CVSS2
Attack Vector
NETWORK
Attack Complexity
LOW
Authentication
SINGLE
Confidentiality Impact
PARTIAL
Integrity Impact
NONE
Availability Impact
PARTIAL
AV:N/AC:L/Au:S/C:P/I:N/A:P
CVSS3
Attack Vector
NETWORK
Attack Complexity
LOW
Privileges Required
LOW
User Interaction
NONE
Scope
UNCHANGED
Confidentiality Impact
HIGH
Integrity Impact
NONE
Availability Impact
HIGH
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H
EPSS
Percentile
76.8%
The implementation of shape inference for ReverseSequence
does not fully validate the value of batch_dim
and can result in a heap OOB read:
import tensorflow as tf
@tf.function
def test():
y = tf.raw_ops.ReverseSequence(
input = ['aaa','bbb'],
seq_lengths = [1,1,1],
seq_dim = -10,
batch_dim = -10 )
return y
test()
There is a check to make sure the value of batch_dim
does not go over the rank of the input, but there is no check for negative values:
const int32_t input_rank = c->Rank(input);
if (batch_dim >= input_rank) {
return errors::InvalidArgument(
"batch_dim must be < input rank: ", batch_dim, " vs. ", input_rank);
}
// ...
DimensionHandle batch_dim_dim = c->Dim(input, batch_dim);
Negative dimensions are allowed in some cases to mimic Pythonโs negative indexing (i.e., indexing from the end of the array), however if the value is too negative then the implementation of Dim
would access elements before the start of an array:
DimensionHandle Dim(ShapeHandle s, int64_t idx) {
if (!s.Handle() || s->rank_ == kUnknownRank) {
return UnknownDim();
}
return DimKnownRank(s, idx);
}
ยท
static DimensionHandle DimKnownRank(ShapeHandle s, int64_t idx) {
CHECK_NE(s->rank_, kUnknownRank);
if (idx < 0) {
return s->dims_[s->dims_.size() + idx];
}
return s->dims_[idx];
}
We have patched the issue in GitHub commit 37c01fb5e25c3d80213060460196406c43d31995.
The fix will be included in TensorFlow 2.8.0. We will also cherrypick this commit on TensorFlow 2.7.1, TensorFlow 2.6.3, and TensorFlow 2.5.3, as these are also affected and still in supported range.
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
This vulnerability has been reported by Yu Tian of Qihoo 360 AIVul Team.
Vendor | Product | Version | CPE |
---|---|---|---|
tensorflow | gpu | 2.7.0 | cpe:2.3:a:tensorflow:gpu:2.7.0:*:*:*:*:*:*:* |
tensorflow | gpu | * | cpe:2.3:a:tensorflow:gpu:*:*:*:*:*:*:*:* |
tensorflow | cpu | 2.7.0 | cpe:2.3:a:tensorflow:cpu:2.7.0:*:*:*:*:*:*:* |
tensorflow | cpu | * | cpe:2.3:a:tensorflow:cpu:*:*:*:*:*:*:*:* |
tensorflow | tensorflow | 2.7.0 | cpe:2.3:a:tensorflow:tensorflow:2.7.0:*:*:*:*:*:*:* |
tensorflow | tensorflow | * | cpe:2.3:a:tensorflow:tensorflow:*:*:*:*:*:*:*:* |
github.com/advisories/GHSA-6gmv-pjp9-p8w8
github.com/tensorflow/tensorflow/blob/5100e359aef5c8021f2e71c7b986420b85ce7b3d/tensorflow/core/framework/shape_inference.h#L415-L428
github.com/tensorflow/tensorflow/blob/5100e359aef5c8021f2e71c7b986420b85ce7b3d/tensorflow/core/ops/array_ops.cc#L1636-L1671
github.com/tensorflow/tensorflow/commit/37c01fb5e25c3d80213060460196406c43d31995
github.com/tensorflow/tensorflow/security/advisories/GHSA-6gmv-pjp9-p8w8
nvd.nist.gov/vuln/detail/CVE-2022-21728
CVSS2
Attack Vector
NETWORK
Attack Complexity
LOW
Authentication
SINGLE
Confidentiality Impact
PARTIAL
Integrity Impact
NONE
Availability Impact
PARTIAL
AV:N/AC:L/Au:S/C:P/I:N/A:P
CVSS3
Attack Vector
NETWORK
Attack Complexity
LOW
Privileges Required
LOW
User Interaction
NONE
Scope
UNCHANGED
Confidentiality Impact
HIGH
Integrity Impact
NONE
Availability Impact
HIGH
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H
EPSS
Percentile
76.8%