⭐ 欢迎来到虫虫下载站! | 📦 资源下载 📁 资源专辑 ℹ️ 关于我们
⭐ 虫虫下载站

📄 zrot_sse2.s

📁 Optimized GotoBLAS libraries
💻 S
📖 第 1 页 / 共 2 页
字号:
/*********************************************************************//*                                                                   *//*             Optimized BLAS libraries                              *//*                     By Kazushige Goto <kgoto@tacc.utexas.edu>     *//*                                                                   *//* Copyright (c) The University of Texas, 2005. All rights reserved. *//* UNIVERSITY EXPRESSLY DISCLAIMS ANY AND ALL WARRANTIES CONCERNING  *//* THIS SOFTWARE AND DOCUMENTATION, INCLUDING ANY WARRANTIES OF      *//* MERCHANTABILITY, FITNESS FOR ANY PARTICULAR PURPOSE,              *//* NON-INFRINGEMENT AND WARRANTIES OF PERFORMANCE, AND ANY WARRANTY  *//* THAT MIGHT OTHERWISE ARISE FROM COURSE OF DEALING OR USAGE OF     *//* TRADE. NO WARRANTY IS EITHER EXPRESS OR IMPLIED WITH RESPECT TO   *//* THE USE OF THE SOFTWARE OR DOCUMENTATION.                         *//* Under no circumstances shall University be liable for incidental, *//* special, indirect, direct or consequential damages or loss of     *//* profits, interruption of business, or related expenses which may  *//* arise from use of Software or Documentation, including but not    *//* limited to those resulting from defects in Software and/or        *//* Documentation, or loss or inaccuracy of data of any kind.         *//*********************************************************************/#define ASSEMBLER#include "common.h"#define N	ARG1	/* rdi */#define X	ARG2	/* rsi */#define INCX	ARG3	/* rdx */#define Y	ARG4	/* rcx */#ifndef WINDOWS_ABI#define INCY	ARG5	/* r8  */#else#define INCY	%r10#endif#undef MOVSD#ifdef PENTIUM4#define PREFETCH	prefetcht0#define PREFETCH_SIZE	144#define MOVSD		movsd#endif#if defined(OPTERON) || defined(BARCELONA)#define PREFETCH	prefetchw#define PREFETCH_SIZE	144#ifdef OPTERON#define MOVSD		movlpd#else#define MOVSD		movsd#endif#endif#define C	%xmm14#define S	%xmm15	PROLOGUE	PROFCODE#ifdef WINDOWS_ABI	movq	40(%rsp), INCY	movsd	48(%rsp), %xmm0	movsd	56(%rsp), %xmm1#endif	SAVEREGISTERS	salq	$ZBASE_SHIFT, INCX	salq	$ZBASE_SHIFT, INCY	pshufd	$0x44, %xmm0, C	pshufd	$0x44, %xmm1, S	cmpq	$0, N	jle	.L999	cmpq	$2 * SIZE, INCX	jne	.L50	cmpq	$2 * SIZE, INCY	jne	.L50.L10:	testq	$SIZE, X	jne	.L30	testq	$SIZE, Y	jne	.L20	movq	N,  %rax	sarq	$3, %rax	jle	.L14	movapd	0 * SIZE(Y), %xmm1	movapd	2 * SIZE(Y), %xmm3	movapd	4 * SIZE(Y), %xmm9	movapd	6 * SIZE(Y), %xmm11	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(X), %xmm2	movapd	4 * SIZE(X), %xmm8	movapd	6 * SIZE(X), %xmm10		decq	%rax	jle	.L12	ALIGN_3.L11:#ifdef PENTIUM4	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif	movapd	%xmm1, %xmm4	mulpd	S, %xmm1	movapd	%xmm3, %xmm6	mulpd	S, %xmm3	movapd	%xmm0, %xmm5	mulpd	C, %xmm0	movapd	%xmm2, %xmm7	mulpd	C, %xmm2	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	movapd	 8 * SIZE(Y), %xmm1	addpd	%xmm3, %xmm2	movapd	10 * SIZE(Y), %xmm3	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(Y)#endif	movapd	%xmm0, 0 * SIZE(X)	movapd	 8 * SIZE(X), %xmm0	movapd	%xmm2, 2 * SIZE(X)	movapd	10 * SIZE(X), %xmm2	movapd	%xmm4, 0 * SIZE(Y)	movapd	%xmm6, 2 * SIZE(Y)	movapd	%xmm9, %xmm4	mulpd	S, %xmm9	movapd	%xmm8, %xmm5	mulpd	C, %xmm8	movapd	%xmm11, %xmm6	mulpd	S, %xmm11	movapd	%xmm10, %xmm7	mulpd	C, %xmm10	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm9,  %xmm8	movapd	12 * SIZE(Y), %xmm9	addpd	%xmm11, %xmm10	movapd	14 * SIZE(Y), %xmm11	subpd	%xmm5,  %xmm4	subpd	%xmm7,  %xmm6	movapd	%xmm8, 4 * SIZE(X)	movapd	12 * SIZE(X), %xmm8	movapd	%xmm10,6 * SIZE(X)	movapd	14 * SIZE(X), %xmm10	movapd	%xmm4, 4 * SIZE(Y)	movapd	%xmm6, 6 * SIZE(Y)#ifdef PENTIUM4	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(Y)#endif#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 8) * SIZE(X)#endif	movapd	%xmm1, %xmm4	mulpd	S, %xmm1	movapd	%xmm3, %xmm6	mulpd	S, %xmm3	movapd	%xmm0, %xmm5	mulpd	C, %xmm0	movapd	%xmm2, %xmm7	mulpd	C, %xmm2	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	movapd	16 * SIZE(Y), %xmm1	addpd	%xmm3, %xmm2	movapd	18 * SIZE(Y), %xmm3	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0,  8 * SIZE(X)	movapd	16 * SIZE(X), %xmm0	movapd	%xmm2, 10 * SIZE(X)	movapd	18 * SIZE(X), %xmm2	movapd	%xmm4,  8 * SIZE(Y)	movapd	%xmm6, 10 * SIZE(Y)#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 8) * SIZE(Y)#endif	movapd	%xmm9, %xmm4	mulpd	S, %xmm9	movapd	%xmm8, %xmm5	mulpd	C, %xmm8	movapd	%xmm11, %xmm6	mulpd	S, %xmm11	movapd	%xmm10, %xmm7	mulpd	C, %xmm10	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm9,  %xmm8	movapd	20 * SIZE(Y), %xmm9	addpd	%xmm11, %xmm10	movapd	22 * SIZE(Y), %xmm11	subpd	%xmm5,  %xmm4	subpd	%xmm7,  %xmm6	movapd	%xmm8,  12 * SIZE(X)	movapd	20 * SIZE(X), %xmm8	movapd	%xmm10, 14 * SIZE(X)	movapd	22 * SIZE(X), %xmm10	movapd	%xmm4,  12 * SIZE(Y)	movapd	%xmm6,  14 * SIZE(Y)	addq	$16 * SIZE, X	addq	$16 * SIZE, Y	decq	%rax	jg	.L11	ALIGN_3.L12:	movapd	%xmm1, %xmm4	mulpd	S, %xmm1	movapd	%xmm3, %xmm6	mulpd	S, %xmm3	movapd	%xmm0, %xmm5	mulpd	C, %xmm0	movapd	%xmm2, %xmm7	mulpd	C, %xmm2	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	movapd	 8 * SIZE(Y), %xmm1	addpd	%xmm3, %xmm2	movapd	10 * SIZE(Y), %xmm3	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	 8 * SIZE(X), %xmm0	movapd	%xmm2, 2 * SIZE(X)	movapd	10 * SIZE(X), %xmm2	movapd	%xmm4, 0 * SIZE(Y)	movapd	%xmm6, 2 * SIZE(Y)	movapd	%xmm9, %xmm4	mulpd	S, %xmm9	movapd	%xmm8, %xmm5	mulpd	C, %xmm8	movapd	%xmm11, %xmm6	mulpd	S, %xmm11	movapd	%xmm10, %xmm7	mulpd	C, %xmm10	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm9,  %xmm8	movapd	12 * SIZE(Y), %xmm9	addpd	%xmm11, %xmm10	movapd	14 * SIZE(Y), %xmm11	subpd	%xmm5,  %xmm4	subpd	%xmm7,  %xmm6	movapd	%xmm8, 4 * SIZE(X)	movapd	12 * SIZE(X), %xmm8	movapd	%xmm10,6 * SIZE(X)	movapd	14 * SIZE(X), %xmm10	movapd	%xmm4, 4 * SIZE(Y)	movapd	%xmm6, 6 * SIZE(Y)	movapd	%xmm1, %xmm4	mulpd	S, %xmm1	movapd	%xmm3, %xmm6	mulpd	S, %xmm3	movapd	%xmm0, %xmm5	mulpd	C, %xmm0	movapd	%xmm2, %xmm7	mulpd	C, %xmm2	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0,  8 * SIZE(X)	movapd	%xmm2, 10 * SIZE(X)	movapd	%xmm4,  8 * SIZE(Y)	movapd	%xmm6, 10 * SIZE(Y)	movapd	%xmm9, %xmm4	mulpd	S, %xmm9	movapd	%xmm8, %xmm5	mulpd	C, %xmm8	movapd	%xmm11, %xmm6	mulpd	S, %xmm11	movapd	%xmm10, %xmm7	mulpd	C, %xmm10	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm9,  %xmm8	addpd	%xmm11, %xmm10	subpd	%xmm5,  %xmm4	subpd	%xmm7,  %xmm6	movapd	%xmm8,  12 * SIZE(X)	movapd	%xmm10, 14 * SIZE(X)	movapd	%xmm4,  12 * SIZE(Y)	movapd	%xmm6,  14 * SIZE(Y)	addq	$16 * SIZE, X	addq	$16 * SIZE, Y	ALIGN_3.L14:	testq	$7, N	jle	.L999	testq	$4, N	jle	.L15	movapd	0 * SIZE(Y), %xmm1	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(Y), %xmm3	movapd	2 * SIZE(X), %xmm2	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 2 * SIZE(X)	movapd	%xmm4, 0 * SIZE(Y)	movapd	%xmm6, 2 * SIZE(Y)	movapd	4 * SIZE(Y), %xmm1	movapd	4 * SIZE(X), %xmm0	movapd	6 * SIZE(Y), %xmm3	movapd	6 * SIZE(X), %xmm2	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 4 * SIZE(X)	movapd	%xmm2, 6 * SIZE(X)	movapd	%xmm4, 4 * SIZE(Y)	movapd	%xmm6, 6 * SIZE(Y)	addq	$8 * SIZE, X	addq	$8 * SIZE, Y	ALIGN_3.L15:	testq	$2, N	jle	.L16	movapd	0 * SIZE(Y), %xmm1	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(Y), %xmm3	movapd	2 * SIZE(X), %xmm2	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 2 * SIZE(X)	movapd	%xmm4, 0 * SIZE(Y)	movapd	%xmm6, 2 * SIZE(Y)	addq	$4 * SIZE, X	addq	$4 * SIZE, Y	ALIGN_3.L16:	testq	$1, N	jle	.L999	movapd	0 * SIZE(Y), %xmm1	movapd	0 * SIZE(X), %xmm0	movapd	%xmm1, %xmm2	movapd	%xmm0, %xmm3	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	addpd	%xmm1, %xmm0	subpd	%xmm3, %xmm2	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 0 * SIZE(Y)	jmp	.L999	ALIGN_3.L20:	movapd	-1 * SIZE(Y), %xmm1	movq	N,  %rax	sarq	$3, %rax	jle	.L24	ALIGN_3.L21:#ifdef PENTIUM4	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif	movapd	1 * SIZE(Y), %xmm3	movapd	3 * SIZE(Y), %xmm8	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(X), %xmm2	SHUFPD_1 %xmm3, %xmm1	SHUFPD_1 %xmm8, %xmm3	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 2 * SIZE(X)	movlpd	%xmm4, 0 * SIZE(Y)	movhpd	%xmm4, 1 * SIZE(Y)	movlpd	%xmm6, 2 * SIZE(Y)	movhpd	%xmm6, 3 * SIZE(Y)#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(Y)#endif	movapd	5 * SIZE(Y), %xmm9	movapd	7 * SIZE(Y), %xmm1	movapd	4 * SIZE(X), %xmm0	movapd	6 * SIZE(X), %xmm2	SHUFPD_1 %xmm9, %xmm8	SHUFPD_1 %xmm1, %xmm9	movapd	%xmm8, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm9, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm8	mulpd	C, %xmm2	mulpd	S, %xmm9	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm8, %xmm0	addpd	%xmm9, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 4 * SIZE(X)	movapd	%xmm2, 6 * SIZE(X)	movlpd	%xmm4, 4 * SIZE(Y)	movhpd	%xmm4, 5 * SIZE(Y)	movlpd	%xmm6, 6 * SIZE(Y)	movhpd	%xmm6, 7 * SIZE(Y)#ifdef PENTIUM4	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(Y)#endif#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 8) * SIZE(X)#endif	movapd	 9 * SIZE(Y), %xmm3	movapd	11 * SIZE(Y), %xmm8	movapd	 8 * SIZE(X), %xmm0	movapd	10 * SIZE(X), %xmm2	SHUFPD_1 %xmm3, %xmm1	SHUFPD_1 %xmm8, %xmm3	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0,  8 * SIZE(X)	movapd	%xmm2, 10 * SIZE(X)	movlpd	%xmm4,  8 * SIZE(Y)	movhpd	%xmm4,  9 * SIZE(Y)	movlpd	%xmm6, 10 * SIZE(Y)	movhpd	%xmm6, 11 * SIZE(Y)#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 8) * SIZE(Y)#endif	movapd	13 * SIZE(Y), %xmm9	movapd	15 * SIZE(Y), %xmm1	movapd	12 * SIZE(X), %xmm0	movapd	14 * SIZE(X), %xmm2	SHUFPD_1 %xmm9, %xmm8	SHUFPD_1 %xmm1, %xmm9	movapd	%xmm8, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm9, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm8	mulpd	C, %xmm2	mulpd	S, %xmm9	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm8, %xmm0	addpd	%xmm9, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 12 * SIZE(X)	movapd	%xmm2, 14 * SIZE(X)	movlpd	%xmm4, 12 * SIZE(Y)	movhpd	%xmm4, 13 * SIZE(Y)	movlpd	%xmm6, 14 * SIZE(Y)	movhpd	%xmm6, 15 * SIZE(Y)	addq	$16 * SIZE, X	addq	$16 * SIZE, Y	decq	%rax	jg	.L21	ALIGN_3.L24:	testq	$7, N	jle	.L999	testq	$4, N	jle	.L25	movapd	1 * SIZE(Y), %xmm3	movapd	3 * SIZE(Y), %xmm8	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(X), %xmm2	