wine-devel

wine-devel@winehq.org

2 participants
84469 discussions

Re: dsound tests propset.c windows compile patch
by Robert Reif 19 Jul '04

19 Jul '04

Robert Reif wrote: > Fix so it compiles with VC7 on XP. > > > Here is the proper fix. Don't include dsconf IIDs in dxguid.lib. Index: dlls/dsound/tests/propset.c =================================================================== RCS file: /home/wine/wine/dlls/dsound/tests/propset.c,v retrieving revision 1.5 diff -u -r1.5 propset.c --- dlls/dsound/tests/propset.c 5 Sep 2003 23:08:38 -0000 1.5 +++ dlls/dsound/tests/propset.c 19 Jul 2004 04:14:11 -0000 @@ -29,9 +29,12 @@ #include "windef.h" #include "wingdi.h" #include "dsound.h" +#include "initguid.h" #include "dsconf.h" -#include "initguid.h" +#ifndef DSBCAPS_CTRLDEFAULT +#define DSBCAPS_CTRLDEFAULT DSBCAPS_CTRLFREQUENCY|DSBCAPS_CTRLPAN|DSBCAPS_CTRLVOLUME +#endif DEFINE_GUID(DSPROPSETID_VoiceManager,0x62A69BAE,0xDF9D,0x11D1,0x99,0xA6,0x00,0xC0,0x4F,0xC9,0x9D,0x46); DEFINE_GUID(DSPROPSETID_EAX20_ListenerProperties,0x306a6a8,0xb224,0x11d2,0x99,0xe5,0x0,0x0,0xe8,0xd8,0xc7,0x22); Index: dlls/dxguid/dxguid.c =================================================================== RCS file: /home/wine/wine/dlls/dxguid/dxguid.c,v retrieving revision 1.2 diff -u -r1.2 dxguid.c --- dlls/dxguid/dxguid.c 3 Jun 2004 00:03:23 -0000 1.2 +++ dlls/dxguid/dxguid.c 19 Jul 2004 04:14:11 -0000 @@ -37,7 +37,6 @@ #include "mmsystem.h" #include "dsound.h" #include "dsdriver.h" -#include "dsconf.h" #include "dmusici.h" #include "dmusics.h" #include "dplay.h"

1 0

Re: dsound tests propset.c windows compile patch
by Robert Reif 19 Jul '04

19 Jul '04

Please disregard this patch. It breaks wine compiles. I'll dig into a better fix that works for both wine and windows. Robert Reif wrote: > Fix so it compiles with VC7 on XP. > > >

1 0

Re: dsound test: Remove unneeded Caps/3D test
by Robert Reif 19 Jul '04

19 Jul '04

Francois Gouget wrote: > This fixes another dsound Windows crash: we first create a buffer > without DSBCAPS_CTRL3D, then we query the IDirectSound3DBuffer on it > which obviously fails so that we try to do an AddRef on a NULL pointer > (buffer3d). > Fixes the right bugs and keeps intent of test. Adds more reference count checks. Tested on XP with DX9. Index: dlls/dsound/tests/dsound.c =================================================================== RCS file: /home/wine/wine/dlls/dsound/tests/dsound.c,v retrieving revision 1.27 diff -u -r1.27 dsound.c --- dlls/dsound/tests/dsound.c 17 Jul 2004 00:04:17 -0000 1.27 +++ dlls/dsound/tests/dsound.c 19 Jul 2004 00:34:23 -0000 @@ -342,14 +342,21 @@ init_format(&wfx,WAVE_FORMAT_PCM,11025,8,1); ZeroMemory(&bufdesc, sizeof(bufdesc)); bufdesc.dwSize=sizeof(bufdesc); - bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2 | DSBCAPS_CTRLVOLUME; + bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2 | DSBCAPS_CTRL3D; bufdesc.dwBufferBytes=wfx.nAvgBytesPerSec*BUFFER_LEN/1000; bufdesc.lpwfxFormat=&wfx; rc=IDirectSound_CreateSoundBuffer(dso,&bufdesc,&secondary,NULL); ok(rc==DS_OK && secondary!=NULL,"CreateSoundBuffer failed to create a secondary buffer 0x%lx\n", rc); if (rc==DS_OK && secondary!=NULL) { - /* add some more refs */ - IDirectSoundBuffer_AddRef(secondary); + LPDIRECTSOUND3DBUFFER buffer3d; + rc=IDirectSound_QueryInterface(secondary, &IID_IDirectSound3DBuffer, (void **)&buffer3d); + ok(rc==DS_OK && buffer3d!=NULL,"QueryInterface failed: %s\n",DXGetErrorString9(rc)); + if (rc==DS_OK && buffer3d!=NULL) { + ref=IDirectSound3DBuffer_AddRef(buffer3d); + ok(ref==2,"IDirectSound3DBuffer_AddRef has %d references, should have 2\n",ref); + } + ref=IDirectSoundBuffer_AddRef(secondary); + ok(ref==2,"IDirectSoundBuffer_AddRef has %d references, should have 2\n",ref); } /* release with buffer */ ref=IDirectSound_Release(dso); @@ -460,14 +467,27 @@ init_format(&wfx,WAVE_FORMAT_PCM,11025,8,1); ZeroMemory(&bufdesc, sizeof(bufdesc)); bufdesc.dwSize=sizeof(bufdesc); - bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2 | DSBCAPS_CTRLVOLUME; + bufdesc.dwFlags=DSBCAPS_GETCURRENTPOSITION2 | DSBCAPS_CTRL3D; bufdesc.dwBufferBytes=wfx.nAvgBytesPerSec*BUFFER_LEN/1000; bufdesc.lpwfxFormat=&wfx; rc=IDirectSound8_CreateSoundBuffer(dso,&bufdesc,&secondary,NULL); ok(rc==DS_OK && secondary!=NULL,"CreateSoundBuffer failed to create a secondary buffer 0x%lx\n", rc); if (rc==DS_OK && secondary!=NULL) { - /* add some more refs */ - IDirectSoundBuffer8_AddRef(secondary); + LPDIRECTSOUND3DBUFFER buffer3d; + LPDIRECTSOUNDBUFFER8 buffer8; + rc=IDirectSound_QueryInterface(secondary, &IID_IDirectSound3DBuffer, (void **)&buffer3d); + ok(rc==DS_OK && buffer3d!=NULL,"QueryInterface failed: %s\n",DXGetErrorString9(rc)); + if (rc==DS_OK && buffer3d!=NULL) { + ref=IDirectSound3DBuffer_AddRef(buffer3d); + ok(ref==2,"IDirectSound3DBuffer_AddRef has %d references, should have 2\n",ref); + } + rc=IDirectSound_QueryInterface(secondary, &IID_IDirectSoundBuffer8, (void **)&buffer8); + if (rc==DS_OK && buffer8!=NULL) { + ref=IDirectSoundBuffer8_AddRef(buffer8); + ok(ref==3,"IDirectSoundBuffer8_AddRef has %d references, should have 3\n",ref); + } + ref=IDirectSoundBuffer_AddRef(secondary); + ok(ref==4,"IDirectSoundBuffer_AddRef has %d references, should have 4\n",ref); } /* release with buffer */ ref=IDirectSound8_Release(dso);

1 0

Yard sale windows app fun: Lotus SmartSuite '96. 1-2-3 works, WordPro doesn't.
by Dan Kegel 18 Jul '04

18 Jul '04

I went to a yard sale today, and that's always dangerous. Today I ended up buying a copy of Lotus SmartSuite '96. I then installed Wine-20040615.tar.gz and tried installing my new disc with $ wine d:\\install.exe It got all confused, started iterating through every file underneath my home directory, and printed out err:shell:SHGetFileInfoW pidl is null! for each one! On a hunch, I made my home directory look empty, $ mkdir .blort $ mv * .blort $ wine d:\\install.exe That seemed much better; Wine only printed out the SHGetFileInfoW error once, presumably for the lotustmp.000 directory it created in my home directory. The next problem was when a dialog box "Select a Program Group..." popped up. At about the same time, an error box with title "Setup Message" and body "Setup has encountered a problem. Please report File='COMMON\SYMTAB11.C',Line='618' to Microsoft Product Support Services." This looks a bit lot like http://support.microsoft.com/default.aspx?scid=kb;EN-US;113252 which happens when GetWindowsDirectory() returns a string with lowercase letters, but clicking OK let me past with no ill affects, and I was able to install. At least when all I installed was Word Pro. When I also tried installing 1-2-3 and ScreenCam, I got the error "Lotus Setup" : "Install encountered an error trying to start Regedit.exe." That popped up twice, after which install thought for a long time, then warned fixme:ole:CoCreateInstance no classfactory created for CLSID {00021401-0000-0000-c000-000000000046}, hres is 0x80040154 and seemed to get stuck. Still, things seemed to be more or less well installed, so I tried running two of the apps. $ cd c/lotus/123/programs $ wine 123w.exe worked quite nicely. (Well, it did say the date was 6/17/104, but nobody's perfect!) Next I tried Word Pro: $ cd ~/.wine/c/lotus/wordpro $ wine wordpro.exe That failed immediately, so quickly that WINEDEBUG=+all wine wordpro.exe was practical to run. Here's part of its output. It looks as if the very first instruction accesses a null pointer in a strange way. 0009:Starting process L"C:\\lotus\\wordpro\\wordpro.exe" (entryproc=0x7a6aee) 0009:trace:seh:EXC_RtlRaiseException code=c0000005 flags=0 addr=0x7a6aee 0009:trace:seh:EXC_RtlRaiseException info[0]=00000001 0009:trace:seh:EXC_RtlRaiseException info[1]=00000000 0009:trace:seh:EXC_RtlRaiseException eax=00000000 ebx=403f1668 ecx=400c0be8 edx=400c0c37 esi=400c1000 edi=007a6aee 0009:trace:seh:EXC_RtlRaiseException ebp=4070fff4 esp=4070ff30 cs=0023 ds=002b es=002b fs=003b gs=0033 flags=00010246 0009:trace:seh:EXC_CallHandler calling handler at 0x403e1708 code=c0000005 flags=0 0009:Call ntdll.NtQueryVirtualMemory(ffffffff,00000000,00000000,4070fac0,0000001c,4070fa84) ret=4037aeb8 0009:Ret ntdll.NtQueryVirtualMemory() retval=00000000 ret=4037aeb8 0009:Call ntdll.RtlImageDirectoryEntryToData(00000000,00000001,00000002,4070fabc) ret=4033fbbe 0009:trace:seh:EXC_RtlRaiseException code=c0000005 flags=0 addr=0x40091368 0009:trace:seh:EXC_RtlRaiseException info[0]=00000000 0009:trace:seh:EXC_RtlRaiseException