue als occur when using radeonhd driver (which is our default)?
(In reply to comment #4) > You're using the radeon driver for Radeon X1400. Does this issue als occur when > using radeonhd driver (which is our default)? I tried the ati driver and it does the same thing. The radeonhd driver does not work with my card (X1400)
The ati driver loads the radeon driver, so that's no new information. X1400 is supported by radeonhd driver. If there are issues with that support, that would be worth being reported.
(In reply to comment #6) > The ati driver loads the radeon driver, so that's no new information. X1400 is > supported by radeonhd driver. If there are issues with that support, that would > be worth being reported. OK, I managed to get a newer radeonhd working (this one: http://download.opensuse.org/repositories/X11:/Drivers:/Video/openSUSE_11.2/i586/xorg-x11-driver-video-radeonhd-1.3.0_20100216_79a0ab2-2.1.i586.rpm) I have the same issue with radeonhd, plus a few more. I will file a separate bug report for those other issues. I tried the vesa driver as well, same problem.
Additional information: - The recent update from KDE 4.3.1 to KDE 4.3.5 did not fix the problem - I removed the option to start the screen saver automatically on all accounts and the problem remains. Hence it does not seem to be related to the screen saver
Unfortunately I no longer have time to address bugs for already released openSUSE/SLE products. If possible verify whether the issue still exists with currently developed openSUSE Factory/SLE11-SP1. If it does please reopen the bugreport and set the product accordingly. Of course you can reopen the bugreport nevertheless, but that doesn't change the situation and the bugreport will remain open for the time being without being addressed.
Seeing exactly the same thing with an up to date 11.2. I think it's best to keep the bug open as far as it's not confirmed to be fixed. This might help to reduce the number of new duplicates (searching for a closed bug is usually not done by default). Please correct the whiteboard content I am adding if inadequate. This is indeed what we need: someone who has the opportunity to test this issue on the development version.
So please go ahead and give 11.3 a try-
11.3 M7 is available for testing. I suggest to reopen in case you can still reproduce that issue.
I tried with 11.3 RC1 and the problem doesn't happen for me anymore.