SHUFPD_1 %xmm3, %xmm1	SHUFPD_1 %xmm8, %xmm3	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 2 * SIZE(X)	movlpd	%xmm4, 0 * SIZE(Y)	movhpd	%xmm4, 1 * SIZE(Y)	movlpd	%xmm6, 2 * SIZE(Y)	movhpd	%xmm6, 3 * SIZE(Y)	movapd	5 * SIZE(Y), %xmm9	movapd	7 * SIZE(Y), %xmm1	movapd	4 * SIZE(X), %xmm0	movapd	6 * SIZE(X), %xmm2	SHUFPD_1 %xmm9, %xmm8	SHUFPD_1 %xmm1, %xmm9	movapd	%xmm8, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm9, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm8	mulpd	C, %xmm2	mulpd	S, %xmm9	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm8, %xmm0	addpd	%xmm9, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 4 * SIZE(X)	movapd	%xmm2, 6 * SIZE(X)	movlpd	%xmm4, 4 * SIZE(Y)	movhpd	%xmm4, 5 * SIZE(Y)	movlpd	%xmm6, 6 * SIZE(Y)	movhpd	%xmm6, 7 * SIZE(Y)	addq	$8 * SIZE, X	addq	$8 * SIZE, Y	ALIGN_3.L25:	testq	$2, N	jle	.L26	movapd	1 * SIZE(Y), %xmm3	movapd	3 * SIZE(Y), %xmm8	movapd	0 * SIZE(X), %xmm0	movapd	2 * SIZE(X), %xmm2	SHUFPD_1 %xmm3, %xmm1	SHUFPD_1 %xmm8, %xmm3	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movapd	%xmm0, 0 * SIZE(X)	movapd	%xmm2, 2 * SIZE(X)	movlpd	%xmm4, 0 * SIZE(Y)	movhpd	%xmm4, 1 * SIZE(Y)	movlpd	%xmm6, 2 * SIZE(Y)	movhpd	%xmm6, 3 * SIZE(Y)	movapd	%xmm8, %xmm1	addq	$4 * SIZE, X	addq	$4 * SIZE, Y	ALIGN_3.L26:	testq	$1, N	jle	.L999	movapd	1 * SIZE(Y), %xmm4	movapd	0 * SIZE(X), %xmm0	SHUFPD_1 %xmm4, %xmm1	movapd	%xmm1, %xmm2	movapd	%xmm0, %xmm3	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	addpd	%xmm1, %xmm0	subpd	%xmm3, %xmm2	movapd	%xmm0, 0 * SIZE(X)	movlpd	%xmm2, 0 * SIZE(Y)	movhpd	%xmm2, 1 * SIZE(Y)	jmp	.L999	ALIGN_3.L30:	testq	$SIZE, Y	jne	.L40	movapd	-1 * SIZE(X), %xmm0	movq	N,  %rax	sarq	$3, %rax	jle	.L34	ALIGN_3.L31:#ifdef PENTIUM4	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(X)#endif	movapd	1 * SIZE(X), %xmm2	movapd	3 * SIZE(X), %xmm8	movapd	0 * SIZE(Y), %xmm1	movapd	2 * SIZE(Y), %xmm3	SHUFPD_1 %xmm2, %xmm0	SHUFPD_1 %xmm8, %xmm2	movapd	%xmm1, %xmm4	movapd	%xmm0, %xmm5	movapd	%xmm3, %xmm6	movapd	%xmm2, %xmm7	mulpd	C, %xmm0	mulpd	S, %xmm1	mulpd	C, %xmm2	mulpd	S, %xmm3	mulpd	C, %xmm4	mulpd	S, %xmm5	mulpd	C, %xmm6	mulpd	S, %xmm7	addpd	%xmm1, %xmm0	addpd	%xmm3, %xmm2	subpd	%xmm5, %xmm4	subpd	%xmm7, %xmm6	movlpd	%xmm0, 0 * SIZE(X)	movhpd	%xmm0, 1 * SIZE(X)	movlpd	%xmm2, 2 * SIZE(X)	movhpd	%xmm2, 3 * SIZE(X)	movapd	%xmm4, 0 * SIZE(Y)	movapd	%xmm6, 2 * SIZE(Y)#if defined(OPTERON) || defined(BARCELONA)	PREFETCH	(PREFETCH_SIZE + 0) * SIZE(Y)#endif

⌨️ 快捷键说明

复制代码 Ctrl + C
搜索代码 Ctrl + F
全屏模式 F11
切换主题 Ctrl + Shift + D
显示快捷键 ?
增大字号 Ctrl + =
减小字号 Ctrl + -