September 2003 - Wine-Devel

Regedit
by Ivan Leo Murray-Smith Sept. 15, 2003

Sept. 15, 2003

Hi list, I'm translating regedit to italian, and I wanted to make it as similar as possible to the windows regedit. The only problem is that the wine regedit is based on the english m$ regedit, but m$ has written the italian version in a slightly different way. For example, what in the m$ english version and in the wine version is import registry file in the m$ italian version is translated into import file of the configuration registry Now I'm not sure if I should base the italian translation of the wine registry on the english version of the registry(I suppose it would be more consistent with other translations), or if I should translate it as m$ has done(What I think is best, as any user coming from an italian windows installation would find familiar menus), any advice would be greatly appreciated.

2 1

keyboard broken in DOS (real mode)
by Michael Stefaniuc Sept. 15, 2003

Sept. 15, 2003

Hello, the installer of the free IDA pro version (http://www.simtel.com/pub/pd/27629.html) dosn't work with wine due missing keyboard input. Pressing some keys makes wine to spit out following error: err:int:DOSVM_Int09SendScan keyboard queue overflow The emulated hardware keyboard buffer overflows because the keys don't make it to the emulated BIOS keyboard buffer. DOSVM_Int09SendScan queues an event for IRQ1 but no handler for int 09 is ever called (the int 09 handler copies the keys from the hardware to the BIOS keyboard buffer). (Thx to Ove for the explanation how things should work) >From the +int,+int21,+relay debug output it seems like the event queued by DOSVM_Int09SendScan is stuck in the queue. I had a look at DOSVM_QueueEvent and the other dosvm and interrupt functions but i didn't see the light of how this stuff works and why the queued event is never processed. Some help with that would be grateful. Btw the keyboard input seems to work in protected mode (tested with the setup program of Imperium Romanum using the DOS32A extender). bye michael -- Michael Stefaniuc Tel.: +49-711-96437-199 System Administration Fax.: +49-711-96437-111 Red Hat GmbH Email: mstefani(a)redhat.com Hauptstaetterstr. 58 http://www.redhat.de/ D-70178 Stuttgart

