Log execution timeouts to Dashboard

[rarkins]

Describe the proposed change(s).

Related discussion: #31982

The Dependency Dashboard should indicate if a repo has been aborted due to an execution timeout, because such timeouts may be "permanent" instead of temporary. Generally we don't want to update the Dashboard for most temporary errors because they could result in wrong data, but I think in the case of an execution timeout causing the job to abort then it would be good to capture this.

BTW in some environments the entire job might be aborted if it goes too long, so this handles the case where Renovate knowingly cancels a child process package manager command and can gracefully exit in time to update.