Re: [PATCH 4/7] KVM: x86: Don't load/put vCPU when unloading its MMU during teardown

From: Sean Christopherson <seanjc@google.com>
To: Yan Zhao <yan.y.zhao@intel.com>
Cc: Marc Zyngier <maz@kernel.org>,
	Oliver Upton <oliver.upton@linux.dev>,
	 Tianrui Zhao <zhaotianrui@loongson.cn>,
	Bibo Mao <maobibo@loongson.cn>,
	 Huacai Chen <chenhuacai@kernel.org>,
	Madhavan Srinivasan <maddy@linux.ibm.com>,
	 Anup Patel <anup@brainfault.org>,
	Paul Walmsley <paul.walmsley@sifive.com>,
	 Palmer Dabbelt <palmer@dabbelt.com>,
	Albert Ou <aou@eecs.berkeley.edu>,
	 Christian Borntraeger <borntraeger@linux.ibm.com>,
	Janosch Frank <frankja@linux.ibm.com>,
	 Claudio Imbrenda <imbrenda@linux.ibm.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	 linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev,
	 kvm@vger.kernel.org, loongarch@lists.linux.dev,
	linux-mips@vger.kernel.org,  linuxppc-dev@lists.ozlabs.org,
	kvm-riscv@lists.infradead.org,  linux-riscv@lists.infradead.org,
	linux-kernel@vger.kernel.org,
	 Aaron Lewis <aaronlewis@google.com>,
	Jim Mattson <jmattson@google.com>,
	 Rick P Edgecombe <rick.p.edgecombe@intel.com>,
	Kai Huang <kai.huang@intel.com>,
	 Isaku Yamahata <isaku.yamahata@intel.com>
Subject: Re: [PATCH 4/7] KVM: x86: Don't load/put vCPU when unloading its MMU during teardown
Date: Tue, 25 Feb 2025 06:44:58 -0800	[thread overview]
Message-ID: <Z73XaiRZMIi_vyvK@google.com> (raw)
In-Reply-To: <Z71tlzQJISk6PFAL@yzhao56-desk.sh.intel.com>

On Tue, Feb 25, 2025, Yan Zhao wrote:
> On Mon, Feb 24, 2025 at 03:55:39PM -0800, Sean Christopherson wrote:
> > Don't load (and then put) a vCPU when unloading its MMU during VM
> > destruction, as nothing in kvm_mmu_unload() accesses vCPU state beyond the
> > root page/address of each MMU, i.e. can't possible need to run with the
> > vCPU loaded.
> > 
> > Signed-off-by: Sean Christopherson <seanjc@google.com>
> > ---
> >  arch/x86/kvm/x86.c | 9 +--------
> >  1 file changed, 1 insertion(+), 8 deletions(-)
> > 
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 045c61cc7e54..9978ed4c0917 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -12767,13 +12767,6 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
> >  	return ret;
> >  }
> >  
> > -static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
> > -{
> > -	vcpu_load(vcpu);
> > -	kvm_mmu_unload(vcpu);
> > -	vcpu_put(vcpu);
> > -}
> > -
> >  static void kvm_unload_vcpu_mmus(struct kvm *kvm)
> >  {
> >  	unsigned long i;
> > @@ -12781,7 +12774,7 @@ static void kvm_unload_vcpu_mmus(struct kvm *kvm)
> >  
> >  	kvm_for_each_vcpu(i, vcpu, kvm) {
> >  		kvm_clear_async_pf_completion_queue(vcpu);
> > -		kvm_unload_vcpu_mmu(vcpu);
> > +		kvm_mmu_unload(vcpu);
> What about just dropping kvm_unload_vcpu_mmu() here?
> kvm_mmu_unload() will be invoked again in kvm_mmu_destroy().
> 
> kvm_arch_vcpu_destroy() --> kvm_mmu_destroy() --> kvm_mmu_unload().

Ugh, I missed that there's yet another call to kvm_mmu_unload().  I definitely
agree with dropping the first kvm_mmu_load(), but I'll do it in a follow-up patch
so that all three changes are isolated (not doing the load/put, doing unload as
part of vCPU destruction, doing unload only once at the end).

And looking at both calls to kvm_mmu_unload(), I suspect that grabbing kvm->srcu
around kvm_mmu_destroy() is unnecessary.  I'll try cleaning that up as well.

-- 
kvm-riscv mailing list
kvm-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kvm-riscv