Implement S/G IO for batched sends and eliminate another frame copy #2874
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Outdated suggestions cannot be applied.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.
Description
This is the final part of #2867 which implements scatter/gather IO for
platf::send_batch()
and eliminates the frame copy infec::encode()
by FEC encoding and sending directly out of the payload buffer. This is done by introducing support for payload buffer lists containing multiple contiguous payloads buffers (necessary because data and parity shards are in different buffers). It also introduces the ability to begin sending messages from a specified offset which avoids having to adjust the buffer lists when we send a FEC block in multiple parts for frame pacing.Testing checklist:
sendmsg()
+ encryptionsendmmsg()
+ encryptionsendmsg()
+ no encryptionsendmmsg()
+ no encryptionWSASendMsg()
+ encryptionWSASendMsg()
+ no encryptionScreenshot
Issues Fixed or Closed
Type of Change
.github/...
)Checklist
Branch Updates
LizardByte requires that branches be up-to-date before merging. This means that after any PR is merged, this branch
must be updated before it can be merged. You must also
Allow edits from maintainers.