[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <7d6bdd4fca0dfdd6528e416510877c3bfd0cbfac.camel@redhat.com>
Date: Thu, 28 Sep 2023 13:42:06 +0300
From: Maxim Levitsky <mlevitsk@...hat.com>
To: Paolo Bonzini <pbonzini@...hat.com>, kvm@...r.kernel.org
Cc: Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
linux-kernel@...r.kernel.org, "H. Peter Anvin" <hpa@...or.com>,
Sean Christopherson <seanjc@...gle.com>,
Ingo Molnar <mingo@...hat.com>,
Thomas Gleixner <tglx@...utronix.de>,
Borislav Petkov <bp@...en8.de>
Subject: Re: [PATCH v2 4/4] KVM: x86: add new nested vmexit tracepoints
On, 2023-09-26 у 18:50 +0200, Paolo Bonzini writes:
> On 9/24/23 14:44, Maxim Levitsky wrote:
> > + trace_kvm_nested_page_fault(fault->address,
> > + vcpu->arch.ept_fault_error_code,
>
> Any reason to include vcpu->arch.ept_fault_error_code rather than the
> injected exit qualification?
The vcpu->arch.ept_fault_error_code is the original exit qualification
which I want to trace as the host error code
captured on ept fault to a new field, but I see new that I can use vmx_get_exit_qual()
to retrieve it again, thus avoiding the overhead.
Also for the injected error code I can use the 'exit_qualification' instead, I think
that I intended to so so but forgot to update the patch.
So something like that works fine for me:
trace_kvm_nested_page_fault(fault->address,
vmx_get_exit_qual(vcpu),
exit_qualification);
Best regards,
Maxim Levitsky
>
> Paolo
>
> > + fault->error_code);
Powered by blists - more mailing lists