linux-kernel - Re: [PATCH v2 02/11] KVM: x86: emulator: introduce update_emulation

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]

Date:   Thu, 21 Jul 2022 14:23:23 +0000
From:   Sean Christopherson <seanjc@...gle.com>
To:     Maxim Levitsky <mlevitsk@...hat.com>
Cc:     kvm@...r.kernel.org, x86@...nel.org,
        Kees Cook <keescook@...omium.org>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        linux-kernel@...r.kernel.org, "H. Peter Anvin" <hpa@...or.com>,
        Borislav Petkov <bp@...en8.de>, Joerg Roedel <joro@...tes.org>,
        Ingo Molnar <mingo@...hat.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Vitaly Kuznetsov <vkuznets@...hat.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Jim Mattson <jmattson@...gle.com>
Subject: Re: [PATCH v2 02/11] KVM: x86: emulator: introduce
 update_emulation_mode

On Thu, Jul 21, 2022, Maxim Levitsky wrote:
> On Wed, 2022-07-20 at 23:44 +0000, Sean Christopherson wrote:
> > On Tue, Jun 21, 2022, Maxim Levitsky wrote:
> > > +	if (!ctxt->ops->get_cr(ctxt, 0) & X86_CR0_PE) {
> > > +		/* Real mode. cpu must not have long mode active */
> > > +		if (efer & EFER_LMA)
> > > +			return X86EMUL_UNHANDLEABLE;
> > 
> > If we hit this, is there any hope of X86EMUL_UNHANDLEABLE doing the right thing?
> > Ah, SMM and the ability to swizzle SMRAM state.  Bummer.  I was hoping we could
> > just bug the VM.
> 
> I just tried to be a good citizen here, it is probably impossible to hit this case.
> (RSM ignores LMA bit in the EFER in the SMRAM).

The reason I asked is because if all of the X86EMUL_UNHANDLEABLE paths are impossible
then my preference would be to refactor this slightly to:

	static int emulator_calc_cpu_mode(const struct x86_emulate_ctxt *ctxt)

and return the mode instead of success/failure, and turn those checks into:

	KVM_EMULATOR_BUG_ON(efer & EFER_LMA);

with the callers being:

	ctxt->mode = emulator_calc_cpu_mode(ctxt);

But I think this one:

	if (!ctxt->ops->get_segment(ctxt, &selector, &cs, &base3, VCPU_SREG_CS))
		return X86EMUL_UNHANDLEABLE;

is reachable in the em_rsm() case :-/