May 2016 - Wine-devel - list.winehq.org

Re: [PATCH 6/7] d3drm: Partially implement IDirect3DRMTexture*::InitFromFile. (v3)
by Stefan Dösinger 16 May '16

16 May '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Am 2016-05-13 um 22:21 schrieb Aaryaman Vasishta: > static HRESULT WINAPI d3drm_texture1_InitFromFile(IDirect3DRMTexture *iface, const char *filename) > { > - FIXME("iface %p, filename %s stub!\n", iface, debugstr_a(filename)); > + struct d3drm_texture *texture = impl_from_IDirect3DRMTexture(iface); > + D3DRMIMAGE *image; > + HRESULT hr; > > - return E_NOTIMPL; > + TRACE("iface %p, filename %s.\n", iface, debugstr_a(filename)); > + > + if (texture->image) > + { > + /* d3drm intentionally leaks a reference IDirect3DRM if texture is already initialized. */ > + IDirect3DRM_AddRef(texture->d3drm); > + return D3DRMERR_BADOBJECT; > + } > + if (FAILED(hr = d3drm_texture_load(texture, filename, FALSE, &image))) > + return hr; > + > + hr = IDirect3DRMTexture3_InitFromImage(&texture->IDirect3DRMTexture3_iface, image); I think you forgot to remove the extra check here :-) . (https://www.winehq.org/pipermail/wine-devel/2016-May/113212.html) > - IDirect3DRM_AddRef(texture->d3drm); > + if (texture->d3drm) > + IDirect3DRM_AddRef(texture->d3drm); This if check looks like it is an accidental leftover from before you reordered patch 1 too. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJXOiCxAAoJEN0/YqbEcdMwpSkP/ifqzhJnNWI0ZRqMQa9/rOUg MgbEJEGLC2xFFWwGWHL3biuf+ibZOkmmAPBUznU9+E70ekzegvvRGquc2HTHP0y+ 43PXR5A9TXFoOEDvkIiYF6KGiTyEoC2K3N3KzMabm5Iz8ReDLqvF4MmwlggTBsWQ RS1hbDgP4KKSGRECJFKmI3F0WJk52T2TtPjVXmHrYLYgSKQ8QmCj0BSCGBLFlqWF uQtH6sM/dI4w6LiqjG83JAlt7VmKZ5hrWJIY7IakF7vZLIOwni4x9+CJyh9b6dc6 gAY27wojqiBtk7x8weQXG2o3kToCL0+dpHTcLNKuJy80tPZa5QVTmXNF5B+Eqtjf 7u3ZWOOEKGdVKKDaztnoEtxlXHfZDn4ST+2pdUhbOW9X5gcxxSQWyour9FnD/fGj i/OkgvM4gszu1RwcZlpbUY8CkyboOJyziVSUxt7cnW2gHDXY0FwhQMvfpHoPSxiq +UmsuTFwopoJzK30kh4kwmJSkBGtackb2nA0i34tb69ixIAfQy55cqYTgozEzclh +2T/wa5q2DhOlHBScjPlUu46j3AqI8B0mtfrLuURB2UWzxKOZpZYtVsTLErVSFqT 745Z+bPbYWGETBJYx+BVxFE4bpGpwqJTI5IH/jITDv4qmNxoLgHKeYnxU9dwayhx PcvR3vpmTKw/C2N84u+n =3f/e -----END PGP SIGNATURE-----

2 1

Re: [PATCH 4/7] d3drm: Implement IDirect3DRM*_CreateTexture. (v6)
by Stefan Dösinger 16 May '16

16 May '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Am 2016-05-13 um 22:21 schrieb Aaryaman Vasishta: > - *texture = &object->IDirect3DRMTexture3_iface; > + if (FAILED(IDirect3DRMTexture3_InitFromImage(*texture, image))) > + { > + IDirect3DRMTexture3_Release(*texture); > + return D3DRMERR_BADVALUE; > + } This will leave a non-NULL pointer stored in *texture, which is most likely wrong. Please extend your tests to check for this. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJXOgmWAAoJEN0/YqbEcdMw4PQP/1Ri28V/xXTb90aSLvHLi3QQ 47cGk1M2C9/z3bj9kZUKfYGADCJVgdf52bcFQUb4yBBEkxtY2RrAYMAY2218l0PB QcNpkFhVEnauRII4CsoLShByEtKLfLo6sqflABUKoBfK9BELjz4QpDEUiMvyKHXk vRtesZBK16VaFnwbpQekFb233/sID9oCLitNzIXJ6ntNLEvGYCgfWmaT9VdXNGvv zvnQfWh8ayXbJGulUclNectDqA2ww+vRxHmC5dkgAVGHsfxOp1dY3Z45b6CNPksE iyxn6A3mAVgy6hhjbm8AjgN/8agkvwDRhL2774ZdemBCL1FAq39kFobSNju4evAx YH6r8bFiN1y3814hxMe74Zk2oXJsCN4ZL5ZxOOabJHdWHCxTDWU8yikordtlfdF2 q1Va/GVi9aKLXiWhfNjusP5bG/DY8hK14dRe570eWP2U9fKbDaFLc8iVhwxZRs0N 3EJddB286uvLjAGEbGEe5iSPTuTKCER0L/8rRTliczYt5rJeX+2955bTZu3ek5Uu XjmOzBP5374IiBDai1dhHX2EwTD3UxisN2olDUfFGee8DkDjzowKt5hFBYVT7x8j hDU7rkRKht2loO7vKEqX+uM9twplIWVIIcQWY2ZXT7gZ8x07bCqVa8eHhKy63cGp J1YMc8wPUUENq+iVYZ1W =sm0l -----END PGP SIGNATURE-----

2 1

Re: [PATCH 1/7] d3drm: Use CreateObject in IDirect3DRM*::LoadTexture.
by Stefan Dösinger 16 May '16

16 May '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Am 2016-05-13 um 22:21 schrieb Aaryaman Vasishta: > - if (FAILED(hr = d3drm_texture_create(&object))) > + if (FAILED(hr = IDirect3DRM_CreateObject(iface, &CLSID_CDirect3DRMTexture, NULL, &IID_IDirect3DRMTexture, (void **)texture))) This line is fairly long and could profit from a line break IMHO. > - if (FAILED(hr = d3drm_texture_create(&object))) > + if (FAILED(hr = IDirect3DRM2_CreateObject(iface, &CLSID_CDirect3DRMTexture, NULL, &IID_IDirect3DRMTexture2, (void **)texture))) I think here you could already put the 2 -> 3 thunk from patch 7. It's a minor issue though. Overall I'd say wait if Henri has any other comments before resending. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJXOh4hAAoJEN0/YqbEcdMwjeMP+wRn3MLnrgMlB85ZfafutK3g wml22BVCwzHvHjbPKFhqbrPUIjS4z/PcZ133djjS6dysL2sBfMK4zxKKqVDRp+0V RRphWDVQehg6dpfaSBEWcEqLvy5eUXbQlGQ9eWGT40++OVZQ/C5rmIDmPt1/pX/H f2vnjCIKhGaiDS7M29fnakQckr3sjGuQAJXJuAJqyUHyPo4Edu340UJQweLCOIXX kDXTAJM0ep17xln5FD8RDjukRc1SRg3P+/PzM1PsED0St/talBFdN16DnfNwzOWO XbE+wlTQnMxeu2ycY5gyTdFxWUq/lqCYL6Z0KYTwBUHIqryz0K2XPnb1fGZUTuKk B+XsxtOJYkcvJEoStUuEQSwsTEGipuOStj+fSFEoYnOHBqZWl6midQGWrU6Xyd39 TlOFMew6IldjQBgbfEbu9dSGr3xPi4jD5dPfThdTZFTMOmD++K2APKGmhVmc5bon QOJ+M6uJJQvnyCqnik63uNQcX5UlBf5Guq1op2K6Zbm3V1neipVBQj8+8mUr7weV HQDccTIyzErBJuEdSDoCuw4womh+UGcFbusWpRjP5lRaa57X9poZO58Ofw61f/HZ Tqm9bbnuVjwk4hLa8z5wVbd2bQAytkK0ISFP9bnmGIVk8SZIwuU471pqunF6UMsF Yl/jFc8sJKIeRAe+bGq2 =LMsi -----END PGP SIGNATURE-----

1 0

Re: [PATCH] qedit: Don't require SampleGrabber to be playing to call callback
by Marvin 16 May '16

16 May '16

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=23019 Your paranoid android. === w2008s64 (32 bit mediadet) === mediadet.c:148: Test failed: CoCreateInstance failed: 80040154 mediadet: unhandled exception c0000005 at 00401A5F === w2008s64 (64 bit mediadet) === mediadet.c:148: Test failed: CoCreateInstance failed: 80040154 mediadet: unhandled exception c0000005 at 0000000000401995

