[Akka.Actor][Perf] Reduce Envelope
copying in message-processing hotpaths
#6830
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Outdated suggestions cannot be applied.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.
Changes
Depends on #6829
While working on some performance fixes for Phobos we noticed some significant improvements by passing all
struct
s handled inside a tight loop asin
parameters into subsequent methods - this helps reduce message copying by allowing the "lower stack" to reference the memory previously allocated in the "upper stack"Since this handling is all synchronous, there's no cost to using
in
parameters in these areas - and we saw a significant upside in processing performance, as much as 10% in a single instance so far.Based on that work, I decided to give that a try with how we handle the
Envelope
payloads inside Akka.Actor - as these are alsostruct
s that are passed around in tight loops.This PR introduces some API changes but these should all be binary compatible, minus any libraries that intentionally override these methods - those will need to be updated to also include the
in
parameter on those methods.Checklist
For significant changes, please ensure that the following have been completed (delete if not relevant):
Latest
dev
BenchmarksFrom the
PingPong
benchmark inside Akka.Benchmarks:This PR's Benchmarks
IMHO - kind of looks like the benefits of this fall within the margin of error.