August 2004 - Wine-devel - list.winehq.org

Re: mixing SHELL32.DLL and SHFOLDER.DLL
by Juan Lang 31 Aug '04

31 Aug '04

Salius wrote: > I am sure shfolder.dll from win98se installation > calls GetProcAddress() to get the address of > SHGetFolderPathW @ shell32.dll. Moreover, my > shell32.dll doesn't export SHGetSpecialFolderPath* > by a name, also. Hm. I'm sorry, I wasn't clear enough. From MSDN on SHGetFolderPath: This function is a superset of SHGetSpecialFolderPath, included with earlier versions of the Shell. On systems preceeding those including Shell32.dll version 5.0 (Windows Millennium Edition (Windows Me) and Windows 2000), SHGetFolderPath was obtained through SHFolder.dll, distributed with Microsoft Internet Explorer 4.0 and later versions. SHFolder.dll always calls the current platform's version of this function. If that fails, it will try to simulate the appropriate behavior. So, I think your patch is a good first step, to use GetProcAddress rather than letting the linker do it. A next step would be, I think, to do what I suggested: if this fails, try to GetProcAddress for SHGetSpecialFolderPath, since this was available on older versions of shell32. You might have to search by ordinal if searching by name fails. So, I don't think my suggestion is that different from Juergen's, just that I suggest an implementation of "try to simulate the appropriate behavior" if the function is not in fact in shell32. --Juan _______________________________ Do you Yahoo!? Win 1 of 4,000 free domain names from Yahoo! Enter now. http://promotions.yahoo.com/goldrush

3 3

E-mail adress
by Davide Giannotti 31 Aug '04

31 Aug '04

I'm trying to fight spam like anybody else, and first thing is to remove mail adresses from every page on the internet. A google search on my adress shows only 2 pages where i compare, they both come frome wine archive: http://www.winehq.com/hypermail/wine-devel/2003.04.txt http://www.spinics.net/lists/wine/msg08863.html Is there a way to make them remove from there? Just a suggestion: i'm conscious that mail adress need to be post into such a mailing list, but wouldn't be good to keep a script running that alter them into something like 'myname {at} mydomain' ? Just a little sacrifice for everybody to help fighting spam. Think about.. pls... -- Davide Giannotti ITS Planet s.n.c davide.giannotti {at} itsplanet.com www.itsplanet.com

3 2

Re: Set last error values compatible with Win9x implementation of GetFileVersionInfoSize
by Uwe Bonnes 31 Aug '04

31 Aug '04

>>>>> "Dmitry" == Dmitry Timoshkov <dmitry(a)baikal.ru> writes: Dmitry> Hello, installer of Quicken depends on last error values set by Dmitry> GetFileVersionInfoSize and expects them to be set to something Dmitry> different than ERROR_RESOURCE_DATA_NOT_FOUND set on NT Dmitry> platforms. Dmitry> I hope that there is no need to introduce several code paths Dmitry> depending on the GetVersion return value. A test programm for our test suite will easily show (and document) that behaviour. Bye -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

2 1

Re: KERNEL32: implement CallNamedPipe
by Dmitry Timoshkov 31 Aug '04

31 Aug '04

"Mike McCormack" <mike(a)codeweavers.com> wrote: > + r = WaitNamedPipeW( lpNamedPipeName, nTimeout ); > + if( !r ) > + return r; win2k kernel32 does not do WaitNamedPipeW. > + handle = CreateFileW( lpNamedPipeName, GENERIC_READ|GENERIC_WRITE, > + 0, NULL, OPEN_EXISTING, 0, NULL ); > + if( !handle ) > + return FALSE; CreateFileW returns INVALID_HANDLE_VALUE on failure, not 0. > + r = WriteFile( handle, lpOutput, dwOutputSize, &count, NULL ); > + if( r ) > + r = ReadFile( handle, lpInput, dwInputSize, lpBytesRead, NULL ); win2k kernel32 calls SetNamedPipeHandleState and TransactNamedPipe (not implemented in Wine) here. > + CloseHandle( handle ); > + > + return r; > } -- Dmitry.

