http://bugs.winehq.org/show_bug.cgi?id=19456
Summary: Mount Manager crashes with optical discs since commit 15acd35 Product: Wine Version: 1.1.26 Platform: PC OS/Version: Mac OS X 10.5 Status: UNCONFIRMED Severity: normal Priority: P2 Component: ntdll AssignedTo: wine-bugs@winehq.org ReportedBy: cdavis@mines.edu
Created an attachment (id=22602) --> (http://bugs.winehq.org/attachment.cgi?id=22602) Wine Log with backtrace showing problem
Recently, the Mount Manager driver (mountmgr.sys) has been crashing on Mac OS. (Good thing it's not a real ring 0 driver...). I've attached a log showing what happens. From the log, it looks like there's some heap corruption going on here.
It only crashes when there's a disc in the optical disc drive. And even then, it doesn't crash the first time it sees the disc. The SECOND time and subsequent times, when Wine already knows about the disc, it crashes.
I did a regression test and found that this is the offending commit:
commit 15acd35fa27d09af01698083af21ef3772edcae1 Author: Alexandre Julliard julliard@winehq.org Date: Wed Jul 22 20:56:04 2009 +0200
mountmgr: Try to match and reuse volumes when creating/updating a drive.
For some reason, reusing volumes fails spectacularly on Mac OS (at least on my system), leading to possible heap corruption and a crash.
Even worse, the driver goes down while holding its critical section. Somehow the CS isn't released. Then, programs that try to use Mount Manager end up deadlocking:
err:ntdll:RtlpWaitForCriticalSection section 0x4040e500 "device.c: device_section" wait timed out in thread 0017, blocked by 0018, retrying (60 sec)
Eventually, Wine gives up on the CS, but then weird things start happening to programs that rely on Mount Manager. For example, winecfg shows no drives on the Drives tab (because Mount Manager is down and can't handle IOCTL requests).
Any help on this would be appreciated.
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #1 from Charles Davis cdavis@mines.edu 2009-07-25 04:33:43 --- I forgot to mention: the log is from me running winecfg and selecting the Drives tab.
http://bugs.winehq.org/show_bug.cgi?id=19456
Jeff Zaroyko jeffz@jeffz.name changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |ead1234@hotmail.com
--- Comment #2 from Jeff Zaroyko jeffz@jeffz.name 2009-07-25 04:41:37 --- *** Bug 19453 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=19456
Jeff Zaroyko jeffz@jeffz.name changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #3 from Jeff Zaroyko jeffz@jeffz.name 2009-07-25 04:45:58 --- backtrace with debug symbols: http://bugs2.winehq.org/attachment.cgi?id=22599
http://bugs.winehq.org/show_bug.cgi?id=19456
André H. nerv@dawncrow.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW CC| |nerv@dawncrow.de Ever Confirmed|0 |1
--- Comment #4 from André H. nerv@dawncrow.de 2009-07-28 14:09:31 --- confirming, got the same problem on PC-x64 in Linux
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #5 from Johan Gill johan.gill@gmail.com 2009-07-29 05:19:03 --- Created an attachment (id=22672) --> (http://bugs.winehq.org/attachment.cgi?id=22672) +mountmgr log with extra logging by me
mountmgr deletes the device entry for the volume, and then tries to access it.
Here's a log that shows the thing happening. I added some logging for clarity.
http://bugs.winehq.org/show_bug.cgi?id=19456
Johan Gill johan.gill@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |johan.gill@gmail.com
--- Comment #6 from Johan Gill johan.gill@gmail.com 2009-07-29 05:58:22 --- Can someone change the version field for this bug? It was introduced after 1.1.26.
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #7 from Dmitry Timoshkov dmitry@codeweavers.com 2009-07-29 06:34:33 --- (In reply to comment #6)
Can someone change the version field for this bug? It was introduced after 1.1.26.
To which value? What 'wine --version' does say for you?
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #8 from Johan Gill johan.gill@gmail.com 2009-07-29 11:37:58 --- The bug seems to emerge in version wine-1.1.26-140-g15acd35
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #9 from Johan Gill johan.gill@gmail.com 2009-07-29 11:42:56 --- Hmm, I just checked the possible values, and I see that we don't have the CVS/GIT option anymore. However, there is "Unspecified". Not really a question specific to this bug, but what are the current semantics of the version field?
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #10 from Nikolay Sivov bunglehead@gmail.com 2009-07-29 11:53:00 --- (In reply to comment #9)
Hmm, I just checked the possible values, and I see that we don't have the CVS/GIT option anymore. However, there is "Unspecified". Not really a question specific to this bug, but what are the current semantics of the version field?
Currently CVS/GIT is a kind of obsolete value with request left in comment to use explicit version bug was reported for. Now latest release version used for new bugs even if problem occurred between releases. You shouldn't care too much about that.
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #11 from Johan Gill johan.gill@gmail.com 2009-07-29 16:02:04 --- Created an attachment (id=22690) --> (http://bugs.winehq.org/attachment.cgi?id=22690) Patch to refresh the volume pointer before it is passed to create_dos_device, so the volume is recreated if needed
Does this help?
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #12 from Charles Davis cdavis@mines.edu 2009-07-29 16:17:13 --- (In reply to comment #11)
Created an attachment (id=22690)
--> (http://bugs.winehq.org/attachment.cgi?id=22690) [details]
Patch to refresh the volume pointer before it is passed to create_dos_device, so the volume is recreated if needed
Does this help?
Yup. Doesn't crash anymore. Thanks!
Don't know what AJ will make of it, though. And to find out, we'll have to wait 'till next week.
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #13 from Johan Gill johan.gill@gmail.com 2009-07-29 16:19:43 --- I'm submitting it to wine-patches, and then we'll see what he makes of it.
http://bugs.winehq.org/show_bug.cgi?id=19456
Ken Sharp kennybobs@o2.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |kennybobs@o2.co.uk
http://bugs.winehq.org/show_bug.cgi?id=19456
Jeff Zaroyko jeffz@jeffz.name changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com
--- Comment #14 from Jeff Zaroyko jeffz@jeffz.name 2009-08-03 02:50:43 --- *** Bug 19555 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=19456
--- Comment #15 from Johan Gill johan.gill@gmail.com 2009-08-04 13:40:41 --- It seems Alexandre fixed the underlying cause of the volume getting deleted in the first place. Retest current git.
http://bugs.winehq.org/show_bug.cgi?id=19456
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #16 from Alexandre Julliard julliard@winehq.org 2009-08-06 10:31:54 --- Should be fixed.
http://bugs.winehq.org/show_bug.cgi?id=19456
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #17 from Alexandre Julliard julliard@winehq.org 2009-08-07 12:59:53 --- Closing bugs fixed in 1.1.27.
http://bugs.winehq.org/show_bug.cgi?id=19456
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- OS/Version|Mac OS X 10.5 |Mac OS X
-
wine-bugs＠winehq.org