February 2015 - Wine-devel - list.winehq.org

Re: [PATCH 4/6] d3d9/tests: Test rhw depth clip with z enabled but no depth buffer set.
by Marvin 11 Feb '15

11 Feb '15

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=11494 Your paranoid android. === w8 (32 bit visual) === visual.c:15503: Test failed: Expected color 0x0000ff00 at 80, 60, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 560, 60, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 80, 180, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 560, 180, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 80, 300, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 560, 300, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 80, 420, got 0x00ff0000, test 0. visual.c:15503: Test failed: Expected color 0x0000ff00 at 560, 420, got 0x00ff0000, test 0.

1 0

Re: [PATCH 2/6] dxgi: Add interface for IDXGISurface1 [try 2]
by Stefan Dösinger 11 Feb '15

11 Feb '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Martell, This patch (and http://source.winehq.org/patches/data/108999) look good to me from a content point of view. The E-Mail submission is a bit broken though. If I pipe the raw mail into git am I get a commit message that has some HTML and other junk in it. I recommend to set up git send-email. It will create the proper numbering for you and make sure that there are no line wraps or HTML junk in the mail. If you want to send only some part of your patches with send-email you can e.g. create a temporary branch, cherry-pick what you want to send, and then send from that branch. Also I recommend to send one patch after another, wait for it to be applied, then send the next one. Cheers, Stefan Am 2015-02-10 um 13:11 schrieb Martell Malone: > I'm going to take this slowly, with this there should be no issues > with he first 2 patches. > > I went with dxgi: in this commit message as that's what henri said > > The prefix for the dxgi IDL patches should be "dxgi:", for the > d3d11 one it should be "d3d11:". > > > Before I resubmit 4 and 6 however I need to know if I should have > "include:" or "dxgi:" + "d3d11:" in them as these create new files -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJU20OEAAoJEN0/YqbEcdMwYhQP/3+K4X1/NH/K5ke0MdCCXzNO af8QxlcLJjNg6XH6bParDEDvs4u9pax3oGULzT07uu9hNpk1/zkssxV6tc3wJIs+ Sxx2A/EOU2ggtJJ5Dn3RPRKFJAiXJCwRClJcePj94lp5S11zNpoBcEJyn9Y9rZZc afvK2cPjj5w5ym8b2Ml9us16JbmN+JyAIr7gX59jZy+67+W6GpZdbB1/OV55Up4y bJH0Sl83vh6GzsZi0QyQ/oRE8OWxhYstOTmPEUMFX66qJh4wzgOy5+oRRmro0PoR X9bshnHGwQPEZs74UtCxpPusyGL6ApcZ5TAdsXiy+J6qhsnhv+bGLbQqwOp0RtPn mMs8V4S50h+84vvO/v/JhsPxMyKt0dC8HqhsHYn9Zl4Idv3bT6Q3wTbaovVKRbTM X0nB6HVoQPVKPRRXnRo2g8DTumrW4Qyg44Ymx3RsW8BNmnPy7YdJFjYIlW0fMwPP Q7MhrKv2lvJy7239Cm+iR1huYkMZgND+u7c45qTl+03SagOy5T+E4PzsZvAaf08O 89v1D1ogIgyvSnswuhY1o3XgQ4hMEm/I9BDespEXk9iedXwK9H4/+kvW+5kgvHJs TaDc0MNA97IiGz+Nvv4LEKk7kydqFmsIgaeKuTibrUs1+vyFWsCMHz+kcmxvA3Db uW3zEYIjKdG+i2xmwmmp =OaYZ -----END PGP SIGNATURE-----

1 0

Re: version: Fix wrong length reported when selecting a block.
by Stefan Dösinger 11 Feb '15

