source/libs/gmp/gmp-src/mpn/x86_64/mul_2.asm

   1 dnl  AMD64 mpn_mul_2 -- Multiply an n-limb vector with a 2-limb vector and
   2 dnl  store the result in a third limb vector.
   3
   4 dnl  Copyright 2008, 2011, 2012 Free Software Foundation, 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 C            cycles/limb
  35 C AMD K8,K9      2.275
  36 C AMD K10        2.275
  37 C Intel P4      13.5
  38 C Intel core2    4.0
  39 C Intel corei    3.8
  40 C Intel atom     ?
  41 C VIA nano       ?
  42
  43 C This code is the result of running a code generation and optimization tool
  44 C suite written by David Harvey and Torbjorn Granlund.
  45
  46 C TODO
  47 C  * Work on feed-in and wind-down code.
  48 C  * Convert "mov $0" to "xor".
  49 C  * Adjust initial lea to save some bytes.
  50 C  * Perhaps adjust n from n_param&3 value?
  51 C  * Replace with 2.25 c/l sequence.
  52
  53 C INPUT PARAMETERS
  54 define(`rp',     `%rdi')
  55 define(`up',     `%rsi')
  56 define(`n_param',`%rdx')
  57 define(`vp',     `%rcx')
  58
  59 define(`v0', `%r8')
  60 define(`v1', `%r9')
  61 define(`w0', `%rbx')
  62 define(`w1', `%rcx')
  63 define(`w2', `%rbp')
  64 define(`w3', `%r10')
  65 define(`n',  `%r11')
  66
  67 ABI_SUPPORT(DOS64)
  68 ABI_SUPPORT(STD64)
  69
  70 ASM_START()
  71         TEXT
  72         ALIGN(16)
  73 PROLOGUE(mpn_mul_2)
  74         FUNC_ENTRY(4)
  75         push    %rbx
  76         push    %rbp
  77
  78         mov     (vp), v0
  79         mov     8(vp), v1
  80
  81         mov     (up), %rax
  82
  83         mov     n_param, n
  84         neg     n
  85         lea     -8(up,n_param,8), up
  86         lea     -8(rp,n_param,8), rp
  87
  88         and     $3, R32(n_param)
  89         jz      L(m2p0)
  90         cmp     $2, R32(n_param)
  91         jc      L(m2p1)
  92         jz      L(m2p2)
  93 L(m2p3):
  94         mul     v0
  95         xor     R32(w3), R32(w3)
  96         mov     %rax, w1
  97         mov     %rdx, w2
  98         mov     8(up,n,8), %rax
  99         add     $-1, n
 100         mul     v1
 101         add     %rax, w2
 102         jmp     L(m23)
 103 L(m2p0):
 104         mul     v0
 105         xor     R32(w2), R32(w2)
 106         mov     %rax, w0
 107         mov     %rdx, w1
 108         jmp     L(m20)
 109 L(m2p1):
 110         mul     v0
 111         xor     R32(w3), R32(w3)
 112         xor     R32(w0), R32(w0)
 113         xor     R32(w1), R32(w1)
 114         add     $1, n
 115         jmp     L(m2top)
 116 L(m2p2):
 117         mul     v0
 118         xor     R32(w0), R32(w0)
 119         xor     R32(w1), R32(w1)
 120         mov     %rax, w2
 121         mov     %rdx, w3
 122         mov     8(up,n,8), %rax
 123         add     $-2, n
 124         jmp     L(m22)
 125
 126
 127         ALIGN(32)
 128 L(m2top):
 129         add     %rax, w3
 130         adc     %rdx, w0
 131         mov     0(up,n,8), %rax
 132         adc     $0, R32(w1)
 133         mov     $0, R32(w2)
 134         mul     v1
 135         add     %rax, w0
 136         mov     w3, 0(rp,n,8)
 137         adc     %rdx, w1
 138         mov     8(up,n,8), %rax
 139         mul     v0
 140         add     %rax, w0
 141         adc     %rdx, w1
 142         adc     $0, R32(w2)
 143 L(m20): mov     8(up,n,8), %rax
 144         mul     v1
 145         add     %rax, w1
 146         adc     %rdx, w2
 147         mov     16(up,n,8), %rax
 148         mov     $0, R32(w3)
 149         mul     v0
 150         add     %rax, w1
 151         mov     16(up,n,8), %rax
 152         adc     %rdx, w2
 153         adc     $0, R32(w3)
 154         mul     v1
 155         add     %rax, w2
 156         mov     w0, 8(rp,n,8)
 157 L(m23): adc     %rdx, w3
 158         mov     24(up,n,8), %rax
 159         mul     v0
 160         mov     $0, R32(w0)
 161         add     %rax, w2
 162         adc     %rdx, w3
 163         mov     w1, 16(rp,n,8)
 164         mov     24(up,n,8), %rax
 165         mov     $0, R32(w1)
 166         adc     $0, R32(w0)
 167 L(m22): mul     v1
 168         add     %rax, w3
 169         mov     w2, 24(rp,n,8)
 170         adc     %rdx, w0
 171         mov     32(up,n,8), %rax
 172         mul     v0
 173         add     $4, n
 174         js      L(m2top)
 175
 176
 177         add     %rax, w3
 178         adc     %rdx, w0
 179         adc     $0, R32(w1)
 180         mov     (up), %rax
 181         mul     v1
 182         mov     w3, (rp)
 183         add     %rax, w0
 184         adc     %rdx, w1
 185         mov     w0, 8(rp)
 186         mov     w1, %rax
 187
 188         pop     %rbp
 189         pop     %rbx
 190         FUNC_EXIT()
 191         ret
 192 EPILOGUE()