2 1

Re: kernel32/tests: Fix a typo in comment
by Marvin 16 May '16

16 May '16

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=23020 Your paranoid android. === w864 (64 bit process) === process.c:1192: Test failed: expected 1, got 0 process.c:1195: Test failed: Child process termination process.c:1199: Test failed: Toolhelp:cntUsage expected 0, but got 1 process.c:1200: Test failed: Toolhelp:th32DefaultHeapID expected 0, but got 1 process.c:1201: Test failed: Toolhelp:th32ModuleID expected 0, but got 1 process.c:1202: Test failed: Toolhelp:th32ParentProcessID expected 280, but got 0 process.c:1204: Test failed: Toolhelp:dwFlags expected 0, but got 1

1 0

Rebased command stream patches
by Stefan Dösinger 15 May '16

15 May '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Hi, In the past weeks I have rebased my command stream patches on top of Henri's recent texture restructuring work. They can be found as usual on https://github.com/stefand/wine/ (git repo URL is https://github.com/stefand/wine.git) They are essentially the same state as the last ones, just rebased, and thus the surface and volume handling code is quite a bit changed. The new patches do a better, but not perfect, job at keeping the tests working during the rework. The tests pass with CSMT enabled, except for a d3d10/11 test. I will hopefully investigate this in the next days and push an updated version. Performance seems to be generally unchanged. That's a good sign I guess. I tested a number of games, but I am sure there are some regressions compared to the last CSMT version. I probably will not do another rebase of this magnitude. Keeping up with the ongoing d3d work requires a lot of time, and I'd rather spend the time I have available for hacking on Wine on smaller tasks that I have a better chance of completing and upstreaming. Though if you run into problems I am interested in getting feedback. I just don't promise that I'll reply or fix anything :-) . Cheers, Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJXOMITAAoJEN0/YqbEcdMwmfwP/j0VXX+0YnFsZCG/u7Bt5Dth LLpbYrUeWcGnS9ddHppdb9VkU0sbrqnXo6j/PA/tXMdli2CFHsC8/DFoPi/A4dKE 5kLDagUE6CPsFOxjmCXdgvdb7yqK2EToz/0sBGJEHXl2tN/B+/yPpIGUfHz5prWb KE1M4w065Wxe39opuV3EkYKk1rFT4wrLUKaH7iBl2rVN4hqvZ4rxS2wFFLFlZk9S E6CR3SDC/HhGZU4ZPpdkMWk3pYi/YDDg1yRESd7PJZNHSVRAo7wgQu4KkQM2aGnG onqmpN6Prw06MKjixSdVF4jcQcPeYiY1dYVB61IslbR9Y7FRSmrxsGnrY51sqTCi pGITgl9loktSbA+hxouMa4xhaz80v6iVvpRkjR1vqnOGZxBRJUC9U9eqRAbdzoZs KhujuEVR74DQ54U5dZbX6+Z682WbPPPqdFMXZrRt0eR3y1TfOhC38Zky3ri7X7/p BleoWUQ4Hut4LK/Td0hhxxkPvfjOGvm6Jm+JPZO4+JnBVZl13WyWat1OOSSOrM2u jcZ7IPkBqeVFu7G1j3YCMoc0hQfa5nPzoHIcbb6DSKythILKSdBZ79KnKzlprvOy pc9EtJFlmSvFi/JiBSJTJuHeUyA9mokmkYtrBsEvh3KBXU/zHtRCLI7qlDpj9yZb +GAQ6tGuZmoUlW0VIwrN =pE16 -----END PGP SIGNATURE-----

1 0

Re: ddraw/tests: A spelling fix for a d3d ok() message.
by Marvin 15 May '16

15 May '16

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=23014 Your paranoid android. === w1064 (32 bit d3d) === d3d.c:2326: Test failed: Unexpected destination texture level pixels; 2031 differences at 0 level d3d.c:2326: Test failed: Unexpected destination texture level pixels; 518 differences at 1 level d3d.c:2326: Test failed: Unexpected destination texture level pixels; 138 differences at 2 level d3d.c:2326: Test failed: Unexpected destination texture level pixels; 46 differences at 3 level d3d.c:2510: Test failed: Unexpected destination texture level pixels; 257 differences d3d.c:2657: Test failed: Unexpected destination texture level pixels; 256 differences at 2 level d3d.c:2657: Test failed: Unexpected destination texture level pixels; 64 differences at 3 level d3d.c:2657: Test failed: Unexpected destination texture level pixels; 16 differences at 4 level d3d.c:2657: Test failed: Unexpected destination texture level pixels; 4 differences at 5 level

1 0

Re: [PATCH] kernel32: Set C3_KASHIDA only for kashida character
by Marvin 15 May '16

15 May '16

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=23006 Your paranoid android. === w1064 (32 bit locale) === locale.c:2367: Test failed: LCMapStringEx should fail with bad locale name locale.c:2368: Test failed: unexpected error code -559038737 locale.c:2459: Test failed: Expected lcid == 0, got got 00001000, error -559038737 locale.c:4189: Test failed: IsValidLocaleName should have failed locale.c:4191: Test failed: IsValidLocaleName should have failed locale.c:4556: Test failed: For id 7, expected ret == 4, got 1, error 87 locale.c:4558: Test failed: For id 7, Expected IVC, got '' === w1064 (64 bit locale) === locale.c:2367: Test failed: LCMapStringEx should fail with bad locale name locale.c:2368: Test failed: unexpected error code -559038737 locale.c:2459: Test failed: Expected lcid == 0, got got 00001000, error -559038737 locale.c:4189: Test failed: IsValidLocaleName should have failed locale.c:4191: Test failed: IsValidLocaleName should have failed locale.c:4556: Test failed: For id 7, expected ret == 4, got 1, error 87 locale.c:4558: Test failed: For id 7, Expected IVC, got ''

2 1

Re: [PATCH] kernel32: Handle NULL locale name in IsValidLocaleName()
by Austin English 15 May '16

15 May '16

On May 15, 2016 2:30 AM, "Nikolay Sivov" <nsivov(a)codeweavers.com> wrote: > > Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> > --- > dlls/kernel32/locale.c | 3 +++ > dlls/kernel32/tests/locale.c | 2 ++ > 2 files changed, 5 insertions(+) > > diff --git a/dlls/kernel32/locale.c b/dlls/kernel32/locale.c > index a69e76f..636187f 100644 > --- a/dlls/kernel32/locale.c > +++ b/dlls/kernel32/locale.c > @@ -2656,6 +2656,9 @@ BOOL WINAPI IsValidLocaleName( LPCWSTR locale ) > { > struct locale_name locale_name; > > + if (!locale) > + return FALSE; > + > /* string parsing */ > parse_locale_name( locale, &locale_name ); > > diff --git a/dlls/kernel32/tests/locale.c b/dlls/kernel32/tests/locale.c > index 20f52d2..70944a2 100644 > --- a/dlls/kernel32/tests/locale.c > +++ b/dlls/kernel32/tests/locale.c > @@ -4179,6 +4179,8 @@ static void test_IsValidLocaleName(void) > ok(!ret, "IsValidLocaleName should have failed\n"); > ret = pIsValidLocaleName(LOCALE_NAME_INVARIANT); > ok(ret, "IsValidLocaleName failed\n"); > + ret = pIsValidLocaleName(NULL); > + ok(!ret, "IsValidLocaleName should have failed\n"); > } > > static void test_CompareStringOrdinal(void) > -- > 2.8.1 Hi Nikolay, I'm just curious what application this fixes? Or was this something you found while writing tests for something else?

2 1

Re: [PATCH] kernel32: Handle NULL locale name in IsValidLocaleName()
by Marvin 15 May '16

15 May '16

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=23005 Your paranoid android. === w1064 (32 bit locale) === locale.c:2367: Test failed: LCMapStringEx should fail with bad locale name locale.c:2368: Test failed: unexpected error code -559038737 locale.c:2459: Test failed: Expected lcid == 0, got got 00001000, error -559038737 locale.c:4177: Test failed: IsValidLocaleName should have failed locale.c:4179: Test failed: IsValidLocaleName should have failed locale.c:4546: Test failed: For id 7, expected ret == 4, got 1, error 87 locale.c:4548: Test failed: For id 7, Expected IVC, got '' === w1064 (64 bit locale) === locale.c:2367: Test failed: LCMapStringEx should fail with bad locale name locale.c:2368: Test failed: unexpected error code -559038737 locale.c:2459: Test failed: Expected lcid == 0, got got 00001000, error -559038737 locale.c:4177: Test failed: IsValidLocaleName should have failed locale.c:4179: Test failed: IsValidLocaleName should have failed locale.c:4546: Test failed: For id 7, expected ret == 4, got 1, error 87 locale.c:4548: Test failed: For id 7, Expected IVC, got ''

1 0