On Mon Sep 25 15:51:27 2023 +0000, Francisco Casas wrote:
I think that the blocker is just the first point: "Replace sm4_register with vkd3d_shader_register" (the first part is !319), we can certainly support non-constant offset dereferences before solving the second point, which may take a while. I am not sure I understand the third point, besides what is addressed by 2/4.
This is still our only blocker for native D3D11 support, and thanks to
Recent Developments I should be able to start helping with HLSL and TPF features again soon! Nice!
That all makes sense to me - with that in mind I tried doing a manual rebase of this based on !370, but got lost a bit at around the codegen stage. Happy to test out any WIPs of a refreshed version of this MR ASAP!