KVM: X86: Cache CR3 in prev_roots when PCID is disabled

The commit 21823fbda5522 ("KVM: x86: Invalidate all PGDs for the
current PCID on MOV CR3 w/ flush") invalidates all PGDs for the specific
PCID and in the case of PCID is disabled, it includes all PGDs in the
prev_roots and the commit made prev_roots totally unused in this case.

Not using prev_roots fixes a problem when CR4.PCIDE is changed 0 -> 1
before the said commit:

	(CR4.PCIDE=0, CR4.PGE=1; CR3=cr3_a; the page for the guest
	 RIP is global; cr3_b is cached in prev_roots)

	modify page tables under cr3_b
		the shadow root of cr3_b is unsync in kvm
	INVPCID single context
		the guest expects the TLB is clean for PCID=0
	change CR4.PCIDE 0 -> 1
	switch to cr3_b with PCID=0,NOFLUSH=1
		No sync in kvm, cr3_b is still unsync in kvm
	jump to the page that was modified in step 1
		shadow page tables point to the wrong page

It is a very unlikely case, but it shows that stale prev_roots can be
a problem after CR4.PCIDE changes from 0 to 1.  However, to fix this
case, the commit disabled caching CR3 in prev_roots altogether when PCID
is disabled.  Not all CPUs have PCID; especially the PCID support
for AMD CPUs is kind of recent.  To restore the prev_roots optimization
for CR4.PCIDE=0, flush the whole MMU (including all prev_roots) when
CR4.PCIDE changes.

Signed-off-by: Lai Jiangshan <laijs@linux.alibaba.com>
Message-Id: <20211019110154.4091-3-jiangshanlai@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

1 files changed, 27 insertions, 2 deletions

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index f7806e3f3019..9d25ef7d4d53 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1022,10 +1022,27 @@ EXPORT_SYMBOL_GPL(kvm_is_valid_cr4);
 
 void kvm_post_set_cr4(struct kvm_vcpu *vcpu, unsigned long old_cr4, unsigned long cr4)
 {
+	/*
+	 * If any role bit is changed, the MMU needs to be reset.
+	 *
+	 * If CR4.PCIDE is changed 1 -> 0, the guest TLB must be flushed.
+	 * If CR4.PCIDE is changed 0 -> 1, there is no need to flush the TLB
+	 * according to the SDM; however, stale prev_roots could be reused
+	 * incorrectly in the future after a MOV to CR3 with NOFLUSH=1, so we
+	 * free them all.  KVM_REQ_MMU_RELOAD is fit for the both cases; it
+	 * is slow, but changing CR4.PCIDE is a rare case.
+	 *
+	 * If CR4.PGE is changed, the guest TLB must be flushed.
+	 *
+	 * Note: resetting MMU is a superset of KVM_REQ_MMU_RELOAD and
+	 * KVM_REQ_MMU_RELOAD is a superset of KVM_REQ_TLB_FLUSH_GUEST, hence
+	 * the usage of "else if".
+	 */
 	if ((cr4 ^ old_cr4) & KVM_MMU_CR4_ROLE_BITS)
 		kvm_mmu_reset_context(vcpu);
-	else if (((cr4 ^ old_cr4) & X86_CR4_PGE) ||
-		 (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE)))
+	else if ((cr4 ^ old_cr4) & X86_CR4_PCIDE)
+		kvm_make_request(KVM_REQ_MMU_RELOAD, vcpu);
+	else if ((cr4 ^ old_cr4) & X86_CR4_PGE)
 		kvm_make_request(KVM_REQ_TLB_FLUSH_GUEST, vcpu);
 }
 EXPORT_SYMBOL_GPL(kvm_post_set_cr4);
@@ -1095,6 +1112,14 @@ static void kvm_invalidate_pcid(struct kvm_vcpu *vcpu, unsigned long pcid)
 		kvm_make_request(KVM_REQ_TLB_FLUSH_CURRENT, vcpu);
 	}
 
+	/*
+	 * If PCID is disabled, there is no need to free prev_roots even if the
+	 * PCIDs for them are also 0, because MOV to CR3 always flushes the TLB
+	 * with PCIDE=0.
+	 */
+	if (!kvm_read_cr4_bits(vcpu, X86_CR4_PCIDE))
+		return;
+
 	for (i = 0; i < KVM_MMU_NUM_PREV_ROOTS; i++)
 		if (kvm_get_pcid(vcpu, mmu->prev_roots[i].pgd) == pcid)
 			roots_to_free |= KVM_MMU_ROOT_PREVIOUS(i);