Lars Eggert
Lars Eggert
Yes; this is possible if the connections use non-zero connection IDs. See RFC9000.
Initial QUIC packets have a unique packet type that identifies them.
Should you bump the crate version?
+1 that something like `memchr_inv` would be great for efficiently skipping over sequences of zeros (which in my use case - QUIC - are padding frames in a packet).
Yup. Cc @rdanyliw and @ghwood
Now that this got fixed, would you mind cutting a release?
There also seems to be an issue with passing `--locked` to nextest: https://github.com/mozilla/neqo/actions/runs/14932174735/job/41950945459?pr=2607#step:7:36
It'd be great to land this, so https://github.com/mozilla/mtu/pull/29 would work on the BSDs.
Bench on `main`: ``` test large_data_10_streams ... bench: 27,558,791 ns/iter (+/- 13,459,810) = 380 MB/s test large_data_1_stream ... bench: 24,324,266 ns/iter (+/- 19,219,937) = 43 MB/s test small_data_100_streams ... bench:...
We've also found on neqo that multi-packet RX without multi-packet TX has limited benefits, since the RX batch size will be very small.