Subject: Re: CVS commit: src/sbin/sysctl
To: Matt Thomas <matt@3am-software.com>
From: Andrew Brown <atatat@atatdot.net>
List: source-changes
Date: 01/22/2004 19:10:25
>> >> >Modified Files:
>> >> >       src/sbin/sysctl: sysctl.8
>> >> >
>> >> >Log Message:
>> >> >Document net.bpf.maxbufsize in sysctl(8).
>> >> >NB: bpf isn't a PF_, so where to list it in sysctl(3)?
>> >>
>> >>hmm...good point.  my first thought was that bpf had to do with
>> >>networking under the "net" sub-tree, but perhaps it fits better with
>> >>devices under the "dev" sub-tree?
>> >
>> >bpf should be use pseudo_AF_HDRCMPLT as it's address family since
>> >that exists mostly for bpf's usage.  Calling that net.bpf seems fine
>> >to me.
>>
>>alternately, put bpf under net.unspec (CTL_NET.PF_UNSPEC)?
>
>Ewww.  Unspec is a wildcard.  Don't overload it.

i was thinking more about how bpf doesn't apply to any particular
protocol family, but operates on all of them.  so to speak.

>CTL_NET.pseudo_AF_HDRCMPLT is the right choice for bpf.

pseudo_PF_HDRCMPLT perhaps (thought we'd have to add it)?

-- 
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org             * "ah!  i see you have the internet
twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
werdna@squooshy.com       * "information is power -- share the wealth."