Garrett D'Amore wrote:
Peter Memishian wrote:
> > What about PIL related concerns? Are we guaranteed that the
crypto > doesn't require an interrupt service routine to to run that
won't be > blocked because we might already be in interrupt context?
Indeed -- e.g., memory allocation triggering disk paging. I think
disk is
typically at a lower interrupt priority (5) than networking (6), so it
seems like that would be a problem.
I was thinking the other way... i.e. we're running in interrupt
context from a NIC with a higher priority PIL than the crypto
processor uses. Meaning that the crypto interrupt cannot ever get
processed, and we deadlock.
The scenario that gives me nightmares is when you have swap
over NFS...
Darren
_______________________________________________
networking-discuss mailing list
[email protected]