info[1]=00000000 0009:trace:seh:EXC_RtlRaiseException eax=00000000 ebx=400beec4 ecx=00000001 edx=00000000 esi=00000000 edi=00000002 0009:trace:seh:EXC_RtlRaiseException ebp=4070f9e4 esp=4070f928 cs=0023 ds=002b es=002b fs=003b gs=0033 flags=00010246 0009:trace:seh:EXC_CallHandler calling handler at 0x40089948 code=c0000005 flags=0 0009:trace:seh:EXC_RtlUnwind code=c0000005 flags=2 0009:trace:seh:EXC_CallHandler calling handler at 0x40088fc0 code=c0000005 flags=2 0009:trace:seh:EXC_CallHandler handler returned 1 0009:Ret ntdll.RtlImageDirectoryEntryToData() retval=00000000 ret=4033fbbe 0009:Call ntdll.NtCreateEvent(4070fac0,001f0003,4070fac4,00000001,00000000) ret=4033f836 0009:Ret ntdll.NtCreateEvent() retval=00000000 ret=4033f836 wine: Unhandled exception (thread 0009), starting debugger... ... Unhandled exception: page fault on write access to 0x00000000 in 32-bit code (0x007a6aee). In 32 bit mode. Register dump: CS:0023 SS:002b DS:002b ES:002b FS:003b GS:0033 EIP:007a6aee ESP:4070ff30 EBP:4070fff4 EFLAGS:00010246( - 00 -RIZP1) EAX:00000000 EBX:403f1668 ECX:400c0be8 EDX:400c0c37 ESI:400c1000 EDI:007a6aee Stack dump: 0x00000000: 00000000 00000000 00000000 00000000 0x00000010: 00000000 00000000 00000000 00000000 0x00000020: 00000000 00000000 00000000 00000000 0x00000030: 00000000 00000000 00000000 00000000 0x00000040: 00000000 00000000 00000000 00000000 0x00000050: 00000000 00000000 00000000 00000000 0x00000060: Backtrace: =>1 0x007a6aee (0x4070fff4) 2 0x4002e83d (0x00000000) 0x007a6aee: addb %al,0x0(%eax) Modules: 0x40070000-400c2000 Deferred ntdll 0x40320000-403f3000 Deferred kernel32 0x00400000-00888000 --none-- wordpro 0x402f0000-403f3000 Deferred kernel32<elf> 0x40019000-40025000 Deferred libnss_files.so.2 0x401b5000-401d7000 Deferred libm.so.6 0x400c2000-401b5000 Deferred libwine_unicode.so.1 0x40053000-400c2000 Deferred ntdll<elf> 0x40000000-40016000 Deferred ld-linux.so.2 0x4004f000-40052000 Deferred libdl.so.2 0x420155f0-841485f0 Deferred libc.so.6 0x40042000-4004f000 Deferred libpthread.so.0 0x4002a000-40042000 Sym libwine.so.1 0x420155f0-841485f0 Deferred libc.so.6 Module Address Debug info Name (14 modules) PE 0x00400000-00888000 --none-- wordpro ELF 0x40000000-40016000 Deferred ld-linux.so.2 ELF 0x40019000-40025000 Deferred libnss_files.so.2 ELF 0x4002a000-40042000 Sym libwine.so.1 ELF 0x40042000-4004f000 Deferred libpthread.so.0 ELF 0x4004f000-40052000 Deferred libdl.so.2 ELF 0x40053000-400c2000 Deferred ntdll<elf> \-PE 0x40070000-400c2000 Deferred ntdll ELF 0x400c2000-401b5000 Deferred libwine_unicode.so.1 ELF 0x401b5000-401d7000 Deferred libm.so.6 ELF 0x402f0000-403f3000 Deferred kernel32<elf> \-PE 0x40320000-403f3000 Deferred kernel32 ELF 0x420155f0-841485f0 Deferred libc.so.6 ELF 0x420155f0-841485f0 Deferred libc.so.6 Threads: process tid prio (all id:s are in hex) 0000000b (D) C:\lotus\wordpro\wordpro.exe 00000009 0 <== 00000010 00000011 0 WineDbg terminated on pid 0xb So I guess I got a couple hours worth of entertainment out of my $5, and I can now run Lotus 1-2-3 if I feel like it. Cool. - Dan -- My technical stuff: http://kegel.com My politics: see http://www.misleader.org for examples of why I'm for regime change

3 8

HttpOpenRequest: Who has to clean up lpszObjectName?
by Uwe Bonnes 18 Jul '04

18 Jul '04

Hallo, appended test fails with builtin wininet, but succeeds with native wininet. The failing call, triggered but the ending \r, seems to be caused by the www-server side handling. www.winehq.org seems to cope with the ending \r, so the test has to be carried out on the xilinx server. Ise webupdate fails caused by this error. What function has to clean up the ending \r? Thanks -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Index: wine/dlls/wininet/tests/http.c =================================================================== RCS file: /home/wine/wine/dlls/wininet/tests/http.c,v retrieving revision 1.13 diff -u -r1.13 http.c --- wine/dlls/wininet/tests/http.c 23 Jan 2004 22:44:26 -0000 1.13 +++ wine/dlls/wininet/tests/http.c 18 Jul 2004 20:36:28 -0000 @@ -275,10 +275,30 @@ ok((strcmp(TEST_URL_PATH,path) == 0),"path cracked wrong\n"); } +void InternetOpenRequest_test() +{ + DWORD rc; + HINTERNET hi, hic = 0, hor = 0; + const char *types[2] = { "*/*", NULL }; + hi = InternetOpenA("",0x0,0x0,0x0,0x0); + ok((hi != 0x0),"InternetOpen Failed\n"); + trace("InternetOpenA -->\n"); + hic=InternetConnectA(hi,"www.xilinx.com",0x0, 0x0, 0x0, 0x3,0x0, 0xdeadbeef); + ok((hic != 0x0),"InternetConnect Failed\n"); + trace("InternetConnect Succeeded\n"); + hor = HttpOpenRequestA(hic, "GET", + "/direct/swhelp/series6/6.2i_latestservicepack.dat\r", + "HTTP/1.0", + 0x0,types,04000000,00000000); + rc= HttpSendRequestA(hor,0x0,0x0,0x0,0x0); + ok((rc != 0x0),"HttpSendRequestA Failed\n"); +} + START_TEST(http) { winapi_test(0x10000000); winapi_test(0x00000000); InternetCrackUrl_test(); InternetOpenUrlA_test(); + InternetOpenRequest_test(); }

1 0

Re: msiexec.exe replacement
by Robert Shearman 18 Jul '04

18 Jul '04

Vincent Béron wrote: >Note: Much functionnality is still absent, but it's already able to >begin the installation of iTunes.msi and netfx.msi. >I plan on updating this in the coming days. > > Nice work. >Changelog: >First draft for a msiexec.exe replacement. > >Vincent > > ... >+ printf("Dll %s does not implement function %s\n", DllName, ProcName); > > Minor nitpick: this (and similar messages) should be fprintf(stderr,...) >+static void DllRegisterServer(LPCSTR DllName) >+{ >+ HRESULT hr; >+ DLLREGISTERSERVER pfDllRegisterServer = NULL; >+ HMODULE DllHandle = NULL; >+ >+ pfDllRegisterServer = LoadProc(DllName, "DllRegisterServer", &DllHandle); >+ >+ hr = pfDllRegisterServer(); >+ if(FAILED(hr)) >+ { >+ printf("Failed to register dll %s\n", DllName); >+ ExitProcess(1); >+ } >+ printf("Successfully registered dll %s\n", DllName); >+ if(DllHandle) >+ FreeLibrary(DllHandle); >+} >+ >+static void DllUnregisterServer(LPCSTR DllName) >+{ >+ HRESULT hr; >+ DLLUNREGISTERSERVER pfDllUnregisterServer = NULL; >+ HMODULE DllHandle = NULL; >+ >+ pfDllUnregisterServer = LoadProc(DllName, "DllUnregisterServer", &DllHandle); >+ >+ hr = pfDllUnregisterServer(); >+ if(FAILED(hr)) >+ { >+ printf("Failed to unregister dll %s\n", DllName); >+ ExitProcess(1); >+ } >+ printf("Successfully unregistered dll %s\n", DllName); >+ if(DllHandle) >+ FreeLibrary(DllHandle); >+} >+ > > ... >+ >+ ExitProcess(0); >+ return 0; >+} > > The ExitProcess call should not be needed. Rob

1 0

Re: PATCH: msrel32 new compiler
by Dmitry Timoshkov 18 Jul '04

18 Jul '04

"Marcus Meissner" <marcus(a)jet.franken.de> wrote: > --- dlls/msrle32/msrle32.c 27 Feb 2004 21:51:13 -0000 1.8 > +++ dlls/msrle32/msrle32.c 16 Jul 2004 12:36:10 -0000 > @@ -572,7 +572,8 @@ > > /* add EOL -- end of line */ > lpbiOut->biSizeImage += 2; > - *((LPWORD)lpOut)++ = 0; > + *(LPWORD)lpOut = 0; > + lpOut += sizeof(DWORD); Shouldn't this one and others be lpOut += sizeof(WORD) and not DWORD? -- Dmitry.

2 1

Re: [Darwine] Remove multi implementation of NtCurrentTeb
by maillard.emmanuel＠libertysurf.fr 18 Jul '04

18 Jul '04

Sorry the problem with NtAllocateVirtualMemory is in RtlCreateUserThread. emmanuel > NB : creating thread with wine-pthread still not work, because calling > NtAllocateVirtualMemory in start_thread always return > STATUS_INVALID_PARAMETER.

1 0

Missing RPMs on SourceForge
by Shachar Shemesh 18 Jul '04

18 Jul '04

Hi all, It seems that the RPMs for Redhat (at least redhat 9) on sourceforge are a fiction. When I actually try to download any of them (src, i386 or i686), I get a notification that the file is not on the mirror. This happens regardless of what mirror I select (I tried at least 5). Is this a problem we can solve? Shachar

2 1

{00021401-0000-0000-c000-000000000046}
by Shachar Shemesh 18 Jul '04

18 Jul '04

Or "ShellCreateIcon", if I'm not mistaken (I likely am). It appears we don't have that OLE registered in our system. This strikes me as odd, as I know that some installers do actually create icons. How do they do that? If someone can point me to our current icon creating code, I may sit down and create this object. Shachar -- Shachar Shemesh Lingnu Open Source Consulting ltd. http://www.lingnu.com/

3 2

Ogre3D based game crashes on exit
by Saulius Krasuckas 18 Jul '04

18 Jul '04

The game [1] crashes when I select DX7 engine at its initial screen, go the gameplay and then simply exit. I have entered a new bug report to provide more details [2] on this issue. My attention was caught by the ending part of the output [3] (quite large, about 400kB in size): trace:module:LdrUnloadDll (L"RenderSystem_Direct3D9.dll") - START trace:module:MODULE_DecRefCount (L"RenderSystem_Direct3D9.dll") ldr.LoadCount: 0 trace:module:MODULE_DecRefCount (L"OgreMain.dll") ldr.LoadCount: 3 trace:module:MODULE_DecRefCount (L"d3d9.dll") ldr.LoadCount: 0 trace:module:MODULE_DecRefCount (L"user32.dll") ldr.LoadCount: 13 trace:module:MODULE_DecRefCount (L"gdi32.dll") ldr.LoadCount: 8 trace:module:MODULE_DecRefCount (L"kernel32.dll") ldr.LoadCount: 23 trace:module:MODULE_DecRefCount (L"kernel32.dll") ldr.LoadCount: 22 trace:module:MODULE_DecRefCount (L"user32.dll") ldr.LoadCount: 12 trace:module:MODULE_DecRefCount (L"gdi32.dll") ldr.LoadCount: 7 trace:module:MODULE_DecRefCount (L"advapi32.dll") ldr.LoadCount: 8 trace:module:MODULE_DecRefCount (L"MSVCP71.dll") ldr.LoadCount: 4 trace:module:MODULE_DecRefCount (L"MSVCR71.dll") ldr.LoadCount: 5 trace:module:MODULE_InitDLL (0x41ca0000 L"RenderSystem_Direct3D9.dll",PROCESS_DETACH,(nil)) - CALL trace:module:MODULE_InitDLL (0x41ca0000,PROCESS_DETACH,(nil)) - RETURN 1 trace:module:MODULE_InitDLL (0x41d30000 L"d3d9.dll",PROCESS_DETACH,(nil)) - CALL trace:module:MODULE_InitDLL (0x41d30000,PROCESS_DETACH,(nil)) - RETURN 1 trace:module:MODULE_FlushModrefs unloading L"F:\\spba\\pluginsr\\RenderSystem_Direct3D9.dll" trace:module:MODULE_FlushModrefs unloading L"c:\\windows\\system\\d3d9.dll" trace:module:LdrUnloadDll END wine: Unhandled exception (thread 0009), starting debugger... <skip /> WineDbg starting on pid 0x8 Unhandled exception: page fault on write access to 0x00000000 in 32-bit code (0x430dc1e9). In 32 bit mode. 0x430dc1e9: sbbl %ebp,0x0(%ebx) Wine-dbg>Backtrace: =>1 0x430dc1e9 (0x4074fe86) Any ideas on how to investigate further on this crash? [1] http://lib.su.lt/~ramonas/spacebandits/ [2] http://bugs.winehq.org/show_bug.cgi?id=2365 [3] http://bugs.winehq.org/attachment.cgi?id=630&action=view

1 1

winetest files
by Ivan 18 Jul '04

18 Jul '04

In the future please avoid having 22 winetest packages set to active, it doubles the size of the sourceforge download page. Ivan.

1 0

PathSearchAndQualify patch (from 2003/12/22) not applied?
by Jeroen Janssen 17 Jul '04

17 Jul '04

Hello, I have an application that tries to use PathSearchAndQualify. However, since the current implementation is just stubbed (it contains only a FIXME + return FALSE), it doesn't work. Now searching in the gmane wine patch archive I found an email dated back to 2003/12/22 that contains a possible implementation for PathSearchAndQualify (see also at http://article.gmane.org/gmane.comp.emulators.wine.patches/6810/match=paths… ). However, it seems that this patch was never applied and I'm wondering why. Note that I haven't tried to apply this patch and test it locally. Can someone clarify if this patch is ok or not (and why it's not applied yet?) Best regards, Jeroen Janssen p.s. I hope this message does not appear multiple times, since I had some problems with gmane

2 1

Test contacts
by Mike Hearn 17 Jul '04

17 Jul '04

Hi, The test reports don't seem to include an email address for the reporters. I'd love to contact Nicolas Escuder or Dimitris Koukoravas to try and find out what their drive G: is and why it makes GetDiskFreeSpace fail. Does anybody know why the test system isn't getting their email addresses? thanks -mike

1 0

Re: TOOLBAR: TOOLBAR_DrawPattern
by Robert Shearman 17 Jul '04

17 Jul '04

Filip Navara wrote: > Changelog: > - Make the pattern drawing in TOOLBAR_DrawPattern look better. > > Filip Navara, navaraf(a)reactos.com > >------------------------------------------------------------------------ > >--- dlls/comctl32/toolbar.c 16 Jun 2004 06:46:13 -0000 1.15 >+++ dlls/comctl32/toolbar.c 16 Jul 2004 19:17:24 -0000 >@@ -634,7 +634,7 @@ > INT cy = lpRect->bottom - lpRect->top; > clrTextOld = SetTextColor(hdc, tbcd->clrBtnHighlight); > clrBkOld = SetBkColor(hdc, tbcd->clrBtnFace); >- PatBlt (hdc, lpRect->left, lpRect->top, cx, cy, PATCOPY); >+ PatBlt (hdc, lpRect->left + 2, lpRect->top + 2, cx - 4, cy - 4, PATCOPY); > SetBkColor(hdc, clrBkOld); > SetTextColor(hdc, clrTextOld); > SelectObject (hdc, hbr); > > This should use GetSystemMetrics(SM_CXEDGE and SM_CYEDGE). I'll submit a patch to fix this since it has already been committed. Rob

2 1

final fix collate icon behavior in printdlg
by James Hawkins 17 Jul '04

17 Jul '04

Hi, Sorry for the backtracking, but this patch conforms to the printer dialog in Windows, unlike the last patch i sent. The behavior is: Initially, Number of copies is 1 and collate is unchecked and shaded (icon is no_collate). If number of copies changes to > 1, collate becomes unshaded if collate is checked, icon changes to collate. if number of copies > 1 and you drop it to one, collate is shaded and does not check or uncheck so the icon stays the same as well. -- James Hawkins

1 0

Re: PrintDlg collate question
by J. Wesley Cleveland 17 Jul '04

17 Jul '04

Hi, With this patch, the behavior of the Copies section of the print dialog box is correct. The new behavior is: If number of copies is 1, collate option is unchecked and shaded because collate is not possible, and collate icon is removed whether it was set to collate or not collate. else if number of copies > 2, collate option is unchecked but not shaded, and no_collate icon is displayed. There is only one change that im not sure should go in or not, and that is when the collate option is checked and we drop the number of copies to one, should we uncheck the collate option as well as shade it? On my windows 2000 box, the collate option comes up checked and grayed as default, and the collate icon does not change as the number of copies is changed. The collate checkbox is grayed and ungrayed as the number of copies changes to and from 1, but the checked status remains whatever I set it to.

1 0

Re: fix collate icon behavior in printdlg
by Robert Shearman 16 Jul '04

16 Jul '04

James Hawkins wrote: >Hi, > > With this patch, the behavior of the Copies section of the print >dialog box is correct. The new behavior is: > >If number of copies is 1, > collate option is unchecked and shaded because collate is not possible, and > collate icon is removed whether it was set to collate or not collate. >else if number of copies > 2, > collate option is unchecked but not shaded, and > no_collate icon is displayed. > >There is only one change that im not sure should go in or not, and >that is when the collate option is checked and we drop the number of >copies to one, should we uncheck the collate option as well as shade >it? > > It is a nice touch that shows the user that the option doesn't mean anything in that state. However, if the user decreases the copies to 1 and then increases up to 2 their setting has been lost. Also, the dialog doesn't look right when in its default state of only 1 copy. There is a large gap where the collate image would have been. It would be nice if we could put another image there of say, only one sheet of paper, showing why the collate option is disabled. Rob

1 0

PrintDlg collate question
by James Hawkins 16 Jul '04

16 Jul '04

1 0

Debugger bugs
by Mike Hearn 16 Jul '04

16 Jul '04

Hi guys, I've got a few problems with various debuggers. Can anybody help? 1) gdb doesn't seem to catch signals to sent to Wine. In a crash, winedbg is invoked without gdb ever being aware of it. 2) winedbg can't locate files anymore, the paths seem to be stripped: 0x401cdb48 fixup_imports [loader.c:509] in ntdll: movl 0x0(%eax),%edx Unable to open file loader.c Wine-dbg>list Wine-dbg> 3) winedbg won't print the values of local variables: Wine-dbg>print i fixme:dbghelp:SymGetTypeInfo Someone didn't properly set ModBase (0x4078ebfc) fixme:winedbg:print_value ---error thanks -mike

