Richard Henderson writes:
> Previously this was only supported for roundAndPackFloat64.
> Include support in round_canonical, round_to_int, roundAndPackFloat32,
> roundAndPackInt32, roundAndPackInt64, roundAndPackUint64.
If we extend the fp-test case we can exercise these routines:
./fp-tes
On 14.02.19 07:11, Richard Henderson wrote:
> Previously this was only supported for roundAndPackFloat64.
> Include support in round_canonical, round_to_int, roundAndPackFloat32,
> roundAndPackInt32, roundAndPackInt64, roundAndPackUint64.
>
> This does not include any of the floatx80 routines, as
Previously this was only supported for roundAndPackFloat64.
Include support in round_canonical, round_to_int, roundAndPackFloat32,
roundAndPackInt32, roundAndPackInt64, roundAndPackUint64.
This does not include any of the floatx80 routines, as we
do not have users for that rounding mode there.
Si
