On 11/21/2014 02:00 PM, Marc Zyngier wrote:
> On 21/11/14 12:53, Christoffer Dall wrote:
>> On Mon, Nov 17, 2014 at 10:36:13AM +0000, Marc Zyngier wrote:
>>> On 17/11/14 09:27, wanghaibin wrote:
>>>> When vgic_update_irq_pending with level-sensitive false, it is need to
>>>> deactivates an interrupt, and, it can go to out directly.
>>>> Here return a false value, because it will be not need to kick.
>>>>
>>>> Signed-off-by: wanghaibin <wanghaibin.wang@xxxxxxxxxx>
>>>> ---
>>>> virt/kvm/arm/vgic.c | 3 +++
>>>> 1 files changed, 3 insertions(+), 0 deletions(-)
>>>>
>>>> diff --git a/virt/kvm/arm/vgic.c b/virt/kvm/arm/vgic.c
>>>> index 52f3bfa..b663140 100644
>>>> --- a/virt/kvm/arm/vgic.c
>>>> +++ b/virt/kvm/arm/vgic.c
>>>> @@ -1665,6 +1665,9 @@ static bool vgic_update_irq_pending(struct kvm *kvm, int cpuid,
>>>> } else {
>>>> vgic_dist_irq_clear_pending(vcpu, irq_num);
>>>> }
Hi all,
As we are going to touch this function, I think the above else section
is dead code. Also the test of level_triggered in !level is useless.
In case we have a edge sensitive IRQ, level cannot be 0 here since
vgic_validate_injection would have caused a return before.
I tested all changes on Calxeda.
Best Regards
Eric
>>>> +
>>>> + ret = false;
>>>> + goto out;
>>>> }
>>>>
>>>> enabled = vgic_irq_is_enabled(vcpu, irq_num);
>>>>
>>>
>>> Have you actually tested this on real hardware? This looks like an
>>> interesting optimization, but I want to see some actual data.
>>>
>> Do you see any harm in merging this? I think it looks reasonable and
>> non-disruptive?
>
> Agreed. I'll take it.
>
> M.
>
_______________________________________________
kvmarm mailing list
kvmarm@xxxxxxxxxxxxxxxxxxxxx
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
