March 2003 - Wine-Devel - list.winehq.org

(no subject)
by Alireza Mahini March 9, 2003

March 9, 2003

I would like to use a win32 DLL in what is otherwise a native Linux application. I want to do this by using WINE. According to the documentation that comes with the DLL, a Windows program can load it with the following code (the names have been changed to protect the innocent -- me): HMODULE hDLL; FARPROC lpfnFunc1, lpfnFunc2; hDLL = LoadLibrary ("library.dll"); if (hDLL == (HMODULE) NULL) return ERROR; lpfnFunc1 = (LPAPPFUNC1) GetProcAddress (m_hDLL, "Function1"); lpfnFunc2 = (LPAPPFUNC2) GetProcAddress (m_hDLL, "Function2"); I created the following test program: #include <wine/wine/winbase.h> int main (int argc, const char *arvg[]) { HMODULE hDLL; hDLL = LoadLibrary ("library.dll"); if (hDLL == (HMODULE) NULL) return 1; else return 0; } When I try to build it, however, I get an "undefined reference to 'LoadLibraryA'". Linking libwine.so does not fix this.Do i try with additional flags for g++ compiler? Or My code is wrong !?please ,guide me to solve this problem .RegardsAlireza Mahini _______________________________________________ Join Excite! - http://www.excite.com The most personalized portal on the Web!

1 0

winedbg trouble
by Dan Kegel March 8, 2003

March 8, 2003

OK, finally tried the new winedbg, and I'm afraid I'm mystified. I wanted to debug my pipe regression test. I tried winedbg pipe.exe This starts out ok, printing stuff to stdout like I expect, but then it does something very odd: it switches to reverse video right before it prints "In 32-bit mode." The "Wine-dbg>" prompt comes up, but it doesn't respond to the keyboard. Also, the display is stair-stepped, as if the newline char isn't being interpreted as CR-LF. Sounds like stdin is being put in raw mode. I have to do 'killall wine' from another window. It's just totally, mysteriously broken... Before, I could actually use it, but now I'm just confused. On the theory that for some unfathomable reason, the correct command is now wineconsole winedbg pipe.exe I tried that. That behaved somewhat better in that it responded to keystrokes, but it scrolls at line 25 no matter what size the window is, and trying to do 'cont' or 'run' fails with messages like No process loaded, cannot execute 'cont' And it's in that horrible reverse video. Why can't winedbg just behave like gdb with respect to stdin and stdout? I'm at a loss. - Dan -- Dan Kegel http://www.kegel.com http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045

2 5

Re: wineconsole
by Andreas Mohr March 8, 2003

March 8, 2003

On Sat, Mar 08, 2003 at 09:23:59PM +0100, Eric Pouech wrote: > another regression fixed... > > this time sounds like I forgot to put back this line in after a couple > of other testings... :-/ > > this would make Dan even angrier, but also get rid of the several > win32-consoles on the same unix-terminal issue Dan was > experie^H^H^H^H^H^H^H banging his head into While you are at it, you could also fix the busy-loop in the cmdline parsing resulting from feeding an unknown parameter string ;-) (such as --BaCkEnD=XXX) (better use automatic advancing to the next argument instead of only manually advancing if the parsing of the last parameter was successful) -- Help prevent Information Technology Fascism! - before it's too late... http://www.againsttcpa.com

1 0

Crash in DirectSound
by Mike Hearn March 8, 2003

March 8, 2003

Hi, There's a crash on line 207 of mixer.c, the second memcpy in the code that deals with buffer wrapping. I'm afraid I can't reliably get a useful stack trace, for some reason winedbg goes crazy with this app, I was just trying and couldn't reproduce the trace I got earlier. Anyway, if there are any DirectSound gurus reading this, do you have any advice on what might be causing the crash? thanks -mike

1 0

Should crosstesting work?
by Ferenc Wagner March 8, 2003

March 8, 2003

Hello, I'm trying to cross-compile some conformance tests. The Winelib tests work, Wine runs the cross-compiled tests, I can produce Win32 apps with MinGW, but I can't run the cross-compiled tests under Win95. Simply running them in a DOS window produces the list of possible tests as expected, but running a real test fails with an Illegal Operation window. What am I missing here? Thanks: Feri.

4 8

err:dc:CreateDCA no driver found for ,,
by Gert Bartho March 8, 2003

March 8, 2003

Dear All, I am not sure whether I have picked the correct mailing-list for my question, please let me know. Okay, yesterday I installed Wine (Wine 20030219) on my Linux Box (RH8.0). I would like to run the Windows programs without Windows partition, which I do not have by the way. First I installed Microsoft Bookshelf (British Reference Collection), and that went almost straight out of the box. However, another programs almost works -- I hope. After placing the correct DLL's in the program's directory, I stumbled upon the following message (screen output): --------------------------------------- [gb(a)localhost vandale]$ wine VDEN32.exe fixme:win32:PE_CreateModule Security directory ignored err:x11drv:X11DRV_CreateWindow invalid window height -335345751 err:x11drv:X11DRV_CreateWindow invalid window height -335345751 fixme:winspool:OpenPrinterW (printerName: NULL, pDefault (nil) Ret: False err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, wine: Unhandled exception, starting debugger... err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:dc:CreateDCA no driver found for ,, err:seh:EXC_DefaultHandling Unhandled exception code c0000096 flags 0 addr 0x40e3278f Terminated --------------------------------------- (Note I deleted a couple of the 'err:dc:CreateDCA no driver found for ,,' lines to save space. Does anyone have a suggestion? I have searched with google for all the errors, but I could not find anything substantial that helped me. I would be very pleased if someone could point me in the right direction. Best wishes Gerard _________________________________________________________________ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=features/junkmail

1 0

Re: (4th) Janitorial dlls/advapi32/registry.c W->A cleanup
by Dimitrie O. Paun March 8, 2003

March 8, 2003

On March 7, 2003 05:09 pm, Tony Lambregts wrote: > This one is implementaed using Dimi's wrapper function *asciiz_to_unicode It ain't mine, it's Dmitry's :) -- Dimi.

1 0

Re: (4th) Janitorial dlls/advapi32/registry.c W->A cleanup
by Alexandre Julliard March 8, 2003

March 8, 2003

Tony Lambregts <tony_lambregts(a)telusplanet.net> writes: > + len = WideCharToMultiByte(CP_ACP, 0, nameW, -1, NULL, 0, NULL, NULL); > + nameA = HeapAlloc(GetProcessHeap(), 0, len); > + WideCharToMultiByte(CP_ACP, 0, nameW, -1, nameA, len, NULL, NULL); > + > for (;;) > { > - sprintf( name, "reg%04x.tmp", count++ ); > - handle = CreateFileA( buffer, GENERIC_WRITE, 0, NULL, > + sprintf( nameA, "reg%04x.tmp", count++ ); > + handle = CreateFileW( buffer, GENERIC_WRITE, 0, NULL, This can't work, the sprintf is supposed to modify the buffer, not a copy of it. Also note that registry functions are not supposed to set last error, so you probably don't want to use Dmitry's wrapper here. -- Alexandre Julliard julliard(a)winehq.com

3 3

Re: Adding shmserver (not subscribed)
by juan＠foxpaw.com March 7, 2003

March 7, 2003

Regarding the comments about shmserver's safety, I wanted to point out that running the shmserver doesn't mean that you have to run multiple processes in it. You can still run multiple wine instances. The shmserver shows a performance improvement for a single application (granted special kinds of applications), not just for several. The shmserver reminds me of how Windows NT and OS/2 handle 16-bit apps: they can each run in their own process, or can several can run in the same process to improve performance. I realize this comparison isn't exact, but I think the analogy is apt. Why not include both servers? If, as Gavriel asserts, the changes to implement the shmserver can be small, maintaining both in the source tree shouldn't be horrendous. --Juan

1 0

MSVCIRT and _pioinfo()
by Uwe Bonnes March 7, 2003

March 7, 2003

Hallo, the MS Standard C++ Libraries implementation in MSVCIRT directly accesses low level MSVCRT structures, like _pioinfo(), the pointer to an array of ioinfo structures. A google/search search brings up some information, e.g. (*1). Some program in the Xilinx suite silently crashes when this _pioinfo() is accessed inside MSVCIRT with builtin msvcrt. How should we handle this case? Implement this array, in the same way as Microsoft did? Or can we construct a builtin MSVCIRT perhaps by using wrappers around some other Standard C++ Libraries implementation? At least a more visible way to crash when unimplemted structures are accessed would be fine. Comments welcome. Bye (*1): from: http://www.geocrawler.com/archives/3/1783/2001/8/0/6525036/ (squid cvs) +/* internal to CTRLIB */ +typedef struct { + long osfhnd; /* underlying OS file HANDLE */ + char osfile; /* attributes of file (e.g., open in text mode?) */ + char pipech; /* one char buffer for handles opened on pipes */ +#ifdef _MT + int lockinitflag; + CRITICAL_SECTION lock; +#endif /* _MT */ + } ioinfo; +extern _CRTIMP ioinfo * __pioinfo[]; +#define IOINFO_L2E 5 +#define IOINFO_ARRAY_ELTS (1 < IOINFO_L2E) +#define _pioinfo(i) ( __pioinfo[(i) > IOINFO_L2E] + ((i) &(IOINFO_ARRAY_ELTS - \ + 1)) ) +#define _osfile(i) ( _pioinfo(i)->osfile ) + -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0