1/* Function sincosf vectorized with AVX2, wrapper version.
2 Copyright (C) 2014-2023 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4
5 The GNU C Library is free software; you can redistribute it and/or
6 modify it under the terms of the GNU Lesser General Public
7 License as published by the Free Software Foundation; either
8 version 2.1 of the License, or (at your option) any later version.
9
10 The GNU C Library is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 Lesser General Public License for more details.
14
15 You should have received a copy of the GNU Lesser General Public
16 License along with the GNU C Library; if not, see
17 <https://www.gnu.org/licenses/>. */
18
19#include <sysdep.h>
20#include "svml_s_wrapper_impl.h"
21
22 .section .text.avx2, "ax", @progbits
23ENTRY (_ZGVdN8vl4l4_sincosf)
24WRAPPER_IMPL_AVX_fFF _ZGVbN4vl4l4_sincosf
25END (_ZGVdN8vl4l4_sincosf)
26libmvec_hidden_def (_ZGVdN8vl4l4_sincosf)
27
28/* AVX2 ISA version as wrapper to SSE ISA version (for vector
29 function declared with #pragma omp declare simd notinbranch). */
30.macro WRAPPER_IMPL_AVX2_fFF_vvv callee
31#ifndef __ILP32__
32 pushq %rbp
33 cfi_adjust_cfa_offset (8)
34 cfi_rel_offset (%rbp, 0)
35 movq %rsp, %rbp
36 cfi_def_cfa_register (%rbp)
37 andq $-32, %rsp
38 subq $224, %rsp
39 vmovups %ymm0, 192(%rsp)
40 lea (%rsp), %rdi
41 vmovdqu %ymm1, 64(%rdi)
42 vmovdqu %ymm2, 96(%rdi)
43 vmovdqu %ymm3, 128(%rdi)
44 vmovdqu %ymm4, 160(%rdi)
45 lea 32(%rsp), %rsi
46 vzeroupper
47 call HIDDEN_JUMPTARGET(\callee)
48 vmovups 208(%rsp), %xmm0
49 lea 16(%rsp), %rdi
50 lea 48(%rsp), %rsi
51 call HIDDEN_JUMPTARGET(\callee)
52 movq 64(%rsp), %rdx
53 movq 72(%rsp), %rsi
54 movq 80(%rsp), %r8
55 movq 88(%rsp), %r10
56 movl (%rsp), %eax
57 movl 4(%rsp), %ecx
58 movl 8(%rsp), %edi
59 movl 12(%rsp), %r9d
60 movl %eax, (%rdx)
61 movl %ecx, (%rsi)
62 movq 96(%rsp), %rax
63 movq 104(%rsp), %rcx
64 movl %edi, (%r8)
65 movl %r9d, (%r10)
66 movq 112(%rsp), %rdi
67 movq 120(%rsp), %r9
68 movl 16(%rsp), %r11d
69 movl 20(%rsp), %edx
70 movl 24(%rsp), %esi
71 movl 28(%rsp), %r8d
72 movl %r11d, (%rax)
73 movl %edx, (%rcx)
74 movq 128(%rsp), %r11
75 movq 136(%rsp), %rdx
76 movl %esi, (%rdi)
77 movl %r8d, (%r9)
78 movq 144(%rsp), %rsi
79 movq 152(%rsp), %r8
80 movl 32(%rsp), %r10d
81 movl 36(%rsp), %eax
82 movl 40(%rsp), %ecx
83 movl 44(%rsp), %edi
84 movl %r10d, (%r11)
85 movl %eax, (%rdx)
86 movq 160(%rsp), %r10
87 movq 168(%rsp), %rax
88 movl %ecx, (%rsi)
89 movl %edi, (%r8)
90 movq 176(%rsp), %rcx
91 movq 184(%rsp), %rdi
92 movl 48(%rsp), %r9d
93 movl 52(%rsp), %r11d
94 movl 56(%rsp), %edx
95 movl 60(%rsp), %esi
96 movl %r9d, (%r10)
97 movl %r11d, (%rax)
98 movl %edx, (%rcx)
99 movl %esi, (%rdi)
100 movq %rbp, %rsp
101 cfi_def_cfa_register (%rsp)
102 popq %rbp
103 cfi_adjust_cfa_offset (-8)
104 cfi_restore (%rbp)
105 ret
106#else
107 leal 8(%rsp), %r10d
108 .cfi_def_cfa 10, 0
109 andl $-32, %esp
110 pushq -8(%r10d)
111 pushq %rbp
112 .cfi_escape 0x10,0x6,0x2,0x76,0
113 movl %esp, %ebp
114 pushq %r12
115 leal -80(%rbp), %esi
116 pushq %r10
117 .cfi_escape 0xf,0x3,0x76,0x70,0x6
118 .cfi_escape 0x10,0xc,0x2,0x76,0x78
119 leal -112(%rbp), %edi
120 movq %rsi, %r12
121 pushq %rbx
122 .cfi_escape 0x10,0x3,0x2,0x76,0x68
123 movq %rdi, %rbx
124 subl $184, %esp
125 vmovdqa %ymm1, -144(%ebp)
126 vmovdqa %ymm2, -176(%ebp)
127 vmovaps %ymm0, -208(%ebp)
128 vzeroupper
129 call HIDDEN_JUMPTARGET(\callee)
130 leal 16(%r12), %esi
131 vmovups -192(%ebp), %xmm0
132 leal 16(%rbx), %edi
133 call HIDDEN_JUMPTARGET(\callee)
134 movl -144(%ebp), %eax
135 vmovss -112(%ebp), %xmm0
136 vmovss %xmm0, (%eax)
137 movl -140(%ebp), %eax
138 vmovss -108(%ebp), %xmm0
139 vmovss %xmm0, (%eax)
140 movl -136(%ebp), %eax
141 vmovss -104(%ebp), %xmm0
142 vmovss %xmm0, (%eax)
143 movl -132(%ebp), %eax
144 vmovss -100(%ebp), %xmm0
145 vmovss %xmm0, (%eax)
146 movl -128(%ebp), %eax
147 vmovss -96(%ebp), %xmm0
148 vmovss %xmm0, (%eax)
149 movl -124(%ebp), %eax
150 vmovss -92(%ebp), %xmm0
151 vmovss %xmm0, (%eax)
152 movl -120(%ebp), %eax
153 vmovss -88(%ebp), %xmm0
154 vmovss %xmm0, (%eax)
155 movl -116(%ebp), %eax
156 vmovss -84(%ebp), %xmm0
157 vmovss %xmm0, (%eax)
158 movl -176(%ebp), %eax
159 vmovss -80(%ebp), %xmm0
160 vmovss %xmm0, (%eax)
161 movl -172(%ebp), %eax
162 vmovss -76(%ebp), %xmm0
163 vmovss %xmm0, (%eax)
164 movl -168(%ebp), %eax
165 vmovss -72(%ebp), %xmm0
166 vmovss %xmm0, (%eax)
167 movl -164(%ebp), %eax
168 vmovss -68(%ebp), %xmm0
169 vmovss %xmm0, (%eax)
170 movl -160(%ebp), %eax
171 vmovss -64(%ebp), %xmm0
172 vmovss %xmm0, (%eax)
173 movl -156(%ebp), %eax
174 vmovss -60(%ebp), %xmm0
175 vmovss %xmm0, (%eax)
176 movl -152(%ebp), %eax
177 vmovss -56(%ebp), %xmm0
178 vmovss %xmm0, (%eax)
179 movl -148(%ebp), %eax
180 vmovss -52(%ebp), %xmm0
181 vmovss %xmm0, (%eax)
182 addl $184, %esp
183 popq %rbx
184 popq %r10
185 .cfi_def_cfa 10, 0
186 popq %r12
187 popq %rbp
188 leal -8(%r10), %esp
189 .cfi_def_cfa 7, 8
190 ret
191#endif
192.endm
193
194ENTRY (_ZGVdN8vvv_sincosf)
195WRAPPER_IMPL_AVX2_fFF_vvv _ZGVbN4vl4l4_sincosf
196END (_ZGVdN8vvv_sincosf)
197
198#ifndef USE_MULTIARCH
199 libmvec_hidden_def (_ZGVdN8vvv_sincosf)
200#endif
201