1 0

load library problem
by Dimitrie O. Paun Sept. 14, 2003

Sept. 14, 2003

Hi folks, I have flagged this before, but got no answer. Once again, this a C++ Winelib app (one of the wxWindows samples): [dimi(a)dimi calendar]$ ls calendar* calendar calendar_calendar.o calendar.cpp calendar.exe.so calendar.o calendar.rc calendar-wrap.dll.so calendar.bkl calendar_calendar_rc.o calendar.dsp calendarM5.xml calendar.pro calendar_resources.o [dimi(a)dimi calendar]$ export WINEDLLPATH=`pwd` [dimi(a)dimi calendar]$ echo $WINEDLLPATH /home/dimi/dev/wine/wxWindows/samples/calendar [dimi(a)dimi calendar]$ wine --debugmsg +module,+loaddll calendar.exe.so 2>log [dimi(a)dimi calendar]$ tail log trace:module:GetModuleFileNameW L"F:\\dev\\wine\\wxWindows\\samples\\calendar\\calendar.exe" trace:module:GetModuleFileNameW L"F:\\dev\\wine\\wxWindows\\samples\\calendar\\calendar.exe" trace:module:MODULE_GetLoadOrder looking for C:\WINDOWS\SYSTEM\calendar-wrap.dll trace:module:GetModuleFileNameW L"F:\\dev\\wine\\wxWindows\\samples\\calendar\\calendar.exe" trace:module:open_app_key searching 'calendar-wrap' in AppDefaults\calendar.exe\DllOverrides trace:module:MODULE_GetLoadOrder got standard wildcard "b,n" for "calendar-wrap.dll" trace:module:load_dll Trying built-in 'C:\WINDOWS\SYSTEM\calendar-wrap.dll' trace:module:load_dll Trying native dll 'C:\WINDOWS\SYSTEM\calendar-wrap.dll' warn:module:load_dll Failed to load module 'C:\WINDOWS\SYSTEM\calendar-wrap.dll'; status=c000000f Error: Could not load the application's module calendar-wrap.dll (126) The piece of code that fails is from the automatically generated wrapper[1]. Here is the part that fails: LoadLibrary("calendar-wrap.dll"); I have the WINEDLLPATH set, what gives? Alexandre, I can't really track this easily back, but I have a gut feeling it got triggered by one of the reorganizations that you did around the loader... [1] That is how the app was linked: [dimi(a)dimi calendar]$ winewrap -v -k -mgui -C -o calendar.exe calendar_calendar.o calendar_calendar_rc.o -L../../lib -lwx_msw-2.5 -lwxexpat -lz -lrpcrt4 -loleaut32 -lole32 -lwine_uuid -lwinspool -lwinmm -lshell32 -lcomctl32 -lcomdlg32 -lctl3d32 -ladvapi32 -lwsock32 -lgdi32 -lpng -ljpeg -ltiff -lcomdlg32 -ladvapi32 -lshell32 Creating file /tmp/wapp9rVK4g.spec winebuild -o /tmp/wapp9rVK4g.spec.c -F calendar-wrap.dll --spec /tmp/wapp9rVK4g.spec -L/usr/local/lib/wine -L../../lib -lrpcrt4 -loleaut32 -lole32 -lwinspool -lwinmm -lshell32 -lcomctl32 -lcomdlg32 -lctl3d32 -ladvapi32 -lwsock32 -lgdi32 -lcomdlg32 -ladvapi32 -lshell32 -lgdi32 -luser32 -lkernel32 calendar_calendar.o calendar_calendar_rc.o ../../lib/libwx_msw-2.5.a ../../lib/libwxexpat.a /usr/lib/libz.a /usr/local/lib/libwine_uuid.a ../../lib/libpng.a ../../lib/libjpeg.a ../../lib/libtiff.a gcc -fPIC -o /tmp/wapp9rVK4g.spec.o -c /tmp/wapp9rVK4g.spec.c g++ -shared -Wl,-Bsymbolic,-z,defs -lwine -lm -L/usr/local/lib/wine -L../../lib -o calendar-wrap.dll.so /tmp/wapp9rVK4g.spec.o calendar_calendar.o ../../lib/libwx_msw-2.5.a ../../lib/libwxexpat.a /usr/lib/libz.a /usr/local/lib/libwine_uuid.a ../../lib/libpng.a ../../lib/libjpeg.a ../../lib/libtiff.a Creating file /tmp/wwrpBOB32M.c gcc -fPIC -I/usr/local/include/wine/windows -o /tmp/wwrpBOB32M.o -c /tmp/wwrpBOB32M.c winebuild -o /tmp/wwrpBOB32M.spec.c --exe calendar.exe -mgui /tmp/wwrpBOB32M.o -L/usr/local/lib/wine -L../../lib -lrpcrt4 -loleaut32 -lole32 -lwinspool -lwinmm -lshell32 -lcomctl32 -lcomdlg32 -lctl3d32 -ladvapi32 -lwsock32 -lgdi32 -lcomdlg32 -ladvapi32 -lshell32 -lgdi32 -luser32 -lkernel32 gcc -fPIC -o /tmp/wwrpBOB32M.spec.o -c /tmp/wwrpBOB32M.spec.c g++ -shared -Wl,-Bsymbolic,-z,defs -lwine -ldl -o calendar.exe.so /tmp/wwrpBOB32M.spec.o /tmp/wwrpBOB32M.o Creating file calendar [dimi(a)dimi calendar]$ cat /tmp/wapp9rVK4g.spec @ stdcall WinMain(ptr long ptr long) WinMain [dimi(a)dimi calendar]$ cat /tmp/wwrpBOB32M.c /* * Copyright 2000 Francois Gouget <fgouget(a)codeweavers.com> for CodeWeavers * Copyright 2002 Dimitrie O. Paun <dpaun(a)rogers.com> */ #include <stdio.h> #include <windows.h> /* * Describe the wrapped application */ /* The app name */ #define APPNAME "calendar" /** * This is either 0 for a console based application or * 1 for a regular windows application. */ #define GUIEXE 1 /** * This is the name of the library containing the application, * e.g. 'hello-wrap.dll' if the application is called 'hello.exe'. */ static char* appName = APPNAME "-wrap.dll"; /** * This is the name of the application's Windows module. If left NULL * then appName is used. */ static char* appModule = NULL; /** * This is the application's entry point. This is usually 'WinMain' for a * gui app and 'main' for a console application. */ #if GUIEXE static char* appInit = "WinMain"; #else static char* appInit = "main"; #endif /** * This is either non-NULL for MFC-based applications and is the name of the * MFC's module. This is the module in which we will take the 'WinMain' * function. */ static char* mfcModule = NULL; void error(const char *format, ...) { va_list ap; char msg[4096]; va_start(ap, format); vsnprintf(msg, sizeof(msg), format, ap); fprintf(stderr, "Error: %s\n", msg); /*MessageBox(NULL, msg, "Error", MB_OK);*/ va_end(ap); exit(1); } #if GUIEXE typedef int WINAPI (*WinMainFunc)(HINSTANCE hInstance, HINSTANCE hPrevInstance, PSTR szCmdLine, int iCmdShow); #else typedef int WINAPI (*MainFunc)(int argc, char** argv, char** envp); #endif #if GUIEXE int WINAPI WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, PSTR szCmdLine, int iCmdShow) #else int WINAPI main(int argc, char** argv, char** envp) #endif { HINSTANCE hApp = 0, hMFC = 0, hMain = 0; void* appMain; int retcode; /* Then if this application is MFC based, load the MFC module */ if (mfcModule) { hMFC = LoadLibrary(mfcModule); if (!hMFC) error("Could not load the MFC module %s (%d)", mfcModule, GetLastError()); /* For MFC apps, WinMain is in the MFC library */ hMain = hMFC; } /* Load the application's module */ if (!appModule) appModule = appName; hApp = LoadLibrary(appModule); if (!hApp) error("Could not load the application's module %s (%d)", appModule, GetLastError()); if (!hMain) hMain = hApp; /* Get the address of the application's entry point */ appMain = GetProcAddress(hMain, appInit); if (!appMain) error("Could not get the address of %s (%d)", appInit, GetLastError()); /* And finally invoke the application's entry point */ #if GUIEXE retcode = (*((WinMainFunc)appMain))(hApp, hPrevInstance, szCmdLine, iCmdShow); #else retcode = (*((MainFunc)appMain))(argc, argv, envp); #endif /* Cleanup and done */ FreeLibrary(hApp); FreeLibrary(hMFC); return retcode; } -- Dimi.

