codekingpro/portable-devtools

1// Copyright 2017 The Go Authors. All rights reserved.2// Use of this source code is governed by a BSD-style3// license that can be found in the LICENSE file.4 5#define	Ln2Hi	6.93147180369123816490e-016#define	Ln2Lo	1.90821492927058770002e-107#define	Log2e	1.44269504088896338700e+008#define	Overflow	7.09782712893383973096e+029#define	Underflow	-7.45133219101941108420e+0210#define	Overflow2	1.0239999999999999e+0311#define	Underflow2	-1.0740e+0312#define	NearZero	0x3e30000000000000	// 2**-2813#define	PosInf	0x7ff000000000000014#define	FracMask	0x000fffffffffffff15#define	C1	0x3cb0000000000000	// 2**-5216#define	P1	1.66666666666666657415e-01	// 0x3FC55555; 0x5555555517#define	P2	-2.77777777770155933842e-03	// 0xBF66C16C; 0x16BEBD9318#define	P3	6.61375632143793436117e-05	// 0x3F11566A; 0xAF25DE2C19#define	P4	-1.65339022054652515390e-06	// 0xBEBBBD41; 0xC5D26BF120#define	P5	4.13813679705723846039e-08	// 0x3E663769; 0x72BEA4D021 22// Exp returns e**x, the base-e exponential of x.23// This is an assembly implementation of the method used for function Exp in file exp.go.24//25// func Exp(x float64) float6426TEXT ·archExp(SB),$0-1627	FMOVD	x+0(FP), F0	// F0 = x28	FCMPD	F0, F029	BNE	isNaN		// x = NaN, return NaN30	FMOVD	$Overflow, F131	FCMPD	F1, F032	BGT	overflow	// x > Overflow, return PosInf33	FMOVD	$Underflow, F134	FCMPD	F1, F035	BLT	underflow	// x < Underflow, return 036	MOVD	$NearZero, R037	FMOVD	R0, F238	FABSD	F0, F339	FMOVD	$1.0, F1	// F1 = 1.040	FCMPD	F2, F341	BLT	nearzero	// fabs(x) < NearZero, return 1 + x42	// argument reduction, x = k*ln2 + r,  |r| <= 0.5*ln243	// computed as r = hi - lo for extra precision.44	FMOVD	$Log2e, F245	FMOVD	$0.5, F346	FNMSUBD	F0, F3, F2, F4	// Log2e*x - 0.547	FMADDD	F0, F3, F2, F3	// Log2e*x + 0.548	FCMPD	$0.0, F049	FCSELD	LT, F4, F3, F3	// F3 = k50	FCVTZSD	F3, R1		// R1 = int(k)51	SCVTFD	R1, F3		// F3 = float64(int(k))52	FMOVD	$Ln2Hi, F4	// F4 = Ln2Hi53	FMOVD	$Ln2Lo, F5	// F5 = Ln2Lo54	FMSUBD	F3, F0, F4, F4	// F4 = hi = x - float64(int(k))*Ln2Hi55	FMULD	F3, F5		// F5 = lo = float64(int(k)) * Ln2Lo56	FSUBD	F5, F4, F6	// F6 = r = hi - lo57	FMULD	F6, F6, F7	// F7 = t = r * r58	// compute y59	FMOVD	$P5, F8		// F8 = P560	FMOVD	$P4, F9		// F9 = P461	FMADDD	F7, F9, F8, F13	// P4+t*P562	FMOVD	$P3, F10	// F10 = P363	FMADDD	F7, F10, F13, F13	// P3+t*(P4+t*P5)64	FMOVD	$P2, F11	// F11 = P265	FMADDD	F7, F11, F13, F13	// P2+t*(P3+t*(P4+t*P5))66	FMOVD	$P1, F12	// F12 = P167	FMADDD	F7, F12, F13, F13	// P1+t*(P2+t*(P3+t*(P4+t*P5)))68	FMSUBD	F7, F6, F13, F13	// F13 = c = r - t*(P1+t*(P2+t*(P3+t*(P4+t*P5))))69	FMOVD	$2.0, F1470	FSUBD	F13, F1471	FMULD	F6, F13, F1572	FDIVD	F14, F15	// F15 = (r*c)/(2-c)73	FSUBD	F15, F5, F15	// lo-(r*c)/(2-c)74	FSUBD	F4, F15, F15	// (lo-(r*c)/(2-c))-hi75	FSUBD	F15, F1, F16	// F16 = y = 1-((lo-(r*c)/(2-c))-hi)76	// inline Ldexp(y, k), benefit:77	// 1, no parameter