]>
Commit | Line | Data |
---|---|---|
a6336cc4 | 1 | /* Function sincosf vectorized in AVX ISA as wrapper to SSE4 ISA version. |
688903eb | 2 | Copyright (C) 2014-2018 Free Software Foundation, Inc. |
a6336cc4 AS |
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 | <http://www.gnu.org/licenses/>. */ | |
18 | ||
19 | #include <sysdep.h> | |
20 | #include "svml_s_wrapper_impl.h" | |
21 | ||
22 | .text | |
ee2196bb AS |
23 | ENTRY (_ZGVcN8vl4l4_sincosf) |
24 | WRAPPER_IMPL_AVX_fFF _ZGVbN4vl4l4_sincosf | |
25 | END (_ZGVcN8vl4l4_sincosf) | |
26 | ||
27 | /* AVX ISA version as wrapper to SSE ISA version (for vector | |
28 | function declared with #pragma omp declare simd notinbranch). */ | |
29 | .macro WRAPPER_IMPL_AVX_fFF_vvv callee | |
30 | #ifndef __ILP32__ | |
31 | pushq %rbp | |
32 | movq %rsp, %rbp | |
33 | andq $-32, %rsp | |
34 | subq $224, %rsp | |
35 | vmovups %ymm0, 64(%rsp) | |
36 | lea (%rsp), %rdi | |
37 | vmovdqu %xmm1, 96(%rdi) | |
38 | vmovdqu %xmm2, 112(%rdi) | |
39 | vmovdqu %xmm3, 128(%rdi) | |
40 | vmovdqu %xmm4, 144(%rdi) | |
41 | vmovdqu %xmm5, 160(%rdi) | |
42 | lea 32(%rsp), %rsi | |
43 | vmovdqu %xmm6, 144(%rsi) | |
44 | vmovdqu %xmm7, 160(%rsi) | |
45 | vzeroupper | |
46 | call HIDDEN_JUMPTARGET(\callee) | |
47 | vmovdqu 80(%rsp), %xmm0 | |
48 | lea 16(%rsp), %rdi | |
49 | lea 48(%rsp), %rsi | |
50 | call HIDDEN_JUMPTARGET(\callee) | |
51 | movq 96(%rsp), %rdx | |
52 | movq 104(%rsp), %rsi | |
53 | movq 112(%rsp), %r8 | |
54 | movq 120(%rsp), %r10 | |
55 | movl (%rsp), %eax | |
56 | movl 4(%rsp), %ecx | |
57 | movl 8(%rsp), %edi | |
58 | movl 12(%rsp), %r9d | |
59 | movl %eax, (%rdx) | |
60 | movl %ecx, (%rsi) | |
61 | movq 128(%rsp), %rax | |
62 | movq 136(%rsp), %rcx | |
63 | movl %edi, (%r8) | |
64 | movl %r9d, (%r10) | |
65 | movq 144(%rsp), %rdi | |
66 | movq 152(%rsp), %r9 | |
67 | movl 16(%rsp), %r11d | |
68 | movl 20(%rsp), %edx | |
69 | movl 24(%rsp), %esi | |
70 | movl 28(%rsp), %r8d | |
71 | movl %r11d, (%rax) | |
72 | movl %edx, (%rcx) | |
73 | movq 160(%rsp), %r11 | |
74 | movq 168(%rsp), %rdx | |
75 | movl %esi, (%rdi) | |
76 | movl %r8d, (%r9) | |
77 | movq 176(%rsp), %rsi | |
78 | movq 184(%rsp), %r8 | |
79 | movl 32(%rsp), %r10d | |
80 | movl 36(%rsp), %eax | |
81 | movl 40(%rsp), %ecx | |
82 | movl 44(%rsp), %edi | |
83 | movl %r10d, (%r11) | |
84 | movl %eax, (%rdx) | |
85 | movq 192(%rsp), %r10 | |
86 | movq 200(%rsp), %rax | |
87 | movl %ecx, (%rsi) | |
88 | movl %edi, (%r8) | |
89 | movq 16(%rbp), %rcx | |
90 | movq 24(%rbp), %rdi | |
91 | movl 48(%rsp), %r9d | |
92 | movl 52(%rsp), %r11d | |
93 | movl 56(%rsp), %edx | |
94 | movl 60(%rsp), %esi | |
95 | movl %r9d, (%r10) | |
96 | movl %r11d, (%rax) | |
97 | movl %edx, (%rcx) | |
98 | movl %esi, (%rdi) | |
99 | movq %rbp, %rsp | |
100 | popq %rbp | |
101 | ret | |
102 | #else | |
103 | leal 8(%rsp), %r10d | |
104 | .cfi_def_cfa 10, 0 | |
105 | andl $-32, %esp | |
