I have a “py314” environment on Linux Mint 22.3 and it works well for several months, it is created previously using command:
conda create --name py314 python=3.14
Recently I updated this environment, it is success.
conda activate py314
conda update --all -y
But this environment is not working anymore, the python command reports error:
Illegal instruction (core dumped)
I tried to recreated this environment, same error. Then reinstall with latest Anaconda installer from scratch and recreate the environment, same error again. Here is example output in terminal:
(base) bobyuan@LenovoG455:~$ conda --version
conda 26.5.3
(base) bobyuan@LenovoG455:~$ python --version
Python 3.14.6
(base) bobyuan@LenovoG455:~$ conda activate py314
(py314) bobyuan@LenovoG455:~$ ls -l $(which python)
lrwxrwxrwx 1 bobyuan bobyuan 10 Sep 2 19:56 /home/bobyuan/anaconda3/envs/py314/bin/python -> python3.14
(py314) bobyuan@LenovoG455:~$ cd /home/bobyuan/anaconda3/envs/py314/bin
(py314) bobyuan@LenovoG455:~/anaconda3/envs/py314/bin$ ls -l python*
lrwxrwxrwx 1 bobyuan bobyuan 10 Sep 2 19:56 python -> python3.14
lrwxrwxrwx 1 bobyuan bobyuan 10 Sep 2 19:56 python3 -> python3.14
lrwxrwxrwx 1 bobyuan bobyuan 10 Sep 2 19:56 python3.1 -> python3.14
-rwxrwxr-x 1 bobyuan bobyuan 36334720 Sep 2 19:56 python3.14
-rwxrwxr-x 1 bobyuan bobyuan 3626 Sep 2 19:56 python3.14-config
lrwxrwxrwx 1 bobyuan bobyuan 17 Sep 2 19:56 python3-config -> python3.14-config
(py314) bobyuan@LenovoG455:~/anaconda3/envs/py314/bin$ which python
/home/bobyuan/anaconda3/envs/py314/bin/python
(py314) bobyuan@LenovoG455:~/anaconda3/envs/py314/bin$ python
Illegal instruction (core dumped)
This laptop is an old machine, I use it as a home server, it is very stable:
Static hostname: LenovoG455
Icon name: computer-laptop
Chassis: laptop 💻
Operating System: Linux Mint 22.3
Kernel: Linux 7.0.0-30-generic
Architecture: x86-64
Hardware Vendor: Lenovo
Hardware Model: Lenovo G455
Firmware Version: 2CCN19WW(V1.08)
Firmware Date: Tue 2010-02-23
Firmware Age: 16y 6month 1w 2d
I searched and find out "CPU Architecture Mismatch" is the top reason, so I am wondering if the conda has bug?