July 2002 - Wine-Devel - list.winehq.org

Re: DOSEMU team
by Shachar Shemesh July 10, 2002

July 10, 2002

WOW, if wine is advanced enough to catch and spread outlook bugs, we must be really moving along. Seriously, though. If anyone here knows "friest", and corresponded with him on the subject of "DOSEMU team", and is using Windows, please check your system. You are infected with a variant of the "klez" virus. Shachar

1 0

RFC: Re: [ros-kernel] Shell32 and dplax
by Steven Edwards July 10, 2002

July 10, 2002

Note: Forwarded to wine-devel I'm more in support of us chaning wine rather then adapting win9x calls to ReactOS. That being said I have enough trouble writing a 3 line patch so my voice doesnt count to much..."You guys are the programmers, I just break shit." IMHO the way WINE handles certain things is really fubar. A good example besides MapSL and ConvertToGlobalHandle is the way WINE implements PSAPI. All wine psapi does it convert the toolhelp functions already implemented in to the needed format with a wineserver call or two thrown in. If we are going to use wine and not fork the crap out of it then we should fix the 9xisms as we find them. Currently wine has all sorts of ways for doing things, you have win16 apis being called in some places, ntdll in others and then 9xisms all over the place. So far we have MapSL ConvertToGlobalHandle PSAPI DriectX HAL interfaces ? idunno I guess it comes to what can we do to better share code. Currently there is not a chance in hell of us sharing PSAPI in its current state. I started two bugs in wines bugzilla a few weeks ago regarding it. Mabey we should have a meta-bug for making WINE more NT like in "certain" areas. I dont expect the wine projects to make all sorts of changes just for us and not as they are trying to get a 1.0 release. Anyone got any feedback? Mabey we will get more support for this once David and Jason are done with the basic Windowing. Steven > MapSL is in wine kernel32 also, Oh yeah, it's part > of windows 9x > kernel32.dll too. > James __________________________________________________ Do You Yahoo!? Sign up for SBC Yahoo! Dial - First Month Free http://sbc.yahoo.com

4 4

RE: Any advice on how to debug this?
by Medland, Bill July 10, 2002

July 10, 2002

Thanks Eric > -----Original Message----- > From: Eric Pouech [mailto:eric.pouech(a)wanadoo.fr] > Sent: Wednesday, July 10, 2002 12:50 PM > To: Medland, Bill > Subject: Re: Any advice on how to debug this? > > > "Medland, Bill" a écrit : > > > > > -----Original Message----- > > So now what do we do. > > > > Having messed about, got a repeatable case and generally > bashed my head > > around I have got to the following unfortunate situation. > > > > It looks to me like: > > > > comctl32.dll includes a shared data segment. > > One of the entries in that segment, presumably during the > DllEntry, is > > populated with the address of GetSystemMetrics > > > > A second copy of comctl32 then tries to use that address to call > > GetSystemMetrics > > > > Unfortunately the second process has user32 mapped > somewhere else and so the > > call goes somewhere weird and things fall over. > > > > Any ideas as to what we can do about it? > hmm bad news... > > did you try to run your app with a NT version (instead of a > Win9x), that > could help (maybe) No, but I have now. (and ignored all the complaints about mixed versions). It works!! I think that this will be the way to go; see Alexandre's reply. > another thought, when your processes disapear under the > debugger, is the > wineserver still up and running (if it crashes, it may explain the > vanishing processes) Yes it was > another solution would be to disable the support for the shared data > segment... but in your case, it could fix this issue but open another > one TOO TRUE!! The app is very dependent upon shared data segments. > > btw, are you sure user32.dll is mapped at another address in > the second > process Yes > another fix would be to force some DLLs to be loaded at fixed > addresses > (user32 could be one of them)... > another question, are you sure the two processes load the > same user32 ? Yes > (could it be that two wine versions coexist and that a > different one is > loaded in each process) ? > > A+ > Thanks for all the help Bill

