diff options
Diffstat (limited to 'sysdeps/x86_64/fpu/svml_s_sincosf4_core.S')
| -rw-r--r-- | sysdeps/x86_64/fpu/svml_s_sincosf4_core.S | 128 |
1 files changed, 125 insertions, 3 deletions
diff --git a/sysdeps/x86_64/fpu/svml_s_sincosf4_core.S b/sysdeps/x86_64/fpu/svml_s_sincosf4_core.S index 1a7d2733af..5daa5118d6 100644 --- a/sysdeps/x86_64/fpu/svml_s_sincosf4_core.S +++ b/sysdeps/x86_64/fpu/svml_s_sincosf4_core.S @@ -1,5 +1,5 @@ /* Function sincosf vectorized with SSE2. - Copyright (C) 2014-2016 Free Software Foundation, Inc. + Copyright (C) 2014-2018 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 @@ -16,13 +16,135 @@ License along with the GNU C Library; if not, see <http://www.gnu.org/licenses/>. */ - #include <sysdep.h> #include "svml_s_wrapper_impl.h" .text -ENTRY (_ZGVbN4vvv_sincosf) +ENTRY (_ZGVbN4vl4l4_sincosf) WRAPPER_IMPL_SSE2_fFF sincosf +END (_ZGVbN4vl4l4_sincosf) +libmvec_hidden_def (_ZGVbN4vl4l4_sincosf) + +/* SSE2 ISA version as wrapper to scalar (for vector + function declared with #pragma omp declare simd notinbranch). */ +.macro WRAPPER_IMPL_SSE2_fFF_vvv callee +#ifndef __ILP32__ + subq $120, %rsp + cfi_adjust_cfa_offset(120) + movaps %xmm0, 96(%rsp) + lea (%rsp), %rdi + movdqa %xmm1, 32(%rdi) + lea 16(%rsp), %rsi + movdqa %xmm2, 32(%rsi) + movdqa %xmm3, 48(%rsi) + movdqa %xmm4, 64(%rsi) + call JUMPTARGET(\callee) + movss 100(%rsp), %xmm0 + lea 4(%rsp), %rdi + lea 20(%rsp), %rsi + call JUMPTARGET(\callee) + movss 104(%rsp), %xmm0 + lea 8(%rsp), %rdi + lea 24(%rsp), %rsi + call JUMPTARGET(\callee) + movss 108(%rsp), %xmm0 + lea 12(%rsp), %rdi + lea 28(%rsp), %rsi + call JUMPTARGET(\callee) + movq 32(%rsp), %rdx + movq 40(%rsp), %rsi + movq 48(%rsp), %r8 + movq 56(%rsp), %r10 + movl (%rsp), %eax + movl 4(%rsp), %ecx + movl 8(%rsp), %edi + movl 12(%rsp), %r9d + movl %eax, (%rdx) + movl %ecx, (%rsi) + movq 64(%rsp), %rax + movq 72(%rsp), %rcx + movl %edi, (%r8) + movl %r9d, (%r10) + movq 80(%rsp), %rdi + movq 88(%rsp), %r9 + movl 16(%rsp), %r11d + movl 20(%rsp), %edx + movl 24(%rsp), %esi + movl 28(%rsp), %r8d + movl %r11d, (%rax) + movl %edx, (%rcx) + movl %esi, (%rdi) + movl %r8d, (%r9) + addq $120, %rsp + cfi_adjust_cfa_offset(-120) + ret +#else + pushq %rbp + .cfi_def_cfa_offset 16 + .cfi_offset 6, -16 + pushq %rbx + .cfi_def_cfa_offset 24 + .cfi_offset 3, -24 + subl $88, %esp + .cfi_def_cfa_offset 112 + leal 64(%rsp), %esi + movaps %xmm1, (%esp) + leal 48(%rsp), %edi + movaps %xmm2, 16(%esp) + movq %rsi, %rbp + movq %rdi, %rbx + movaps %xmm0, 32(%esp) + call JUMPTARGET(\callee) + movups 36(%esp), %xmm0 + leal 4(%rbp), %esi + leal 4(%rbx), %edi + call JUMPTARGET(\callee) + movups 40(%esp), %xmm0 + leal 8(%rbp), %esi + leal 8(%rbx), %edi + call JUMPTARGET(\callee) + movups 44(%esp), %xmm0 + leal 12(%rbp), %esi + leal 12(%rbx), %edi + call JUMPTARGET(\callee) + movq (%esp), %rax + movss 48(%esp), %xmm0 + movdqa (%esp), %xmm4 + movdqa 16(%esp), %xmm7 + movss %xmm0, (%eax) + movss 52(%esp), %xmm0 + pextrd $1, %xmm4, %eax + movss %xmm0, (%eax) + movq 8(%esp), %rax + movss 56(%esp), %xmm0 + movss %xmm0, (%eax) + movss 60(%esp), %xmm0 + pextrd $3, %xmm4, %eax + movss %xmm0, (%eax) + movq 16(%esp), %rax + movss 64(%esp), %xmm0 + movss %xmm0, (%eax) + movss 68(%esp), %xmm0 + pextrd $1, %xmm7, %eax + movss %xmm0, (%eax) + movq 24(%esp), %rax + movss 72(%esp), %xmm0 + movss %xmm0, (%eax) + movss 76(%esp), %xmm0 + pextrd $3, %xmm7, %eax + movss %xmm0, (%eax) + addl $88, %esp + .cfi_def_cfa_offset 24 + popq %rbx + .cfi_def_cfa_offset 16 + popq %rbp + .cfi_def_cfa_offset 8 + ret +#endif +.endm + +ENTRY (_ZGVbN4vvv_sincosf) +WRAPPER_IMPL_SSE2_fFF_vvv sincosf END (_ZGVbN4vvv_sincosf) #ifndef USE_MULTIARCH |