May 2013 - Wine-devel - list.winehq.org

Re: Christian Costa : d3dx9_36: Implement skin mesh loading.
by gurketsky 29 May '13

29 May '13

Hi, just saw that a bit late. Is there a reason for the rewrite of hr? Could any call return anything other than D3D_OK and succeed (maybe S_FALSE)? It may be possible to drop "if (FAILED(hr)) goto end;" (for both cases) and "hr = D3D_OK"... Also the function (parse_mesh) seems to overwrite the result anyway to D3D_OK on success, so I see no reason to change the hr to D3D_OK in parse_skin_mesh_info. Cheers Rico On 28.05.2013 22:16, Alexandre Julliard wrote: > + if (SUCCEEDED(hr)) > + hr = mesh_data->skin_info->lpVtbl->SetBoneOffsetMatrix(mesh_data->skin_info, index, > + (const D3DMATRIX*)(data + nb_influences * (sizeof(DWORD) + sizeof(FLOAT)))); > + if (FAILED(hr)) > + goto end; > + } > + > + hr = D3D_OK;

2 1

Re: Increment, Decrement, and Assignment Operators
by Alexandre Julliard 29 May '13

29 May '13

Stephen Leibowitz <librestephen(a)gmail.com> writes: > Hello: > > This is my first patch. The changes are minor. If it works out, I may > attempt something more substantial. Doing cosmetic changes like this across the code base is not a good idea. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] gdi32: Ask for a newer FreeType version
by Alexandre Julliard 29 May '13

29 May '13

Detlef Riekenberg <wine.dev(a)web.de> writes: > @@ -3686,7 +3686,8 @@ static BOOL init_freetype(void) > WINE_MESSAGE( > "Wine cannot find the FreeType font library. To enable Wine to\n" > "use TrueType fonts please install a version of FreeType greater than\n" > - "or equal to 2.0.5.\n" > + "or equal to 2.1.4.\n" > + "(Fonts with vertical metrics need at least version 2.4 of FreeType)\n" > "http://www.freetype.org\n"); > return FALSE; > } > @@ -3749,6 +3750,7 @@ sym_not_found: > "Wine cannot find certain functions that it needs inside the FreeType\n" > "font library. To enable Wine to use TrueType fonts please upgrade\n" > "FreeType to at least version 2.1.4.\n" > + "(Fonts with vertical metrics need at least version 2.4 of FreeType)\n" That's not useful. There are probably very few users who still have such an old Freetype, and even if they do, telling them about vertical metrics will only confuse them. -- Alexandre Julliard julliard(a)winehq.org

1 0

re: Increment, Decrement, and Assignment Operators
by Dan Kegel 28 May '13

28 May '13

Watch http://source.winehq.org/patches/ for your patch's status (see legend at bottom). It might be rejected since it doesn't actually fix any behavioral problem in Wine. IIRC trivial cleanups in code are frowned upon, especially if they cross many modules. Better to pick some small real problem and fix it. How about http://bugs.winehq.org/show_bug.cgi?id=32922 ?

1 0

Re: gdi32/tests: Do not try to test OS/2 win metrics on non-windows fonts
by Dmitry Timoshkov 28 May '13

28 May '13

Aric Stewart <aric(a)codeweavers.com> wrote: > + ascent = GET_BE_WORD(tt_os2.usWinAscent); > + descent = GET_BE_WORD(tt_os2.usWinDescent); > + cell_height = ascent + descent; > + ok(ntm->ntmCellHeight == cell_height, "%s: ntmCellHeight %u != %u, os2.usWinAscent/os2.usWinDescent %u/%u\n", > + font_name, ntm->ntmCellHeight, cell_height, ascent, descent); > + > + SetLastError(0xdeadbeef); The SetLastError() call should be left at its previous position. -- Dmitry.

