netdev - Re: [PATCH] Disable rp_filter for IPsec packets

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20110314.144156.245399355.davem@davemloft.net>
Date:	Mon, 14 Mar 2011 14:41:56 -0700 (PDT)
From:	David Miller <davem@...emloft.net>
To:	msmith@...co.com
Cc:	netdev@...r.kernel.org
Subject: Re: [PATCH] Disable rp_filter for IPsec packets

From: Michael Smith <msmith@...co.com>
Date: Mon, 14 Mar 2011 17:29:43 -0400

> David Miller wrote:
>> Secondly, fib_validate_source() already takes too many damn arguments.
>> Find another, less costly, way to pass this information down there.
> 
> What would be a less costly way to pass it? Could I just hand it the
> whole skb?

I don't see how passing a pointer is better than passing an interger.
In both cases you're adding an extra argument to the function.

I was trying to get you to think out of the box and come up with
something clever, but that isn't working. :-)

>> Frankly, I think RPF should be disabled completely by default.  When
>> it doesn't do anything useful, it's making route lookups twice as
>> expensive as they need to be.
> 
> Yeah, it's disabled by default. It's an easy way of preventing
> spoofing of internal source addresses from the Internet, so I like it.

It is not "disabled by default".  fib_validate_source() still does a
limited validation of the reverse path, even with the sysctl is set to
zero.

I want it to do absolutely nothing, and instead just use inet_select_addr()
to calculate spec_dst.

Even the spec_dst calculation is spurious, necessary only in limited
situations, and even in that case only takes on special values for
multicast and broadcast addresses.

In short, fib_validate_source() is nothing but completely unnecessary
overhead in the common case.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html