November 2016 - Wine-Devel - list.winehq.org

Re: Testing Direct3D

by Stefan Dösinger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, We're using real Windows machines, yeah. The driver is a big part of d3d, so you want an AMD or Nvidia GPU, not virtualbox (which uses wined3d) or vmware. Intel GPUs have known problems that break some games. Stefan Am 24. November 2016 16:18:29 GMT+00:00, schrieb Fabian Maurer <dark.shadow4(a)web.de>: >Hello Wine-Devel, > >It might be a stupid question, but how do you test the Direct3D >implementation? I don't mean the conformance tests, these too need to >get the >data somewhere. So you probably have a windows running somewhere. > >I currently I don't have a windows machine to test on though. Only a >Virtualbox VM, and this doesn't offer (reliable) 3D acceleration for DX >> 9. >Do you have your own windows machine dedicated to this, or how do you >do it? - -- Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail gesendet. -----BEGIN PGP SIGNATURE----- Version: APG v1.1.1 iQJGBAEBCgAwBQJYNz34KRxTdGVmYW4gRPZzaW5nZXIgPHN0ZWZhbmRvZXNpbmdl ckBnbXguYXQ+AAoJEN0/YqbEcdMwTIgP/0QrNY4Foy9P6VqnIxLpk2ldAXKqqvL2 GhVPWhV/o3Xljeqhk1mmxoHCLmeuLsrb+nyRU06epxVOW6iCiGQ/M1wymvJOnvCv VA1h4DCc5FebXdaMhow3eknOpTXXieegPiUTXizDyefWLIVbJk95UsRTXMdGESA8 Vj5+h4T7/8oZ95NAccNRIaETPo2iuBBhXSXSqDKB+3vyMPIwUSfU3tkRW8gh/SVc /A6V0xx7qLL2QbKVfvK8k5mKxEdUJDpFQXDvpGADMM04DI8mHGkD7yrtokBCGAXn APZt8E7VOGgLgbr0ycj4vCodhIcrVYZWwWHGwB/YCidKGjVKmss9kcV9dTm6nk6a U3806npg8PrqzeGQy3StTK43paWB77g0w69diHs/CcMboaSWmRfn6hv5mHXflGKz +PQbQyt4l7+vXkAakHkd4IC6ODpXV/hn1ZpbrgLFZRkt0Uas7FLriliE0xnbGnz7 sdbmqhW+LPBnl9rlS0aNEpblk8hlwDW+CJEvZt/DZnx8VI4DGfRdWzd5ND3EzY3P 9NYHZ/vu5oRi97I5AyXj3aOfR9aZXWzPxx64DHFH1XeW0YQTWf6mQmhzfewJAY2w JzQBnlmHSZYmW3vXRQ6tSU+Rt5r1ku/TMGLWGkJyHQpD1A7EZKkseGrP6RjAJOsC SC/NX9yJA6k9 =f3u4 -----END PGP SIGNATURE-----

Nov. 24, 2016

Re: WineTest reference Linux machines

by Francois Gouget

On Wed, 23 Nov 2016, Francois Gouget wrote: [...] > Note that we did not get the WoW tests. I looked at > wt-bot-cw1-hd6800-wow32.report and it looks like the X server (or fvwm) > crashed: > > Starting up > err:winediag:nulldrv_CreateWindow Application tried to create a window, > but no driver could be loaded. > err:winediag:nulldrv_CreateWindow Make sure that your X server is > running and that $DISPLAY is set correctly. > Error: Unable to create a window, the display driver is not working. > Command exited with non-zero status 1 > > > We'll see if it still happens with click to focus. I think I found why the X server was disppearing: lightdm sometimes fails to autologin so I have a script that checks whether the Xfce session is running which restarts lightdm if not. But now that it's fvwm instead of Xfce I had to update the script. I also tweaked wt-daily to better detect when the tests don't complete or the report upload fails and to not count the tests as done in those cases. That will give them another chance to run should the X server really crash. So we still have a failure for d3d9:device and d3d9:visual. Are those still related to the window manager? -- Francois Gouget <fgouget(a)codeweavers.com>

Nov. 24, 2016

Re: Github mirror seems to be out-of-sync

by Nikolay Sivov

On 24.11.2016 19:19, Akihiro Sagawa wrote: > Source code admin, > > Official Github mirror, https://github.com/wine-mirror/wine, seems to be > not updated for a few days. Could you take a look? > Thanks in advance. > I think github automatically maintains this mirror on its own, it wasn't Wine project initiative to have it. > Akihiro Sagawa

Nov. 24, 2016

Github mirror seems to be out-of-sync

by Akihiro Sagawa

Source code admin, Official Github mirror, https://github.com/wine-mirror/wine, seems to be not updated for a few days. Could you take a look? Thanks in advance. Akihiro Sagawa

Nov. 24, 2016

Testing Direct3D

by Fabian Maurer

Hello Wine-Devel, It might be a stupid question, but how do you test the Direct3D implementation? I don't mean the conformance tests, these too need to get the data somewhere. So you probably have a windows running somewhere. I currently I don't have a windows machine to test on though. Only a Virtualbox VM, and this doesn't offer (reliable) 3D acceleration for DX > 9. Do you have your own windows machine dedicated to this, or how do you do it?

