kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: Sean Christopherson <seanjc@google.com>,
	Vitaly Kuznetsov <vkuznets@redhat.com>,
	Wanpeng Li <wanpengli@tencent.com>,
	Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
	kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	Peter Gonda <pgonda@google.com>,
	Maxim Levitsky <mlevitsk@redhat.com>
Subject: [PATCH 2/2] KVM: x86: Allow userspace to update tracked sregs for protected guests
Date: Fri,  7 May 2021 09:59:47 -0700	[thread overview]
Message-ID: <20210507165947.2502412-3-seanjc@google.com> (raw)
In-Reply-To: <20210507165947.2502412-1-seanjc@google.com>

Allow userspace to set CR0, CR4, CR8, and EFER via KVM_SET_SREGS for
protected guests, e.g. for SEV-ES guests with an encrypted VMSA.  KVM
tracks the aforementioned registers by trapping guest writes, and also
exposes the values to userspace via KVM_GET_SREGS.  Skipping the regs
in KVM_SET_SREGS prevents userspace from updating KVM's CPU model to
match the known hardware state.

Fixes: 5265713a0737 ("KVM: x86: Update __get_sregs() / __set_sregs() to support SEV-ES")
Reported-by: Peter Gonda <pgonda@google.com>
Cc: stable@vger.kernel.org
Cc: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
---
 arch/x86/kvm/x86.c | 73 ++++++++++++++++++++++++++--------------------
 1 file changed, 42 insertions(+), 31 deletions(-)

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 3bf52ba5f2bb..1b7d0e97c82b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -9963,21 +9963,25 @@ static int __set_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 	if (kvm_set_apic_base(vcpu, &apic_base_msr))
 		goto out;
 
-	if (vcpu->arch.guest_state_protected)
-		goto skip_protected_regs;
+	if (!vcpu->arch.guest_state_protected) {
+		dt.size = sregs->idt.limit;
+		dt.address = sregs->idt.base;
+		static_call(kvm_x86_set_idt)(vcpu, &dt);
+		dt.size = sregs->gdt.limit;
+		dt.address = sregs->gdt.base;
+		static_call(kvm_x86_set_gdt)(vcpu, &dt);
 
-	dt.size = sregs->idt.limit;
-	dt.address = sregs->idt.base;
-	static_call(kvm_x86_set_idt)(vcpu, &dt);
-	dt.size = sregs->gdt.limit;
-	dt.address = sregs->gdt.base;
-	static_call(kvm_x86_set_gdt)(vcpu, &dt);
-
-	vcpu->arch.cr2 = sregs->cr2;
-	mmu_reset_needed |= kvm_read_cr3(vcpu) != sregs->cr3;
-	vcpu->arch.cr3 = sregs->cr3;
-	kvm_register_mark_available(vcpu, VCPU_EXREG_CR3);
+		vcpu->arch.cr2 = sregs->cr2;
+		mmu_reset_needed |= kvm_read_cr3(vcpu) != sregs->cr3;
+		vcpu->arch.cr3 = sregs->cr3;
+		kvm_register_mark_available(vcpu, VCPU_EXREG_CR3);
+	}
 
+	/*
+	 * Writes to CR0, CR4, CR8, and EFER are trapped (after the instruction
+	 * completes) for SEV-EV guests, thus userspace is allowed to set them
+	 * so that KVM's model can be updated to mirror hardware state.
+	 */
 	kvm_set_cr8(vcpu, sregs->cr8);
 
 	mmu_reset_needed |= vcpu->arch.efer != sregs->efer;
@@ -9990,35 +9994,42 @@ static int __set_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 	mmu_reset_needed |= kvm_read_cr4(vcpu) != sregs->cr4;
 	static_call(kvm_x86_set_cr4)(vcpu, sregs->cr4);
 
