Skip to content
GitLab
  • Menu
Projects Groups Snippets
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
    • Contribute to GitLab
  • Sign in
  • T torflow
  • Project information
    • Project information
    • Activity
    • Labels
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
  • Issues 0
    • Issues 0
    • List
    • Boards
    • Service Desk
    • Milestones
  • CI/CD
    • CI/CD
    • Pipelines
    • Jobs
    • Schedules
  • Deployments
    • Deployments
    • Environments
    • Releases
  • Monitor
    • Monitor
    • Incidents
  • Analytics
    • Analytics
    • Value stream
    • CI/CD
    • Repository
  • Wiki
    • Wiki
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
Collapse sidebar
  • The Tor Project
  • Network Health
  • torflow
  • Issues
  • #3701
Closed
Open
Created Aug 08, 2011 by Mike Perry@mikeperryOwner

PyTorCtl shouldn't os.kill() itself

In order to ensure we died properly from uncaught exceptions in user-written event threads, I hacked in an os.kill() into the event loop that catches these exceptions, because sys.exit() only exits the thread, and os._exit() doesn't call atexit handlers. os.kill() does call the atexit handlers, but it causes the return value of the interpreter to be the signal used. This exit value breaks out bw authority parent/child model, which relies on exit codes to properly detect child death and respawn.

Assignee
Assign to
Time tracking