[v4,00/10] evtchn: (not so) recent XSAs follow-on

Message

Jan Beulich Jan. 5, 2021, 1:07 p.m. UTC

These are grouped into a series largely because of their origin,
not so much because there are heavy dependencies among them.
Compared to v3, there are several a new patches (in particular
the simultaneous locking of two domains' event locks gets done
away with) and movement of a controversial one to the end. See
also the individual patches.

1: use per-channel lock where possible
2: bind-interdomain doesn't need to hold both domains' event locks
3: convert domain event lock to an r/w one
4: don't call Xen consumer callback with per-channel lock held
5: closing of vIRQ-s doesn't require looping over all vCPU-s
6: slightly defer lock acquire where possible
7: add helper for port_is_valid() + evtchn_from_port()
8: closing of ports doesn't need to hold both domains' event locks
9: type adjustments
10: drop acquiring of per-channel lock from send_guest_{global,vcpu}_virq()

Jan