Searched refs:invocations (Results 1 – 25 of 424) sorted by relevance
12345678910>>...17
41 a set of boolean conditions across a group of shader invocations. These44 invocations across which boolean conditions are evaluated is46 how individual shader invocations are assigned to such sets. In47 particular, the set of shader invocations has no necessary relationship48 with the compute shader local work group -- a pair of shader invocations54 non-compute shader invocations and execute them in a SIMD fashion. When63 where <condition> diverges between invocations, a SIMD implementation64 might first call do_fast_path() for the invocations where <condition> is65 true and leave the other invocations dormant. Once do_fast_path()66 returns, it might call do_general_path() for invocations where <condition>[all …]
41 a set of boolean conditions across a group of shader invocations. These44 invocations across which boolean conditions are evaluated is46 how individual shader invocations are assigned to such sets. In47 particular, the set of shader invocations has no necessary relationship48 with the compute shader workgroup -- a pair of shader invocations54 non-compute shader invocations and execute them in a SIMD fashion. When63 where <condition> diverges between invocations, a SIMD implementation64 might first call do_fast_path() for the invocations where <condition> is65 true and leave the other invocations dormant. Once do_fast_path()66 returns, it might call do_general_path() for invocations where <condition>[all …]
51 across shader invocations within a sub-group.87 specified value <v> across all active shader invocations in the sub-group131 …| genIType minInvocationsAMD(genIType v) | invocations in the sub-group with <Reduce…136 …| genIType minInvocationsNonUniformAMD(genIType v) | invocations in the sub-group with <Reduce…141 …| genIType minInvocationsInclusiveScanAMD(genIType v) | invocations in the sub-group with <Inclus…150 …| genType v) | invocations in the sub-group with <Inclus…159 …| genIType minInvocationsExclusiveScanAMD(genIType v) | invocations in the sub-group with <Exclus…168 …| genType v) | invocations in the sub-group with <Exclus…177 …| genIType maxInvocationsAMD(genIType v) | invocations in the sub-group with <Reduce…182 …| genIType maxInvocationsNonUniformAMD(genIType v) | invocations in the sub-group with <Reduce…[all …]
61 a set of boolean conditions across a group of shader invocations. These64 invocations across which boolean conditions are evaluated is66 how individual shader invocations are assigned to such sets. In67 particular, the set of shader invocations has no necessary relationship68 with the compute shader workgroup -- a pair of shader invocations74 non-compute shader invocations and execute them in a SIMD fashion. When83 where <condition> diverges between invocations, a SIMD implementation84 might first call do_fast_path() for the invocations where <condition> is85 true and leave the other invocations dormant. Once do_fast_path()86 returns, it might call do_general_path() for invocations where <condition>[all …]
62 This extension provides the ability for a group of invocations which103 A sub-group is a collection of invocations which execute in lockstep.104 The variable <gl_SubGroupSizeARB> is the maximum number of invocations124 <gl_SubGroupSizeARB> is the total number of invocations in a sub-group.126 The <gl_SubGroup??MaskARB> variables provide a bitmask for all invocations,148 evaluating the expression <value> in all active invocations in the150 call. The sub-group may have inactive invocations for example due to152 invocations may be represented by the return value of ballotARB(). Bits156 to invocations that are not active or that do not exist in the sub group161 set for all active invocations in the sub-group.[all …]
61 large number of fragment shader invocations that perform loads and65 invocations is largely undefined. For algorithms that use shader76 invocations from touching the same memory concurrently.81 invocations with "overlapping" coverage in a given pixel, the OpenGL97 guarantees that the critical section for multiple shader invocations with107 worrying about other invocations for the same pixel accessing the data152 ordering of the execution of shader invocations between calls to the164 By default, fragment shader invocations are generally executed in165 undefined order. Multiple fragment shader invocations may be executed166 concurrently, including multiple invocations corresponding to a single[all …]
51 a set of boolean conditions across a group of shader invocations. These54 invocations across which boolean conditions are evaluated is56 how individual shader invocations are assigned to such sets. In57 particular, the set of shader invocations has no necessary relationship58 with the compute shader local work group -- a pair of shader invocations64 non-compute shader invocations and execute them in a SIMD fashion. When73 where <condition> diverges between invocations, a SIMD implementation74 might first call do_fast_path() for the invocations where <condition> is75 true and leave the other invocations dormant. Once do_fast_path()76 returns, it might call do_general_path() for invocations where <condition>[all …]
52 This extension provides the ability for a group of invocations which93 A sub-group is a collection of invocations which execute in lockstep.94 The variable <gl_SubGroupSizeARB> is the maximum number of invocations114 <gl_SubGroupSizeARB> is the total number of invocations in a sub-group.116 The <gl_SubGroup??MaskARB> variables provide a bitmask for all invocations,138 evaluating the expression <value> in all active invocations in the140 call. The sub-group may have inactive invocations for example due to142 invocations may be represented by the return value of ballotARB(). Bits146 to invocations that are not active or that do not exist in the sub group151 set for all active invocations in the sub-group.[all …]
51 large number of fragment shader invocations that perform loads and55 invocations is largely undefined. For algorithms that use shader66 invocations from touching the same memory concurrently.71 invocations with "overlapping" coverage in a given pixel, the OpenGL87 guarantees that the critical section for multiple shader invocations with97 worrying about other invocations for the same pixel accessing the data142 ordering of the execution of shader invocations between calls to the154 By default, fragment shader invocations are generally executed in155 undefined order. Multiple fragment shader invocations may be executed156 concurrently, including multiple invocations corresponding to a single[all …]
35 shader invocations that are running concurrently (a _subgroup_).49 individual shader invocations are assigned to subgroups.51 shader _local workgroup_ -- any pair of shader invocations in a compute57 invocations and execute them concurrently.69 where code:condition diverges between invocations, an implementation might70 first execute code:do_fast_path() for the invocations where code:condition71 is true and leave the other invocations dormant.73 invocations where code:condition is code:false and leave the other74 invocations dormant.76 might be better off just using the general path for all invocations.[all …]
31 invocations.32 The code:DerivativeGroupQuadsNV execution mode assembles shader invocations35 The code:DerivativeGroupLinearNV execution mode assembles shader invocations48 (1) Should we specify that the groups of four shader invocations used for49 derivatives in a compute shader are the same groups of four invocations that
29 This extension provides the ability for a group of invocations, which39 * code:SubgroupGeMaskKHR, containing the subgroup mask of the invocations41 * code:SubgroupGtMaskKHR, containing the subgroup mask of the invocations43 * code:SubgroupLeMaskKHR, containing the subgroup mask of the invocations45 * code:SubgroupLtMaskKHR, containing the subgroup mask of the invocations49 * code:SubgroupSize, containing the maximum number of invocations in a
153 EXPECT_EQ(35, urbg.invocations()); in TEST()159 EXPECT_EQ(35, urbg.invocations()); in TEST()165 EXPECT_EQ(35, urbg.invocations()); in TEST()171 EXPECT_EQ(35, urbg.invocations()); in TEST()200 EXPECT_EQ(12, urbg.invocations()); in TEST()206 EXPECT_EQ(12, urbg.invocations()); in TEST()212 EXPECT_EQ(12, urbg.invocations()); in TEST()
408 EXPECT_EQ(13, urbg.invocations()); in TEST()421 EXPECT_EQ(13, urbg.invocations()); in TEST()476 EXPECT_EQ(1, urbg.invocations()) << box << " " << std::hex << v; in TEST()491 EXPECT_EQ(1, urbg.invocations()) << box << " " << std::hex << v; in TEST()513 EXPECT_EQ(3, urbg.invocations()); in TEST()522 EXPECT_EQ(3, urbg.invocations()); in TEST()537 EXPECT_EQ(2, urbg.invocations()); in TEST()546 EXPECT_EQ(2, urbg.invocations()); in TEST()559 EXPECT_EQ(3, urbg.invocations()); in TEST()
67 where the corresponding bits are set for all active invocations71 that return true in some invocations and false in others. Perform73 across invocations. Verify returned value.80 passed as parameter. Final color should be same for all invocations -81 first invocation completed dictates result for other invocations.87 for all invocations - first invocation completed dictates result88 for other invocations.
35 shader invocations within one group to generate single result68 as a parameter should diverge between invocations. All pixels72 as a parameter should be true for all invocations. All pixels88 passed as a parameter should diverge between invocations.92 passed as a parameter should be true for all invocations.96 passed as a parameter should be false for all invocations.
24 Compute shaders operate on compute invocations in a workgroup.196 At each stage of the pipeline, multiple invocations of a shader may: execute198 Further, invocations of a single shader produced as the result of different200 The relative execution order of invocations of the same shader type is202 Shader invocations may: complete in a different order than that in which the207 The relative execution order of invocations of different shader types is210 pipeline stage, the shader invocations from the previous stage are221 fragment), even the number of shader invocations that may: perform loads and232 * The relative execution order of invocations of the same shader type is239 shader invocations are not.[all …]
61 large number of fragment shader invocations that perform loads and65 invocations is largely undefined. For algorithms that use shader76 invocations from touching the same memory concurrently.80 critical section of fragment shader code. For pairs of shader invocations97 guarantees that the critical section for multiple shader invocations with107 worrying about other invocations for the same pixel accessing the data156 ordering of the execution of shader invocations between calls to the168 By default, fragment shader invocations are generally executed in169 undefined order. Multiple fragment shader invocations may be executed170 concurrently, including multiple invocations corresponding to a single[all …]
128 A subgroup is a set of invocations that can synchronize and share data149 visible by other invocations within the subgroup. A _subgroupBarrier_ can156 The vote subgroup operations allow invocations within a subgroup to159 * Do all active subgroup invocations agree that an expression is true?160 * Do any active subgroup invocations evaluate an expression to true?161 * Do all active subgroup invocations have the same value of an expression?170 The arithmetic subgroup operations allow invocations to perform scan175 scan, cumulatively applying the operation across the invocations in a181 The ballot subgroup operations allow invocations to perform more183 all invocations within a subgroup to provide a boolean value and get[all …]
56 shader invocation until invocations from previous primitives that map to 59 fragment shader invocations are made visible to the fragment shader 97 invocations that overlap in xy window coordinates. Calling 99 transactions issued by shader invocations from previous primitives, 125 execution until completion of all shader invocations from previous 128 invocations mapped to same xy window coordinates are made visible to 131 regardless of whether previous and current invocations cover overlapping 134 shader invocations mapped to the same sample number under same window xy
3 invocations = -121 invocations = -142 invocations = 1
2 let invocations = 0; variable10 console.log(invocations++, c);