ospf6d: revert PR #8622

[eqvinox]

Revert #8622; see comments there.

While the revert doesn't fix this, it backs out code that isn't part of the correct solution. Also I'd rather crash one router than (possibly) break the network.

[LabN-CI]

💚 Basic BGPD CI results: SUCCESS, 0 tests failed

Results table

_	_
Result	SUCCESS git merge/8944 0c3a104
Date	06/29/2021
Start	11:02:38
Finish	11:28:07
Run-Time	25:29
Total	1815
Pass	1815
Fail	0
Valgrind-Errors
Valgrind-Loss
Details	vncregress-2021-06-29-11:02:38.txt
Log	autoscript-2021-06-29-11:03:46.log.bz2
Memory	515 517 413

For details, please contact louberger

[idryzhov]

It looks like after the recent OSPFv3 socket rework, ospf6_make_lsupdate_interface does the right thing and sends multiple packets when there's not enough space in a single packet. There's a probability that this revert won't even reintroduce the crash.

[eqvinox]

It looks like after the recent OSPFv3 socket rework, ospf6_make_lsupdate_interface does the right thing and sends multiple packets when there's not enough space in a single packet. There's a probability that this revert won't even reintroduce the crash.

That'd be great, I didn't dig that far… We still need a 64k limit tough, for seriously degenerate cases. And some tests... (for large LSAs generally, whereever they come from)

[also, CI hung due to a server crash, it's rerunning now]

[NetDEF-CI]

Continuous Integration Result: SUCCESSFUL

Congratulations, this patch passed basic tests

Tested-by: NetDEF / OpenSourceRouting.org CI System

CI System Testrun URL: https://ci1.netdef.org/browse/FRR-FRRPULLREQ-19908/

This is a comment from an automated CI system.
For questions and feedback in regards to this CI system, please feel free to email
Martin Winter - mwinter (at) opensourcerouting.org.