On 1/19/21 6:13 PM, Henri Verbeet wrote:
On Mon, 18 Jan 2021 at 13:59, Rémi Bernon rbernon@codeweavers.com wrote:
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=49395 Signed-off-by: Rémi Bernon rbernon@codeweavers.com
dlls/d2d1/d2d1_private.h | 6 ++++- dlls/d2d1/device.c | 54 ++++++++++++++++++++++++++++++++++++++++ dlls/d2d1/tests/d2d1.c | 2 +- 3 files changed, 60 insertions(+), 2 deletions(-)
Is the design of this series purely a consequence of the decision to not implement state swapping in d3d11_immediate_context_SwapDeviceContextState(), or is there some other consideration I'm overlooking? In particular, without having done an attempt at implementing this myself, the "natural" implementation to me would seem to simply replace existing usage of d3d10 interfaces with their d3d11 equivalents, where the ID3DDeviceContextState interface would be the equivalent of the ID3D10StateBlock interface.
Well, I imagined D2D1 should still have to work over D3D10-only devices? Although I'm not sure if there's such a thing it didn't seem obvious to me that every D3D10 device is also a D3D11 device.