1 /*

     2  * Copyright © 2004 Red Hat, Inc.

     3  * Copyright © 2005 Trolltech AS

     4  *

     5  * Permission to use, copy, modify, distribute, and sell this software and its

     6  * documentation for any purpose is hereby granted without fee, provided that

     7  * the above copyright notice appear in all copies and that both that

     8  * copyright notice and this permission notice appear in supporting

     9  * documentation, and that the name of Red Hat not be used in advertising or

    10  * publicity pertaining to distribution of the software without specific,

    11  * written prior permission.  Red Hat makes no representations about the

    12  * suitability of this software for any purpose.  It is provided "as is"

    13  * without express or implied warranty.

    14  *

    15  * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS

    16  * SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND

    17  * FITNESS, IN NO EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY

    18  * SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES

    19  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN

    20  * AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING

    21  * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS

    22  * SOFTWARE.

    23  *

    24  * Author:  Søren Sandmann (sandmann@redhat.com)

    25  *          Lars Knoll (lars@trolltech.com)

    26  * 

    27  * Based on work by Owen Taylor

    28  */

    29 

    30 

    31 #define MmxNegate(x) _mm_xor_si64((x), mmx_4x00ff)

    32 #define MmxAlpha(x) _mm_shuffle_pi16 ((x), _MM_SHUFFLE(3, 3, 3, 3));

    33 #define MmxTo(x) _mm_unpacklo_pi8 (_mm_cvtsi32_si64((x)), mmx_0)

    34 #define MmxFrom(x) (uint32_t)_mm_cvtsi64_si32(_mm_packs_pu16((x), mmx_0))

    35 #define MmxAdd(x, y) _mm_adds_pu16 ((x), (y))

    36 

    37 #define MmxMulAdd(x, a, y) do {                     \

    38         x = _mm_mullo_pi16 (x, a);                  \

    39         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \

    40         x = _mm_adds_pu16 (x, mmx_4x0080);           \

    41         x = _mm_srli_pi16 (x, 8);                   \

    42         x = _mm_adds_pu16(x, y);                    \

    43     } while (0)

    44 

    45 #define MmxMul(x, a) do {                           \

    46         x = _mm_mullo_pi16 (x, a);                  \

    47         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \

    48         x = _mm_adds_pu16 (x, mmx_4x0080);           \

    49         x = _mm_srli_pi16 (x, 8);                   \

    50     } while (0)

    51 

    52 #define MmxAddMul(x, a, y, b) do {                  \

    53         x = _mm_mullo_pi16 (x, a);                  \

    54         y = _mm_mullo_pi16 (y, b);                  \

    55         x = _mm_srli_pi16(x, 1);                    \

    56         y = _mm_srli_pi16(y, 1);                    \

    57         x = _mm_adds_pu16 (x, y);                    \

    58         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \

    59         x = _mm_adds_pu16 (x, mmx_4x0080);           \

    60         x = _mm_srli_pi16 (x, 7);                   \

    61     } while (0)

    62