4 8

Re: [comctl32] Fix listview item activate
by Dimitrie O. Paun Sept. 14, 2003

Sept. 14, 2003

On September 14, 2003 09:36 am, Maxime Bellengé wrote: > + nmia.uNewState = 0; > + nmia.uOldState = 0; > + nmia.uChanged = 0; > + nmia.lParam = 0; > + nmia.uKeyFlags = 0; What about these flags, shouldn't we fill these in as well? At the very least we need an entry in the TODO list at the top of the file... -- Dimi.

1 0

Link winlib with windows DLL help????
by Kevin Atkinson Sept. 14, 2003

Sept. 14, 2003

I am having a hell of a time getting a simple program to link with a windows dll. I am really sorry for the newbie question but I can't seam to find an answer anywhere and the mailing list archive is not searchable. Here is what I got: avs2yuv4mpeg.cpp - the main program avisynth_c.h - header for DLL I need to link with avisynth_c.lib - lib " " " And I want to compile this into an executable avs2yuv4mpeg I try "winemaker -iavisynth_c ." But when making I get "could not open .def file for avisynth_c" And I am not sure how to create this .def file. Please help. Once again really sorry. But the mailing list is not searchable AND the winlib docs are really confusing and do not specially cover this situation. -- http://kevin.atkinson.dhs.org

3 8

Re: wineproc: a new wine utility
by Dimitrie O. Paun Sept. 14, 2003

Sept. 14, 2003

On September 3, 2003 03:53 pm, Eric Pouech wrote: > any comment welcomed!! Sorry for the late reply -- I just got a chance to look into it. It looks good, I hope to see (some version of) it in the tree soon. One thing that bug me were the refreshes -- too much flicker, loss of selection etc. I think we need to be smarter and not change the display if nothing changed. But of course, this can happen once we have something integrated into the tree... -- Dimi.

1 0

Re: FAQ: Update the ports section
by Ivan Leo Murray-Smith Sept. 14, 2003

Sept. 14, 2003

>On Thu, 11 Sep 2003, Ivan Leo Murray-Smith wrote: > >> The original sentence that started this is >> >The following list are the x86 operating systems supported by >[...] >> "The list is current as of XX/XX/XXXX" is a solution that should make everybody >> happy. > >Not really because as the above clearly illustrates, this is not what >the original sentence said or meant to say. I really don't know why >everyone focused on this 'up-to-date' thing. > >*Sigh* This is just getting so boring, just use "The following lists the x86 operating systems supported by" and that can be the end of that (Hoping nobody starts discussing again about there great knowledge of slang and ignorance of english...)

