5b952d2cbf
We were passing arrays by value which the compiler fails to optimize, causing abyssal performance. To fix this, we need to consider that descriptors can be in constant or const device address spaces. Also, lone descriptors are passed by value, so we explicitly remove address space qualifiers. One failure case is when shader passes a texture/sampler array as an argument. It's all UniformConstant in SPIR-V, but in MSL it might be thread, const device or constant, so that won't work ... Global variable use works fine though, and that should cover 99.9999999% of use cases.
28 lines
880 B
GLSL
28 lines
880 B
GLSL
#pragma clang diagnostic ignored "-Wmissing-prototypes"
|
|
|
|
#include <metal_stdlib>
|
|
#include <simd/simd.h>
|
|
|
|
using namespace metal;
|
|
|
|
struct main0_out
|
|
{
|
|
float4 FragColor [[color(0)]];
|
|
};
|
|
|
|
static inline __attribute__((always_inline))
|
|
float4 load_subpasses(texture2d_ms<float> uInput, thread uint& gl_SampleID, thread float4& gl_FragCoord)
|
|
{
|
|
float4 _24 = uInput.read(uint2(gl_FragCoord.xy), gl_SampleID);
|
|
return _24;
|
|
}
|
|
|
|
fragment main0_out main0(texture2d_ms<float> uSubpass0 [[texture(0)]], texture2d_ms<float> uSubpass1 [[texture(1)]], uint gl_SampleID [[sample_id]], float4 gl_FragCoord [[position]])
|
|
{
|
|
main0_out out = {};
|
|
gl_FragCoord.xy += get_sample_position(gl_SampleID) - 0.5;
|
|
out.FragColor = (uSubpass0.read(uint2(gl_FragCoord.xy), 1) + uSubpass1.read(uint2(gl_FragCoord.xy), 2)) + load_subpasses(uSubpass0, gl_SampleID, gl_FragCoord);
|
|
return out;
|
|
}
|
|
|