March 2005 - Wine-devel - list.winehq.org

Re: winetest, winrash and new volunteers
by Steven Edwards 16 Mar '05

16 Mar '05

--- Chris Morgan <cmorgan(a)alum.wpi.edu> wrote: > winetest to upload the testing results. The winetest building is black magic > to me :-) Right now I am interested in the winetest side of things. Like how to get winetest and the server to talk to each other and what is needed in the build and url files. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

2 1

MSN Webcam patch
by Vladdy Impaler 16 Mar '05

16 Mar '05

RIght now I'm working on a patch so I can use my webcam under wine without having to switch to unix or use non-working solutions.. Right now there are 2 patches needed to get MSN to connect to the internet: first is to disable SSL, second strcpy's a value at urlmon getuseragent or something To get this to work, a few modifications to existing files are needed: First of all, if the catagory exists, let devenum return all from videoinput catagory: diff -Nru /root/wine-20050211/dlls/devenum/createdevenum.c wine-20050211/dlls/devenum/createdevenum.c --- /root/wine-20050211/dlls/devenum/createdevenum.c 2005-01-25 11:56:39.000000000 +0100 +++ wine-20050211/dlls/devenum/createdevenum.c 2005-03-15 00:15:47.000000000 +0100 @@ -117,7 +117,8 @@ if (IsEqualGUID(clsidDeviceClass, &CLSID_AudioRendererCategory) || IsEqualGUID(clsidDeviceClass, &CLSID_AudioInputDeviceCategory) || - IsEqualGUID(clsidDeviceClass, &CLSID_MidiRendererCategory)) + IsEqualGUID(clsidDeviceClass, &CLSID_MidiRendererCategory) || + IsEqualGUID(clsidDeviceClass, &CLSID_VideoInputDeviceCategory)) { hbasekey = HKEY_CURRENT_USER; strcpyW(wszRegKey, wszActiveMovieKey); and a few other patches to devenum main (Not sure if those are needed) diff -Nru /root/wine-20050211/dlls/devenum/devenum_main.c wine-20050211/dlls/devenum/devenum_main.c --- /root/wine-20050211/dlls/devenum/devenum_main.c 2004-12-07 15:37:11.000000000 +0100 +++ wine-20050211/dlls/devenum/devenum_main.c 2005-03-15 00:15:47.000000000 +0100 @@ -122,7 +122,7 @@ {&CLSID_AudioCompressorCategory, acmcat, TRUE}, {&CLSID_VideoCompressorCategory, vidcat, TRUE}, {&CLSID_LegacyAmFilterCategory, filtcat, TRUE}, - {&CLSID_VideoInputDeviceCategory, vfwcat, FALSE}, + {&CLSID_VideoInputDeviceCategory, vfwcat, TRUE}, {&CLSID_AudioInputDeviceCategory, wavein, FALSE}, {&CLSID_AudioRendererCategory, waveout, FALSE}, {&CLSID_MidiRendererCategory, midiout, FALSE}, @@ -156,7 +156,7 @@ pMapper = (IFilterMapper2*)mapvptr; - IFilterMapper2_CreateCategory(pMapper, &CLSID_VideoInputDeviceCategory, MERIT_DO_NOT_USE, friendlyvidcap); + IFilterMapper2_CreateCategory(pMapper, &CLSID_VideoInputDeviceCategory, MERIT_NORMAL, friendlyvidcap); IFilterMapper2_CreateCategory(pMapper, &CLSID_LegacyAmFilterCategory, MERIT_NORMAL, friendlydshow); IFilterMapper2_CreateCategory(pMapper, &CLSID_VideoCompressorCategory, MERIT_DO_NOT_USE, friendlyvidcomp); IFilterMapper2_CreateCategory(pMapper, &CLSID_AudioInputDeviceCategory, MERIT_DO_NOT_USE, friendlyaudcap); I wanted to put all custom code into qcap, but because of the pins that's not possible. I wrote a basic stub for CaptureGraphBuilder and added it and regsvr calls to qcap. There s no current implementation of qcap, so therefore I added my gzipped qcap.tgz as an attachment (To keep things simple) My current implementation of the actual interface was based on filesource.c, so i made some functions and struct defs global in quartz: diff -Nru /root/wine-20050211/dlls/quartz/filesource.c wine-20050211/dlls/quartz/filesource.c --- /root/wine-20050211/dlls/quartz/filesource.c 2005-01-06 20:36:47.000000000 +0100 +++ wine-20050211/dlls/quartz/filesource.c 2005-03-15 00:15:47.000000000 +0100 @@ -663,17 +663,6 @@ FileSource_GetCurFile }; - -/* the dwUserData passed back to user */ -typedef struct DATAREQUEST -{ - IMediaSample * pSample; /* sample passed to us by user */ - DWORD_PTR dwUserData; /* user data passed to us */ - OVERLAPPED ovl; /* our overlapped structure */ - - struct DATAREQUEST * pNext; /* next data request in list */ -} DATAREQUEST; - void queue(DATAREQUEST * pHead, DATAREQUEST * pItem) { DATAREQUEST * pCurrent; diff -Nru /root/wine-20050211/dlls/quartz/quartz_private.h wine-20050211/dlls/quartz/quartz_private.h --- /root/wine-20050211/dlls/quartz/quartz_private.h 2005-02-10 18:13:18.000000000 +0100 +++ wine-20050211/dlls/quartz/quartz_private.h 2005-03-15 00:15:47.000000000 +0100 @@ -52,8 +52,8 @@ HRESULT QUARTZ_CreateSystemClock(IUnknown * pUnkOuter, LPVOID * ppv); HRESULT ACMWrapper_create(IUnknown * pUnkOuter, LPVOID * ppv); HRESULT WAVEParser_create(IUnknown * pUnkOuter, LPVOID * ppv); - HRESULT EnumMonikerImpl_Create(IMoniker ** ppMoniker, ULONG nMonikerCount, IEnumMoniker ** ppEnum); +HRESULT VfwCapture_create(IUnknown * pUnkOuter, LPVOID * ppv); typedef struct tagENUMPINDETAILS { @@ -80,4 +80,14 @@ BOOL CompareMediaTypes(const AM_MEDIA_TYPE * pmt1, const AM_MEDIA_TYPE * pmt2, BOOL bWildcards); void dump_AM_MEDIA_TYPE(const AM_MEDIA_TYPE * pmt); +typedef struct DATAREQUEST +{ + IMediaSample * pSample; /* sample passed to us by user */ + DWORD_PTR dwUserData; /* user data passed to us */ + OVERLAPPED ovl; /* our overlapped structure */ + + struct DATAREQUEST * pNext; /* next data request in list */ +} DATAREQUEST; + +void queue(DATAREQUEST * pHead, DATAREQUEST * pItem); #endif /* __QUARTZ_PRIVATE_INCLUDED__ */ As you can see here, a VfwCapture class is being created to use it, we must register it and qcap with regsvr32 diff -Nru /root/wine-20050211/dlls/quartz/regsvr.c wine-20050211/dlls/quartz/regsvr.c --- /root/wine-20050211/dlls/quartz/regsvr.c 2005-02-10 18:13:18.000000000 +0100 +++ wine-20050211/dlls/quartz/regsvr.c 2005-03-15 00:15:47.000000000 +0100 @@ -922,6 +922,12 @@ "quartz.dll", "Both" }, + { &CLSID_VfwCapture, + "Video for wine capture interface", + NULL, + "quartz.dll", + "Both" + }, { NULL } /* list terminator */ }; @@ -1112,6 +1118,18 @@ { 0xFFFFFFFF }, } }, + { &CLSID_VfwCapture, + &CLSID_VideoInputDeviceCategory, + {'V','i','d','e','o',' ','F','o','r',' ','W','i','n','e',' ','(','V','4','W',')', 0}, + 0x800000, + { { REG_PINFLAG_B_OUTPUT, + { { &MEDIATYPE_Stream, &GUID_NULL }, + { NULL } + }, + }, + { 0xFFFFFFFF }, + } + }, /* This creates a fake device for us with an ugly name */ { NULL } /* list terminator */ }; Also, main.c needs to be informed of a new class: diff -Nru /root/wine-20050211/dlls/quartz/main.c wine-20050211/dlls/quartz/main.c --- /root/wine-20050211/dlls/quartz/main.c 2005-03-14 23:47:05.000000000 +0100 +++ wine-20050211/dlls/quartz/main.c 2005-03-15 00:15:47.000000000 +0100 @@ -71,7 +71,8 @@ { &CLSID_AVIDec, AVIDec_create }, { &CLSID_SystemClock, &QUARTZ_CreateSystemClock }, { &CLSID_ACMWrapper, &ACMWrapper_create }, - { &CLSID_WAVEParser, &WAVEParser_create } + { &CLSID_WAVEParser, &WAVEParser_create }, + { &CLSID_VfwCapture, &VfwCapture_create } }; static HRESULT WINAPI And at last, we need the interface of course: diff -Nru /root/wine-20050211/dlls/quartz/Makefile.in wine-20050211/dlls/quartz/Makefile.in --- /root/wine-20050211/dlls/quartz/Makefile.in 2005-02-10 18:13:18.000000000 +0100 +++ wine-20050211/dlls/quartz/Makefile.in 2005-03-15 00:15:47.000000000 +0100 @@ -28,7 +28,8 @@ systemclock.c \ transform.c \ videorenderer.c \ - waveparser.c + waveparser.c \ + v4wsource.c (v4wsource.c added as attachment) Before you think 'yey i can use a webcam now', I have to warn you, you won't What did is made from figuring out what MSN did at every step, for this to work and keeping it debuggable I ONLY used builtin dlls minimum required overrides in wine config: "quartz" = "builtin";FUCKING WORK!!!! "amstream" = "builtin" "qcap" = "builtin";You go girl..... "ole32" = "builtin";Not really needed, but makes debugging easier Currently this is NOT working, and MSN crashes for a reason I do not understand, Iif you want to help me continue working on webcam support i would greatly appreciate it if you could help me pinpoint the source of the crash.. WINEDEBUG=+quartz,+qcap is recommended.. and for now I'm only Currently the code mostly exists out of ugly hacks, what I was working on before msn crashed were the video formats (IKSPropertySet and IAMStreamConfig capabilities), if someone can help me figure out why and were it crashes, I could continue on my webcam project :=). You could try to get it to work through 'Extra' 'Configure webcam' (Or something like that, since I'm using Dutch version), and it crashes for a reason unknown to me.. PS: I accidentally deleted my first patch file, you need to add IAMStreamConfig and perhaps some other class into the include/*.idl files, I'll release a patch when I'm at least 1% awake again :/

