[PATCH] Update sparc ULPs.
David Miller
davem@davemloft.net
Sat Nov 17 23:38:00 GMT 2012
From: "Joseph S. Myers" <joseph@codesourcery.com>
Date: Sat, 17 Nov 2012 23:21:23 +0000
> On Sat, 17 Nov 2012, David Miller wrote:
>
>> That's all fine and well, but I should note that the float versions of
>> these routines use the same exact value of tiny as the double ones.
>
> It's safe (other than possibly getting spurious underflows in some cases)
> to use a value that's smaller than necessary (e.g. using a double value in
> a float function), but not to use one that's too large (e.g. using a
> double value in a long double function).
Fair enough.
I'll adjust the float value to be more accurate and adjust my
long-double variant as well.
Thanks.
More information about the Libc-alpha
mailing list