⭐ 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

📄 zgemm_kernel.s

📁 Optimized GotoBLAS libraries
💻 S
📖 第 1 页 / 共 5 页
字号:
#else	nop	__LINE__#endif	(p3) FMA	f82   = f42, f58, f82	// A3 * B3	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f78 = [C5], SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f83   = f42, f59, f83	// A3 * B4	nop	__LINE__	}	;;/* 43 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f75 = [C1], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA	f98   = f42, f60, f98	// A3 * B5	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f79 = [C5], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f99  = f42, f61, f99	// A3 * B6	nop	__LINE__	}	;;/* 44 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f88 = [C2], SIZE#else	nop	__LINE__#endif	(p3) FMA	f114  = f42, f62, f114	// A3 * B7	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f92 = [C6], SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f115  = f42, f63, f115	// A3 * B8	nop	__LINE__	}	;;/* 45 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f89 = [C2], SIZE#else	nop	__LINE__#endif	(p3) FMA	f67   = f43, f56, f67	// A4 * B1	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f93 = [C6], SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f66   = f43, f57, f66	// A4 * B2	nop	__LINE__	}	;;/* 46 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f90 = [C2], SIZE#else	nop	__LINE__#endif	(p3) FMA	f83   = f43, f58, f83	// A4 * B3	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f94 = [C6], SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f82   = f43, f59, f82	// A4 * B4	nop	__LINE__	}	;;/* 47 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f91 = [C2], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA	f99   = f43, f60, f99	// A4 * B5	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f95 = [C6], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f98  = f43, f61, f98	// A4 * B6	nop	__LINE__	}	;;/* 48 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f104 = [C3], SIZE#else	nop	__LINE__#endif	(p3) FMA	f115  = f43, f62, f115	// A4 * B7	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f108 = [C7], SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f114  = f43, f63, f114	// A4 * B8	nop	__LINE__	}	;;/* 49 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f105 = [C3], SIZE#else	nop	__LINE__#endif	(p3) FMA	f68   = f44, f56, f68	// A5 * B1	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f109 = [C7], SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f69   = f44, f57, f69	// A5 * B2	nop	__LINE__	}	;;/* 50 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f106 = [C3], SIZE#else	nop	__LINE__#endif	(p3) FMA	f84   = f44, f58, f84	// A5 * B3	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f110 = [C7], SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f85   = f44, f59, f85	// A5 * B4	nop	__LINE__	}	;;/* 51 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f107 = [C3], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA	f100  = f44, f60, f100	// A5 * B5	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f111 = [C7], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f101  = f44, f61, f101	// A5 * B6	nop	__LINE__	}	;;/* 52 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f120 = [C4], SIZE#else	nop	__LINE__#endif	(p3) FMA	f116  = f44, f62, f116	// A5 * B7	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f124 = [C8], SIZE#else	nop	__LINE__#endif	(p3) FMA_B	f117  = f44, f63, f117	// A5 * B8	nop	__LINE__	}	;;/* 53 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f121 = [C4], SIZE#else	nop	__LINE__#endif	(p3) FMA	f69   = f45, f56, f69	// A6 * B1	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f125 = [C8], SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f68   = f45, f57, f68	// A6 * B2	nop	__LINE__	}	;;/* 54 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f122 = [C4], SIZE#else	nop	__LINE__#endif	(p3) FMA	f85   = f45, f58, f85	// A6 * B3	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f126 = [C8], SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f84   = f45, f59, f84	// A6 * B4	nop	__LINE__	}	;;/* 55 */	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f123 = [C4], -3 * SIZE#else	nop	__LINE__#endif 	(p3) FMA	f101  = f45, f60, f101	// A6 * B5	nop	__LINE__	}	{ .mfb#if! defined(TRMMKERNEL) && !defined(BETAZERO)	(p5) LDFD	f127 = [C8], -3 * SIZE#else	nop	__LINE__#endif	(p3) FMA_A	f100  = f45, f61, f100	// A6 * B6	nop	__LINE__	}	;;/* 56 */	{ .mfb	nop	__LINE__	(p3) FMA	f117  = f45, f62, f117	// A6 * B7	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_A	f116  = f45, f63, f116	// A6 * B8	nop	__LINE__	}	;;/* 57 */	{ .mfb	nop	__LINE__	(p3) FMA	f70   = f46, f56, f70	// A7 * B1	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_B	f71   = f46, f57, f71	// A7 * B2	nop	__LINE__	}	;;/* 58 */	{ .mfb	nop	__LINE__	(p3) FMA	f86   = f46, f58, f86	// A7 * B3	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_B	f87   = f46, f59, f87	// A7 * B4	nop	__LINE__	}	;;/* 59 */	{ .mfb	nop	__LINE__	(p3) FMA	f102  = f46, f60, f102	// A7 * B5	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_B	f103  = f46, f61, f103	// A7 * B6	nop	__LINE__	}	;;/* 60 */	{ .mfb	nop	__LINE__	(p3) FMA	f118  = f46, f62, f118	// A7 * B7	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_B	f119  = f46, f63, f119	// A7 * B8	nop	__LINE__	}	;;/* 61 */	{ .mfb	nop	__LINE__	(p3) FMA	f71   = f47, f56, f71	// A8 * B1	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_A	f70   = f47, f57, f70	// A8 * B2	nop	__LINE__	}	;;/* 62 */	{ .mfb	nop	__LINE__	(p3) FMA	f87   = f47, f58, f87	// A8 * B3	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_A	f86   = f47, f59, f86	// A8 * B4	nop	__LINE__	}	;;/* 63 */	{ .mfb	nop	__LINE__	(p3) FMA	f103  = f47, f60, f103	// A8 * B5	nop	__LINE__	}	{ .mfb	nop	__LINE__	(p3) FMA_A	f102  = f47, f61, f102	// A8 * B6	nop	__LINE__	}	;;/* 64 */	{ .mfi	nop	__LINE__	(p3) FMA	f119  = f47, f62, f119	// A8 * B7	adds	L = -1, L	}	{ .mfb	nop	__LINE__	(p3) FMA_A	f118  = f47, f63, f118	// A8 * B8	br.cloop.sptk.few .L012	}	;;#if! defined(TRMMKERNEL) && !defined(BETAZERO)	{ .mfb	nop	__LINE__	FMA	f72  = ALPHA_R, f64, f72	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f76 = ALPHA_R, f68, f76	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_C	f73  = ALPHA_R, f65, f73	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_C	f77 = ALPHA_R, f69, f77	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f74 = ALPHA_R, f66, f74	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f78 = ALPHA_R, f70, f78	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_C	f75 = ALPHA_R, f67, f75	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_C	f79 = ALPHA_R, f71, f79	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f72  = ALPHA_I, f65, f72	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f76 = ALPHA_I, f69, f76	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f73  = ALPHA_I, f64, f73	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f77 = ALPHA_I, f68, f77	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f74 = ALPHA_I, f67, f74	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f78 = ALPHA_I, f71, f78	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f75 = ALPHA_I, f66, f75	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f79 = ALPHA_I, f70, f79	nop	__LINE__	}	;;	{ .mfb	STFD	[C1] = f72, SIZE	FMA	f88 = ALPHA_R, f80, f88	nop	__LINE__	}	{ .mfb	STFD	[C5] = f76, SIZE	FMA	f92 = ALPHA_R, f84, f92	nop	__LINE__	}	;;	{ .mfb	STFD	[C1] = f73, SIZE	FCALC_C	f89 = ALPHA_R, f81, f89	nop	__LINE__	}	{ .mfb	STFD	[C5] = f77, SIZE	FCALC_C	f93 = ALPHA_R, f85, f93	nop	__LINE__	}	;;	{ .mfb	STFD	[C1] = f74, SIZE	FMA	f90 = ALPHA_R, f82, f90	nop	__LINE__	}	{ .mfb	STFD	[C5] = f78, SIZE	FMA	f94 = ALPHA_R, f86, f94	nop	__LINE__	}	;;	{ .mfb	STFD	[C1] = f75, 5 * SIZE	FCALC_C	f91 = ALPHA_R, f83, f91	nop	__LINE__	}	{ .mfb	STFD	[C5] = f79, 5 * SIZE	FCALC_C	f95 = ALPHA_R, f87, f95	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f88 = ALPHA_I, f81, f88	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f92 = ALPHA_I, f85, f92	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f89 = ALPHA_I, f80, f89	