You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Have you reproduced the bug with TensorFlow Nightly?
Yes
Source
source
TensorFlow version
tf 2.16.1
Custom code
Yes
OS platform and distribution
Ubuntu 20.04
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?
Illegal size will trigger a segfault.
Standalone code to reproduce the issue
# Signal --4;2024-05-13 05:48:16.272992: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.2024-05-13 05:48:16.273352: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.2024-05-13 05:48:16.277715: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.2024-05-13 05:48:16.331241: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.2024-05-13 05:48:17.210828: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT2024-05-13 05:48:17.792175: E tensorflow/core/util/util.cc:131] oneDNN supports DT_HALF only on platforms with AVX-512. Falling back to the default Eigen-based implementation if present.
# FusedResizeConv2DUsingGemmOp
import tensorflow as tf
mode = "REFLECT"
strides = [1, 1, 1, 1]
padding = "VALID"
resize_align_corners = False
input = tf.constant(0, shape=[1,2,3,2], dtype=tf.float16)
size = tf.constant([65534,65534], shape=[2], dtype=tf.int32)
paddings = tf.constant(0, shape=[4,2], dtype=tf.int32)
filter = tf.constant(0, shape=[1,2,2,2], dtype=tf.float16)
tf.raw_ops.FusedResizeAndPadConv2D(input=input, size=size, paddings=paddings, filter=filter, mode=mode, strides=strides, padding=padding, resize_align_corners=resize_align_corners)
Relevant log output
2024-05-14 08:31:20.675971: E tensorflow/core/util/util.cc:131] oneDNN supports DT_HALF only on platforms with AVX-512. Falling back to the default Eigen-based implementation if present.
Segmentation fault (core dumped)
Issue type
Bug
Have you reproduced the bug with TensorFlow Nightly?
Yes
Source
source
TensorFlow version
tf 2.16.1
Custom code
Yes
OS platform and distribution
Ubuntu 20.04
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?
Illegal
size
will trigger a segfault.Standalone code to reproduce the issue
Relevant log output
2024-05-14 08:31:20.675971: E tensorflow/core/util/util.cc:131] oneDNN supports DT_HALF only on platforms with AVX-512. Falling back to the default Eigen-based implementation if present. Segmentation fault (core dumped)
ASAN report:
The text was updated successfully, but these errors were encountered: