August 2007 - Wine-Devel - list.winehq.org

Re: [5/5] WineD3D: texkill ignores the .w only in ps 1.x
by Tomas Carnecky Sept. 1, 2007

Sept. 1, 2007

+ + hr = IDirect3DDevice9_Clear(device, 0, NULL, D3DCLEAR_TARGET, 0xff00ff00, 0.0, 0); + ok(hr == D3D_OK, "IDirect3DDevice9_CreatePixelShader returned %s\n", DXGetErrorString9(hr)); Is that worth changing? tom

2 1

Re: [4/5] WineD3D: Implement texldb
by H. Verbeet Aug. 31, 2007

Aug. 31, 2007

On 31/08/2007, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > Not as much a comment on this patch, but rather something I just thought of and I'll otherwise probably forget: our d3d9types.h doesn't contain definitions for D3DSI_TEXLD_PROJECT and D3DSI_TEXLD_BIAS, so we should add those.

1 0

Re: [2/5] WineD3D: Fix the z range
by Stefan Dösinger Aug. 31, 2007

Aug. 31, 2007

Am Freitag, 31. August 2007 17:20 schrieb Stefan Dösinger: Do not apply this patch yet, it breaks half life 2. There's something about the minZ and maxZ values in the viewport that we have to take into account.

1 1

Re: [3/5] WineD3D: Fix texdepth instruction
by Alexandre Julliard Aug. 31, 2007

Aug. 31, 2007

Stefan Dösinger <stefan(a)codeweavers.com> writes: > The DCT pixel shader test spotted a problem with texdepth. The test > included in this patch demonstrates texdepth's undocumented clamping > behavior, and > verifies the documented r5.r == 0.0 special case. This doesn't work for me: ../../../tools/runtest -q -P wine -M d3d9.dll -T ../../.. -p d3d9_test.exe.so visual.c && touch visual.ok visual.c:2281: Test failed: Pixel 158(25% - 2 pixel) has color 00ffffff, expected 0x000000ff visual.c:2303: Test failed: Pixel 318(50% - 2 pixel) has color 00ffff00, expected 0x000000ff visual.c:2344: Test failed: Pixel 318(50% - 2 pixel) has color 0000ff00, expected 0x000000ff make: *** [visual.ok] Error 3 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: winex11.drv[1/4]: wintab: restructure WTInfoA to dump a category by dumping field by field (resend)
by Alexandre Julliard Aug. 31, 2007

Aug. 31, 2007

"Mikolaj Zalewski" <mikolajz(a)google.com> writes: > I see that my patch to skip non-pressure-sensitive devices didn't > get it. I will try to look into it next week while now I'll resend the > Unicode patches that are easier. Here the first patch could be > controversial as I use a lot of macros. But that allows me not to use > long parameter lists. It's ugly. I don't see why you'd need long parameter lists at all, simply adding a current offset in CopyTabletData should be enough. Also, always going through all the cases even when a specific index is specified is not nice. And the winex11 function should always return Unicode, the A translation can be done in the generic code. -- Alexandre Julliard julliard(a)winehq.org

1 0

Vacation
by Alexandre Julliard Aug. 31, 2007

Aug. 31, 2007

Hi folks, I'll be on vacation all of next week, so there won't be any commits. If you need to synchronize development with others in the meantime, I encourage you to create a tree on repo.or.cz. The 0.9.45 release will also be delayed by a week, so it will be on Sept 14th. Have fun while I'm gone... -- Alexandre Julliard julliard(a)winehq.org

1 0

RegDeleteTree and older versions of windows?
by Maarten Lankhorst Aug. 31, 2007

Aug. 31, 2007

Now that most functions get replaced with RegDeleteTree, I'm wondering if it is possible to link against that function statically for mingw/msvc dlls, so wine dlls can be still be used on win9x and windows nt and win2k. The problem is I don't know what the best solution would be to make sure the function makes it in, in case of mingw. I think it would be best to add it to include/wine/regdeltree.h and then include it from all dlls that need it, and probably from dlls/advapi32/registry.c as well, so code duplication is avoided. Something along the lines of: #if defined(__MINGW32__) || defined(__MSVC__) || defined(EXPORTREGDELTREE) (insert regdeltree functions here) #endif I'm open for suggestions though, since I really can't come up with any good ways to make sure wine dlls will keep working for older versions of windows. Cheers, Maarten

2 1

Re: winex11.drv: fix memory leak in create_cursor (found by Smatch)
by James Hawkins Aug. 31, 2007

Aug. 31, 2007

On 8/31/07, Lionel_Debroux <lionel_debroux(a)yahoo.fr> wrote: > create_cursor leaks some heap memory in an error path. Found in > Michael Stefaniuc's list of Wine potential bugs detected by Smatch. > > I'm almost sure that the other potential memory leak detected by Smatch, > in the same function, is NOTABUG. Indeed, bitMask32 is conditionally > freed at line 850, if the number of bits per pixel (ptr->bBitsPerPixel) > is 32 - which is the condition which led to bitMask32 being allocated, > at line 639. > > 2007-08-31 Lionel Debroux <lionel_debroux(a)yahoo.fr> > * dlls/winex11.drv/mouse.c: > winex11.drv: Fix memory leak in create_cursor (found by Smatch) > > >From d62a874a6d399546940a97fc3d800a692c08b362 Mon Sep 17 00:00:00 2001 > From: Lionel Debroux <lionel_debroux(a)yahoo.fr> > Date: Fri, 31 Aug 2007 08:16:36 +0200 > Subject: winex11.drv: fix memory leak in create_cursor (found by Smatch). > > --- > dlls/winex11.drv/mouse.c | 2 ++ > 1 files changed, 2 insertions(+), 0 deletions(-) > > diff --git a/dlls/winex11.drv/mouse.c b/dlls/winex11.drv/mouse.c > index 86dad89..d8bf30a 100644 > --- a/dlls/winex11.drv/mouse.c > +++ b/dlls/winex11.drv/mouse.c > @@ -763,6 +763,8 @@ static Cursor create_cursor( Display *display, CURSORICONINFO *ptr ) > pixmapBits = XCreateBitmapFromData( display, root_window, (char *)pXorBits, xmax, ymax ); > if (!pixmapBits) > { > + if (bitMask32) > + HeapFree( GetProcessHeap(), 0, bitMask32 ); > XFreePixmap( display, pixmapAll ); > You shouldn't be checking bitMask32 for NULL before freeing it. -- James Hawkins

1 0

Re: crypt32(1/17): Greatly expand certificate chain tests
by Alexandre Julliard Aug. 31, 2007

Aug. 31, 2007

"Juan Lang" <juan.lang(a)gmail.com> writes: > In case the big version doesn't get through in time, here's a gzipped version. It fails for me: ../../../tools/runtest -q -P wine -M crypt32.dll -T ../../.. -p crypt32_test.exe.so chain.c && touch chain.ok chain.c:1187: Test failed: Chain 12: CertGetCertificateChain failed: 80092004 make[2]: *** [chain.ok] Error 1 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [1/3] WineD3D: Fix texdp3tex in glsl
by H. Verbeet Aug. 31, 2007

Aug. 31, 2007

On 30/08/2007, Stefan Dösinger <stefandoesinger(a)gmx.at> wrote: > The preferred way to do that is to use shader_glsl_get_sample_function(). Also note that previously the docs just said "The result is used to sample the texture at texture stage m by performing a 1D lookup." The addition of using 0.0 to extend the vector is apparently recent.

2 2