Fixed CVE-2021-45115 -- Prevented DoS vector in UserAttributeSimilarityValidator.

Thanks Chris Bailey for the report.

Co-authored-by: Adam Johnson <me@adamj.eu>

1 files changed, 38 insertions, 2 deletions

diff --git a/django/contrib/auth/password_validation.py b/django/contrib/auth/password_validation.py
index bc70a84543..2e64bf1586 100644
--- a/django/contrib/auth/password_validation.py
+++ b/django/contrib/auth/password_validation.py
@@ -115,6 +115,36 @@ class MinimumLengthValidator:
         ) % {'min_length': self.min_length}
 
 
+def exceeds_maximum_length_ratio(password, max_similarity, value):
+    """
+    Test that value is within a reasonable range of password.
+
+    The following ratio calculations are based on testing SequenceMatcher like
+    this:
+
+    for i in range(0,6):
+      print(10**i, SequenceMatcher(a='A', b='A'*(10**i)).quick_ratio())
+
+    which yields:
+
+    1 1.0
+    10 0.18181818181818182
+    100 0.019801980198019802
+    1000 0.001998001998001998
+    10000 0.00019998000199980003
+    100000 1.999980000199998e-05
+
+    This means a length_ratio of 10 should never yield a similarity higher than
+    0.2, for 100 this is down to 0.02 and for 1000 it is 0.002. This can be
+    calculated via 2 / length_ratio. As a result we avoid the potentially
+    expensive sequence matching.
+    """
+    pwd_len = len(password)
+    length_bound_similarity = max_similarity / 2 * pwd_len
+    value_len = len(value)
+    return pwd_len >= 10 * value_len and value_len < length_bound_similarity
+
+
 class UserAttributeSimilarityValidator:
     """
     Validate that the password is sufficiently different from the user's
@@ -130,19 +160,25 @@ class UserAttributeSimilarityValidator:
 
     def __init__(self, user_attributes=DEFAULT_USER_ATTRIBUTES, max_similarity=0.7):
         self.user_attributes = user_attributes
+        if max_similarity < 0.1:
+            raise ValueError('max_similarity must be at least 0.1')
         self.max_similarity = max_similarity
 
     def validate(self, password, user=None):
         if not user:
             return
 
+        password = password.lower()
         for attribute_name in self.user_attributes:
             value = getattr(user, attribute_name, None)
             if not value or not isinstance(value, str):
                 continue
-            value_parts = re.split(r'\W+', value) + [value]
+            value_lower = value.lower()
+            value_parts = re.split(r'\W+', value_lower) + [value_lower]
             for value_part in value_parts:
-                if SequenceMatcher(a=password.lower(), b=value_part.lower()).quick_ratio() >= self.max_similarity:
+                if exceeds_maximum_length_ratio(password, self.max_similarity, value_part):
+                    continue
+                if SequenceMatcher(a=password, b=value_part).quick_ratio() >= self.max_similarity:
                     try:
                         verbose_name = str(user._meta.get_field(attribute_name).verbose_name)
                     except FieldDoesNotExist: