about summary refs log tree commit diff
path: root/sysdeps/x86_64/fpu/multiarch/svml_s_exp2f4_core_sse4.S
blob: f4ddfbe93299682b958bf4404a77fd1c8e93b066 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
/* Function exp2f vectorized with SSE4.
   Copyright (C) 2021-2023 Free Software Foundation, Inc.
   This file is part of the GNU C Library.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
   License as published by the Free Software Foundation; either
   version 2.1 of the License, or (at your option) any later version.

   The GNU C Library is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   Lesser General Public License for more details.

   You should have received a copy of the GNU Lesser General Public
   License along with the GNU C Library; if not, see
   https://www.gnu.org/licenses/.  */

/*
 * ALGORITHM DESCRIPTION:
 *
 *   exp2(x)  = 2^n * T[j] * (1 + P(y))
 *   where
 *        x = m*(1/K) + y,    y in [-1/K..1/K]
 *        m = n*K + j,           m, n,j - signed integer, j in [-K/2..K/2]
 *
 *        values of 2^j/K are tabulated
 *
 *        P(y) is a minimax polynomial approximation of exp2(x)-1
 *        on small interval [-1/K..1/K]
 *
 *  Special cases:
 *
 *   exp2(NaN)  = NaN
 *   exp2(+INF) = +INF
 *   exp2(-INF) = 0
 *   exp2(x)    = 1 for subnormals
 *   For IEEE float
 *     if x >= 128.0 then exp2f(x) overflow
 *     if x < -151.0 then exp2f(x) underflow
 *
 */

/* Offsets for data table __svml_sexp2_data_internal
 */
#define _sShifter			0
#define _sPC0				16
#define _sPC1				32
#define _sPC2				48
#define _sPC3				64
#define _sPC4				80
#define _sPC5				96
#define _sPC6				112
#define _iAbsMask			128
#define _iDomainRange			144

#include <sysdep.h>

	.section .text.sse4, "ax", @progbits
ENTRY(_ZGVbN4v_exp2f_sse4)
	subq	$72, %rsp
	cfi_def_cfa_offset(80)

	/* Check for overflow\underflow  */
	movups	__svml_sexp2_data_internal(%rip), %xmm1

	/*  Implementation  */
	movaps	%xmm1, %xmm5

	/*  Polynomial  */
	movups	_sPC6+__svml_sexp2_data_internal(%rip), %xmm4
	addps	%xmm0, %xmm5
	movaps	%xmm5, %xmm3

	/*  2^N  */
	pslld	$23, %xmm5

	/* Check for overflow\underflow  */
	movdqu	_iAbsMask+__svml_sexp2_data_internal(%rip), %xmm2
	subps	%xmm1, %xmm3

	/*  R  */
	movaps	%xmm0, %xmm1
	pand	%xmm0, %xmm2
	pcmpgtd	_iDomainRange+__svml_sexp2_data_internal(%rip), %xmm2
	subps	%xmm3, %xmm1
	movmskps %xmm2, %edx
	mulps	%xmm1, %xmm4
	addps	_sPC5+__svml_sexp2_data_internal(%rip), %xmm4
	mulps	%xmm1, %xmm4
	addps	_sPC4+__svml_sexp2_data_internal(%rip), %xmm4
	mulps	%xmm1, %xmm4
	addps	_sPC3+__svml_sexp2_data_internal(%rip), %xmm4
	mulps	%xmm1, %xmm4
	addps	_sPC2+__svml_sexp2_data_internal(%rip), %xmm4
	mulps	%xmm1, %xmm4
	addps	_sPC1+__svml_sexp2_data_internal(%rip), %xmm4
	mulps	%xmm4, %xmm1
	addps	_sPC0+__svml_sexp2_data_internal(%rip), %xmm1

	/*  Reconstruction  */
	paddd	%xmm5, %xmm1
	testl	%edx, %edx

	/* Go to special inputs processing branch */
	jne	L(SPECIAL_VALUES_BRANCH)
	# LOE rbx rbp r12 r13 r14 r15 edx xmm0 xmm1

	/* Restore registers
	 * and exit the function
	 */

L(EXIT):
	movaps	%xmm1, %xmm0
	addq	$72, %rsp
	cfi_def_cfa_offset(8)
	ret
	cfi_def_cfa_offset(80)

