KVM: X86: Don't track dirty for KVM_SET_[TSS_ADDR|IDENTITY_MAP_ADDR]

Originally, we have three code paths that can dirty a page without
vcpu context for X86:

  - init_rmode_identity_map
  - init_rmode_tss
  - kvmgt_rw_gpa

init_rmode_identity_map and init_rmode_tss will be setup on
destination VM no matter what (and the guest cannot even see them), so
it does not make sense to track them at all.

To do this, allow __x86_set_memory_region() to return the userspace
address that just allocated to the caller.  Then in both of the
functions we directly write to the userspace address instead of
calling kvm_write_*() APIs.

Another trivial change is that we don't need to explicitly clear the
identity page table root in init_rmode_identity_map() because no
matter what we'll write to the whole page with 4M huge page entries.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20201001012044.5151-4-peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

1 files changed, 42 insertions, 46 deletions

diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index fd70bdf63844..46b32aa43811 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -3515,42 +3515,33 @@ bool __vmx_guest_state_valid(struct kvm_vcpu *vcpu)
 	return true;
 }
 
-static int init_rmode_tss(struct kvm *kvm)
+static int init_rmode_tss(struct kvm *kvm, void __user *ua)
 {
-	gfn_t fn;
-	u16 data = 0;
-	int idx, r;
+	const void *zero_page = (const void *) __va(page_to_phys(ZERO_PAGE(0)));
+	u16 data;
+	int i;
+
+	for (i = 0; i < 3; i++) {
+		if (__copy_to_user(ua + PAGE_SIZE * i, zero_page, PAGE_SIZE))
+			return -EFAULT;
+	}
 
-	idx = srcu_read_lock(&kvm->srcu);
-	fn = to_kvm_vmx(kvm)->tss_addr >> PAGE_SHIFT;
-	r = kvm_clear_guest_page(kvm, fn, 0, PAGE_SIZE);
-	if (r < 0)
-		goto out;
 	data = TSS_BASE_SIZE + TSS_REDIRECTION_SIZE;
-	r = kvm_write_guest_page(kvm, fn++, &data,
-			TSS_IOPB_BASE_OFFSET, sizeof(u16));
-	if (r < 0)
-		goto out;
-	r = kvm_clear_guest_page(kvm, fn++, 0, PAGE_SIZE);
-	if (r < 0)
-		goto out;
-	r = kvm_clear_guest_page(kvm, fn, 0, PAGE_SIZE);
-	if (r < 0)
-		goto out;
+	if (__copy_to_user(ua + TSS_IOPB_BASE_OFFSET, &data, sizeof(u16)))
+		return -EFAULT;
+
 	data = ~0;
-	r = kvm_write_guest_page(kvm, fn, &data,
-				 RMODE_TSS_SIZE - 2 * PAGE_SIZE - 1,
-				 sizeof(u8));
-out:
-	srcu_read_unlock(&kvm->srcu, idx);
-	return r;
+	if (__copy_to_user(ua + RMODE_TSS_SIZE - 1, &data, sizeof(u8)))
+		return -EFAULT;
+
+	return 0;
 }
 
 static int init_rmode_identity_map(struct kvm *kvm)
 {
 	struct kvm_vmx *kvm_vmx = to_kvm_vmx(kvm);
 	int i, r = 0;
-	kvm_pfn_t identity_map_pfn;
+	void __user *uaddr;
 	u32 tmp;
 
 	/* Protect kvm_vmx->ept_identity_pagetable_done. */
@@ -3561,24 +3552,24 @@ static int init_rmode_identity_map(struct kvm *kvm)
 
 	if (!kvm_vmx->ept_identity_map_addr)
 		kvm_vmx->ept_identity_map_addr = VMX_EPT_IDENTITY_PAGETABLE_ADDR;
-	identity_map_pfn = kvm_vmx->ept_identity_map_addr >> PAGE_SHIFT;
 
-	r = __x86_set_memory_region(kvm, IDENTITY_PAGETABLE_PRIVATE_MEMSLOT,
-				    kvm_vmx->ept_identity_map_addr, PAGE_SIZE);
-	if (r < 0)
+	uaddr = __x86_set_memory_region(kvm,
+					IDENTITY_PAGETABLE_PRIVATE_MEMSLOT,
+					kvm_vmx->ept_identity_map_addr,
+					PAGE_SIZE);
+	if (IS_ERR(uaddr)) {
+		r = PTR_ERR(uaddr);
 		goto out;
+	}
 
-	r = kvm_clear_guest_page(kvm, identity_map_pfn, 0, PAGE_SIZE);
-	if (r < 0)
-		goto out;
 	/* Set up identity-mapping pagetable for EPT in real mode */
 	for (i = 0; i < PT32_ENT_PER_PAGE; i++) {
 		tmp = (i << 22) + (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER |
 			_PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_PSE);
-		r = kvm_write_guest_page(kvm, identity_map_pfn,
-				&tmp, i * sizeof(tmp), sizeof(tmp));
-		if (r < 0)
+		if (__copy_to_user(uaddr + i * sizeof(tmp), &tmp, sizeof(tmp))) {
+			r = -EFAULT;
 			goto out;
+		}
 	}
 	kvm_vmx->ept_identity_pagetable_done = true;
 
@@ -3605,19 +3596,22 @@ static void seg_setup(int seg)
 static int alloc_apic_access_page(struct kvm *kvm)
 {
 	struct page *page;
-	int r = 0;
+	void __user *hva;
+	int ret = 0;
 
 	mutex_lock(&kvm->slots_lock);
 	if (kvm->arch.apic_access_page_done)
 		goto out;
-	r = __x86_set_memory_region(kvm, APIC_ACCESS_PAGE_PRIVATE_MEMSLOT,
-				    APIC_DEFAULT_PHYS_BASE, PAGE_SIZE);
-	if (r)
+	hva = __x86_set_memory_region(kvm, APIC_ACCESS_PAGE_PRIVATE_MEMSLOT,
+				      APIC_DEFAULT_PHYS_BASE, PAGE_SIZE);
+	if (IS_ERR(hva)) {
+		ret = PTR_ERR(hva);
 		goto out;
+	}
 
 	page = gfn_to_page(kvm, APIC_DEFAULT_PHYS_BASE >> PAGE_SHIFT);
 	if (is_error_page(page)) {
-		r = -EFAULT;
+		ret = -EFAULT;
 		goto out;
 	}
 
@@ -3629,7 +3623,7 @@ static int alloc_apic_access_page(struct kvm *kvm)
 	kvm->arch.apic_access_page_done = true;
 out:
 	mutex_unlock(&kvm->slots_lock);
-	return r;
+	return ret;
 }
 
 int allocate_vpid(void)
@@ -4638,7 +4632,7 @@ static int vmx_interrupt_allowed(struct kvm_vcpu *vcpu, bool for_injection)
 
 static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr)
 {
-	int ret;
+	void __user *ret;
 
 	if (enable_unrestricted_guest)
 		return 0;
@@ -4648,10 +4642,12 @@ static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr)
 				      PAGE_SIZE * 3);
 	mutex_unlock(&kvm->slots_lock);
 
-	if (ret)
-		return ret;
+	if (IS_ERR(ret))
+		return PTR_ERR(ret);
+
 	to_kvm_vmx(kvm)->tss_addr = addr;
-	return init_rmode_tss(kvm);
+
+	return init_rmode_tss(kvm, ret);
 }
 
 static int vmx_set_identity_map_addr(struct kvm *kvm, u64 ident_addr)