Oliver Yang writes:
> When someone asked me whether we need do zone testing for NIC driver, I 
> found need a good understanding of how drivers work with regards to
> zones. Does anybody can explain it to me?

At least until something fancy like Crossbow arrives, network
interfaces know nothing at all about Zones.  The Zone ID is an issue
in the IP stack itself.

> My conclusion is, we just need test NIC interface can be work properly 
> with "physical-unit:logical-unit" form in a global zone.

I agree.  One can never have "too much" testing, but I don't see a
point to having special nonglobal zone tests of network interface
drivers.

There are some complicated issues related to Zones that you may need
to look into, though.  One of them is that zone shutdown depends on
having the last reference to the zone dropped.  Each cred_t holds a
reference on the zone.  And each dblk_t can have a reference to a
cred_t.  Thus, if you don't strip the cred_ts from the dblks before
enqueuing for output, or if you use desballoc() and friends to get a
callback on free of blocks you allocate, you can end up delaying zone
operations indefinitely.

These aren't just Zones issues, though.  Similar design issues can
affect DR.

-- 
James Carlson, KISS Network                    <[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