wine-devel August 2005

wine-devel@winehq.org

132 participants
254 discussions

Re: Loading resources from wine dlls
by Michael Jung 26 Aug '05

26 Aug '05

Hi Rob, On Thursday 25 August 2005 06:50, you wrote: > Anything that uses SearchPath is similarly broken when needing to find a > builtin dll or exe. We should add a hack^H^H^H^Hextension to that > function to make it be able to find installed .so files with a similar > name. I think that's just part of the problem. If I create a symbolic link from .wine/drive_c/windows/system32/shell32.dll to /usr/local/lib/wine/shell32.dll.so SearchPath succeeds. But then USER32_GetResourceTable (dlls/user/exticon.c, line 128) fails (I guess it can't handle ELF wrapped PEs?). Would it be ok to look for an ELF header and then for an enclosed PE header in USER32_GetResourceTable (and then return IMAGE_NT_SIGNATURE and the correct pointer)? Where can I find informations about the format of ELF wrapped PEs? Bye, -- Michael Jung mjung(a)iss.tu-darmstadt.de

2 2

Re: wine/dlls/commdlg printdlg.c
by Vitaly Lipatov 26 Aug '05

26 Aug '05

В сообщении от 25 Август 2005 23:23 Alexandre Julliard написал(a): > Log message: > Vitaly Lipatov <lav(a)etersoft.ru> > Restore paper orientation when pagesetupdlg is initializing. > > Patch: http://cvs.winehq.org/patch.py?id=19818 Dear Alexandre! You are removed dm checking for NULL... But it can be NULL in some cases... <<<<<<< printdlg.c /* Set orientation radiobutton properly */ dm = GlobalLock(pda->dlga->hDevMode); if (dm && dm->u1.s1.dmOrientation == DMORIENT_LANDSCAPE) ======= /* Set orientation radiobutton properly */ dm = GlobalLock(pda->dlga->hDevMode); if (dm->u1.s1.dmOrientation == DMORIENT_LANDSCAPE) >>>>>>> 1.93 -- Vitaly Lipatov, ALT Linux Team Russia, Saint-Petersburg, www.etersoft.ru

2 1

DLL loading
by Bojan �ffff8aernek 26 Aug '05

26 Aug '05

hello, I'd like to load a DLL from a memory buffer. The regular Win32 API function to load DLLs, however, takes a path name. So, I'm looking for resources about this kind of thing, but it has to work both under win32 and unix+wine. There used to be a web page about just this thing, but I lost the link. Any ideas? Thanks. ____________________________________________________ Start your day with Yahoo! - make it your home page http://www.yahoo.com/r/hs

2 1

undocumented shell32 stuff
by Saulius Krasuckas 26 Aug '05

26 Aug '05

I have stumbled upon a unimplemented function shell32.dll.SHLocalAlloc . OpenRCE page says it is related to kernel32.dll.LocalAlloc [1], and authors of the NSElib (The NameSpace Extension Library) defines it taking two int-params [2]. They define the function in ShlExt.h file which says: // This file can be considered to be in the public domain. // The latest version can normally be found here: // http://www.whirlingdervishes.com/nselib/public // The file is based on various sources, and includes definitions // that were taken over from various places, including // Microsoft documentation and header files. It also contains // many definitions that were introduced by Whirling Dervishes // Software. OTOH, there is one page on the web [3], which declares itself as a homepage of the ShlExt.h and which describes history of the file. It seems as prototype which was extended by Whirling Dervishes Software and based on which their NSElib was built. Earlier ShlExt.h was beeing included in the PSDK, but now it isn't AFAICS. And this is another quote from [2]: // februari 2003: Henk Devos: Made compatible with the // october 2003 Platform SDK // januari 2003: Henk Devos: added some more undefined flags, // including the ones for CDefFolderMenu_Create2. // september 2002: Henk Devos: Made compatible with new Platform SDK // januari 2002: Henk Devos: Added comments with ordinals of // all exports in Windows 2000 version of shell32.dll // 24/06/99 Some changes made by Henk Devos It stays above the changelog of the old file [4]. What developers think about this story? Can we start importing it into our tree? If not, should I declare SHLocalAlloc() somewhere? And where should I put it's code, maybe dlls/shell32/utility.c ? [1] http://www.openrce.org/reference_library/win32_call_chains/2000SP4/SHLWAPI/… [2] http://www.whirlingdervishes.com/nselib/mfc/samples/source.php [3] http://www.mvps.org/user32/shlext.html#shlext [4] http://www.mvps.org/user32/shlext.zip

2 2

Help with debugging needed
by Stefan Dösinger 26 Aug '05

26 Aug '05

Hello, I am trying(once again) to get Empire Earth [1] running with wine. After solving a few problems[2], I've run into a strange access violation crash. It looks like a buffer overrun, followed by a return to an invalid adress. A +relay,+snoop trace gives this: ... 0009:CALL Low-Level Engine.?Deactivate@GERasterizer@@UAEJXZ() ret=7e2b9a14 0009:CALL Low-Level Engine.?Dereference@UReferenceCounter@@QAEXXZ() ret=7e2bd6c5 0009:RET Low-Level Engine.?Dereference@UReferenceCounter@@QAEXXZ() retval=7f24fa70 ret=7e2bd6c5 0009:Call ntdll.RtlEnterCriticalSection(7f074a40) ret=7f0294c5 0009:Ret ntdll.RtlEnterCriticalSection() retval=00000000 ret=7f0294c5 0009:Call ntdll.RtlLeaveCriticalSection(7f074a40) ret=7f029506 0009:Ret ntdll.RtlLeaveCriticalSection() retval=00000000 ret=7f029506 0009:CALL MSVCRT._except_handler3(<unknown, check return>) ret=7beb649b 0009:CALL MSVCRT._XcptFilter(<unknown, check return>) ret=00665947 0009:Call kernel32.GetLastError() ret=7800385f 0009:Ret kernel32.GetLastError() retval=000005b4 ret=7800385f 0009:Call kernel32.TlsGetValue(00000000) ret=7800386d 0009:Ret kernel32.TlsGetValue() retval=7f180f80 ret=7800386d 0009:Call kernel32.SetLastError(000005b4) ret=7800387e 0009:Ret kernel32.SetLastError() retval=000005b4 ret=7800387e 0009:Call kernel32.UnhandledExceptionFilter(7fc2f81c) ret=7800ed71 0009:Call ntdll.NtQueryVirtualMemory(ffffffff,9803a11c,00000000,7fc2f6e8,0000001c,7fc2f6a8) ret=7fd1aae1 0009:Ret ntdll.NtQueryVirtualMemory() retval=00000000 ret=7fd1aae1 0009:Call kernel32.IsBadCodePtr(7800add3) ret=22d2b743 0009:Ret kernel32.IsBadCodePtr() retval=00000000 ret=22d2b743 0009:Call ntdll.NtCreateEvent(7fc2f6e8,001f0003,7fc2f6ec,00000001,00000000) ret=7fccb9b9 0009:Ret ntdll.NtCreateEvent() retval=00000000 ret=7fccb9b9 wine: Unhandled exception (thread 0009), starting debugger... ... The crash messages look a little bit different every time, with illegal instructions / access violations or even a Segmentation Fault without starting winedbg at various addreses(Attached file crashes). On very rare occasions, this crash doesn't occur, and the game continues to load and crashes later in some ddraw function. Instead of crashing it complains about a corrupted heap: err:heap:HEAP_ValidateInUseArena Heap 0x7fd80000: prev arena 0x7fe01640 is not prev for in-use 0x7fe01cb0 I've looked at a +heap trace, but I couldn't find anything usefull. (See ee-nocrash for a log). I've also attached a normal log without any special debug flags set(ee-normal.log.gz). I've added a few ERR statements for testing in some functions. So my questions are: *Am I right with my suspection that the problems are caused by a incorrect return? *How can I get a disassembly of Low-Level Engine.?Deactivate@GERasterizer@@UAEJXZ or simmilar functions. I didn't find this symbol. Thanks for your help, Stefan Dösinger [1] A demo is available at http://www.vugames.de/scripts/download.aspx?content=4&id=53721&nf=21&url=ft…, but I haven't checked wheter it produces the same results. [2] Other users reported a msvc runtime error, apparently caused by an error returned by Main_DirectDraw_SetCooperativeLevel. EE calls this function with cooplevel == DDSCL_SETFOCUSWINDOW. I made the function simply return DD_OK in this case. Furthermore it checks for sound support and the OS version(setting it to Win98 or WinME is recommended) and it needs native msvcrt(It even seems to ship it's own version)

5 20

Re: PATCH: sensap IsNetworkAlive
by Vijay Kiran Kamuju 26 Aug '05

26 Aug '05

Hi, According to MSDN, >+#define NETWORK_ALIVE_RAS 2 must be #define NETWORK_ALIVE_WAN 2 I think so, see http://msdn.microsoft.com/library/default.asp?url=/library/en-us/syncmgr/sy… Thanks, Vijay

1 0

More compatible memory arrangement ?: dlls/winsock.c: WS_dup_pe()
by Krzysztof C. 26 Aug '05

26 Aug '05

Hi, I was having a problem with a game, it couldn't connect to the server, so I started debugging things, and it ofcourse brought me to dlls/winsock.c but the problem seems to be on the game side ( unfortunately its closed-source, so I can only guess what's happening ). The debugging showed that when the game was getting server's ip from DNS (WS_gethostbyname()), it was getting wrong values. Then I started looking for the cause, and I found out that the game was incorrectly reading the hostent/WS_hostent structure. It was assuming that the ip address was at particular offset from the beginning of the structure, it didn't follow the pointers and thus it was getting wrong data. Then I rewrote few lines of the code and now it looks like: "pointer | pointer | pointer | ... data | data | data | .... " and everything works fine. So I assume that Microsoft's winsock arrange this structure this way, but I am not sure if it is 100% compatible way solving "one application problem", although few other programs dependent on this function worked fine with this change. Anyway I left the topic with question mark because I would like to know someone else's opinion on that. ( btw. I am new to mailing lists and it is my first post, so please don't be too rough on me in your responses : ) Index: dlls/winsock/socket.c =================================================================== RCS file: /home/wine/wine/dlls/winsock/socket.c,v retrieving revision 1.188 diff -u -p -r1.188 socket.c --- dlls/winsock/socket.c 25 Aug 2005 10:22:12 -0000 1.188 +++ dlls/winsock/socket.c 25 Aug 2005 22:20:58 -0000 @@ -3343,16 +3343,16 @@ static struct WS_hostent *WS_dup_he(cons p_to->h_addrtype = p_he->h_addrtype; p_to->h_length = p_he->h_length; - p = (char *)(p_to + 1); + p = (char *)p_to + sizeof(struct hostent); + p_to->h_addr_list = (char **)p; + p += list_dup(p_he->h_addr_list, p_to->h_addr_list, p_he->h_length); + p_to->h_name = p; - strcpy(p, p_he->h_name); - p += strlen(p) + 1; + strcpy(p_to->h_name, p_he->h_name); + p += strlen(p_he->h_name) + 1; p_to->h_aliases = (char **)p; - p += list_dup(p_he->h_aliases, p_to->h_aliases, 0); - - p_to->h_addr_list = (char **)p; - list_dup(p_he->h_addr_list, p_to->h_addr_list, p_he->h_length); + list_dup(p_he->h_aliases, p_to->h_aliases, 0); return p_to; }

1 0

Re: Need some help with a patch for version.dll
by Stefan Dösinger 25 Aug '05

25 Aug '05

Hi, > I can't use a LPCSTR, cause: > > BOOL WINAPI VerQueryValueA( LPVOID pBlock, LPCSTR lpSubBlock, > LPVOID *lplpBuffer, UINT *puLen ) > > lplpBuffer points to a place in pBlock where the value is. This value is > in most cases a UNICODE string. This string has to be converted to ANSI > and placed on a different place in pBlock (+ info->wLength + 4). > > Hope this makes sense. Oh, sure, you can't change it here. I guess you need a typecast here to supress the warning. But there are some helper functions for converting a unicode string to an ansi string. If you can use them, this might save you some troubles. Stefan

1 0

Need some help with a patch for version.dll
by Paul Vriens 25 Aug '05

25 Aug '05

Hi, I've finally nailed the last (?) bugs in version.dll. The attached patch shows how. There is however a warning during compile and I have no clue (still learning C :-) ) how to fix that: info.c: In function `VerQueryValueA': info.c:754: warning: pointer of type `void *' used in arithmetic info.c:754: warning: pointer of type `void *' used in arithmetic info.c: In function `VerQueryValueW': info.c:817: warning: pointer of type `void *' used in subtraction info.c:817: warning: pointer of type `void *' used in arithmetic Any idea's ? Cheers, Paul.

2 1

Compiling winelib dependant code with g++
by Gaudet Michael-MGAU01 25 Aug '05

25 Aug '05

wineg++ is using g++, and is passing its options to g++. How can I link some winelib-dependant code with g++ ? compilation is okay, but at linking time, all wine functions are unresolved. // ### here's the output :### [root@udp015809uds Wine2]# g++ -lwine registry.o registry_wrap.o -o registry.out registry.o(.text+0x2e): In function `CreateKey(char const*)': : undefined reference to `RegCreateKeyA' registry.o(.text+0x58): In function `CreateKey(char const*)': : undefined reference to `RegQueryValueExA' registry.o(.text+0x8f): In function `CreateKey(char const*)': : undefined reference to `RegQueryValueExA' registry.o(.text+0xa5): In function `CreateKey(char const*)': : undefined reference to `MessageBoxA' collect2: ld returned 1 exit status // ### And the same code compiled with wineg++ ### [root@udp015809uds Wine2]# wineg++ -lwine registry.o registry_wrap.o -o registry.out // program runs okay. For information, I need to use g++ in order to assimilate wine lib framework to SWIG, to generate Win32 wrappers in C# for mono. Thanks.

2 1

← Newer
1
...
5
6
7
8
9
10
11
...
26
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-devel August 2005