September 2008 - Wine-devel

Re: DirectDraw: fix compilation on mingw
by Henri Verbeet 29 Sep '08

29 Sep '08

2008/9/28 Roderick Colenbrander <thunderbird2k(a)gmx.net>: > From: Roderick Colenbrander <roderick(a)roderick-laptop.(none)> You're certainly not the only one who does this, but you're supposed to have a proper email address in there.

1 0

Re: wined3d: Convert between A8R8G8B8 and X8R8G8B8
by Henri Verbeet 29 Sep '08

29 Sep '08

2008/9/29 Kjell Rune Skaaraas <kjella79(a)yahoo.no>: > Hello, > > A simple patch which adds converting between A8R8G8B8 and X8R8G8B8 that > gives me one less warning to deal with while trying to make a game function. > It zeroes out the alpha channel to be on the safe side, though it should > never be used in the destination format anyway. > What does this fix, in practice?

1 0

Re: dsound: GetVolume should check error from driver (resend)
by Kirill K. Smirnov 29 Sep '08

29 Sep '08

Hi, Dan! >ALSA_CheckSetVolume returns right early because hctl is NULL, >but that error is not propagated through dsound properly. >While we should figure out the alsa part, for now, let's at least >propagate the error, and skip tests that can't be done without >sensing volume. >If this patch is accepted, I'll send a second one to add a >similar skip() to tests/ds3d8.c. 0) Prehistory. I've fought with this issue too and sent a patch: http://www.winehq.org/pipermail/wine-patches/2008-July/058124.html Reply was: http://www.winehq.org/pipermail/wine-devel/2008-July/067803.html 1) About your patch. There are 4 places which suffers from this issue: @@ -599,9 +599,14 @@ static HRESULT WINAPI PrimaryBufferImpl_SetVolume( @@ -640,9 +645,14 @@ static HRESULT WINAPI PrimaryBufferImpl_GetVolume( @@ -924,9 +934,14 @@ static HRESULT WINAPI PrimaryBufferImpl_SetPan( @@ -965,9 +980,14 @@ static HRESULT WINAPI PrimaryBufferImpl_GetPan( Your patch fixes only two of them: @@ -640,7 +641,12 @@ static HRESULT WINAPI PrimaryBufferImpl_GetVolume( @@ -965,7 +973,11 @@ static HRESULT WINAPI PrimaryBufferImpl_GetPan( 2) About this bug. This behaviour was introduced by this patch: Author: Francois Gouget <fgouget(a)codeweavers.com> Date: 23.07.2004 23:06:31 Parent: Sort the winmm test formats by increasing sampling rate. Child: Added Italian strings. Branch: master (ntdll: Add stubbed implementations for NtQueryEaFile and...) origin (ntdll: Add stubbed implementations for NtQueryEaFile and...) Follows: wine-20040716 (Release 20040716.) Precedes: wine-20040813 (Release 20040813.) Fix the volume and balance of primary buffers so it is the system's volume and balance, like on Windows. In that time both alsa/oss drivers asked dsound to open mmdevice by setting DSDDESC_DOMMSYSTEMOPEN flag, so it was correct. But three years later, alsa driver stopped to use mmsystem: Author: Maarten Lankhorst <m.b.lankhorst(a)gmail.com> Date: 02.06.2007 19:03:24 Parent: rsaenh: Allow CRYPT_NEWKEYSET and CRYPT_VERIFYCONTEXT at... Child: mshtml: Use the implementation of IHlink in hlink.dll ins... Branch: master (ntdll: Add stubbed implementations for NtQueryEaFile and...) origin (ntdll: Add stubbed implementations for NtQueryEaFile and...) Follows: wine-0.9.38 (Release 0.9.38.) Precedes: wine-0.9.39 (Release 0.9.39.) winealsa: Detach directsound implementation from waveout implementation. As far as I understand, the most correct way to fix this issue is to force alsa (and others) drivers to use DSDDESC_DOMMSYSTEMOPEN flag. Reporting error to high-level (as you and Alexandre suggest) or using default values (as I tried) are just hiding the bug instead of fixing it. I tried this last way, but alsa began to complain: "Why are we called with DirectSound flag? It doesn't use MMSYSTEM any more" Thus I stopped fighting and switched to bug-free OSS ;-) -- Kirill

1 0

Getting Coverity access
by Juan Lang 28 Sep '08

28 Sep '08

As we're talking about this at WineConf and not everyone has an account, you can get one by emailing scan-admin(a)coverity.com and asking. There are fewer than 100 bugs that we haven't looked at it, and I'd appreciate the help! --Juan

2 1

Re: [PATCH 1/2] gdiplus: Allocate enough space for BITMAPINFO in GdipCreateTextureIA.
by Nikolay Sivov 28 Sep '08

28 Sep '08

Lei Zhang wrote: > @@ -512,8 +512,11 @@ GpStatus WINGDIPAPI GdipCreateTextureIA(GpImage *image, > IPicture_get_CurDC(image->picture, &hdc); > bm_is_selected = (hdc != 0); > > - bmi.bmiHeader.biSize = sizeof(BITMAPINFOHEADER); > - bmi.bmiHeader.biBitCount = 0; > + pbmi = GdipAlloc(sizeof(BITMAPINFOHEADER) + 256 * sizeof(RGBQUAD)); Why 256?

