Date: Thu, 6 Mar 1997 10:42:21 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: julian@whistle.com (Julian Elischer) Cc: Hackers@FreeBSD.ORG Subject: Re: Network gurus? in_rmx.c.. why? Message-ID: <199703061742.KAA13666@phaeton.artisoft.com> In-Reply-To: <331E65E2.41C67EA6@whistle.com> from "Julian Elischer" at Mar 5, 97 10:36:18 pm
next in thread | previous in thread | raw e-mail | index | archive | help
Well, I'm no Garrett, but I have an opinion... 8-). > As those of you who've been following things probably know.. > We at Whistle have a need for things such as interfaces > to come and go, and for routes to be cleanly destroyed when we > reassingn addresses here and there in a dynamic manner. Everyone needs this to work reliably, even if they never use the feature, IMO. > i.e. it assumes (and needs to assume) that only references > generated by rtalloc1() are counted.. In other words > rt_refcnt is only a count of references by open sessions. Why can't you turn this around: consider the other references to be sessions? Then the count of all references *is* the count of "sessions". This is sort of like what I've been suggesting for vnode reference counts for about threee years now, with the name cache entries counting as a "reference". > Am I right in thinking that the only reason in_rmx needs to know > when a route is not in use is so that it can keep the number of > kernel routes in a place like wcarchive under control? This feels like a flag, not a reference, to me. Regards, Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199703061742.KAA13666>