[Bug 12365] New: wineserver : High CPU usage and fails to close after application with wine 0.9.59
http://bugs.winehq.org/show_bug.cgi?id=12365
Summary: wineserver : High CPU usage and fails to close after application with wine 0.9.59 Product: Wine Version: CVS/GIT Platform: Other URL: http://winehq.org OS/Version: other Status: UNCONFIRMED Severity: major Priority: P2 Component: kernel32 AssignedTo: wine-bugs@winehq.org ReportedBy: kj.allcock@virgin.net
Created an attachment (id=11858) --> (http://bugs.winehq.org/attachment.cgi?id=11858) High CPU on 0.9.59 trace
After an update to git version 0.9.59 (make uninstall;make clean;make depend && make; make install), at every startup of an application with wine, the wineserver and 'new' services.exe take up 100% CPU. In my case on dual core PC means 50% so applications still manage to run but incredibly slowly.
After closing whatever application it was, wineserver and services.exe continue to run. I have to do a manual wineserver -k to remove them.
Trying to look at the console output with the +relay flag shows there is crash in kernel32 (ntoskrnl.exe.KeQueryActiveProcessors), and then many "Waits".
I have tried 3 different applications previously known to work (Steam, TeamSpeak, DVDShrink).
As its a wineserver behaviour i'm not sure how to get you more details ...
I'm on: Fedora 8 (latest kernel) Intel Dual Core NVidia GForce 7100 with nvidia driver 169.12
http://bugs.winehq.org/show_bug.cgi?id=12365
--- Comment #1 from Keith kj.allcock@virgin.net 2008-04-05 01:47:41 --- Created an attachment (id=11859) --> (http://bugs.winehq.org/attachment.cgi?id=11859) Another log but with +all (4000 lines)
Hi,
I have added a more comprehensive log in case it helps ... Uses the +all flag, so have only attached the first 4000 lines
http://bugs.winehq.org/show_bug.cgi?id=12365
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- URL|http://winehq.org | Severity|major |normal Component|kernel32 |-unknown
--- Comment #2 from Dmitry Timoshkov dmitry@codeweavers.com 2008-04-05 03:58:46 --- If that's a regression please perform the regression test:
http://wiki.winehq.org/RegressionTesting
http://bugs.winehq.org/show_bug.cgi?id=12365
--- Comment #3 from Dmitry Timoshkov dmitry@codeweavers.com 2008-04-05 03:59:57 --- Also please try with clean ~/.wine first.
http://bugs.winehq.org/show_bug.cgi?id=12365
--- Comment #4 from Keith kj.allcock@virgin.net 2008-04-05 04:26:08 --- Yes, I have been working on this, Its the first time I have done the regression test.
Good news is that the clean .wine directory has stopped the CPU usage, but is that something that ideally still needs to be looked at ? Its not good to have to reinstall everything from scratch after each update.
It still looks like the wineserver failing to shut down still exists though. Could this be by design now ? If its a small resident program to speed up wine loads or something it might be intended ?
http://bugs.winehq.org/show_bug.cgi?id=12365
James Hawkins truiken@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |INVALID
--- Comment #5 from James Hawkins truiken@gmail.com 2008-04-05 04:29:38 --- (In reply to comment #4)
Yes, I have been working on this, Its the first time I have done the regression test.
Good news is that the clean .wine directory has stopped the CPU usage, but is that something that ideally still needs to be looked at ? Its not good to have to reinstall everything from scratch after each update.
Wine is beta software (or alpha, can't remember)...but the point remains. Until we have a stable release, you can expect to remove and recreate your wine prefix after an update.
It still looks like the wineserver failing to shut down still exists though. Could this be by design now ? If its a small resident program to speed up wine loads or something it might be intended ?
Yes it's be design. Marking this bug as invalid.
http://bugs.winehq.org/show_bug.cgi?id=12365
James Hawkins truiken@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #6 from James Hawkins truiken@gmail.com 2008-04-05 04:29:51 --- Closing.
http://bugs.winehq.org/show_bug.cgi?id=12365
--- Comment #7 from Austin English austinenglish@gmail.com 2008-04-05 05:09:51 --- (In reply to comment #4)
Yes, I have been working on this, Its the first time I have done the regression test.
Good news is that the clean .wine directory has stopped the CPU usage, but is that something that ideally still needs to be looked at ? Its not good to have to reinstall everything from scratch after each update.
It still looks like the wineserver failing to shut down still exists though. Could this be by design now ? If its a small resident program to speed up wine loads or something it might be intended ?
If you backed up the directory before erasing it, could you see if running $ wineprefixcreate
fixes the issue as well? If not, don't worry about it.
http://bugs.winehq.org/show_bug.cgi?id=12365
Detlef Riekenberg wine.dev@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wine.dev@web.de Status|CLOSED |UNCONFIRMED Component|-unknown |ntoskrnl OS/Version|other |Linux Platform|Other |PC-x86-64 Resolution|INVALID | Summary|wineserver : High CPU usage |ntoskrnl: crash on |and fails to close after |unimplemented API: |application with wine 0.9.59|KeQueryActiveProcessors / | |KeQueryInterruptTime | |(starforce)
--- Comment #8 from Detlef Riekenberg wine.dev@web.de 2008-04-05 15:23:06 --- (In reply to comment #3)
Also please try with clean ~/.wine first.
(In reply to comment #5)
Wine is beta software (or alpha, can't remember)...but the point remains. Until we have a stable release, you can expect to remove and recreate your wine prefix after an update.
(In reply to comment #7)
could you see if running $ wineprefixcreate fixes the issue as well? If not, don't worry about it
This is no longer valid: ---- cut ---- I think we should stop telling people to blow away their .wine, and have them run wineprefixcreate instead. ---- cut ---- http://www.winehq.org/pipermail/wine-devel/2008-March/064252.html #################################################################
Please read the Bug, before you close it. The description from Keith has already mentioned the crash-location: ntoskrnl.exe.KeQueryActiveProcessors
From the crashlog: PE 450000- 461000 Deferred sfdrv01.sys
This is the Starforce Copy Protection.
Keith, what was the version of Wine before your Update? Do you still have the affected .wine directory? What Programs did you installed? (Might be a Game and can already be uninstalled in the meantime)
http://bugs.winehq.org/show_bug.cgi?id=12365
Vitaliy Margolen vitaliy@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #9 from Vitaliy Margolen vitaliy@kievinfo.com 2008-04-05 17:07:14 --- StarForce's drivers won't work in Wine regardless how hard you try. However there are lots of drivers that won't work right now on Wine.
What should be fixed is wineserver/services.exe/ntoskrnl to gracefully handle failure of the starting a driver.
http://bugs.winehq.org/show_bug.cgi?id=12365
killertux killertux@suomi24.fi changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |obfuscation
http://bugs.winehq.org/show_bug.cgi?id=12365
--- Comment #10 from Detlef Riekenberg wine.dev@web.de 2008-04-06 14:31:28 --- (In reply to comment #9)
StarForce's drivers won't work in Wine regardless how hard you try.
I know.
However there are lots of drivers that won't work right now on Wine.
I send Patches for the missing functions. KeQueryActiveProcessors: http://www.winehq.org/pipermail/wine-patches/2008-April/052826.html
KeQueryInterruptTime: http://www.winehq.org/pipermail/wine-patches/2008-April/052828.html
What should be fixed is wineserver/services.exe/ntoskrnl to gracefully handle failure of the starting a driver.
I think, you mean an Exception-Handler?
http://bugs.winehq.org/show_bug.cgi?id=12365
Detlef Riekenberg wine.dev@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #11 from Detlef Riekenberg wine.dev@web.de 2008-04-08 17:37:26 --- Patches for KeQueryActiveProcessors and KeQueryInterruptTime are commited.
Use current gt or wait for Wine-0.9.60
I like the Idea of an Exception-Handler, but I have no game to test.
http://bugs.winehq.org/show_bug.cgi?id=12365
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #12 from Alexandre Julliard julliard@winehq.org 2008-04-18 11:04:36 --- Closing bugs reported fixed in 0.9.60.
http://bugs.winehq.org/show_bug.cgi?id=12365
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|CVS/GIT |unspecified
http://bugs.winehq.org/show_bug.cgi?id=12365
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |28553fa8a79fe1dbc0c2cc15a16 | |72d88bb7d1e66 CC| |focht@gmx.net Version|unspecified |0.9.59.
-
wine-bugs＠winehq.org