[BusyBox] bug#1158: can of worms

[Erik Andersen]

On Fri Apr 27, 2001 at 07:22:48PM -0500, Manuel Novoa III wrote:
> Larry Doolittle wrote:
> 
> > I volunteer to code a version of m_h_r_s that never overflows,
> > and doesn't use float.  It might look a little like the
> > computation in init.c/check_free_memory(), that is robust in
> > the face of a unit that is not even a power of two.  Or maybe
> > not.
> 
> Take a look at the old utility.c in the cvs archive attic; specificly
> the function format in version 1.202.  I had coded a version which
> didn't use floating point and wouldn't overflow.  Apparently it was
> replaced by floating point code in 1.217.  I don't know why.  It worked
> fine when I tested it in isolation.

The problem was that du, ls, and df all has different opinions on how to use
the interface, so fixing it for one broke the other two.  I would be glad to go
back to using your version (and avoid the float problems) as long as du, ls,
and df all work as expected -- all at the same time.

 -Erik

--
Erik B. Andersen   email:  andersen at lineo.com
--This message was written using 73% post-consumer electrons--