2 3

DefPushButton management patch
by Katia Maculan 16 Jul '04

16 Jul '04

I found and fixed 2 problems in DEAFULTPUSHBUTTON management: 1) In function DIALOG_CreateControls32 if a default push button exists the dlgInfo->idResult is set with the DefPushButtonId. If I change the focus to other controls of the dialog with the TAB key, each time I change the focus the function DEFDLG_SetDefButton is called (and that is correct) and the dlgInfo->idResult is set with the new control id also if it is NOT a button (i.e. "edit control"). As a result, pressing the 'Enter'key, nothing happens. I reviewed the code and changed the following: 1. added a member in the DIALOGINFO structure (int OriDefButID) where the original default button ID is stored. 2. modified the DEFDG_SetDefButton function (windows/defdlg.c). 2) Pressing the Enter key in a Dialog without a default push-button, a wrong button ID is returned (always '1'). I modifiied the code (DIALOG_CreateControls32 function in ./windows/dialog.c), so that at least an existing button-ID is stored as a default. NB: I can't use CVS (for a login problem) -- Katia Maculan Developer ISIS Papyrus Italy Srl Tel: (+39) 0125 6455400 Fax: (+39) 0125 6455150 E-mail: katia.maculan(a)isis-papyrus.com Info: info(a)isis-papyrus.com Hotline: +43-2236-27551-111 Visit our Website: http://www.isis-papyrus.com diff -urN oldwinedir/dlls/user/controls.h newwinedir/dlls/user/controls.h --- oldwinedir/dlls/user/controls.h 2004-03-04 02:41:11.000000000 +0100 +++ newwinedir/dlls/user/controls.h 2004-07-15 14:05:44.000000000 +0200 @@ -126,6 +126,7 @@ INT idResult; /* EndDialog() result / default pushbutton ID */ UINT flags; /* EndDialog() called for this dialog */ HGLOBAL16 hDialogHeap; + INT OriDefButID; /* the original DEFPUSHBUTTON */ } DIALOGINFO; #define DF_END 0x0001 diff -urN oldwinedir/windows/defdlg.c newwinedir/windows/defdlg.c --- oldwinedir/windows/defdlg.c 2004-04-20 06:02:35.000000000 +0200 +++ newwinedir/windows/defdlg.c 2004-07-15 14:07:12.000000000 +0200 @@ -135,13 +135,27 @@ { DWORD dlgcode=0; /* initialize just to avoid a warning */ HWND hwndNew = GetDlgItem(hwndDlg, wParam); - - dlgInfo->idResult = wParam; + //more tests are necessary before save that id like a defpushbutton dlgInfo->idResult = wParam; if (hwndNew && !((dlgcode=SendMessageW(hwndNew, WM_GETDLGCODE, 0, 0 )) & (DLGC_UNDEFPUSHBUTTON | DLGC_BUTTON))) + { + //if destination is not a push button the original def push button is set like default + if (dlgInfo->idResult) + { + HWND hwndOld = GetDlgItem( hwndDlg, dlgInfo->idResult ); + if (hwndOld && (SendMessageA( hwndOld, WM_GETDLGCODE, 0, 0) & DLGC_DEFPUSHBUTTON)) + SendMessageA( hwndOld, BM_SETSTYLE, BS_PUSHBUTTON, TRUE ); + } + dlgInfo->idResult = dlgInfo->OriDefButID; + hwndNew = GetDlgItem (hwndDlg, dlgInfo->OriDefButID); + if (hwndNew) + SendMessageA( hwndNew, BM_SETSTYLE, BS_DEFPUSHBUTTON, TRUE ); return FALSE; /* Destination is not a push button */ + } + if (dlgInfo->idResult == wParam) //if new defbutton is the actual nothing change + return TRUE; if (dlgInfo->idResult) /* There's already a default pushbutton */ { HWND hwndOld = GetDlgItem( hwndDlg, dlgInfo->idResult ); @@ -150,8 +164,11 @@ } if (hwndNew) { - if(dlgcode==DLGC_UNDEFPUSHBUTTON) + if(dlgcode & DLGC_UNDEFPUSHBUTTON) //it is DLGC_UNDEFBUTTON && DLGC_BUTTON + { SendMessageA( hwndNew, BM_SETSTYLE, BS_DEFPUSHBUTTON, TRUE ); + dlgInfo->idResult = wParam; // now is correct to save the new def push button + } } return TRUE; } diff -urN oldwinedir/windows/dialog.c newwinedir/windows/dialog.c --- oldwinedir/windows/dialog.c 2004-05-18 22:45:22.000000000 +0200 +++ newwinedir/windows/dialog.c 2004-07-15 14:07:05.000000000 +0200 @@ -292,6 +292,7 @@ DLG_CONTROL_INFO info; HWND hwndCtrl, hwndDefButton = 0; INT items = dlgTemplate->nbItems; + DWORD dlgcode=0; //to have the control information TRACE(" BEGIN\n" ); while (items--) @@ -353,7 +354,8 @@ /* Send initialisation messages to the control */ if (dlgInfo->hUserFont) SendMessageA( hwndCtrl, WM_SETFONT, (WPARAM)dlgInfo->hUserFont, 0 ); - if (SendMessageA(hwndCtrl, WM_GETDLGCODE, 0, 0) & DLGC_DEFPUSHBUTTON) + dlgcode = SendMessageA(hwndCtrl, WM_GETDLGCODE, 0, 0); //save the result for future test + if (dlgcode & DLGC_DEFPUSHBUTTON) { /* If there's already a default push-button, set it back */ /* to normal and use this one instead. */ @@ -361,7 +363,21 @@ SendMessageA( hwndDefButton, BM_SETSTYLE, BS_PUSHBUTTON, FALSE ); hwndDefButton = hwndCtrl; dlgInfo->idResult = GetWindowLongA( hwndCtrl, GWL_ID ); + dlgInfo->OriDefButID = dlgInfo->idResult; //save the ID like the original DEFPUSHBUTTON } + else + { + // if no defpushbutton is found and the control is a normal pushbutton + // this button is set like defpushbutton + if ((dlgInfo->idResult == 0) && (dlgcode & DLGC_UNDEFPUSHBUTTON)) + { + hwndDefButton = hwndCtrl; + dlgInfo->idResult = GetWindowLongA( hwndCtrl, GWL_ID ); + dlgInfo->OriDefButID = dlgInfo->idResult; + SendMessageA( hwndDefButton, BM_SETSTYLE, BS_DEFPUSHBUTTON, FALSE ); + } + + } } TRACE(" END\n" ); return TRUE;