2 1

Kernel pipe test segfaults?
by Rolf Kalbermatter Sept. 14, 2003

Sept. 14, 2003

Mike Hearn <mike(a)theoretic.com> wrote: >Maybe this is a known issue, I haven't really followed the test case >work up until now, but does the kernel32 pipe test segfault for anybody >else? I have problems, too. It used to crash somewhere in kernel32 with processes and now with pipes, where the test hangs for several minutes and then is aborted with returning an error 139. This is with Suse 8.2 and an Athlon optimized 2.4.20 Standard Suse kernel. Rolf Kalbermatter CIT Engineering Nederland BV tel: +31 (070) 415 9190 Treubstraat 7H fax: +31 (070) 415 9191 2288 EG Rijswijk http://www.citengineering.com Netherlands mailto:rolf.kalbermatter(a)citeng.com

1 0

Really slow named pipes
by Duane Clark Sept. 13, 2003

Sept. 13, 2003

This is a bit old, from May, but anyhow... The patch to implement anonymous pipes on top of named pipes: http://www.winehq.com/hypermail/wine-cvs/2003/05/0305.html along with a corresponding bug fix: http://www.winehq.com/hypermail/wine-cvs/2003/06/0123.html causes Xilinx ISE to run really, really slow. It has a window to which messages are sent via a pipe, and what I see now is that one line comes out every 2/3 of a second or so, as regular as a clock at least by eye. I can revert those two patches, and it goes back to spitting out the messages very fast. Since there are a lot of messages being sent, the current version slows the app down to a crawl. It does not look like the patch itself is the problem; I am guessing that it is the underlying named pipe implementation that is the problem, which was exposed by the patch. Also, reverting to the full CVS 5/20, just after the pipe patches, has the same problem. So it appears that whatever is wrong with named pipes (if that really is the problem) was already there on 5/20, rather than it being something introduced since then. Any hints on what I could look for would be welcome.

4 10

Re: kernel32.GetProcessIoCounters basic implementation
by Eric Pouech Sept. 13, 2003

Sept. 13, 2003

Eric Pouech wrote: > TSIA > A+ oops, wrong file... use this one instead :-( A+ -- Eric Pouech Name: iocnt ChangeLog: Implemented GetIoProcessCounters License: X11 GenDate: 2003/09/13 12:50:39 UTC ModifiedFiles: dlls/kernel/kernel32.spec dlls/kernel/process.c =================================================================== RCS file: /home/cvs/cvsroot/wine/wine/dlls/kernel/kernel32.spec,v retrieving revision 1.106 diff -u -u -r1.106 kernel32.spec --- dlls/kernel/kernel32.spec 11 Sep 2003 03:00:11 -0000 1.106 +++ dlls/kernel/kernel32.spec 13 Sep 2003 07:18:18 -0000 @@ -452,7 +452,7 @@ @ stdcall GetProcessFlags(long) @ stdcall GetProcessHeap() @ stdcall GetProcessHeaps(long ptr) -@ stub GetProcessIoCounters +@ stdcall GetProcessIoCounters(long ptr) @ stdcall GetProcessShutdownParameters(ptr ptr) @ stdcall GetProcessTimes(long ptr ptr ptr ptr) @ stdcall GetProcessVersion(long) Index: dlls/kernel/process.c =================================================================== RCS file: /home/cvs/cvsroot/wine/wine/dlls/kernel/process.c,v retrieving revision 1.4 diff -u -u -r1.4 process.c --- dlls/kernel/process.c 5 Sep 2003 23:08:37 -0000 1.4 +++ dlls/kernel/process.c 13 Sep 2003 07:17:56 -0000 @@ -424,6 +424,16 @@ return !status; } +/****************************************************************** + * GetProcessIoCounters (KERNEL32.@) + * + * + */ +BOOL WINAPI GetProcessIoCounters(HANDLE hProcess, PIO_COUNTERS ioc) +{ + memset(ioc, 0, sizeof(ioc)); + return TRUE; +} /*********************************************************************** * RegisterServiceProcess (KERNEL.491)

1 0

Here is one - Linux on Cygwin on Linux
by Steven Edwards Sept. 13, 2003

Sept. 13, 2003

Here is a good one for someone to test. The User Mode Linux people have done it. They are working on a port of User Mode Linux to Win32. This has been something I have been wanting to see for a while so we can use Linux as our Posix subystem in ReactOS by moving parts of UMLwin32 in to back to kernel mode. Right now you still need cygwin but if someone wants to try running it under Cygwin on WINE here is the url: http://umlwin32.sourceforge.net Thanks Steven __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com

3 2

Re: Word-splitting in the FAQ
by Dan Kegel Sept. 13, 2003

Sept. 13, 2003

On the general subject of spellchecking, there are a few typo lists and spellcheckers at http://www.kegel.com/kerspell/ which specialize in typos in C program comments... might be of some interest. One of them was contributed by Francois and has already been run on the Wine tree, I think... -- Dan Kegel http://www.kegel.com http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045

2 1

Re: FAQ: Update the ports section
by Ivan Leo Murray-Smith Sept. 13, 2003

Sept. 13, 2003

The original sentence that started this is >The following list are the x86 operating systems supported by from twickline_at_skybest.com http://www.winehq.org/hypermail/wine-devel/2003/09/0356.html He claims he knows 3 types of slang, and is from west virginia, but he never claimed to be "A native english speaker", so you don't have (Such a big) reason to be bothered. I was the first to claim a native english speaker (As far as I know) and I have tried explaining to Tom that there is a thing called grammar, but he had a few problems understanding that. "The list is current as of XX/XX/XXXX" is a solution that should make everybody happy.

