July 2015 - Wine-Devel - list.winehq.org

Re: [6/12] ntdll/tests: Add tests for threadpool timer merging.
by Sebastian Lackner July 3, 2015

July 3, 2015

On 03.07.2015 01:59, Sebastian Lackner wrote: > --- > dlls/ntdll/tests/threadpool.c | 117 ++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 117 insertions(+) > The test failure on Windows 7 seems to be random, it worked fine on a second run (and also on several test runs before): https://newtestbot.winehq.org/JobDetails.pl?Key=14968 If it occurs too often in practice we can increase the timeouts, but for now I think it should be fine. BTW: Forgot to mention, the Sleeps() in this patch are necessary because there is no way to do any other kind of synchronization with the Windows internal timer queue thread.

1 0

Re: [6/12] ntdll/tests: Add tests for threadpool timer merging.
by Marvin July 3, 2015

July 3, 2015

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 https://testbot.winehq.org/JobDetails.pl?Key=14965 Your paranoid android. === w7u (32 bit threadpool) === threadpool.c:856: Test failed: WaitForSingleObject returned 258 threadpool.c:897: Test failed: expected that ticks are nonzero threadpool.c:899: Test failed: expected that timers are merged

1 0

Re: [PATCH 2/4] wined3d: Implement vertex blending in glsl_vertex_pipe.
by Stefan Dösinger July 2, 2015

July 2, 2015

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I like this patchset. I can't find anything wrong in it, but Matteo knows the fixed function pipeline replacement better than I do, so I'll wait for his comments. I tested it with various DirectX SDK samples (bend from dx7, SkinnedMesh and VertexBlend from dx8). It works very well, but I noticed a lighting oddity in SkinnedMesh. The lighting differs slightly between your code ("D3D non-indexed" option) and the test program's vertex shader or d3dx9's software vertex blending ("D3D indexed (VS)" / "Software Skinning"). This difference cannot be seen on Windows. I have attached screenshots. Cheers, Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJVlPPNAAoJEN0/YqbEcdMwUcIP/jwJsie6/93Ui3RQTqjT8oOP q0ips3E83dHpi0ceL8xXujr0pyO1x7cVuUX+hZ7Tf08bZu86FpieKZkTYZxRKbKj V5+x/bVlWGM+ogB06E7tHx2oM8tBFlh10wwh9M/QnitjtzCM+E3WKfE01W5NsRRL csUMxE1mBHvnp08Me2CZyzgtVWwQuTWH6IMUtmqQPfOmmOgzIoofjjRowjhZovFi D9cufKD/D+CBsuv/Py5AMYWfjv7gTZqB09sotaIzYWuyI/e6GRRFT2wzX1yOIl+I fhOr3AsHeFLWOAYd41DjMuqmqvQ5e/UigYxkDtssOCHVig7VtaQ0RKTEv6mMfDF4 6Ljgyi7RdGrgle41cDo6DBXTr0WZw6Y2eTLs0G+WCNkJ8/OMR9hLAt/YBg1mPEPo QUNTsqwNQ1Poxuxmo5fWqD0MhLAao7Wgm6dZTiECNPNNBsC91XRYR9jXF29xaB3h 3ekVCTdtI1+wUQolGw1/HZ7whaHkWGrxOY4qqiZsgYRYz1yl52Bt4sJDJYUHknbd GZ4JRkVQXD86D5txRLf500e03QCmDTXxjqEKSsNBNCtcpER1mp+3Hs/SImp/adRr 3tNuEMJduA/5jm27Nt+K13DwsazQXCJMDv5dTYlggBI9c3dhFDmjXNNGSvrvkBkf xW9FbSiHRRDoLR3UMuoR =Ghq/ -----END PGP SIGNATURE-----

2 1

Re: [3/3] msvcirt: Implement ios constructors and assignment
by Marvin July 2, 2015

July 2, 2015

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 https://testbot.winehq.org/JobDetails.pl?Key=14949 Your paranoid android. === wvistau64 (64 bit msvcirt) === No test summary line found === w2008s64 (64 bit msvcirt) === No test summary line found

1 0

Re: [2/3] msvcirt/tests: Add initial tests for ios
by Marvin July 2, 2015

July 2, 2015

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 https://testbot.winehq.org/JobDetails.pl?Key=14948 Your paranoid android. === wvistau64 (64 bit msvcirt) === No test summary line found === w2008s64 (64 bit msvcirt) === No test summary line found

1 0

A wrapper around the GTK+ rendering engine
by Ivan Akulinchev July 2, 2015

July 2, 2015

Hello, I just want to thank you for Wine and show you my own uxtheme.dll implementation, which uses the GTK+ rendering engine. Thus Wine applications look really native for users. Here is the link: https://github.com/akulinchev/uxthemegtk (see the screenshots at the bottom of the page) If you want, I can help you to add it to the Wine code. If not, I will develop it alone as a third-party project, but then I have a question: what is the best way to replace a Wine library? I think, I shouldn't change anything in /usr/lib/wine. I know about "native libs", but it's an ELF binary, not PE. Regards, Ivan

