jasonmolenda added a comment. I think the m_thread_pcs.clear() in UpdateThreadIDList() is to clear out any old thread pc values that we might have populated in the past. The only way I can see this happening is if the remote stub SOMETIMES returned the thread-pcs: list in the stop packet. UpdateThreadPCsFromStopReplyThreadsValue() which we call if we did get a thread-pcs: in this stop reply packet, clears m_thread_pcs so the only reason to have this also in UpdateThreadIDList() is if we had a stub that was inconsistent.
I agree that the m_thread_pcs.clear() in UpdateThreadIDsFromStopReplyThreadsValue() is wrong, and your patch here is good. I don't have strong feelings about removing the m_thread_pcs() in UpdateThreadIDList(), but I think I would prefer leaving it in to changing it unless we know it's causing problems. Does this cause problems in your environment? https://reviews.llvm.org/D48868 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
