[Bug 19125] New: log file corruption on dual-core MacOS

wine-bugs at winehq.org wine-bugs at winehq.org
Mon Jun 29 09:51:42 CDT 2009


http://bugs.winehq.org/show_bug.cgi?id=19125

           Summary: log file corruption on dual-core MacOS
           Product: Wine
           Version: 1.1.24
          Platform: PC
        OS/Version: Mac OS X 10.5
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: wineserver
        AssignedTo: wine-bugs at winehq.org
        ReportedBy: hoehle at users.sourceforge.net


Created an attachment (id=22088)
 --> (http://bugs.winehq.org/attachment.cgi?id=22088)
excerpts of +reg traces

The attached log snippet reveals how 2 simultaneous threads (one for font, one
for desktop entries) access the registry. This somehow causes buffer contention
(maybe a static buffer is used somewhere, instead of a stack-allocated one?)
and garbage is printed to the log.

trace:reg:NtEnumerateValueKey (0xe4,95,1,0x74d6bc,256)
e 4096, minappaddr 0x10000, maxappaddr 0x7ffeffff, act.cpumask 3, numcpus 2,
CPU type 586, allocgran. 65536, CPU level 6, CPU rev 5898
^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@trace:reg:NtCreateKey
(0x0,L"Machine\\Hardware\\Description\\System",<null>,0,f003f,0x32fc48)

This snippet might give a clue about where the bogus comes from, as
"maxappaddr" normally does not occur in a registry trace. OTOH, it may be
random buffer content.

Nothing was traced but WINEDEBUG=+reg in the attached log.  It's been produced
with "The Westerner" v1.3.

Maybe this buffer reuse is specific to MacOS.  OTOH, maybe it's caused by the
dual core Intel? I had never seen this before on a single core Ubuntu system.

-- 
Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email
Do not reply to this email, post in Bugzilla using the
above URL to reply.
------- You are receiving this mail because: -------
You are watching all bug changes.



More information about the wine-bugs mailing list