/*********************************************************************/
/* 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 16
#define ARGS 0
#define STACK_M 4 + STACK + ARGS(%esp)
#define STACK_X 24 + STACK + ARGS(%esp)
#define STACK_INCX 28 + STACK + ARGS(%esp)
#define STACK_Y 32 + STACK + ARGS(%esp)
#define STACK_INCY 36 + STACK + ARGS(%esp)
#define M %edx
#define X %esi
#define Y %edi
#define INCX %ebx
#define INCY %ecx
#include "l1param.h"
PROLOGUE
PROFCODE
pushl %ebp
pushl %edi
pushl %esi
pushl %ebx
movl STACK_M, M
movl STACK_X, X
movl STACK_Y, Y
movl STACK_INCX, INCX
movl STACK_INCY, INCY
leal (, INCX, SIZE), INCX
leal (, INCY, SIZE), INCY
cmpl $SIZE, INCX
jne .L40
cmpl $SIZE, INCY
jne .L40
testl $SIZE, Y
je .L10
movsd 0 * SIZE(X), %xmm0
movsd 0 * SIZE(Y), %xmm1
movsd %xmm1, 0 * SIZE(X)
movsd %xmm0, 0 * SIZE(Y)
addl $1 * SIZE, X
addl $1 * SIZE, Y
decl M
jle .L19
ALIGN_4
.L10:
subl $-16 * SIZE, X
subl $-16 * SIZE, Y
testl $SIZE, X
jne .L20
movl M, %eax
sarl $4, %eax
jle .L13
ALIGN_3
.L11:
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 0) - PREOFFSET(X)
#endif
movaps -16 * SIZE(X), %xmm0
movaps -16 * SIZE(Y), %xmm1
movaps %xmm0, -16 * SIZE(Y)
movaps %xmm1, -16 * SIZE(X)
movaps -14 * SIZE(X), %xmm0
movaps -14 * SIZE(Y), %xmm1
movaps %xmm0, -14 * SIZE(Y)
movaps %xmm1, -14 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 0) - PREOFFSET(Y)
#endif
movaps -12 * SIZE(X), %xmm0
movaps -12 * SIZE(Y), %xmm1
movaps %xmm0, -12 * SIZE(Y)
movaps %xmm1, -12 * SIZE(X)
movaps -10 * SIZE(X), %xmm0
movaps -10 * SIZE(Y), %xmm1
movaps %xmm0, -10 * SIZE(Y)
movaps %xmm1, -10 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 64) - PREOFFSET(X)
#endif
movaps -8 * SIZE(X), %xmm0
movaps -8 * SIZE(Y), %xmm1
movaps %xmm0, -8 * SIZE(Y)
movaps %xmm1, -8 * SIZE(X)
movaps -6 * SIZE(X), %xmm0
movaps -6 * SIZE(Y), %xmm1
movaps %xmm0, -6 * SIZE(Y)
movaps %xmm1, -6 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 64) - PREOFFSET(Y)
#endif
movaps -4 * SIZE(X), %xmm0
movaps -4 * SIZE(Y), %xmm1
movaps %xmm0, -4 * SIZE(Y)
movaps %xmm1, -4 * SIZE(X)
movaps -2 * SIZE(X), %xmm0
movaps -2 * SIZE(Y), %xmm1
movaps %xmm0, -2 * SIZE(Y)
movaps %xmm1, -2 * SIZE(X)
subl $-16 * SIZE, Y
subl $-16 * SIZE, X
decl %eax
jg .L11
ALIGN_3
.L13:
testl $8, M
jle .L14
movaps -16 * SIZE(X), %xmm0
movaps -16 * SIZE(Y), %xmm1
movaps %xmm0, -16 * SIZE(Y)
movaps %xmm1, -16 * SIZE(X)
movaps -14 * SIZE(X), %xmm0
movaps -14 * SIZE(Y), %xmm1
movaps %xmm0, -14 * SIZE(Y)
movaps %xmm1, -14 * SIZE(X)
movaps -12 * SIZE(X), %xmm0
movaps -12 * SIZE(Y), %xmm1
movaps %xmm0, -12 * SIZE(Y)
movaps %xmm1, -12 * SIZE(X)
movaps -10 * SIZE(X), %xmm0
movaps -10 * SIZE(Y), %xmm1
movaps %xmm0, -10 * SIZE(Y)
movaps %xmm1, -10 * SIZE(X)
addl $8 * SIZE, X
addl $8 * SIZE, Y
ALIGN_3
.L14:
testl $4, M
jle .L15
movaps -16 * SIZE(X), %xmm0
movaps -16 * SIZE(Y), %xmm1
movaps %xmm0, -16 * SIZE(Y)
movaps %xmm1, -16 * SIZE(X)
movaps -14 * SIZE(X), %xmm0
movaps -14 * SIZE(Y), %xmm1
movaps %xmm0, -14 * SIZE(Y)
movaps %xmm1, -14 * SIZE(X)
addl $4 * SIZE, X
addl $4 * SIZE, Y
ALIGN_3
.L15:
testl $2, M
jle .L16
movaps -16 * SIZE(X), %xmm0
movaps -16 * SIZE(Y), %xmm1
movaps %xmm0, -16 * SIZE(Y)
movaps %xmm1, -16 * SIZE(X)
addl $2 * SIZE, X
addl $2 * SIZE, Y
ALIGN_3
.L16:
testl $1, M
jle .L19
movsd -16 * SIZE(X), %xmm0
movsd -16 * SIZE(Y), %xmm1
movlps %xmm1, -16 * SIZE(X)
movlps %xmm0, -16 * SIZE(Y)
ALIGN_3
.L19:
popl %ebx
popl %esi
popl %edi
popl %ebp
ret
ALIGN_3
.L20:
movhps -16 * SIZE(X), %xmm0
movaps -16 * SIZE(Y), %xmm1
movlps %xmm1, -16 * SIZE(X)
decl M
jle .L29
movl M, %eax
sarl $4, %eax
jle .L23
ALIGN_4
.L21:
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 0) - PREOFFSET(X)
