linux-kernel - Re: [PATCH 19/61] KVM: VMX: Add helpers to query Intel PT mode

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] [day] [month] [year] [list]

Date:   Tue, 3 Mar 2020 14:41:51 -0800
From:   Sean Christopherson <sean.j.christopherson@...el.com>
To:     Paolo Bonzini <pbonzini@...hat.com>
Cc:     Vitaly Kuznetsov <vkuznets@...hat.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Jim Mattson <jmattson@...gle.com>,
        Joerg Roedel <joro@...tes.org>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 19/61] KVM: VMX: Add helpers to query Intel PT mode

Disclaimer: I'm going off a few lines in the SDM and the original patches,
everything I say could be completely wrong :-)

On Tue, Feb 25, 2020 at 03:54:21PM +0100, Paolo Bonzini wrote:
> On 24/02/20 23:18, Sean Christopherson wrote:
> >>>  {
> >>>  	u32 vmexit_ctrl = vmcs_config.vmexit_ctrl;
> >>> -	if (pt_mode == PT_MODE_SYSTEM)
> >>> +	if (vmx_pt_mode_is_system())
> >> ... and here? I.e. to cover the currently unsupported 'host-only' mode.
> > Hmm, good question.  I don't think so?  On VM-Enter, RTIT_CTL would need to
> > be loaded to disable PT.  Clearing RTIT_CTL on VM-Exit would be redundant
> > at that point[1].  And AIUI, the PIP for VM-Enter/VM-Exit isn't needed
> > because there is no context switch from the decoder's perspective.
> 
> How does host-only mode differ from "host-guest but don't expose PT to
> the guest"?  So I would say that host-only mode is a special case of
> host-guest, not of system mode.

AIUI, host-guest needs a special packet for VM-Enter/VM-Exit so that the
trace analyzer understands there was a context switch.  With host-only, the
packet isn't needed because tracing stops entirely.  So it's not that
host-only is a special case of system mode, but rather it doesn't need the
VM-Exit control enabled to generate the special packet.