ip6: Drop dead code in ip6_input_hbh()
ClosedPublic
Actions

Authored by markj on Fri, May 29, 7:33 PM.

Details

Reviewers

thj
glebius
kp
tuexen
pouria
zlei

Group Reviewers

network

Commits

rGacf0be6e5192: ip6: Drop dead code in ip6_input_hbh()

Summary

After commit 069a67374ed9, ip6_input() quickly rejects packets with
plen == 0, before ip6_input_hbh() is called. So, there is no need to
check this condition again in the helper function.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 73608
Build 70491: arc lint + arc unit

Event Timeline

markj created this revision.Fri, May 29, 7:33 PM

Herald added a reviewer: pouria. · View Herald TranscriptFri, May 29, 7:33 PM

Herald added subscribers: melifaro, ae, imp. · View Herald Transcript

markj requested review of this revision.Fri, May 29, 7:33 PM

Harbormaster completed remote builds in B73557: Diff 178940.Fri, May 29, 7:33 PM

tuexen accepted this revision.Fri, May 29, 8:00 PM

This revision is now accepted and ready to land.Fri, May 29, 8:00 PM

bms added a subscriber: bms.Fri, May 29, 8:06 PM

bms added a comment.Fri, May 29, 8:48 PM

This comment was removed by bms.

pouria accepted this revision.Fri, May 29, 9:47 PM

zlei added a subscriber: zlei.Mon, Jun 1, 4:51 PM

zlei added inline comments.

sys/netinet6/ip6_input.c
434	Shall those stats be moved to `ip6_input()`, after the checking `if (plen == 0)` ?

markj added inline comments.Mon, Jun 1, 5:02 PM

sys/netinet6/ip6_input.c
434	`ifs6_in_discard` is handled already. We should increment `ifs6_in_hdrerr` too I think. Is `ip6s_badoptions` still correct? Probably it should be `ip6s_tooshort` instead?

increment more counters
add a branch annotation

This revision now requires review to proceed.Mon, Jun 1, 5:07 PM

Harbormaster completed remote builds in B73608: Diff 179066.Mon, Jun 1, 5:07 PM

Looks good to me.

sys/netinet6/ip6_input.c
434	The `ifs6_in_hdrerr` was to indicate a packet without `JumboPayload option`, since If the payload length field is 0 and the next header field indicates Hop-by-Hop Options header Since we do not support / check `JumboPayload option`, then `ip6s_tooshort` looks reasonable. CC the original author @bz to see if he has any thoughts about it.