2 1

ikernel.exe error still not fix in 11_9_03.tar.gz
by lineric＠qwest.net Sept. 12, 2003

Sept. 12, 2003

1 0

Re: [ros-kernel] an interesting idea
by Steven Edwards Sept. 12, 2003

Sept. 12, 2003

This is what I have been thinking. We can move parts of UML back to a subsytem mode and them implement a Xsever in to Win32k.sys/GDI32. You wouldnt even need to recompile X applications provided we implement X properly. All that would be needed is a modded Xserver and X libs. The client libs should still work using X protocal and Xtenstions. This type of hybrid system could then run any application Windows or Linux and would not suffer from the same problems WINE does because it would be a real Linux Kerenl and GLIBC running. Thanks Steven --- Jonathan Wilson <jonwil(a)tpgi.com.au> wrote: > Something we could do (once we get UML going without cygwin and down into > the kernel) is to build a Linux subsystem (mabie even with ports of > GTK/QT/X/etc on top of GDI) that could run linux apps just as easily as it > can run windows apps. > > __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com

2 1

Re: [winecfg 8] Complete drive edit dialog, warn on lack of C drive, implement drive delete, code cleanups
by Dimitrie O. Paun Sept. 12, 2003

Sept. 12, 2003

On 11 Sep 2003, Mike Hearn wrote: > ChangeLog: > - All settings in the drive edit dialog are now instant apply This is a cool feature, but we should keep it optional. Our primary audience are Windows users, and they are not used to that. Heck, even I am not used to it, I am used to be able to play around with stuff in dialogs, and than press CANCEL. In fact, KDE is not instant apply, so I would argue that even most Linux users are used to that. I'd argue that we need to support instant apply to integrate with GNOME, but to default to the classic OK/CANCEL for now, until we can figure how and when to enable one or the other behaviour. -- Dimi.

5 8

Kernel pipe test segfaults?
by Mike Hearn Sept. 12, 2003

Sept. 12, 2003

Maybe this is a known issue, I haven't really followed the test case work up until now, but does the kernel32 pipe test segfault for anybody else? thanks -mike

2 1

Re: WIne help - get rid of W->A calls
by Eric Pouech Sept. 12, 2003

Sept. 12, 2003

hatky wrote: > My first patch, so a deep check is welcome > > Change log: > - Remove W->A call from WineHelpA/W > - This patch is dedicated to dimi that guided me I'm afraid the patch is wrong. the current implentation of WinHelp, which sends data from any app to the program winhelp, requires that the data block passed between those two apps encodes strings in ASCII, not unicode. Your patch modifies the sender so that it encodes in Unicode, but the receiver is left untouched. I'd suggest anyway that we remain as close as possible as Windows do. - we don't completly handle this protocol as Windows do (Windows send a WM_HELP - registered - message, we only use a WM_COPYDATA), but the format of the message (parameter block) is quite the same (we just add some extra fields to ensure that we don't mess with a spurious WM_COPYDATA message in winhelp). - as Windows encode strings in this message block in ASCII, I'd suggest keeping the encoding of strings inside the message as ASCII. A+ -- Eric Pouech

3 2

configure new wine
by eric Sept. 12, 2003

Sept. 12, 2003

Dear cvs wine users: By reading the doc's install page, I success uninstall cvs 's wine reinstall 8/13' wine, then test again, it indeed repeat the ikernel problem, but it at least configure correct by default. So I copy /root/.wine/config to /root/oldwineconfig. Then uninstall it again as doc indicate, then install cvs again, after that I cp oldwineconfig /etc/wine/config and /root/.wine/config then run my app, it still show segmentation fault please help, thanks in advance Sincerely Eric

1 0

Can not uninstall CVS wine
by eric Sept. 12, 2003

Sept. 12, 2003

Dear Wine users: After I install CVS wine, and find it is not work, it type rpm -e wine reinstall my old wine, then if I did not do it by force, it showed error ------------------ [root(a)localhost x86]# rpm -vhi /root/wine-2* Preparing... ########################################### [100%] file /usr/bin/widl from install of wine-20030813-1rh9winehq conflicts with file from package wine-cvs-unstripped-20030911-1 file /usr/bin/wine from install of wine-20030813-1rh9winehq conflicts with file from package wine-cvs-unstripped-20030911-1 ------------------ if I use force to install old package, it still can not run my app, segmentation fault, and I check rpm -q wine, it reponse wine package in not installed highly appreciate any help to remove that CVS's wine throughly and thanks in advance Sincerely Eric

1 0

openbsd supports MAP_FIXED too
by Ted Unangst Sept. 12, 2003

Sept. 12, 2003

sorry no patch. i'm not sure what the status of wine on openbsd is, although i came across a notice that wine was capable of using MAP_FIXED. i committed what amounted to todd's (netbsd) patch to openbsd shortly after it went into netbsd. don't know if this helps / who cares, but fyi. patch below could probably be || defined (__OpenBSD__) as well. http://www.winehq.org/hypermail/wine-patches/2003/08/0278.html -- someone's writing down your mistakes someone's documenting your downfall

1 0

scheduler related complete breakage on FreeBSD
by Gerald Pfeifer Sept. 12, 2003

Sept. 12, 2003

On FreeBSD 4.8 (using -lc_r instead of -lc to get pthreads) Wine crashes upon startup: /sw/test/wine% wine FIXME:pthread_cond_init Segmentation fault (core dumped) % gdb53 miscemu/wine wine.core GNU gdb 5.3 (FreeBSD) Reading symbols from /sw/test/wine/dlls/libntdll.dll.so...done. Loaded symbols for /sw/wine/dlls/libntdll.dll.so Reading symbols from /sw/test/wine/libs/libwine.so.1...done. Loaded symbols for /sw/test/wine/libs/libwine.so.1 Reading symbols from /sw/test/wine/libs/libwine_unicode.so.1...done. Loaded symbols for /sw/test/wine/libs/libwine_unicode.so.1 Reading symbols from /usr/lib/libc.so.4...done. Loaded symbols for /usr/lib/libc.so.4 Reading symbols from /usr/lib/libm.so.2...done. Loaded symbols for /usr/lib/libm.so.2 Reading symbols from /usr/lib/libc_r.so.4...done. Loaded symbols for /usr/lib/libc_r.so.4 Reading symbols from /sw/test/wine/dlls/kernel32.dll.so...done. Loaded symbols for /sw/test/wine/dlls/kernel32.dll.so Reading symbols from /usr/libexec/ld-elf.so.1...done. Loaded symbols for /usr/libexec/ld-elf.so.1 (gdb) bt #0 0x5c251aac in .cerror () from /usr/lib/libc.so.4 #1 0x5c2feca6 in flock () from /usr/lib/libc_r.so.4 #2 0x5c231919 in _yp_dobind () from /usr/lib/libc.so.4 #3 0x5c231fa8 in yp_match () from /usr/lib/libc.so.4 #4 0x5c215c12 in endpwent () from /usr/lib/libc.so.4 #5 0x5c2153b8 in getpwuid () from /usr/lib/libc.so.4 #6 0x5c0d06d1 in init_paths () at config.c:112 #7 0x5c0d0a4e in wine_get_server_dir () at config.c:191 #8 0x5c07e5ae in CLIENT_InitServer () at ../../scheduler/client.c:641 #9 0x5c07f4f4 in process_init (argv=0xbfbff690) at ../../scheduler/process.c:306 #10 0x5c07fa19 in __wine_process_init (argc=1, argv=0xbfbff690) at ../../scheduler/process.c:480 #11 0x5c0d21c4 in wine_init (argc=1, argv=0xbfbff690, error=0xbfbff240 "", error_size=1024) at loader.c:428 #12 0x3c000621 in main (argc=1, argv=0xbfbff690) at main.c:33 #13 0x3c000566 in _start () Any help will be appreciated. Gerald

1 1

FIXME:pthread_cond_init
by Gerald Pfeifer Sept. 12, 2003

Sept. 12, 2003

When building Wine on FreeBSD with libc_r (r for reentrant) instead of libc, I now trigger the following FIXME in scheduler/pthread.c: /***** CONDITIONS *****/ /* not implemented right now */ int __pthread_cond_init(pthread_cond_t *cond, const pthread_condattr_t *cond_attr) { P_OUTPUT("FIXME:pthread_cond_init\n"); return 0; } What's the proper way to address this? (I'm afraid, it's beyond what I'll be able to implement, though.) Gerald

