From: Pablo Neira Ayuso <pablo@netfilter.org>
To: Phil Sutter <phil@nwl.cc>, netfilter-devel@vger.kernel.org
Subject: Re: [iptables PATCH v2 00/17] Eliminate dedicated arptables-nft parser
Date: Fri, 15 Oct 2021 13:25:29 +0200 [thread overview]
Message-ID: <YWllKW6PmE9FJHhn@salvia> (raw)
In-Reply-To: <20211015110124.GL1668@orbyte.nwl.cc>
On Fri, Oct 15, 2021 at 01:01:24PM +0200, Phil Sutter wrote:
> Hi Pablo,
>
> On Thu, Oct 14, 2021 at 10:56:09PM +0200, Pablo Neira Ayuso wrote:
> > On Thu, Sep 30, 2021 at 04:04:02PM +0200, Phil Sutter wrote:
> > > Commandline parsing was widely identical with iptables and ip6tables.
> > > This series adds the necessary code-changes to unify the parsers into a
> > > common one.
> > >
> > > Changes since v1:
> > > - Fix patch 12, the parser has to check existence of proto_parse
> > > callback before dereferencing it. Otherwise arptables-nft segfaults if
> > > '-p' option is given.
> >
> > LGTM.
>
> Thanks for your review!
>
> > > - Patches 13-17 add all the arptables quirks to restore compatibility
> > > with arptables-legacy. I didn't consider them important enough to push
> > > them unless someone complains. Yet breaking existing scripts is bad
> > > indeed. Please consider them RFC: If you consider (one of) them not
> > > important, please NACk and I will drop them before pushing.
> >
> > For patch 13-16, you could display a warning for people to fix their
> > scripts, so this particular (strange) behaviour in some cases can be
> > dropped (at least, 13-15 look like left-over/bugs). For the
> > check_inverse logic, I'd suggest to display a warning too, this is
> > what it was done in iptables time ago to address this inconsistency.
>
> I wonder how likely it is for someone to rely upon the behaviour. I can
> imagine a script passing an empty interface name and expecting the
> argument to be ignored (patch 13). Though what are the odds someone
> actually calls arptables with '-m something' (patch 14) or a bogus table
> name (patch 15)?
I think this is a leftover/bug that has been there for long time.
> > I'd probably keep back patch 17/17, the max chain name length was
> > reduced by when the revision field was introduced and this resulted in
> > no issue being reported.
>
> If that's OK with you, I would turn the empty interface name error into
> a warning for arptables-nft, reintroduce the warning for intrapositioned
> negations and drop the remaining quirks as they are likely hiding a bug.
Agreed.
prev parent reply other threads:[~2021-10-15 11:25 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-30 14:04 [iptables PATCH v2 00/17] Eliminate dedicated arptables-nft parser Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 01/17] nft: Introduce builtin_tables_lookup() Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 02/17] xshared: Store optstring in xtables_globals Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 03/17] nft-shared: Introduce init_cs family ops callback Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 04/17] xtables: Simplify addr_mask freeing Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 05/17] nft: Add family ops callbacks wrapping different nft_cmd_* functions Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 06/17] xtables-standalone: Drop version number from init errors Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 07/17] libxtables: Introduce xtables_globals print_help callback Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 08/17] arptables: Use standard data structures when parsing Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 09/17] nft-arp: Introduce post_parse callback Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 10/17] nft-shared: Make nft_check_xt_legacy() family agnostic Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 11/17] xtables: Derive xtables_globals from family Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 12/17] nft: Merge xtables-arp-standalone.c into xtables-standalone.c Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 13/17] xtables: arptables doesn't warn about empty interface Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 14/17] xtables: arptables accepts but ignores '-m' Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 15/17] xtables: arptables ignores wrong -t values Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 16/17] xtables: Support '!' betwen option and argument Phil Sutter
2021-09-30 14:04 ` [iptables PATCH v2 17/17] nft: Store maximum allowed chain name length in family ops Phil Sutter
2021-10-14 20:56 ` [iptables PATCH v2 00/17] Eliminate dedicated arptables-nft parser Pablo Neira Ayuso
2021-10-15 11:01 ` Phil Sutter
2021-10-15 11:25 ` Pablo Neira Ayuso [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YWllKW6PmE9FJHhn@salvia \
--to=pablo@netfilter.org \
--cc=netfilter-devel@vger.kernel.org \
--cc=phil@nwl.cc \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).