You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Unlike in HLSL, HLSL++ will still pack all types in 16b n128, even if running in HLSLPP_SCALAR. That breaks the C++<->HLSL interop with constant/structured buffer packing which is an useful use case. Any ideas on how it could be (easily) hacked to support HLSL-like alignment? :) Thanks!