March 2004 - Wine-Devel - list.winehq.org

Re: Win32 packages released on sourceforge
by Ivan Leo Murray-Smith March 19, 2004

March 19, 2004

> Can you please post the list of errors to wine-devel, so people are aware of the > problems (and hopefully send patches to fix them :))? Complete compile error log is attached. > Now that we have this build, can we get rid of: > Wine programs for Windows > from the "Support Files" package? Done. Ivan.

5 5

Re: urlmon: Implement FindMimeFromData
by Juan Lang March 19, 2004

March 19, 2004

Kevin, you said: IMHO we should really keep the list that is checked to a minimum (being the list Windows checks) as this code is basically doing "If a web server reports a MIME type in this list, ignore it and make our own guess" which is totally wrong standards-wise and a major cause of many problems under IE on Windows.. (Apologies for the lack of quoting.) According to that appendix, IE should only be overriding MIME types that are text/plain, application/octet-stream, or empty. It does so because it considers these "ambiguous". I think your patch may be useful for the small set of types you suggest, but shouldn't you only override in these cases too? --Juan __________________________________ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.com

3 2

wine.inf
by Ivan Leo Murray-Smith March 18, 2004

March 18, 2004

Hi list, I've installed wine.inf, but the system.reg file is very short, it only contains font information and computer name, this isn't exactly what I got with winedefault.reg I need this because I'm going to release a new version of winesetuptk that doesn't add the "Filesystem" parameter to the drive config, as AJ has removed it, and the winesetuptk build tree wants system.reg, user.reg and userdef.reg. But system.reg, well, I've described it, userdef.reg is empty and user.reg only contains the Control Panel\\Colors and Control Panel\\International entries. Am I doing something wrong? Ivan.

1 0

urlmon: Implement FindMimeFromData
by Kevin Koltzau March 18, 2004

March 18, 2004

Here is an implementation of FindMimeFromData, based off of MSDN http://msdn.microsoft.com/workshop/networking/moniker/overview/appendix_a.a… There is lots of debate on the 'net regarding this "feature" in IE, most of it suggesting this is a bad idea.. One possible change that would work more "standards compliant" rather then "Windows like" would be to return the proposed MIME type unchanged if one is given, otherwise go through the detection methods..let me know if that would be a preferred solution Changelog Implement FindMimeFromData

2 2

Re: [RESENT] TIME_GetBias
by Alexandre Julliard March 18, 2004

March 18, 2004

Uwe Bonnes <bon(a)elektron.ikp.physik.tu-darmstadt.de> writes: > Changelog: > wine/dlls/ntdll/time.c: TIME_GetBias > Calculate the Bias before calling gmtime This cannot be right, the bias is supposed to be the difference between local time and GMT, so you have to use gmtime() not localtime(). -- Alexandre Julliard julliard(a)winehq.org

5 15

Re: Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec)
by Christian Costa March 18, 2004

March 18, 2004

> Message du 18/03/04 00:51 > De : Marcus Meissner > A : Christian Costa > Copie à : wined > Objet : Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec) > On Wed, Mar 17, 2004 at 10:37:01PM +0000, Christian Costa wrote: > > Hi, > > > > When executing the DllMain of the native IR32_32.DLL (VFW), Wine crashes. > > After some investigations, I found that a part of the code uses the EBX > > register at one point without saving > > and restore it afterwards. Since this register contains the PLT pointer > > for standard unix functions like libc ones, this > > does not take a long time for Wine to crashing. > > I tested with the ugly hack attached that saves and restores the EBX > > register between the call to DllMain and that fixed the problem and > > the DLL works fine then. > > This might happen. That ebx can be scrapped? > > Your fix is not really ok, you must not modify the stackpointer in asm statements. > > Try using: > __asm__ __volatile__ ("":::"ebx"); > > after entry(), which tells the compiler that ebx has been scrapped. Thanks for this! :-) Christian

