numpy umath.cpython SIGILL (1.14.2) #10787
Description
Downgrading numpy to 1.11.3 helped
I'm getting a segfault when using numpy 1.14.2. Here's the gdb trace:
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `python tests/test_db_bq_loader.py'.
Program terminated with signal SIGILL, Illegal instruction.
#0 0x00007fb03004e400 in LONG_add_avx2 (args=<optimized out>,
dimensions=<optimized out>, steps=<optimized out>,
__NPY_UNUSED_TAGGEDfunc=<optimized out>)
from /opt/conda/envs/xx/lib/python3.6/site-packages/numpy/core/umath.cpython-36m-x86_64-linux-gnu.so
(gdb)
Here is the code that can replicate this:
import numpy as np
import pandas as pd
from numpy.random import randint
data = [{'ds': '2017-01-01', 'a': 10, 'b': 20, 'c': 30}, {'ds': '2017-01-01', 'a': 11, 'b': 21, 'c': 31},{'ds': '2017-01-01', 'a': 12, 'b': 22, 'c': 32},{'ds': '2017-01-02', 'a': 13, 'b': 23, 'c': 33},{'ds': '2017-01-02', 'a': 14, 'b': 24, 'c': 34},{'ds': '2017-01-03', 'a': 15, 'b': 25, 'c': 35},{'ds': '2017-01-04', 'a': randint(1, 100), 'b': randint(1, 100), 'c': randint(1, 100)}]
df = pd.DataFrame(data)
np.array_equal(df.sort_values(['ds', 'a', 'b', 'c']).values, df.sort_values(['ds', 'a', 'b', 'c']).values)
- It's running in a docker container. proc/cpuinfo says GenuineIntel Xeon E5-2660
Python version and pip freeze
root@31dedfee1f82:/app/xxx# python --version
Python 3.6.4
root@31dedfee1f82:/app/xxx# pip freeze
asn1crypto==0.24.0
bcrypt==3.1.3
blinker==1.4
cachetools==2.0.1
certifi==2018.1.18
cffi==1.11.5
chardet==3.0.4
cryptography==2.1.4
future==0.16.0
google-api-core==0.1.4
google-api-python-client==1.6.4
google-auth==1.4.1
google-auth-httplib2==0.0.2
google-auth-oauthlib==0.1.0
google-cloud-bigquery==0.29.0
google-cloud-core==0.28.1
google-cloud-storage==1.6.0
google-resumable-media==0.3.1
googleapis-common-protos==1.5.3
httplib2==0.10.3
idna==2.6
numpy==1.14.2
oauth2client==4.1.2
oauthlib==2.0.6
pandas==0.22.0
pandas-gbq==0.3.0
paramiko==2.4.1
protobuf==3.5.2
psutil==5.4.3
psycopg2==2.7.4
py4j==0.10.6
pyarrow==0.8.0
pyasn1==0.4.2
pyasn1-modules==0.2.1
pycparser==2.18
PyHive==0.5.1
PyJWT==1.5.3
pymssql==2.1.4.dev1
PyMySQL==0.8.0
PyNaCl==1.1.2
pyOpenSSL==17.5.0
pysftp==0.2.9
PySocks==1.6.8
pyspark==2.3.0
python-dateutil==2.7.0
pytz==2018.3
PyYAML==3.12
requests==2.18.4
requests-oauthlib==0.8.0
rsa==3.4.2
sasl==0.2.1
scikit-learn==0.19.1
scipy==1.0.0
simplejson==3.13.2
six==1.11.0
SQLAlchemy==1.2.4
thrift==0.11.0
thrift-sasl==0.3.0
uritemplate==3.0.0
urllib3==1.22