From: Henri Verbeet Subject: Re: [PATCH] d2d1/tests: Add tests for 2D affine effect. Message-Id: Date: Thu, 9 Sep 2021 17:35:15 +0200 In-Reply-To: <20210903033332.641821-1-zhui@codeweavers.com> References: <20210903033332.641821-1-zhui@codeweavers.com> On Fri, 3 Sept 2021 at 05:33, Ziqing Hui wrote: > + const DWORD image_4x4[] = > + { > + 0xfdcba987, 0xffff0000, 0x98765432, 0xffffffff, > + 0x4b4b4b4b, 0x89abcdef, 0xdeadbeef, 0xabcdef01, > + 0x7f000011, 0x40ffffff, 0x12345678, 0xaabbccdd, > + 0x44444444, 0xff7f7f7f, 0x1221abba, 0x00000000 > + }; I think that can be static const. > + ID2D1DeviceContext_BeginDraw(context); > + ID2D1DeviceContext_Clear(context, 0); > + ID2D1DeviceContext_DrawImage(context, output, NULL, NULL, 0, 0); > + hr = ID2D1DeviceContext_EndDraw(context, NULL, NULL); > + ok(hr == S_OK, "Got unexpected hr %#x.\n", hr); > + get_surface_readback(&ctx, &rb); > + colour = get_readback_colour(&rb, 0, 0); > + todo_wine > + ok(compare_colour(colour, test->origin_colour, 1), > + "Got unexpected colour %#x, expected %#x.\n", colour, test->origin_colour); > + "rb" is never released. I'm not sure how helpful checking only a single pixel of the result really is. This seems like something where we'd want to use compare_figure(), or perhaps compare_surface()/compare_sha1() if we're careful about the transformation matrices we use. (Compare test_bitmap_brush().)