Strange thing with iptables

Strange thing with iptables

[Szabolcs Gyurko]

Hi all,


I found a strange issue with the iptables. One of my friend showed me a  
rule
which looked like:

iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT


What I was surprised on is the netmask. Is this a feature or a bug? I mean
this is quite strange netmask for me.

Thanks for any answers,

-- 
Szabolcs Gyurko

Re: Strange thing with iptables

[Martin Josefsson]

[-- Attachment #1: Type: text/plain, Size: 522 bytes --]

On Thu, 2004-09-09 at 13:49, Szabolcs Gyurko wrote:
> Hi all,
> 
> 
> I found a strange issue with the iptables. One of my friend showed me a  
> rule
> which looked like:
> 
> iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT
> 
> 
> What I was surprised on is the netmask. Is this a feature or a bug? I mean
> this is quite strange netmask for me.

It's a feature :)
It doesn't make the current code any more complicated.
And ther are actually people using it to do weird stuff...

-- 
/Martin

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Strange thing with iptables

[Alexey Toptygin]

On Thu, 9 Sep 2004, Martin Josefsson wrote:

>> which looked like:
>>
>> iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT
>>
>>
>> What I was surprised on is the netmask. Is this a feature or a bug? I mean
>> this is quite strange netmask for me.
>
> It's a feature :)
> It doesn't make the current code any more complicated.
> And ther are actually people using it to do weird stuff...

Do you mean that one can use arbitrary bitmasks wherever netfilter wants a 
netmask value?
So, one might select all IPs with the LSB set with 0.0.0.1/0.0.0.1?

 			Alexey

Re: Strange thing with iptables

[Szabolcs Gyurko]

Sure. That is what seems to me an absolute discarding of the ipv4 rules.
But it's a feature, so...


On Thu, 9 Sep 2004 14:33:06 +0000 (UTC), Alexey Toptygin  
<alexeyt@freeshell.org> wrote:

> On Thu, 9 Sep 2004, Martin Josefsson wrote:
>
>>> which looked like:
>>>
>>> iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT
>>>
>>>
>>> What I was surprised on is the netmask. Is this a feature or a bug? I  
>>> mean
>>> this is quite strange netmask for me.
>>
>> It's a feature :)
>> It doesn't make the current code any more complicated.
>> And ther are actually people using it to do weird stuff...
>
> Do you mean that one can use arbitrary bitmasks wherever netfilter wants  
> a netmask value?
> So, one might select all IPs with the LSB set with 0.0.0.1/0.0.0.1?
>
>  			Alexey
>



-- 
Szabolcs Gyurko

Re: Strange thing with iptables

[Martin Josefsson]

[-- Attachment #1: Type: text/plain, Size: 510 bytes --]

On Thu, 2004-09-09 at 16:33, Alexey Toptygin wrote:

> > It's a feature :)
> > It doesn't make the current code any more complicated.
> > And ther are actually people using it to do weird stuff...
> 
> Do you mean that one can use arbitrary bitmasks wherever netfilter wants a 
> netmask value?
> So, one might select all IPs with the LSB set with 0.0.0.1/0.0.0.1?

Yes you can.
Although this might change in the (distant) future when diffrent
algorithms are used for rule-lookup.

-- 
/Martin

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Strange thing with iptables

[Jozsef Kadlecsik]

On Thu, 9 Sep 2004, Szabolcs Gyurko wrote:

> Sure. That is what seems to me an absolute discarding of the ipv4 rules.
> But it's a feature, so...

Search google for "netmask 255.255.0.255" and read the first two hits
(netfilter archive from 2001 October :-).
(And do not top-post, please :-().

> On Thu, 9 Sep 2004 14:33:06 +0000 (UTC), Alexey Toptygin
> <alexeyt@freeshell.org> wrote:
>
> > On Thu, 9 Sep 2004, Martin Josefsson wrote:
> >
> >>> which looked like:
> >>>
> >>> iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT
> >>>
> >>>
> >>> What I was surprised on is the netmask. Is this a feature or a bug? I
> >>> mean
> >>> this is quite strange netmask for me.
> >>
> >> It's a feature :)
> >> It doesn't make the current code any more complicated.
> >> And ther are actually people using it to do weird stuff...
> >
> > Do you mean that one can use arbitrary bitmasks wherever netfilter wants
> > a netmask value?
> > So, one might select all IPs with the LSB set with 0.0.0.1/0.0.0.1?

Best regards,
Jozsef
-
E-mail  : kadlec@blackhole.kfki.hu, kadlec@sunserv.kfki.hu
PGP key : http://www.kfki.hu/~kadlec/pgp_public_key.txt
Address : KFKI Research Institute for Particle and Nuclear Physics
          H-1525 Budapest 114, POB. 49, Hungary

Re: Strange thing with iptables

[Willy Tarreau]

Hi,

On Thu, Sep 09, 2004 at 02:18:15PM +0200, Martin Josefsson wrote:
> > 
> > iptables -A FORWARD -s $machine/255.255.0.255 -j ACCEPT
> > 
> > 
> > What I was surprised on is the netmask. Is this a feature or a bug? I mean
> > this is quite strange netmask for me.
> 
> It's a feature :)
> It doesn't make the current code any more complicated.
> And ther are actually people using it to do weird stuff...

I second this. I actually had to use the same principle on some equipment
(alteon) which also supports this, and it saved me a lot of filters when
writing anti-spoofing rules on a port where two IP networks coexist.

Cheers,
Willy