pass overhead.78	// 2, skip unnecessary checks for Inf/NaN/Zero79	FMOVD	F16, R080	AND	$FracMask, R0, R2	// fraction81	LSR	$52, R0, R5	// exponent82	ADD	R1, R5		// R1 = int(k)83	CMP	$1, R584	BGE	normal85	ADD	$52, R5		// denormal86	MOVD	$C1, R887	FMOVD	R8, F1		// m = 2**-5288normal:89	ORR	R5<<52, R2, R090	FMOVD	R0, F091	FMULD	F1, F0		// return m * x92	FMOVD	F0, ret+8(FP)93	RET94nearzero:95	FADDD	F1, F096isNaN:97	FMOVD	F0, ret+8(FP)98	RET99underflow:100	MOVD	ZR, ret+8(FP)101	RET102overflow:103	MOVD	$PosInf, R0104	MOVD	R0, ret+8(FP)105	RET106 107 108// Exp2 returns 2**x, the base-2 exponential of x.109// This is an assembly implementation of the method used for function Exp2 in file exp.go.110//111// func Exp2(x float64) float64112TEXT ·archExp2(SB),$0-16113	FMOVD	x+0(FP), F0	// F0 = x114	FCMPD	F0, F0115	BNE	isNaN		// x = NaN, return NaN116	FMOVD	$Overflow2, F1117	FCMPD	F1, F0118	BGT	overflow	// x > Overflow, return PosInf119	FMOVD	$Underflow2, F1120	FCMPD	F1, F0121	BLT	underflow	// x < Underflow, return 0122	// argument reduction; x = r*lg(e) + k with |r| <= ln(2)/2123	// computed as r = hi - lo for extra precision.124	FMOVD	$0.5, F2125	FSUBD	F2, F0, F3	// x + 0.5126	FADDD	F2, F0, F4	// x - 0.5127	FCMPD	$0.0, F0128	FCSELD	LT, F3, F4, F3	// F3 = k129	FCVTZSD	F3, R1		// R1 = int(k)130	SCVTFD	R1, F3		// F3 = float64(int(k))131	FSUBD	F3, F0, F3	// t = x - float64(int(k))132	FMOVD	$Ln2Hi, F4	// F4 = Ln2Hi133	FMOVD	$Ln2Lo, F5	// F5 = Ln2Lo134	FMULD	F3, F4		// F4 = hi = t * Ln2Hi135	FNMULD	F3, F5		// F5 = lo = -t * Ln2Lo136	FSUBD	F5, F4, F6	// F6 = r = hi - lo137	FMULD	F6, F6, F7	// F7 = t = r * r138	// compute y139	FMOVD	$P5, F8		// F8 = P5140	FMOVD	$P4, F9		// F9 = P4141	FMADDD	F7, F9, F8, F13	// P4+t*P5142	FMOVD	$P3, F10	// F10 = P3143	FMADDD	F7, F10, F13, F13	// P3+t*(P4+t*P5)144	FMOVD	$P2, F11	// F11 = P2145	FMADDD	F7, F11, F13, F13	// P2+t*(P3+t*(P4+t*P5))146	FMOVD	$P1, F12	// F12 = P1147	FMADDD	F7, F12, F13, F13	// P1+t*(P2+t*(P3+t*(P4+t*P5)))148	FMSUBD	F7, F6, F13, F13	// F13 = c = r - t*(P1+t*(P2+t*(P3+t*(P4+t*P5))))149	FMOVD	$2.0, F14150	FSUBD	F13, F14151	FMULD	F6, F13, F15152	FDIVD	F14, F15	// F15 = (r*c)/(2-c)153	FMOVD	$1.0, F1	// F1 = 1.0154	FSUBD	F15, F5, F15	// lo-(r*c)/(2-c)155	FSUBD	F4, F15, F15	// (lo-(r*c)/(2-c))-hi156	FSUBD	F15, F1, F16	// F16 = y = 1-((lo-(r*c)/(2-c))-hi)157	// inline Ldexp(y, k), benefit:158	// 1, no parameter pass overhead.159	// 2, skip unnecessary checks for Inf/NaN/Zero160	FMOVD	F16, R0161	AND	$FracMask, R0, R2	// fraction162	LSR	$52, R0, R5	// exponent163	ADD	R1, R5		// R1 = int(k)164	CMP	$1, R5165	BGE	normal166	ADD	$52, R5		// denormal167	MOVD	$C1, R8168	FMOVD	R8, F1		// m = 2**-52169normal:170	ORR	R5<<52, R2, R0171	FMOVD	R0, F0172	FMULD	F1, F0		// return m * x173isNaN:174	FMOVD	F0, ret+8(FP)175	RET176underflow:177	MOVD	ZR, ret+8(FP)178	RET179overflow:180	MOVD	$PosInf, R0181	MOVD	R0, ret+8(FP)182	RET183