[RFA/PATCH] PR/9711: quadratic slowdown for deep stack traces

[Joel Brobecker]

> Fine with me, although I wouldn't mind a reference to which is the
> the "another function" talked about here (--- my thinking is that it
> should be "easy" to get here when touching the problematic caller in
> question.  If it is grep-easy, the merrier.  If the reference in the
> comment ends up out-of-date at some point, then it just means the
> that a good time to rethink the stash as been
> reached).   Just a "(see foo_func)" would be fine.  Anyway, thanks!

I usually am not a big fan of putting references to other calling
functions inside comments, but I don't mind much. New version attached.
I am hoping to commit this on Wednesday (still hoping that we have
every blocking item dealt with by then, although it looks like it's
going to be a bit of a stretch).

-- 
Joel