Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 30 Oct 2013 12:10:41 +0400
From:      Gleb Smirnoff <glebius@FreeBSD.org>
To:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   Re: svn commit: r256512 - head/usr.bin/netstat
Message-ID:  <20131030081041.GU52889@FreeBSD.org>
In-Reply-To: <20131028204113.GA37879@caravan.chchile.org>
References:  <201310150955.r9F9t8pq084143@svn.freebsd.org> <20131028204113.GA37879@caravan.chchile.org>

next in thread | previous in thread | raw e-mail | index | archive | help
  Jeremie,

On Mon, Oct 28, 2013 at 09:41:13PM +0100, Jeremie Le Hen wrote:
J> On Tue, Oct 15, 2013 at 09:55:08AM +0000, Gleb Smirnoff wrote:
J> > Log:
J> >     Rewrite netstat/if.c to use getifaddrs(3) and getifmaddrs(3) instead of
J> >   libkvm digging in kernel memory. This is possible since r231506 made
J> >   getifaddrs(3) to supply if_data for each ifaddr.
J> >   
J> >     The pros of this change is that now netstat(1) doesn't know about kernel
J> >   struct ifnet and struct ifaddr. And these structs are about to change
J> >   significantly in head soon. New netstat binary will work well with 10.0
J> >   and any future kernel.
J> >   
J> >     The cons is that now it isn't possible to obtain interface statistics
J> >   from a vmcore.
J> 
J> Is there a kgdb macro or something like that that could be used for
J> debugging purpose?

No there is no such thing.

I've been debugging network related problems a lot, but I never had a need
to do 'netstat -i' on a vmcore. What we usually are interested at is
consistency of pointers and so on, not packets/bytes counters of an already
dead system.

Such a tool can be developed and put into src/tools, but I don't see enough
demand for it.

-- 
Totus tuus, Glebius.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20131030081041.GU52889>