This is what has always happened in the past, people were preaching
for TOE back when 100Mbit ethernet was "new and fast". But you
certainly don't see anyone trying to justify TOE for those link
speeds today. The same will happen for 1Gbit and 10Gbit links
a year or so from now, the cpu, memory, and PCI bus will be fast
enough.
TOE is therefore by definition a technology which we know will will
be deprecated for current link technologies over time. It is a
specialized hack, and once it's in we can never take it out of
the kernel. Why put in a specialized hack when the fully functional,
fully featureful, general purpose net stack is "good enough"?
This seems to say it all. If the community had agreed upon a *generic*
solution for TOE in the early days - 100Mbit or even 1Gbit this discussion
would not have been there. It is true that once the cpu, memory and the
PCI bus are fast enough, high network speeds will not be a problem.
But that takes time and technologies like TOE can supplement the network
stack till the supporting infrastructure (cpu, memory etc..) pulls up
its socks.
And seeing what has happened during 100Mbit, 1Gbit and 10Gbit it seems
reqirements for networking are always "one step ahead" and the cpu, memory,
bus-bandwidth will take time to match the requirements. Had this evllution
been fast enough I wonder if we would have included TSO/LSO support or
for that matter even the checksum offload in the stack. The full featured,
generalised network stack would have "done it all".
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email
______________________________________________________________________
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
