Re: [PATCH] fix float formatting of some exact halfway cases

[Rich Felker <dalias@libc.org>]

On Wed, Oct 12, 2016 at 12:49:59AM +0200, Szabolcs Nagy wrote:
> in nearest rounding mode exact halfway cases were not following the
> round to even rule if the rounding happened at a base 1000000000 digit
> boundary of the internal representation and the previous digit was odd.
> 
> e.g. printf("%.0f", 1.5) printed 1 instead of 2.
> ---
>  src/stdio/vfprintf.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/src/stdio/vfprintf.c b/src/stdio/vfprintf.c
> index e439a07..ff65295 100644
> --- a/src/stdio/vfprintf.c
> +++ b/src/stdio/vfprintf.c
> @@ -343,7 +343,8 @@ static int fmt_fp(FILE *f, long double y, int w, int p, int fl, int t)
>  		if (x || d+1!=z) {
>  			long double round = 2/LDBL_EPSILON;
>  			long double small;
> -			if (*d/i & 1) round += 2;
> +			if ((*d/i & 1) || (i==1000000000 && d>a && (d[-1]&1)))
> +				round += 2;
>  			if (x<i/2) small=0x0.8p0;
>  			else if (x==i/2 && d+1==z) small=0x1.0p0;
>  			else small=0x1.8p0;
> -- 
> 2.10.0

Thanks, committed.

Rich