linux-kernel - Re: [PATCH v3 5/6] KVM: x86: make several AVIC callbacks optional

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <Yg/ZZMAz7XZ6wn/u@google.com>
Date:   Fri, 18 Feb 2022 17:37:40 +0000
From:   Sean Christopherson <seanjc@...gle.com>
To:     Paolo Bonzini <pbonzini@...hat.com>
Cc:     linux-kernel@...r.kernel.org, kvm@...r.kernel.org
Subject: Re: [PATCH v3 5/6] KVM: x86: make several AVIC callbacks optional

On Fri, Feb 18, 2022, Paolo Bonzini wrote:
> On 2/18/22 17:23, Sean Christopherson wrote:
> > The "AVIC" callbacks are being deleted, not
> > made optional, it's kvm_x86_ops' APICv hooks that are becoming optional.
> 
> Maybe "make several APIC virtualization callbacks optional".

Works for me.

> > > +KVM_X86_OP_OPTIONAL(apicv_post_state_restore)
> > 
> > apicv_post_state_restore() isn't conditional, it's implemented and wired up
> > unconditionally by both VMX and SVM.
> 
> True, on the other hand there's no reason why a hypothetical third vendor
> would have to support it.  The call is conditional to apicv_active being
> true.

Ah, right, even if the the static_call_cond() is unnecessary because we want to
make the hook mandatory if APICv is supported, APICv itself may not be supported.

With the new shortlog,

Reviewed-by: Sean Christopherson <seanjc@...gle.com>