	/* Branch to process
	 * special inputs
	 */

L(SPECIAL_VALUES_BRANCH):
	movups	%xmm0, 32(%rsp)
	movups	%xmm1, 48(%rsp)
	# LOE rbx rbp r12 r13 r14 r15 edx

	xorl	%eax, %eax
	movq	%r12, 16(%rsp)
	cfi_offset(12, -64)
	movl	%eax, %r12d
	movq	%r13, 8(%rsp)
	cfi_offset(13, -72)
	movl	%edx, %r13d
	movq	%r14, (%rsp)
	cfi_offset(14, -80)
	# LOE rbx rbp r15 r12d r13d

	/* Range mask
	 * bits check
	 */

L(RANGEMASK_CHECK):
	btl	%r12d, %r13d

	/* Call scalar math function */
	jc	L(SCALAR_MATH_CALL)
	# LOE rbx rbp r15 r12d r13d

	/* Special inputs
	 * processing loop
	 */

L(SPECIAL_VALUES_LOOP):
	incl	%r12d
	cmpl	$4, %r12d

	/* Check bits in range mask */
	jl	L(RANGEMASK_CHECK)
	# LOE rbx rbp r15 r12d r13d

	movq	16(%rsp), %r12
	cfi_restore(12)
	movq	8(%rsp), %r13
	cfi_restore(13)
	movq	(%rsp), %r14
	cfi_restore(14)
	movups	48(%rsp), %xmm1

	/* Go to exit */
	jmp	L(EXIT)
	cfi_offset(12, -64)
	cfi_offset(13, -72)
	cfi_offset(14, -80)
	# LOE rbx rbp r12 r13 r14 r15 xmm1

	/* Scalar math fucntion call
	 * to process special input
	 */

L(SCALAR_MATH_CALL):
	movl	%r12d, %r14d
	movss	32(%rsp, %r14, 4), %xmm0
	call	exp2f@PLT
	# LOE rbx rbp r14 r15 r12d r13d xmm0

	movss	%xmm0, 48(%rsp, %r14, 4)

	/* Process special inputs in loop */
	jmp	L(SPECIAL_VALUES_LOOP)
	# LOE rbx rbp r15 r12d r13d
END(_ZGVbN4v_exp2f_sse4)

	.section .rodata, "a"
	.align	16

#ifdef __svml_sexp2_data_internal_typedef
typedef unsigned int VUINT32;
typedef struct {
	__declspec(align(16)) VUINT32 _sShifter[4][1];
	__declspec(align(16)) VUINT32 _sPC0[4][1];
	__declspec(align(16)) VUINT32 _sPC1[4][1];
	__declspec(align(16)) VUINT32 _sPC2[4][1];
	__declspec(align(16)) VUINT32 _sPC3[4][1];
	__declspec(align(16)) VUINT32 _sPC4[4][1];
	__declspec(align(16)) VUINT32 _sPC5[4][1];
	__declspec(align(16)) VUINT32 _sPC6[4][1];
	__declspec(align(16)) VUINT32 _iAbsMask[4][1];
	__declspec(align(16)) VUINT32 _iDomainRange[4][1];
} __svml_sexp2_data_internal;
#endif
__svml_sexp2_data_internal:
	.long	0x4b400000, 0x4b400000, 0x4b400000, 0x4b400000 /* _sShifter */
	.align	16
	.long	0x3F800000, 0x3F800000, 0x3F800000, 0x3F800000 /* _sPC0 */
	.align	16
	.long	0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218 /* _sPC1 */
	.align	16
	.long	0x3e75fdef, 0x3e75fdef, 0x3e75fdef, 0x3e75fdef /* _sPC2 */
	.align	16
	.long	0x3d6357cf, 0x3d6357cf, 0x3d6357cf, 0x3d6357cf /* _sPC3 */
	.align	16
	.long	0x3c1d962c, 0x3c1d962c, 0x3c1d962c, 0x3c1d962c /* _sPC4 */
	.align	16
	.long	0x3aaf7a51, 0x3aaf7a51, 0x3aaf7a51, 0x3aaf7a51 /* _sPC5 */
	.align	16
	.long	0x39213c8c, 0x39213c8c, 0x39213c8c, 0x39213c8c /* _sPC6 */
	//common
	.align	16
	.long	0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff /* _iAbsMask */
	.align	16
	.long	0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000 /* _iDomainRange=126.0 */
	.align	16
	.type	__svml_sexp2_data_internal, @object
	.size	__svml_sexp2_data_internal, .-__svml_sexp2_data_internal