Inner Daemons
Inner Daemons
@amarhassam Yeah, I try to post updates to #7197 every now and then, and you can track the status there. I've had 2 people offer to write the HLSL writer...
@cwfitzgerald All comments should be addressed, ready for round 2! (=2)
I'll work on fixing the tests... at some point
For future me to enjoy and ponder: ``` 0x78e9e40b7010 100000 0x78e9e1da32f0 72 0x78e9e40b7010 100000 0x78e9e1d87610 25 0x78e9e40b7010 1
I also have some AMD GPU code from windows that I'll upload at some point but its absolutely unreadable to my eyes.
For future reference: I have taken the generated code with debug symbols and put it through SPIRV-Opt, I have put it through SPIRV-Cross and then back through glslc, and both...
Update: got this to work by modifying the body of main() in the generated GLSL. Therefore, the interface is fine, and its just an issue with something goofy.
These comments are spammy if anyone is listening so you don't need to I have narrowed it down to this little bit of code that breaks stuff even if the...
The following when replacing the body of the generated GLSL makes it terminate with sigsegv: ``` void main() { mesh_output.primitives[0u].cull = !taskPayload.visible; SetMeshOutputsEXT(3, 1); bool cull = mesh_output.primitives[gl_LocalInvocationIndex].cull; gl_MeshPrimitivesEXT[0].gl_CullPrimitiveEXT =...
``` #version 450 #extension GL_EXT_mesh_shader : require layout(local_size_x = 1, local_size_y = 1, local_size_z = 1) in; layout(max_vertices = 3, max_primitives = 1, triangles) out; struct TaskPayload { vec4 colorMask;...