2 1

Re: Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec)
by Christian Costa March 18, 2004

March 18, 2004

> Message du 18/03/04 00:50 > De : Mike Hearn > A : wine-devel(a)winehq.com > Copie à : > Objet : Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec) > On Wed, 17 Mar 2004 22:37:01 +0000, Christian Costa wrote: > > When executing the DllMain of the native IR32_32.DLL (VFW), Wine crashes. > > After some investigations, I found that a part of the code uses the EBX > > register at one point without saving and restore it afterwards. > > What calling convention is the entrypoint declared with? How can I retreive this info from the dll? I guess it is stdcall like it should be. > I thought EBX was > a register that had to be saved/restored according to the ABI and the > windows compilers would always do that? > I don't know but that would explain what is happening altough if it was the case I wonder why noone has encountered this problem yet. Bye, Christian

2 1

Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec)
by Christian Costa March 18, 2004

March 18, 2004

> Message du 18/03/04 01:02 > De : Alexandre Julliard > A : Christian Costa > Copie à : wined > Objet : Re: PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec) > Christian Costa writes: > > > When executing the DllMain of the native IR32_32.DLL (VFW), Wine crashes. > > After some investigations, I found that a part of the code uses the > > EBX register at one point without saving > > and restore it afterwards. Since this register contains the PLT > > pointer for standard unix functions like libc ones, this > > does not take a long time for Wine to crashing. > > Is the dll really modifying ebx explicitly? Isn't it rather a > consequence of a stack overflow or some similar bug? Yes, ebx is modified explicity and not poped from the stack. There is no save/restore statements. Bye, Christian

1 0

wineinstall hangs
by Fabian Cenedese March 18, 2004

March 18, 2004

Hi I just tried on a new computer to checkout wine from cvs and do a install. Compilation worked fine, but there was again a problem with the regedit: --- snip --- Preparing to install default Wine registry entries... Installing default Wine registry entries... Warning: Language 'us' was not recognized, defaulting to English. Could not stat /mnt/fd0 (No such file or directory), ignoring drive A: Xlib: extension "GLX" missing on display ":1.0". fixme:urlmon:URLMON_DllRegisterServer (void): stub wine: Unhandled exception (thread 0009), starting debugger... Warning: Language 'us' was not recognized, defaulting to English. Could not stat /mnt/fd0 (No such file or directory), ignoring drive A: (hang, CTRL-C) Registry entries successfully installed. err:seh:setup_exception nested exception on signal stack in thread 0009 eip 40108aa6 esp 40031c28 stack 0x405a0000-0x406a0000 Installation complete for now. Good luck (this is still alpha software). If you have problems with WINE, please read the documentation first, as many kinds of potential problems are explained there. --- snip --- But after that I could start wine regedit and the keys seem to have been generated. So I don't know what really failed. bye Fabi

3 4

Re: Avoid a possible deadlock when stopping playback
by Christian Costa March 18, 2004

March 18, 2004

Hi Dmitry, Are you sure this is usefull. How can SetEvent cause a deadlock? Bye, Christian > Message du 16/03/04 14:54 > De : Dmitry Timoshkov > A : wine-patches(a)winehq.com > Copie à : > Objet : Avoid a possible deadlock when stopping playback > Hello, > > looks like it's a good idea to do the same when signalling hStopEvent. > > Changelog: > Dmitry Timoshkov > Avoid a possible deadlock when stopping playback. > > --- cvs/hq/wine/dlls/winmm/mciavi/mciavi.c 2004-03-16 18:37:03.000000000 +0800 > +++ wine/dlls/winmm/mciavi/mciavi.c 2004-03-16 19:38:42.000000000 +0800 > @@ -606,7 +606,9 @@ static DWORD MCIAVI_mciStop(UINT wDevID, > switch (wma->dwStatus) { > case MCI_MODE_PLAY: > case MCI_MODE_RECORD: > + LeaveCriticalSection(&wma->cs); > SetEvent(wma->hStopEvent); > + EnterCriticalSection(&wma->cs); > /* fall through */ > case MCI_MODE_PAUSE: > /* Since our wave notification callback takes the lock, > > > > >

