Summary: | Kwin dreadfully crashed EVERY DAMN THING I TRY TO DO!!! | ||
---|---|---|---|
Product: | [Plasma] kwin | Reporter: | Christian <over1pixel> |
Component: | general | Assignee: | KWin default assignee <kwin-bugs-null> |
Status: | RESOLVED DUPLICATE | ||
Severity: | crash | ||
Priority: | NOR | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Arch Linux | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
Christian
2010-09-10 21:58:49 UTC
a) a crashreport w/o a backtrace is like takling about the weather - nice but pointless ;-) b) do not - NEVER! - collect bugs, that's not helpfull to handle them (i skipped reading the latter ones) c) - the crash is likely the DRI2 issue as in bug #241402, try to add "export LIBGL_ALWAYS_INDIRECT=1" to ~/.xprofile (create the file in doubt) - disable the blur plugin, your "GPU" is not capable of this at all - add your GPU to the blacklist for the lanzos filter (like in bug #243181) to fix overexposed thumbnails and bad thumbnailing performance (it's btw. exposé - the accent goes the other direction... and KDE calls it "present windows" :-) i'll mark this as "duplicate" of the dri bug, check what remains and open new bugs FOR EACH remaining issue in case. thanks. ps: this is no helpdesk, bug i don't even know what "package module" is supposed to be... *** This bug has been marked as a duplicate of bug 241402 *** ehm... I didn't attached any bug report because on every crash, the wizard told me that the report was unusefull (I have gbd installed)
the ~/.xprofile helped a lot
I don't know how to add my card to the blacklist, glxinfo says:
<<
OpenGL vendor string: Tungsten Graphics, Inc
OpenGL renderer string: Mesa DRI Intel(R) G45/G43 GEM 20100328 2010Q1 x86/MMX/SSE2
OpenGL version string: 1.4 (2.1 Mesa 7.8.2)
>>
I modified the blacklist in kwinrc in:
[Blacklist][Lanczos]
Advanced Micro Devices=DRI R600:-:7.8.1,DRI R600:-:7.8.2
Intel=GM45 Express Chipset GEM 20100328:-:7.8.2,GM45 Express Chipset GEM 20091221:-:7.7.1,965GM GEM 20100328 2010Q1:-:7.8.2,965GM GEM 20091221 2009Q4:-:7.7.1,Ironlake Mobile GEM 20100328:-:7.8.2,Mesa DRI Intel(R) G45/G43 GEM 20100328 2010Q1 x86/MMX/SSE2
adding the last part after the last comma, but with NO improvements on improvvise freeze of the desktop with the X process at 100%
(sorry for have been used you like an helpdesk :P )
try ",G45/G43 GEM 20100328 2010Q1:-:7.8.2" then restart "kwin --replace &" while this is no helpdesk, in order to get any help anyway you'll at least have to explain your opinion of "package module" - stuff presented by lsmod are always kernel modules (the kernel won't load other items :-) even with modifying kwinrc as you suggest I still have several HUGE kde slow downs if I have more than 1 user logged in different kde sessions I guess it doesn't happen so random, for example if I rapidly open 2-3 heavy tabs in firefox (facebook, gmail, google reader) the system freezes for... 20 seconds, every time!! could it be related to shared video memory that, maybe not correctly managed, is not enough for so many instances of kde? a) if it's related to compositing*, suspend it for the non active sessions (SHIFT+ALT+F12 - also toggles resume) b) firefox is no way a kde application, but 20 seconds rather sound like a HDD invocation to me, ie. FF's caching. Last time i had such lags reg. FF, there were bad blocks on the HDD -> run smartctl for your own good. c) please elaborate on "freezes the system" - just the client? (FF) the entire server (X11, all GUI) or _really_ the system ("linux" - what heppens with eg. music playback from a CLI player (eg. mplayer) during this) *compositing requires to more or less double the X11 ersource amount, you can check your current resource stats using "xrestop" and compare them to your HW - also 'm currently under the impression that there're issues with your gpu/driver handling multiple gl contexts... I think the real problem is the interaction between kde and graphic driver, and so through X when I say "freeze" I mean that there's one or two processes called "X" which suck all my dual core cpu! since I suppose X has high priority, this hangs up the whole system now I try smartctl but I guess it's not an hdd fault |