Re: Future of CONNMARK (was Re: MASQUERADE: Route sent us somewhere else (was Re: Fw: Rusty's brain broke!)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, 18 Jan 2004, Henrik Nordstrom wrote:

> On Sat, 17 Jan 2004, Tom Eastep wrote:
>
> > I am very much in favor of the change you propose. The ability to set
> > individual bits would allow Netfilter configuration tools like Shorewall to
> > make internal use of packet marking by reserving part of the mark field for
> > use by the tool and the remainder of the field for use by the user.
> >
> > Given that the current MARK target lacks this capability, I am not able to
> > make effective use of that target in Shorewall.
>
> Please note that we are talkning about the CONNMARK target which is quite
> different from MARK. The two operates on different values. The discussed
> change will NOT add mask operations to the standard MARK target.
>

I realize that. I was simply stating that any type of packet/connection
marking facility that doesn't support the setting of individual bits
is of limited use to higher-level tools.

> If you need mask operations in the standard MARK target then nothing stops
> you from writing an extended MARK target having mask operations. It is
> just that it can not be done easily in the standard kernel due to binary
> compatibility issues.
>
> Why there is not a extra class p-o-m patch to add mask capability to MARK
> I do not know. I am pretty sure the netfilter team would not mind if such
> patch is submitted, but as indicated above it can not progress beyond
> "extra" due to the frozen nature of the existing MARK target, at least not
> unless a different target name is used.
>

I have strongly resisted making anything in Shorewall dependent on p-o-m
features. Supporting Shorewall takes enough of my time without having to
hand-hold newbies through kernel and iptables builds and installs.

> It is a pity the iptables match/target interface does not have versioning
> support of the target/match structures.

All the more reason to avoid p-o-m dependencies.

> Maybe 2.7 will..

We can hope...

-Tom
--
Tom Eastep    \ Nothing is foolproof to a sufficiently talented fool
Shoreline,     \ http://shorewall.net
Washington USA  \ teastep@shorewall.net
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Netdev]     [Ethernet Bridging]     [Linux 802.1Q VLAN]     [Linux Wireless]     [Kernel Newbies]     [Security]     [Linux for Hams]     [Netfilter]     [Git]     [Bugtraq]     [Yosemite News and Information]     [MIPS Linux]     [ARM Linux]     [Linux RAID]     [Linux PCI]     [Linux Admin]     [Samba]

  Powered by Linux