September 2018 - Wine-Devel

Re: [PATCH 3/3] ddraw: Check for DDRAW_NO3D before checking for a video memory surface in d3d_device_create().

by Henri Verbeet

On 28 September 2018 at 04:07, Mathew Hodson <mathew.hodson(a)gmail.com> wrote: > On Thu, 27 Sep 2018 at 11:46, Henri Verbeet <hverbeet(a)codeweavers.com> wrote: >> - return DDERR_NO3D; >> + return DDERR_OUTOFMEMORY; > > Is it supposed to return out of memory when DDRAW_NO3D is set? I > didn't get that from the commit message. > Yeah. The commit message didn't quite spell it out, but see e.g. [1] and [2]. [1] https://test.winehq.org/data/d8f3f1b71726b793f2f9a86225a73a7d286e62b3/win7_… [2] https://source.winehq.org/git/wine.git/blob/HEAD:/dlls/ddraw/tests/d3d.c#l1…

Sept. 27, 2018

Re: [PATCH] wined3d: Modernize the macOS spelling.

by Henri Verbeet

On 28 September 2018 at 04:02, Francois Gouget <fgouget(a)free.fr> wrote: > On Fri, 28 Sep 2018, Henri Verbeet wrote: >> I don't necessarily want to block this going in if you really care >> about this, but I also think the name-change is completely ridiculous. > > You mean Apple's decision to change the name or the patch ;-) > Apple's decision. > The thing is that "MacOS X" is a misspelled form of "Mac OS X", the > former OS name, so my script reports it as a typo[1]. And it is with > spelling issues as with warnings or test failures: if you let them > accumulate you soon cannot tell what's new from what's not[2]. > > So I can either change that to one of the official names or disable this > specific check for Wine. So even though it's a minor issue I typically > prefer to just send a patch once in a while. But I can disable this > check if you prefer. > I guess what I'm saying is that I'm fine with this going in, but have reservations about endorsing it with a Signed-off-by.

Sept. 27, 2018

[PATCH] testbot/Build: Only rebuild impacted test executables.

by Francois Gouget

If a patch impacts a module but not its tests, don't try to rebuild the corresponding test executable. This fixes build failures when the module does not have tests in the first place. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/bin/build/Build.pl | 1 + 1 file changed, 1 insertion(+) diff --git a/testbot/bin/build/Build.pl b/testbot/bin/build/Build.pl index 1287f6db5..54b1aeae1 100755 --- a/testbot/bin/build/Build.pl +++ b/testbot/bin/build/Build.pl @@ -78,6 +78,7 @@ sub BuildTestExecutables($$$) my (@BuildDirs, @TestExes); foreach my $TestInfo (values %{$Impacts->{Tests}}) { + next if (!$TestInfo->{UnitCount}); push @BuildDirs, $TestInfo->{Path}; my $TestExe = "$TestInfo->{Path}/$TestInfo->{ExeBase}.exe"; push @TestExes, $TestExe; -- 2.19.0

Sept. 27, 2018

Re: [PATCH 3/3] ddraw: Check for DDRAW_NO3D before checking for a video memory surface in d3d_device_create().

by Mathew Hodson

On Thu, 27 Sep 2018 at 11:46, Henri Verbeet <hverbeet(a)codeweavers.com> wrote: > > While slightly odd, this matches what the tests expect. > > Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com> > --- > dlls/ddraw/device.c | 14 +++++++------- > 1 file changed, 7 insertions(+), 7 deletions(-) > > diff --git a/dlls/ddraw/device.c b/dlls/ddraw/device.c > index 5db0b81a819..7866c281465 100644 > --- a/dlls/ddraw/device.c > +++ b/dlls/ddraw/device.c > @@ -6986,18 +6986,18 @@ HRESULT d3d_device_create(struct ddraw *ddraw, struct ddraw_surface *target, IUn > return DDERR_NOPALETTEATTACHED; > } > > - if (!(target->surface_desc.ddsCaps.dwCaps & DDSCAPS_VIDEOMEMORY)) > - { > - WARN("Surface %p is not in video memory.\n", target); > - return D3DERR_SURFACENOTINVIDMEM; > - } > - > if (ddraw->flags & DDRAW_NO3D) > { > ERR_(winediag)("The application wants to create a Direct3D device, " > "but the current DirectDrawRenderer does not support this.\n"); > > - return DDERR_NO3D; > + return DDERR_OUTOFMEMORY; Is it supposed to return out of memory when DDRAW_NO3D is set? I didn't get that from the commit message. > + } > + > + if (!(target->surface_desc.ddsCaps.dwCaps & DDSCAPS_VIDEOMEMORY)) > + { > + WARN("Surface %p is not in video memory.\n", target); > + return D3DERR_SURFACENOTINVIDMEM; > } > > if (ddraw->d3ddevice) > -- > 2.11.0 -- Sincerely, Mathew Hodson

Sept. 27, 2018

Re: [PATCH] wined3d: Modernize the macOS spelling.

by Francois Gouget

On Fri, 28 Sep 2018, Henri Verbeet wrote: > On 27 September 2018 at 10:55, Francois Gouget <fgouget(a)free.fr> wrote: > > Signed-off-by: Francois Gouget <fgouget(a)free.fr> > > --- > > dlls/wined3d/adapter_gl.c | 14 +++++++------- > > 1 file changed, 7 insertions(+), 7 deletions(-) > > > I don't necessarily want to block this going in if you really care > about this, but I also think the name-change is completely ridiculous. You mean Apple's decision to change the name or the patch ;-) The thing is that "MacOS X" is a misspelled form of "Mac OS X", the former OS name, so my script reports it as a typo[1]. And it is with spelling issues as with warnings or test failures: if you let them accumulate you soon cannot tell what's new from what's not[2]. So I can either change that to one of the official names or disable this specific check for Wine. So even though it's a minor issue I typically prefer to just send a patch once in a while. But I can disable this check if you prefer. [1] Although the X is part of 'X server' the script is still right: it's my understanding that the OS name is either 'Mac OS X' or 'OS X' or now 'macOS', but cannot be 'MacOS'. [2] And people also copy / paste the bad spelling or get influenced by it; both cases leading to the spread of the misspelled form. -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ Nouvelle version : les anciens bogues ont \xC3\xA9t\xC3\xA9 remplac\xC3\xA9s par de nouveaux.

Sept. 27, 2018

Re: [PATCH] wined3d: Remove extraneous flush from EndScene.

by Henri Verbeet

Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com>

Sept. 27, 2018

Re: [PATCH 2/2] d2d1: Implement newer CreateBitmap() variant.

by Henri Verbeet

Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com>

Sept. 27, 2018

Re: [PATCH 1/2] d2d1/tests: Add some tests for compatible target size handling.

by Henri Verbeet

On 27 September 2018 at 17:47, Nikolay Sivov <nsivov(a)codeweavers.com> wrote: > +static void test_compatible_target_size(ID2D1RenderTarget *rt) > +{ > + float dpi_x, dpi_y, rt_dpi_x, rt_dpi_y; > + ID2D1BitmapRenderTarget *bitmap_rt; > + ID2D1DeviceContext *context; > + D2D1_SIZE_U pixel_size; > + D2D1_SIZE_F size; > + HRESULT hr; > + > + ID2D1RenderTarget_GetDpi(rt, &rt_dpi_x, &rt_dpi_y); > + > + pixel_size.height = pixel_size.width = 0; > + hr = ID2D1RenderTarget_CreateCompatibleRenderTarget(rt, NULL, &pixel_size, NULL, > + D2D1_COMPATIBLE_RENDER_TARGET_OPTIONS_NONE, &bitmap_rt); > +todo_wine > + ok(SUCCEEDED(hr), "Failed to create render target, hr %#x.\n", hr); > + if (FAILED(hr)) > + return; > + I'd typically suggest to use ok_() for helpers like this one. In this case, that doesn't help a lot for the "target dpi" tests though. Any chance you could make those table-driven and then prefix the ok-message with the index?

Sept. 27, 2018

Re: [PATCH vkd3d 1/8] vkd3d: Simplify error handling in vkd3d_create_vk_device().

by Henri Verbeet

On 27 September 2018 at 15:00, Józef Kucia <joseph.kucia(a)gmail.com> wrote: > @@ -1149,12 +1144,17 @@ static HRESULT vkd3d_create_vk_device(struct d3d12_device *device, > { > ERR("Failed to create queues, hr %#x.\n", hr); > device->vk_procs.vkDestroyDevice(vk_device, NULL); > - return hr; > + goto fail; > } > > TRACE("Created Vulkan device %p.\n", vk_device); > > - return S_OK; > +fail: > + if (queue_properties) > + vkd3d_free(queue_properties); > + if (queue_info) > + vkd3d_free(queue_info); > + return hr; > } I guess this is fine, but I'd feel a little better about it if the label was called e.g. "done" in this case, since we're also going through there on success.

Sept. 27, 2018

Re: [PATCH vkd3d 8/8] tests: Add test for initial pipeline state.

by Henri Verbeet

Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com>

Sept. 27, 2018