Re: [PATCH] ddraw: add dsurface dimension tests, try 2

[Patrick Rudolph]

Am 2013-11-18 16:44, schrieb Stefan Dösinger:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi,
>
> Overall, did you check if there is a capability reporting problem? The
> game may have the bad size from our DDSCAPS structure.

I'm not sure. It successfully creates a primary surface with dimension 
320x240 and it creates offscreen surfaces with dimension 320x240 and 
DDSCAPS_SYSTEMMEMORY flag set. Only offscreen surfaces with 
DDSCAPS_VIDEOMEMORY flag set are out of range. I was wondering: 15728640 
= 320 << 16
and 20971520 = 240 << 16. Where does the two byte offset come from ?
> Even if this is a capability problem I think this test is relevant.
>
> Am 2013-11-17 20:19, schrieb Patrick Rudolph:
> > dlls/ddraw/tests/dsurface.c | 88
>
> +++++++++++++++++++++++++++++++++++++++++++++
>
> Please add new tests to ddraw{1, 2, 4, 7}.c, and if possible implement
> them for all ddraw versions.

I can do that as well.
> > - ZeroMemory(&desc, sizeof(desc));
>
> memset(desc, 0, sizeof(desc)); is preferred.
> > - if(!(ddcaps.ddsCaps.dwCaps & DDSCAPS_VIDEOMEMORY)) +    { +
> >
> > todo_wine ok(ret == DDERR_INVALIDPARAMS, "Creating an
>
> offscreen plain surface with width 0xffff and height 1 returned
> %08x\n", ret);
> > - }
>
> Is there any situation in which wine enters this codepath? I think the
> todo_wine is not needed here.

You are right, wine always presents a direct draw device with 
DDSCAPS_VIDEOMEMORY. I removed ligGL, ddraw still "works", but doesn't 
render anything.
> > - /* Test out of memory surface size */ +    ZeroMemory(&desc,
> >
> > sizeof(desc)); +    desc.dwSize = sizeof(desc); +    desc.dwFlags =
> > DDSD_CAPS | DDSD_HEIGHT | DDSD_WIDTH; +    desc.ddsCaps.dwCaps |=
> > DDSCAPS_OFFSCREENPLAIN |
>
> DDSCAPS_VIDEOMEMORY;
> > - desc.dwHeight = 0xffff; +    desc.dwWidth = 0xffff; +    /*
> >
> > this should allocate 4.29 Gbyte of video memory */
>
> Testing DDSCAPS_SYSTEMMEMORY might show some interesting results as
> well, especially in a Win64 executable.

I can add tests for this.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.22 (GNU/Linux)
> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>
> iQIcBAEBAgAGBQJSijXyAAoJEN0/YqbEcdMwolYP+wWH7o7Dn9LjmjJq0dca5ZD9
> EKfyGNxrtg/Th1o2qKdP4rmsl1G8UC96Xo2BP9Fi5tH9BZTzAXTiq2pfSiKGf0Z7
> l48UbVGVuVd/nxBG8ULZDS7e8b9h+NwcprM3flWW6frbqsDMxf1LPhtNsEQU8gnE
> 0/hFXW/T33FkPP3XiBK6vCgWjP1umK9clcQHvit7DitFpFBmwOj5SxktoIsoPVpL
> TcOiaaXKcABTbKCz4xclOaAMM7k94cH0ByJBVu+8OXNTOHzd0xbtrnFl6UZi74jH
> dQyqLIyQ4uDs1760He7WtwgMEeF4NhV+u3A6qRP3eLceSf59OhEhSUNGKu5yduVb
> F16yGgQiTIKHpqtD1sT1RPEDAZgUYL9Uq8BnUrPbNZ+FJI5ZwIbY7KghozAhUweA
> /TC9VMLwUnJKk8v/8xHPY1WI3474Pht4joahUNWkTQ8QuN7ms44IyyyALBsWM3gF
> eOSTBNEcWZz+3uRaS68/ZDeqzo+R+dwmwU8XTeWEMTxuy30x1OdCTp3BgR8YNXMa
> 2uLWa5mL4pkOAVJ6gI6HHPpX8T582lB3PcZgRMd5QirwIg76dwvnk6zpiv4RqFS5
> 3HCCey4c3FO4aFhDcelkNsdcpCMbaGyv4lqtLDzBe7bR5xOrtWvmdl5r3vvlCbO2
> uERNPhePZ8ogI6buwboM
> =v8j0
> -----END PGP SIGNATURE-----