Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Thomas Gleixner <tglx@linutronix.de>
To:     Stefano De Venuto <stefano.devenuto99@gmail.com>,
        linux-kernel@vger.kernel.org
Cc:     pbonzini@redhat.com, seanjc@google.com, vkuznets@redhat.com,
        wanpengli@tencent.com, jmattson@google.com, x86@kernel.org,
        hpa@zytor.com, kvm@vger.kernel.org, rostedt@goodmis.org,
        y.karadz@gmail.com,
        Stefano De Venuto <stefano.devenuto99@gmail.com>,
        Dario Faggioli <dfaggioli@suse.com>
Subject: Re: [PATCH] Move VMEnter and VMExit tracepoints closer to the actual event
In-Reply-To: <20210519182303.2790-1-stefano.devenuto99@gmail.com>
References: <20210519182303.2790-1-stefano.devenuto99@gmail.com>
Date:   Thu, 20 May 2021 09:21:24 +0200
Message-ID: <875yzddg5n.ffs@nanos.tec.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On Wed, May 19 2021 at 20:23, Stefano De Venuto wrote:
> diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
> index 05eca131eaf2..c77d4866e239 100644
> --- a/arch/x86/kvm/svm/svm.c
> +++ b/arch/x86/kvm/svm/svm.c
> @@ -3275,8 +3275,6 @@ static int handle_exit(struct kvm_vcpu *vcpu, fastpath_t exit_fastpath)
>  	struct kvm_run *kvm_run = vcpu->run;
>  	u32 exit_code = svm->vmcb->control.exit_code;
>  
> -	trace_kvm_exit(exit_code, vcpu, KVM_ISA_SVM);
> -
>  	/* SEV-ES guests must use the CR write traps to track CR registers. */
>  	if (!sev_es_guest(vcpu->kvm)) {
>  		if (!svm_is_intercept(svm, INTERCEPT_CR0_WRITE))
> @@ -3707,6 +3705,8 @@ static noinstr void svm_vcpu_enter_exit(struct kvm_vcpu *vcpu)
>  
>  	kvm_guest_enter_irqoff();
>  
> +	trace_kvm_entry(vcpu);

No. This violates the noinstr rules and will make objtool complain on a
full validation run.

So this wants to be:

+       instrumentation_begin();
+	trace_kvm_entry(vcpu);
+       instrumentation_end();

  	kvm_guest_enter_irqoff();

and on the exit side the trace wants to be post kvm_guest_exit_irqoff().

Thanks,

        tglx