2 * Copyright © 2015 Intel Corporation
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
32 #if defined(__SSE__) || defined(_MSC_VER)
33 /* MSVC always has SSE nowadays */
34 #include <xmmintrin.h>
35 #include <emmintrin.h>
39 #include <smmintrin.h>
42 /* The C standard library has functions round()/rint()/nearbyint() that round
43 * their arguments according to the rounding mode set in the floating-point
44 * control register. While there are trunc()/ceil()/floor() functions that do
45 * a specific operation without modifying the rounding mode, there is no
46 * roundeven() in any version of C.
48 * Technical Specification 18661 (ISO/IEC TS 18661-1:2014) adds roundeven(),
49 * but it's unfortunately not implemented by glibc.
51 * This implementation differs in that it does not raise the inexact exception.
53 * We use rint() to implement these functions, with the assumption that the
54 * floating-point rounding mode has not been changed from the default Round
59 * \brief Rounds \c x to the nearest integer, with ties to the even integer.
62 _mesa_roundevenf(float x
)
66 __m128 m
= _mm_load_ss(&x
);
67 m
= _mm_round_ss(m
, m
, _MM_FROUND_CUR_DIRECTION
| _MM_FROUND_NO_EXC
);
68 _mm_store_ss(&ret
, m
);
76 * \brief Rounds \c x to the nearest integer, with ties to the even integer.
79 _mesa_roundeven(double x
)
83 __m128d m
= _mm_load_sd(&x
);
84 m
= _mm_round_sd(m
, m
, _MM_FROUND_CUR_DIRECTION
| _MM_FROUND_NO_EXC
);
85 _mm_store_sd(&ret
, m
);
93 * \brief Rounds \c x to the nearest integer, with ties to the even integer,
94 * and returns the value as a long int.
97 _mesa_lroundevenf(float x
)
99 #if defined(__SSE__) || defined(_MSC_VER)
100 #if LONG_MAX == INT64_MAX
101 return _mm_cvtss_si64(_mm_load_ss(&x
));
102 #elif LONG_MAX == INT32_MAX
103 return _mm_cvtss_si32(_mm_load_ss(&x
));
105 #error "Unsupported long size"
113 * \brief Rounds \c x to the nearest integer, with ties to the even integer,
114 * and returns the value as a long int.
117 _mesa_lroundeven(double x
)
119 #if defined(__SSE__) || defined(_MSC_VER)
120 #if LONG_MAX == INT64_MAX
121 return _mm_cvtsd_si64(_mm_load_sd(&x
));
122 #elif LONG_MAX == INT32_MAX
123 return _mm_cvtsd_si32(_mm_load_sd(&x
));
125 #error "Unsupported long size"