⚡️zb: don't clone signature when creating body

[TTWNO]

We saw performance uplifts in the range of 10-15% in parsing events just by not cloning, reparsing the signature every time a body was created.

Hopefully it's a small enough change that it's easy to review.

This directly reduces the cost of calling Body::new(...) by 90-95%.

[TTWNO]

We saw performance uplifts in the range of 10-15% in parsing events just by not cloning, reparsing the signature every time a body was created.

Hopefully it's a small enough change that it's easy to review.

This directly reduces the cost of calling Body::new(...) by 90-95%.

[zeenix]

We saw performance uplifts in the range of 10-15% in parsing events just by not cloning, reparsing the signature every time a body was created.

Hopefully it's a small enough change that it's easy to review.

This directly reduces the cost of calling Body::new(...) by 90-95%.

Wow, cool. Could you please:

1. put this in the description of the PR and commit?
2. add a benchmark for this (unless it's covered by the existing ones?)?

[TTWNO]

Added benchmarks. It's measured in nanoseconds so don't worry too much if there is a bit of variation here and there. But if it jumps into the hundreds of ns... then there might be a problem.

[zeenix]

Added benchmarks.

Awesome! I'll move the commit order so it's possible to see the change. :)

It's measured in nanoseconds so don't worry too much if there is a bit of variation here and there. But if it jumps into the hundreds of ns... then there might be a problem.

Right, we already have a few benches that are in ns so I'm familiar with the drill. :)

[zeenix]

@TTWNO I took the liberty of merging the new benchmark into the existing msg-de one and editting the commit log accordingly. I hope you don't mind. :)

[TTWNO]

All good by me. Thanks!