From: Dave Airlie
Doing a straight uint/int->fp32->fp64 conversion causes
some precision issues, Roland suggested splitting the
integer into two portions and doing two separate
int->fp32->fp64 conversions then adding the results.
This passes the tests in CTS and piglit.
Signed-off-by: Dave Airl
Am 02.02.2018 um 05:56 schrieb Dave Airlie:
> From: Dave Airlie
>
> Doing a straight uint/int->fp32->fp64 conversion causes
> some precision issues, Roland suggested splitting the
> integer into two portions and doing two separate
> int->fp32->fp64 conversions then adding the results.
>
> This p