CVE-2021-29550

TensorFlow is an end-to-end open source platform for machine learning. An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool. This is because the implementation(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L85-L89) computes a divisor quantity by dividing two user controlled values. The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger. Later, these computed values are used as arguments(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L96-L99) to GeneratePoolingSequence(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_pool_common.cc#L100-L108). There, the first computation is a division in a modulo operation. Since output_length can be 0, this results in runtime crashing. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

TensorFlow is an end-to-end open source platform for machine learning. An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool. This is because the implementation(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L85-L89) computes a divisor quantity by dividing two user controlled values. The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger. Later, these computed values are used as arguments(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L96-L99) to GeneratePoolingSequence(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_pool_common.cc#L100-L108). There, the first computation is a division in a modulo operation. Since output_length can be 0, this results in runtime crashing. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

Impact

An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool:

import tensorflow as tf

value = tf.constant([60], shape=[1, 1, 1, 1], dtype=tf.int32)
pooling_ratio = [1.0, 1.0000014345305555, 1.0, 1.0]
pseudo_random = False
overlapping = False
deterministic = False
seed = 0
seed2 = 0

tf.raw_ops.FractionalAvgPool(
  value=value, pooling_ratio=pooling_ratio, pseudo_random=pseudo_random,
  overlapping=overlapping, deterministic=deterministic, seed=seed, seed2=seed2)

This is because the implementation computes a divisor quantity by dividing two user controlled values:

for (int i = 0; i < tensor_in_and_out_dims; ++i) {
  output_size[i] = static_cast<int>(std::floor(input_size[i] / pooling_ratio_[i]));
  DCHECK_GT(output_size[i], 0); 
} 

The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger.

Later, these computed values are used as arguments to GeneratePoolingSequence. There, the first computation is a division in a modulo operation:

std::vector<int64> GeneratePoolingSequence(int input_length, int output_length,
                                           GuardedPhiloxRandom* generator,
                                           bool pseudo_random) {
  ...
  if (input_length % output_length == 0) {
    diff = std::vector<int64>(output_length, input_length / output_length);
  }
  ...
}

Since output_length can be 0, this results in runtime crashing.

Patches

We have patched the issue in GitHub commit 548b5eaf23685d86f722233d8fbc21d0a4aecb96.

The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by Ying Wang and Yakun Zhang of Baidu X-Team.

TensorFlow is an end-to-end open source platform for machine learning. An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool. This is because the implementation(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L85-L89) computes a divisor quantity by dividing two user controlled values. The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger. Later, these computed values are used as arguments(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L96-L99) to GeneratePoolingSequence(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_pool_common.cc#L100-L108). There, the first computation is a division in a modulo operation. Since output_length can be 0, this results in runtime crashing. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

TensorFlow is an end-to-end open source platform for machine learning. An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool. This is because the implementation(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L85-L89) computes a divisor quantity by dividing two user controlled values. The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger. Later, these computed values are used as arguments(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_avg_pool_op.cc#L96-L99) to GeneratePoolingSequence(https://github.com/tensorflow/tensorflow/blob/acc8ee69f5f46f92a3f1f11230f49c6ac266f10c/tensorflow/core/kernels/fractional_pool_common.cc#L100-L108). There, the first computation is a division in a modulo operation. Since output_length can be 0, this results in runtime crashing. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

Impact

An attacker can cause a runtime division by zero error and denial of service in tf.raw_ops.FractionalAvgPool:

import tensorflow as tf

value = tf.constant([60], shape=[1, 1, 1, 1], dtype=tf.int32)
pooling_ratio = [1.0, 1.0000014345305555, 1.0, 1.0]
pseudo_random = False
overlapping = False
deterministic = False
seed = 0
seed2 = 0

tf.raw_ops.FractionalAvgPool(
  value=value, pooling_ratio=pooling_ratio, pseudo_random=pseudo_random,
  overlapping=overlapping, deterministic=deterministic, seed=seed, seed2=seed2)

This is because the implementation computes a divisor quantity by dividing two user controlled values:

for (int i = 0; i < tensor_in_and_out_dims; ++i) {
  output_size[i] = static_cast<int>(std::floor(input_size[i] / pooling_ratio_[i]));
  DCHECK_GT(output_size[i], 0); 
} 

The user controls the values of input_size[i] and pooling_ratio_[i] (via the value.shape() and pooling_ratio arguments). If the value in input_size[i] is smaller than the pooling_ratio_[i], then the floor operation results in output_size[i] being 0. The DCHECK_GT line is a no-op outside of debug mode, so in released versions of TF this does not trigger.

Later, these computed values are used as arguments to GeneratePoolingSequence. There, the first computation is a division in a modulo operation:

std::vector<int64> GeneratePoolingSequence(int input_length, int output_length,
                                           GuardedPhiloxRandom* generator,
                                           bool pseudo_random) {
  ...
  if (input_length % output_length == 0) {
    diff = std::vector<int64>(output_length, input_length / output_length);
  }
  ...
}

Since output_length can be 0, this results in runtime crashing.

Patches

We have patched the issue in GitHub commit 548b5eaf23685d86f722233d8fbc21d0a4aecb96.

The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by Ying Wang and Yakun Zhang of Baidu X-Team.