Re: Problem report #18

["Stuart D. Herring" <herring@lanl.gov>]

Another simplification (unfortunately -- not a solution):  The majority of
this report is a red herring.  The (possible) issue is quite simple:
before the while loop, save_wv==prev_wv==0.  This code in the loop:

> 1847 		  if (prev_wv)
> 1848 		    prev_wv->next = wv;
> 1849 		  else
>
> Event var_deref_op: Variable "save_wv" tracked as NULL was dereferenced.
> Also see events: [assign_zero][alias_transfer]
>
> 1850 		    save_wv->contents = wv;

insists on dereferencing one of them.  If this branch (the else of a
vector-element-is-which-type? conditional) gets taken first time through
the loop, SIGSEGV.  But we're analyzing the vector menu_items here... if
we know that its structure precludes the first iteration of this loop
taking the last case (that is, if (aref menu_items start) is t, nil,
'lambda, or 'quote), then it's fine.

Hope it helps,
Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or
too sparse, it is because mass-energy conversion has occurred during
shipping.