Comment in ccosh()

Comment in ccosh()

[Damian McGuckin]

There is a comment in ccoshf()

 	 /* x < 88.7: expf(|x|) won't overflow */

where is correct

The equivalent in ccosh() which says:

 	/* x < 710: exp(|x|) won't overflow */

This is far less concise than the earlier case.  More strictly

 	 /* x < 709.78: exp(|x|) won't overflow */

Am I being overly pedantic?

Thanks - Damian

Re: [musl] Comment in ccosh()

[Szabolcs Nagy]

* Damian McGuckin <damianm@esi.com.au> [2025-07-02 15:17:20 +1000]:

> 
> There is a comment in ccoshf()
> 
> 	 /* x < 88.7: expf(|x|) won't overflow */
> 
> where is correct
> 
> The equivalent in ccosh() which says:
> 
> 	/* x < 710: exp(|x|) won't overflow */
> 
> This is far less concise than the earlier case.  More strictly
> 
> 	 /* x < 709.78: exp(|x|) won't overflow */
> 
> Am I being overly pedantic?

i guess the comment is sloppy

i think if you submit such a patch it can be accepted.

(or reword it in another way so it is not actually false)

> 
> Thanks - Damian

[musl] Re: spam>Re: [musl] Comment in ccosh()

[Damian McGuckin]

On Sun, 6 Jul 2025, Szabolcs Nagy wrote:

>> Am I being overly pedantic?
>
> i guess the comment is sloppy
>
> i think if you submit such a patch it can be accepted.
>
> (or reword it in another way so it is not actually false)

I will ponder what seems optimal and make a suggestion.

Mind you, MUSL is way in front of GNU libc which simply took comments from 
the double version of sinh() and cosh() and used then in the float and 
long double variants.  So the float routine talks about maxdouble but uses 
constants which are related to maxfloat!! And likewise in the long double 
routine.

Thanks - Damian