Home | History | Annotate | Line # | Download | only in generic
toom_eval_pm2.c revision 1.1.1.1
      1 /* mpn_toom_eval_pm2 -- Evaluate a polynomial in +2 and -2
      2 
      3    Contributed to the GNU project by Niels Mller and Marco Bodrato
      4 
      5    THE FUNCTION IN THIS FILE IS INTERNAL WITH A MUTABLE INTERFACE.  IT IS ONLY
      6    SAFE TO REACH IT THROUGH DOCUMENTED INTERFACES.  IN FACT, IT IS ALMOST
      7    GUARANTEED THAT IT WILL CHANGE OR DISAPPEAR IN A FUTURE GNU MP RELEASE.
      8 
      9 Copyright 2009 Free Software Foundation, Inc.
     10 
     11 This file is part of the GNU MP Library.
     12 
     13 The GNU MP Library is free software; you can redistribute it and/or modify
     14 it under the terms of the GNU Lesser General Public License as published by
     15 the Free Software Foundation; either version 3 of the License, or (at your
     16 option) any later version.
     17 
     18 The GNU MP Library is distributed in the hope that it will be useful, but
     19 WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
     20 or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
     21 License for more details.
     22 
     23 You should have received a copy of the GNU Lesser General Public License
     24 along with the GNU MP Library.  If not, see http://www.gnu.org/licenses/.  */
     25 
     26 #include "gmp.h"
     27 #include "gmp-impl.h"
     28 
     29 /* DO_addlsh2(d,a,b,n,cy) computes cy,{d,n} <- {a,n} + 4*(cy,{b,n}), it
     30    can be used as DO_addlsh2(d,a,d,n,d[n]), for accumulation on {d,n+1}. */
     31 #if HAVE_NATIVE_mpn_addlsh2_n
     32 #define DO_addlsh2(d, a, b, n, cy)	\
     33 do {					\
     34   (cy) <<= 2;				\
     35   (cy) += mpn_addlsh2_n(d, a, b, n);	\
     36 } while (0)
     37 #else
     38 #if HAVE_NATIVE_mpn_addlsh_n
     39 #define DO_addlsh2(d, a, b, n, cy)	\
     40 do {					\
     41   (cy) <<= 2;				\
     42   (cy) += mpn_addlsh_n(d, a, b, n, 2);	\
     43 } while (0)
     44 #else
     45 /* The following is not a general substitute for addlsh2.
     46    It is correct if d == b, but it is not if d == a.	*/
     47 #define DO_addlsh2(d, a, b, n, cy)	\
     48 do {					\
     49   (cy) <<= 2;				\
     50   (cy) += mpn_lshift(d, b, n, 2);	\
     51   (cy) += mpn_add_n(d, d, a, n);	\
     52 } while (0)
     53 #endif
     54 #endif
     55 
     56 /* Evaluates a polynomial of degree 2 < k < GMP_NUMB_BITS, in the
     57    points +2 and -2. */
     58 int
     59 mpn_toom_eval_pm2 (mp_ptr xp2, mp_ptr xm2, unsigned k,
     60 		   mp_srcptr xp, mp_size_t n, mp_size_t hn, mp_ptr tp)
     61 {
     62   int i;
     63   int neg;
     64   mp_limb_t cy;
     65 
     66   ASSERT (k >= 3);
     67   ASSERT (k < GMP_NUMB_BITS);
     68 
     69   ASSERT (hn > 0);
     70   ASSERT (hn <= n);
     71 
     72   /* The degree k is also the number of full-size coefficients, so
     73    * that last coefficient, of size hn, starts at xp + k*n. */
     74 
     75   cy = 0;
     76   DO_addlsh2 (xp2, xp + (k-2) * n, xp + k * n, hn, cy);
     77   if (hn != n)
     78     cy = mpn_add_1 (xp2 + hn, xp + (k-2) * n + hn, n - hn, cy);
     79   for (i = k - 4; i >= 0; i -= 2)
     80     DO_addlsh2 (xp2, xp + i * n, xp2, n, cy);
     81   xp2[n] = cy;
     82 
     83   k--;
     84 
     85   cy = 0;
     86   DO_addlsh2 (tp, xp + (k-2) * n, xp + k * n, n, cy);
     87   for (i = k - 4; i >= 0; i -= 2)
     88     DO_addlsh2 (tp, xp + i * n, tp, n, cy);
     89   tp[n] = cy;
     90 
     91   if (k & 1)
     92     ASSERT_NOCARRY(mpn_lshift (tp , tp , n + 1, 1));
     93   else
     94     ASSERT_NOCARRY(mpn_lshift (xp2, xp2, n + 1, 1));
     95 
     96   neg = (mpn_cmp (xp2, tp, n + 1) < 0) ? ~0 : 0;
     97 
     98 #if HAVE_NATIVE_mpn_add_n_sub_n
     99   if (neg)
    100     mpn_add_n_sub_n (xp2, xm2, tp, xp2, n + 1);
    101   else
    102     mpn_add_n_sub_n (xp2, xm2, xp2, tp, n + 1);
    103 #else /* !HAVE_NATIVE_mpn_add_n_sub_n */
    104   if (neg)
    105     mpn_sub_n (xm2, tp, xp2, n + 1);
    106   else
    107     mpn_sub_n (xm2, xp2, tp, n + 1);
    108 
    109   mpn_add_n (xp2, xp2, tp, n + 1);
    110 #endif /* !HAVE_NATIVE_mpn_add_n_sub_n */
    111 
    112   ASSERT (xp2[n] < (1<<(k+2))-1);
    113   ASSERT (xm2[n] < ((1<<(k+3))-1 - (1^k&1))/3);
    114 
    115   neg ^= ((k & 1) - 1);
    116 
    117   return neg;
    118 }
    119 
    120 #undef DO_addlsh2
    121