1 0

Re: Add some padding in WINPROC_wrapper
by Mike McCormack 31 Aug '04

31 Aug '04

Uwe Bonnes wrote: > Changelog > windows/winproc.c: WINPROC_wrapper > Add some padding between saved registers and the winproc > arguments. Some winprocs may trash the register area. > > As Mike Hearn suggested, adding some padding helped with my problem with > WebPACK_42wp30_full_installer.exe. I hope, I found the right way to add that > padding. It may work for one application, but it doesn't look like a propper fix. I'd suggest there's probably something else going on that's causing stack corruption. Mike

3 7

Re: wineserver epoll
by Alexandre Julliard 31 Aug '04

31 Aug '04

Shachar Shemesh <wine-patches(a)shemesh.biz> writes: > After much talk, the patch seems ready for commit. Since there are many > systems with slightly older glibc that have the header but do not have > the implementation in glibc, or that have a stub implementation that is > guaranteed to fail, we are calling the kernel functions directly. This > also allows compiling on older systems than the code actually ships to. > > Accordingly, the autoconf checks for epoll_create in glibc were dropped. That's wrong; if the libc function exists you should use it, direct system calls should only be used as a fallback. And when using system calls you must not depend on the libc header, you have to define the structures yourself. -- Alexandre Julliard julliard(a)winehq.org

5 7

Debug Channels trashed
by Uwe Bonnes 31 Aug '04

31 Aug '04

Hallo, I am trying to run the WebPACK_42wp30_full_installer.exe : http://direct.xilinx.com/direct/webpack/42wp3/\ : WebPACK_42wp30_full_installer.exe (a 110 MByte Download). Running it crashes quite early, while testing the debug channels: err:font:ReadFontDir Can't open directory "/usr/X11R6/lib/X11/fonts/TrueType" err:font:ReadFontDir Can't open directory "/usr/X11R6/lib/X11/fonts/TT" wine: Unhandled exception (thread 0009), starting debugger... fixme:console:SetConsoleCtrlHandler (0x405ac2f0,1) - no error checking or testing yet WineDbg starting on pid 0x8 Unhandled exception: page fault on read access to 0xc0685566 in 32-bit code (0x40710c60). In 32 bit mode. 0x40710c60 WINPROC_CallWndProc [/home/wine/wine/dlls/user/../../windows/winproc.c:232] in user32: testb $0x8,0x0(%eax) 232 if (TRACE_ON(relay)) 232 if (TRACE_ON(relay)) Wine-dbg>Backtrace: =>1 0x40710c60 WINPROC_CallWndProc+0x70(wParam=0x1002c, lParam=0x0) [/home/wine/wine/dlls/user/../../windows/winproc.c:232] in user32 (0x4067fb7c) 2 0x4071578d CallWindowProcA+0xed(func=0x4078656c, hwnd=0x10022, msg=0x110, wParam=0x1002c, lParam=0x0) [/home/wine/wine/dlls/user/../../windows/winproc.c:3123] in user32 (0x4067fba0) 3 0x406eb083 DefDlgProcA+0x63(hwnd=0x10022, msg=0x110, wParam=0x1002c, lParam=0x0) [/home/wine/wine/dlls/user/../../windows/defdlg.c:451] in user32 (0x4067fbd0) 4 0x40710437 WINPROC_wrapper+0x17 in user32 (0x4067fbf4) 5 0x40710c4c WINPROC_CallWndProc+0x5c(wParam=0x1002c, lParam=0x0) [/home/wine/wine/dlls/user/../../windows/winproc.c:230] in user32 (0x4067fc30) 6 0x4071578d CallWindowProcA+0xed(func=0x406eb020, hwnd=0x10022, msg=0x110, wParam=0x1002c, lParam=0x0) [/home/wine/wine/dlls/user/../../windows/winproc.c:3123] in user32 (0x4067fc54) 7 0x40725b8f call_window_proc+0x11f(lparam=0x0, unicode=0x0, same_thread=0x1) [/home/wine/wine/dlls/user/message.c:1461] in user32 (0x4067fcb8) 8 0x4072839e SendMessageTimeoutA+0x1be(hwnd=0x10022, msg=0x110, wparam=0x1002c, lparam=0x0, flags=0x0, timeout=0xffffffff, res_ptr=0x4067fd3c) [/home/wine/wine/dlls/user/message.c:1968] in user32 (0x4067fd18) 9 0x40728425 SendMessageA+0x35(hwnd=0x10022, msg=0x110, wparam=0x1002c, lparam=0x0) [/home/wine/wine/dlls/user/message.c:2012] in user32 (0x4067fd44) 10 0x406efc5c DIALOG_CreateIndirect+0x107c(dlgProc=0x4012df, param=0x0, unicode=0x0, modal=0x1) [/home/wine/wine/dlls/user/../../windows/dialog.c:672] in user32 (0x4067fe54) 11 0x406f04be DialogBoxParamA+0x4e(hInst=0x400000, name=0x12c, owner=0x0, dlgProc=0x4012df, param=0x0) [/home/wine/wine/dlls/user/../../windows/dialog.c:821] in user32 (0x4067fe78) 12 0x0040184d (0x004137a2) It seems the location of the debug channels is picked up from the ebx register Wine-dbg>disass 0x40710c52 0x40710c52 WINPROC_CallWndProc [../windows/winproc.c:231] in user32: call 0x40707c70 WIN_RestoreWndsLock [/home/wine/wine/dlls/user/../../windows/user.c:183] in user32 0x40710c57 WINPROC_CallWndProc [../windows/winproc.c:232] in user32: movl 0x190(%ebx),%eax 0x40710c5d WINPROC_CallWndProc [../windows/winproc.c:232] in user32: addl $24,%esp 0x40710c60 WINPROC_CallWndProc [../windows/winproc.c:232] in user32: testb $0x8,0x0(%eax) I wonder what can trash the testing of the debug channels. I ran with +heap to look for heap corruption, but couldn't find something. Uncommenting the offending code > if (TRACE_ON(relay)) > DPRINTF( "%04lx:Ret window proc %p (hwnd=%p,msg=%s,wp=%08x,lp=%08lx) retval=%08lx\n", > GetCurrentThreadId(), proc, hwnd, SPY_GetMsgName(msg, hwnd), wParam, lParam, retvalue ); makes the application proceed. Bye -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

