sysdeps/i386/i586/memcopy.h

   1 /* memcopy.h -- definitions for memory copy functions.  Pentium version.
   2    Copyright (C) 1994-2014 Free Software Foundation, Inc.
   3    Contributed by Torbjorn Granlund (tege@sics.se).
   4    This file is part of the GNU C Library.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 /* Get the i386 definitions.  We will override some of them below.  */
  21 #include <sysdeps/i386/memcopy.h>
  22
  23 /* Written like this, the Pentium pipeline can execute the loop at a
  24    sustained rate of 2 instructions/clock, or asymptotically 480
  25    Mbytes/second at 60Mhz.  */
  26
  27 #undef  WORD_COPY_FWD
  28 #define WORD_COPY_FWD(dst_bp, src_bp, nbytes_left, nbytes)              \
  29   do                                                                    \
  30     {                                                                   \
  31       asm volatile ("subl       $32,%2\n"                               \
  32                     "js         2f\n"                                   \
  33                     "movl       0(%0),%%edx\n"  /* alloc dest line */   \
  34                     "1:\n"                                              \
  35                     "movl       28(%0),%%eax\n" /* alloc dest line */   \
  36                     "subl       $32,%2\n"       /* decr loop count */   \
  37                     "movl       0(%1),%%eax\n"  /* U pipe */            \
  38                     "movl       4(%1),%%edx\n"  /* V pipe */            \
  39                     "movl       %%eax,0(%0)\n"  /* U pipe */            \
  40                     "movl       %%edx,4(%0)\n"  /* V pipe */            \
  41                     "movl       8(%1),%%eax\n"                          \
  42                     "movl       12(%1),%%edx\n"                         \
  43                     "movl       %%eax,8(%0)\n"                          \
  44                     "movl       %%edx,12(%0)\n"                         \
  45                     "movl       16(%1),%%eax\n"                         \
  46                     "movl       20(%1),%%edx\n"                         \
  47                     "movl       %%eax,16(%0)\n"                         \
  48                     "movl       %%edx,20(%0)\n"                         \
  49                     "movl       24(%1),%%eax\n"                         \
  50                     "movl       28(%1),%%edx\n"                         \
  51                     "movl       %%eax,24(%0)\n"                         \
  52                     "movl       %%edx,28(%0)\n"                         \
  53                     "leal       32(%1),%1\n"    /* update src ptr */    \
  54                     "leal       32(%0),%0\n"    /* update dst ptr */    \
  55                     "jns        1b\n"                                   \
  56                     "2: addl    $32,%2" :                               \
  57                     "=r" (dst_bp), "=r" (src_bp), "=r" (nbytes_left) :  \
  58                     "0" (dst_bp), "1" (src_bp), "2" (nbytes) :          \
  59                     "ax", "dx");                                        \
  60     } while (0)
  61
  62 #undef  WORD_COPY_BWD
  63 #define WORD_COPY_BWD(dst_ep, src_ep, nbytes_left, nbytes)              \
  64   do                                                                    \
  65     {                                                                   \
  66       asm volatile ("subl       $32,%2\n"                               \
  67                     "js         2f\n"                                   \
  68                     "movl       -4(%0),%%edx\n"                         \
  69                     "1:\n"                                              \
  70                     "movl       -32(%0),%%eax\n"                        \
  71                     "subl       $32,%2\n"                               \
  72                     "movl       -4(%1),%%eax\n"                         \
  73                     "movl       -8(%1),%%edx\n"                         \
  74                     "movl       %%eax,-4(%0)\n"                         \
  75                     "movl       %%edx,-8(%0)\n"                         \
  76                     "movl       -12(%1),%%eax\n"                        \
  77                     "movl       -16(%1),%%edx\n"                        \
  78                     "movl       %%eax,-12(%0)\n"                        \
  79                     "movl       %%edx,-16(%0)\n"                        \
  80                     "movl       -20(%1),%%eax\n"                        \
  81                     "movl       -24(%1),%%edx\n"                        \
  82                     "movl       %%eax,-20(%0)\n"                        \
  83                     "movl       %%edx,-24(%0)\n"                        \
  84                     "movl       -28(%1),%%eax\n"                        \
  85                     "movl       -32(%1),%%edx\n"                        \
  86                     "movl       %%eax,-28(%0)\n"                        \
  87                     "movl       %%edx,-32(%0)\n"                        \
  88                     "leal       -32(%1),%1\n"                           \
  89                     "leal       -32(%0),%0\n"                           \
  90                     "jns        1b\n"                                   \
  91                     "2: addl    $32,%2" :                               \
  92                     "=r" (dst_ep), "=r" (src_ep), "=r" (nbytes_left) :  \
  93                     "0" (dst_ep), "1" (src_ep), "2" (nbytes) :          \
  94                     "ax", "dx");                                        \
  95     } while (0)