Return to README CVS log

Up to [local] / OpenXM_contrib / gmp / mpn / alpha / ev6 / nails

Annotation of OpenXM_contrib/gmp/mpn/alpha/ev6/nails/README, Revision 1.1.1.1

1.1       ohara       1: Copyright 2002 Free Software Foundation, Inc.
                      2:
                      3: This file is part of the GNU MP Library.
                      4:
                      5: The GNU MP Library is free software; you can redistribute it and/or modify it
                      6: under the terms of the GNU Lesser General Public License as published by the
                      7: Free Software Foundation; either version 2.1 of the License, or (at your
                      8: option) any later version.
                      9:
                     10: The GNU MP Library is distributed in the hope that it will be useful, but
                     11: WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
                     12: FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public License
                     13: for more details.
                     14:
                     15: You should have received a copy of the GNU Lesser General Public License along
                     16: with the GNU MP Library; see the file COPYING.LIB.  If not, write to the Free
                     17: Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307,
                     18: USA.
                     19:
                     20:
                     21:
                     22:
                     23:
                     24: This directory contains preliminary assembly code for nails-enabled
                     25: 21264.  The code is not very well optimized.
                     26:
                     27: 1. The *mul_* functions use too many registers.  Several temps could be
                     28:    combined.
                     29: 2. The *mul_1 functions load from up and rp very early.  The code could be
                     30:    considerably simplified with later loads.
                     31: 3. The addmul_1 and submul_1 functions run 1.25 cycles/limb slower than
                     32:    possible with better U0/U1/L0/L1 scheduling.