2 3

Rundll32 needs X
by Vincent Béron March 18, 2004

March 18, 2004

Hi folks, Since wine.inf is now used as the source for the default values for the registry (via rundll32), importing it absolutely needs X (whereas the older regedit method didn't when importing a .reg file). This can be a problem for people running wineinstall via ssh from a Windows box. Is there a reason rundll32 needs user32 and cannot delayimport it? Vincent

3 3

PLT pointer destroyed problem when loading IR32_32.DLL (Indeo codec)
by Christian Costa March 17, 2004

March 17, 2004

Hi, When executing the DllMain of the native IR32_32.DLL (VFW), Wine crashes. After some investigations, I found that a part of the code uses the EBX register at one point without saving and restore it afterwards. Since this register contains the PLT pointer for standard unix functions like libc ones, this does not take a long time for Wine to crashing. I tested with the ugly hack attached that saves and restores the EBX register between the call to DllMain and that fixed the problem and the DLL works fine then. Can someone tell me what happens? Thanks in advance, Christian Index: loader.c =================================================================== RCS file: /home/wine/wine/dlls/ntdll/loader.c,v retrieving revision 1.66 diff -u -r1.66 loader.c --- loader.c 16 Mar 2004 03:10:07 -0000 1.66 +++ loader.c 17 Mar 2004 21:19:39 -0000 @@ -690,7 +690,9 @@ else TRACE("(%p %s,%s,%p) - CALL\n", module, debugstr_w(wm->ldr.BaseDllName.Buffer), reason_names[reason], lpReserved ); + __asm("push %ebx"); retv = entry( module, reason, lpReserved ); + __asm("pop %ebx"); /* The state of the module list may have changed due to the call to the dll. We cannot assume that this module has not been

4 3

Re: [AppDB] - Script the table creation, help users to setup local appdb
by Paul van Schayck March 17, 2004

March 17, 2004

Chris, On Wednesday 17 March 2004 19:16, Chris Morgan wrote: > README now has some instructions on how to setup a local > appdb. You should also note that magic_quotes_gpc should be off. Otherwise you would end up inside your database with \ everywhere. And for development display_errors is very handy. And great work on those tables. I was already wondered wy I couldn't login. Regards, Paul

1 0

Win32 packages released on sourceforge
by Ivan Leo Murray-Smith March 17, 2004

March 17, 2004

The win32 packages of wine are on sourceforge, you can get them from http://sourceforge.net/project/showfiles.php?group_id=6241 or to view the win32 packages only http://sourceforge.net/project/showfiles.php?group_id=6241&package_id=112520 Ivan.

2 1

Re: Stubs needed by some Windows Update Installs
by Dmitry Timoshkov March 17, 2004

March 17, 2004

"Aric Stewart" <aric(a)codeweavers.com> wrote: > @@ -297,7 +297,7 @@ > @ stub SetupSetSourceListA > @ stub SetupSetSourceListW > @ stdcall SetupTermDefaultQueueCallback(ptr) > -@ stub SetupTerminateFileLog > +@ stdcall SetupTerminateFileLog(ptr) The argument here should be 'long'. -- Dmitry.

