Return to copyi.asm CVS log

Up to [local] / OpenXM_contrib / gmp / mpn / x86 / pentium4

Annotation of OpenXM_contrib/gmp/mpn/x86/pentium4/copyi.asm, Revision 1.1.1.1

1.1       ohara       1: dnl  Pentium-4 mpn_copyi -- copy limb vector, incrementing.
                      2: dnl
                      3:
                      4: dnl  Copyright 1999, 2000, 2001 Free Software Foundation, Inc.
                      5: dnl
                      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
                      9: dnl  modify it under the terms of the GNU Lesser General Public License as
                     10: dnl  published by the Free Software Foundation; either version 2.1 of the
                     11: dnl  License, or (at your option) any later version.
                     12: dnl
                     13: dnl  The GNU MP Library is distributed in the hope that it will be useful,
                     14: dnl  but WITHOUT ANY WARRANTY; without even the implied warranty of
                     15: dnl  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
                     16: dnl  Lesser General Public License for more details.
                     17: dnl
                     18: dnl  You should have received a copy of the GNU Lesser General Public
                     19: dnl  License along with the GNU MP Library; see the file COPYING.LIB.  If
                     20: dnl  not, write to the Free Software Foundation, Inc., 59 Temple Place -
                     21: dnl  Suite 330, Boston, MA 02111-1307, USA.
                     22:
                     23:
                     24: dnl  The rep/movsl is very slow for small blocks on pentium4.  Its startup
                     25: dnl  time seems to be about 110 cycles.  It then copies at a rate of one
                     26: dnl  limb per cycle.  We therefore fall back to an open-coded 2 c/l copying
                     27: dnl  loop for smaller sizes.
                     28:
                     29: dnl  Ultimately, we may want to use 64-bit movd or 128-bit movdqu in some
                     30: dnl  nifty unrolled arrangement.  Clearly, that could reach much higher
                     31: dnl  speeds, at least for large blocks.
                     32:
                     33: include(`../config.m4')
                     34:
                     35:
                     36: defframe(PARAM_SIZE, 12)
                     37: defframe(PARAM_SRC, 8)
                     38: defframe(PARAM_DST,  4)
                     39:
                     40:        TEXT
                     41:        ALIGN(8)
                     42:
                     43: PROLOGUE(mpn_copyi)
                     44: deflit(`FRAME',0)
                     45:
                     46:        movl    PARAM_SIZE, %ecx
                     47:        cmpl    $150, %ecx
                     48:        jg      L(replmovs)
                     49:
                     50:        movl    PARAM_SRC, %eax
                     51:        movl    PARAM_DST, %edx
                     52:        movl    %ebx, PARAM_SIZE
                     53:        testl   %ecx, %ecx
                     54:        jz      L(end)
                     55:
                     56: L(loop):
                     57:        movl    (%eax), %ebx
                     58:        leal    4(%eax), %eax
                     59:        addl    $-1, %ecx
                     60:        movl    %ebx, (%edx)
                     61:        leal    4(%edx), %edx
                     62:
                     63:        jnz     L(loop)
                     64:
                     65: L(end):
                     66:        movl    PARAM_SIZE, %ebx
                     67:        ret
                     68:
                     69: L(replmovs):
                     70:        cld     C better safe than sorry, see mpn/x86/README
                     71:
                     72:        movl    %esi, %eax
                     73:        movl    PARAM_SRC, %esi
                     74:        movl    %edi, %edx
                     75:        movl    PARAM_DST, %edi
                     76:
                     77:        rep
                     78:        movsl
                     79:
                     80:        movl    %eax, %esi
                     81:        movl    %edx, %edi
                     82:
                     83:        ret
                     84:
                     85: EPILOGUE()