-	idx = srcu_read_lock(&vcpu->kvm->srcu);
-	if (is_pae_paging(vcpu)) {
+	/*
+	 * PDPTEs, like regular PTEs, are always encrypted, thus reading them
+	 * will return garbage.  Shadow paging, including nested NPT, isn't
+	 * compatible with protected guests, so ignoring the PDPTEs is a-ok.
+	 */
+	if (!vcpu->arch.guest_state_protected && is_pae_paging(vcpu)) {
+		idx = srcu_read_lock(&vcpu->kvm->srcu);
 		load_pdptrs(vcpu, vcpu->arch.walk_mmu, kvm_read_cr3(vcpu));
+		srcu_read_unlock(&vcpu->kvm->srcu, idx);
+
 		mmu_reset_needed = 1;
 	}
-	srcu_read_unlock(&vcpu->kvm->srcu, idx);
 
 	if (mmu_reset_needed)
 		kvm_mmu_reset_context(vcpu);
 
-	kvm_set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
-	kvm_set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
-	kvm_set_segment(vcpu, &sregs->es, VCPU_SREG_ES);
-	kvm_set_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
-	kvm_set_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
-	kvm_set_segment(vcpu, &sregs->ss, VCPU_SREG_SS);
+	if (!vcpu->arch.guest_state_protected) {
+		kvm_set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
+		kvm_set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
+		kvm_set_segment(vcpu, &sregs->es, VCPU_SREG_ES);
+		kvm_set_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
+		kvm_set_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
+		kvm_set_segment(vcpu, &sregs->ss, VCPU_SREG_SS);
 
-	kvm_set_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
-	kvm_set_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);
+		kvm_set_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
+		kvm_set_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);
 
-	update_cr8_intercept(vcpu);
+		update_cr8_intercept(vcpu);
 
-	/* Older userspace won't unhalt the vcpu on reset. */
-	if (kvm_vcpu_is_bsp(vcpu) && kvm_rip_read(vcpu) == 0xfff0 &&
-	    sregs->cs.selector == 0xf000 && sregs->cs.base == 0xffff0000 &&
-	    !is_protmode(vcpu))
-		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
+		/* Older userspace won't unhalt the vcpu on reset. */
+		if (kvm_vcpu_is_bsp(vcpu) && kvm_rip_read(vcpu) == 0xfff0 &&
+		    sregs->cs.selector == 0xf000 &&
+		    sregs->cs.base == 0xffff0000 && !is_protmode(vcpu))
+			vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
+	}
 
-skip_protected_regs:
 	max_bits = KVM_NR_INTERRUPTS;
 	pending_vec = find_first_bit(
 		(const unsigned long *)sregs->interrupt_bitmap, max_bits);
-- 
2.31.1.607.g51e8a6a459-goog


  parent reply	other threads:[~2021-05-07 17:00 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-07 16:59 [PATCH 0/2] KVM: x86: Fixes for SEV-ES state tracking Sean Christopherson
2021-05-07 16:59 ` [PATCH 1/2] KVM: SVM: Update EFER software model on CR0 trap for SEV-ES Sean Christopherson
2021-05-07 23:15   ` Tom Lendacky
2021-05-07 16:59 ` Sean Christopherson [this message]
2021-05-07 23:21   ` [PATCH 2/2] KVM: x86: Allow userspace to update tracked sregs for protected guests Tom Lendacky
2021-05-10 16:10     ` Sean Christopherson
2021-05-10 18:07       ` Tom Lendacky
2021-05-10 21:02         ` Sean Christopherson
2021-05-10 21:23           ` Tom Lendacky
2021-05-10 22:40             ` Sean Christopherson
2021-05-14 14:19       ` Peter Gonda

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=20210507165947.2502412-3-seanjc@google.com \
    --to=seanjc@google.com \
    --cc=jmattson@google.com \
    --cc=joro@8bytes.org \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mlevitsk@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=pgonda@google.com \
    --cc=vkuznets@redhat.com \
    --cc=wanpengli@tencent.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;
as well as URLs for NNTP newsgroup(s).