Skip to content

Segmentation fault when combining complex Variable operations with tf.raw_ops.Conj and assign_add #105367

@qiqicliff

Description

@qiqicliff

Issue type

Bug

Have you reproduced the bug with TensorFlow Nightly?

Yes

Source

source

TensorFlow version

tf2.17

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 behavior?

Description:
A segmentation fault occurs when performing complex number operations involving:
Complex64/complex128 variables
tf.raw_ops.Conjoperation
Variable.assign_add()method
The crash happens during the execution of the operation chain.

Expected behavior:
Operations should execute successfully, performing complex conjugate addition.
Actual behavior:
Segmentation fault (core dumped) during operation execution.

Standalone code to reproduce the issue

import tensorflow as tf
input_data = tf.constant([1 + 2j, 3 + 4j], dtype=tf.complex64)
var = tf.Variable(input_data, dtype=tf.complex64)
conj_result = tf.raw_ops.Conj(input=input_data)
assign_add_op = var.assign_add(conj_result)

Relevant log output

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions