[Bug 2096] Quicken 2000 Deluxe crashes/Bad parm to CreateBitmap

Wine Bugs wine-bugs at winehq.org
Wed Apr 21 07:26:36 CDT 2004


http://bugs.winehq.org/show_bug.cgi?id=2096





------- Additional Comments From carl.sopchak at cegis123.com  2004-21-04 07:26 -------
I did the regression testing that Tony pointed me to.  There are three files that were 
updated 12/12/2003 at 00:01:20 CST: internet.c, urlcache.c and wininet.spec, all in 
dlls/wininet.  A quick look at the changes leads me to believe they are related to 
implementing unicode, but (a) I could be wrong, and (b) there may be other changes 
unrelated to unicode.  Before these changes are incorporated, Quicken 2000 Deluxe 
does not crash in the scenario originally posted.  After integration, it does. 
 
The changes made in this update were by no means trivial, so it is well beyond my 
capability to try to debug any potential problem with them in this case. 
 
I also have to mention that I find it hard to believe that these changes are DIRECTLY 
causing the problem (although I don't know nearly enough to be certain!).  What part of 
printing and/or display management uses internet related code??  (Remember, the error 
I'm getting is an X resource error, although I guess any memory allocation is supposed to 
be done through X...)  Using --debugmsg +wininet shows no additional messages on 
stdout. 
 
One thing I did notice during the testing is that I get the fixme message from 
CreateCompatibleBitmap showing huge X and Y values even prior to the 12/12 update, 
but for some reason (which I have not determined) Quicken does not crash.  Also, the 
values shown in the message are not necessarily the same when trying different cvs 
versions of wine.  (65,535 +/- seems to come up relatively frequently, as well as 2 [along 
with a number in the millions].) 
 
My gut is telling me that the problem really is that bad value(s) are being passed to 
CreateCompatibleBitmap (or some function which calls same), perhaps due to using a 
wrong data type for the parameter(s), which values get changed based on the size of the 
wine program, or where it happens to be loaded into memory.  I have no real basis for this 
assertion, with the exception of the fact that I have seen LONGs used where there should 
be INTs (see original report).  (I don't know C - or wine coding - well enough to know if 
this is a real problem or not...) 
 
One other thing, I think I have a reporducible scenario:  Set up a Quicken file with 
reminders turned on, and add a transaction to a checking account dated "today" that is a 
check to be printed.  When Quicken is started up, the reminders page is displayed.  
Double click on the check that needs printing.  When the printing dialog is displayed, 
press Alt-K (select checks dated through today), Alt-O (one check on initial page), 
Ctrl-Enter (accept selections).  Quicken will then ask if the tear-off stub is still attached.  
Press Space (selecting Yes).  Quicken crashes hard...  This seems to crash every time 
(for me). 
 
Lastly, I'm trying to determine (via the regression test methodology) when the fixme 
messages began to appear, however I'm having difficulty trying to compile wine using 
dates in 2002.  I'll post what I find... 
 
HTH, 
 
Carl 

-- 
Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.



More information about the wine-bugs mailing list