File 0001-gvariant-Fix-checking-arithmetic-for-tuple... of Package glib2.17285

Overview Repositories Revisions Requests Users Attributes Meta

File 0001-gvariant-Fix-checking-arithmetic-for-tuple-element-e.patch of Package glib2.17285

From 7ec1e7e7c8e085454927bb102faaad0c74cf5966 Mon Sep 17 00:00:00 2001
From: Philip Withnall <withnall@endlessm.com>
Date: Thu, 16 Aug 2018 20:12:02 +0100
Subject: [PATCH 01/15] gvariant: Fix checking arithmetic for tuple element
 ends
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

When checking whether a serialised GVariant tuple is in normal form,
it’s possible for `offset_ptr -= offset_size` to underflow and wrap
around, resulting in gvs_read_unaligned_le() reading memory outside the
serialised GVariant bounds.

See §(Tuples) in gvariant-serialiser.c for the documentation on how
tuples are serialised. Briefly, all variable-length elements in the
tuple have an offset to their end stored in an array of offsets at the
end of the tuple. The width of each offset is in offset_size. offset_ptr
is added to the start of the serialised tuple to get the offset which is
currently being examined. The offset array is in reverse order compared
to the tuple elements, hence the subtraction.

The bug can be triggered if a tuple contains a load of variable-length
elements, each of whose length is actually zero (i.e. empty arrays).

Includes a unit test.

oss-fuzz#9801

Signed-off-by: Philip Withnall <withnall@endlessm.com>
---
 glib/gvariant-serialiser.c |  3 +++
 glib/tests/gvariant.c      | 28 ++++++++++++++++++++++++++++
 2 files changed, 31 insertions(+)

diff --git a/glib/gvariant-serialiser.c b/glib/gvariant-serialiser.c
index 69f183121..96df54e23 100644
--- a/glib/gvariant-serialiser.c
+++ b/glib/gvariant-serialiser.c
@@ -1065,6 +1065,9 @@ gvs_tuple_is_normal (GVariantSerialised value)
           break;
 
         case G_VARIANT_MEMBER_ENDING_OFFSET:
+          if (offset_ptr < offset_size)
+            return FALSE;
+
           offset_ptr -= offset_size;
 
           if (offset_ptr < offset)
diff --git a/glib/tests/gvariant.c b/glib/tests/gvariant.c
index de8e42d0b..a5095a380 100644
--- a/glib/tests/gvariant.c
+++ b/glib/tests/gvariant.c
@@ -4631,6 +4631,30 @@ test_stack_dict_init (void)
   g_variant_unref (variant);
 }
 
+/* Test checking arbitrary binary data for normal form. This time, it’s a tuple
+ * with invalid element ends. */
+static void
+test_normal_checking_tuples (void)
+{
+  const guint8 data[] = {
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0x00,
+    'a', '(', 'a', 'o', 'a', 'o', 'a', 'a', 'o', 'a', 'a', 'o', ')'
+  };
+  gsize size = sizeof (data);
+  GVariant *variant = NULL;
+  GVariant *normal_variant = NULL;
+
+  variant = g_variant_new_from_data (G_VARIANT_TYPE_VARIANT, data, size,
+                                     FALSE, NULL, NULL);
+  g_assert_nonnull (variant);
+
+  normal_variant = g_variant_get_normal_form (variant);
+  g_assert_nonnull (normal_variant);
+
+  g_variant_unref (normal_variant);
+  g_variant_unref (variant);
+}
+
 int
 main (int argc, char **argv)
 {
@@ -4692,5 +4716,9 @@ main (int argc, char **argv)
 
   g_test_add_func ("/gvariant/stack-builder-init", test_stack_builder_init);
   g_test_add_func ("/gvariant/stack-dict-init", test_stack_dict_init);
+
+  g_test_add_func ("/gvariant/normal-checking/tuples",
+                   test_normal_checking_tuples);
+
   return g_test_run ();
 }
-- 
2.14.4

Places

File 0001-gvariant-Fix-checking-arithmetic-for-tuple-element-e.patch of Package glib2.17285

Places