Re: build failure (alpha-current): floating point exception 8

[Iain Hibbert <plunky%rya-online.net@localhost>]

On Tue, 3 Mar 2009, Christos Zoulas wrote:

> Does this work?
>
> Index: lib.c
> ===================================================================
> RCS file: /cvsroot/src/dist/nawk/lib.c,v
> retrieving revision 1.15
> diff -u -u -r1.15 lib.c
> --- lib.c     1 Mar 2009 23:30:52 -0000       1.15
> +++ lib.c     3 Mar 2009 17:06:02 -0000
> @@ -706,11 +706,11 @@
>  #include <math.h>
>  int is_number(const char *s)
>  {
> -     double r;
> +     double r, hv = HUGE_VAL;
>       char *ep;
>       errno = 0;
>       r = strtod(s, &ep);
> -     if (ep == s || r == HUGE_VAL || errno == ERANGE)
> +     if (ep == s || memcmp(&r, &hv, sizeof(hv)) == 0 || errno == ERANGE)
>               return 0;

it looks to me as if this was doing the wrong thing in the first place,
since HUGE_VAL is a valid number but not the only potentially invalid
result? The actual test should be

    (errno == ERANGE && (r == HUGE_VAL || r == -HUGE_VAL || r == 0))

but since the original code was failing whenever (errno == ERANGE),
perhaps you could forget the HUGE_VAL stuff altogether and hope the
implementation never sets that unless it found an error?

iain