[PATCH 4.19 1/2] KVM: do not assume PTE is writable after follow_pfn

[PATCH 4.19 1/2] KVM: do not assume PTE is writable after follow_pfn

[Ovidiu Panait]

From: Paolo Bonzini <pbonzini@redhat.com>

commit bd2fae8da794b55bf2ac02632da3a151b10e664c upstream.

In order to convert an HVA to a PFN, KVM usually tries to use
the get_user_pages family of functinso.  This however is not
possible for VM_IO vmas; in that case, KVM instead uses follow_pfn.

In doing this however KVM loses the information on whether the
PFN is writable.  That is usually not a problem because the main
use of VM_IO vmas with KVM is for BARs in PCI device assignment,
however it is a bug.  To fix it, use follow_pte and check pte_write
while under the protection of the PTE lock.  The information can
be used to fail hva_to_pfn_remapped or passed back to the
caller via *writable.

Usage of follow_pfn was introduced in commit add6a0cd1c5b ("KVM: MMU: try to fix
up page faults before giving up", 2016-07-05); however, even older version
have the same issue, all the way back to commit 2e2e3738af33 ("KVM:
Handle vma regions with no backing page", 2008-07-20), as they also did
not check whether the PFN was writable.

Fixes: 2e2e3738af33 ("KVM: Handle vma regions with no backing page")
Reported-by: David Stevens <stevensd@google.com>
Cc: 3pvd@google.com
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
[OP: backport to 4.19, adjust follow_pte() -> follow_pte_pmd()]
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
---
 virt/kvm/kvm_main.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 1ecb27b3421a..6aeac96bf147 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1495,9 +1495,11 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
 			       kvm_pfn_t *p_pfn)
 {
 	unsigned long pfn;
+	pte_t *ptep;
+	spinlock_t *ptl;
 	int r;
 
-	r = follow_pfn(vma, addr, &pfn);
+	r = follow_pte_pmd(vma->vm_mm, addr, NULL, NULL, &ptep, NULL, &ptl);
 	if (r) {
 		/*
 		 * get_user_pages fails for VM_IO and VM_PFNMAP vmas and does
@@ -1512,14 +1514,19 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
 		if (r)
 			return r;
 
-		r = follow_pfn(vma, addr, &pfn);
+		r = follow_pte_pmd(vma->vm_mm, addr, NULL, NULL, &ptep, NULL, &ptl);
 		if (r)
 			return r;
+	}
 
+	if (write_fault && !pte_write(*ptep)) {
+		pfn = KVM_PFN_ERR_RO_FAULT;
+		goto out;
 	}
 
 	if (writable)
-		*writable = true;
+		*writable = pte_write(*ptep);
+	pfn = pte_pfn(*ptep);
 
 	/*
 	 * Get a reference here because callers of *hva_to_pfn* and
@@ -1534,6 +1541,8 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
 	 */ 
 	kvm_get_pfn(pfn);
 
+out:
+	pte_unmap_unlock(ptep, ptl);
 	*p_pfn = pfn;
 	return 0;
 }
-- 
2.25.1

[PATCH 4.19 2/2] KVM: do not allow mapping valid but non-reference-counted pages

[Ovidiu Panait]

From: Nicholas Piggin <npiggin@gmail.com>

commit f8be156be163a052a067306417cd0ff679068c97 upstream.

It's possible to create a region which maps valid but non-refcounted
pages (e.g., tail pages of non-compound higher order allocations). These
host pages can then be returned by gfn_to_page, gfn_to_pfn, etc., family
of APIs, which take a reference to the page, which takes it from 0 to 1.
When the reference is dropped, this will free the page incorrectly.

Fix this by only taking a reference on valid pages if it was non-zero,
which indicates it is participating in normal refcounting (and can be
released with put_page).

