Re: [PATCH 13/34] KVM: x86: hyper-v: optimize and cleanup kvm_hv_process_stimers()

[Yury Norov]

On Mon, Nov 20, 2023 at 03:26:08PM +0100, Vitaly Kuznetsov wrote:
> Yury Norov <yury.norov@xxxxxxxxx> writes:
> 
> > The function traverses stimer_pending_bitmap n a for-loop bit by bit.
> > We can do it faster by using atomic find_and_set_bit().
> >
> > While here, refactor the logic by decreasing indentation level
> > and dropping 2nd check for stimer->config.enable.
> >
> > Signed-off-by: Yury Norov <yury.norov@xxxxxxxxx>
> > ---
> >  arch/x86/kvm/hyperv.c | 39 +++++++++++++++++++--------------------
> >  1 file changed, 19 insertions(+), 20 deletions(-)
> >
> > diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
> > index 238afd7335e4..460e300b558b 100644
> > --- a/arch/x86/kvm/hyperv.c
> > +++ b/arch/x86/kvm/hyperv.c
> > @@ -870,27 +870,26 @@ void kvm_hv_process_stimers(struct kvm_vcpu *vcpu)
> >  	if (!hv_vcpu)
> >  		return;
> >  
> > -	for (i = 0; i < ARRAY_SIZE(hv_vcpu->stimer); i++)
> > -		if (test_and_clear_bit(i, hv_vcpu->stimer_pending_bitmap)) {
> > -			stimer = &hv_vcpu->stimer[i];
> > -			if (stimer->config.enable) {
> > -				exp_time = stimer->exp_time;
> > -
> > -				if (exp_time) {
> > -					time_now =
> > -						get_time_ref_counter(vcpu->kvm);
> > -					if (time_now >= exp_time)
> > -						stimer_expiration(stimer);
> > -				}
> > -
> > -				if ((stimer->config.enable) &&
> > -				    stimer->count) {
> > -					if (!stimer->msg_pending)
> > -						stimer_start(stimer);
> > -				} else
> > -					stimer_cleanup(stimer);
> > -			}
> > +	for_each_test_and_clear_bit(i, hv_vcpu->stimer_pending_bitmap,
> > +					ARRAY_SIZE(hv_vcpu->stimer)) {
> > +		stimer = &hv_vcpu->stimer[i];
> > +		if (!stimer->config.enable)
> > +			continue;
> > +
> > +		exp_time = stimer->exp_time;
> > +
> > +		if (exp_time) {
> > +			time_now = get_time_ref_counter(vcpu->kvm);
> > +			if (time_now >= exp_time)
> > +				stimer_expiration(stimer);
> >  		}
> > +
> > +		if (stimer->count) {
> 
> You can't drop 'stimer->config.enable' check here as stimer_expiration()
> call above actually changes it. This is done on purpose: oneshot timers
> fire only once so 'config.enable' is reset to 0.

Ok, I see. Will fix in v2