/*********************************************************************/

/* Copyright 2009, 2010 The University of Texas at Austin.           */

/* All rights reserved.                                              */

/*                                                                   */

/* Redistribution and use in source and binary forms, with or        */

/* without modification, are permitted provided that the following   */

/* conditions are met:                                               */

/*                                                                   */

/*   1. Redistributions of source code must retain the above         */

/*      copyright notice, this list of conditions and the following  */

/*      disclaimer.                                                  */

/*                                                                   */

/*   2. Redistributions in binary form must reproduce the above      */

/*      copyright notice, this list of conditions and the following  */

/*      disclaimer in the documentation and/or other materials       */

/*      provided with the distribution.                              */

/*                                                                   */

/*    THIS  SOFTWARE IS PROVIDED  BY THE  UNIVERSITY OF  TEXAS AT    */

/*    AUSTIN  ``AS IS''  AND ANY  EXPRESS OR  IMPLIED WARRANTIES,    */

/*    INCLUDING, BUT  NOT LIMITED  TO, THE IMPLIED  WARRANTIES OF    */

/*    MERCHANTABILITY  AND FITNESS FOR  A PARTICULAR  PURPOSE ARE    */

/*    DISCLAIMED.  IN  NO EVENT SHALL THE UNIVERSITY  OF TEXAS AT    */

/*    AUSTIN OR CONTRIBUTORS BE  LIABLE FOR ANY DIRECT, INDIRECT,    */

/*    INCIDENTAL,  SPECIAL, EXEMPLARY,  OR  CONSEQUENTIAL DAMAGES    */

/*    (INCLUDING, BUT  NOT LIMITED TO,  PROCUREMENT OF SUBSTITUTE    */

/*    GOODS  OR  SERVICES; LOSS  OF  USE,  DATA,  OR PROFITS;  OR    */

/*    BUSINESS INTERRUPTION) HOWEVER CAUSED  AND ON ANY THEORY OF    */

/*    LIABILITY, WHETHER  IN CONTRACT, STRICT  LIABILITY, OR TORT    */

/*    (INCLUDING NEGLIGENCE OR OTHERWISE)  ARISING IN ANY WAY OUT    */

/*    OF  THE  USE OF  THIS  SOFTWARE,  EVEN  IF ADVISED  OF  THE    */

/*    POSSIBILITY OF SUCH DAMAGE.                                    */

/*                                                                   */

/* The views and conclusions contained in the software and           */

/* documentation are those of the authors and should not be          */

/* interpreted as representing official policies, either expressed   */

/* or implied, of The University of Texas at Austin.                 */

/*********************************************************************/

#define ASSEMBLER

#include "common.h"

#define M	ARG1	/* rdi */

#define X	ARG2	/* rsi */

#define INCX	ARG3	/* rdx */

#define I	%rax

#include "l1param.h"

	PROLOGUE

	PROFCODE

	SAVEREGISTERS

	pxor	%xmm0, %xmm0

	testq	M, M

	jle	.L999

	testq	INCX, INCX

	jle	.L999

	pxor	%xmm1, %xmm1

	pxor	%xmm2, %xmm2

	pxor	%xmm3, %xmm3

	pcmpeqb	%xmm15, %xmm15

	psrld	$1, %xmm15

	salq	$ZBASE_SHIFT, INCX

	cmpq	$2 * SIZE, INCX

	jne	.L100

	subq	$-32 * SIZE, X

	addq	M, M

	cmpq	$3, M

	jle	.L18

	testq	$4, X

	je	.L05

	movss	-32 * SIZE(X), %xmm0

	andps	%xmm15, %xmm0

	addq	$SIZE, X

	decq	M

	jle	.L998

	ALIGN_3

.L05:

	testq	$8, X

	je	.L10

#ifdef movsd

	xorps	%xmm1, %xmm1

#endif

	movsd	-32 * SIZE(X), %xmm1

	andps	%xmm15, %xmm1

	addq	$2 * SIZE, X

	subq	$2, M

	jle	.L998

	ALIGN_3

.L10:

	movq	M,  I

	sarq	$5, I

	jle	.L14

	movaps	-32 * SIZE(X), %xmm4

	movaps	-28 * SIZE(X), %xmm5

	movaps	-24 * SIZE(X), %xmm6

	movaps	-20 * SIZE(X), %xmm7

	movaps	-16 * SIZE(X), %xmm8

	movaps	-12 * SIZE(X), %xmm9

	movaps	 -8 * SIZE(X), %xmm10

	movaps	 -4 * SIZE(X), %xmm11

	decq	I

	jle	.L12

	ALIGN_3

.L11:

#ifdef PREFETCH

	PREFETCH (PREFETCHSIZE +  0) - PREOFFSET(X)

#endif

	andps	%xmm15, %xmm4

	addps	%xmm4,  %xmm0

	movaps	  0 * SIZE(X), %xmm4

	andps	%xmm15, %xmm5

	addps	%xmm5,  %xmm1

	movaps	  4 * SIZE(X), %xmm5

	andps	%xmm15, %xmm6

	addps	%xmm6,  %xmm2

	movaps	  8 * SIZE(X), %xmm6

	andps	%xmm15, %xmm7

	addps	%xmm7,  %xmm3

	movaps	 12 * SIZE(X), %xmm7

#if defined(PREFETCH) && !defined(FETCH128)

	PREFETCH (PREFETCHSIZE +  64) - PREOFFSET(X)

#endif

	andps	%xmm15, %xmm8

	addps	%xmm8,  %xmm0

	movaps	 16 * SIZE(X), %xmm8

	andps	%xmm15, %xmm9

	addps	%xmm9,  %xmm1

	movaps	 20 * SIZE(X), %xmm9

	andps	%xmm15, %xmm10

	addps	%xmm10, %xmm2

	movaps	 24 * SIZE(X), %xmm10

	andps	%xmm15, %xmm11

	addps	%xmm11, %xmm3

	movaps	 28 * SIZE(X), %xmm11

	subq	$-32 * SIZE, X

	decq	I

	jg	.L11

	ALIGN_3

.L12:

	andps	%xmm15, %xmm4

	addps	%xmm4,  %xmm0

	andps	%xmm15, %xmm5

	addps	%xmm5,  %xmm1

	andps	%xmm15, %xmm6

	addps	%xmm6,  %xmm2

	andps	%xmm15, %xmm7

	addps	%xmm7,  %xmm3

	andps	%xmm15, %xmm8

	addps	%xmm8,  %xmm0

	andps	%xmm15, %xmm9

	addps	%xmm9,  %xmm1

	andps	%xmm15, %xmm10

	addps	%xmm10, %xmm2

	andps	%xmm15, %xmm11

	addps	%xmm11, %xmm3

	addq	$32 * SIZE, X

	ALIGN_3

.L14:

	testq	$31,  M

	jle	.L998

.L15:

	testq	$16, M

	je	.L16

	movaps	-32 * SIZE(X), %xmm4

	andps	%xmm15, %xmm4

	addps	%xmm4, %xmm0

	movaps	-28 * SIZE(X), %xmm5

	andps	%xmm15, %xmm5

	addps	%xmm5, %xmm1

	movaps	-24 * SIZE(X), %xmm4

	andps	%xmm15, %xmm4

	addps	%xmm4, %xmm0

	movaps	-20 * SIZE(X), %xmm5

	andps	%xmm15, %xmm5

	addps	%xmm5, %xmm1

	addq	$16 * SIZE, X

	ALIGN_3

.L16:

	testq	$8, M

	je	.L17

	movaps	-32 * SIZE(X), %xmm4

	andps	%xmm15, %xmm4

	addps	%xmm4, %xmm0

	movaps	-28 * SIZE(X), %xmm5

	andps	%xmm15, %xmm5

	addps	%xmm5, %xmm1

	addq	$8 * SIZE, X

	ALIGN_3

.L17:

	testq	$4, M

	je	.L18

	movaps	-32 * SIZE(X), %xmm6

	andps	%xmm15, %xmm6

	addps	%xmm6, %xmm2

	addq	$4 * SIZE, X

	ALIGN_3

.L18:

	testq	$2, M

	je	.L19

#ifdef movsd

	xorps	%xmm7, %xmm7

#endif

	movsd	-32 * SIZE(X), %xmm7

	andps	%xmm15, %xmm7

	addps	%xmm7, %xmm3

	addq	$2 * SIZE, X

	ALIGN_3

.L19:

	testq	$1, M

	je	.L998

	movss	-32 * SIZE(X), %xmm6

	andps	%xmm15, %xmm6

	addps	%xmm6, %xmm2

	jmp	.L998

	ALIGN_4

.L100:

	movq	M,  I

	sarq	$2, I

	jle	.L105

	ALIGN_4

.L101:

	movsd	(X), %xmm4

	addq	INCX, X

	movhps	(X), %xmm4

	addq	INCX, X

	andps	%xmm15, %xmm4

	addps	%xmm4,  %xmm0

	movsd	(X), %xmm5

	addq	INCX, X

	movhps	(X), %xmm5

	addq	INCX, X

	andps	%xmm15, %xmm5

	addps	%xmm5,  %xmm1

	decq	I

	jg	.L101

	ALIGN_4

.L105:

#ifdef movsd

	xorps	%xmm4, %xmm4

#endif

	andq	$3,  M

	jle	.L998

	ALIGN_4

.L106:

	movsd	(X), %xmm4

	andps	%xmm15, %xmm4

	addps	%xmm4, %xmm0

	addq	INCX, X

	decq	M

	jg	.L106

	ALIGN_4

.L998:

	addps	%xmm1, %xmm0

	addps	%xmm3, %xmm2

	addps	%xmm2, %xmm0

#ifndef HAVE_SSE3

	movhlps	%xmm0, %xmm1

	addps	%xmm1, %xmm0

	movaps	%xmm0, %xmm1

	shufps  $1, %xmm0, %xmm0

	addss	 %xmm1, %xmm0

#else

	haddps	%xmm0, %xmm0

	haddps	%xmm0, %xmm0

#endif	

	ALIGN_4

.L999:

	RESTOREREGISTERS

	ret

	EPILOGUE