> (puzzlement number one: why is there a 20-30 second delay on
> initializing CPU #1.
This issue was fixed in this changeset in -unstable:
It's an obvious candidate to be pulled into 3.0-testing, and hence 3.0.1
Xen-users mailing list