This is the mail archive of the
newlib@sourceware.org
mailing list for the newlib project.
aarch64-none : TR1 Math floating point functions
- From: Mahesh Bodapati <maheshbodapati90 at gmail dot com>
- To: newlib at sourceware dot org
- Date: Fri, 4 May 2018 15:38:29 +0530
- Subject: aarch64-none : TR1 Math floating point functions
Hi,
I got errors when I tried to use std::remainder and std::round for
aarch64-none target and the the same test case is working for arm-none..
*#include <iostream>*
*#include <cmath>*
*#include <cfenv>*
*int main()*
*{*
* float x=10,y=20,z;*
* int a=10,b=15,c;*
* z= std::remainder(x, y); // Error : error: ‘remainder’ is not a member
of 'std'*
* c= std::round(b); // error: 'round' is not a member of 'std'*
* z= std::fmod(x, y);*
* return (int)z;*
*}*
When I went through the newlib sources, these FP functions are defined only
if LDBL_EQ_DBL is set but for aarch64 this macro will not be set because of
different sizes of long double and double.
So these functions are not supported in aarch-none tool chain?
Thanking you in advance…
Thanks,
Mahesh B