pkg/rootless: correctly handle proxy signals on reexec

[Luap99]

There are quite a lot of places in podman were we have some signal
handlers, most notably libpod/shutdown/handler.go.

However when we rexec we do not want any of that and just send all
signals we get down to the child obviously. So before we install our
signal handler we must first reset all others with signal.Reset().

Also while at it fix a problem were the joinUserAndMountNS() code path
would not forward signals at all. This code path is used when you have
running containers but the pause process was killed.

Fixes #16091
Given that signal handlers run in different goroutines parallel it would
explain why it flakes sometimes in CI. However to my understanding this
flake can only happen when the pause process is dead before we run the
podman command. So the question still is what kills the pause process?

Does this PR introduce a user-facing change?

Fixed a bug in the rootless podman reexec logic were signals were not forwarded correctly.

[Luap99]

@giuseppe @edsantiago PTAL

[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

[edsantiago]

I haven't reviewed the code, but I tested it on a 1mt VM and it passes \o/.

Failing because of missing tests. Maybe you could tweak 032.bats to kill the rootless pause process, if running?

[Luap99]

Failing because of missing tests. Maybe you could tweak 032.bats to kill the rootless pause process, if running?

Yeah I am thinking about something like that. I would prefer to not change the existing test but I think I can add something like this to 550-pause-process.

[edsantiago]

So the question still is what kills the pause process?

New pause-related flake filed, #18685

[Luap99]

@edsantiago updated with tests for both code paths I changed. I confirmed that both fail on main right now and pass with this patch.

[edsantiago]

Might want to skip_if_remote:

Error: cannot use command "podman-remote unshare" with the remote podman client

(Also, much lower pri, might want to fix the duplication in that message; I'll file an issue for it later)

Other failures are flakes, including unlinkat-ebusy. I restarted them before realizing that the third flake (this one) was a hard failure.

[Luap99]

Might want to skip_if_remote:

yes I also saw that, I just waited so you can take a look at the tests before I repush.

[vrothberg]

LGTM, nice catch!

[edsantiago]

Am looking at tests, and confused, but nothing that should block this.

(Confusion: I'm having trouble understanding the difference between the two new tests, and how the 2nd new test differs from the single pause process test. But again, that is my lack of comprehension, and not a blocker nor a request to change anything)

[TomSweeneyRedHat]

LGTM

[TomSweeneyRedHat]

/lgtm