5 8

Re: [PATCH 4/4] d3d9/tests: Add basic tests for vertex blending.
by Henri Verbeet July 2, 2015

July 2, 2015

On 2 July 2015 at 08:26, Józef Kucia <joseph.kucia(a)gmail.com> wrote: > + static const float quad_upper_right[] = > + { > + -1.0f, -1.0f, 0.0f, 0.0f, 0.0f, 0.0f, > + -1.0f, 1.0f, 0.0f, 0.0f, 0.0f, 0.0f, > + 1.0f, -1.0f, 0.0f, 0.0f, 0.0f, 0.0f, > + 1.0f, 1.0f, 0.0f, 0.0f, 0.0f, 0.0f, > + }, I think it's slightly more readable to specify these as e.g. static const struct { struct vec3 position; struct vec3 blendweights; } quad_upper_right[] = { ... }, instead of an otherwise unstructured collection of floats. You can also potentially fold these into the tests[] array. > + static const struct { > + const float *vertex_data; > + const POINT *quad_points; > + const POINT *empty_points; > + } > + tests[] = Position of the upper brace.

1 0

Re: [PATCH 2/4] wined3d: Implement vertex blending in glsl_vertex_pipe.
by Henri Verbeet July 2, 2015

July 2, 2015

On 2 July 2015 at 08:26, Józef Kucia <joseph.kucia(a)gmail.com> wrote: > @@ -3071,7 +3075,7 @@ static inline BOOL shader_constant_is_local(const struct wined3d_shader *shader, > > void get_identity_matrix(struct wined3d_matrix *mat) DECLSPEC_HIDDEN; > void get_modelview_matrix(const struct wined3d_context *context, const struct wined3d_state *state, > - struct wined3d_matrix *mat) DECLSPEC_HIDDEN; > + int i, struct wined3d_matrix *mat) DECLSPEC_HIDDEN; I think we'll want the index to be unsigned. (Also applies to shader_glsl_ffp_vertex_normalmatrix_uniform() in patch 112562) Looks pretty good otherwise.

1 0

Re: [PATCH 1/2] comctl32: Set the default separator width when notifying a -ve cmd id.
by Marvin July 2, 2015

July 2, 2015

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 https://testbot.winehq.org/JobDetails.pl?Key=14914 Your paranoid android. === wvistau64 (64 bit toolbar) === toolbar.c:519: Test failed: Images count mismatch - 26 vs 10 toolbar.c:519: Test failed: Icon size mismatch - 8x8 vs 20x20 toolbar.c:523: Test failed: Images count mismatch - 28 vs 22 toolbar.c:523: Test failed: Icon size mismatch - 16x16 vs 20x20 toolbar.c:533: Test failed: Icon size mismatch - 8x8 vs 16x16 toolbar.c:536: Test failed: Icon size mismatch - 30x30 vs 16x16 toolbar.c:544: Test failed: Images count mismatch - 8 vs 13 toolbar.c:544: Test failed: Icon size mismatch - 30x30 vs 16x16 toolbar.c:548: Test failed: Images count mismatch - 208 vs 13 toolbar.c:548: Test failed: Icon size mismatch - 1x1 vs 16x16 toolbar.c:551: Test failed: Images count mismatch - 208 vs 13 toolbar.c:551: Test failed: Icon size mismatch - 1x5 vs 16x16 toolbar.c:554: Test failed: Images count mismatch - 41 vs 13 toolbar.c:554: Test failed: Icon size mismatch - 5x1 vs 16x16 toolbar.c:1331: Test failed: Icon size mismatch - 14x12 vs 16x15

2 1

Re: [PATCH 2/2] comctl32: Implement the legacy notifications during restore.
by Marvin July 1, 2015

July 1, 2015

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 https://testbot.winehq.org/JobDetails.pl?Key=14915 Your paranoid android. === wvistau64 (64 bit toolbar) === toolbar.c:584: Test failed: Images count mismatch - 26 vs 10 toolbar.c:584: Test failed: Icon size mismatch - 8x8 vs 20x20 toolbar.c:588: Test failed: Images count mismatch - 28 vs 22 toolbar.c:588: Test failed: Icon size mismatch - 16x16 vs 20x20 toolbar.c:598: Test failed: Icon size mismatch - 8x8 vs 16x16 toolbar.c:601: Test failed: Icon size mismatch - 30x30 vs 16x16 toolbar.c:609: Test failed: Images count mismatch - 8 vs 13 toolbar.c:609: Test failed: Icon size mismatch - 30x30 vs 16x16 toolbar.c:613: Test failed: Images count mismatch - 208 vs 13 toolbar.c:613: Test failed: Icon size mismatch - 1x1 vs 16x16 toolbar.c:616: Test failed: Images count mismatch - 208 vs 13 toolbar.c:616: Test failed: Icon size mismatch - 1x5 vs 16x16 toolbar.c:619: Test failed: Images count mismatch - 41 vs 13 toolbar.c:619: Test failed: Icon size mismatch - 5x1 vs 16x16 toolbar.c:1396: Test failed: Icon size mismatch - 14x12 vs 16x15

1 0