Unable to restore a container to an existing network namespace

[Rexyyj]

Is this a BUG REPORT or FEATURE REQUEST? (leave only one on its own line)

/kind bug

Description
When a container is created by joining a network namespace with --net ns:<path> option, this container can be checkpointed successfully but will not be able to restore.

During the restoring, the following error appears:

sudo podman --runtime /usr/bin/runc container restore -i ./ckp2.tar.gz  --print-stats

Error: cannot set static IP or MAC address if not creating a network namespace: invalid argument

Steps to reproduce the issue:

1.Create a network namespace and setup veth interface

ip netns add test
ip netns exec test /bin/bash
ip link add enp2s0 type veth peer name eth0
ip addr add 10.1.1.0/24 dev eth0
ip link set eth0 up
exit

2.Run container by joining the network namespace

sudo podman run --name test2 --runtime /usr/bin/runc -d --network ns:/run/netns/test alpine sh -c 'while sleep 3600; do :; done'

3.Checkpoint the container

sudo podman --runtime /usr/bin/runc container checkpoint --compress=none -e ./ckp.tar.gz test2  --print-stats

4.Restore the container with checkpoint image and get the error

sudo podman --runtime /usr/bin/runc container restore -i ./ckp2.tar.gz  --print-stats

Error: cannot set static IP or MAC address if not creating a network namespace: invalid argument

Describe the results you received:
From the checkpointed image, the container did not create a network namespace, so the restore failed.

Describe the results you expected:
The container can restore to the original namespace. Or there is an option for user to specify a new namespace to restore the container to.

Output of podman version:

Client:       Podman Engine
Version:      4.1.0
API Version:  4.1.0
Go Version:   devel go1.19-5f2fdbe7ed Wed May 18 01:02:54 2022 +0000
Built:        Wed May 18 17:11:36 2022
OS/Arch:      linux/amd64

Output of podman info --debug:

host:
  arch: amd64
  buildahVersion: 1.26.1
  cgroupControllers: []
  cgroupManager: cgroupfs
  cgroupVersion: v1
  conmon:
    package: Unknown
    path: /usr/local/libexec/podman/conmon
    version: 'conmon version 2.1.1, commit: 546e4dbe78c72754053d4029593e7d9fc59adff1'
  cpuUtilization:
    idlePercent: 99.76
    systemPercent: 0.1
    userPercent: 0.14
  cpus: 2
  distribution:
    codename: focal
    distribution: ubuntu
    version: "20.04"
  eventLogger: file
  hostname: rex-vm
  idMappings:
    gidmap:
    - container_id: 0
      host_id: 1000
      size: 1
    - container_id: 1
      host_id: 100000
      size: 65536
    uidmap:
    - container_id: 0
      host_id: 1000
      size: 1
    - container_id: 1
      host_id: 100000
      size: 65536
  kernel: 5.8.18-050818-generic
  linkmode: dynamic
  logDriver: k8s-file
  memFree: 2447998976
  memTotal: 4123381760
  networkBackend: cni
  ociRuntime:
    name: runc
    package: Unknown
    path: /usr/bin/runc
    version: |-
      runc version 1.1.0+dev
      commit: v1.1.0-179-g95243661
      spec: 1.0.2-dev
      go: devel go1.19-5f2fdbe7ed Wed May 18 01:02:54 2022 +0000
      libseccomp: 2.5.1
  os: linux
  remoteSocket:
    path: /run/user/1000/podman/podman.sock
  security:
    apparmorEnabled: false
    capabilities: CAP_AUDIT_WRITE,CAP_CHOWN,CAP_DAC_OVERRIDE,CAP_FOWNER,CAP_FSETID,CAP_KILL,CAP_MKNOD,CAP_NET_BIND_SERVICE,CAP_NET_RAW,CAP_SETFCAP,CAP_SETGID,CAP_SETPCAP,CAP_SETUID,CAP_SYS_CHROOT
    rootless: true
    seccompEnabled: true
    seccompProfilePath: ""
    selinuxEnabled: false
  serviceIsRemote: false
  slirp4netns:
    executable: ""
    package: ""
    version: ""
  swapFree: 1469710336
  swapTotal: 1469710336
  uptime: 7h 23m 45.2s (Approximately 0.29 days)
plugins:
  log:
  - k8s-file
  - none
  - passthrough
  network:
  - bridge
  - macvlan
  - ipvlan
  volume:
  - local
registries:
  search:
  - registry.fedoraproject.org
  - registry.access.redhat.com
  - docker.io
  - quay.io
store:
  configFile: /home/rex/.config/containers/storage.conf
  containerStore:
    number: 0
    paused: 0
    running: 0
    stopped: 0
  graphDriverName: vfs
  graphOptions: {}
  graphRoot: /home/rex/.local/share/containers/storage
  graphRootAllocated: 31040933888
  graphRootUsed: 14684463104
  graphStatus: {}
  imageCopyTmpDir: /var/tmp
  imageStore:
    number: 0
  runRoot: /run/user/1000/containers
  volumePath: /home/rex/.local/share/containers/storage/volumes
version:
  APIVersion: 4.1.0
  Built: 1652886696
  BuiltTime: Wed May 18 17:11:36 2022
  GitCommit: ""
  GoVersion: devel go1.19-5f2fdbe7ed Wed May 18 01:02:54 2022 +0000
  Os: linux
  OsArch: linux/amd64
  Version: 4.1.0

Package info (e.g. output of rpm -q podman or apt list podman):

None. Built from source.

Have you tested with the latest version of Podman and have you checked the Podman Troubleshooting Guide? (https://github.com/containers/podman/blob/main/troubleshooting.md)

Yes, tested with version 4.1.0

Additional environment details (AWS, VirtualBox, physical, etc.):
The test is done in following environment in KVM:
(Note: the ubuntu kernel version is very important. Other kernel version may cause checkpoint failed.)

$ lsb_release -a

Distributor ID:	Ubuntu
Description:	Ubuntu 20.04.4 LTS
Release:	20.04
Codename:	focal

$ uname -r
5.8.18-050818-generic

[Luap99]

Maybe a bit simpler to reproduce it fails the same way when --network none is used instead of the custom netns.