Got asked an interesting question by @matt77hias about why enkiTS doesn't use std::future in the API and I think my answer might be worth sharing:https://github.com/dougbinks/enkiTS/issues/27 …
-
That can't happen in enkiTS, or we're misunderstanding each other (easy on twitter). I also test on 1 to hardware thread count threads.
You're right about twitter. Hope this helps. In the two scenarios, depending on which core picks up Task D, a deadlock *may* occur. Certainly, one a single core it's almost guaranteed. This is why wait-with-assist is multithreading code smell.pic.twitter.com/NijYw9eL1j
-
-
The solution here is to replace Waits with continuations. It's not always an easy change to make, but it's worth it. Especially timing (gpu/network/disc async calls?) start mixing things up even more.
-
The other solution is to reschedule the task, depending on how early the wait is in the task. Ideally you just don't schedule the task until it would not be blocked (execution graphs basically do this), but that's not always possible (ie, waiting on the GPU).
- Näytä vastaukset
Uusi keskustelu -
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.