CVE-2020-26271Medium· 4.4▾ SunlitHeap out of bounds access in MakeEdge in TensorFlow
▾ Sunlit zone — Low / medium · no exploitation signal
impact 24.2 · likelihood 0 · exploitation 0
Need a working PoC? Pro members can cast a request and our team develops one — it lands right here.
Exploit-prediction probability, daily snapshots since Jul 8.
Disclosure to exploitation, from the record and what we observed since indexing it.
Disclosed via OSV
Last analysed / modified upstream
0.2%
0.2% → 0.2%
Under certain cases, loading a saved model can result in accessing uninitialized memory while building the computation graph. The MakeEdge function creates an edge between one output tensor of the src node (given by output_index) and the input slot of the dst node (given by input_index). This is only possible if the types of the tensors on both sides coincide, so the function begins by obtaining the corresponding DataType values and comparing these for equality:
DataType src_out = src->output_type(output_index);
DataType dst_in = dst->input_type(input_index);
//...
However, there is no check that the indices point to inside of the arrays they index into. Thus, this can result in accessing data out of bounds of the corresponding heap allocated arrays.
In most scenarios, this can manifest as unitialized data access, but if the index points far away from the boundaries of the arrays this can be used to leak addresses from the library.
We have patched the issue in GitHub commit 0cc38aaa4064fd9e79101994ce9872c6d91f816b and will release TensorFlow 2.4.0 containing the patch. TensorFlow nightly packages after this commit will also have the issue resolved.
Since this issue also impacts TF versions before 2.4, we will patch all releases between 1.15 and 2.3 inclusive.
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
tensorflow < 1.15.5tensorflow >= 2.0.0, < 2.0.4tensorflow >= 2.1.0, < 2.1.3tensorflow >= 2.2.0, < 2.2.2tensorflow >= 2.3.0, < 2.3.2tensorflow-cpu < 1.15.5tensorflow-cpu >= 2.0.0, < 2.0.4tensorflow-cpu >= 2.1.0, < 2.1.3tensorflow-cpu >= 2.2.0, < 2.2.2tensorflow-cpu >= 2.3.0, < 2.3.2tensorflow-gpu < 1.15.5tensorflow-gpu >= 2.0.0, < 2.0.4tensorflow-gpu >= 2.1.0, < 2.1.3tensorflow-gpu >= 2.2.0, < 2.2.2tensorflow-gpu >= 2.3.0, < 2.3.2Upgrade to a patched release:
tensorflow 1.15.5tensorflow 2.0.4tensorflow 2.1.3tensorflow 2.2.2tensorflow 2.3.2tensorflow-cpu 1.15.5tensorflow-cpu 2.0.4tensorflow-cpu 2.1.3tensorflow-cpu 2.2.2tensorflow-cpu 2.3.2tensorflow-gpu 1.15.5tensorflow-gpu 2.0.4tensorflow-gpu 2.1.3tensorflow-gpu 2.2.2tensorflow-gpu 2.3.2Connected by shared product, vendor, weakness, or advisory.
CVE-2021-29541Low· 2.5Null pointer dereference in `StringNGrams`
CVE-2021-29525Low· 2.5Division by 0 in `Conv2DBackpropInput`
CVE-2021-29527Low· 2.5Division by 0 in `QuantizedConv2D`
CVE-2021-29524Low· 2.5Division by 0 in `Conv2DBackpropFilter`
CVE-2020-15207High· 8.7Segfault and data corruption in tensorflow-lite
CVE-2021-29535Low· 2.5Heap buffer overflow in `QuantizedMul`