This repository has been archived by the owner on Nov 1, 2023. It is now read-only.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The manual implementation of
Stream
forDirectoryMonitor
misused theWaker
provided bypoll_next()
. In practice, this resulted in telling the executor to poll the stream again immediately after any non-terminaltry_recv()
. The end result was equivalent to callingtry_recv()
in a tight loop. This pegged agent CPU usage as soon as we started monitoring directories for new files.Now, we take a much simpler approach. The private
into_async()
shim function spawns a thread that blocks on new (synchronous)Receiver
messages, and then immediately sends them to an unbounded asyncReceiver
. All subsequent API surface can thus be effortlessly async.DirectoryMonitor
now exports a publicnext_file()
async method, which is equivalent in practice to implementingStream
. We aren't using anyStream
combinators, we don't lose anything in doing this. We can also use this to implStream
again later, if needed.