/*********************************************************************/
/* 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 STACK 8
#define ARGS 0
#define STACK_M 4 + STACK + ARGS(%esp)
#define STACK_X 8 + STACK + ARGS(%esp)
#define STACK_INCX 12 + STACK + ARGS(%esp)
#define I %eax
#define M %ecx
#define X %esi
#define INCX %ebx
#define xmm8 xmm4
#define xmm9 xmm5
#define xmm10 xmm6
#define xmm11 xmm7
#include "l1param.h"
PROLOGUE
PROFCODE
pushl %esi
pushl %ebx
movl STACK_M, M
movl STACK_X, X
movl STACK_INCX, INCX
xorps %xmm0, %xmm0
xorps %xmm1, %xmm1
testl M, M
jle .L999
testl INCX, INCX
jle .L999
pcmpeqb %xmm3, %xmm3
psrlq $1, %xmm3
sall $ZBASE_SHIFT, INCX
cmpl $2 * SIZE, INCX
jne .L40
subl $-16 * SIZE, X
addl M, M
testl $SIZE, X
je .L05
movsd -16 * SIZE(X), %xmm0
addl $SIZE, X
andps %xmm3, %xmm0
subl $1, M
jle .L999
ALIGN_3
.L05:
movl M, I
sarl $4, I
jle .L20
movaps -16 * SIZE(X), %xmm4
movaps -14 * SIZE(X), %xmm5
movaps -12 * SIZE(X), %xmm6
movaps -10 * SIZE(X), %xmm7
decl I
jle .L11
ALIGN_4
.L10:
#ifdef PREFETCH
PREFETCH (PREFETCHSIZE + 0) - PREOFFSET(X)
#endif
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
movaps -8 * SIZE(X), %xmm4
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
movaps -6 * SIZE(X), %xmm5
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
movaps -4 * SIZE(X), %xmm6
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
movaps -2 * SIZE(X), %xmm7
#if defined(PREFETCH) && !defined(FETCH128)
PREFETCH (PREFETCHSIZE + 64) - PREOFFSET(X)
#endif
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
movaps 0 * SIZE(X), %xmm4
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
movaps 2 * SIZE(X), %xmm5
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
movaps 4 * SIZE(X), %xmm6
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
movaps 6 * SIZE(X), %xmm7
subl $-16 * SIZE, X
decl I
jg .L10
ALIGN_4
.L11:
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
movaps -8 * SIZE(X), %xmm4
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
movaps -6 * SIZE(X), %xmm5
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
movaps -4 * SIZE(X), %xmm6
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
movaps -2 * SIZE(X), %xmm7
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
subl $-16 * SIZE, X
ALIGN_3
.L20:
andl $15, M
jle .L999
testl $8, M
je .L21
movaps -16 * SIZE(X), %xmm4
movaps -14 * SIZE(X), %xmm5
movaps -12 * SIZE(X), %xmm6
movaps -10 * SIZE(X), %xmm7
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
addl $8 * SIZE, X
ALIGN_3
.L21:
testl $4, M
je .L22
movaps -16 * SIZE(X), %xmm4
movaps -14 * SIZE(X), %xmm5
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
addl $4 * SIZE, X
ALIGN_3
.L22:
testl $2, M
je .L23
movaps -16 * SIZE(X), %xmm4
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
addl $2 * SIZE, X
.L23:
testl $1, M
je .L999
#ifdef movsd
xorps %xmm4, %xmm4
#endif
movsd -16 * SIZE(X), %xmm4
andps %xmm3, %xmm4
addsd %xmm4, %xmm0
jmp .L999
ALIGN_3
.L40:
movl M, I
sarl $2, I
jle .L60
ALIGN_4
.L50:
movsd 0 * SIZE(X), %xmm4
movhps 1 * SIZE(X), %xmm4
addl INCX, X
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
movsd 0 * SIZE(X), %xmm5
movhps 1 * SIZE(X), %xmm5
addl INCX, X
andps %xmm3, %xmm5
addpd %xmm5, %xmm1
movsd 0 * SIZE(X), %xmm6
movhps 1 * SIZE(X), %xmm6
addl INCX, X
andps %xmm3, %xmm6
addpd %xmm6, %xmm0
movsd 0 * SIZE(X), %xmm7
movhps 1 * SIZE(X), %xmm7
addl INCX, X
andps %xmm3, %xmm7
addpd %xmm7, %xmm1
decl I
jg .L50
ALIGN_4
.L60:
andl $3, M
jle .L999
ALIGN_4
.L61:
movsd 0 * SIZE(X), %xmm4
movhps 1 * SIZE(X), %xmm4
andps %xmm3, %xmm4
addpd %xmm4, %xmm0
addl INCX, X
decl M
jg .L61
ALIGN_4
.L999:
addpd %xmm1, %xmm0
#ifndef HAVE_SSE3
movaps %xmm0, %xmm1
unpckhpd %xmm0, %xmm0
addsd %xmm1, %xmm0
#else
haddpd %xmm0, %xmm0
#endif
movsd %xmm0, STACK_M
fldl STACK_M
popl %ebx
popl %esi
ret
EPILOGUE