linux-kernel - Re: [RFC v2 2/2] KVM: VMX: Enable bus lock VM exit

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:   Wed, 2 Sep 2020 18:25:06 -0700
From:   Sean Christopherson <sean.j.christopherson@...el.com>
To:     Xiaoyao Li <xiaoyao.li@...el.com>
Cc:     Vitaly Kuznetsov <vkuznets@...hat.com>,
        Chenyi Qiang <chenyi.qiang@...el.com>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org, Paolo Bonzini <pbonzini@...hat.com>,
        Wanpeng Li <wanpengli@...cent.com>,
        Jim Mattson <jmattson@...gle.com>,
        Joerg Roedel <joro@...tes.org>
Subject: Re: [RFC v2 2/2] KVM: VMX: Enable bus lock VM exit

On Thu, Sep 03, 2020 at 08:52:41AM +0800, Xiaoyao Li wrote:
> On 9/3/2020 6:44 AM, Sean Christopherson wrote:
> > On Tue, Sep 01, 2020 at 10:43:12AM +0200, Vitaly Kuznetsov wrote:
> > > > +			vcpu->arch.bus_lock_detected = true;
> > > > +		} else {
> > > > +			vcpu->arch.bus_lock_detected = false;
> > > 
> > > This is a fast path so I'm wondering if we can move bus_lock_detected
> > > clearing somewhere else.
> > 
> > Why even snapshot vmx->exit_reason.bus_lock_detected?  I don't see any
> > reason why vcpu_enter_guest() needs to handle the exit to userspace, e.g.
> > it's just as easily handled in VMX code.
> 
> Because we want to handle the exit to userspace only in one place, i.e.,
> after kvm_x86_ops.handle_exit(vcpu, exit_fastpath). Otherwise, we would have
> to check vmx->exit_reason.bus_lock_detected in every other handler, at least
> in those can preempt the bus lock VM-exit theoretically.

That's not hard to do in vmx.c, e.g.

diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index 0316b86bad43a..ea2fed7f21565 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -5677,7 +5677,7 @@ void dump_vmcs(void)
  * The guest has exited.  See if we can fix it or if we need userspace
  * assistance.
  */
-static int vmx_handle_exit(struct kvm_vcpu *vcpu, fastpath_t exit_fastpath)
+static int __vmx_handle_exit(struct kvm_vcpu *vcpu, fastpath_t exit_fastpath)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
        u32 exit_reason = vmx->exit_reason;
@@ -5822,6 +5822,22 @@ static int vmx_handle_exit(struct kvm_vcpu *vcpu, fastpath_t exit_fastpath)
        return 0;
 }

+static int vmx_handle_exit(struct kvm_vcpu *vcpu, fastpath_t exit_fastpath)
+{
+       int ret = __vmx_handle_exit(vcpu, exit_fastpath);
+
+       if (vmx->exit_reason.bus_lock_detected) {
+               if (ret)
+                       vcpu->run->exit_reason = KVM_EXIT_BUS_LOCK;
+               else
+                       vcpu->run->flags |= KVM_RUN_BUS_LOCK;
+               return 0;
+       }
+
+       vcpu->run->flags &= ~KVM_RUN_BUS_LOCK;
+       return ret;
+}
+
 /*
  * Software based L1D cache flush which is used when microcode providing
  * the cache control MSR is not loaded.