libpod: setupNetNS() correctly mount netns

[Luap99]

The netns dir has a special logic to bind mout itself and make itslef
shared. This code here didn't which lead to catastrophic bug during
netns unmounting as we were unable to unmount the netns as the mount got
duplicated and had the wrong parent mount. This caused us to loop forever
trying to remove the file.

Fixes https://issues.redhat.com/browse/RHEL-59620
Fixes #23685

Does this PR introduce a user-facing change?

Fixed a bug that made it impossible to unmount our netns on cleanup which causes our commands to hang.

[giuseppe]

I think we need to modify check_netns_files to ignore the mount point itself

[Luap99]

I think we need to modify check_netns_files to ignore the mount point itself

I don't understand what you are trying to say with that.

[giuseppe]

I think we need to modify check_netns_files to ignore the mount point itself

I don't understand what you are trying to say with that.

I was looking at the failing test:

[+0814s] # --- /tmp/CI_lOcy/bats-run-jyH39Y/suite/netns-pre	2024-09-20 11:50:25.729240197 +0000
[+0814s] # +++ /tmp/CI_lOcy/bats-run-jyH39Y/suite/netns-post	2024-09-20 12:03:37.699435052 +0000
[+0814s] # @@ -0,0 +1 @@
[+0814s] # +netns-66ea18a9-066d-f924-bf03-f3fed67b8660
[+0814s] # 
[+0814s] # ^^^^^ Leaks found in /run/netns ^^^^^
[+0814s] # bats warning: Executed 333 instead of expected 332 tests
[+0814s] make: *** [Makefile:691: localsystem] Error 1
[12:03:37] END - [+0814s] total duration since 2024-09-20T12:03:37Z

it seems related to this change, and kind of makes sense since we are adding a new mount on /run/netns now, which wasn't done before in the Podman code

[Luap99]

I don't think the mount matters here in any way for this check the file is visible with or without the mount. And we already had the mount setup until the first ctr was one with --userns so it should not have changed much.

The error above clearly complains about netns-66ea18a9-066d-f924-bf03-f3fed67b8660 which is a ctr netns file that we didn't remove (unfortunately there is no way to know which test leaked it as it is just a random ID)

In general the leak check was just added two days ago, #2399, so I would not be surprised if this is a pre-existing issue that just causes some flakes. I at least cannot see anything wrong on the netns creating part that would cause new leaks

[Luap99]

Test passed on rerun I keep an eye out on the flakes with the netns cleanup.

[giuseppe]

LGTM

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: giuseppe, Luap99

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [Luap99,giuseppe]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[mheon]

LGTM

[mheon]

/lgtm