#24587: Reset bootstrapping state on shutdown -------------------------------------------------+------------------------- Reporter: nickm | Owner: nickm Type: defect | Status: | merge_ready Priority: Medium | Milestone: Tor: | 0.3.4.x-final Component: Core Tor/Tor | Version: Severity: Normal | Resolution: Keywords: tor-mobile, s8-api, review-group-31 | Actual Points: Parent ID: #23847 | Points: Reviewer: | Sponsor: -------------------------------------------------+------------------------- Changes (by asn):
* status: needs_review => merge_ready Comment: Hey, did some review here as part of r-g-31. Commit `6033538` looks very good to me. Commit `3809036` also looks OK, but not very happy with it. I feel like resetting all the global statics on `tor_free_all()` makes sense but it's all a very brittle logic. The moment someone adds a new global static in that file and doesn't know about this convention of wiping at `tor_free_all()`, it will introduce a bug IIUC. Furthermore, the fact that some of those vars get reset to 0 and others to 1 is kinda arbitrary (and you need to look at their definitions to see if it's correct). I wonder what we could do about `3809036` to make it better. Perhaps we should de-global those variables, put them in a struct, and initialize them in a function, then call that function from some entry-point and `tor_free_all()`? That seems like a better approach but probably not so trivial. Maybe subject for a future ticket on this area? I'll mark this as merge_ready because I think it's correct, but we should think about the above concerns. -- Ticket URL: <https://trac.torproject.org/projects/tor/ticket/24587#comment:4> Tor Bug Tracker & Wiki <https://trac.torproject.org/> The Tor Project: anonymity online
_______________________________________________ tor-bugs mailing list tor-bugs@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs