ReceiveTempSMS
/
Sweden-Number
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

d3d9/tests: Accept Nvidia results in test_sample_mask as broken. 

Signed-off-by: Stefan Dösinger <stefan@codeweavers.com>
Signed-off-by: Henri Verbeet <hverbeet@codeweavers.com>
Signed-off-by: Alexandre Julliard <julliard@winehq.org>
This commit is contained in:
Stefan Dösinger 2022-01-21 12:58:12 +03:00 committed by Alexandre Julliard
parent 47f5a50260
commit fc98f7d9f9
1 changed files with 9 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
dlls/d3d9/tests/visual.c
View File

@ -26911,7 +26911,15 @@ static void test_sample_mask(void)
ok(hr == D3D_OK, "Got unexpected hr %#x.\n", hr);
get_rt_readback(rt, &rb);
colour = get_readback_color(&rb, 64, 64);
ok(color_match(colour, 0xffff8080, 1), "Got unexpected colour %08x.\n", colour);
/* Multiple generations of Nvidia cards return broken results.
* A mask with no bits or all bits set produce the expected results (0x00 / 0xff),
* but any other mask behaves almost as if the result is 0.5 + (enabled / total)
* samples. It's not quite that though (you'd expect 0xbf or 0xc0 instead of 0xbc).
*
* I looked at a few other possible problems: Incorrectly enabled Z test, alpha test,
* culling, the multisample mask affecting CopyRects. Neither of these make a difference. */
ok(color_match(colour, 0xffff8080, 1) || broken(color_match(colour, 0xffffbcbc, 1)),
"Got unexpected colour %08x.\n", colour);
release_surface_readback(&rb);
hr = IDirect3DDevice9_EndScene(device);