1 0

Re: [PATCH] winhttp: initialize schemes (Coverity) (resubmit)
by Hans Leidekker 28 May '13

28 May '13

On Tue, 2013-05-28 at 22:46 +0200, Marcus Meissner wrote: > 1020943 Uninitialized scalar variable > > schemes is |= in the called function, but not > initialized before. > > Was submitted 2 weeks ago already, I think it is good. It's good but query_auth_schemes is called from more places, so it would be better to fix it in that helper, without touching the parameter on error.

1 0

Re: [PATCH] dsound: Copy SetFormat format more directly if we're not in PRIMARYWRITE mode
by Marvin 28 May '13

28 May '13

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=25741 Your paranoid android. === WXPPROSP2 (32 bit dsound) === dsound.c:1433: Test failed: validbits: 0 === WXPX64 (32 bit dsound) === dsound.c:1433: Test failed: validbits: 0 === W2K3R2SESP2 (32 bit dsound) === dsound.c:1433: Test failed: validbits: 0 === WXPX64 (64 bit dsound) === dsound.c:1433: Test failed: validbits: 0

1 0

Re: [5/5] kernel32: Don't move fixed memory area in GlobalReAlloc. (try 2)
by Alexandre Julliard 28 May '13

28 May '13

Akihiro Sagawa <sagawa.aki(a)gmail.com> writes: > This closes bug 14819. > > try 2: > * tests/heap.c was updated to avoid dependency on heap reallocation > strategies. > * Use HeapAlloc instead of LocalAlloc in some places. In those cases, > they want large memory blocks, not pinned memory addresses. They are using LMEM_MOVEABLE though. A few more tests for this flag would probably be a good idea. -- Alexandre Julliard julliard(a)winehq.org

1 0

dxgi, directx 11 and code duplication
by John Yani 28 May '13

28 May '13

Hi. I'm trying to set up a minimal implementation that will allow to run Assasin's Creed III. It uses these 2 functions: dxgi.dll.CreateDXGIFactory1 d3d11.dll.D3D11CreateDevice See http://bugs.winehq.org/show_bug.cgi?id=24517 I need an advice on how to implement CreateDXGIFactory1 without duplicating a lot of code. There are problems with a straight-forward implementation: interface IWineDXGIFactory : IDXGIFactory { struct wined3d *get_wined3d(); } interface IWineDXGIFactory1 : IDXGIFactory1 { struct wined3d *get_wined3d(); } The problem is how to reuse functions where IWineDXGIFactory is used. Obviously we can't pass IWineDXGIFactory1 object where IWineDXGIFactory is required. Because the next time we call get_wined3d, a crash will occur (a different function will be called). Is there a more elegant solution than duplicating code and replacing all occurencdes of IWineDXGIFactory by IWineDXGIFactory1? Btw, D3D11CreateDevice accepts both IDXGIFactory and IDXGIFactory1: http://msdn.microsoft.com/en-us/library/windows/desktop/ff476082(v=vs.85).a…

2 1

dxgi, directx 11 and code duplication
by John Yani 27 May '13

27 May '13

Hi. I'm trying to set up a minimal implementation that will allow to run Assasin's Creed III. It uses these 2 functions: dxgi.dll.CreateDXGIFactory1 d3d11.dll.D3D11CreateDevice See http://bugs.winehq.org/show_bug.cgi?id=24517 I need an advice on how to implement CreateDXGIFactory1 without duplicating a lot of code. There are problems with a straight-forward implementation: interface IWineDXGIFactory : IDXGIFactory { struct wined3d *get_wined3d(); } interface IWineDXGIFactory1 : IDXGIFactory1 { struct wined3d *get_wined3d(); } The problem is how to reuse functions where IWineDXGIFactory is used. Obviously we can't pass IWineDXGIFactory1 object where IWineDXGIFactory is required. Because the next time we call get_wined3d, a crash will occur (a different function will be called). Is there a more elegant solution than duplicating code and replacing all occurencdes of IWineDXGIFactory by IWineDXGIFactory1? Btw, D3D11CreateDevice accepts both IDXGIFactory and IDXGIFactory1: http://msdn.microsoft.com/en-us/library/windows/desktop/ff476082(v=vs.85).a…

1 0