On 03 Jun 2014, at 12:02, Alexander Graf wrote:
> On 06/03/2014 11:14 AM, Tristan Gingold wrote:
>> Remove the code that reduce the result to float32 as the frsqrte
>> instruction is defined to return a double-precision estimate of
>> the reciprocal square root.
>>
>> Although reducing the frac
Remove the code that reduce the result to float32 as the frsqrte
instruction is defined to return a double-precision estimate of
the reciprocal square root.
Although reducing the fractional part is harmless (as the estimation
must have at least 12 bits of precision according to the old PEM),
reduc
On 06/03/2014 11:14 AM, Tristan Gingold wrote:
Remove the code that reduce the result to float32 as the frsqrte
instruction is defined to return a double-precision estimate of
the reciprocal square root.
Although reducing the fractional part is harmless (as the estimation
must have at least 12 b
