source/libs/gmp/gmp-src/mpn/x86_64/pentium4/lshiftc.asm

   1 dnl  x86-64 mpn_lshiftc optimized for Pentium 4.
   2
   3 dnl  Copyright 2003, 2005, 2007, 2008, 2010, 2012 Free Software Foundation,
   4 dnl  Inc.
   5
   6 dnl  This file is part of the GNU MP Library.
   7 dnl
   8 dnl  The GNU MP Library is free software; you can redistribute it and/or modify
   9 dnl  it under the terms of either:
  10 dnl
  11 dnl    * the GNU Lesser General Public License as published by the Free
  12 dnl      Software Foundation; either version 3 of the License, or (at your
  13 dnl      option) any later version.
  14 dnl
  15 dnl  or
  16 dnl
  17 dnl    * the GNU General Public License as published by the Free Software
  18 dnl      Foundation; either version 2 of the License, or (at your option) any
  19 dnl      later version.
  20 dnl
  21 dnl  or both in parallel, as here.
  22 dnl
  23 dnl  The GNU MP Library is distributed in the hope that it will be useful, but
  24 dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
  25 dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
  26 dnl  for more details.
  27 dnl
  28 dnl  You should have received copies of the GNU General Public License and the
  29 dnl  GNU Lesser General Public License along with the GNU MP Library.  If not,
  30 dnl  see https://www.gnu.org/licenses/.
  31
  32 include(`../config.m4')
  33
  34
  35 C            cycles/limb
  36 C AMD K8,K9      ?
  37 C AMD K10        ?
  38 C Intel P4       4.15
  39 C Intel core2    ?
  40 C Intel corei    ?
  41 C Intel atom     ?
  42 C VIA nano       ?
  43
  44 C INPUT PARAMETERS
  45 define(`rp',`%rdi')
  46 define(`up',`%rsi')
  47 define(`n',`%rdx')
  48 define(`cnt',`%cl')
  49
  50 ABI_SUPPORT(DOS64)
  51 ABI_SUPPORT(STD64)
  52
  53 ASM_START()
  54         TEXT
  55         ALIGN(32)
  56 PROLOGUE(mpn_lshiftc)
  57         FUNC_ENTRY(4)
  58         mov     -8(up,n,8), %rax
  59         pcmpeqd %mm6, %mm6              C 0xffff...fff
  60         movd    R32(%rcx), %mm4
  61         neg     R32(%rcx)               C put rsh count in cl
  62         and     $63, R32(%rcx)
  63         movd    R32(%rcx), %mm5
  64
  65         lea     1(n), R32(%r8)
  66
  67         shr     R8(%rcx), %rax          C function return value
  68
  69         and     $3, R32(%r8)
  70         je      L(rol)                  C jump for n = 3, 7, 11, ...
  71
  72         dec     R32(%r8)
  73         jne     L(1)
  74 C       n = 4, 8, 12, ...
  75         movq    -8(up,n,8), %mm2
  76         psllq   %mm4, %mm2
  77         movq    -16(up,n,8), %mm0
  78         pxor    %mm6, %mm2
  79         psrlq   %mm5, %mm0
  80         pandn   %mm2, %mm0
  81         movq    %mm0, -8(rp,n,8)
  82         dec     n
  83         jmp     L(rol)
  84
  85 L(1):   dec     R32(%r8)
  86         je      L(1x)                   C jump for n = 1, 5, 9, 13, ...
  87 C       n = 2, 6, 10, 16, ...
  88         movq    -8(up,n,8), %mm2
  89         psllq   %mm4, %mm2
  90         movq    -16(up,n,8), %mm0
  91         pxor    %mm6, %mm2
  92         psrlq   %mm5, %mm0
  93         pandn   %mm2, %mm0
  94         movq    %mm0, -8(rp,n,8)
  95         dec     n
  96 L(1x):
  97         cmp     $1, n
  98         je      L(ast)
  99         movq    -8(up,n,8), %mm2
 100         psllq   %mm4, %mm2
 101         movq    -16(up,n,8), %mm3
 102         psllq   %mm4, %mm3
 103         movq    -16(up,n,8), %mm0
 104         movq    -24(up,n,8), %mm1
 105         pxor    %mm6, %mm2
 106         psrlq   %mm5, %mm0
 107         pandn   %mm2, %mm0
 108         pxor    %mm6, %mm3
 109         psrlq   %mm5, %mm1
 110         pandn   %mm3, %mm1
 111         movq    %mm0, -8(rp,n,8)
 112         movq    %mm1, -16(rp,n,8)
 113         sub     $2, n
 114
 115 L(rol): movq    -8(up,n,8), %mm2
 116         psllq   %mm4, %mm2
 117         movq    -16(up,n,8), %mm3
 118         psllq   %mm4, %mm3
 119
 120         sub     $4, n
 121         jb      L(end)
 122         ALIGN(32)
 123 L(top):
 124         C finish stuff from lsh block
 125         movq    16(up,n,8), %mm0
 126         pxor    %mm6, %mm2
 127         movq    8(up,n,8), %mm1
 128         psrlq   %mm5, %mm0
 129         psrlq   %mm5, %mm1
 130         pandn   %mm2, %mm0
 131         pxor    %mm6, %mm3
 132         movq    %mm0, 24(rp,n,8)
 133         movq    (up,n,8), %mm0
 134         pandn   %mm3, %mm1
 135         movq    %mm1, 16(rp,n,8)
 136         movq    -8(up,n,8), %mm1
 137         C start two new rsh
 138         psrlq   %mm5, %mm0
 139         psrlq   %mm5, %mm1
 140
 141         C finish stuff from rsh block
 142         movq    8(up,n,8), %mm2
 143         pxor    %mm6, %mm0
 144         movq    (up,n,8), %mm3
 145         psllq   %mm4, %mm2
 146         psllq   %mm4, %mm3
 147         pandn   %mm0, %mm2
 148         pxor    %mm6, %mm1
 149         movq    %mm2, 8(rp,n,8)
 150         movq    -8(up,n,8), %mm2
 151         pandn   %mm1, %mm3
 152         movq    %mm3, (rp,n,8)
 153         movq    -16(up,n,8), %mm3
 154         C start two new lsh
 155         sub     $4, n
 156         psllq   %mm4, %mm2
 157         psllq   %mm4, %mm3
 158
 159         jae     L(top)
 160
 161 L(end): pxor    %mm6, %mm2
 162         movq    8(up), %mm0
 163         psrlq   %mm5, %mm0
 164         pandn   %mm2, %mm0
 165         pxor    %mm6, %mm3
 166         movq    (up), %mm1
 167         psrlq   %mm5, %mm1
 168         pandn   %mm3, %mm1
 169         movq    %mm0, 16(rp)
 170         movq    %mm1, 8(rp)
 171
 172 L(ast): movq    (up), %mm2
 173         psllq   %mm4, %mm2
 174         pxor    %mm6, %mm2
 175         movq    %mm2, (rp)
 176         emms
 177         FUNC_EXIT()
 178         ret
 179 EPILOGUE()