Re: [PATCH] Add cast to unsigned char to strverscmp

On Thu, Nov 7, 2024 at 1:06 PM <[email protected]> wrote:

> On 2024-11-07 09:53, Jeremy Bettis wrote:
> > GCC 14.2 with -Werror=sign-compare fails on this code.
> >
> > Signed-off-by: Jeremy Bettis <[email protected] <mailto:
> [email protected]>>
> > ---
> >   newlib/libc/string/strverscmp.c | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/newlib/libc/string/strverscmp.c
> b/newlib/libc/string/strverscmp.c
> > index 55966335f..e86718faa 100644
> > --- a/newlib/libc/string/strverscmp.c
> > +++ b/newlib/libc/string/strverscmp.c
> > @@ -76,7 +76,7 @@ int strverscmp(const char *l0, const char *r0)
> >    else if (c!='0') z=0;
> >    }
> >
> > - if (l[dp]<'1' + 9U && r[dp]<'1' + 9U) {
>
> I'd call that a false positive?
>

Is this a side-effect of the signedness of char being unspecified and
varying by architecture?

I'd lean to trying to taking the U off the 9.


> > + if ((unsigned char)(l[dp]-'1')<9U && (unsigned char)(r[dp]-'1')<9U) {
>
> What happens when chars are '\0' or '\x80'?
>
> >    /* If we're looking at non-degenerate digit sequences starting
> >    * with nonzero digits, longest digit string is greater. */
> >    for (j=i; isdigit(l[j]); j++)
>

--joel

>
> --
> Take care. Thanks, Brian Inglis              Calgary, Alberta, Canada
>
> La perfection est atteinte                   Perfection is achieved
> non pas lorsqu'il n'y a plus rien à ajouter  not when there is no more to
> add
> mais lorsqu'il n'y a plus rien à retirer     but when there is no more to
> cut
>                                  -- Antoine de Saint-Exupéry
>