netdev - Re: bond_eth_hash only uses the fifth byte of MAC address

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] [day] [month] [year] [list]

Message-ID: <1518704.1749579144@famine>
Date: Tue, 10 Jun 2025 11:12:24 -0700
From: Jay Vosburgh <jv@...sburgh.net>
To: Haylin Moore <hmoore@...ulo.com>
cc: netdev@...r.kernel.org
Subject: Re: bond_eth_hash only uses the fifth byte of MAC address

Haylin Moore <hmoore@...ulo.com> wrote:

>Hello All,
>
>I am currently digging into the source code powering layer2+3 bonding
>and saw that for bond_eth_hash only the 5th byte of both the source
>and destination MAC address is used in the XOR.
>
>Is there a reason for this? I was not able to find anything searching
>the mailing lists or the web. This functionality while documented just
>feels weird to me.

	My recollection, which could be totally wrong, is that the
described algorithm was chosen to mimic what was available on Cisco
Etherchannel at the time.  Bonding's balance-xor mode was originally
developed to interoperate with Etherchannel, before LACP became popular,
so this would have been implemented circa 2000-2002.

	Development on bonding was done out of tree at that time, and
updates were distributed on sourceforge, with discussion primarily on
the bonding-devel mailing list.  marc.info has the bonding-devel list
archived; you might be able to find the original discussion there.

	-J

---
	-Jay Vosburgh, jv@...sburgh.net