2 3

Current CVS msi build failure
by Ferenc Wagner March 17, 2004

March 17, 2004

Hi, with current CVS I get: [...] bison -p SQL_ -d ../../../src/dlls/msi/sql.y -o sql.tab.c ../../../src/dlls/msi/sql.y contains 4 shift/reduce conflicts and 10 reduce/reduce conflicts. bison -p SQL_ -d ../../../src/dlls/msi/sql.y -o sql.tab.c ../../../src/dlls/msi/sql.y contains 4 shift/reduce conflicts and 10 reduce/reduce conflicts. gcc -c -I../../../src/dlls/msi -I. -I../../../src/include -I../../include -D__WINESRC__ -D_REENTRANT -fPIC -Wall -pipe -mpreferred-stack-boundary=2 -fno-strict-aliasing -gstabs+ -Wpointer-arith -g -O2 -o tokenize.o ../../../src/dlls/msi/tokenize.c ../../../src/dlls/msi/tokenize.c:58: `TK_CHAR' undeclared here (not in a function) ../../../src/dlls/msi/tokenize.c:58: initializer element is not constant ../../../src/dlls/msi/tokenize.c:58: (near initialization for `aKeywordTable[12].tokenType') ../../../src/dlls/msi/tokenize.c:59: `TK_CHAR' undeclared here (not in a function) ../../../src/dlls/msi/tokenize.c:59: initializer element is not constant ../../../src/dlls/msi/tokenize.c:59: (near initialization for `aKeywordTable[13].tokenType') ../../../src/dlls/msi/tokenize.c:90: `TK_HOLD' undeclared here (not in a function) ../../../src/dlls/msi/tokenize.c:90: initializer element is not constant [etc., similar ones for TK_INT, TK_LONG, TK_LONGCHAR, TK_OBJECT and TK_SHORT] These IDs are really missing from my sql.tab.h. Any clues? -- Feri.

2 2

Re: Stub for SignalObjectAndWait
by Mike Hearn March 17, 2004

March 17, 2004

On Wed, 17 Mar 2004 01:50:40 +0100, Raphael wrote: > PS: i'm trying to get MSN Messenger working, how we can register some com dlls > into registry ? The app installer should do that for you (last time I tried MSN Messenger I didn't have any problems with missing registrations) but you can run regsvr32 on a DLL to reregister it if you want to. thanks -mike

1 0

Re: REBAR: prevent recursions with WM_SIZE reason
by Mike McCormack March 17, 2004

March 17, 2004

Alexandre, please don't apply this patch... it doesn't work the right way. Juan Lang pointed out that I have a = instead of a |= ... and after fixing it, it doesn't do what it mean to do. thanks, Mike Mike McCormack wrote: > > ChangeLog: > * prevent recursions with WM_SIZE reason

1 0

Re: tools/winetest winetest.cron
by Ferenc Wagner March 16, 2004

March 16, 2004

Jeremy Newman <jnewman(a)winehq.org> writes: > ChangeSet ID: 11483 > > Log message: > set the lang Did you check if it was needed? I thought my patch took care of this in a language/locale neutral way by opening the report file in :raw mode. I'd like to not if it's not the case. -- Feri.

2 1

Browsing the wine source
by Blake Leverett March 16, 2004

March 16, 2004

Hello, I have been getting wine from CVS for a while now, and have been looking through the source. How do people who work with the wine code browse around in the source tree? I found that 'ctags -e -R' can be used with emacs, which works OK, but I was wondering if there is a better method. Thanks, Blake Leverett

2 1

Re: cards.dll improvements
by Tom March 16, 2004

March 16, 2004

Sam wrote: > Improvements to cards.dll based on documentation at > http://www.microsoft.com/mind/0396/games.asp. > I'm not sure if you have read this post or not but here it is. http://www.winehq.com/hypermail/wine-devel/2003/07/0691.html Tom

1 0

how to call setupapi / wine.inf
by Marcus Meissner March 16, 2004

March 16, 2004

Hi, I am trying to setup wine with wine.inf, with the following call: /usr/bin/rundll32 setupapi.dll,InstallHinfSection DefaultInstall 128 /usr/share/wine/wine.inf However, the /usr/share/wine/wine.inf path does not seem to be translated correctly. 0009:Call kernel32.GetFullPathNameW(406efb6e L"/usr/share/wine/wine.inf",00000000,00000000,00 000000) ret=40c8333c 0009:Ret kernel32.GetFullPathNameW() retval=0000001b ret=40c8333c 0009:Call ntdll.RtlAllocateHeap(40360000,00000000,00000036) ret=40c83542 0009:Ret ntdll.RtlAllocateHeap() retval=4039a0e0 ret=40c83542 0009:Call kernel32.GetFullPathNameW(406efb6e L"/usr/share/wine/wine.inf",0000001b,4039a0e0,00 000000) ret=40c83571 0009:Ret kernel32.GetFullPathNameW() retval=0000001a ret=40c83571 0009:Call kernel32.CreateFileW(4039a0e0 L"Y:\\usr\\share\\wine\\wine.inf",80000000,00000001,0 0000000,00000003,00000000,00000000) ret=40c835af 0009:Ret kernel32.CreateFileW() retval=ffffffff ret=40c835af The config script has: [Drive Y] "Type" = "network" "Path" = "%HOME%" [Drive Z] "Type" = "network" "Path" = "/" So it should have been using Z:\\ Is this a problem in GetFullPathName or should I better use: /usr/bin/rundll32 setupapi.dll,InstallHinfSection DefaultInstall 128 z:\\usr\\share\\wine\\wine.inf ? Ciao, Marcus

4 4

dlls/ntdll/directory.c portability breakage
by Gerald Pfeifer March 16, 2004

March 16, 2004

The new file dlls/ntdll/directory.c added by date: 2004/03/16 01:32:02; author: julliard; state: Exp; Implemented NtQueryDirectoryFile (partly based on a patch by Eric Pouech). broke non-Linux platforms (like FreeBSD): /usr/bin/gcc -c -I. -I. -I../../include -I../../include -D__WINESRC__ -D_NTSYSTEM_ -D_REENTRANT -fPIC -Wall -pipe -mpreferred-stack-boundary=2 -fno-strict-aliasing -gstabs+ -Wpointer-arith -g -O2 -o directory.o directory.c directory.c: In function `NtQueryDirectoryFile': directory.c:446: error: `O_DIRECTORY' undeclared (first use in this function) directory.c:446: error: (Each undeclared identifier is reported only once directory.c:446: error: for each function it appears in.) There *is* code to #define O_DIRECTORY /* To avoid blocking on non-directories in DOSFS_OpenDir_VFAT*/ #ifndef O_DIRECTORY # define O_DIRECTORY 0200000 /* must be directory */ #endif but it's in #ifdef linux only! Gerald -- Gerald Pfeifer (Jerry) gerald(a)pfeifer.com http://www.pfeifer.com/gerald/

1 0

Re: cards.dll
by Joerg Mayer March 16, 2004

March 16, 2004

On Tue, Mar 16, 2004 at 12:09:24AM +0200, Shachar Shemesh wrote: > Can someone remind me again why we can't have GPL DLLs in the tree? > Doesn't the "call through documented interface blah blah blah not > derived work" argument cause the license for each two DLLs in Wine to be > independant? Basically it goes like this: LGPLed code can be run as a library, i.e. a function of an LGPLed dll can be called as a function call. GPLed code must be run as a (Unix) process if you want to use it from non-GPL-compatible code. See the GPL and especially the GPL FAQ for details. Ciao Jörg -- Joerg Mayer <jmayer(a)loplof.de> We are stuck with technology when what we really want is just stuff that works. Some say that should read Microsoft instead of technology.

2 1

Re: shell32 patch 23 d [correction]
by Alexandre Julliard March 15, 2004

March 15, 2004

"Martin Fuchs" <martin-fuchs(a)gmx.net> writes: > Alexandre - is there any thing wrong with this patch? Yes there has to be something wrong, you create a sei_tmp structure but you never actually use it. I'm not sure what your intent was. -- Alexandre Julliard julliard(a)winehq.org

1 0