Jim - thanks for this comprehensive and detailed post. On 5 September 2013 22:28, <[email protected]> wrote: > ... > The other option would be to do some timing kind of heuristic, where we let > only one thread run with a timeout, and if it doesn't return after that > timeout we interrupt it resume it with all threads running. That might be > worth trying, but you'd have to put in place a mechanism for ThreadPlans to > register a timeout and a callback on timeout that could adjust and restart > the process, and that isn't in place yet. Not sure this would appear any > more arbitrary than what we do now, however, though it might allow more code > to run on only a single thread.
Yes, in my opinion that would seem even more arbitrary which is what I'd like to avoid (and not the actual behaviour itself). Do you think it'd be reasonable to provide a user-facing message when we run other threads during stepping? _______________________________________________ lldb-dev mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev
