- Wine-Devel - list.winehq.org

RE: Strange CVS problem (help!!!)
by Patrik Stridvall Sept. 6, 2002

Sept. 6, 2002

> On September 6, 2002 05:16 pm, Lionel Ulmer wrote: > > > As of lately, this no longer happens: I get a big fat > conflic for all > > > the changes I submit! This is *very* annoying. Anyone knows > > > how to fix this behaviour? Alexandre, are you doing anything > > > different? > > > > Maybe automatic white-space stripping at commit time ? > > I was thinking of that, but I get a conflict on _every_ hunk... > How can I avoid it, this is really, really bad. In particular, > if I submit a large patch, and then I continue working on > a different part of the file, I am in trouble... Try running winapi_cleanup before generating the patch. Running it before doing "cvs update" might also help. winapi_cleanup strips trailing white-space from .c and .h files. Note that you have to install it first. cd tools/winapi make install Then you can run it in whatever directory you like and it converts the .c and .h files. You can also run it on specific files like: winapi_cleanup foo.c bar.c baz.h PS. It is "intelligent" and only actually modifies files that needs to be modified so the timestamp on the files that needn't be modified remains.

1 0

Re: PATCH: almost ready with the conversion of HWND to a void*
by Alexandre Julliard Sept. 6, 2002

Sept. 6, 2002

Michael Stefaniuc <mstefani(a)redhat.de> writes: > the attached patch is independent of the previous sent HWND4.diff and > should almost finish the conversion of HWND to a void*. There are still > some "warning: assignment makes pointer from integer without a cast" but > they occur between SERVER_START_REQ and SERVER_END_REQ and i didn't > figured out to what to cast the HWND's (that code is a little bit too > obfuscated for me). Actually you don't need to cast them at all. When handles are made void* server handles will be void* too so it will all work fine. -- Alexandre Julliard julliard(a)winehq.com

2 1

Patch: winemaker-options
by Martin Wilck Sept. 6, 2002

Sept. 6, 2002

This patch fixes winemaker's annoying habit to insist on --mfc even if --nomfc was specified, and adds an option to not add any DLLs automatically. Patch: winemaker-options.diff Martin Wilck <Martin.Wilck(a)fujitsu-siemens.com> Modified files: tools: winemaker - fix --nomfc option which is currrently broken - add --nodlls option for small apps This patch is incremental over Francois Gouget's latest winemaker patch: http://www.winehq.com/hypermail/wine-patches/2002/09/0028.html and my previous patch (winemaker-bugs.diff) --- CVS/wine/tools/winemaker Fri Sep 6 17:35:39 2002 +++ TMP/wine/tools/winemaker Fri Sep 6 17:36:46 2002 @@ -234,6 +234,14 @@ my $TF_MFC=4; ## +# User has specified --nomfc option for this target or globally +my $TF_NOMFC=8; + +## +# --nodlls option: Do not use standard DLL set +my $TF_NODLLS=16; + +## # Initialize a target: # - set the target type to TT_SETTINGS, i.e. no real target will # be generated. @@ -436,23 +444,37 @@ push @{@$target[$T_LIBRARY_PATH]},$option; } elsif ($option =~ /^-l/) { push @{@$target[$T_LIBRARIES]},"$'"; - } elsif (@$target[$T_TYPE] != $TT_DLL and $option =~ /^--wrap/) { - @$target[$T_FLAGS]|=$TF_WRAP; - } elsif (@$target[$T_TYPE] != $TT_DLL and $option =~ /^--nowrap/) { - @$target[$T_FLAGS]&=~$TF_WRAP; - } elsif ($option =~ /^--mfc/) { - @$target[$T_FLAGS]|=$TF_MFC; + } elsif ($option =~ /^--wrap/) { if (@$target[$T_TYPE] != $TT_DLL) { @$target[$T_FLAGS]|=$TF_WRAP; + } else { + print STDERR "warning: option --wrap is illegal for DLLs - ignoring"; + }; + } elsif ($option =~ /^--nowrap/) { + if (@$target[$T_TYPE] != $TT_DLL) { + @$target[$T_FLAGS]&=~$TF_WRAP; + } else { + print STDERR "warning: option --nowrap is illegal for DLLs - ignoring"; } + } elsif ($option =~ /^--mfc/) { + @$target[$T_FLAGS]|=$TF_MFC; + @$target[$T_FLAGS]&=~$TF_NOMFC; } elsif ($option =~ /^--nomfc/) { @$target[$T_FLAGS]&=~$TF_MFC; - @$target[$T_FLAGS]&=~($TF_MFC|$TF_WRAP); + @$target[$T_FLAGS]|=$TF_NOMFC; + } elsif ($option =~ /^--nodlls/) { + @$target[$T_FLAGS]|=$TF_NODLLS; } else { print STDERR "error: unknown option \"$option\"\n"; return 0; } } + if (@$target[$T_TYPE] != $TT_DLL && + @$target[$T_FLAGS] & $TF_MFC && + !(@$target[$T_FLAGS] & $TF_WRAP)) { + print STDERR "info: option --mfc requires --wrap"; + @$target[$T_FLAGS]|=$TF_WRAP; + } return 1; } @@ -543,7 +565,7 @@ } elsif ($dentry =~ /\.c$/i and $dentry !~ /\.spec\.c$/) { push @sources_c,"$dentry"; } elsif ($dentry =~ /\.(cpp|cxx)$/i) { - if ($dentry =~ /^stdafx.cpp$/i) { + if ($dentry =~ /^stdafx.cpp$/i && !(@$project_settings[$T_FLAGS] & $TF_NOMFC)) { push @sources_misc,"$dentry"; @$project_settings[$T_FLAGS]|=$TF_MFC; } else { @@ -553,7 +575,7 @@ push @sources_rc,"$dentry"; } elsif ($dentry =~ /\.(h|hxx|hpp|inl|rc2|dlg)$/i) { push @sources_misc,"$dentry"; - if ($dentry =~ /^stdafx.h$/i) { + if ($dentry =~ /^stdafx.h$/i && !(@$project_settings[$T_FLAGS] & $TF_NOMFC)) { @$project_settings[$T_FLAGS]|=$TF_MFC; } } elsif ($dentry =~ /\.dsp$/i) { @@ -779,7 +801,11 @@ # which we don't have in Wine. Also I add ntdll which seems # necessary for Winelib. my @std_dlls=qw(advapi32.dll comdlg32.dll gdi32.dll kernel32.dll ntdll.dll odbc32.dll ole32.dll oleaut32.dll shell32.dll user32.dll winspool.drv); - @$target[$T_DLLS]=\@std_dlls; + if (@$target[$T_FLAGS] & $TF_NODLLS == 0) { + @$target[$T_DLLS]=\@std_dlls; + } else { + @$target[$T_DLLS]=[]; + } push @{@$project[$P_TARGETS]},$target; # Ask for target-specific options @@ -2169,7 +2195,7 @@ print STDERR "Usage: winemaker [--nobanner] [--backup|--nobackup] [--nosource-fix]\n"; print STDERR " [--lower-none|--lower-all|--lower-uppercase]\n"; print STDERR " [--lower-include|--nolower-include]\n"; - print STDERR " [--guiexe|--windows|--cuiexe|--console|--dll]\n"; + print STDERR " [--guiexe|--windows|--cuiexe|--console|--dll|--nodlls]\n"; print STDERR " [--wrap|--nowrap] [--mfc|--nomfc]\n"; print STDERR " [-Dmacro[=defn]] [-Idir] [-Pdir] [-idll] [-Ldir] [-llibrary]\n"; print STDERR " [--interactive] [--single-target name]\n"; @@ -2182,9 +2208,6 @@ exit (2); } - -project_init(\@main_project,""); - while (@ARGV>0) { my $arg=shift @ARGV; # General options @@ -2245,11 +2268,13 @@ $opt_flags&=~$TF_WRAP; } elsif ($arg eq "--mfc") { $opt_flags|=$TF_MFC; - $opt_flags|=$TF_MFC|$TF_WRAP; $needs_mfc=1; } elsif ($arg eq "--nomfc") { - $opt_flags&=~($TF_MFC|$TF_WRAP); + $opt_flags&=~$TF_MFC; + $opt_flags|=$TF_NOMFC; $needs_mfc=0; + } elsif ($arg eq "--nodlls") { + $opt_flags|=$TF_NODLLS; # Catch errors } else { @@ -2264,6 +2289,11 @@ usage(); } } + + if ($opt_flags & $TF_MFC && $opt_target_type != $TT_DLL) { + print STDERR "info: option --mfc requires --wrap\n"; + $opt_flags |= $TF_WRAP; + }; } if (!defined $opt_work_dir) { @@ -2279,6 +2309,8 @@ print_banner(); } +project_init(\@main_project,""); + # Fix the file and directory names fix_file_and_directory_names("."); -- Martin Wilck Phone: +49 5251 8 15113 Fujitsu Siemens Computers Fax: +49 5251 8 20409 Heinz-Nixdorf-Ring 1 mailto:Martin.Wilck(a)Fujitsu-Siemens.com D-33106 Paderborn http://www.fujitsu-siemens.com/primergy

1 0

Winelib apps containing upper case letters don't run
by Martin Wilck Sept. 6, 2002

Sept. 6, 2002

The patch below fixes a problem introduced August 14 (process.c 1.196): Names of "builtin" apps (into which category fall also Winelib apps) aren't found if they aren't lowercase-only. Explanation: Create a directory "Test" and there a file foo.cpp like this: #include <iostream.h> int main (void) { cout << "Hello world!" <<endl; return 0; } In the "Test" directory, run $ winemaker --cuiexe .; ./configure; make; ./Test => error. This fails because the "Test" shell script (aka "wineloader") calls "wine Test.exe", and this ends up searching "test.exe.so" in the DLL path. My patch below fixes that, but I'm not sure if it is the right thing to do. Perhaps wineloader should simply look for "Test.exe.so" rather than for "Test.exe". This is only one of several problems I have with the current winelib/winemaker. Martin Patch: process.diff Martin Wilck <Martin.Wilck(a)fujitsu-siemens.com> scheduler: process.c open_builtin_exe_file(): Allow running Winelib apps whose names contain upper case letters. Index: scheduler/process.c =================================================================== RCS file: /home/wine/wine/scheduler/process.c,v retrieving revision 1.198 diff -u -r1.198 process.c --- scheduler/process.c 27 Aug 2002 01:13:59 -0000 1.198 +++ scheduler/process.c 6 Sep 2002 12:05:17 -0000 @@ -265,11 +265,15 @@ { char exename[MAX_PATH], *p; const char *basename = get_basename(name); + void *dll = NULL; if (strlen(basename) >= sizeof(exename)) return NULL; strcpy( exename, basename ); for (p = exename; *p; p++) *p = FILE_tolower(*p); - return wine_dll_load_main_exe( exename, error, error_size, test_only ); + dll = wine_dll_load_main_exe( exename, error, error_size, test_only ); + if (!dll) + dll = wine_dll_load_main_exe( basename, error, error_size, test_only ); + return dll; } -- Martin Wilck Phone: +49 5251 8 15113 Fujitsu Siemens Computers Fax: +49 5251 8 20409 Heinz-Nixdorf-Ring 1 mailto:Martin.Wilck(a)Fujitsu-Siemens.com D-33106 Paderborn http://www.fujitsu-siemens.com/primergy

1 0

Re: Comctrl generic notification support
by Andreas Mohr Sept. 6, 2002

Sept. 6, 2002

Hi, On Thu, Sep 05, 2002 at 01:54:50PM -0400, Dimitrie O. Paun wrote: > ChangeLog > Add generic notification support to comctl32 > > Index: dlls/comctl32/comctl32.h > =================================================================== > RCS file: /var/cvs/wine/dlls/comctl32/comctl32.h,v > retrieving revision 1.16 > diff -u -r1.16 comctl32.h > --- dlls/comctl32/comctl32.h 31 May 2002 23:25:43 -0000 1.16 > +++ dlls/comctl32/comctl32.h 5 Sep 2002 17:33:49 -0000 > @@ -4,6 +4,7 @@ > * > * Copyright 1999 Thuy Nguyen > * Copyright 1999 Eric Kohl > + * Copyright 2002 Dimitrie O. Paun > * > * This library is free software; you can redistribute it and/or > * modify it under the terms of the GNU Lesser General Public > @@ -23,6 +24,8 @@ > #ifndef __WINE_COMCTL32_H > #define __WINE_COMCTL32_H > > +#include "commctrl.h" > + > extern HMODULE COMCTL32_hModule; > extern HBRUSH COMCTL32_hPattern55AABrush; > > @@ -125,4 +128,132 @@ > #define WINE_FILEVERSION 5, COMCTL32_VERSION_MINOR, 0, 0 > #define WINE_FILEVERSIONSTR "5.00" > > +/* Notification support */ > + > +inline static LRESULT __send_notify(HWND hwnd, UINT code, NMHDR *hdr) Isn't using double underscore for such non-system functions rather pretty weird ? (or IOW: even strongly discouraged ?) -- Microsoft Licensing 6.0: "Pay us now in advance, so that we can own you later."

4 3

Re: Atom test fix
by Dmitry Timoshkov Sept. 5, 2002

Sept. 5, 2002

"Francois Gouget" <fgouget(a)free.fr> wrote: > Conformance tests must not use Wine-specific headers (like > wine/unicode.h) otherwise they will not compile on Windows. > > wine/test.h is the only exception. > > Changelog: > > * dlls/kernel/tests/atom.c > > Do not include wine/unicode.h > Duplicate strlenW I would suggest to simply replace it by lstrlenW, instead of duplicating code. -- Dmitry.

1 0

Re: VersionNumber reg. key
by Jeff Smith Sept. 5, 2002

Sept. 5, 2002

My $0.02: I say don't create, and dynamically 'lie' about it's value. I'm sure it's easier said than done, but that's my thoughts, anyway. -- Jeff S >From: Ender <winedev(a)admdev.com> >To: wine-devel(a)winehq.com >Subject: VersionNumber reg. key >Date: Wed, 4 Sep 2002 21:28:32 -0500 (CDT) >read the HKLM\\Software\\Microsoft\\Windows\\CurrentVersion\\VersionNumber >key. And of course, as this is a 'required' key in Windows, there was no >code to cope with failure :) >whether or not Wine should dynamically 'lie' about this registry key, >depending what winver the current thread is running in. > - Ender _________________________________________________________________ Chat with friends online, try MSN Messenger: http://messenger.msn.com

3 2

[Bug 448] - need help debugging 16-32-bit parameter passing
by Martin Wilck Sept. 5, 2002

