2015-11-02 17:08+0100, Paolo Bonzini:
> On 02/11/2015 15:59, Radim Krcmar wrote:
>>> We do not want to do too much work in atomic context, in particular
>>> not walking all the VCPUs of the virtual machine. So we want
>>> to distinguish the architecture-specific injection function for irqfd
>>> from kvm_set_msi. Since it's still empty, reuse the newly added
>>> kvm_arch_set_irq and rename it to kvm_arch_set_irq_inatomic.
>>
>> kvm/queue uses kvm_arch_set_irq since b7184313f4b9 ("kvm/x86: Hyper-V
>> synthetic interrupt controller").
>>
>> Is synic going to be dropped before this patch is merged?
>
> Yes. Both because the Virtuozzo people confirmed that kvm_arch_set_irq
> isn't needed for synic, and because synic is currently broken with APICv.
Thanks.
(We can add direct delivery for |online vcpus| < X if performance with
low number of VCPUs happens to regress because of the schedule_work,)
Reviewed-by: Radim Krčmář <rkrcmar@xxxxxxxxxx>
[2/3] and [3/3] are good too.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
