From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Paul E. McKenney" Subject: Re: [PATCH v2] kvm/x86: Inform RCU of quiescent state when entering guest mode Date: Wed, 11 Jul 2018 11:20:53 -0700 Message-ID: <20180711182053.GA3593@linux.vnet.ibm.com> References: <20180711174843.GX3593@linux.vnet.ibm.com> <20180711180101.3711464-1-dwmw2@infradead.org> Reply-To: paulmck@linux.vnet.ibm.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: peterz@infradead.org, mhillenb@amazon.de, linux-kernel@vger.kernel.org, kvm@vger.kernel.org To: David Woodhouse Return-path: Content-Disposition: inline In-Reply-To: <20180711180101.3711464-1-dwmw2@infradead.org> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On Wed, Jul 11, 2018 at 07:01:01PM +0100, David Woodhouse wrote: > From: David Woodhouse > > RCU can spend long periods of time waiting for a CPU which is actually in > KVM guest mode, entirely pointlessly. Treat it like the idle and userspace > modes, and don't wait for it. > > Signed-off-by: David Woodhouse And idiot here forgot about some of the debugging code in RCU's dyntick-idle code. I will reply with a fixed patch. The code below works just fine as long as you don't enable CONFIG_RCU_EQS_DEBUG, so should be OK for testing, just not for mainline. Thanx, Paul > --- > arch/x86/kvm/x86.c | 2 ++ > include/linux/rcupdate.h | 7 +++++++ > kernel/rcu/tree.c | 2 ++ > 3 files changed, 11 insertions(+) > > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c > index 0046aa70205a..b0c82f70afa7 100644 > --- a/arch/x86/kvm/x86.c > +++ b/arch/x86/kvm/x86.c > @@ -7458,7 +7458,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) > vcpu->arch.switch_db_regs &= ~KVM_DEBUGREG_RELOAD; > } > > + rcu_kvm_enter(); > kvm_x86_ops->run(vcpu); > + rcu_kvm_exit(); > > /* > * Do this here before restoring debug registers on the host. And > diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h > index 65163aa0bb04..1325f9d9ce00 100644 > --- a/include/linux/rcupdate.h > +++ b/include/linux/rcupdate.h > @@ -212,6 +212,13 @@ do { \ > #error "Unknown RCU implementation specified to kernel configuration" > #endif > > +/* > + * These are currently identical to the _idle_ versions but let's > + * explicitly have separate copies to keep Paul honest in future. > + */ > +static inline void rcu_kvm_enter(void) { rcu_idle_enter(); } > +static inline void rcu_kvm_exit(void) { rcu_idle_exit(); } > + > /* > * The init_rcu_head_on_stack() and destroy_rcu_head_on_stack() calls > * are needed for dynamic initialization and destruction of rcu_head > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c > index aa7cade1b9f3..d38c381bf4e3 100644 > --- a/kernel/rcu/tree.c > +++ b/kernel/rcu/tree.c > @@ -760,6 +760,7 @@ void rcu_idle_enter(void) > lockdep_assert_irqs_disabled(); > rcu_eqs_enter(false); > } > +EXPORT_SYMBOL_GPL(rcu_idle_enter); > > #ifdef CONFIG_NO_HZ_FULL > /** > @@ -913,6 +914,7 @@ void rcu_idle_exit(void) > rcu_eqs_exit(false); > local_irq_restore(flags); > } > +EXPORT_SYMBOL_GPL(rcu_idle_exit); > > #ifdef CONFIG_NO_HZ_FULL > /** > -- > 2.17.1 >