1 0

RE: Any advice on how to debug this?
by Medland, Bill July 10, 2002

July 10, 2002

> -----Original Message----- > From: Eric Pouech [mailto:eric.pouech(a)wanadoo.fr] > Sent: Wednesday, July 10, 2002 4:59 AM > To: Medland, Bill > Cc: Wine devel (E-mail) > Subject: Re: Any advice on how to debug this? > > > > The stack traces in both cases are similar and indicate to > me that the > > processes have ended up executing data. > > However by the time I get to disassemble the code referred to by the > > backtrace the code isn't there any longer. > is most of the cases, it means that the program is no longer > running (it > may > have been killed by the other program) > > > What's my best bet for tracking this down? > as the processes seem to be furtive, using relay traces (and friends) > may be > a better bet to locate the issue, and then perhaps set the bp where > needed > > A+ > Thanks Eric. So now what do we do. Having messed about, got a repeatable case and generally bashed my head around I have got to the following unfortunate situation. It looks to me like: comctl32.dll includes a shared data segment. One of the entries in that segment, presumably during the DllEntry, is populated with the address of GetSystemMetrics A second copy of comctl32 then tries to use that address to call GetSystemMetrics Unfortunately the second process has user32 mapped somewhere else and so the call goes somewhere weird and things fall over. Any ideas as to what we can do about it? Bill

2 1

Meta-bug conversions
by Francois Gouget July 10, 2002

July 10, 2002

Hi, I propose to use what would hopefully be better concepts to replace the following meta-bugs: * bug 35 - Wine 0.9.0 Meta Bug http://bugs.winehq.com/showdependencytree.cgi?id=35 Instead we could tag all the bug that should be fixed in Wine 0.9.0 with the 0.9.0 target milestone. * bug 395 - Tasklist http://bugs.winehq.com/showdependencytree.cgi?id=395 * bug 455 - FIXMEs http://bugs.winehq.com/showdependencytree.cgi?id=455 * bug 406- Tasklets http://bugs.winehq.com/showdependencytree.cgi?id=406 These three could be replaced by keywords: Tasklist, FIXMEs and Tasklets. Then these bugs can be replaced with a query that returns the bugs that have these keywords. In all cases, the advantages of the alternative methods are: * the above meta-bugs introduce artificial dependencies between bugs * each time a bug changes status, you get an email for the bug, and an email for the corresponding meta-bug * similarly, moving a bug moving from one category to another generates three emails * the new concepts seem a better match Does the above proposal look ok? If so then, once the required keywords are created, I can tag the relevant bugs and retire the meta-bugs (which will create a flurry of emails to wine-bugs in the short term). -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ I haven't lost my mind, it's backed up on tape around here somewhere...

2 1

about XIM support on wine
by rain shaw July 10, 2002

July 10, 2002

Hi all, I do apply some patches to the wine source according to some tips in bugs library,and I can input chinese in notepad now. But,every time window was created,will the input method popup.For example,if a message box popped up,the xim will popup as well. It seems that we called the XIM related Xlib function every time new window was created. Would u please give me some suggestions about how to disable create and open XIM while create non-text-input related window? in other word,how to determin the x-window that going to be created is a control that need input method ? Thanx!

2 1

Debugger
by Massimo July 10, 2002

July 10, 2002

Hmmm.... I fiddled a bit with winedbg and found it awful to use.... Maybe I'm too comfortable with visual debuggers :-) There's no plan to make a graphical front-end to it, for example extending DDD or better GVD ? I had a look at GVD sources and docs, and it shouldn't be too difficult to add a supported debugger... It has a quite good support for many external debuggers.... I see only une problem on it, as it seems to me that winedbg opens a new xterm and translates his i/o on it, which can confuse GVD as stated in docs. So, 2 questions : 1) - There's a way to instruct winedbg to not change xterm ? 2) - There's a volunteer with enough skills on winedbg that can help me to write GVD patch ? :-) Regards Max

