Follow-up Comment #2, bug #22230 (project freeciv):
I've tried, but can't replicate the specific crash here, but I have been able
to get some other pathfinding crashes. All appear to involve ptile being NULL
(as do the three attached stacktraces), which can happen (at least) when
initialising the (ignored) start node after application of patch #4768. My
crashes could be stopped by adjusting pf_*_map_move_cost() to return
PF_IMPOSSIBLE_MC if passed a null ptile, but I suspect the bug is somewhere
else, and that this only hides it, rather than fixing it.
Given the change in tile_index() from patch #4678, it may be wise to add
"fc_assert(ptile);" to the pf_*_map_move_cost() functions if the bug is fixed,
as there is no sensible way to trap the error during variable assignment while
iterating, and this may provide more useful guidance if similar issues appear
in the future.
Reply to this item at:
Message sent via/by Gna!
Freeciv-dev mailing list