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

WineHQ: email trouble

by Dimitrie O. Paun

Folks, Whazzup with winehq.org? I get duplicated emails, and more importantly the latency on the wine-*(a)winehq.org mailing lists is measured in (many) hours instead of seconds. What gives? -- Dimi.

April 30, 2003

Callbacks while app is looping

by Mark Westcott

Hi all, I've been working on fixing the crash in 'Gens' when trying to reassign keys. The app basically does a: while (1) { IDIRECTINPUTINTERFACE::GetDeviceState(); } Until it notices a key has been pressed. DirectInput hooks onto system keyboard events with SetWindowsHookEx(WH_KEYBOARD_LL........), but while the app is doing its nasty while(1) loop, DirectInput's callback procedure never gets called. This is as far as I can get (new to wine) - any ideas anyone? I threw a test case app that illustrates the problem (btw, the WH_KEYBOARD_LL hook has the note 'should use SendMessage instead', but I don't get what this means or know if using SendMessage instead would have anything to do with fixing this problem) -------- Mark Westcott

April 30, 2003

Native ms progs and mybe wine 1.1

by hatky

I opened a metabug that I hope some of you will like where I try to "Get native windows programs working" bug 1424 Think of it: dxdiag is a great test case for the directX part (direct draw,direct 3d, sound, music, direct play...) ftp,telnet,ping,tracert can show who good the network code is defrag will mybe show who great low level hardware accses is ? what will it do with a non fat paritation? do I hear a challange? I don't know what the rest of you think, but a big bug tree looks cool to me becouse you know you can fix something little and to how many other things it helps so this tree may get a bit big.... I don't hear no objections :) I hope some of you will help with it a bit and I think it can be a nice start for the wine 1.1 guildlines hoping to get excited responces Hatky. __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com

April 30, 2003

Re: WINE flag

by Shachar Shemesh

KK singh wrote: >hi, > >I have a win32 application that plays some hashing >tricks with HWND while running on win98 and uses the >fact that on win98 HWND is always 16bit. > >This application is crashing on wine configured as >win98 because HWND is something bigger than 16bit on >wine. > >Now Can I Check from inside my application code, >whether the application is running on linux or win98 >If yes then I might be able to avoid those hashing >tricks and solve the crash > > I think the last time the subject came up, the consensus was that the existance of HKLM/Programs/Wine is the best way to check whether you are running on Wine. A different matter (and one which I am not qualified to attest to) is whether it is smart for wine to be 100% win98 compatible in that regard, and whether it is smart on your program's side to do those tricks. >Regards >kks > > > Hope this helps. Shachar

April 30, 2003

Winevdm eats first parameter

by Sylvain Petreolle

Trying to run a dos program with parameters, the first parameter is eaten by winevdm. at wcmd prompt, i run : dos32a d2demo.exe and dos32a answers : syntax is DOS32A <execname.xxx> which never happened before. A small asm tutorial program about dos psp shows this problem. (attached) run "wine winevdm.exe psp.com 1 2 3 4" , it will say : Les param?tres suivants ont ?t? transmis : 2 3 4 ===== Sylvain Petreolle (spetreolle at users dot sourceforge dot net) ICQ #170597259 No more War ! "What if tomorrow the War could be over ?" Morpheus, in "Reloaded". For the Law of Oil and Fire, Im an European that lives in France. For all my Brothers and friends, Im a human living on Earth. ___________________________________________________________ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com

April 30, 2003

Re: Wine Regression

by Rein Klazes

On Wed, 30 Apr 2003 00:53:40 -0400, you wrote: > With the current cvs tree parts of the Agent news reader are broken. > > I don't know the exact date it last worked but it wasn't that long ago. > > When double clicking on a group that is empty, the dialog you would normally > see that asks if youw ant to download x headers, or all available headers, > etc displays, but only the border and title bar are drawn and then wine > crashes. Agent crashes here as well, but long before that, at startup before the splash screen is drawn. My problem is fixed by Gerald Pfeifers patch at http://www.integrita.com/cgi-local/lwgate.pl/WINE-PATCHES/archives/2003-04/… Rein. -- Rein Klazes rklazes(a)xs4all.nl