#endif
movaps -15 * SIZE(X), %xmm2
movaps -14 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -16 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -15 * SIZE(X)
movaps -13 * SIZE(X), %xmm0
movaps -12 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -14 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -13 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 0) - PREOFFSET(Y)
#endif
movaps -11 * SIZE(X), %xmm2
movaps -10 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -12 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -11 * SIZE(X)
movaps -9 * SIZE(X), %xmm0
movaps -8 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -10 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -9 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 64) - PREOFFSET(X)
#endif
movaps -7 * SIZE(X), %xmm2
movaps -6 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -8 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -7 * SIZE(X)
movaps -5 * SIZE(X), %xmm0
movaps -4 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -6 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -5 * SIZE(X)
#ifdef PREFETCHW
PREFETCHW (PREFETCHSIZE + 64) - PREOFFSET(Y)
#endif
movaps -3 * SIZE(X), %xmm2
movaps -2 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -4 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -3 * SIZE(X)
movaps -1 * SIZE(X), %xmm0
movaps 0 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -2 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -1 * SIZE(X)
subl $-16 * SIZE, X
subl $-16 * SIZE, Y
decl %eax
jg .L21
ALIGN_3
.L23:
testl $8, M
jle .L24
movaps -15 * SIZE(X), %xmm2
movaps -14 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -16 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -15 * SIZE(X)
movaps -13 * SIZE(X), %xmm0
movaps -12 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -14 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -13 * SIZE(X)
movaps -11 * SIZE(X), %xmm2
movaps -10 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -12 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -11 * SIZE(X)
movaps -9 * SIZE(X), %xmm0
movaps -8 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -10 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -9 * SIZE(X)
addl $8 * SIZE, X
addl $8 * SIZE, Y
ALIGN_3
.L24:
testl $4, M
jle .L25
movaps -15 * SIZE(X), %xmm2
movaps -14 * SIZE(Y), %xmm3
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -16 * SIZE(Y)
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -15 * SIZE(X)
movaps -13 * SIZE(X), %xmm0
movaps -12 * SIZE(Y), %xmm1
SHUFPD_1 %xmm0, %xmm2
movaps %xmm2, -14 * SIZE(Y)
SHUFPD_1 %xmm1, %xmm3
movaps %xmm3, -13 * SIZE(X)
addl $4 * SIZE, X
addl $4 * SIZE, Y
ALIGN_3
.L25:
testl $2, M
jle .L26
movaps -15 * SIZE(X), %xmm2
movaps -14 * SIZE(Y), %xmm3
SHUFPD_1 %xmm3, %xmm1
movaps %xmm1, -15 * SIZE(X)
SHUFPD_1 %xmm2, %xmm0
movaps %xmm0, -16 * SIZE(Y)
movaps %xmm2, %xmm0
movaps %xmm3, %xmm1
addl $2 * SIZE, X
addl $2 * SIZE, Y
ALIGN_3
.L26:
testl $1, M
jle .L29
movhps %xmm0, -16 * SIZE(Y)
movhps -15 * SIZE(X), %xmm0
movhps %xmm1, -15 * SIZE(X)
addl $SIZE, X
addl $SIZE, Y
ALIGN_3
.L29:
movhps %xmm0, -16 * SIZE(Y)
popl %ebx
popl %esi
popl %edi
popl %ebp
ret
ALIGN_3
.L40:
movl M, %eax
sarl $3, %eax
jle .L45
ALIGN_3
.L41:
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
addl INCX, X
movsd %xmm0, (Y)
addl INCY, Y
decl %eax
jg .L41
ALIGN_3
.L45:
movl M, %eax
andl $7, %eax
jle .L47
ALIGN_3
.L46:
movsd (X), %xmm0
movsd (Y), %xmm1
movsd %xmm1, (X)
movsd %xmm0, (Y)
addl INCX, X
addl INCY, Y
decl %eax
jg .L46
ALIGN_3
.L47:
popl %ebx
popl %esi
popl %edi
popl %ebp
ret
EPILOGUE