Description
Describe the bug
This is the problematic docker image: rapidsai/rapidsai-core-dev:22.06-cuda11.5-devel-ubuntu20.04-py3.9. After pulling this image and running this container using the instructions given in the release selector, trying to use basic commands like nano or htop from the command line results in `Segmentation fault (core dumped).
Steps/Code to reproduce bug
docker pull rapidsai/rapidsai-core-dev:22.06-cuda11.5-devel-ubuntu20.04-py3.9
sudo docker run --gpus all --rm -it --shm-size=1g --ulimit memlock=-1 -p 8888:8888 -p 8787:8787 -p 8786:8786 rapidsai/rapidsai-core-dev:22.06-cuda11.5-devel-ubuntu20.04-py3.9This will open a terminal inside the container. In the container try doing:
(rapids) root@bd7436c2aa8b:/rapids/notebooks# cd
(rapids) root@bd7436c2aa8b:~# touch a b c d
(rapids) root@bd7436c2aa8b:~# ls
a b c d
(rapids) root@bd7436c2aa8b:~# watch -n 1 ls
Segmentation fault (core dumped)Additionally, let's say we try installing stuff like nano or htop:
$ apt-get update
$ apt-get install nano htop
$ nano a
Segmentation fault (core dumped)
$ htop
Segmentation fault (core dumped)
Expected behavior
People using this docker image should be able to use basic tools like these.
Additional Context:
I tried installing GDB and running nano/htop/watch with gdb to see what happens:
$ gdb /usr/bin/nano
+ gdb /usr/bin/nano
(gdb) run
Starting program: /usr/bin/nano
warning: Error disabling address space randomization: Operation not permitted
Program received signal SIGSEGV, Segmentation fault.
0x00007f195a257b9f in termattrs_sp () from /opt/conda/envs/rapids/lib/libncursesw.so.6
(gdb) run
Starting program: /usr/bin/htop
warning: Error disabling address space randomization: Operation not permitted
Program received signal SIGSEGV, Segmentation fault.
0x00007f0d45772b9f in termattrs_sp () from /opt/conda/envs/rapids/lib/libncursesw.so.6
Also, this seems to be a problem affecting the development images only. For example trying to run nano/htop/watch on the base container causes no issues. Docker run command used for base container:
docker run --gpus all --rm -it --shm-size=1g --ulimit memlock=-1 rapidsai/rapidsai-core:22.06-cuda11.5-base-ubuntu20.04-py3.9
Environment details (please complete the following information):
- Environment location: Cloud (Azure) - though it's a dedicated VM.
- Method of install: Docker (docker pull and run commands provided above)