Dave Marquardt writes:
> So how would you tell sendfile where the packet/datagram boundaries
> are?  That's the big issue that comes to my mind.

I suspect the idea here is to call sendfile() a zillion times in a
loop, with each call's 'len' parameter set to the size of the datagram
to send -- thus using it just for its zero-copy behavior, and not for
its bulk-transfer behavior.

Though it would probably "work," I'm not sure this is really the right
or best solution to the problem.  It seems like it'd be better to have
some kind of kernel-level periodic transmission support -- something
more tailored to the protocol being supported rather than deferred to
the application.

-- 
James Carlson, Solaris Networking              <[EMAIL PROTECTED]>
Sun Microsystems / 1 Network Drive         71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to