3 2

comctl32:listview crash on Win98
by Francois Gouget 28 Sep '08

28 Sep '08

Listview has a test to ensure that sending a LVM_SETIMAGELIST between WM_NCCREATE and WM_CREATE works fine. The problem is that this causes the conformance test to crash on my Windows 98 OSR0 VM. Right on this line: static LRESULT CALLBACK create_test_wndproc(HWND hwnd, UINT uMsg, WPARAM wParam, LPARAM lParam) { [...] return CallWindowProc(listviewWndProc, hwnd, uMsg, wParam, lParam); } I cannot check here but the logs from Windows 95 VM also really look like it crashes there. So I suspect this happens on all comctl32.dll < 5.0. The problem is I'm not sure how to check for this so I can skip the test. Checking the Windows or comctl32 version seems likely to be rejected. Any other suggestion? -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ f u kn rd ts, ur wy 2 gky 4 ur wn gd.

1 0

Re: howto fix bug 13462
by Louis. Lenders 28 Sep '08

28 Sep '08

Rob Shearman <robertshearman <at> gmail.com> writes: > > 2008/9/15 Louis Lenders <xerox_xerox2000 <at> yahoo.co.uk>: > > Rob Shearman <robertshearman <at> gmail.com> writes: > I suggest you write a test case that replicates what the > application does and either send it to wine-patches > with appropriate todo_wine's or attach it to the bug. > > Well, it looks like the app is only checking if the uuid of the > interface is defined using Context_QueryInterface, > and that's about it. > It doesn't really seem to try to use anything from it, as > Erich's initial hack is already enough to get the apps running as > well (this one:http://bugs.winehq.org/attachment.cgi?id=13379) > > That's fine. You can still add a test that does the same > > as what the app does. > Hi, as there isn't really much activity in the bug (understatement haha) i tried to hack up a test myself. Rob, is the test below more or less what you meant? or am i way off... (The test passes on winxp, and fails in wine, so afaics shows we should support the interface ) diff --git a/dlls/ole32/tests/compobj.c b/dlls/ole32/tests/compobj.c index bdc34d9...49c3276 100644 --- a/dlls/ole32/tests/compobj.c +++ b/dlls/ole32/tests/compobj.c @@ -28,6 +28,7 @@ #include "objbase.h" #include "shlguid..h" #include "urlmon.h" /* for CLSID_FileProtocol */ +//#include "ctxtcall.h" #include "wine/test.h" @@ -61,6 +62,15 @@ static const IID IID_IWineTest = 0x4fd0, {0xa1, 0xa2, 0x5d, 0x5a, 0x36, 0x54, 0xd3, 0xbd} }; /* 5201163f-8164-4fd0-a1a2-5d5a3654d3bd */ + +static const IID IID_IContextCallback = +{ + 0x000001da , + 0x0000, + 0x0000, + {0xc0, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x46 } +}; + static const CLSID CLSID_WineOOPTest = { 0x5201163f, 0x8164, @@ -1022,6 +1032,26 @@ static void test_CoGetObjectContext(void) CoUninitialize(); } +static void test_ContextCallBack(void) +{ + + HRESULT hr; + IComThreadingInfo *pComThreadingInfo; + + if (!pCoGetObjectContext) + { + skip("CoGetObjectContext not present\n"); + return; + } + + pCoInitializeEx(NULL, COINIT_APARTMENTTHREADED); + + hr = pCoGetObjectContext(&IID_IContextCallback, (void **)&pComThreadingInfo); + + ok_ole_success(hr, "CoGetObjectContext"); + CoUninitialize(); +} + START_TEST(compobj) { HMODULE hOle32 = GetModuleHandle("ole32"); @@ -1050,4 +1080,5 @@ START_TEST(compobj) test_registered_object_thread_affinity(); test_CoFreeUnusedLibraries(); test_CoGetObjectContext(); + test_ContextCallBack(); }

4 4

Re: user32/tests: Move shell test to the end - killing explorer affects lots of tests
by Dmitry Timoshkov 28 Sep '08

28 Sep '08

Vitaliy Margolen <wine-patches(a)kievinfo.com> wrote: > This fixes all win tests except 2 on win2k. If it will affect other test(s) > following win test then we either shouldn't kill explorer or restart it or > more correctly configured shell. Probably adding a comment requesting to keep the test at the end would be a good idea. -- Dmitry.

1 0

Re: [PATCH] shell32: print the correct GUIDs in the shellpath tests.
by Juan Lang 28 Sep '08

28 Sep '08

Hi Reece, I'm not nak'ing your patch. I just wanted to point out a simpler way: -static const char *printGUID(const GUID *guid) +static char *printGUID(const GUID *guid) { - static char guidSTR[39]; + char *guidSTR = HeapAlloc(GetProcessHeap(), 0, 39); Rather than allocating the necessary buffer, you can pass a pointer to a buffer to printGUID, and have printGUID return that buffer. Have the caller declare two GUID buffers, and pass one for the expected value and the other for the received value. Cheers, and thanks for looking into the failing tests, --Juan

3 4

Patchwatcher: boldly going where MS has gone before
by Dan Kegel 28 Sep '08

28 Sep '08

Pre-commit tests were done commonly at Microsoft as early as 2001 by a system named Gauntlet, see http://blogs.msdn.com/misampso/archive/2005/03/14/395374.aspx http://blogs.msdn.com/vcblog/archive/2006/11/02/gauntlet-a-peek-into-visual… Well, better late than never!

1 0