4 5

Linux registry
by Robert van Herk 31 Aug '04

31 Aug '04

Not knowing how much off topic this would be, I just thought I paste a link here: http://registry.sourceforge.net/ These people seem to be implementing a kinda Windows registry clone into Linux, but then secure, stable 'n' stuff... Since the many registry implementation discussions here, I thought, why not throw this into the discussion too? It might be worth to take a look here and perhaps think about using some of their stuff. Greetings, Robert van Herk

3 2

Re: SHAllocShared fixes
by Alexandre Julliard 30 Aug '04

30 Aug '04

Francois Gouget <fgouget(a)free.fr> writes: > This actually calls for two more notes: > * This looks very much what one would get by calling the corresponding > functions in shlwapi (which seems much more complete). However in > shell32 there is talk of sending messages around (WM_USER+2) so maybe > it's the same concept but a different implementation? The WM_USER+2 thing is something a specific caller was doing, it's not part of the function itself. I think the shell32 functions should simply forward to the shlwapi ones. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: mixing SHELL32.DLL and SHFOLDER.DLL
by Juan Lang 30 Aug '04

30 Aug '04

Saulius wrote: > Lets try. I am not sure about many things: > > 1, whether is better to duplicate code > SHGetFolderPathW() or to try finding it by an > ordinal its value (the case for the older versions > of SHELL32.DLL) as it was stated by Juergen? What are you trying to accomplish? For example, it seems reasonable to assert that you must use the builtin shell32 and shfolder.dll in combination, or not at all, as Alexandre once suggested. Or, in order for Wine's shfolder.dll to succeed on older (native) versions of shell32.dll, it seems reasonable to have it call SHGetSpecialFolderPath* from SHGetFolderPath*, as MS' version does. In other words, why do you need the combination of native shell32 and builtin shfolder.dll? --Juan __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! http://promotions.yahoo.com/new_mail

3 6