1 1

BuildCommDCB conformance test
by Kevin Groeneveld Sept. 12, 2003

Sept. 12, 2003

Here is my first attempt at a conformance test for the BuildCommDCB functions. It does not cover very many cases yet, but I thought I would post what I have so far in case I am going about this the wrong way. I compiled this with MSVC and it ran with no errors on both Windows 95 and Windows 2000. It also compiles with libwine in Linux, but for this to work you need to add a missing member to the DCB structure in winbase.h. If I run this test with the existing implementation of BuildCommDCB it page faults, probably because it is writing to a constant string. If my previous patch to BuildCommDCB is applied, then the test runs ok. Kevin diff -urN wine-20030813/dlls/kernel/tests/Makefile.in wine-20030813.patched/dlls/kernel/tests/Makefile.in --- wine-20030813/dlls/kernel/tests/Makefile.in 2003-08-23 23:12:31.000000000 -0400 +++ wine-20030813.patched/dlls/kernel/tests/Makefile.in 2003-09-10 23:31:59.000000000 -0400 @@ -8,6 +8,7 @@ CTESTS = \ alloc.c \ atom.c \ + build_dcb.c \ codepage.c \ console.c \ directory.c \ diff -urN wine-20030813/dlls/kernel/tests/build_dcb.c wine-20030813.patched/dlls/kernel/tests/build_dcb.c --- wine-20030813/dlls/kernel/tests/build_dcb.c 1969-12-31 19:00:00.000000000 -0500 +++ wine-20030813.patched/dlls/kernel/tests/build_dcb.c 2003-09-11 01:00:19.000000000 -0400 @@ -0,0 +1,171 @@ +/* Unit test suite for BuildCommDCBA/W & BuildCommDCBAndTimeoutsA/W + * + * Copyright 2003 Kevin Groeneveld + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "wine/test.h" +#include "winbase.h" + +typedef struct +{ + char *string; + BOOL result; + DCB dcb1, dcb2; + COMMTIMEOUTS timeouts1, timeouts2; +} TEST; + +TEST test[] = +{ + { + "baud=9600", + TRUE, + { 0x00000000, 9600, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 9600, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0x00, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff } + }, + { + " baud=9600", + FALSE, + { 0x00000000, 0x00000000, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 0xffffffff, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff } + }, + { + "to=on", + TRUE, + { 0x00000000, 0x00000000, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 0xffffffff, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0x00, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x0000EA60 }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x0000EA60 } + }, +}; + +#define TEST_COUNT (sizeof(test) / sizeof(TEST)) + +void check_result(char *function, char *string, int initial_value, BOOL result1, BOOL result2) +{ + DWORD LastError = GetLastError(); + DWORD CorrectError = (result2 ? 0xdeadbeef : ERROR_INVALID_PARAMETER); + + ok(LastError == CorrectError, "%s(\"%s\"), 0x%02x: GetLastError() returned 0x%08lx, should be 0x%08lx", function, string, initial_value, LastError, CorrectError); + ok(result1 == result2, "%s(\"%s\"), 0x%02x: return value should be %s", function, string, initial_value, result2 ? "TRUE" : "FALSE"); +} + +#define check_dcb_member(a,b) ok(pdcb1->a == pdcb2->a, "%s(\"%s\"), 0x%02x: "#a" is "b", should be "b, function, string, initial_value, pdcb1->a, pdcb2->a); + +void check_dcb(char *function, char *string, int initial_value, DCB *pdcb1, DCB *pdcb2) +{ + /* DCBlength is a special case since Win 9x sets it but NT does not. + We will accept either as correct. */ + ok(pdcb1->DCBlength == pdcb2->DCBlength || pdcb1->DCBlength == sizeof(DCB), "%s(\"%s\"), 0x%02x: DCBlength is %lu, should be %lu or %lu", function, string, initial_value, pdcb1->DCBlength, pdcb2->DCBlength, (DWORD)sizeof(DCB)); + + /* Check all other DCB members */ + check_dcb_member(BaudRate, "%lu"); + check_dcb_member(fBinary, "%u"); + check_dcb_member(fParity, "%u"); + check_dcb_member(fOutxCtsFlow, "%u"); + check_dcb_member(fOutxDsrFlow, "%u"); + check_dcb_member(fParity, "%u"); + check_dcb_member(fOutxCtsFlow, "%u"); + check_dcb_member(fOutxDsrFlow, "%u"); + check_dcb_member(fDtrControl, "%u"); + check_dcb_member(fDsrSensitivity, "%u"); + check_dcb_member(fTXContinueOnXoff, "%u"); + check_dcb_member(fOutX, "%u"); + check_dcb_member(fInX, "%u"); + check_dcb_member(fErrorChar, "%u"); + check_dcb_member(fNull, "%u"); + check_dcb_member(fRtsControl, "%u"); + check_dcb_member(fAbortOnError, "%u"); + check_dcb_member(fDummy2, "%u"); + check_dcb_member(wReserved, "%u"); + check_dcb_member(XonLim, "%u"); + check_dcb_member(XoffLim, "%u"); + check_dcb_member(ByteSize, "%u"); + check_dcb_member(Parity, "%u"); + check_dcb_member(StopBits, "%u"); + check_dcb_member(XonChar, "%d"); + check_dcb_member(XoffChar, "%d"); + check_dcb_member(ErrorChar, "%d"); + check_dcb_member(EofChar, "%d"); + check_dcb_member(EvtChar, "%d"); + check_dcb_member(wReserved1, "%u"); +} + +#define check_timeouts_member(a) ok(ptimeouts1->a == ptimeouts2->a, "%s(\"%s\"), 0x%02x: "#a" is %lu, should be %lu", function, string, initial_value, ptimeouts1->a, ptimeouts2->a); + +void check_timeouts(char *function, char *string, int initial_value, COMMTIMEOUTS *ptimeouts1, COMMTIMEOUTS *ptimeouts2) +{ + check_timeouts_member(ReadIntervalTimeout); + check_timeouts_member(ReadTotalTimeoutMultiplier); + check_timeouts_member(ReadTotalTimeoutConstant); + check_timeouts_member(WriteTotalTimeoutMultiplier); + check_timeouts_member(WriteTotalTimeoutConstant); +} + +void test_BuildCommDCBA(char* string, BOOL expected_result, DCB *pexpected_dcb, int initial_value) +{ + BOOL result; + DCB dcb; + + /* set initial conditions */ + memset(&dcb, initial_value, sizeof(DCB)); + SetLastError(0xdeadbeef); + + result = BuildCommDCBA(string, &dcb); + + /* check results */ + check_result("BuildCommDCBA", string, initial_value, result, expected_result); + check_dcb("BuildCommDCBA", string, initial_value, &dcb, pexpected_dcb); +} + +void test_BuildCommDCBAndTimeoutsA(char* string, BOOL expected_result, DCB *pexpected_dcb, COMMTIMEOUTS *pexpected_timeouts, int initial_value) +{ + BOOL result; + DCB dcb; + COMMTIMEOUTS timeouts; + + /* set initial conditions */ + memset(&dcb, initial_value, sizeof(DCB)); + memset(&timeouts, initial_value, sizeof(COMMTIMEOUTS)); + SetLastError(0xdeadbeef); + + result = BuildCommDCBAndTimeoutsA(string, &dcb, &timeouts); + + /* check results */ + check_result("BuildCommDCBAndTimeoutsA", string, initial_value, result, expected_result); + check_dcb("BuildCommDCBAndTimeoutsA", string, initial_value, &dcb, pexpected_dcb); + check_timeouts("BuildCommDCBAndTimeoutsA", string, initial_value, &timeouts, pexpected_timeouts); +} + +void do_test(TEST *ptest) +{ + test_BuildCommDCBA(ptest->string, ptest->result, &ptest->dcb1, 0x00); + test_BuildCommDCBA(ptest->string, ptest->result, &ptest->dcb2, 0xFF); + test_BuildCommDCBAndTimeoutsA(ptest->string, ptest->result, &ptest->dcb1, &ptest->timeouts1, 0x00); + test_BuildCommDCBAndTimeoutsA(ptest->string, ptest->result, &ptest->dcb2, &ptest->timeouts2, 0xFF); +} + +START_TEST(build_dcb) +{ + int i; + + for(i = 0; i < TEST_COUNT; i++) + do_test(&test[i]); +} diff -urN wine-20030813/include/winbase.h wine-20030813.patched/include/winbase.h --- wine-20030813/include/winbase.h 2003-08-23 23:12:51.000000000 -0400 +++ wine-20030813.patched/include/winbase.h 2003-09-08 20:19:39.000000000 -0400 @@ -976,6 +976,7 @@ char ErrorChar; char EofChar; char EvtChar; + WORD wReserved1; } DCB, *LPDCB; typedef struct tagCOMMCONFIG {

4 4

Re: missing ordinals in avifil32
by Alexandre Julliard Sept. 12, 2003

Sept. 12, 2003

Dave Miller <compsol(a)ptd.net> writes: > The first 59 ordinals seem to match in all windows versions. Add > ordinals to avifil32.spec to match windows. Unless ordinals are actually needed, it's better not to add them. In that case the functions are sorted so it's very likely that Windows is using automatic allocation too. -- Alexandre Julliard julliard(a)winehq.com

3 3