Add cwd argument for hooks exec Run func

[fangpenlin]

For fixing the wrong cwd bug in podman:

containers/podman#18907

[rhatdan]

Shouldn't we be calling this from somewhere in the code? I did not notice that they calls were all in test function.
What should a caller to this function set the CWD to?

[fangpenlin]

@rhatdan The Run function is called from podman and buildah , here

https://github.com/containers/podman/pull/18921/files#diff-694b11f8e3815ed5986dc6e81cbc40ecff32b50938a38b44c8d0cb405adb3330

and here

https://github.com/containers/buildah/pull/4871/files

I checked in the common code base, it's not called anywhere except from tests.

[Luap99]

Also please squash the commits.

@vrothberg PTAL

[Luap99]

Maybe taking a step back, is it anywhere defined what the cwd when executing hooks should be?
It is not mentioned here: https://github.com/opencontainers/runtime-spec/blob/v1.0.2/config.md#posix-platform-hooks
But is says you should get the state via stdin. And the state contains the bundle path: https://github.com/opencontainers/runtime-spec/blob/v1.0.2/runtime.md#state.

[vrothberg]

That's a fair point, @Luap99. Technically, the cwd shouldn't matter and if the hook relies on this behavior it may work with Podman/crun but not other container engines.

[fangpenlin]

The only reason I created the issue is because I saw a difference between hook executable invoked by crun and podman (poststop in this case). It's a bit odd to see a bit different behaivor for the hooks. But I also understand one probably should not rely on cwd for reading config.json or other files as it's not in the spec any, as the state from stdin should be used.

I guess because before making the crun call by podman, the cwd will be set to use the bundle dir, so the hook calls coming from crun is going to be the bundle dir. But for poststop hook call, it's made by podman directly, so no explicit dir set, as a result, the current working dir of podman will be used.

Please feel free to close my issues and PRs as if this is not part of the spec and not really a bug 😅

[Luap99]

I think it is a good idea overall to have a stable cwd for the hooks and not depend on the cmd of the calling process.

[rhatdan]

I agree.

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: fangpenlin
Once this PR has been reviewed and has the lgtm label, please assign luap99 for approval. For more information see the Kubernetes Code Review Process.

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

Needs approval from an approver in each of these files:

• OWNERS

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

[fangpenlin]

Close in favor of #1521