/* Function cosf vectorized with AVX2.
   Copyright (C) 2014-2018 Free Software Foundation, Inc.
   This file is part of the GNU C Library.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
   License as published by the Free Software Foundation; either
   version 2.1 of the License, or (at your option) any later version.

   The GNU C Library is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   Lesser General Public License for more details.

   You should have received a copy of the GNU Lesser General Public
   License along with the GNU C Library; if not, see
   <http://www.gnu.org/licenses/>.  */


#include <sysdep.h>
#include "svml_s_trig_data.h"

	.text
ENTRY (_ZGVdN8v_cosf_avx2)
/*
  ALGORITHM DESCRIPTION:

  1) Range reduction to [-Pi/2; +Pi/2] interval
    a) We remove sign using AND operation
    b) Add Pi/2 value to argument X for Cos to Sin transformation
    c) Getting octant Y by 1/Pi multiplication
    d) Add "Right Shifter" value
    e) Treat obtained value as integer for destination sign setting.
       Shift first bit of this value to the last (sign) position
    f) Subtract "Right Shifter"  value
    g) Subtract 0.5 from result for octant correction
    h) Subtract Y*PI from X argument, where PI divided to 4 parts:
         X = X - Y*PI1 - Y*PI2 - Y*PI3 - Y*PI4;
  2) Polynomial (minimax for sin within [-Pi/2; +Pi/2] interval)
    a) Calculate X^2 = X * X
    b) Calculate polynomial:
         R = X + X * X^2 * (A3 + x^2 * (A5 + .....
  3) Destination sign setting
    a) Set shifted destination sign using XOR operation:
         R = XOR( R, S );
 */
        pushq     %rbp
        cfi_adjust_cfa_offset (8)
        cfi_rel_offset (%rbp, 0)
        movq      %rsp, %rbp
        cfi_def_cfa_register (%rbp)
        andq      $-64, %rsp
        subq      $448, %rsp
        movq      __svml_s_trig_data@GOTPCREL(%rip), %rax
        vmovaps   %ymm0, %ymm2
        vmovups __sRShifter(%rax), %ymm5
        vmovups __sPI1_FMA(%rax), %ymm7

/* b) Add Pi/2 value to argument X for Cos to Sin transformation */
        vaddps __sHalfPI(%rax), %ymm2, %ymm4

/*
  1) Range reduction to [-Pi/2; +Pi/2] interval
  c) Getting octant Y by 1/Pi multiplication
  d) Add "Right Shifter" (0x4B000000) value
 */
        vfmadd132ps __sInvPI(%rax), %ymm5, %ymm4

/* f) Subtract "Right Shifter" (0x4B000000) value */
        vsubps    %ymm5, %ymm4, %ymm6

/*
  e) Treat obtained value as integer for destination sign setting.
  Shift first bit of this value to the last (sign) position (S << 31)
 */
        vpslld    $31, %ymm4, %ymm0

/* g) Subtract 0.5 from result for octant correction */
        vsubps __sOneHalf(%rax), %ymm6, %ymm4

/* Check for large and special arguments */
        vandps __sAbsMask(%rax), %ymm2, %ymm3
        vcmpnle_uqps __sRangeReductionVal(%rax), %ymm3, %ymm1

/*
  h) Subtract Y*PI from X argument, where PI divided to 4 parts:
  X = X - Y*PI1 - Y*PI2 - Y*PI3
 */
        vmovaps   %ymm2, %ymm3
        vfnmadd231ps %ymm4, %ymm7, %ymm3
        vfnmadd231ps __sPI2_FMA(%rax), %ymm4, %ymm3
        vfnmadd132ps __sPI3_FMA(%rax), %ymm3, %ymm4

/* a) Calculate X^2 = X * X */
        vmulps    %ymm4, %ymm4, %ymm5

/*
  3) Destination sign setting
  a) Set shifted destination sign using XOR operation:
  R = XOR( R, S );
 */
        vxorps    %ymm0, %ymm4, %ymm6
        vmovups __sA9_FMA(%rax), %ymm0

