1 /* mpn_mul -- Multiply two natural numbers.
2 
3 Copyright (C) 1991-2022 Free Software Foundation, Inc.
4 
5 This file is part of the GNU MP Library.
6 
7 The GNU MP Library is free software; you can redistribute it and/or modify
8 it under the terms of the GNU Lesser General Public License as published by
9 the Free Software Foundation; either version 2.1 of the License, or (at your
10 option) any later version.
11 
12 The GNU MP Library is distributed in the hope that it will be useful, but
13 WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
14 or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
15 License for more details.
16 
17 You should have received a copy of the GNU Lesser General Public License
18 along with the GNU MP Library; see the file COPYING.LIB.  If not, see
19 <https://www.gnu.org/licenses/>.  */
20 
21 #include <gmp.h>
22 #include "gmp-impl.h"
23 
24 /* Multiply the natural numbers u (pointed to by UP, with USIZE limbs)
25    and v (pointed to by VP, with VSIZE limbs), and store the result at
26    PRODP.  USIZE + VSIZE limbs are always stored, but if the input
27    operands are normalized.  Return the most significant limb of the
28    result.
29 
30    NOTE: The space pointed to by PRODP is overwritten before finished
31    with U and V, so overlap is an error.
32 
33    Argument constraints:
34    1. USIZE >= VSIZE.
35    2. PRODP != UP and PRODP != VP, i.e. the destination
36       must be distinct from the multiplier and the multiplicand.  */
37 
38 /* If KARATSUBA_THRESHOLD is not already defined, define it to a
39    value which is good on most machines.  */
40 #ifndef KARATSUBA_THRESHOLD
41 #define KARATSUBA_THRESHOLD 32
42 #endif
43 
44 mp_limb_t
mpn_mul(mp_ptr prodp,mp_srcptr up,mp_size_t usize,mp_srcptr vp,mp_size_t vsize)45 mpn_mul (mp_ptr prodp,
46 	 mp_srcptr up, mp_size_t usize,
47 	 mp_srcptr vp, mp_size_t vsize)
48 {
49   mp_ptr prod_endp = prodp + usize + vsize - 1;
50   mp_limb_t cy;
51   mp_ptr tspace;
52   TMP_DECL (marker);
53 
54   if (vsize < KARATSUBA_THRESHOLD)
55     {
56       /* Handle simple cases with traditional multiplication.
57 
58 	 This is the most critical code of the entire function.  All
59 	 multiplies rely on this, both small and huge.  Small ones arrive
60 	 here immediately.  Huge ones arrive here as this is the base case
61 	 for Karatsuba's recursive algorithm below.  */
62       mp_size_t i;
63       mp_limb_t cy_limb;
64       mp_limb_t v_limb;
65 
66       if (vsize == 0)
67 	return 0;
68 
69       /* Multiply by the first limb in V separately, as the result can be
70 	 stored (not added) to PROD.  We also avoid a loop for zeroing.  */
71       v_limb = vp[0];
72       if (v_limb <= 1)
73 	{
74 	  if (v_limb == 1)
75 	    MPN_COPY (prodp, up, usize);
76 	  else
77 	    MPN_ZERO (prodp, usize);
78 	  cy_limb = 0;
79 	}
80       else
81 	cy_limb = mpn_mul_1 (prodp, up, usize, v_limb);
82 
83       prodp[usize] = cy_limb;
84       prodp++;
85 
86       /* For each iteration in the outer loop, multiply one limb from
87 	 U with one limb from V, and add it to PROD.  */
88       for (i = 1; i < vsize; i++)
89 	{
90 	  v_limb = vp[i];
91 	  if (v_limb <= 1)
92 	    {
93 	      cy_limb = 0;
94 	      if (v_limb == 1)
95 		cy_limb = mpn_add_n (prodp, prodp, up, usize);
96 	    }
97 	  else
98 	    cy_limb = mpn_addmul_1 (prodp, up, usize, v_limb);
99 
100 	  prodp[usize] = cy_limb;
101 	  prodp++;
102 	}
103       return cy_limb;
104     }
105 
106   TMP_MARK (marker);
107 
108   tspace = (mp_ptr) TMP_ALLOC (2 * vsize * BYTES_PER_MP_LIMB);
109   MPN_MUL_N_RECURSE (prodp, up, vp, vsize, tspace);
110 
111   prodp += vsize;
112   up += vsize;
113   usize -= vsize;
114   if (usize >= vsize)
115     {
116       mp_ptr tp = (mp_ptr) TMP_ALLOC (2 * vsize * BYTES_PER_MP_LIMB);
117       do
118 	{
119 	  MPN_MUL_N_RECURSE (tp, up, vp, vsize, tspace);
120 	  cy = mpn_add_n (prodp, prodp, tp, vsize);
121 	  mpn_add_1 (prodp + vsize, tp + vsize, vsize, cy);
122 	  prodp += vsize;
123 	  up += vsize;
124 	  usize -= vsize;
125 	}
126       while (usize >= vsize);
127     }
128 
129   /* True: usize < vsize.  */
130 
131   /* Make life simple: Recurse.  */
132 
133   if (usize != 0)
134     {
135       mpn_mul (tspace, vp, vsize, up, usize);
136       cy = mpn_add_n (prodp, prodp, tspace, vsize);
137       mpn_add_1 (prodp + vsize, tspace + vsize, usize, cy);
138     }
139 
140   TMP_FREE (marker);
141   return *prod_endp;
142 }
143