Nov. 24, 2016

Re: [2/5] strmbase: Add validation checks when updating source rectangle.

by Akihiro Sagawa

On Mon, 21 Nov 2016 10:19:59 -0600, Andrew Eikum wrote: > On Sun, Nov 20, 2016 at 03:33:22PM +0900, Akihiro Sagawa wrote: > > @@ -53,6 +53,26 @@ HRESULT WINAPI BaseControlVideo_Destroy(BaseControlVideo *pControlVideo) > > return BaseDispatch_Destroy(&pControlVideo->baseDispatch); > > } > > > > +static HRESULT BaseControlVideoImpl_CheckSourceRect(BaseControlVideo *This, RECT *pSourceRect) > > +{ > > + LONG VideoWidth, VideoHeight; > > + HRESULT hr; > > + > > + if (pSourceRect->top < 0 || pSourceRect->left < 0 || > > + pSourceRect->top >= pSourceRect->bottom || > > + pSourceRect->left >= pSourceRect->right) > > + return E_INVALIDARG; > > I'm not opposed to spelling it out, but you could use IsRectEmpty() > here instead. Thanks for reviewing. I sent v2 patch set based on your helpful comments. It also includes fix of underlying copy-paste error. > > + > > + hr = BaseControlVideoImpl_GetVideoSize((IBasicVideo *)This, &VideoWidth, &VideoHeight); > > + if (FAILED(hr)) > > + return hr; > > + > > + if (pSourceRect->bottom > VideoHeight || pSourceRect->right > VideoWidth) > > + return E_INVALIDARG; > > + > > Shouldn't these use (bottom - top) and (right - left)? I'd like to say no at this point because native doesn't accept source rectangle that points outside of the original video area. tests/filtergraph.c's line 94 and 104 test this case. Akihiro Sagawa

Nov. 24, 2016

Re: 2 of 13 [PATCH] po: Update Simplified Chinese translation (v3).

by Jactry Zeng

This looks good to me. Thanks for your contribution! 2016-11-23 0:09 GMT+08:00 Mingcong Bai <jeffbaichina(a)gmail.com>: > Signed-off-by: Mingcong Bai <jeffbai(a)aosc.xyz> > --- > po/zh_CN.po | 1202 ++++++++++++++++++++++++++++++ > ++++------------------------- > 1 file changed, 687 insertions(+), 515 deletions(-) > > > > -- Regards, Jactry Zeng

Nov. 24, 2016

Re: [PATCH 1/2] shlwapi: Remove dot segments for URLs starting with /

by Jacek Caban

Hi Daniel, On 24.11.2016 00:44, Daniel Lehman wrote: > diff --git a/dlls/shlwapi/url.c b/dlls/shlwapi/url.c > index 879a734..5de9076 100644 > --- a/dlls/shlwapi/url.c > +++ b/dlls/shlwapi/url.c > @@ -372,6 +372,9 @@ HRESULT WINAPI UrlCanonicalizeW(LPCWSTR pszUrl, LPWSTR pszCanonicalized, > dwFlags |= URL_ESCAPE_UNSAFE; > state = 5; > is_file_url = TRUE; > + } else if(url[0] == '/') { > + state = 2; Wouldn't state 5 be more appropriate? > + is_file_url = TRUE; It would be interesting to have tests using URL_WININET_COMPATIBILITY and URL_FILE_USE_PATHURL flags. is_file_url usually affects slashes conversion in such cases, but you set it too late to take such effect. It may be right, but tests would be nice. Thanks, Jacek

Nov. 24, 2016

Re: [PATCH 04/11] wined3d: Introduce wined3d_device_set_unordered_access_view().

by Henri Verbeet

On 23 November 2016 at 14:36, Józef Kucia <jkucia(a)codeweavers.com> wrote: > + FIXME("Unordered access view unsupported yet (layers %u-%u).\n", > + desc->u.texture.layer_idx, desc->u.texture.layer_count); I don't think you can do that in English. You'd either have "UAV unsupported" or "UAV not supported yet". In the specific case here, I'd probably use something like "Unsupported layer selection %u+%u.\n" or something along those lines though.

Nov. 24, 2016

Re: [PATCH] d2d1: Implement retrieving dash pattern data for stroke styles

by Henri Verbeet

On 24 November 2016 at 12:13, Nikolay Sivov <bunglehead(a)gmail.com> wrote: > Oh, right. That means GetGradientStops() is broken too. Are you okay > with keeping memset() or it's better to explicitly set them to 0.0f? > Memset() is fine with me.

Nov. 24, 2016