SetTimestampDelta overflow issue?

[tbarbette]

When transforming a trace, I have observed that SetTimestampDelta messes up the timing using a pcap with nano timestamp.

https://github.com/tbarbette/fastclick/blob/e5260ccffd4822bb3e21208c029d3504abf5b342/elements/analysis/settimestampdelta.cc#L61

One should create test cases and try to reproduce, then fix :) Good first issue !

[XueSongTap]

Can I work on this ？

[tbarbette]

I would be happy if you do :)

[XueSongTap]

Hi, @tbarbette

I tried crafting a nano‑timestamp PCAP with several 1 ns‑spaced packets. With the current code I only see all nanos being cut to µs (so every delta prints 0).

The bit that feels suspicious is the plain tv -= _first; line in SetTimestampDelta: when the underlying Timestamp is micro‑second‑based, any nano input has already been flattened before this subtraction.

Does that match what you’re seeing, or is there another code path that actually scrambles the order/values?

Thanks.

[tbarbette]

There's a configure option, --enable-nanotimestamp, is that activated?