106 | pushq -8(%r10d) | |
107 | pushq %rbp | |
108 | .cfi_escape 0x10,0x6,0x2,0x76,0 | |
109 | movl %esp, %ebp | |
110 | pushq %r12 | |
111 | leal -80(%rbp), %esi | |
112 | pushq %r10 | |
113 | .cfi_escape 0xf,0x3,0x76,0x70,0x6 | |
114 | .cfi_escape 0x10,0xc,0x2,0x76,0x78 | |
115 | leal -112(%rbp), %edi | |
116 | movq %rsi, %r12 | |
117 | pushq %rbx | |
118 | .cfi_escape 0x10,0x3,0x2,0x76,0x68 | |
119 | movq %rdi, %rbx | |
120 | subl $184, %esp | |
121 | vmovaps %xmm1, -128(%ebp) | |
122 | vmovaps %xmm2, -144(%ebp) | |
123 | vmovaps %xmm3, -160(%ebp) | |
124 | vmovaps %xmm4, -176(%ebp) | |
125 | vmovaps %ymm0, -208(%ebp) | |
126 | vzeroupper | |
127 | call HIDDEN_JUMPTARGET(\callee) | |
128 | leal 16(%r12), %esi | |
129 | vmovups -192(%ebp), %xmm0 | |
130 | leal 16(%rbx), %edi | |
131 | call HIDDEN_JUMPTARGET(\callee) | |
132 | movq -128(%ebp), %rax | |
133 | vmovss -112(%ebp), %xmm0 | |
134 | vmovdqa -128(%ebp), %xmm7 | |
135 | vmovdqa -144(%ebp), %xmm3 | |
136 | vmovss %xmm0, (%eax) | |
137 | vmovss -108(%ebp), %xmm0 | |
138 | vpextrd $1, %xmm7, %eax | |
139 | vmovss %xmm0, (%eax) | |
140 | movq -120(%ebp), %rax | |
141 | vmovss -104(%ebp), %xmm0 | |
142 | vmovss %xmm0, (%eax) | |
143 | vmovss -100(%ebp), %xmm0 | |
144 | vpextrd $3, %xmm7, %eax | |
145 | vmovdqa -160(%ebp), %xmm7 | |
146 | vmovss %xmm0, (%eax) | |
147 | movq -144(%ebp), %rax | |
148 | vmovss -96(%ebp), %xmm0 | |
149 | vmovss %xmm0, (%eax) | |
150 | vmovss -92(%ebp), %xmm0 | |
151 | vpextrd $1, %xmm3, %eax | |
152 | vmovss %xmm0, (%eax) | |
153 | movq -136(%ebp), %rax | |
154 | vmovss -88(%ebp), %xmm0 | |
155 | vmovss %xmm0, (%eax) | |
156 | vmovss -84(%ebp), %xmm0 | |
157 | vpextrd $3, %xmm3, %eax | |
158 | vmovss %xmm0, (%eax) | |
159 | movq -160(%ebp), %rax | |
160 | vmovss -80(%ebp), %xmm0 | |
161 | vmovss %xmm0, (%eax) | |
162 | vmovss -76(%ebp), %xmm0 | |
163 | vpextrd $1, %xmm7, %eax | |
164 | vmovss %xmm0, (%eax) | |
165 | movq -152(%ebp), %rax | |
166 | vmovss -72(%ebp), %xmm0 | |
167 | vmovss %xmm0, (%eax) | |
168 | vmovss -68(%ebp), %xmm0 | |
169 | vpextrd $3, %xmm7, %eax | |
170 | vmovss %xmm0, (%eax) | |
171 | movq -176(%ebp), %rax | |
172 | vmovss -64(%ebp), %xmm0 | |
173 | vmovdqa -176(%ebp), %xmm3 | |
174 | vmovss %xmm0, (%eax) | |
175 | vmovss -60(%ebp), %xmm0 | |
176 | vpextrd $1, %xmm3, %eax | |
177 | vmovss %xmm0, (%eax) | |
178 | movq -168(%ebp), %rax | |
179 | vmovss -56(%ebp), %xmm0 | |
180 | vmovss %xmm0, (%eax) | |
181 | vmovss -52(%ebp), %xmm0 | |
182 | vpextrd $3, %xmm3, %eax | |
183 | vmovss %xmm0, (%eax) | |
184 | addl $184, %esp | |
185 | popq %rbx | |
186 | popq %r10 | |
187 | .cfi_def_cfa 10, 0 | |
188 | popq %r12 | |
189 | popq %rbp | |
190 | leal -8(%r10), %esp | |
191 | .cfi_def_cfa 7, 8 | |
192 | ret | |
193 | #endif | |
194 | .endm | |
195 | ||
196 | ENTRY (_ZGVcN8vvv_sincosf) | |
197 | WRAPPER_IMPL_AVX_fFF_vvv _ZGVbN4vl4l4_sincosf | |
198 | END (_ZGVcN8vvv_sincosf) |