/*
  b) Calculate polynomial:
  R = X + X * X^2 * (A3 + x^2 * (A5 + x^2 * (A7 + x^2 * (A9))))
 */
        vfmadd213ps __sA7_FMA(%rax), %ymm5, %ymm0
        vfmadd213ps __sA5_FMA(%rax), %ymm5, %ymm0
        vfmadd213ps __sA3(%rax), %ymm5, %ymm0
        vmulps    %ymm5, %ymm0, %ymm0
        vmovmskps %ymm1, %ecx
        vfmadd213ps %ymm6, %ymm6, %ymm0
        testl     %ecx, %ecx
        jne       .LBL_1_3

.LBL_1_2:
        cfi_remember_state
        movq      %rbp, %rsp
        cfi_def_cfa_register (%rsp)
        popq      %rbp
        cfi_adjust_cfa_offset (-8)
        cfi_restore (%rbp)
        ret

.LBL_1_3:
        cfi_restore_state
        vmovups   %ymm2, 320(%rsp)
        vmovups   %ymm0, 384(%rsp)
        je        .LBL_1_2

        xorb      %dl, %dl
        xorl      %eax, %eax
        vmovups   %ymm8, 224(%rsp)
        vmovups   %ymm9, 192(%rsp)
        vmovups   %ymm10, 160(%rsp)
        vmovups   %ymm11, 128(%rsp)
        vmovups   %ymm12, 96(%rsp)
        vmovups   %ymm13, 64(%rsp)
        vmovups   %ymm14, 32(%rsp)
        vmovups   %ymm15, (%rsp)
        movq      %rsi, 264(%rsp)
        movq      %rdi, 256(%rsp)
        movq      %r12, 296(%rsp)
        cfi_offset_rel_rsp (12, 296)
        movb      %dl, %r12b
        movq      %r13, 288(%rsp)
        cfi_offset_rel_rsp (13, 288)
        movl      %ecx, %r13d
        movq      %r14, 280(%rsp)
        cfi_offset_rel_rsp (14, 280)
        movl      %eax, %r14d
        movq      %r15, 272(%rsp)
        cfi_offset_rel_rsp (15, 272)
        cfi_remember_state

.LBL_1_6:
        btl       %r14d, %r13d
        jc        .LBL_1_12

.LBL_1_7:
        lea       1(%r14), %esi
        btl       %esi, %r13d
        jc        .LBL_1_10

.LBL_1_8:
        incb      %r12b
        addl      $2, %r14d
        cmpb      $16, %r12b
        jb        .LBL_1_6

        vmovups   224(%rsp), %ymm8
        vmovups   192(%rsp), %ymm9
        vmovups   160(%rsp), %ymm10
        vmovups   128(%rsp), %ymm11
        vmovups   96(%rsp), %ymm12
        vmovups   64(%rsp), %ymm13
        vmovups   32(%rsp), %ymm14
        vmovups   (%rsp), %ymm15
        vmovups   384(%rsp), %ymm0
        movq      264(%rsp), %rsi
        movq      256(%rsp), %rdi
        movq      296(%rsp), %r12
        cfi_restore (%r12)
        movq      288(%rsp), %r13
        cfi_restore (%r13)
        movq      280(%rsp), %r14
        cfi_restore (%r14)
        movq      272(%rsp), %r15
        cfi_restore (%r15)
        jmp       .LBL_1_2

.LBL_1_10:
        cfi_restore_state
        movzbl    %r12b, %r15d
        vmovss    324(%rsp,%r15,8), %xmm0
        vzeroupper

        call      JUMPTARGET(cosf)

        vmovss    %xmm0, 388(%rsp,%r15,8)
        jmp       .LBL_1_8

.LBL_1_12:
        movzbl    %r12b, %r15d
        vmovss    320(%rsp,%r15,8), %xmm0
        vzeroupper

        call      JUMPTARGET(cosf)

        vmovss    %xmm0, 384(%rsp,%r15,8)
        jmp       .LBL_1_7

END (_ZGVdN8v_cosf_avx2)