11 Feb '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Am 2015-02-09 um 23:20 schrieb learn_more: > From: Mark <learn0more(a)gmail.com> We need to know your full name to accept your patch to be able to show where our code comes from. Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJU2z8oAAoJEN0/YqbEcdMw/tYP/RxWhEnR4ZbwplYloK3vcq6H P08nlylTgQ+MfjUwReQxuPoVU3FNTx2+zD1omcJ37NJvorMONu6EUJ8OtCcBGmWO wmii92XCOZJkZOaOAO5pd/ziSamZeeQNVVG1fhW+NkcEF2Jp/PtuPwWF/eZ4UL9V Rz0mrf/BUZ6xCmCPz8hRO22foMCHpIArVtke1j/9GN0df5LB2bLpMlvcmQrR/YT2 r12xz4+z4FeyBFGS1wZ5s7U6LawlUVR6YUJtCwcBmxrmYAY3byiYoX8fXbYJt+CI rAoXo427GKLf27SArsT21Q2aFYzRff3hm5YQUDO2idm07bmiTTj5rdDAekq7DQKT 3m28wBBjrSyjbJCKAXRBJyvC2au0WdC0cTh/1Ji5FtOC0Csl6pJ2ST3iIrSvNgw/ AJgbizBGTsb9Wgtm1nH+AEJn9w7enqR1BGzIL4WNSNrsoyvSyO2WWR9Cc0demOle 2Z12gKTbiTWxssezxTqvpTWrX6Cqym2gJW8ZbRpK2J/INZbpqH889UwWC3LMw/8v rUDx23MV985/h2AS6h2takGDmvtKU/Wdv6zluaYdA9fpkmM7waLS5G5Lgrjx17rz H8JvB7VbLuf6eIzxVnlsZE6NJo+HXghoZyD00hKvM9zh1wRhNgMfcWgNfA47W1+w 9+id7LnpWmtcTEA99AuH =+oRF -----END PGP SIGNATURE-----

1 0

Re: setupapi: Implement SetupOpenLog(), SetupLogErrorA(), SetupLogErrorW(), SetupCloseLog() (try 3)
by Nikolay Sivov 11 Feb '15

11 Feb '15

This is better, but still not quire right. > +CRITICAL_SECTION setuplog_critical; Take a look at how critical section data is initialized in wine code, name needs to be adjusted to let's say 'setupapi_cs'. > + if (MessageString) > + { > + len = lstrlenA(MessageString) + 1; > + msg = HeapAlloc(GetProcessHeap(), 0, len * sizeof(WCHAR)); > + if (msg == NULL) > + { > + SetLastError(ERROR_NOT_ENOUGH_MEMORY); > + return FALSE; > + } > + MultiByteToWideChar(CP_ACP, 0, MessageString, -1, msg, len); > + } That's not how A->W conversion works, we have tons of examples for that. > OsVersionInfo.dwOSVersionInfoSize = sizeof(OSVERSIONINFOW); > if (!GetVersionExW(&OsVersionInfo)) > return FALSE; > + InitializeCriticalSection(&setuplog_critical); > SETUPAPI_hInstance = hinstDLL; > break; > case DLL_PROCESS_DETACH: > if (lpvReserved) break; > + DeleteCriticalSection(&setuplog_critical); > if (CABINET_hInstance) FreeLibrary(CABINET_hInstance); > break; You won't need these calls once you init it like the rest of Wine code does. Also what about closing files on DLL_PROCESS_DETACH? > + SetLastError(ERROR_ALREADY_EXISTS); > + return TRUE; Is it really important for some application to set this error on success? In general you only need to set it if call failed. > + setupact = CreateFileW(path, GENERIC_WRITE, FILE_SHARE_WRITE | FILE_SHARE_READ, > + NULL, OPEN_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL); Did you check what happens if multiple processes are logging?

2 3

Re: [PATCH 5/5] d3d9/tests: Test upload of user memory textures.
by Stefan Dösinger 11 Feb '15

