avocado.main(): avoid an infinite fork loop bomb [v3]

[clebergnu]

Because the current implementation of avocado.main() creates a job
and runs "sys.argv[0]" to implement standalone mode, it ends up
running itself over and over.

This simple proposed fix, prevents avocado.main() from running
itself again if called from itself. Since they are on different
processes, the mechanism chosen to do this is to set an environment
variable, that will be seen by the next process.

Also, by exiting from main() with an error code, the test first
level test will fail. This will let the user know that the chosen
approach (SIMPLE tests written in Python and calling main()) are
not worthy of a PASS.

This adresses issue #961.

Signed-off-by: Cleber Rosa crosa@redhat.com

---

Changes from v2:

• Return an error code to signal that the used approach (calling main() is not ideal from a SIMPLE test).

Changes from v1:

• Add error message to STDERR, warning that main() will exit to avoid a loop.

[clebergnu]

@ldoktor I believe this is the behavior you were looking for. Can you take a look at it?

[ldoktor]

Very nice, thanks

[clebergnu]

OK, so using process.run() is not a good idea because it can't handle certain situations such as the fork bomb.

Using --job-timeout, now that the nested main() behavior is to exit with error, is also not an option because of another bug which I documented here:

https://trello.com/c/H1AxryDp/568-bug-avocado-does-not-report-the-right-job-status-when-interrupted

Because of that, I implemented the timeout rather manually, using Python's subprocess module, and cleaning up the process (group) in case it doesn't finish in time.

[clebergnu]

Closing in favor of #991