Attention is currently required from: arehbein, fixeria, osmith, pespin.

daniel has posted comments on this change. ( https://gerrit.osmocom.org/c/libosmo-netif/+/33198?usp=email )

Change subject: stream: Add client-side (segmentation) support for IPA ......................................................................

Patch Set 21:

(1 comment)

Patchset:

PS21:

I guess printing the timestamp infront of the log should just be removed? from a quick glance this s […]

I just retriggered jenkins and it succeeds again... The tests uses osmo's gettimeofday_override functionality so it shouldn't cause these weird changes due to (wall clock) time delays. It's counting main loop iterations and the only way a mismatch could happen here (that I can think of) is if either the sending fd wasn't writable after enqueuing the msg or the receiving fd didn't become readable directly after the write.

So it's definitely worrying that we don't know the cause for this issue, but since nobody can reproduce it right now I'd say we continue with this patch (with timestamps) and see if it resurfaces.