mirror of
https://github.com/gfx-rs/wgpu.git
synced 2026-04-22 03:02:01 -04:00
333ed78529
We are using `program_cache.try_lock()` when creting pipelines which is covered by a guard gotten from `context.lock()`. For the `.try_lock()` to always succeed we need to make sure that the other lock acquisitions are also covered by a `context.lock()`. The `wgpu_examples::hello_compute::tests::multithreaded_compute` test has been failing intermittently in CI due to this.