[CF-Devel] Some bug fixes

[crossfire-devel at archives.real-time.com]

Andreas Kirschbaum wrote:
>
     
      Hi,
     
     >
     
     
     >
     
      I'd like to present some bug fixes for the server. (I'll commit them if
     
     >
     
      there are no objections.) The patches correct some out of bounds
     
     >
     
      accesses to arrays that occurred on my local server.
     
     >
     
     
     >
     
      For the first patch (patch-1.diff), I do not have any specific item to
     
     >
     
      reproduce the bug but it occurred frequently that "enc > 20".
     
     
  Looks fine.

>
     
     
     >
     
      The loops in patch-2.diff run one too far, especially for (client) map
     
     >
     
      size 25x25: in this case the loops run from -1 to 25 (inclusive).
     
     
  I'd double check that logic if that change still works properly if the player 
chooses an even map size (eg, 20x20).  Granted, that it is a little odd in that 
the player won't be centered, but I recall some odd issues there.

  Alternative would only be to allow odd map sizes or the like.

>
     
     
     >
     
      The patch patch-3.diff fixes two out of range errors in lines 737+ and
     
     >
     
      1061+. The remaining changes replace the sizeof-expressions with the
     
     >
     
      macro "arraysize". Is this change acceptable or should I correct the
     
     >
     
      bugs only? Maybe the macro should go into a header file (which one?)
     
     >
     
      because there are some other places (server/commands.c,
     
     >
     
      server/alchemy.c, ...) where it could be used.
     
     
  If not seen any problems here, so I'm not positive what that change is fixing.

>
     
     
     >
     
      patch-4.diff fixes some invalid array references when no free spot can
     
     >
     
      be found. (Only the first one in treasure.c did actually crash.) I don't
     
     >
     
      think, the fixes for the exit (and maybe the key) are very good because
     
     >
     
      they drop the exit or place the key in an invalid place.
     
     
  Hard to say - in some cases, it can be better to just crash, because at least 
that allows easier investigation on what caused the problem.

>
     
     
     >
     
      The bug in patch-5.diff did happen in /pup_land/nurnberg/dick/bomb1 with
     
     >
     
      the first titan (it has Str=60). For me, this patch is not yet ready for
     
     >
     
      inclusion because similar problems exist for dex_bonus[] and maybe other
     
     >
     
      xxx_bonus[]. My preferred solution is to create new functions to
     
     >
     
      calculate the bonus. These functions can take care for the correct index
     
     >
     
      values.
     
     
  Well, one solution is to just extend the tables themselves (str_bonus, etc). 
And in fact, that is probably the best solution, because it keeps it simple 
(same code for both monsters & characters).

  Probably still important to have a max_player_stat vs just a max_stat.


_______________________________________________
crossfire-devel mailing list
     
     crossfire-devel at lists.real-time.com
     
     
     https://mailman.real-time.com/mailman/listinfo/crossfire-devel