MIME-Version: 1.0
References: <20211207043100.3357474-1-marcorr@google.com>
In-Reply-To: <20211207043100.3357474-1-marcorr@google.com>
From:   Jim Mattson <jmattson@google.com>
Date:   Mon, 6 Dec 2021 20:49:02 -0800
Message-ID: <CALMp9eR6tORQsSm6fGYumY9QrQVHTo3BxempwYRyOEYrv+zHew@mail.gmail.com>
Subject: Re: [PATCH] KVM: x86: Always set kvm_run->if_flag
To:     Marc Orr <marcorr@google.com>
Cc:     pbonzini@redhat.com, seanjc@google.com, vkuznets@redhat.com,
        wanpengli@tencent.com, joro@8bytes.org, tglx@linutronix.de,
        mingo@redhat.com, bp@alien8.de, dave.hansen@linux.intel.com,
        x86@kernel.org, hpa@zytor.com, thomas.lendacky@amd.com,
        kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Dec 6, 2021 at 8:31 PM Marc Orr <marcorr@google.com> wrote:
>
> The kvm_run struct's if_flag is apart of the userspace/kernel API. The

Typo: 'a part'.

> SEV-ES patches failed to set this flag because it's no longer needed by
> QEMU (according to the comment in the source code). However, other
> hypervisors may make use of this flag. Therefore, set the flag for
> guests with encrypted regiesters (i.e., with guest_state_protected set).

Typo: 'registers'.

> Fixes: f1c6366e3043 ("KVM: SVM: Add required changes to support intercepts under SEV-ES")
> Signed-off-by: Marc Orr <marcorr@google.com>
> ---

> -       /*
> -        * if_flag is obsolete and useless, so do not bother
> -        * setting it for SEV-ES guests.  Userspace can just
> -        * use kvm_run->ready_for_interrupt_injection.
> -        */
> -       kvm_run->if_flag = !vcpu->arch.guest_state_protected
> -               && (kvm_get_rflags(vcpu) & X86_EFLAGS_IF) != 0;
> -
> +       kvm_run->if_flag = static_call(kvm_x86_get_if_flag)(vcpu);

I'm sorry that I missed that change when it first went by. Maintaining
backwards compatibility with existing userspace code is a fundamental
tenet of Linux kernel development.

Acked-by: Jim Mattson <jmattson@google.com>