d1479f871a
This avoids a lot of huge code changes. Arrays generally cannot be copied in and out of buffers, at least no compiler frontend seems to do it. Also avoids a lot of issues surrounding packed vectors and matrices.
34 lines
1.1 KiB
Plaintext
34 lines
1.1 KiB
Plaintext
#pragma clang diagnostic ignored "-Wmissing-prototypes"
|
|
|
|
#include <metal_stdlib>
|
|
#include <simd/simd.h>
|
|
|
|
using namespace metal;
|
|
|
|
struct SSBO
|
|
{
|
|
float4 in_data[1];
|
|
};
|
|
|
|
struct SSBO2
|
|
{
|
|
float4 out_data[1];
|
|
};
|
|
|
|
constant uint3 gl_WorkGroupSize [[maybe_unused]] = uint3(1u);
|
|
|
|
// Implementation of the GLSL mod() function, which is slightly different than Metal fmod()
|
|
template<typename Tx, typename Ty>
|
|
inline Tx mod(Tx x, Ty y)
|
|
{
|
|
return x - y * floor(x / y);
|
|
}
|
|
|
|
kernel void main0(const device SSBO& _23 [[buffer(0)]], device SSBO2& _33 [[buffer(1)]], uint3 gl_GlobalInvocationID [[thread_position_in_grid]])
|
|
{
|
|
_33.out_data[gl_GlobalInvocationID.x] = mod(_23.in_data[gl_GlobalInvocationID.x], _33.out_data[gl_GlobalInvocationID.x]);
|
|
_33.out_data[gl_GlobalInvocationID.x] = as_type<float4>(as_type<uint4>(_23.in_data[gl_GlobalInvocationID.x]) % as_type<uint4>(_33.out_data[gl_GlobalInvocationID.x]));
|
|
_33.out_data[gl_GlobalInvocationID.x] = as_type<float4>(as_type<int4>(_23.in_data[gl_GlobalInvocationID.x]) % as_type<int4>(_33.out_data[gl_GlobalInvocationID.x]));
|
|
}
|
|
|