| 1 | /* Wrapper implementations of vector math functions. |
|---|
| 2 | Copyright (C) 2014-2024 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 | /* SSE2 ISA version as wrapper to scalar. */ |
|---|
| 20 | .macro WRAPPER_IMPL_SSE2 callee |
|---|
| 21 | push %rbx |
|---|
| 22 | cfi_adjust_cfa_offset (8) |
|---|
| 23 | cfi_rel_offset (%rbx, 0) |
|---|
| 24 | subq $16, %rsp |
|---|
| 25 | cfi_adjust_cfa_offset (16) |
|---|
| 26 | movaps %xmm0, (%rsp) |
|---|
| 27 | call JUMPTARGET(\callee) |
|---|
| 28 | movss %xmm0, (%rsp) |
|---|
| 29 | movss 4(%rsp), %xmm0 |
|---|
| 30 | call JUMPTARGET(\callee) |
|---|
| 31 | movss %xmm0, 4(%rsp) |
|---|
| 32 | movss 8(%rsp), %xmm0 |
|---|
| 33 | call JUMPTARGET(\callee) |
|---|
| 34 | movd %xmm0, %ebx |
|---|
| 35 | movss 12(%rsp), %xmm0 |
|---|
| 36 | call JUMPTARGET(\callee) |
|---|
| 37 | movd %ebx, %xmm1 |
|---|
| 38 | unpcklps %xmm0, %xmm1 |
|---|
| 39 | movsd (%rsp), %xmm0 |
|---|
| 40 | unpcklpd %xmm1, %xmm0 |
|---|
| 41 | addq $16, %rsp |
|---|
| 42 | cfi_adjust_cfa_offset (-16) |
|---|
| 43 | popq %rbx |
|---|
| 44 | cfi_adjust_cfa_offset (-8) |
|---|
| 45 | cfi_restore (%rbx) |
|---|
| 46 | ret |
|---|
| 47 | .endm |
|---|
| 48 | |
|---|
| 49 | /* 2 argument SSE2 ISA version as wrapper to scalar. */ |
|---|
| 50 | .macro WRAPPER_IMPL_SSE2_ff callee |
|---|
| 51 | push %rbx |
|---|
| 52 | cfi_adjust_cfa_offset (8) |
|---|
| 53 | cfi_rel_offset (%rbx, 0) |
|---|
| 54 | subq $32, %rsp |
|---|
| 55 | cfi_adjust_cfa_offset (40) |
|---|
| 56 | movaps %xmm0, (%rsp) |
|---|
| 57 | movaps %xmm1, 16(%rsp) |
|---|
| 58 | call JUMPTARGET(\callee) |
|---|
| 59 | movss 20(%rsp), %xmm1 |
|---|
| 60 | movss %xmm0, 0(%rsp) |
|---|
| 61 | movss 4(%rsp), %xmm0 |
|---|
| 62 | call JUMPTARGET(\callee) |
|---|
| 63 | movss 24(%rsp), %xmm1 |
|---|
| 64 | movss %xmm0, 4(%rsp) |
|---|
| 65 | movss 8(%rsp), %xmm0 |
|---|
| 66 | call JUMPTARGET(\callee) |
|---|
| 67 | movss 28(%rsp), %xmm1 |
|---|
| 68 | movd %xmm0, %ebx |
|---|
| 69 | movss 12(%rsp), %xmm0 |
|---|
| 70 | call JUMPTARGET(\callee) |
|---|
| 71 | /* merge 4x results into xmm0. */ |
|---|
| 72 | movd %ebx, %xmm1 |
|---|
| 73 | unpcklps %xmm0, %xmm1 |
|---|
| 74 | movsd (%rsp), %xmm0 |
|---|
| 75 | unpcklpd %xmm1, %xmm0 |
|---|
| 76 | addq $32, %rsp |
|---|
| 77 | cfi_adjust_cfa_offset (-32) |
|---|
| 78 | popq %rbx |
|---|
| 79 | cfi_adjust_cfa_offset (-8) |
|---|
| 80 | cfi_restore (%rbx) |
|---|
| 81 | ret |
|---|
| 82 | .endm |
|---|
| 83 | |
|---|
| 84 | /* 3 argument SSE2 ISA version as wrapper to scalar. */ |
|---|
| 85 | .macro WRAPPER_IMPL_SSE2_fFF callee |
|---|
| 86 | pushq %rbp |
|---|
| 87 | cfi_adjust_cfa_offset (8) |
|---|
| 88 | cfi_rel_offset (%rbp, 0) |
|---|
| 89 | pushq %rbx |
|---|
| 90 | cfi_adjust_cfa_offset (8) |
|---|
| 91 | cfi_rel_offset (%rbx, 0) |
|---|
| 92 | movq %rdi, %rbp |
|---|
| 93 | movq %rsi, %rbx |
|---|
| 94 | subq $24, %rsp |
|---|
| 95 | cfi_adjust_cfa_offset (24) |
|---|
| 96 | movaps %xmm0, (%rsp) |
|---|
| 97 | call JUMPTARGET(\callee) |
|---|
| 98 | movss 4(%rsp), %xmm0 |
|---|
| 99 | leaq 4(%rbp), %rdi |
|---|
| 100 | leaq 4(%rbx), %rsi |
|---|
| 101 | call JUMPTARGET(\callee) |
|---|
| 102 | movss 8(%rsp), %xmm0 |
|---|
| 103 | leaq 8(%rbp), %rdi |
|---|
| 104 | leaq 8(%rbx), %rsi |
|---|
| 105 | call JUMPTARGET(\callee) |
|---|
| 106 | movss 12(%rsp), %xmm0 |
|---|
| 107 | leaq 12(%rbp), %rdi |
|---|
| 108 | leaq 12(%rbx), %rsi |
|---|
| 109 | call JUMPTARGET(\callee) |
|---|
| 110 | addq $24, %rsp |
|---|
| 111 | cfi_adjust_cfa_offset (-24) |
|---|
| 112 | popq %rbx |
|---|
| 113 | cfi_adjust_cfa_offset (-8) |
|---|
| 114 | cfi_restore (%rbx) |
|---|
| 115 | popq %rbp |
|---|
| 116 | cfi_adjust_cfa_offset (-8) |
|---|
| 117 | cfi_restore (%rbp) |
|---|
| 118 | ret |
|---|
| 119 | .endm |
|---|
| 120 | |
|---|
| 121 | #include "svml_sd_wrapper_impl.h" |
|---|
| 122 | |
|---|
