`Segmentation fault` in `tf.raw_ops.LoadAndRemapMatrix` when the value of `num_cols` is too large while `col_remapping` is empty. about tensorflow HOT 3 OPEN

@xl1043237213,
I tried to execute the mentioned code and observed that the issue/error is occuring whenever the input size of num_columns is >=12 digit number. With the <11 digit number as input, the code was executed without any issue and the output is also as expected. Kindly find the gist of it here and also observed that the input value for the num_rows, num_cols and max_rows_in_memory should be int.

Args
num_rows: An int that is >= 0. Number of rows (length of the 1st dimension) in the output matrix.
num_cols: An int that is >= 1. Number of columns (length of the 2nd dimension) in the output matrix.
max_rows_in_memory: An optional int. Defaults to -1.

Thank you!

from tensorflow.

@tilakrayal , The reportee knows that and it was mentioned in the title itself. Users can try with specially crafted inputs to check the behaviour of the APIs and expect meaningful exception. Please note that these are specially crafted inputs and they have spent lot of time to check for which inputs the API is failing.

cc: @mihaimaruseac

from tensorflow.

@tilakrayal , once again, please be more careful in what canned reply gets sent

from tensorflow.