#1594

Tests: synchronize Vulkan occlusion results before CPU readback

Openhosienzade45wants to merge intofrom

Claimed in #1548. Found during the investigation of #1540.

What and why

Both occlusion tests read their query results from a mapped buffer, but neither marked that buffer host_readable. That skipped the replay queue’s transfer-write → host-read barrier. A queue-idle wait completes execution; Vulkan’s fence synchronization note also calls for a memory dependency to make writes visible to the host.

Set host_readable in both test bindings so they use the runtime’s existing readback path.

Tested

  • vk_overlapping_occlusion and vk_binary_occlusion both passed on Linux llvmpipe with Mesa 25.0.7 and 26.1.6, with synchronization validation enabled.
  • A local vkCmdPipelineBarrier2 trace on Mesa 26.1.6 measured 0 before / 1 after transfer-to-host-read barriers in the first overlapping-occlusion submission. All pixel counts and sentinels agreed in both runs.
  • The reported Windows Radeon timeout did not reproduce on llvmpipe. Its cause remains unconfirmed.
  • git diff --check passed.

No game data. AI-assisted implementation and testing.