tree a8de69b06807a4b4347a9979f9655406227cc4ae
parent bd8c04021b6299390debccd5b6e4fb5e2f063d01
author Attila Fazekas <afazekas@redhat.com> 1397937527 +0200
committer Attila Fazekas <afazekas@redhat.com> 1397939281 +0200

Fix stress runner signal related issues

The stress runner with -S option needs to stop all workers on error.

* The workers which terminated successfully (0) must be distinguished
  orm the failed workers, and the successfully terminated workers
  MUST not cause mass termination.
* The SIGCHLD single does not needs to be handled anymore when we terminate all
  workers.
* Does not makes sens to handle again and again, the SIGCHLD signal after
  any worker finished with failure.

Change-Id: Ic563f3acaa4e56f5135c438612f1db990b858552