1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
|
#version 460
vec4 undeclared_errors(vec4 f4)
{
vec4 result;
gl_SubgroupSize; // ERROR, extension not enabled (basic)
gl_SubgroupInvocationID; // ERROR, extension not enabled (basic)
subgroupBarrier(); // ERROR, extension not enabled (basic)
subgroupMemoryBarrier(); // ERROR, extension not enabled (basic)
subgroupMemoryBarrierBuffer(); // ERROR, extension not enabled (basic)
subgroupMemoryBarrierImage(); // ERROR, extension not enabled (basic)
subgroupElect(); // ERROR, extension not enabled (basic)
gl_NumSubgroups; // ERROR, extension not enabled (basic)
gl_SubgroupID; // ERROR, extension not enabled (basic)
subgroupMemoryBarrierShared(); // ERROR, extension not enabled (basic)
subgroupAll(true); // ERROR extension not enabled (vote)
subgroupAny(false); // ERROR extension not enabled (vote)
subgroupAllEqual(f4); // ERROR extension not enabled (vote)
gl_SubgroupEqMask; // ERROR extension not enabled (ballot)
gl_SubgroupGeMask; // ERROR extension not enabled (ballot)
gl_SubgroupGtMask; // ERROR extension not enabled (ballot)
gl_SubgroupLeMask; // ERROR extension not enabled (ballot)
gl_SubgroupLtMask; // ERROR extension not enabled (ballot)
subgroupBroadcast(f4, 0); // ERROR extension not enabled (ballot)
subgroupBroadcastFirst(f4); // ERROR extension not enabled (ballot)
uvec4 ballot = subgroupBallot(false); // ERROR extension not enabled (ballot)
subgroupInverseBallot(uvec4(0x1)); // ERROR extension not enabled (ballot)
subgroupBallotBitExtract(ballot, 0); // ERROR extension not enabled (ballot)
subgroupBallotBitCount(ballot); // ERROR extension not enabled (ballot)
subgroupBallotInclusiveBitCount(ballot); // ERROR extension not enabled (ballot)
subgroupBallotExclusiveBitCount(ballot); // ERROR extension not enabled (ballot)
subgroupBallotFindLSB(ballot); // ERROR extension not enabled (ballot)
subgroupBallotFindMSB(ballot); // ERROR extension not enabled (ballot)
subgroupShuffle(f4, 0); // ERROR extension not enabled (shuffle)
subgroupShuffleXor(f4, 0x1); // ERROR extension not enabled (shuffle)
subgroupShuffleUp(f4, 1); // ERROR extension not enabled (shuffle_relative)
subgroupShuffleDown(f4, 1); // ERROR extension not enabled (shuffle_relative)
result = subgroupAdd(f4); // ERROR, extension not enabled (arith)
subgroupMul(f4); // ERROR, extension not enabled (arith)
subgroupMin(f4); // ERROR, extension not enabled (arith)
subgroupMax(f4); // ERROR, extension not enabled (arith)
subgroupAnd(ballot); // ERROR, extension not enabled (arith)
subgroupOr(ballot); // ERROR, extension not enabled (arith)
subgroupXor(ballot); // ERROR, extension not enabled (arith)
subgroupInclusiveAdd(f4); // ERROR, extension not enabled (arith)
subgroupInclusiveMul(f4); // ERROR, extension not enabled (arith)
subgroupInclusiveMin(f4); // ERROR, extension not enabled (arith)
subgroupInclusiveMax(f4); // ERROR, extension not enabled (arith)
subgroupInclusiveAnd(ballot); // ERROR, extension not enabled (arith)
subgroupInclusiveOr(ballot); // ERROR, extension not enabled (arith)
subgroupInclusiveXor(ballot); // ERROR, extension not enabled (arith)
subgroupExclusiveAdd(f4); // ERROR, extension not enabled (arith)
subgroupExclusiveMul(f4); // ERROR, extension not enabled (arith)
subgroupExclusiveMin(f4); // ERROR, extension not enabled (arith)
subgroupExclusiveMax(f4); // ERROR, extension not enabled (arith)
subgroupExclusiveAnd(ballot); // ERROR, extension not enabled (arith)
subgroupExclusiveOr(ballot); // ERROR, extension not enabled (arith)
subgroupExclusiveXor(ballot); // ERROR, extension not enabled (arith)
subgroupClusteredAdd(f4, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredMul(f4, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredMin(f4, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredMax(f4, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredAnd(ballot, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredOr(ballot, 2); // ERROR, extension not enabled (clustered)
subgroupClusteredXor(ballot, 2); // ERROR, extension not enabled (clustered)
subgroupQuadBroadcast(f4, 0); // ERROR, extension not enabled (quad)
subgroupQuadSwapHorizontal(f4); // ERROR, extension not enabled (quad)
subgroupQuadSwapVertical(f4); // ERROR, extension not enabled (quad)
subgroupQuadSwapDiagonal(f4); // ERROR, extension not enabled (quad)
uvec4 parti = subgroupPartitionNV(f4); // ERROR, extension not enabled (partitioned)
subgroupPartitionedAddNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedMulNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedMinNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedMaxNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedAndNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedOrNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedXorNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveAddNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveMulNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveMinNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveMaxNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveAndNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveOrNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedInclusiveXorNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveAddNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveMulNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveMinNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveMaxNV(f4, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveAndNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveOrNV(ballot, parti); // ERROR, extension not enabled (partitioned)
subgroupPartitionedExclusiveXorNV(ballot, parti); // ERROR, extension not enabled (partitioned)
return result;
}
#define MAX_VIEWS gl_MaxMeshViewCountNV
#define BARRIER() \
memoryBarrierShared(); \
barrier();
#extension GL_NV_mesh_shader : enable
layout(local_size_x = 32) in;
// test use of shared memory in task shaders:
layout(binding=0) writeonly uniform image2D uni_image;
uniform block0 {
uint uni_value;
};
shared vec4 mem[10];
// test use of task memory in task shaders:
taskNV out Task {
vec2 dummy;
vec2 submesh[3];
uint viewID;
} mytask;
void main()
{
uint iid = gl_LocalInvocationID.x;
uint gid = gl_WorkGroupID.x;
uint viewID = gl_MeshViewIndicesNV[gl_MeshViewCountNV%MAX_VIEWS];
// 1. shared memory load and stores
for (uint i = 0; i < 10; ++i) {
mem[i] = vec4(i + uni_value);
}
imageStore(uni_image, ivec2(iid), mem[gid]);
imageStore(uni_image, ivec2(iid), mem[gid+1]);
BARRIER();
// 2. task memory stores
mytask.dummy = vec2(30.0, 31.0);
mytask.submesh[0] = vec2(32.0, 33.0);
mytask.submesh[1] = vec2(34.0, 35.0);
mytask.submesh[2] = mytask.submesh[gid%2];
mytask.viewID = viewID;
BARRIER();
// 3. set task count
gl_TaskCountNV = 3;
}
#extension GL_KHR_shader_subgroup_basic: enable
void basic_works (void)
{
gl_SubgroupSize;
gl_SubgroupInvocationID;
subgroupBarrier();
subgroupMemoryBarrier();
subgroupMemoryBarrierBuffer();
subgroupMemoryBarrierImage();
subgroupElect();
gl_NumSubgroups; // allowed in task
gl_SubgroupID; // allowed in task
subgroupMemoryBarrierShared(); // allowed in task
}
#extension GL_KHR_shader_subgroup_ballot: enable
void ballot_works(vec4 f4) {
gl_SubgroupEqMask;
gl_SubgroupGeMask;
gl_SubgroupGtMask;
gl_SubgroupLeMask;
gl_SubgroupLtMask;
subgroupBroadcast(f4, 0);
subgroupBroadcastFirst(f4);
uvec4 ballot = subgroupBallot(false);
subgroupInverseBallot(uvec4(0x1));
subgroupBallotBitExtract(ballot, 0);
subgroupBallotBitCount(ballot);
subgroupBallotInclusiveBitCount(ballot);
subgroupBallotExclusiveBitCount(ballot);
subgroupBallotFindLSB(ballot);
subgroupBallotFindMSB(ballot);
}
#extension GL_KHR_shader_subgroup_vote: enable
void vote_works(vec4 f4)
{
subgroupAll(true);
subgroupAny(false);
subgroupAllEqual(f4);
}
#extension GL_KHR_shader_subgroup_shuffle: enable
#extension GL_KHR_shader_subgroup_shuffle_relative: enable
void shuffle_works(vec4 f4)
{
subgroupShuffle(f4, 0);
subgroupShuffleXor(f4, 0x1);
subgroupShuffleUp(f4, 1);
subgroupShuffleDown(f4, 1);
}
#extension GL_KHR_shader_subgroup_arithmetic: enable
void arith_works(vec4 f4)
{
uvec4 ballot;
subgroupAdd(f4);
subgroupMul(f4);
subgroupMin(f4);
subgroupMax(f4);
subgroupAnd(ballot);
subgroupOr(ballot);
subgroupXor(ballot);
subgroupInclusiveAdd(f4);
subgroupInclusiveMul(f4);
subgroupInclusiveMin(f4);
subgroupInclusiveMax(f4);
subgroupInclusiveAnd(ballot);
subgroupInclusiveOr(ballot);
subgroupInclusiveXor(ballot);
subgroupExclusiveAdd(f4);
subgroupExclusiveMul(f4);
subgroupExclusiveMin(f4);
subgroupExclusiveMax(f4);
subgroupExclusiveAnd(ballot);
subgroupExclusiveOr(ballot);
subgroupExclusiveXor(ballot);
}
#extension GL_KHR_shader_subgroup_clustered: enable
void clustered_works(vec4 f4)
{
uvec4 ballot = uvec4(0x55,0,0,0);
subgroupClusteredAdd(f4, 2);
subgroupClusteredMul(f4, 2);
subgroupClusteredMin(f4, 2);
subgroupClusteredMax(f4, 2);
subgroupClusteredAnd(ballot, 2);
subgroupClusteredOr(ballot, 2);
subgroupClusteredXor(ballot, 2);
}
#extension GL_KHR_shader_subgroup_quad: enable
void quad_works(vec4 f4)
{
subgroupQuadBroadcast(f4, 0);
subgroupQuadSwapHorizontal(f4);
subgroupQuadSwapVertical(f4);
subgroupQuadSwapDiagonal(f4);
}
#extension GL_NV_shader_subgroup_partitioned: enable
void partitioned_works(vec4 f4)
{
uvec4 parti = subgroupPartitionNV(f4);
uvec4 ballot = uvec4(0x55,0,0,0);
subgroupPartitionedAddNV(f4, parti);
subgroupPartitionedMulNV(f4, parti);
subgroupPartitionedMinNV(f4, parti);
subgroupPartitionedMaxNV(f4, parti);
subgroupPartitionedAndNV(ballot, parti);
subgroupPartitionedOrNV(ballot, parti);
subgroupPartitionedXorNV(ballot, parti);
subgroupPartitionedInclusiveAddNV(f4, parti);
subgroupPartitionedInclusiveMulNV(f4, parti);
subgroupPartitionedInclusiveMinNV(f4, parti);
subgroupPartitionedInclusiveMaxNV(f4, parti);
subgroupPartitionedInclusiveAndNV(ballot, parti);
subgroupPartitionedInclusiveOrNV(ballot, parti);
subgroupPartitionedInclusiveXorNV(ballot, parti);
subgroupPartitionedExclusiveAddNV(f4, parti);
subgroupPartitionedExclusiveMulNV(f4, parti);
subgroupPartitionedExclusiveMinNV(f4, parti);
subgroupPartitionedExclusiveMaxNV(f4, parti);
subgroupPartitionedExclusiveAndNV(ballot, parti);
subgroupPartitionedExclusiveOrNV(ballot, parti);
subgroupPartitionedExclusiveXorNV(ballot, parti);
}
// tests for NV_shader_sm_builtins
void sm_builtins_err()
{
gl_WarpsPerSMNV; // ERROR, no extension
gl_SMCountNV; // ERROR, no extension
gl_WarpIDNV; // ERROR, no extension
gl_SMIDNV; // ERROR, no extension
}
#ifdef GL_NV_shader_sm_builtins
#extension GL_NV_shader_sm_builtins : enable
#endif
void sm_builtins()
{
gl_WarpsPerSMNV;
gl_SMCountNV;
gl_WarpIDNV;
gl_SMIDNV;
}
|