11 Feb '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 2015-02-10 um 20:14 schrieb Matteo Bruni: > + hr = IDirect3DDevice9Ex_CreateTexture(device, 33, 33, 1, 0, D3DFMT_L8, > + D3DPOOL_SYSTEMMEM, &texture, &mem); > ... > + hr = IDirect3DDevice9Ex_SetTexture(device, 0, (IDirect3DBaseTexture9 *)texture); > + ok(SUCCEEDED(hr), "SetTexture failed (0x%08x)\n", hr); Am I missing something, or are you texturing from system memory here? I am surprised this is working. I'll test it on my collection of Windows boxes to be sure. I think it would be a good idea to test the UpdateTexture / UpdateSurface codepath with user memory textures and a strange pitch like in your test. This doesn't have to be part of this patch though. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJU2yz/AAoJEN0/YqbEcdMw/nMP/i61GWBL1ZajYzvS2iohebWp 6nE9k6Yh9OziU7LjJZfkWekvNPGPd8UYN2B5CzdqM6LPdBM3PFj3UFAROzECAPjn JpLHncb8gh25CXFAV0V/rdlaoXuYIKkxQDiZMyiDEP01gZNLTX+bt1rdWmK6JNtJ czFYyoSKyWyzbiUOQSAKponMxYbd2rK3B714Xsn50CRJdA6z26DSgqMfHabZcrl1 R+PWOfwDh7n6es7ax5rVWDNIvip1bFoU97wTahfIxIoL227eSTsFwwRo1pNsLnd4 9KsQmIWdKzIMZqPPkhnKGZz1Gu79Q3GJnlLjbWODYL5fgfbI+Bcv6npMZqHtctvb yFn86dVX2t6WTAMBmC4J3PVm+Igrzlk7nQt6l4X4KeMf7HYP9wDtOh/FIt4TV4qy 7Zu7JpuRDk5yKgnE/y7eew6YHyQGmVy59Y4IL7CR2EiwAevZr/HblvBjmsJ+CfSF NHF1QZkt/VH4v3D2XQzrW9Pln9H5m3luFTzA5cVOsv1v4r62uAY69tJLUmVRguyx icxFGbcuUKPtiFu4ER9Q0dZXx2gscy7agLYGbbwvsNFapBG2dTPaS1qJVyXe7/kD ML82rlfGImYupx6aSE/GAvD/IWaZgICt+XqPj5PnOIBz/mN9o9KziewCSg53Xp+2 Ai+BIdMBa8KOiK2MGPrH =Jvrb -----END PGP SIGNATURE-----

1 0

Re: [PATCH] user32: use InvalidateRect to repaint buttons
by Dmitry Timoshkov 11 Feb '15

11 Feb '15

Mark Harmstone <hellas(a)burntcomma.com> wrote: > At present, whenever a button needs to be redrawn user32 calls its own > internal function. This causes problems if the button is subclassed, as > the superclass then won't receive the WM_PAINT message - this is why > themed buttons lose their theme when you click on them. This patch changes > it to instead call InvalidateRect, which will force a WM_PAINT message > to be sent to the topmost class. Did you run 'make test' with your changes? -- Dmitry.

1 0

Re: [PATCH 3/5] wined3d: Set the surface pitch for user-memory surfaces.
by Henri Verbeet 10 Feb '15

10 Feb '15

On 10 February 2015 at 20:14, Matteo Bruni <mbruni(a)codeweavers.com> wrote: > + { > surface->resource.size = wined3d_format_calculate_size(texture_resource->format, > texture_resource->device->surface_alignment, width, height, 1); > + surface->pitch = wined3d_format_calculate_pitch(texture_resource->format, width); > + } > A small comment mentioning that user memory surfaces don't have the regular surface alignment probably wouldn't hurt.

1 0

Re: connect: add IsInternetConnectedStub
by Bruno Jesus 10 Feb '15

10 Feb '15

On Tue, Feb 10, 2015 at 4:03 PM, Austin English <austinenglish(a)gmail.com> wrote: > > https://bugs.winehq.org/show_bug.cgi?id=26759 Isn't it better to simply add a note like "faking success" and return S_OK? Because returning false will probably not have a useful effect in the application. I didn't test the application though. > -- > -Austin > Best regards, Bruno

2 1

Re: vbscript: Added InStrRev implementation.
by Marvin 10 Feb '15

10 Feb '15

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=11485 Your paranoid android. === w2000pro (32 bit) === No test summary line found === wxppro (32 bit) === No test summary line found === wvistau64 (32 bit) === No test summary line found === w2008s64 (32 bit) === No test summary line found === w7u (32 bit) === No test summary line found === w8 (32 bit) === No test summary line found === wvistau64 (64 bit) === No test summary line found === w2008s64 (64 bit) === No test summary line found

1 0

Re: [PATCH 2/6] dxgi: Add interface for IDXGISurface1 [try 2]
by Jacek Caban 10 Feb '15

10 Feb '15

On 02/10/15 13:11, Martell Malone wrote: > Before I resubmit 4 and 6 however I need to know if I should have > "include:" or "dxgi:" + "d3d11:" in them as these create new files We already have both variants in Git history and there was never a problem with it. If you ask me, use whatever you prefer. Jacek

3 2