Hello Jeff , All , Why couldn't the implementation be setup that
the tcp offload code would be acquired from a kernel build ? ie:
the card presents a filesystem to the user and a script (or ?) is
used to place the necessary components onto that file system &
tested & ... Very difficult yes , But s/b do able .
Thoughts , JimL
On Thu, 5 Jun 2003, Jeff Garzik wrote:
> David S. Miller wrote:
> > Just remember that those customers will come back to you a few months
> > later and ask why there is next to zero performance benefit from the
> > more expensive TOE solution compared to the TSO+checksum offloading
> > solution.
>
> Not to mention the loss of flexibility (netfilter, etc) and the
> near-impossibility to do security evaluation, and the near-impossibility
> to do security errata if a bug is found in the offloaded TCP stack after
> product deployment.
> TCP Offload is, IMO, a legal liability no company with smart lawyers
> would touch...
> Jeff
--
+------------------------------------------------------------------+
| James W. Laferriere | System Techniques | Give me VMS |
| Network Engineer | P.O. Box 854 | Give me Linux |
| babydr@baby-dragons.com | Coudersport PA 16915 | only on AXP |
+------------------------------------------------------------------+
-
: 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
