[big discussion of user-space networking deleted]
> : >
> : > Not really. The device generating the packets runs kernel code, does it
> : > not?
> :
> : No it doesn't. It's entirely in user space -- which makes it an unusual
> : example.
>
> OK, so this is a single user device, right? If the kernel isn't
> arbitrating access to the device then you either have to have a bunch
> of friendly (and smart) users or it's single user.
Or smart hardware with kernel intervention at some earlier point in
time to set the demultiplexing up securely. We've got a NIC that does
this, and you can basically drive CPU utilization to zero while
saturating your PCI bus if you do it right.
This issue of user-space networking strikes me as a red herring, though -
I don't see any reason why user-level networking creates any greater need
for lightweight processes than anything else.
-- ............................................................................ Peter Desnoyers 162 Pleasant St. (617) 661-1979 pjd@fred.cambridge.ma.us Cambridge, Mass. 02139 (978) 461-0402 (work) pjd@giga-net.com- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/