1/* Copyright (C) 2007-2009 Xiph.Org Foundation
2   Copyright (C) 2003-2008 Jean-Marc Valin
3   Copyright (C) 2007-2008 CSIRO
4   Copyright (C) 2013      Parrot */
5/*
6   Redistribution and use in source and binary forms, with or without
7   modification, are permitted provided that the following conditions
8   are met:
9
10   - Redistributions of source code must retain the above copyright
11   notice, this list of conditions and the following disclaimer.
12
13   - Redistributions in binary form must reproduce the above copyright
14   notice, this list of conditions and the following disclaimer in the
15   documentation and/or other materials provided with the distribution.
16
17   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
18   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
19   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
20   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
21   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
22   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
23   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
24   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
25   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
26   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
27   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
28*/
29
30#ifndef FIXED_ARMv5E_H
31#define FIXED_ARMv5E_H
32
33#include "fixed_armv4.h"
34
35/** 16x32 multiplication, followed by a 16-bit shift right. Results fits in 32 bits */
36#undef MULT16_32_Q16
37static OPUS_INLINE opus_val32 MULT16_32_Q16_armv5e(opus_val16 a, opus_val32 b)
38{
39  int res;
40  __asm__(
41      "#MULT16_32_Q16\n\t"
42      "smulwb %0, %1, %2\n\t"
43      : "=r"(res)
44      : "r"(b),"r"(a)
45  );
46  return res;
47}
48#define MULT16_32_Q16(a, b) (MULT16_32_Q16_armv5e(a, b))
49
50
51/** 16x32 multiplication, followed by a 15-bit shift right. Results fits in 32 bits */
52#undef MULT16_32_Q15
53static OPUS_INLINE opus_val32 MULT16_32_Q15_armv5e(opus_val16 a, opus_val32 b)
54{
55  int res;
56  __asm__(
57      "#MULT16_32_Q15\n\t"
58      "smulwb %0, %1, %2\n\t"
59      : "=r"(res)
60      : "r"(b), "r"(a)
61  );
62  return res<<1;
63}
64#define MULT16_32_Q15(a, b) (MULT16_32_Q15_armv5e(a, b))
65
66
67/** 16x32 multiply, followed by a 15-bit shift right and 32-bit add.
68    b must fit in 31 bits.
69    Result fits in 32 bits. */
70#undef MAC16_32_Q15
71static OPUS_INLINE opus_val32 MAC16_32_Q15_armv5e(opus_val32 c, opus_val16 a,
72 opus_val32 b)
73{
74  int res;
75  __asm__(
76      "#MAC16_32_Q15\n\t"
77      "smlawb %0, %1, %2, %3;\n"
78      : "=r"(res)
79      : "r"(b<<1), "r"(a), "r"(c)
80  );
81  return res;
82}
83#define MAC16_32_Q15(c, a, b) (MAC16_32_Q15_armv5e(c, a, b))
84
85/** 16x16 multiply-add where the result fits in 32 bits */
86#undef MAC16_16
87static OPUS_INLINE opus_val32 MAC16_16_armv5e(opus_val32 c, opus_val16 a,
88 opus_val16 b)
89{
90  int res;
91  __asm__(
92      "#MAC16_16\n\t"
93      "smlabb %0, %1, %2, %3;\n"
94      : "=r"(res)
95      : "r"(a), "r"(b), "r"(c)
96  );
97  return res;
98}
99#define MAC16_16(c, a, b) (MAC16_16_armv5e(c, a, b))
100
101/** 16x16 multiplication where the result fits in 32 bits */
102#undef MULT16_16
103static OPUS_INLINE opus_val32 MULT16_16_armv5e(opus_val16 a, opus_val16 b)
104{
105  int res;
106  __asm__(
107      "#MULT16_16\n\t"
108      "smulbb %0, %1, %2;\n"
109      : "=r"(res)
110      : "r"(a), "r"(b)
111  );
112  return res;
113}
114#define MULT16_16(a, b) (MULT16_16_armv5e(a, b))
115
116#endif
117