532f65583e
Remove all shenanigans with propagation, and only consume nonuniform qualifiers exactly where needed (last minute).
56 lines
3.2 KiB
Plaintext
56 lines
3.2 KiB
Plaintext
#version 450
|
|
#extension GL_EXT_buffer_reference : require
|
|
#extension GL_EXT_nonuniform_qualifier : require
|
|
#extension GL_KHR_shader_subgroup_ballot : require
|
|
layout(local_size_x = 1, local_size_y = 1, local_size_z = 1) in;
|
|
|
|
layout(set = 0, binding = 0, std430) restrict readonly buffer SSBO_Offsets
|
|
{
|
|
uvec2 _m0[];
|
|
} _7;
|
|
|
|
layout(push_constant, std430) uniform RootConstants
|
|
{
|
|
uint _m0;
|
|
uint _m1;
|
|
uint _m2;
|
|
uint _m3;
|
|
uint _m4;
|
|
uint _m5;
|
|
uint _m6;
|
|
uint _m7;
|
|
} registers;
|
|
|
|
layout(set = 1, binding = 0) uniform samplerBuffer _8[];
|
|
layout(set = 4, binding = 0, r32f) uniform imageBuffer _9[];
|
|
layout(set = 4, binding = 0, r32ui) uniform uimageBuffer _10[];
|
|
|
|
void main()
|
|
{
|
|
uint _61 = registers._m4 + 2u;
|
|
uint _64 = subgroupBroadcastFirst(_61);
|
|
uint _71 = subgroupBroadcastFirst(registers._m4);
|
|
uint _76 = registers._m1 + 1u;
|
|
uint _79 = subgroupBroadcastFirst(_76);
|
|
uint _88 = gl_GlobalInvocationID.x + 4u;
|
|
uint _99 = gl_GlobalInvocationID.x + 1024u;
|
|
imageStore(_9[registers._m4], int((_99 < _7._m0[_71].y) ? (_99 + _7._m0[_71].x) : 4294967295u), vec4(imageLoad(_9[registers._m4], int((_88 < _7._m0[_71].y) ? (_88 + _7._m0[_71].x) : 4294967295u))));
|
|
uint _106 = gl_GlobalInvocationID.x + 2u;
|
|
uint _117 = gl_GlobalInvocationID.x + 2048u;
|
|
imageStore(_9[registers._m4], int((_117 < _7._m0[_71].y) ? (_117 + _7._m0[_71].x) : 4294967295u), vec4(texelFetch(_8[_76], int((_106 < _7._m0[_79].y) ? (_106 + _7._m0[_79].x) : 4294967295u))));
|
|
uint _130 = imageAtomicAdd(_10[_61], int((gl_GlobalInvocationID.x < _7._m0[_64].y) ? (gl_GlobalInvocationID.x + _7._m0[_64].x) : 4294967295u), 40u);
|
|
uint _137 = imageAtomicCompSwap(_10[_61], int((gl_GlobalInvocationID.y < _7._m0[_64].y) ? (gl_GlobalInvocationID.y + _7._m0[_64].x) : 4294967295u), 40u, 50u);
|
|
imageStore(_9[registers._m4], int((0u < _7._m0[_71].y) ? (0u + _7._m0[_71].x) : 4294967295u), vec4(float(_7._m0[_71].y)));
|
|
imageStore(_9[registers._m4], int((1u < _7._m0[_71].y) ? (1u + _7._m0[_71].x) : 4294967295u), vec4(float(_7._m0[_79].y)));
|
|
uint _11 = registers._m4 + (gl_GlobalInvocationID.z + 0u);
|
|
imageStore(_9[nonuniformEXT(_11)], int((_99 < _7._m0[_11].y) ? (_99 + _7._m0[_11].x) : 4294967295u), vec4(imageLoad(_9[nonuniformEXT(_11)], int((_88 < _7._m0[_11].y) ? (_88 + _7._m0[_11].x) : 4294967295u))));
|
|
uint _13 = registers._m1 + (gl_GlobalInvocationID.z + 0u);
|
|
imageStore(_9[nonuniformEXT(_11)], int((_117 < _7._m0[_11].y) ? (_117 + _7._m0[_11].x) : 4294967295u), vec4(texelFetch(_8[nonuniformEXT(_13)], int((_88 < _7._m0[_13].y) ? (_88 + _7._m0[_13].x) : 4294967295u))));
|
|
uint _15 = registers._m4 + (gl_GlobalInvocationID.z + 0u);
|
|
uint _209 = imageAtomicAdd(_10[nonuniformEXT(_15)], int((gl_GlobalInvocationID.y < _7._m0[_15].y) ? (gl_GlobalInvocationID.y + _7._m0[_15].x) : 4294967295u), 40u);
|
|
uint _215 = imageAtomicCompSwap(_10[nonuniformEXT(_15)], int((gl_GlobalInvocationID.y < _7._m0[_15].y) ? (gl_GlobalInvocationID.y + _7._m0[_15].x) : 4294967295u), 40u, 70u);
|
|
imageStore(_9[registers._m4], int((2u < _7._m0[_71].y) ? (2u + _7._m0[_71].x) : 4294967295u), vec4(float(_7._m0[_11].y)));
|
|
imageStore(_9[registers._m4], int((3u < _7._m0[_71].y) ? (3u + _7._m0[_71].x) : 4294967295u), vec4(float(_7._m0[_13].y)));
|
|
}
|
|
|