This addresses CVE-2021-22543.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Tested-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
---
 virt/kvm/kvm_main.c | 19 +++++++++++++++++--
 1 file changed, 17 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 6aeac96bf147..3559eba5f502 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1489,6 +1489,13 @@ static bool vma_is_valid(struct vm_area_struct *vma, bool write_fault)
 	return true;
 }
 
+static int kvm_try_get_pfn(kvm_pfn_t pfn)
+{
+	if (kvm_is_reserved_pfn(pfn))
+		return 1;
+	return get_page_unless_zero(pfn_to_page(pfn));
+}
+
 static int hva_to_pfn_remapped(struct vm_area_struct *vma,
 			       unsigned long addr, bool *async,
 			       bool write_fault, bool *writable,
@@ -1538,13 +1545,21 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
 	 * Whoever called remap_pfn_range is also going to call e.g.
 	 * unmap_mapping_range before the underlying pages are freed,
 	 * causing a call to our MMU notifier.
+	 *
+	 * Certain IO or PFNMAP mappings can be backed with valid
+	 * struct pages, but be allocated without refcounting e.g.,
+	 * tail pages of non-compound higher order allocations, which
+	 * would then underflow the refcount when the caller does the
+	 * required put_page. Don't allow those pages here.
 	 */ 
-	kvm_get_pfn(pfn);
+	if (!kvm_try_get_pfn(pfn))
+		r = -EFAULT;
 
 out:
 	pte_unmap_unlock(ptep, ptl);
 	*p_pfn = pfn;
-	return 0;
+
+	return r;
 }
 
 /*
-- 
2.25.1

Re: [PATCH 4.19 1/2] KVM: do not assume PTE is writable after follow_pfn

[Paolo Bonzini]

On 23/07/21 22:11, Ovidiu Panait wrote:
> From: Paolo Bonzini <pbonzini@redhat.com>
> 
> commit bd2fae8da794b55bf2ac02632da3a151b10e664c upstream.
> 
> In order to convert an HVA to a PFN, KVM usually tries to use
> the get_user_pages family of functinso.  This however is not
> possible for VM_IO vmas; in that case, KVM instead uses follow_pfn.
> 
> In doing this however KVM loses the information on whether the
> PFN is writable.  That is usually not a problem because the main
> use of VM_IO vmas with KVM is for BARs in PCI device assignment,
> however it is a bug.  To fix it, use follow_pte and check pte_write
> while under the protection of the PTE lock.  The information can
> be used to fail hva_to_pfn_remapped or passed back to the
> caller via *writable.
> 
> Usage of follow_pfn was introduced in commit add6a0cd1c5b ("KVM: MMU: try to fix
> up page faults before giving up", 2016-07-05); however, even older version
> have the same issue, all the way back to commit 2e2e3738af33 ("KVM:
> Handle vma regions with no backing page", 2008-07-20), as they also did
> not check whether the PFN was writable.
> 
> Fixes: 2e2e3738af33 ("KVM: Handle vma regions with no backing page")
> Reported-by: David Stevens <stevensd@google.com>
> Cc: 3pvd@google.com
> Cc: Jann Horn <jannh@google.com>
> Cc: Jason Gunthorpe <jgg@ziepe.ca>
> Cc: stable@vger.kernel.org
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> [OP: backport to 4.19, adjust follow_pte() -> follow_pte_pmd()]
> Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
> ---
>   virt/kvm/kvm_main.c | 15 ++++++++++++---
>   1 file changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 1ecb27b3421a..6aeac96bf147 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -1495,9 +1495,11 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
>   			       kvm_pfn_t *p_pfn)
>   {
>   	unsigned long pfn;
> +	pte_t *ptep;
> +	spinlock_t *ptl;
>   	int r;
>   
> -	r = follow_pfn(vma, addr, &pfn);
> +	r = follow_pte_pmd(vma->vm_mm, addr, NULL, NULL, &ptep, NULL, &ptl);
>   	if (r) {
>   		/*
>   		 * get_user_pages fails for VM_IO and VM_PFNMAP vmas and does
> @@ -1512,14 +1514,19 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
>   		if (r)
>   			return r;
>   
> -		r = follow_pfn(vma, addr, &pfn);
> +		r = follow_pte_pmd(vma->vm_mm, addr, NULL, NULL, &ptep, NULL, &ptl);
>   		if (r)
>   			return r;
> +	}
>   
> +	if (write_fault && !pte_write(*ptep)) {
> +		pfn = KVM_PFN_ERR_RO_FAULT;
> +		goto out;
>   	}
>   
>   	if (writable)
> -		*writable = true;
> +		*writable = pte_write(*ptep);
> +	pfn = pte_pfn(*ptep);
>   
>   	/*
>   	 * Get a reference here because callers of *hva_to_pfn* and
> @@ -1534,6 +1541,8 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
>   	 */
>   	kvm_get_pfn(pfn);
>   
> +out:
> +	pte_unmap_unlock(ptep, ptl);
>   	*p_pfn = pfn;
>   	return 0;
>   }
> 

