diff mbox series

[v4] KVM: emulate: Don't inject #GP when emulating RDMPC if CR0.PE=0

Message ID 1634719951-73285-1-git-send-email-wanpengli@tencent.com (mailing list archive)
State New, archived
Headers show
Series [v4] KVM: emulate: Don't inject #GP when emulating RDMPC if CR0.PE=0 | expand

Commit Message

Wanpeng Li Oct. 20, 2021, 8:52 a.m. UTC 
From: Wanpeng Li <wanpengli@tencent.com>

SDM mentioned that we should #GP for rdpmc if ECX is not valid or 
(CR4.PCE is 0 and CPL is 1, 2, or 3 and CR0.PE is 1).

Let's add the CR0.PE is 1 checking to rdpmc emulate, though this isn't
strictly necessary since it's impossible for CPL to be >0 if CR0.PE=0.

Reviewed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
---
v3 -> v4:
 * add comments instead of pseudocode
v2 -> v3:
 * add the missing 'S'
v1 -> v2:
 * update patch description

 arch/x86/kvm/emulate.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

Paolo Bonzini Oct. 20, 2021, 9:01 a.m. UTC | #1 
On 20/10/21 10:52, Wanpeng Li wrote:
> From: Wanpeng Li<wanpengli@tencent.com>
> 
> SDM mentioned that we should #GP for rdpmc if ECX is not valid or
> (CR4.PCE is 0 and CPL is 1, 2, or 3 and CR0.PE is 1).
> 
> Let's add the CR0.PE is 1 checking to rdpmc emulate, though this isn't
> strictly necessary since it's impossible for CPL to be >0 if CR0.PE=0.
> 
> Reviewed-by: Sean Christopherson<seanjc@google.com>
> Signed-off-by: Wanpeng Li<wanpengli@tencent.com>
> ---
> v3 -> v4:
>   * add comments instead of pseudocode

No, the commit message was fine.  What I meant is there's no need to 
change the code.  Just add a comment about why CR0.PE isn't tested.

Paolo
Wanpeng Li Oct. 20, 2021, 10:15 a.m. UTC | #2 
On Wed, 20 Oct 2021 at 17:02, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> On 20/10/21 10:52, Wanpeng Li wrote:
> > From: Wanpeng Li<wanpengli@tencent.com>
> >
> > SDM mentioned that we should #GP for rdpmc if ECX is not valid or
> > (CR4.PCE is 0 and CPL is 1, 2, or 3 and CR0.PE is 1).
> >
> > Let's add the CR0.PE is 1 checking to rdpmc emulate, though this isn't
> > strictly necessary since it's impossible for CPL to be >0 if CR0.PE=0.
> >
> > Reviewed-by: Sean Christopherson<seanjc@google.com>
> > Signed-off-by: Wanpeng Li<wanpengli@tencent.com>
> > ---
> > v3 -> v4:
> >   * add comments instead of pseudocode
>
> No, the commit message was fine.  What I meant is there's no need to
> change the code.  Just add a comment about why CR0.PE isn't tested.

Just sent out a new version.

    Wanpeng
diff mbox series

Patch

diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index 9a144ca8e146..ab7ec569e8c9 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -4213,6 +4213,7 @@  static int check_rdtsc(struct x86_emulate_ctxt *ctxt)
 static int check_rdpmc(struct x86_emulate_ctxt *ctxt)
 {
 	u64 cr4 = ctxt->ops->get_cr(ctxt, 4);
+	u64 cr0 = ctxt->ops->get_cr(ctxt, 0);
 	u64 rcx = reg_read(ctxt, VCPU_REGS_RCX);
 
 	/*
@@ -4222,7 +4223,7 @@  static int check_rdpmc(struct x86_emulate_ctxt *ctxt)
 	if (enable_vmware_backdoor && is_vmware_backdoor_pmc(rcx))
 		return X86EMUL_CONTINUE;
 
-	if ((!(cr4 & X86_CR4_PCE) && ctxt->ops->cpl(ctxt)) ||
+	if ((!(cr4 & X86_CR4_PCE) && ctxt->ops->cpl(ctxt) && (cr0 & X86_CR0_PE)) ||
 	    ctxt->ops->check_pmc(ctxt, rcx))
 		return emulate_gp(ctxt, 0);