Details

Reviewers

emaste
allanjude
wblock

Commits

rS323390: If the user tries to set kern.randompid to 1 (which is meaningless), set
rS312307: Improve some of the sysctl descriptions added in r299827.

Summary

PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=205879

Add descriptions to most sysctl's in the net.inet6 namespace which doesn't already have one. I used the inet6(4) man page as source.

Diff Detail

Repository

rS FreeBSD src repository - subversion

Lint

Lint Not Applicable

Unit

Tests Not Applicable

Event Timeline

marieheleneka_gmail.com updated this revision to Diff 13449.Feb 18 2016, 11:29 PM

marieheleneka_gmail.com retitled this revision from to Add missing descriptions in sysctl net.inet6 namespace.

marieheleneka_gmail.com updated this object.

marieheleneka_gmail.com edited the test plan for this revision. (Show Details)

marieheleneka_gmail.com added a reviewer: allanjude.

marieheleneka_gmail.com set the repository for this revision to rS FreeBSD src repository - subversion.

Herald added a subscriber: imp. · View Herald TranscriptFeb 18 2016, 11:29 PM

marieheleneka_gmail.com added a subscriber: wblock.Feb 18 2016, 11:41 PM

emaste added a subscriber: emaste.Feb 19 2016, 1:04 AM

wblock added inline comments.Feb 19 2016, 3:44 PM

sys/netinet6/in6_proto.c
534 ↗	(On Diff #13449)	Default number of fragmented packets
535 ↗	(On Diff #13449)	s/setting/value/
536 ↗	(On Diff #13449)	s/setting/value/
560 ↗	(On Diff #13449)	Default maximum number of
574 ↗	(On Diff #13449)	Passive -> active: s/Identifies/Identify/ "this nodes KAME" is actually a possessive: this node's KAME
581 ↗	(On Diff #13449)	"IPv6 router renumbering prefix babysitting" is a lot of adjectives in a row. It's kind of hard to find the noun. Would reordering help clarify? "Default interval in seconds for prefix babysitting of IPv6 router renumbering."?
595 ↗	(On Diff #13449)	The "whether" wording is kind of confusing, but I'm not sure what would be clearer. "Determine whether" might be a little better. Avoid contractions, and clear up a little confusion over who "they" refers to: A value of 0 means sockets listen on both IPv6 and IPv4 addresses.

Thanks for this!

A couple of notes:

descriptions conventionally do not include a period
I think it makes sense to keep the descriptions similar to the IPv4 ones where they have the same meaning -- e.g. Enable IP forwarding between interfaces vs Enable routing/forwarding of IPv6 packets

Updated revision addressing inline comments.
Moved Warren Block and Ed Maste from Subscribers to Reviewers.

wblock added inline comments.Feb 19 2016, 11:18 PM

sys/netinet6/in6_proto.c
595 ↗	(On Diff #13513)	Was the removal of the second sentence intentional? If not, I think this is better than the earlier version I suggested: When 0, sockets listen on both IPv6 and IPv4 addresses.

marieheleneka_gmail.com added inline comments.Feb 19 2016, 11:29 PM

sys/netinet6/in6_proto.c
595 ↗	(On Diff #13513)	It was intentional, but I marked my comment about it as done. Doh. Quoting: "I'll drop the second sentence here, as it's implicated by the first sentence AND documented in the man page." s/implicated/implied :)
581 ↗	(On Diff #13449)	This is a tough one, but I think "Default interval in seconds for babysitting of IPv6 router prefix renumbering." should do the trick.
595 ↗	(On Diff #13449)	I'll drop the second sentence here, as it's implicated by the first sentence AND documented in the man page.

Works for me.

sys/netinet6/in6_proto.c
595 ↗	(On Diff #13513)	Phabricator conscientiously hid that comment from me, apparently. :)

This revision is now accepted and ready to land.Feb 20 2016, 1:41 AM

Please do not commit this yet. Out of a couple thousand sysctl's, only about 130 had a description ending with a period. Although I did not filter out the ones without a description, I still think this is telling.

Therefore, I'll remove the ending periods on these descriptions.

sysctl MIB descriptions now end without punctuation, to be in line with how this is generally done. Also removed ending period of some existing descriptions.

This revision now requires review to proceed.Mar 1 2016, 12:39 PM

wblock added inline comments.Apr 6 2016, 2:47 PM

sys/netinet6/in6_proto.c
526 ↗	(On Diff #13930)	When there is actually more there one complete sentence in a comment, adding a final ending period seems okay. There might be a guideline on how these should be written. Maybe they are supposed to be expressed as a single short description. Or maybe we should create one. This section could be rewritten as: Enable sending ICMPv6 redirects for unforwardable IPv6 packets, ignored unless the node is routing IPv6
561 ↗	(On Diff #13930)	As above, this can just be " incoming IPv6 packets, 0 for no artificial limit");

emaste added inline comments.Apr 6 2016, 2:53 PM

sys/netinet6/in6_proto.c
526 ↗	(On Diff #13930)	There might be a guideline on how these should be written. Maybe they are supposed to be expressed as a single short description. Or maybe we should create one. I don't think it's written down anywhere, but probably should be. My impression is that they should be a single short description, and in fact the "ignored unless" may well be superfluous. Perhaps this could just be `send ICMPv6 redirects for unforwardable IPv6 packets when routing IPv6`