Sept. 5, 2002

Included below are my latest results for Bug 448. The problem there is that winsock connects to port 512 although a service at port 5557 was specified. The winsock functions seem to work correctly: -> getservbyname16 returns port 0x15b5 = 5557 -> connect16 receives port 512 (!) directly after that. (the only thing that happens in between is a socket() call). The interesting part is that 512 in network byte order is 0x0002, i.e. the same as the protocol family (0x0002=AF_INET). Is it possible that something gets overwritten here when connect() is called? My problem is that I have no clue hall the magic relaying between 16 and 32 bit code in wine works. Perhaps someone can give me a hint? Martin ------- Additional Comments From martin.wilck(a)fujitsu-siemens.com 2002-09-05 07:58 ------- I finally got some time to look into this. It does not seem to be a winsock error. By inserting some TRACES, I can clearly see that getservbyname() correctly returns port 5557 (hex 0x15b5): trace:winsock:WINSOCK_getservbyname16 "ctsql", "tcp" trace:winsock:__ws_getservbyname name ctsql proto tcp flag 2 i 6 trace:winsock:WS_dup_se s_name ctsql s_port 15b5 s_proto tcp trace:winsock:WS_dup_se 16 s_port 15b5 trace:winsock:__ws_getservbyname s_name ctsql s_port 15b5 s_proto tcp [ ... ] But when "connect" is called, it is passed port 512 (hex 0200): trace:winsock:WINSOCK_connect16 socket 0048, ptr 0x40c493d6, length 16 sockaddr_in: family 2, address 172.25.186.5, port 512 02 00 02 00 ac 19 ba 05 00 00 00 00 00 00 00 00 Interestingly, (since the port is in network byte order) the hex patterns for address family and port are both 0x0002. This looks as if either - the app is assigning the family to the port number (?) (This would mean that the shouldn't be able to run on Windows either, since you say all was fine in 20011108, there must be another reason) - something goes wrong when the argument is passed from the 16- to 32bit layer. Have to look further. -- Martin Wilck Phone: +49 5251 8 15113 Fujitsu Siemens Computers Fax: +49 5251 8 20409 Heinz-Nixdorf-Ring 1 mailto:Martin.Wilck(a)Fujitsu-Siemens.com D-33106 Paderborn http://www.fujitsu-siemens.com/primergy

1 0

Trying to debug Paradise Poker client
by Per Nystrom Sept. 5, 2002

Sept. 5, 2002

Hi, I filed bug 1007 (kernel32.dll.Heap32* functions unimplemented) on this, but I'm not even sure that's the real problem. One thing that I'm seeing is retval=00000000 (FALSE) from all the calls to advapi32.CryptAcquireContextA() -- since this application relies *heavily* on encryption, I wonder if this is a good place to start. If I run it with the advapi32 dll overridden to use native, I get a different error and the App seems to crash even earlier in its execution. The following is what I got using the native advapi32 dll: fixme:file:VxDCall_VMM Unknown service 00010029 08070158:Ret kernel32.VxDCall4() retval=ffffffff ret=bfe81d8d fs=008f eax=ffffffff ebx=00000000 ecx=4011faa8 edx=405c6a00 esi=00000064 edi=00000001 ebp=405c6d5c esp=405c6d40 ds=002b es=002b gs=0ac7 flags=00000202 08070158:Call kernel32.CreateEventA(00000005,00000064,00000001,71467c2a "\205\300u\020h\024$EqSj\001S\377\327\243LPJqU\350\304\262\377\377;\363t\007V\377\025<\022Eq_^][\303\203=LPJq") ret=00000001 err:win32:CreateEventW Bad security attributes pointer 0x5 08070158:Ret kernel32.CreateEventA() retval=00000000 ret=00000001 wine: Unhandled exception, starting debugger... I'm attaching the entire relay trace (compressed) from running with the builtin advapi32 dll; it's much larger but probably more useful for fixing the problem. Any help is greatly appreciated! Per Nystrom System specifics: wine was built from the very latest in the CVS repository as of tonight Pentium II, 128Mb RAM and 1+ Gb swap running Red Hat 7.1 with updated X (4.2.0) glibc-2.2.4-29 gcc-2.96-81 -- "When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl."

2 1

Re: [PATCH] winsock-test.diff
by Alexandre Julliard Sept. 5, 2002

Sept. 5, 2002

Martin Wilck <Martin.Wilck(a)Fujitsu-Siemens.com> writes: > Rationale: > - Make it easier to add new unit tests. > - Conform better to the usual wine unit testing scheme > (instead of running different tests from the same file, > run separate files testing different functionality) That's not really the usual scheme. It may look that way because we have few tests at the moment, but we can't put each test in a separate file, or we will end up with thousands of them. The right approach is to follow more or less the layout of the source files, that is if two APIs are implemented in the same source file they should be tested in the same test file; so I think the current winsock tests are OK as they are. And in any case #including .c files is strongly discouraged. -- Alexandre Julliard julliard(a)winehq.com

2 1