4 8

Problems with VGA PLANETS 4
by Oliver Ob July 10, 2002

July 10, 2002

My P4_BUG.log 01 _ Installed Wine run with -v gives "Wine release 20000430" and error "fixme: pthread_atfork" 02 _ Edited /etc/wine.conf c:\windows\Profiles\Administrator\user.dat must be c:\windows\user.dat 03 _ Trying wine /wine/mnt/c/Programme/Play/vgap/Planets4/planets4.exe as "user" gives errors: - fixme: pthread_atfork - what does this mean at all? - fixme:reg:_w95_parse_dkv RegSetValueEx returned: 0x0000000e - err:seh:EXC_DefaultHandling Unhandled exception code c000013a flags 0 addr 0x401 36fb4 04 _ Building the font metrics takes a LOT of time, god is that slow! And then it fails! fixme:debug:FlushInstructionCache (0xffffffff,0x41410000,0x00006000):stub last address with 6000, 6000 and with a000 - a popup window from wine says "Unhandled exception 0xc0000005 at address 0x24748b56. Do you want to debug it?" and clicking YES results in - err:seh:UnhandledExceptionFilter Couln't start debugger (winedbg 136366288 1414155735) 05 _ The Temporary Directory (as specified in your config file) is NOT writeable - how can "user" have write access to c:\temp now? What user group has rw access to /mnt/c ? 06 _ Running using su -c "see above" works better, so the errors which do not tell me nothing seem to mostly have to do with access rights. Can someone explain those errors to me? Errors are yet some others - fixme:file:AreFileApisANSI (void): stub - fixme:win32:PE_LoadImage We need to perform base relocations for c:\Programme\Play\vgap\Planets4\gug32.dll - fixme:win32:PE_CreateModule Security directory ignored - err:menu:MENU_GetMenu invalid menu handle=0, ptr(nil), magic=0 - err:menu:MENU_GetMenu invalid menu handle=0, ptr(nil), magic=0 - err:menu:MENU_GetMenu invalid menu handle=4d0, ptr=0x4043f27c, magic=0 - err:menu:MENU_GetMenu invalid menu handle=0, ptr(nil), magic=0 - err:menu:MENU_GetMenu invalid menu handle=0, ptr(nil), magic=0 - fixme:bitmap:X11DRV_DIB_GetImageBits_4 from 24 bit bitmap with mask R,G,B 0,0,0 to 4 bit DIB 07 _ Will try to make my first turn using wine + P4 now. Unpacking the RST works. It opens. I can edit the turn. Try to make a TRN file. But now this makes me worry!!!! A popup window speaks to me that the Planets4.exe has been DAMAGED?!! and that I must replace it with a good copy!??!?!?!?!? 08 _ I quit and see these error messages: - fixme:string:LCMapStringA (0x0400,0x00400100,0x424bf08c,1,(nil),0): unimplemented flags: 0x00400000 - fixme:string:LCMapStringA (0x0400,0x00400100,0x424bf08c,1,0x4044f048,1): unimplemented flags: 0x00400000 - fixme:string:LCMapStringA (0x0400,0x00400100,0x424bf08c,1,(nil),0): unimplemented flags: 0x00800000 - fixme:string:LCMapStringA (0x0400,0x00400100,0x424bf08c,1,0x4044f048,1): unimplemented flags: 0x00800000 - fixme:dc:GetDCEx new hrgnClip[9722] smashes the previous [96c2] - fixme:dc:GetDCEx new hrgnClip[96be] smashes the previous [9726] And what now? And whenever I minimalize the wine window with the program I cannot get it back to full size (normally ALT SPACE + X) how to do that? -- *º¤., ¸¸,.¤º*¨¨¨*¤ =Oliver(a)home= *º¤., ¸¸,.¤º*¨¨*¤ I http://www.bmw-roadster.de/Friends/Olli/olli.html I I http://www.bmw-roadster.de/Friends/friends.html I I http://groups.yahoo.com/group/VGAP-93 I I mailto:VGAP-93-subscribe(a)yahoogroups.com I I http://home.t-online.de/home/spacecraft.portal I >>> Telek0ma iBBMS - now back online +49.4504.TRSi1/TRSi2 <<<

2 1

PSDRV question
by Carl Sopchak July 10, 2002

July 10, 2002

I've been trying to get Quicken 2000 Deluxe running in wine. After I figured out that it wasn't finding a DLL for the internet connection (and got that fixed), I started testing the second-most important function of Quicken that I need: check printing. All went well until I tried to print a sample check (to verify alignment), at which point Wine crashed into winedbg. I figured out why: When reading the ppd file, it sets up an *InputSlot for ManualFeed (even though it's not really an InputSlot, but most Win programs treat it as one...). The problem was that, in ppd.c, a Name wasn't given to the slot (but a FullName was), so when ps.c tried to output the header for the InputSlot it crashed (since strlen(NULL) causes a crash). I hacked the AddSlot call to name the slot "Manual", and this seems to work OK. The question that I have is, is this the way that it should REALLY get fixed (does the name matter?), or should something else be done? I'd be happy to correct my hack, if necessary. In either case, I can post the patch to sourceforge, once I know "the right way" this should be handled. Thanks for the help, Carl P.S., I added some TRACE() statements to help me debug this. These can be left in the patch that I post, right?

2 2

DOSEMU team
by friest July 10, 2002

July 10, 2002

1 0