[PATCH 5.4 41/71] KVM: do not allow mapping valid but non-reference-counted pages

From: Sasha Levin
Date: Mon Jun 28 2021 - 10:47:54 EST


From: Nicholas Piggin <npiggin@xxxxxxxxx>

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@xxxxxxxxx>
Tested-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>
Cc: stable@xxxxxxxxxxxxxxx
Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
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 f83fa0aeeb45..b2287e7d3ba4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1593,6 +1593,13 @@ 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,
@@ -1642,13 +1649,21 @@ 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;
}

/*
--
2.30.2