From: Ladi Prosek <lprosek@redhat.com>
To: kvm@vger.kernel.org
Cc: rkrcmar@redhat.com, pbonzini@redhat.com
Subject: [PATCH v3 4/6] KVM: nVMX: treat CR4.VMXE as reserved in SMM
Date: Mon, 25 Sep 2017 10:09:02 +0200 [thread overview]
Message-ID: <20170925080904.24850-5-lprosek@redhat.com> (raw)
In-Reply-To: <20170925080904.24850-1-lprosek@redhat.com>
Intel SDM 34.14.3 Protection of CR4.VMXE in SMM:
"Under the default treatment, CR4.VMXE is treated as a reserved bit while a
logical processor is in SMM. Any attempt by software running in SMM to set
this bit causes a general-protection exception."
em_rsm() may set CR4.VMXE as part of its loading of saved SMM state so care
must be taken to order this properly (HF_SMM_MASK is required to access the
state-save area but must be cleared before setting CR4.VMXE).
Signed-off-by: Ladi Prosek <lprosek@redhat.com>
---
arch/x86/kvm/emulate.c | 18 +++++++++++++++---
arch/x86/kvm/vmx.c | 4 ++++
2 files changed, 19 insertions(+), 3 deletions(-)
diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index 15f527b44aa7..f12bf51d379a 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -2479,6 +2479,9 @@ static int rsm_load_state_32(struct x86_emulate_ctxt *ctxt, u64 smbase)
ctxt->ops->set_smbase(ctxt, GET_SMSTATE(u32, smbase, 0x7ef8));
+ ctxt->ops->set_hflags(ctxt, ctxt->ops->get_hflags(ctxt) &
+ ~X86EMUL_SMM_MASK);
+
return rsm_enter_protected_mode(ctxt, cr0, cr4);
}
@@ -2531,7 +2534,8 @@ static int rsm_load_state_64(struct x86_emulate_ctxt *ctxt, u64 smbase)
dt.address = GET_SMSTATE(u64, smbase, 0x7e68);
ctxt->ops->set_gdt(ctxt, &dt);
- r = rsm_enter_protected_mode(ctxt, cr0, cr4);
+ /* We're still in SMM so CR4.VMXE is reserved. */
+ r = rsm_enter_protected_mode(ctxt, cr0, cr4 & ~X86_CR4_VMXE);
if (r != X86EMUL_CONTINUE)
return r;
@@ -2541,6 +2545,13 @@ static int rsm_load_state_64(struct x86_emulate_ctxt *ctxt, u64 smbase)
return r;
}
+ /* Out of SMM now and finish off CR4. */
+ ctxt->ops->set_hflags(ctxt, ctxt->ops->get_hflags(ctxt) &
+ ~X86EMUL_SMM_MASK);
+
+ if (ctxt->ops->set_cr(ctxt, 4, cr4))
+ return X86EMUL_UNHANDLEABLE;
+
return X86EMUL_CONTINUE;
}
@@ -2601,9 +2612,10 @@ static int em_rsm(struct x86_emulate_ctxt *ctxt)
if ((ctxt->ops->get_hflags(ctxt) & X86EMUL_SMM_INSIDE_NMI_MASK) == 0)
ctxt->ops->set_nmi_mask(ctxt, false);
+ else
+ ctxt->ops->set_hflags(ctxt, ctxt->ops->get_hflags(ctxt) &
+ ~X86EMUL_SMM_INSIDE_NMI_MASK);
- ctxt->ops->set_hflags(ctxt, ctxt->ops->get_hflags(ctxt) &
- ~(X86EMUL_SMM_INSIDE_NMI_MASK | X86EMUL_SMM_MASK));
return X86EMUL_CONTINUE;
}
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index be4b5f9a84bf..7e350b8d724d 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -4366,6 +4366,10 @@ static int vmx_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
*/
if (!nested_vmx_allowed(vcpu))
return 1;
+
+ /* cr4.VMXE is a reserved bit in SMM */
+ if (is_smm(vcpu))
+ return 1;
}
if (to_vmx(vcpu)->nested.vmxon && !nested_cr4_valid(vcpu, cr4))
--
2.13.5
next prev parent reply other threads:[~2017-09-25 8:09 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-25 8:08 [PATCH v3 0/6] KVM: nested virt SMM fixes Ladi Prosek
2017-09-25 8:08 ` [PATCH v3 1/6] KVM: x86: introduce ISA specific SMM entry/exit callbacks Ladi Prosek
2017-09-25 8:09 ` [PATCH v3 2/6] KVM: x86: introduce ISA specific smi_allowed callback Ladi Prosek
2017-09-25 8:09 ` [PATCH v3 3/6] KVM: nVMX: fix SMI injection in guest mode Ladi Prosek
2017-09-25 8:09 ` Ladi Prosek [this message]
2017-09-25 8:09 ` [PATCH v3 5/6] KVM: nSVM: refactor nested_svm_vmrun Ladi Prosek
2017-09-25 8:09 ` [PATCH v3 6/6] KVM: nSVM: fix SMI injection in guest mode Ladi Prosek
2017-10-03 19:53 ` Radim Krčmář
2017-10-04 10:10 ` Ladi Prosek
2017-10-04 14:42 ` Radim Krčmář
2017-10-10 8:03 ` Ladi Prosek
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170925080904.24850-5-lprosek@redhat.com \
--to=lprosek@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=rkrcmar@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox