ANNOUNCEMENT: Live Wireshark University & Allegro Packets online APAC Wireshark Training Session
April 17th, 2024 | 14:30-16:00 SGT (UTC+8) | Online

Wireshark-dev: Re: [Wireshark-dev] [Wireshark-commits] rev 44485: /trunk/epan/ /trunk/epan/: li

From: Graham Bloice <graham.bloice@xxxxxxxxxxxxx>
Date: Tue, 14 Aug 2012 17:32:24 +0100
> -----Original Message-----
> From: wireshark-dev-bounces@xxxxxxxxxxxxx [mailto:wireshark-dev-
> bounces@xxxxxxxxxxxxx] On Behalf Of Guy Harris
> Sent: 14 August 2012 17:27
> To: wireshark-dev@xxxxxxxxxxxxx
> Subject: Re: [Wireshark-dev] [Wireshark-commits] rev 44485: /trunk/epan/
> /trunk/epan/: libwireshark.def
>
>
> On Aug 14, 2012, at 4:24 AM, grahamb@xxxxxxxxxxxxx wrote:
>
> >
> http://anonsvn.wireshark.org/viewvc/viewvc.cgi?view=rev&revision=44485
> >
> > User: grahamb
> > Date: 2012/08/14 04:24 AM
> >
> > Log:
> > Added proto_tree_free to the libwireshark expoert definitions for use
in
> plugins
>
> What plugin would have a need for it?  Freeing the protocol tree filled
up by a
> dissection done elsewhere would be considered a bit impolite, as it
might
> pull the tree out from under somebody using it.  If the plugin has done
its
> own top-level dissection (which seems a bit odd - dissectors shouldn't
be
> doing that, taps shouldn't need to do that as the tapping should be
doing the
> dissection for it, file format plugins are at a level below libwireshark
and
> shouldn't even know what a protocol tree *is*, and codecs are orthogonal
to
> dissection), it should have done it with the epan_ APIs and thus should
free it
> with a call to epan_dissect_cleanup() or epan_dissect_free().

I did wonder, but someone asked for it.  Easily removed if we think that
having it there will get folks into trouble.