-lquadmath

Description

This option enables the 128-bit quadruple-precision floating-point number library. Add this link option for compilation when using the quadruple-precision floating-point number library.

Usage

Compilation command before the change:

gcc test.c -o test.out -lquadmath -DAARCH64_QUADMATH
./test.out

Compilation command after the change:
You do not need to explicitly specify the -DAARCH64_QUADMATH macro. The __aarch64__ predefined macro is enabled by default on the AArch64 platform.
1 2
gcc test.c -o test.out -lquadmath ./test.out

Suggestions on inventory project adaptation:
No file modifications are required; the added -DAARCH64_QUADMATH macro is invalid and has no impact.

Sample code: test.c

#include <stdio.h>
#include <math.h>
#include <quadmath.h>

#define USE_FLOAT128 1

int main()
{
#ifdef  USE_FLOAT128
    typedef __float128  long_double_t;
#else
    typedef long double long_double_t;
#endif

    long_double_t ld=0.1;
    long_double_t res;

    int* i = (int*) &res;
    i[0] = i[1] = i[2] = i[3] = 0xdeadbeef;

    for(ld = 0.0000000000000001; ld < __LDBL_MAX__; ld += 0.0000000000000001)
    {
        res = tanq(ld);
        printf("%08x-%08x-%08x-%08x\r", i[0], i[1], i[2], i[3]);
    }

    return 0;
}

Result

The compilation result shows that the 128-bit floating-point number calculation result is displayed, indicating that the 128-bit calculation is enabled.

Parent topic: Static Compilation Optimization