Hi all, I was just playing around and found that the highlight net tool is visibly slow. I delved into the code, and saw that finding connected NETLIST_OBJECT involves needlessly looping over NETLIST_OBJECTS.
Is there any performance consideration behind having the NETLIST_OBJECTS as a std::vector? Would it be better to have an std::unordered_map from the net codes to a std::vector of all NETLIST_OBJECT with that net code? Rudimentary instrumentation with callgrind shows that most of the time is spent in SCH_SHEET_PATH comparisons. -- Regards, Gaurav Juvekar _______________________________________________ Mailing list: https://launchpad.net/~kicad-developers Post to : [email protected] Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
