sl@0: /*
sl@0:  * LIBOIL - Library of Optimized Inner Loops
sl@0:  * Copyright (c) 2003,2004 David A. Schleef <ds@schleef.org>
sl@0:  * All rights reserved.
sl@0:  *
sl@0:  * Redistribution and use in source and binary forms, with or without
sl@0:  * modification, are permitted provided that the following conditions
sl@0:  * are met:
sl@0:  * 1. Redistributions of source code must retain the above copyright
sl@0:  *    notice, this list of conditions and the following disclaimer.
sl@0:  * 2. Redistributions in binary form must reproduce the above copyright
sl@0:  *    notice, this list of conditions and the following disclaimer in the
sl@0:  *    documentation and/or other materials provided with the distribution.
sl@0:  * 
sl@0:  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
sl@0:  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
sl@0:  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
sl@0:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT,
sl@0:  * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
sl@0:  * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
sl@0:  * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
sl@0:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
sl@0:  * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
sl@0:  * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
sl@0:  * POSSIBILITY OF SUCH DAMAGE.
sl@0:  */
sl@0: //Portions Copyright (c)  2008-2009 Nokia Corporation and/or its subsidiary(-ies). All rights reserved. 
sl@0: 
sl@0: #ifdef HAVE_CONFIG_H
sl@0: #include "config.h"
sl@0: #endif
sl@0: 
sl@0: #include <liboil/liboilfunction.h>
sl@0: #include "liboil/simdpack/simdpack.h"
sl@0: 
sl@0: #include <math.h>
sl@0: 
sl@0: 
sl@0: static void
sl@0: mult8x8_s16_mmx(int16_t *dest, int16_t *src1, int16_t *src2, int dstr, int sstr1,
sl@0:     int sstr2)
sl@0: {
sl@0: #if !defined(__WINSCW__) && !defined(__WINS__)      
sl@0:         asm volatile(
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0: 
sl@0:             "  add %3, %0 \n"
sl@0:             "  add %4, %1 \n"
sl@0:             "  add %5, %2 \n"
sl@0: 
sl@0:             "  movq 0(%1), %%mm0 \n"
sl@0:             "  pmullw 0(%2), %%mm0 \n"
sl@0:             "  movq %%mm0, 0(%0) \n"
sl@0:             "  movq 8(%1), %%mm1 \n"
sl@0:             "  pmullw 8(%2), %%mm1 \n"
sl@0:             "  movq %%mm1, 8(%0) \n"
sl@0:             "  emms\n "
sl@0: 
sl@0:             : "+r" (dest), "+r" (src1), "+r" (src2)
sl@0:             : "m" (dstr), "m" (sstr1), "m" (sstr2));
sl@0: #endif
sl@0: }
sl@0: 
sl@0: OIL_DEFINE_IMPL_FULL (mult8x8_s16_mmx, mult8x8_s16, OIL_IMPL_FLAG_MMX);
sl@0: 
sl@0: 
sl@0: 
sl@0: #ifdef	__SYMBIAN32__
sl@0:  
sl@0: OilFunctionImpl* __oil_function_impl_mult8x8_s16_mmx, mult8x8_s16() {
sl@0: 		return &_oil_function_impl_mult8x8_s16_mmx, mult8x8_s16;
sl@0: }
sl@0: #endif
sl@0: