| ------------------------------------------------------------------------- |
| drawElements Quality Program Test Specification |
| ----------------------------------------------- |
| |
| Copyright 2014 The Android Open Source Project |
| |
| Licensed under the Apache License, Version 2.0 (the "License"); |
| you may not use this file except in compliance with the License. |
| You may obtain a copy of the License at |
| |
| http://www.apache.org/licenses/LICENSE-2.0 |
| |
| Unless required by applicable law or agreed to in writing, software |
| distributed under the License is distributed on an "AS IS" BASIS, |
| WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| See the License for the specific language governing permissions and |
| limitations under the License. |
| ------------------------------------------------------------------------- |
| Shader Storage / Image Storage synchronization tests |
| |
| Tests: |
| + dEQP-GLES31.functional.synchronization.* |
| |
| Includes: |
| |
| + In-invocation synchronization |
| - Test synchronization primitives within a single invocation. |
| |
| + Inter-invocation synchronization |
| - Test synchronization between work group invocations. |
| |
| + Inter-call synchronization |
| - Test synchronization between draw calls |
| |
| Excludes: |
| |
| - Inter-invocation synchronization without use of barrier() |
| - Use vertex, fragment, or geometry shaders. |
| |
| Description: |
| |
| In-invocation synchronization tests read and write to a coherent image or |
| buffer within a single invocation. Data accessed by a single invocation is not |
| accessed by other invocations. Image accesses are synchronized using |
| memoryBarrierImage(). SSBO accesses do not need to be explicitly synchronized |
| as memory operations execute in order. |
| |
| Inter-invocation synchronization tests read and write to a coherent image or |
| buffer within a single work group. Data accessed by a single invocation is also |
| accessed by another invocation within the work group. Accesses are synchronized |
| using a combination of groupMemoryBarrier() and barrier(). |
| |
| inter_call.with_memory_barrier test cases read and write to an image or buffer. |
| Data written in a single call is read in another call. glMemoryBarrier() is used |
| for synchronization between the calls. |
| |
| inter_call.without_memory_barrier.* inter-call synchronization tests modify a image |
| or buffer declared as volatile with atomic operations or using atomic counters. |
| Multiple computes are dispatched without calling MemoryBarrier(). Result and |
| intermediate results are checked to verify operation atomicy over multiple compute |
| dispatches. |
| |
| inter_call.without_memory_barrier.ssbo_atomic_counter_mixed_* cases test relative |
| atomicy of buffer atomic operations and atomic counter operations. Separate programs |
| modify atomically either a buffer or an atomic counter backed by the same buffer. |
| Multiple instances of programs are dispatched without calling MemoryBarrrier(). Buffer |
| content and intermediate values are checked to verify atomicy between operations. |