You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Given that blocking in k23's scheduler is even less desirable that is the case for "hosted" async runtimes like Tokio we should take all steps to prevent tasks from blocking for too long. To this end we should implement a mechanism like tokio's cooperative scheduling budget where tasks periodically decrease a "budget" counter and when it is exhausted yield back to the runtime. This would be virtually identical to the epoch based system used by the WASM engine but for kernel tasks and futures.
The text was updated successfully, but these errors were encountered:
Given that blocking in k23's scheduler is even less desirable that is the case for "hosted" async runtimes like Tokio we should take all steps to prevent tasks from blocking for too long. To this end we should implement a mechanism like tokio's cooperative scheduling budget where tasks periodically decrease a "budget" counter and when it is exhausted yield back to the runtime. This would be virtually identical to the epoch based system used by the WASM engine but for kernel tasks and futures.
The text was updated successfully, but these errors were encountered: