sl@0: /*
sl@0:  * Copyright © 2004 Red Hat, Inc.
sl@0:  * Copyright © 2005 Trolltech AS
sl@0:  *
sl@0:  * Permission to use, copy, modify, distribute, and sell this software and its
sl@0:  * documentation for any purpose is hereby granted without fee, provided that
sl@0:  * the above copyright notice appear in all copies and that both that
sl@0:  * copyright notice and this permission notice appear in supporting
sl@0:  * documentation, and that the name of Red Hat not be used in advertising or
sl@0:  * publicity pertaining to distribution of the software without specific,
sl@0:  * written prior permission.  Red Hat makes no representations about the
sl@0:  * suitability of this software for any purpose.  It is provided "as is"
sl@0:  * without express or implied warranty.
sl@0:  *
sl@0:  * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS
sl@0:  * SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND
sl@0:  * FITNESS, IN NO EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY
sl@0:  * SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
sl@0:  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
sl@0:  * AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING
sl@0:  * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
sl@0:  * SOFTWARE.
sl@0:  *
sl@0:  * Author:  Søren Sandmann (sandmann@redhat.com)
sl@0:  *          Lars Knoll (lars@trolltech.com)
sl@0:  * 
sl@0:  * Based on work by Owen Taylor
sl@0:  */
sl@0: 
sl@0: 
sl@0: #define MmxNegate(x) _mm_xor_si64((x), mmx_4x00ff)
sl@0: #define MmxAlpha(x) _mm_shuffle_pi16 ((x), _MM_SHUFFLE(3, 3, 3, 3));
sl@0: #define MmxTo(x) _mm_unpacklo_pi8 (_mm_cvtsi32_si64((x)), mmx_0)
sl@0: #define MmxFrom(x) (uint32_t)_mm_cvtsi64_si32(_mm_packs_pu16((x), mmx_0))
sl@0: #define MmxAdd(x, y) _mm_adds_pu16 ((x), (y))
sl@0: 
sl@0: #define MmxMulAdd(x, a, y) do {                     \
sl@0:         x = _mm_mullo_pi16 (x, a);                  \
sl@0:         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \
sl@0:         x = _mm_adds_pu16 (x, mmx_4x0080);           \
sl@0:         x = _mm_srli_pi16 (x, 8);                   \
sl@0:         x = _mm_adds_pu16(x, y);                    \
sl@0:     } while (0)
sl@0: 
sl@0: #define MmxMul(x, a) do {                           \
sl@0:         x = _mm_mullo_pi16 (x, a);                  \
sl@0:         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \
sl@0:         x = _mm_adds_pu16 (x, mmx_4x0080);           \
sl@0:         x = _mm_srli_pi16 (x, 8);                   \
sl@0:     } while (0)
sl@0: 
sl@0: #define MmxAddMul(x, a, y, b) do {                  \
sl@0:         x = _mm_mullo_pi16 (x, a);                  \
sl@0:         y = _mm_mullo_pi16 (y, b);                  \
sl@0:         x = _mm_srli_pi16(x, 1);                    \
sl@0:         y = _mm_srli_pi16(y, 1);                    \
sl@0:         x = _mm_adds_pu16 (x, y);                    \
sl@0:         x = _mm_adds_pu16 (x, _mm_srli_pi16 (x, 8)); \
sl@0:         x = _mm_adds_pu16 (x, mmx_4x0080);           \
sl@0:         x = _mm_srli_pi16 (x, 7);                   \
sl@0:     } while (0)
sl@0: