Page 325 - ARM 64 Bit Assembly Language

314 Chapter 9

9.7.7.2 Operations

Name Effect Description
fcsel if cond is True then Set Fd to Fn or Fm.
Fd ← Fn
else
Fd ← Fm
end if

9.7.7.3 Examples
1 fcsel s0, s1, s2, ge // if ge, then s0<-s1 else s0<-s2

9.8 Floating point sine function

A fixed point implementation of the sine function was discussed in Section 8.6,and shownto
be superior to the floating point sine function provided by GCC. Now that we have covered
the FP instructions, we can write an assembly version using floating point that also performs
better than the routines provided by GCC.

Listing 9.1 Simple scalar implementation of the sinx function using IEEE single precision.

1 //*************************************************************
2 // Name: sincos_a_f.S
3 // Author: Larry Pyeatt
4 // Date: 2/22/2018
5 //*************************************************************
6 // This is a version of the sin functions that uses single
7 // precision floating point with the FP/NEON instruction set.
8 // --------------------------------------------------------------
9 .data
10 // The following is a table of constants used in the
11 // Taylor series approximation for sine
12 .align 5 // Align to cache
13 ctab: .word 0xBE2AAAAA // -1.666666e-01
14 .word 0x3C088889 // 8.333334e-03
15 .word 0xB9500D00 // -1.984126e-04
16 .word 0x3638EF1D // 2.755732e-06
17 .word 0xB2D7322A // -2.505210e-08
18 .equ TERMS,((. - ctab)/4)
19 // --------------------------------------------------------------
20 .text
21 .align 2
22 // float sin_a_f(x)
23 // sin_a_f implements the sine function using IEEE single