2 1

Clipboard regression
by Ron Jensen 16 Mar '05

16 Mar '05

>Hello List, > >This patch from 2005/03/07 causes a Clipboard regression: > >http://www.winehq.org/hypermail/wine-cvs/2005/03/0111.html > >Whenever the clipboard is used the process crashes with an X error. >To reproduce it just start notepad and right click into the text area: This also seems to be the cause of X11 errors I am seeing with IrfanView and Textpad. Both apps fail at startup with an X Error, IrfanView is the simpler case: :~$ WINEDEBUG=clipboard wine C/Program\ Files/IrfanView/i_view32.exe trace:clipboard:X11DRV_CLIPBOARD_InsertClipboardFormat Registering format(49173): L"Rich Text Format" drvData 399 trace:clipboard:X11DRV_CLIPBOARD_InsertClipboardFormat Registering format(49174): L"GIF" drvData 396 trace:clipboard:X11DRV_IsClipboardFormatAvailable (0008) trace:clipboard:X11DRV_EmptyClipboard 0 entries remaining in cache. X Error of failed request: BadMatch (invalid parameter attributes) Major opcode of failed request: 1 (X_CreateWindow) Serial number of failed request: 208 Current serial number in output stream: 209 The offending function seems to be the calls to X11DRV_CLIPBOARD_QueryTargets() in X11DRV_CLIPBOARD_QueryAvailableData(). X11DRV_IsClipboardFormatAvailable() calls X11DRV_CLIPBOARD_UpdateCache() which calls X11DRV_CLIPBOARD_QueryAvailableData() which calls X11DRV_CLIPBOARD_QueryTargets() which throws the X Error. These apps both work if started in a desktop window. Ron Jensen

3 4

Re: shell32: set %AllUsersProfile%, and fix shell folder registration
by Juan Lang 16 Mar '05

16 Mar '05

--- Alexandre Julliard <julliard(a)winehq.org> wrote: > It's not a c:\windows, it's a %10%. setupapi takes care of putting the > right value in there (defaulting to c:\windows if nothing else is > specified). All I see in dirid.c is a call to GetWindowsDirectoryW. That, in turn, checks %windir%. So the windows directory appears to be set on a per-process basis, not on a per-install basis. This doesn't strike me as correct. --Juan __________________________________ Do you Yahoo!? Make Yahoo! your home page http://www.yahoo.com/r/hs

2 1

Re: shell32: set %AllUsersProfile%, and fix shell folder registration
by Juan Lang 16 Mar '05

16 Mar '05

--- Alexandre Julliard <julliard(a)winehq.org> wrote: > Not sure what you mean, the windows dir is definitely being set in > wine.inf, and you can use %10% in there to access it. Where? I don't see a c:\windows anywhere in there. The env var stuff in shellpath.c was always a hack though, and I think the registration bug was holding me back from working on it. I'll try to move all the "special" stuff out of it and into wine.inf. --Juan __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

2 1

Re: RICHEDIT: RTF reader improvements and optimizations (synced to current CVS)
by Alexandre Julliard 16 Mar '05

16 Mar '05

Krzysztof Foltman <kfoltman(a)portal.onet.pl> writes: > @@ -3721,6 +3723,22 @@ > else > RTFRouteToken(info); /* "\*" is ignored with known destinations */ > break; > + case rtfUnicode: > + { > + WCHAR buf[2]; > + buf[0] = info->rtfParam; The indentation is messed up, it looks like you set your tabs to 2 spaces. You should *never* do that, a tab is 8 spaces. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [bug2796] Richedit bug
by Paul Vriens 16 Mar '05

16 Mar '05

On Wed, 2005-03-16 at 18:54, Ann and Jason Edmeades wrote: > Nice bug: > > WCHAR *p = ALLOC_N_OBJ(WCHAR, nLen+1); > #define ALLOC_N_OBJ(type, count) (type *)HeapAlloc(me_heap, 0, > count*sizeof(type)) > > becomes > (WCHAR *)HeapAlloc(me_heap, 0, nLen+1*sizeof(WCHAR)) > > and by precedence rules means we allocate just nLen + 2 bytes rather than 2 > * (nLen+1) > > Changelog > > Correct memory allocation routine > > > Krzysztof - Does this solve the issue in the comment below it, ie the memory > overrun? > > Jason > Hi, this one also fixes a bug with riched20 when trying to install DVD Shrink 3.2.0.15. Thanks, Paul.

1 0

Re: [bug2796] Richedit bug
by Krzysztof Foltman 16 Mar '05

16 Mar '05

Ann and Jason Edmeades wrote: > WCHAR *p = ALLOC_N_OBJ(WCHAR, nLen+1); > #define ALLOC_N_OBJ(type, count) (type *)HeapAlloc(me_heap, 0, > count*sizeof(type)) That is so embarrasing! They teach this in the nursery school :/ By coincidence, I've spent at least half an hour today on finding the memory overwrite elsewhere, which turned out to be caused by the bug above. The buffer overrun comment refers to something else - I'm not sure if WideCharToMultiByte will produce single byte character string for ANSI codepages. If it produced a real multibyte string, it could exceed the buffer supplied by the application. Krzysztof

1 0

Re: IP Address control broken
by Rob Shearman 16 Mar '05

16 Mar '05

Robert Reif wrote: > Rob Shearman wrote: > >>> Disabling the control does not change it's color. >> It doesn't change colour when disabled on Windows. Of course, this is purely a user interface feature so we could make this a Wine-only extenstion. Rob

2 1

wineboot - Multi-User environment
by Joe Baker 16 Mar '05

16 Mar '05

It seems that if user1 is running a wine app and user2 is running a wine app, and user2 wants to perform a wineboot command, wineboot will wait till user1 closes out all wine applications. My observations are based on running winetools as user2 to install software while user1 continues to run a wine application. When it comes time to perform a "Windows Reboot", winetools just hangs there waiting for the wine processes to terminate. I think that wineboot should be modified to only attempt to stop processes owned by the user calling wineboot. -Joe Baker

1 0