| [Freeciv-Dev] (PR#4181) recursion in tech requirements[Top] [All Lists][Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
 
 
[glip - Wed May  7 11:31:00 2003]:
> On Tue, 6 May 2003, Per I. Mathisen wrote:
> 
> > 
> > On Tue, 6 May 2003, Guest wrote:
> > > I am customizing my tech file and Freeciv-cvs as of April broke.
It says
> > > that it starts but then hangs my linux machine. I do not see any error
> > > mesages. It severly slows down my system and still does not display a
> > > map. why is that?
> > 
> > It is because of this:
> > 
> >     [advance_mobile_warfare10]
> >     name     = _("Mobile Warfare10")
> >     req1     = "Mobile Warfare10"
> >     req2     = "None"
> >     flags    = ""
> > 
> > You have this tech as its own requirement. Hence the tech tree code
(which
> > is recursive) loops until it runs out of stack, I guess.
> > 
> > The attached patch will detect this condition and print the following
> > error message:
> > 
> >   0: tech "Mobile Warfare10": requires itself (data/new/techs.ruleset)
> 
> Well detected but a bad fix.  You can still have a loop of length more 
> than 1, eg
> 
> Fusion requires Fission
> and
> Fission requires Fusion
> 
> The proper fix, IMO, is to put 
> if (advances[tech].req[0] == goal  ||  advances[tech].req[1] == goal) {
>   die("%s is an indirect prerequisite for itself", advances[goal].name);
> } 
> just before two recursive calls to 
> build_required_tech_helper in the function
> build_required_tech_helper  (common/tech.c:92)
Indeed.  How about this patch?
Note we should use LOG_FATAL+exit rather than die(), since this isn't a
code error...
jason
 tech_loops.diff Description: tech_loops.diff
 
 
| [Prev in Thread] | Current Thread | [Next in Thread] |  
[Freeciv-Dev] (PR#4181) recursion in tech requirements,
Jason Short <=
 
 |  |