Lucene search

K
osvGoogleOSV:GHSA-Q263-FVXM-M5MW
HistoryDec 10, 2020 - 7:07 p.m.

Heap out of bounds access in MakeEdge in TensorFlow

2020-12-1019:07:34
Google
osv.dev
11
tensorflow
makeedge function
heap out of bounds
security issue
uninitialized memory access
patch
github commit
tf versions
security guide
computation graph

EPSS

0

Percentile

12.8%

Impact

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.

Patches

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.

For more information

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

EPSS

0

Percentile

12.8%