linux-kernel - Re: [PATCH 0/3] Reanme the definitions of INTERRUPT_PENDING, NMI_PENDING and TSC

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <2beeb1fb-7d3a-d829-38e0-ddf76b65bd3c@redhat.com>
Date:   Mon, 9 Dec 2019 16:23:15 +0100
From:   Paolo Bonzini <pbonzini@...hat.com>
To:     Sean Christopherson <sean.j.christopherson@...el.com>,
        Xiaoyao Li <xiaoyao.li@...el.com>
Cc:     Shuah Khan <shuah@...nel.org>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-kselftest@...r.kernel.org
Subject: Re: [PATCH 0/3] Reanme the definitions of INTERRUPT_PENDING,
 NMI_PENDING and TSC_OFFSETING

On 06/12/19 21:47, Sean Christopherson wrote:
>> When reading the codes, I find the definitions of interrupt-window exiting
>> and nmi-window exiting don't match the names in latest intel SDM.
> I prefer KVM's names even though they diverge from the SDM.  The "window
> exiting" terminology is very literal, which is desirable for the SDM
> because it doesn't leave any wiggle room.  But for software, IMO the
> "event pending" terminology is preferable as it's more descriptive of the
> intended use of the control, e.g. KVM sets VIRTUAL_{INTR,NMI}_PENDING when
> it has a virtual event to inject and clears it after injecting said event.
> 

On the other hand:

static void enable_irq_window(struct kvm_vcpu *vcpu)
{
        exec_controls_setbit(to_vmx(vcpu), CPU_BASED_VIRTUAL_INTR_PENDING);
}

static void enable_nmi_window(struct kvm_vcpu *vcpu)
{
        if (!enable_vnmi ||
            vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & GUEST_INTR_STATE_STI) {
                enable_irq_window(vcpu);
                return;
        }

        exec_controls_setbit(to_vmx(vcpu), CPU_BASED_VIRTUAL_NMI_PENDING);
}

so we're already using a lot the "window" nomenclature in KVM.  I've applied Xiaoyao's patches.

Paolo