projects / platform / upstream / VK-GL-CTS.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 29759e0) | patch
Fix multithreaded object counting in internally_synchronized_objects.*
authorJeff Bolz <jbolz@nvidia.com>
Fri, 11 Feb 2022 03:40:03 +0000 (21:40 -0600)
committerJeff Bolz <jbolz@nvidia.com>
Fri, 11 Feb 2022 03:42:40 +0000 (21:42 -0600)
commitf68697a26681082d50655ee349000093fbe1cc30
treec93aea10f1629740725a6be8a01711579e48c10etree | snapshot
parent29759e0fca7966a1cc73875c5b0c55594c2396d9commit | diff
Fix multithreaded object counting in internally_synchronized_objects.*

Copy the barrier code from vktApiObjectManagementTests.cpp and do a
barrier on each iteration in the main process. This guarantees that
CTS will reserve the maximum number of pipeline pool slots we might
need.

Component: VulkanSC

Affects tests: dEQP-VKSC.synchronization.internally_synchronized_objects.*

Change-Id: Icf2ac519076c80803adb3833814e4ab290761105
external/vulkancts/modules/vulkan/synchronization/vktSynchronizationInternallySynchronizedObjectsTests.cpp diff | blob | history
Domain: Graphics System / GL;