Missing interger overflow check in TensorFlow

[edwardyehuang]

Click to expand!

Issue Type

Bug

Source

source

Tensorflow Version

tf 2.10

Custom Code

Yes

OS Platform and Distribution

No response

Mobile device

No response

Python version

No response

Bazel version

No response

GCC/Compiler version

No response

CUDA/cuDNN version

No response

GPU model and memory

No response

Current Behaviour?

Tensorflow does not check integer overflow, which is dangerous and can cause bugs for many users.

Standalone code to reproduce the issue

a = 2147483648
print(tf.convert_to_tensor(a, tf.int32))

BUG case:

print(a* tf.constant(1, tf.int32))

Relevant log output

No response

[mohantym]

@edwardyehuang !
Thanks for the PR #58996. This issue will be closed once PR is merged.
Thank you!

[Rishikesh2338]

If this Pr is not assigned to anyone I can do this.

[mohantym]

Sure @Rishikesh2338 !

The Fix has to be replicated in C++ side as per mihai's comment.
Possible location is ops.cc.

will post an update by End of week.

Thank you!

[mohantym]

Hi @SuryanarayanaY !
Could you look at this issue. Attached gist in 2.9, 2.10, 2.11 for reference.
Thank you!

[edwardyehuang]

Hi @SuryanarayanaY ! Could you look at this issue. Attached gist in 2.9, 2.10, 2.11 for reference. Thank you!

Note that, this is not only the issue in tf.convert_to_tensor.

The example below is also dangerous.

import tensorflow as tf

a = 2147483647
a = tf.convert_to_tensor(a, tf.int32)

print(a) # tf.Tensor(2147483647, shape=(), dtype=int32)

b = a + tf.constant(1, tf.int32)

print(b) # tf.Tensor(-2147483648, shape=(), dtype=int32)

[edwardyehuang]

@mihaimaruseac @mohantym Could you remove the label "awaiting PR merge" please?

[SuryanarayanaY]

@edwardyehuang ,

Removed the label awaiting PR merge as requested. Could you please confirm whether you are willing to work on this issue? Thanks!

[github-actions]

This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.

[github-actions]

This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No