Received-SPF: pass (google.com: domain of linux-kernel+bounces-177123-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
From: Leonardo Bras <leobras@redhat.com>
To: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Leonardo Bras <leobras@redhat.com>,
	Frederic Weisbecker <frederic@kernel.org>,
	"Paul E. McKenney" <paulmck@kernel.org>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Sean Christopherson <seanjc@google.com>,
	linux-kernel@vger.kernel.org,
	kvm@vger.kernel.org
Subject: Re: [RFC PATCH 1/1] kvm: Note an RCU quiescent state on guest exit
Date: Mon, 13 May 2024 00:14:32 -0300
Message-ID: <ZkGFmISfnrKNrUgj@LeoBras>
In-Reply-To: <ZkE4N1X0wglygt75@tpad>
References: <20240511020557.1198200-1-leobras@redhat.com> <ZkE4N1X0wglygt75@tpad>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: 8bit

On Sun, May 12, 2024 at 06:44:23PM -0300, Marcelo Tosatti wrote:
> On Fri, May 10, 2024 at 11:05:56PM -0300, Leonardo Bras wrote:
> > As of today, KVM notes a quiescent state only in guest entry, which is good
> > as it avoids the guest being interrupted for current RCU operations.
> > 
> > While the guest vcpu runs, it can be interrupted by a timer IRQ that will
> > check for any RCU operations waiting for this CPU. In case there are any of
> > such, it invokes rcu_core() in order to sched-out the current thread and
> > note a quiescent state.
> > 
> > This occasional schedule work will introduce tens of microsseconds of
> > latency, which is really bad for vcpus running latency-sensitive
> > applications, such as real-time workloads.
> > 
> > So, note a quiescent state in guest exit, so the interrupted guests is able
> > to deal with any pending RCU operations before being required to invoke
> > rcu_core(), and thus avoid the overhead of related scheduler work.
> 
> This does not properly fix the current problem, as RCU work might be
> scheduled after the VM exit, followed by a timer interrupt.
> 
> Correct?

Correct, for this case, check the note below:

> 
> > 
> > Signed-off-by: Leonardo Bras <leobras@redhat.com>
> > ---
> > 
> > ps: A patch fixing this same issue was discussed in this thread:
> > https://lore.kernel.org/all/20240328171949.743211-1-leobras@redhat.com/
> > 
> > Also, this can be paired with a new RCU option (rcutree.nocb_patience_delay)
> > to avoid having invoke_rcu() being called on grace-periods starting between
> > guest exit and the timer IRQ. This RCU option is being discussed in a
> > sub-thread of this message:
> > https://lore.kernel.org/all/5fd66909-1250-4a91-aa71-93cb36ed4ad5@paulmck-laptop/

^ This one above.
The idea is to use this rcutree.nocb_patience_delay=N :
a new option we added on RCU that allow us to avoid invoking rcu_core() if 
the grace_period < N miliseconds. This only works on nohz_full cpus.

So with both the current patch and the one in above link, we have the same 
effect as we previously had with last_guest_exit, with a cherry on top: we 
can avoid rcu_core() getting called in situations where a grace period just 
started after going into kernel code, and a timer interrupt happened before 
it can report quiescent state again. 

For our nohz_full vcpu thread scenario, we have:

- guest_exit note a quiescent state
- let's say we start a grace period in the next cycle
- If timer interrupts, it requires the grace period to be older than N 
  miliseconds
  - If we configure a proper value for patience, it will never reach the 
    end of patience before going guest_entry, and thus noting a quiescent 
    state

What do you think?

Thanks!
Leo

> > 
> > 
> >  include/linux/context_tracking.h |  6 ++++--
> >  include/linux/kvm_host.h         | 10 +++++++++-
> >  2 files changed, 13 insertions(+), 3 deletions(-)
> > 
> > diff --git a/include/linux/context_tracking.h b/include/linux/context_tracking.h
> > index 6e76b9dba00e..8a78fabeafc3 100644
> > --- a/include/linux/context_tracking.h
> > +++ b/include/linux/context_tracking.h
> > @@ -73,39 +73,41 @@ static inline void exception_exit(enum ctx_state prev_ctx)
> >  }
> >  
> >  static __always_inline bool context_tracking_guest_enter(void)
> >  {
> >  	if (context_tracking_enabled())
> >  		__ct_user_enter(CONTEXT_GUEST);
> >  
> >  	return context_tracking_enabled_this_cpu();
> >  }
> >  
> > -static __always_inline void context_tracking_guest_exit(void)
> > +static __always_inline bool context_tracking_guest_exit(void)
> >  {
> >  	if (context_tracking_enabled())
> >  		__ct_user_exit(CONTEXT_GUEST);
> > +
> > +	return context_tracking_enabled_this_cpu();
> >  }
> >  
> >  #define CT_WARN_ON(cond) WARN_ON(context_tracking_enabled() && (cond))
> >  
> >  #else
> >  static inline void user_enter(void) { }
> >  static inline void user_exit(void) { }
> >  static inline void user_enter_irqoff(void) { }
> >  static inline void user_exit_irqoff(void) { }
> >  static inline int exception_enter(void) { return 0; }
> >  static inline void exception_exit(enum ctx_state prev_ctx) { }
> >  static inline int ct_state(void) { return -1; }
> >  static inline int __ct_state(void) { return -1; }
> >  static __always_inline bool context_tracking_guest_enter(void) { return false; }
> > -static __always_inline void context_tracking_guest_exit(void) { }
> > +static __always_inline bool context_tracking_guest_exit(void) { return false; }
> >  #define CT_WARN_ON(cond) do { } while (0)
> >  #endif /* !CONFIG_CONTEXT_TRACKING_USER */
> >  
> >  #ifdef CONFIG_CONTEXT_TRACKING_USER_FORCE
> >  extern void context_tracking_init(void);
> >  #else
> >  static inline void context_tracking_init(void) { }
> >  #endif /* CONFIG_CONTEXT_TRACKING_USER_FORCE */
> >  
> >  #ifdef CONFIG_CONTEXT_TRACKING_IDLE
> > diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> > index 48f31dcd318a..e37724c44843 100644
> > --- a/include/linux/kvm_host.h
> > +++ b/include/linux/kvm_host.h
> > @@ -480,21 +480,29 @@ static __always_inline void guest_state_enter_irqoff(void)
> >  /*
> >   * Exit guest context and exit an RCU extended quiescent state.
> >   *
> >   * Between guest_context_enter_irqoff() and guest_context_exit_irqoff() it is
> >   * unsafe to use any code which may directly or indirectly use RCU, tracing
> >   * (including IRQ flag tracing), or lockdep. All code in this period must be
> >   * non-instrumentable.
> >   */
> >  static __always_inline void guest_context_exit_irqoff(void)
> >  {
> > -	context_tracking_guest_exit();
> > +	/*
> > +	 * Guest mode is treated as a quiescent state, see
> > +	 * guest_context_enter_irqoff() for more details.
> > +	 */
> > +	if (!context_tracking_guest_exit()) {
> > +		instrumentation_begin();
> > +		rcu_virt_note_context_switch();
> > +		instrumentation_end();
> > +	}
> >  }
> >  
> >  /*
> >   * Stop accounting time towards a guest.
> >   * Must be called after exiting guest context.
> >   */
> >  static __always_inline void guest_timing_exit_irqoff(void)
> >  {
> >  	instrumentation_begin();
> >  	/* Flush the guest cputime we spent on the guest */
> > -- 
> > 2.45.0
> > 
> > 
> > 
>