2 2

printdlg question
by truiken＠gmail.com 16 Jul '04

16 Jul '04

Hi, In the PRINTDLG_WMInitDialog() and PRINTDLG_SetUpPrinterListCombo() functions, what purpose does the 'name' variable server? I know it's the name of the current printer, but what role does it play in setting up the printer list combo? dlls/commdlg/printdlg.c Thanks, James Hawkins

1 0

Re: wine-pthread use the TEB register on PowerPC
by Alexandre Julliard 16 Jul '04

16 Jul '04

"Pierre d'Herbemont" <pidherbemont(a)opendarwin.org> writes: > This patch enables the use of wine-pthread on darwin, and more > generally on PowerPC hosts. > > Pierre. > > ChangeLog: > Use the TEB register on PowerPC hosts. The intent is that wine-pthread should use only portable pthread functions. Why do you need the TEB register to be setup when using wine-pthread? -- Alexandre Julliard julliard(a)winehq.org

1 0

wrong test results on test.winehq.org
by Stefan Leichter 15 Jul '04

15 Jul '04

Hello, is it right that the test result in the directory 200407111000 are build from cvs of the 11th July 2004? If no please delete the mail now ;-) Otherwise please take a look to the failing kernel:profile tests results of e.g win98 (http://test.winehq.org/data/200407111000/98se_JosephBooker/kernel32:profile…) you will see lines like profile.c:82: Test failed: test<01>: ret<0000000000> exp<4294967295> but if you check the source code of the tests (http://cvs.winehq.org/cvsweb/wine/dlls/kernel/tests/profile.c?rev=1.5&conte…) you will find that a second number must be displayed at the end of the line ok((res == profileInt[i].result) || (res == profileInt[i].result9x), "test<%02d>: ret<%010u> exp<%010u><%010u>\n", ^^^^^ here i, res, profileInt[i].result, profileInt[i].result9x); So either the test binary used old source files or the results are sorted into a wrong "wine version". Can please someone have a look Thanks Stefan

2 1

Re: Prevent heap corruption in CryptAcquireContext
by Alexandre Julliard 15 Jul '04

15 Jul '04

Michael Jung <mjung(a)iss.tu-darmstadt.de> writes: > this is a resubmit of a simple patch I've sent yesterday, but which was not > applied to cvs. I would like to give some detail on it: There is a section of > code labeled "error:" at the end of CryptAcquireContext, which releases > occupied resources in the case of an error. Amongst others, the allocated > memory pointed to by "keyname" is released. However, in some cases keyname is > released twice, once in the normal code and once in the error cleanup code. > CRYPT_Free uses LocalFree, which returns "NULL" if the memory was > successfully released. And it does basically nothing if it is called with a > "NULL" pointer. This means that by executing "keyname = > (PSTR)CRYPT_Free(keyname)" the second call to CRYPT_Free does basically > nothing, which is better than corrupting the heap ;) The patch is OK, but it would be better to really avoid calling CRYPT_Free twice, and also to fix all the other places in that function that have the same problem. -- Alexandre Julliard julliard(a)winehq.org

2 1

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