Skip to content

Conversation

@lenary
Copy link
Member

@lenary lenary commented Jul 11, 2025

This commit lines up the implementation fo doRelaxationQCCall with the implementation of relaxations for qc.e.li/qc.li.

The implementation:

  • First checks which relaxations are possible, before trying any.
  • Then applies the possible relaxations in a priority order, starting with the relaxation with the smallest result, and working to relaxations with larger results.

This structure makes remarking about missed relaxations a lot easier, as those messages are only emitted once the relaxation being performed has been chosen, so we also know how many bytes have been saved.

The only change is the order of some error messages, and slightly higher stats about missed relaxations because now RelaxationBytesDeleted + RelaxationBytesMissed should stay constant despite changes to relaxation control flags if any relaxations are enabled.

This commit lines up the implementation fo `doRelaxationQCCall` with the
implementation of relaxations for `qc.e.li`/`qc.li`.

The implementation:
- First checks which relaxations are possible, before trying any.
- Then applies the possible relaxations in a priority order, starting
  with the relaxation with the smallest result, and working to
  relaxations with larger results.

This structure makes remarking about missed relaxations a lot easier, as
those messages are only emitted once the relaxation being performed has
been chosen, so we also know how many bytes have been saved.

The only change is the order of some error messages, and slightly higher
stats about missed relaxations because now `RelaxationBytesDeleted +
RelaxationBytesMissed` should stay constant despite changes to
relaxation control flags if any relaxations are enabled.

Signed-off-by: Sam Elliott <quic_aelliott@quicinc.com>
@lenary
Copy link
Member Author

lenary commented Jul 11, 2025

(I split these changes out of #201 while debugging, but they are not the problem)

Copy link
Contributor

@quic-seaswara quic-seaswara left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

is this PR still needed ?

Can we do the required testing and we can merge this ?

@lenary
Copy link
Member Author

lenary commented Oct 10, 2025

I'll rebase this. It doesn't change the functionality, just some debug messages. Will make sure it is tested.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants