Re: [PATCH RFC v3 3/6] sched/idle: Add a generic poll before enter real idle path

[Quan Xu]

On 2017-11-16 17:53, Thomas Gleixner wrote:
> On Thu, 16 Nov 2017, Quan Xu wrote:
> > On 2017-11-16 06:03, Thomas Gleixner wrote:
> > --- a/drivers/cpuidle/cpuidle.c
> > +++ b/drivers/cpuidle/cpuidle.c
> > @@ -210,6 +210,13 @@ int cpuidle_enter_state(struct cpuidle_device *dev,
> > struct cpuidle_driver *drv,
> >  ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ target_state = &drv->states[index];
> >  ÂÂÂÂÂÂÂ }
> >
> > +#ifdef CONFIG_PARAVIRT
> > +ÂÂÂÂÂÂ paravirt_idle_poll();
> > +
> > +ÂÂÂÂÂÂ if (need_resched())
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂ return -EBUSY;
> > +#endif
> That's just plain wrong. We don't want to see any of this PARAVIRT crap in
> anything outside the architecture/hypervisor interfacing code which really
> needs it.
>
> The problem can and must be solved at the generic level in the first place
> to gather the data which can be used to make such decisions.
>
> How that information is used might be either completely generic or requires
> system specific variants. But as long as we don't have any information at
> all we cannot discuss that.
>
> Please sit down and write up which data needs to be considered to make
> decisions about probabilistic polling. Then we need to compare and contrast
> that with the data which is necessary to make power/idle state decisions.
>
> I would be very surprised if this data would not overlap by at least 90%.

Peter, tglx
Thanks for your comments..

rethink of this patch set,

1. which data needs to considerd to make decisions about probabilistic 
polling

I really need to write up which data needs to considerd to make
decisions about probabilistic polling. At last several months,
I always focused on the data _from idle to reschedule_, then to bypass
the idle loops. unfortunately, this makes me touch scheduler/idle/nohz
code inevitably.

with tglx's suggestion, the data which is necessary to make power/idle
state decisions, is the last idle state's residency time. IIUC this data
is duration from idle to wakeup, which maybe by reschedule irq or other irq.

I also test that the reschedule irq overlap by more than 90% (trace the
need_resched status after cpuidle_idle_call), when I run ctxsw/netperf for
one minute.

as the overlap, I think I can input the last idle state's residency time
to make decisions about probabilistic polling, as @dev->last_residency does.
it is much easier to get data.


2. do a HV specific idle driver (function)

so far, power management is not exposed to guest.. idle is simple for 
KVM guest,
calling "sti" / "hlt"(cpuidle_idle_call() --> default_idle_call())..
thanks Xen guys, who has implemented the paravirt framework. I can 
implement it
as easy as following:

ÂÂÂÂÂÂÂÂÂÂÂÂ --- a/arch/x86/kernel/kvm.c
ÂÂÂÂÂÂÂÂÂÂÂÂ +++ b/arch/x86/kernel/kvm.c
ÂÂÂÂÂÂÂÂÂÂÂÂ @@ -465,6 +465,12 @@ static void __init 
kvm_apf_trap_init(void)
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ update_intr_gate(X86_TRAP_PF, async_page_fault);
ÂÂÂÂÂÂÂÂÂÂÂÂÂ }

ÂÂÂÂÂÂÂÂÂÂÂÂ +static __cpuidle void kvm_safe_halt(void)
ÂÂÂÂÂÂÂÂÂÂÂÂ +{
ÂÂÂ ÂÂÂÂ +ÂÂÂÂÂÂÂ /* 1. POLL, if need_resched() --> return */
ÂÂÂ ÂÂÂÂ +
ÂÂÂÂÂÂÂÂÂÂÂÂ +ÂÂÂÂÂÂÂ asm volatile("sti; hlt": : :"memory"); /* 2. halt */
ÂÂÂÂÂÂÂÂÂÂÂÂ +
ÂÂÂ ÂÂÂÂ +ÂÂÂÂÂÂÂ /* 3. get the last idle state's residency time */
ÂÂÂÂÂÂÂÂÂÂÂÂ +
ÂÂÂ ÂÂÂÂ +ÂÂÂÂÂÂÂ /* 4. update poll duration based on last idle state's 
residency time */
ÂÂÂÂÂÂÂÂÂÂÂÂ +}
ÂÂÂÂÂÂÂÂÂÂÂÂ +
ÂÂÂÂÂÂÂÂÂÂÂÂÂ void __init kvm_guest_init(void)
ÂÂÂÂÂÂÂÂÂÂÂÂÂ {
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ int i;
ÂÂÂÂÂÂÂÂÂÂÂÂ @@ -490,6 +496,8 @@ void __init kvm_guest_init(void)
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ if (kvmclock_vsyscall)
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ kvm_setup_vsyscall_timeinfo();

ÂÂÂÂÂÂÂÂÂÂÂÂ +ÂÂÂÂÂÂ pv_irq_ops.safe_halt = kvm_safe_halt;
ÂÂÂÂÂÂÂÂÂÂÂÂ +
ÂÂÂÂÂÂÂÂÂÂÂÂÂ #ifdef CONFIG_SMP




then, I am no need to introduce a new pvops, and never modify 
schedule/idle/nohz code again.
also I can narrow all of the code down in arch/x86/kernel/kvm.c.

If this is in the right direction, I will send a new patch set next week..

thanks,

Quan
Alibaba Cloud