/* ARM assembly AARCH64 Raspberry PI 3B */
/*  program nroot64.s   */
/* link with gcc. Use C function for display float */

/*******************************************/
/* Constantes file                         */
/*******************************************/
/* for this file see task include a file in language AArch64 assembly*/
.include "../includeConstantesARM64.inc"

/*******************************************/
/* Initialized data                        */
/*******************************************/
.data
szFormat1:         .asciz "Root= %+09.15f\n"
.align 4
qNumberA:          .quad 1024
/*******************************************/
/* UnInitialized data                      */
/*******************************************/
.bss
.align 4
/*******************************************/
/*  code section                           */
/*******************************************/
.text
.global main
main:                                   // entry of program

    /* root 10ieme de 1024  */
    ldr x0,qAdriNumberA                 // number address
    ldr d0,[x0]                         // load number in registre d0
    scvtf d0,d0                         // conversion in float
    mov x0,#10                          // N
    bl nthRoot
    ldr x0,qAdrszFormat1                // format
    bl printf                           // call C function !!!
                                        // Attention register dn lost !!!
    /* square root of 2   */
    fmov d0,2                           // conversion 2 in float register d0
    mov x0,#2                           // N
    bl nthRoot
    ldr x0,qAdrszFormat1                // format
                                        // d0 contains résult
    bl printf                           // call C function !!!

100:                                    // standard end of the program
    mov x0, #0                          // return code
    mov x8, #EXIT                       // request to exit program
    svc 0                               // perform the system call

qAdrszFormat1:           .quad szFormat1
qAdriNumberA:            .quad qNumberA

/******************************************************************/
/*     compute  nth root                                          */
/******************************************************************/
/* x0 contains N   */
/* d0 contains the value                 */
/* x0 return result                      */
nthRoot:
    stp x1,lr,[sp,-16]!            // save  registers
    stp x2,x3,[sp,-16]!            // save  registers
    stp d1,d2,[sp,-16]!            // save float registers
    stp d3,d4,[sp,-16]!            // save float registers
    stp d5,d6,[sp,-16]!            // save float registers
    stp d7,d8,[sp,-16]!            // save float registers
    fmov d6,x0                     //
    scvtf d6,d6                    // N conversion in float double précision (64 bits)
    sub x1,x0,#1                   // N - 1
    fmov d8,x1                     //
    scvtf d4,d8                    //conversion in float double précision
    fmov d2,d0                     // a = A
    fdiv d3,d0,d6                  // b = A/n
    adr x2,dfPrec                  // load précision
    ldr d8,[x2]
1:                                 // begin loop
    fmov d2,d3                     // a <- b
    fmul d5,d3,d4                  // (N-1)*b

    fmov d1,1                      // constante 1 -> float
    mov x2,0                       // loop indice
2:                                 // compute pow (n-1)
    fmul d1,d1,d3                  //
    add x2,x2,1
    cmp x2,x1                      // n -1 ?
    blt 2b                         // no -> loop
    fdiv d7,d0,d1                  // A / b pow (n-1)
    fadd d7,d7,d5                  // + (N-1)*b
    fdiv d3,d7,d6                  // / N -> new b
    fsub d1,d3,d2                  // compute gap
    fabs d1,d1                     // absolute value
    fcmp d1,d8                     // compare float maj FPSCR
    bgt 1b                         // if gap > précision -> loop
    fmov d0,d3                     // end return result in d0
100:
    ldp d7,d8,[sp],16              // restaur  2 float registers
    ldp d5,d6,[sp],16              // restaur  2 float registers
    ldp d3,d4,[sp],16              // restaur  2 float registers
    ldp d1,d2,[sp],16              // restaur  2 float registers
    ldp x2,x3,[sp],16              // restaur  2 registers
    ldp x1,lr,[sp],16              // restaur  2 registers
    ret                            // return to address lr x30
dfPrec:        .double 0f1E-10     // précision
/********************************************************/
/*        File Include fonctions                        */
/********************************************************/
/* for this file see task include a file in language AArch64 assembly */
.include "../includeARM64.inc"