Acked-by: Paolo Bonzini <pbonzini@redhat.com>

Re: [PATCH 4.19 2/2] KVM: do not allow mapping valid but non-reference-counted pages

[Paolo Bonzini]

On 23/07/21 22:11, Ovidiu Panait wrote:
> From: Nicholas Piggin <npiggin@gmail.com>
> 
> commit f8be156be163a052a067306417cd0ff679068c97 upstream.
> 
> It's possible to create a region which maps valid but non-refcounted
> pages (e.g., tail pages of non-compound higher order allocations). These
> host pages can then be returned by gfn_to_page, gfn_to_pfn, etc., family
> of APIs, which take a reference to the page, which takes it from 0 to 1.
> When the reference is dropped, this will free the page incorrectly.
> 
> Fix this by only taking a reference on valid pages if it was non-zero,
> which indicates it is participating in normal refcounting (and can be
> released with put_page).
> 
> This addresses CVE-2021-22543.
> 
> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> Tested-by: Paolo Bonzini <pbonzini@redhat.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
> ---
>   virt/kvm/kvm_main.c | 19 +++++++++++++++++--
>   1 file changed, 17 insertions(+), 2 deletions(-)
> 
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 6aeac96bf147..3559eba5f502 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -1489,6 +1489,13 @@ static bool vma_is_valid(struct vm_area_struct *vma, bool write_fault)
>   	return true;
>   }
>   
> +static int kvm_try_get_pfn(kvm_pfn_t pfn)
> +{
> +	if (kvm_is_reserved_pfn(pfn))
> +		return 1;
> +	return get_page_unless_zero(pfn_to_page(pfn));
> +}
> +
>   static int hva_to_pfn_remapped(struct vm_area_struct *vma,
>   			       unsigned long addr, bool *async,
>   			       bool write_fault, bool *writable,
> @@ -1538,13 +1545,21 @@ static int hva_to_pfn_remapped(struct vm_area_struct *vma,
>   	 * Whoever called remap_pfn_range is also going to call e.g.
>   	 * unmap_mapping_range before the underlying pages are freed,
>   	 * causing a call to our MMU notifier.
> +	 *
> +	 * Certain IO or PFNMAP mappings can be backed with valid
> +	 * struct pages, but be allocated without refcounting e.g.,
> +	 * tail pages of non-compound higher order allocations, which
> +	 * would then underflow the refcount when the caller does the
> +	 * required put_page. Don't allow those pages here.
>   	 */
> -	kvm_get_pfn(pfn);
> +	if (!kvm_try_get_pfn(pfn))
> +		r = -EFAULT;
>   
>   out:
>   	pte_unmap_unlock(ptep, ptl);
>   	*p_pfn = pfn;
> -	return 0;
> +
> +	return r;
>   }
>   
>   /*
> 

Acked-by: Paolo Bonzini <pbonzini@redhat.com>

Re: [PATCH 4.19 1/2] KVM: do not assume PTE is writable after follow_pfn

[Greg KH]

On Fri, Jul 23, 2021 at 11:11:33PM +0300, Ovidiu Panait wrote:
> From: Paolo Bonzini <pbonzini@redhat.com>
> 
> commit bd2fae8da794b55bf2ac02632da3a151b10e664c upstream.
> 

Both now queued up, thanks!

greg k-h