Fix for issue 128

[dimsh99]

No description provided.

[dimsh99]

Thanks, Stephen, I'll check on this.

[dimsh99]

I remembered why I gave up on using selinux_enabled? first time.
If Chef::Recipe.send(:include, Chef::Util::Selinux) added to recipe selinux_enabled? can't be accessed from within resource only_if guard:
undefined method `selinux_enabled?' for Chef::Resource::Bash

I see 2 ways to make it work:

1. Inject Selinux into bash resource:
   Chef::Resource::Bash.send(:include, Chef::Util::Selinux)
2. Condition around resource:
   if selinux_enabled? bash 'Restore security context' do .... end end

The first approach makes all Bash resources in runlist "infected" by Chef::Util::Selinux, the other one is considered bad practice in chef.
So I'm not sure what to choose. Maybe I'm missing something?

[shoekstra]

Now that you mention that, it does sound familiar. I dug into our wrapper cookbook and we ended up with something like this:

extend Chef::Util::Selinux

selinux_on = selinux_enabled?

and then use selinux_on in the resource's guard block? I think this would be the cleanest option IMO.

On a separate note we should discuss SELinux testing, but let's do that in #155.

[shoekstra]

I've rebased and fixed up the guard to test if SELinux is enabled.

Thanks for the contribution!

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.