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f93 = ALPHA_I, f84, f93	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f90 = ALPHA_I, f83, f90	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f94 = ALPHA_I, f87, f94	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f91 = ALPHA_I, f82, f91	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f95 = ALPHA_I, f86, f95	nop	__LINE__	}	;;	{ .mfb	STFD	[C2] = f88, SIZE	FMA	f104 = ALPHA_R, f96,  f104	nop	__LINE__	}	{ .mfb	STFD	[C6] = f92, SIZE	FMA	f108 = ALPHA_R, f100, f108	nop	__LINE__	}	;;	{ .mfb	STFD	[C2] = f89, SIZE	FCALC_C	f105 = ALPHA_R, f97,  f105	nop	__LINE__	}	{ .mfb	STFD	[C6] = f93, SIZE	FCALC_C	f109 = ALPHA_R, f101, f109	nop	__LINE__	}	;;	{ .mfb	STFD	[C2] = f90, SIZE	FMA	f106 = ALPHA_R, f98,  f106	nop	__LINE__	}	{ .mfb	STFD	[C6] = f94, SIZE	FMA	f110 = ALPHA_R, f102, f110	nop	__LINE__	}	;;	{ .mfb	STFD	[C2] = f91, 5 * SIZE	FCALC_C	f107 = ALPHA_R, f99,  f107	nop	__LINE__	}	{ .mfb	STFD	[C6] = f95, 5 * SIZE	FCALC_C	f111 = ALPHA_R, f103, f111	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f104 = ALPHA_I, f97,  f104	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f108 = ALPHA_I, f101, f108	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f105 = ALPHA_I, f96,  f105	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f109 = ALPHA_I, f100, f109	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f106 = ALPHA_I, f99,  f106	nop	__LINE__	}	{ .mfb	nop	__LINE__ 	FCALC_D	f110 = ALPHA_I, f103, f110	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f107 = ALPHA_I, f98,  f107	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f111 = ALPHA_I, f102, f111	nop	__LINE__	}	;;	{ .mfb	STFD	[C3] = f104, SIZE	FMA	f120 = ALPHA_R, f112, f120	nop	__LINE__	}	{ .mfb	STFD	[C7] = f108, SIZE	FMA	f124 = ALPHA_R, f116, f124	nop	__LINE__	}	;;	{ .mfb	STFD	[C3] = f105, SIZE	FCALC_C	f121 = ALPHA_R, f113, f121	nop	__LINE__	}	{ .mfb	STFD	[C7] = f109, SIZE	FCALC_C	f125 = ALPHA_R, f117, f125	nop	__LINE__	}	;;	{ .mfb	STFD	[C3] = f106, SIZE	FMA	f122 = ALPHA_R, f114, f122	nop	__LINE__	}	{ .mfb	STFD	[C7] = f110, SIZE	FMA	f126 = ALPHA_R, f118, f126	nop	__LINE__	}	;;	{ .mfb	STFD	[C3] = f107, 5 * SIZE	FCALC_C	f123 = ALPHA_R, f115, f123	nop	__LINE__	}	{ .mfb	STFD	[C7] = f111, 5 * SIZE	FCALC_C	f127 = ALPHA_R, f119, f127	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f120 = ALPHA_I, f113, f120	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f124 = ALPHA_I, f117, f124	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMA	f121 = ALPHA_I, f112, f121	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMA	f125 = ALPHA_I, f116, f125	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_D	f122 = ALPHA_I, f115, f122	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_D	f126 = ALPHA_I, f119, f126	nop	__LINE__	}	;;	{ .mfi	nop	__LINE__	FMA	f123 = ALPHA_I, f114, f123	cmp.ne	p6, p0 = 1, I	}	{ .mfb	nop	__LINE__	FMA	f127 = ALPHA_I, f118, f127	nop	__LINE__	}	;;	{ .mfi	STFD	[C4] = f120, SIZE	mov	f64  = f0	adds	I = -1, I	}	{ .mfb	STFD	[C8] = f124, SIZE	mov	f65  = f0	nop	__LINE__	}	;;	{ .mfi	STFD	[C4] = f121, SIZE	mov	f80  = f0	and	TEMP = 3, M	}	{ .mfb	STFD	[C8] = f125, SIZE	mov	f81  = f0	nop	__LINE__	}	;;	{ .mfi	STFD	[C4] = f122, SIZE	mov	f96  = f0	cmp.ne	p8, p9 = r0, TEMP	}	{ .mfb	STFD	[C8] = f126, SIZE	mov	f97  = f0	nop	__LINE__	}	;;	{ .mfb	STFD	[C4] = f123, 5 * SIZE	mov	f112 = f0	nop	__LINE__	}	{ .mfb	STFD	[C8] = f127, 5 * SIZE	mov	f113  = f0	(p6)	br.cond.dptk .L011	}	;;#else	{ .mfb	nop	__LINE__	FMPY	f72  = ALPHA_R, f64	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMPY	f76 = ALPHA_R, f68	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_C	f73  = ALPHA_R, f65, f0	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_C	f77 = ALPHA_R, f69, f0	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FMPY	f74 = ALPHA_R, f66	nop	__LINE__	}	{ .mfb	nop	__LINE__	FMPY	f78 = ALPHA_R, f70	nop	__LINE__	}	;;	{ .mfb	nop	__LINE__	FCALC_C	f75 = ALPHA_R, f67, f0	nop	__LINE__	}	{ .mfb	nop	__LINE__	FCALC_C	f79 = ALPHA_R, f71, f0

⌨️ 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 + -