April 30, 2003

Cygwin building part 2

by Steven Edwards

We have 5 bugs that are preventing building on cygwin. Not that it will work after it builds but lets get it to build first and then worry about running. Some of these seem like they are simple while others like the kernel32 building are very complex. I will see if I can fix the winsock bug tonight. 1. We cannot build kernel32 with the current CVS http://www.winehq.org/hypermail/wine-devel/2003/04/0765.html 2. when building wininet on cygwin we get this: internet.o(.text+0x1365): In function `InternetCloseHandle': /cygdrive/g/src/wine-dev/wine/dlls/wininet/internet.c:670: undefined reference t o `___wine_exception_handler' collect2: ld returned 1 exit status dllwrap: gcc exited with status 1 3. With winsock we are getting this: gcc -c -I. -I. -I../../include -I../../include -D_REENTRANT -D__WINESRC__ -DUSE _WS_PREFIX -Wall -mpreferred-stack-boundary=2 -gstabs+ -Wpointer-arith -g -O2 -o socket.o socket.c socket.c:286: `IP_HDRINCL' undeclared here (not in a function) socket.c:286: initializer element is not constant socket.c:286: (near initialization for `ws_ip_map[6][1]') socket.c:286: initializer element is not constant socket.c:286: (near initialization for `ws_ip_map[6]') socket.c:287: initializer element is not constant socket.c:287: (near initialization for `ws_ip_map[7]') socket.c:288: initializer element is not constant socket.c:288: (near initialization for `ws_ip_map[8]') socket.c:289: initializer element is not constant socket.c:289: (near initialization for `ws_ip_map[9]') make[1]: *** [socket.o] Error 1 4. miscemu fails to build main.c:34: warning: implicit declaration of function `exit' gcc -o wine main.o -L../libs/wine -lwine -L../libs/port -lwine_port main.o(.text+0x51): In function `main': /cygdrive/g/src/wine-dev/wine/miscemu/main.c:32: undefined reference to `_wine_i nit' collect2: ld returned 1 exit status 5. There is no get/set_thread_context support. This patch by David Fraser stubs it. Index: context_i386.c =================================================================== RCS file: /home/wine/wine/server/context_i386.c,v retrieving revision 1.28 diff -u -r1.28 context_i386.c --- context_i386.c 22 Mar 2003 21:00:09 -0000 1.28 +++ context_i386.c 30 Apr 2003 16:39:35 -0000 @@ -78,7 +78,26 @@ #define PTRACE_SETDBREGS PT_SETDBREGS #endif -#ifdef linux +#if defined(__CYGWIN__) + +/* retrieve a thread context */ +static void get_thread_context( struct thread *thread, unsigned int +flags, CONTEXT *context ) +{ + /* FIXME: implement this */ + file_set_error(); +} + + +/* set a thread context */ +static void set_thread_context( struct thread *thread, unsigned int +flags, const CONTEXT *context ) +{ + /* FIXME: implement this */ + file_set_error(); +} + +#elif defined(linux) #ifdef HAVE_SYS_USER_H # include <sys/user.h> #endif

April 30, 2003

Re: iphlpapi.dll implementation

by Dimitrie O. Paun

On Wed, 30 Apr 2003, Juan Lang wrote: > Email me if interested. No, it's not vaporware, it's > just not as well tested as I'd like. Just send it over to wine-devel for comments, or wine-patches for inclusion into the tree. -- Dimi.

April 30, 2003

dxdiag gui regression

by Robert Reif

I am having problems with the gui in dxdiag.exe with the latest cvs. It appears to be coming up in the spanish language and the tab labels are all set to the first one. It worked ok with a cvs from a few days ago. Here are the errors: wine dxdiag.exe err:wave:OSS_WaveOutInit /dev/mixer1: not found! err:wave:OSS_WaveInInit /dev/mixer1: not found! fixme:ole:CoCreateInstance no classfactory created for CLSID {4590f811-1d3a-11d0-891f-00aa004b2e24}, hres is 0x80040154 err:module:BUILTIN32_LoadLibraryExA loaded .so but dll user.exe still not found - 16-bit dll or version conflict. err:module:BUILTIN32_LoadLibraryExA loaded .so but dll user.exe still not found - 16-bit dll or version conflict. err:ver:VerQueryValueW called on NE resource! fixme:global:GlobalMemoryStatusEx we should emulate the 4GB bug here, as per MSDN fixme:dsound:IKsPropertySetImpl_Get unsupported property: {f2957840-260c-11d1-a4d8-00c04fc28aca} fixme:system:EnumDisplayDevicesW ((nil),0,0x40711968,0x00000000), stub! fixme:system:EnumDisplayDevicesW (0x4071196c,0,0x4071120c,0x00000000), stub! fixme:dsound:IKsPropertySetImpl_Get unsupported property: {1aeaa606-35f0-11d1-b161-00c04fc28aca} This is a unicode program and it has a lot of problems related to that. Did it just happen to work before and some of the recent changes changed that behavior? The strange thing is that it worked properly when I tried to trace it to see what was happening. Any ideas before I start digging into this deeper? Thanks, Bob.

April 30, 2003

Re: Avoid non-portable signal(SIGCHLD, SIG_IGN)

by Joerg Wunsch

As Gerald Pfeifer wrote: > assertion "libc_sigaction" failed: file "../../scheduler/pthread.c", > line 661 > #3 0x5c16c404 in sigaction (signum=12, act=0xbfbfdce0, oldact=0xbfbfdcc8) at ../../scheduler/pthread.c:661 Signal #12 is SIGSYS (Illegal system call). That should not be related to the SIGCHLD manipulations. > #4 0x5c074f09 in getcwd () from /usr/lib/libc.so.4 Is getcwd() supposed to cause a SIGSYS at all? OTOH, that failing dlsym() syscall could be related to all this... I don't know why that complicated dlsym() is used, but maybe since sigaction() is now invoked directly before, the run-time loader has already loaded it from libc.so before. Maybe Alexandre can shed some light on how that's supposed to work. -- cheers, J"org .-.-. --... ...-- -.. . DL8DTL http://www.sax.de/~joerg/ NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)

April 30, 2003