Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751987AbdHAO6x (ORCPT ); Tue, 1 Aug 2017 10:58:53 -0400 Received: from mail-wm0-f43.google.com ([74.125.82.43]:36751 "EHLO mail-wm0-f43.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751837AbdHAO6u (ORCPT ); Tue, 1 Aug 2017 10:58:50 -0400 Date: Tue, 1 Aug 2017 16:58:44 +0200 From: Christoffer Dall To: Jintack Lim Cc: kvmarm@lists.cs.columbia.edu, Christoffer Dall , Marc Zyngier , Jonathan Corbet , Paolo Bonzini , Radim =?utf-8?B?S3LEjW3DocWZ?= , linux@armlinux.org.uk, Catalin Marinas , Will Deacon , akpm@linux-foundation.org, mchehab@kernel.org, cov@codeaurora.org, Daniel Lezcano , david.daney@cavium.com, mark.rutland@arm.com, Suzuki K Poulose , stefan@hello-penguin.com, Andy Gross , wcohen@redhat.com, Ard Biesheuvel , shankerd@codeaurora.org, vladimir.murzin@arm.com, james.morse@arm.com, linux-doc@vger.kernel.org, lkml - Kernel Mailing List , KVM General , arm-mail-list Subject: Re: [RFC PATCH v2 04/38] KVM: arm/arm64: Check if nested virtualization is in use Message-ID: <20170801145844.GK5176@cbox> References: <1500397144-16232-1-git-send-email-jintack.lim@linaro.org> <1500397144-16232-5-git-send-email-jintack.lim@linaro.org> <20170730195949.GF5176@cbox> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3060 Lines: 85 On Tue, Aug 01, 2017 at 10:07:40AM -0400, Jintack Lim wrote: > On Sun, Jul 30, 2017 at 3:59 PM, Christoffer Dall wrote: > > On Tue, Jul 18, 2017 at 11:58:30AM -0500, Jintack Lim wrote: > >> Nested virtualizaion is in use only if all three conditions are met: > >> - The architecture supports nested virtualization. > >> - The kernel parameter is set. > >> - The userspace uses nested virtualiztion feature. > >> > >> Signed-off-by: Jintack Lim > >> --- > >> arch/arm/include/asm/kvm_host.h | 11 +++++++++++ > >> arch/arm64/include/asm/kvm_host.h | 2 ++ > >> arch/arm64/kvm/nested.c | 17 +++++++++++++++++ > >> virt/kvm/arm/arm.c | 4 ++++ > >> 4 files changed, 34 insertions(+) > >> > >> diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h > >> index 00b0f97..7e9e6c8 100644 > >> --- a/arch/arm/include/asm/kvm_host.h > >> +++ b/arch/arm/include/asm/kvm_host.h > >> @@ -303,4 +303,15 @@ static inline int __init kvmarm_nested_cfg(char *buf) > >> { > >> return 0; > >> } > >> + > >> +static inline int init_nested_virt(void) > >> +{ > >> + return 0; > >> +} > >> + > >> +static inline bool nested_virt_in_use(struct kvm_vcpu *vcpu) > >> +{ > >> + return false; > >> +} > >> + > >> #endif /* __ARM_KVM_HOST_H__ */ > >> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h > >> index 6df0c7c..86d4b6c 100644 > >> --- a/arch/arm64/include/asm/kvm_host.h > >> +++ b/arch/arm64/include/asm/kvm_host.h > >> @@ -387,5 +387,7 @@ static inline void __cpu_init_stage2(void) > >> } > >> > >> int __init kvmarm_nested_cfg(char *buf); > >> +int init_nested_virt(void); > >> +bool nested_virt_in_use(struct kvm_vcpu *vcpu); > >> > >> #endif /* __ARM64_KVM_HOST_H__ */ > >> diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c > >> index 79f38da..9a05c76 100644 > >> --- a/arch/arm64/kvm/nested.c > >> +++ b/arch/arm64/kvm/nested.c > >> @@ -24,3 +24,20 @@ int __init kvmarm_nested_cfg(char *buf) > >> { > >> return strtobool(buf, &nested_param); > >> } > >> + > >> +int init_nested_virt(void) > >> +{ > >> + if (nested_param && cpus_have_const_cap(ARM64_HAS_NESTED_VIRT)) > >> + kvm_info("Nested virtualization is supported\n"); > >> + > >> + return 0; > >> +} > >> + > >> +bool nested_virt_in_use(struct kvm_vcpu *vcpu) > >> +{ > >> + if (nested_param && cpus_have_const_cap(ARM64_HAS_NESTED_VIRT) > >> + && test_bit(KVM_ARM_VCPU_NESTED_VIRT, vcpu->arch.features)) > >> + return true; > > > > you could initialize a bool in init_nested_virt which you then check > > here to avoid duplicating the logic. > > I can make a bool to check the kernel param and the capability. The > third one is per VM given by the userspace, so we don't know it when > we initialize the host hypervisor. We can potentially have a bool in > kvm_vcpu_arch or kvm_arch to cache the whole three conditions, if that > sounds ok. > Yes, that sounds good to me. Thanks, -Christoffer