Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 17 Nov 2022 11:59:51 -0800
From:   Isaku Yamahata <isaku.yamahata@gmail.com>
To:     Binbin Wu <binbin.wu@linux.intel.com>
Cc:     isaku.yamahata@intel.com, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org, isaku.yamahata@gmail.com,
        Paolo Bonzini <pbonzini@redhat.com>, erdemaktas@google.com,
        Sean Christopherson <seanjc@google.com>,
        Sagi Shahar <sagis@google.com>,
        David Matlack <dmatlack@google.com>,
        Kai Huang <kai.huang@intel.com>
Subject: Re: [PATCH v10 049/108] KVM: x86/tdp_mmu: Support TDX private
 mapping for TDP MMU
Message-ID: <20221117195951.GO2350331@ls.amr.corp.intel.com>
References: <cover.1667110240.git.isaku.yamahata@intel.com>
 <9d5595dfe1b5ab77bcb5650bc4d940dd977b0a32.1667110240.git.isaku.yamahata@intel.com>
 <30be6d64-31bd-bfc8-72f7-fb57999e4566@linux.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <30be6d64-31bd-bfc8-72f7-fb57999e4566@linux.intel.com>
Precedence: bulk

On Tue, Nov 08, 2022 at 09:41:44PM +0800,
Binbin Wu <binbin.wu@linux.intel.com> wrote:

> > diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h
> > index f28c9fd72ac4..1b01dc2098b0 100644
> > --- a/arch/x86/include/asm/kvm-x86-ops.h
> > +++ b/arch/x86/include/asm/kvm-x86-ops.h
> > @@ -94,6 +94,11 @@ KVM_X86_OP_OPTIONAL_RET0(set_tss_addr)
> >   KVM_X86_OP_OPTIONAL_RET0(set_identity_map_addr)
> >   KVM_X86_OP_OPTIONAL_RET0(get_mt_mask)
> >   KVM_X86_OP(load_mmu_pgd)
> > +KVM_X86_OP_OPTIONAL(link_private_spt)
> > +KVM_X86_OP_OPTIONAL(free_private_spt)
> > +KVM_X86_OP_OPTIONAL(set_private_spte)
> > +KVM_X86_OP_OPTIONAL(remove_private_spte)
> > +KVM_X86_OP_OPTIONAL(zap_private_spte)
> >   KVM_X86_OP(has_wbinvd_exit)
> >   KVM_X86_OP(get_l2_tsc_offset)
> >   KVM_X86_OP(get_l2_tsc_multiplier)
> > 
> > @@ -509,9 +524,81 @@ static void handle_removed_pt(struct kvm *kvm, tdp_ptep_t pt, bool shared)
> >   		WARN_ON_ONCE(ret);
> >   	}
> > +	if (is_private_sp(sp) &&
> > +	    WARN_ON(static_call(kvm_x86_free_private_spt)(kvm, sp->gfn, sp->role.level,
> > +							  kvm_mmu_private_spt(sp)))) {
> > +		/*
> > +		 * Failed to unlink Secure EPT page and there is nothing to do
> > +		 * further.  Intentionally leak the page to prevent the kernel
> > +		 * from accessing the encrypted page.
> > +		 */
> > +		kvm_mmu_init_private_spt(sp, NULL);
> 
> Do you think is it better to add some statistics for the intentinal leakage?

No because this error case happens only when race condition bug(TDX_OPERAND_BUSY)
or TDX module was shutdown due to bug or attack.  They are (should be) rare.

The race condition bug should be fixed. It's covered by WARN_ON().  In the case
of TDX module shutdown, the current TDX KVM implementation checks unexpected
error and mark it bugged, KVM_BUG_ON(), in vmx/tdx.c.  So adding statistic
counter doesn't help.  It's another story to harden the code against TDX module
shutdown.
-- 
Isaku Yamahata <isaku.yamahata@gmail.com>