On Tue, 4 Jun 2019 14:53:26 +0100 Marc Zyngier <[email protected]> wrote:
> That's to prevent the injection of an interrupt firing on the same CPU > while we're saving the corresponding vcpu interrupt context, among other > things (the whole guest exit path runs with interrupt disabled in order > to avoid this kind of thing). Can't we use a per_cpu local lock for this? -- Steve _______________________________________________ kvmarm mailing list [email protected] https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
