On Fri, 2015-10-23 at 13:38 +0100, David Vrabel wrote:
> On 23/10/15 13:37, Ian Campbell wrote:
> > On Fri, 2015-10-23 at 12:05 +0100, David Vrabel wrote:
> > > When injecting an interrupt for a passthrough device into a guest,
> > > the
> > > per-domain event_lock is held, reducing performance whe
On Fri, 2015-10-23 at 12:05 +0100, David Vrabel wrote:
> When injecting an interrupt for a passthrough device into a guest, the
> per-domain event_lock is held, reducing performance when a guest has
> many VCPUs and high interrupt rates.
Did you CC me due to a possible impact on ARM? If so then I
On 23/10/15 13:37, Ian Campbell wrote:
> On Fri, 2015-10-23 at 12:05 +0100, David Vrabel wrote:
>> When injecting an interrupt for a passthrough device into a guest, the
>> per-domain event_lock is held, reducing performance when a guest has
>> many VCPUs and high interrupt rates.
>
> Did you CC m
When injecting an interrupt for a passthrough device into a guest, the
per-domain event_lock is held, reducing performance when a guest has
many VCPUs and high interrupt rates.
By using a per-interrupt lock in the hot paths, this contention is
eliminated and performance improves (a bit).
For test
