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;
MIME-Version: 1.0
References: <20210412130938.68178-1-david.edmondson@oracle.com>
 <20210412130938.68178-6-david.edmondson@oracle.com> <YHRvchkUSIeU8tRR@google.com>
 <cuno8eisbf9.fsf@oracle.com>
In-Reply-To: <cuno8eisbf9.fsf@oracle.com>
From:   Aaron Lewis <aaronlewis@google.com>
Date:   Tue, 13 Apr 2021 11:45:52 -0700
Message-ID: <CAAAPnDGy2MZF2QVTTdNQgQC3Sh9mOjJx-cetn2nZ4cu6-h1Zvg@mail.gmail.com>
Subject: Re: [PATCH 5/6] KVM: SVM: pass a proper reason in kvm_emulate_instruction()
To:     David Edmondson <david.edmondson@oracle.com>
Cc:     Sean Christopherson <seanjc@google.com>,
        linux-kernel@vger.kernel.org, kvm list <kvm@vger.kernel.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>, x86@kernel.org,
        Thomas Gleixner <tglx@linutronix.de>,
        Joerg Roedel <joro@8bytes.org>,
        Wanpeng Li <wanpengli@tencent.com>,
        Borislav Petkov <bp@alien8.de>,
        Jim Mattson <jmattson@google.com>,
        Ingo Molnar <mingo@redhat.com>,
        Joao Martins <joao.m.martins@oracle.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

>
> > Depending on what you're trying to do with the info, maybe there's a better
> > option.  E.g. Aaron is working on a series that includes passing pass the code
> > stream (instruction bytes) to userspace on emulation failure, though I'm not
> > sure if he's planning on providing the VM-Exit reason.
>
> Having the instruction stream will be good.
>
> Aaron: do you have anything to share now? In what time frame do you
> think you might submit patches?

I should be able to have something out later this week.  There is no
exit reason as Sean indicated, so if that's important it will have to
be reworked afterwards.  For struct internal in kvm_run I use data[0]
for flags to indicate what's contained in the rest of it, I use
data[1] as the instruction size, and I use data[2,3] to store the
instruction bytes.  Hope that helps.

>
> I'm happy to re-work this to make the exit reason available, if that's
> the appropriate direction.
>
> dme.
> --
> And you're standing here beside me, I love the passing of time.