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

RE: [PATCH 1/3] d3drm/tests: Add test for IDirect3DRM*::CreateDeviceFromClipper (try 5).
by Aaryaman Vasishta 25 Jun '15

25 Jun '15

> > This is by no means a full review, just a couple of obvious things I noticed. > > On 24 June 2015 at 17:02, Aaryaman Vasishta <jem456.vasishta at gmail.com <https://www.winehq.org/mailman/listinfo/wine-devel>> wrote: > >* +static HRESULT CALLBACK surface_callback(IDirectDrawSurface *surface, DDSURFACEDESC *desc, void *context) > *>* +{ > *>* + IDirectDrawClipper *d3drm_clipper = NULL; > *Initializing this to NULL is pointless, you initialize it before all > uses with the IDirectDrawSurface_GetClipper() call below. > > >* + if (context) > *>* + { > *"context" should always be non-NULL here. > > >* + hr = IDirectDrawSurface_GetClipper(surface, &d3drm_clipper); > *>* + ok(hr == DD_OK, "Cannot get attached clipper from primary surface (hr = %x).\n", hr); > *>* + if (SUCCEEDED(hr)) > *You have an ok() one line above that states "hr == DD_OK" at this > point. If the GetClipper() call failed you'd get a test failure. > > Once you've found the primary you can of course return DDENUMRET_CANCEL. > > I think you're making things too complicated though. I'd just call > EnumSurfaces() with "&primary" as context pointer, set it in the > callback, and then just do the rest of the tests after EnumSurfaces() > returns, instead of trying to fit those into the callback. > > Thank you for the suggestions! I have decided to return the primary surface directly, and set it to NULL if it doesn't exist via the context, something like this, which seems to run fine on my windows machine: > static HRESULT CALLBACK surface_callback(IDirectDrawSurface *surface, > DDSURFACEDESC *desc, void **context) > { > IDirectDrawClipper *d3drm_clipper; > HRESULT hr; > > if (desc->ddsCaps.dwCaps & DDSCAPS_PRIMARYSURFACE) > { > *context = surface; > return DDENUMRET_CANCEL; > } > IDirectDrawSurface_Release(surface); > > return DDENUMRET_OK; > } > Then in the tests, simply check if the context returned is NULL, if yes then perform related tests on it inside the test. Thank you!

1 1

Re: [3/14]hid: translated HidD_ functions to their IOCTLs
by Nikolay Sivov 25 Jun '15

25 Jun '15

On 25.06.2015 18:15, Aric Stewart wrote: > BOOLEAN WINAPI HidD_SetFeature(HANDLE HidDeviceObject, PVOID ReportBuffer, ULONG ReportBufferLength) > { > - FIXME("(%p %p %u) stub\n", HidDeviceObject, ReportBuffer, ReportBufferLength); > - return FALSE; > + TRACE("(%p %p %u)\n", HidDeviceObject, ReportBuffer, ReportBufferLength); > + return DeviceIoControl(HidDeviceObject, IOCTL_HID_SET_OUTPUT_REPORT, ReportBuffer, ReportBufferLength, NULL, 0, NULL, NULL); > } I can't say I know anything about that, but this looks like a typo, maybe it should use IOCTL_HID_SET_FEATURE?

1 0

Re: d3dx9-36: Save multiplications to improve performance
by Matteo Bruni 25 Jun '15

25 Jun '15

2015-06-25 4:07 GMT+02:00 David Adam <david.adam.cnrs(a)gmail.com>: > Hello, > > with my dual-core, for sides=65000 the gain is about 40%. It probably has more to do with avoiding the division than the multiplication. Can you add a test for a polygon with that many faces and see if the result matches native better with the simplification or without?

1 0

Re: [1/3] d3dcompiler: Share the source with d3dcompiler_46 (try 2)
by Matteo Bruni 25 Jun '15

25 Jun '15

2015-06-25 12:30 GMT+02:00 Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com>: > Hi, > Rename define to D3D_COMPILER_VERSION > > Changelog: > d3dcompiler: Share the source with d3dcompiler_46 > > > Best Regards > Alistair Leslie-Hughes Series looks good to me, thanks!

1 0

Re: [PATCH 1/3] d3drm/tests: Add test for IDirect3DRM*::CreateDeviceFromClipper (try 5).
by Henri Verbeet 25 Jun '15

25 Jun '15

This is by no means a full review, just a couple of obvious things I noticed. On 24 June 2015 at 17:02, Aaryaman Vasishta <jem456.vasishta(a)gmail.com> wrote: > +static HRESULT CALLBACK surface_callback(IDirectDrawSurface *surface, DDSURFACEDESC *desc, void *context) > +{ > + IDirectDrawClipper *d3drm_clipper = NULL; Initializing this to NULL is pointless, you initialize it before all uses with the IDirectDrawSurface_GetClipper() call below. > + if (context) > + { "context" should always be non-NULL here. > + hr = IDirectDrawSurface_GetClipper(surface, &d3drm_clipper); > + ok(hr == DD_OK, "Cannot get attached clipper from primary surface (hr = %x).\n", hr); > + if (SUCCEEDED(hr)) You have an ok() one line above that states "hr == DD_OK" at this point. If the GetClipper() call failed you'd get a test failure. Once you've found the primary you can of course return DDENUMRET_CANCEL. I think you're making things too complicated though. I'd just call EnumSurfaces() with "&primary" as context pointer, set it in the callback, and then just do the rest of the tests after EnumSurfaces() returns, instead of trying to fit those into the callback.

1 0

Re: wineboot: Create MachineGuid registry value.
by Alexandre Julliard 25 Jun '15

25 Jun '15

Sebastian Lackner <sebastian(a)fds-team.de> writes: > The MachineGuid registry key should exist even without calling CryptAcquireContextW. > Optionally we could remove the remove the corresponding code from advapi32 afterwards. This doesn't need to run at every boot, so you can probably stick it into the registration of some crypto dll like wintrust. And please also remove the advapi32 code. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: [1/3] wininet/tests: Add tests for overriding host header.
by Marvin 25 Jun '15

25 Jun '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=14698 Your paranoid android. === w7u (32 bit http) === http.c:2891: Test failed: HttpSendRequest failed http.c:2896: Test failed: HttpQueryInfo failed http.c:2897: Test failed: got status 0, expected 200 http.c:5813: Test failed: HttpAddRequestHeadersA failed 12019 http.c:5819: Test failed: HttpQueryInfoA failed 12019 http.c:5820: Test failed: header missing

1 0

Re: [1/3] wininet/tests: Add tests for overriding host header.
by Sebastian Lackner 25 Jun '15

25 Jun '15

On 25.06.2015 12:27, Sebastian Lackner wrote: > From: Michael Müller <michael(a)fds-team.de> > > --- > dlls/wininet/tests/http.c | 109 ++++++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 109 insertions(+) > The testbot already shows a test failure on Win7, but it seems to be random. It doesn't occur in a second run: https://newtestbot.winehq.org/JobDetails.pl?Key=14699.

1 0

Re: [PATCH (resend)] cmd: Consistently wrap lines to 75 characters.
by Alexandre Julliard 25 Jun '15

25 Jun '15

Alex Henrie <alexhenrie24(a)gmail.com> writes: > 75 characters is the maximum length of a line before gettext breaks it, > so this convention will cause each translatable string to fit snugly > into the po files. It's also about what we were doing anyway, just not > consistently. > > The biggest advantage to consistent line wrapping is that translators > will no longer have to guess where they should insert line breaks. You'll need to also update the duplicate text in programs/start, and it would be nice if you could try to fix at least some of the translations so that we don't have to invalidate all of them. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: d3drm: Mark DllCanUnloadNow and DllGetClassObject as private.
by Alexandre Julliard 25 Jun '15

25 Jun '15

Sebastian Lackner <sebastian(a)fds-team.de> writes: > @@ -19,5 +19,5 @@ > @ stdcall D3DRMVectorScale(ptr ptr float) > @ stdcall D3DRMVectorSubtract(ptr ptr ptr) > @ stdcall Direct3DRMCreate(ptr) > -@ stub DllCanUnloadNow > -@ stub DllGetClassObject > +@ stub -private DllCanUnloadNow > +@ stub -private DllGetClassObject Private is not useful on a stub. -- Alexandre Julliard julliard(a)winehq.org

1 0