diff options
author | Andrew Senkevich <andrew.senkevich@intel.com> | 2015-06-09 18:29:47 +0300 |
---|---|---|
committer | Andrew Senkevich <andrew.senkevich@intel.com> | 2015-06-09 18:29:47 +0300 |
commit | 04f496d6025753058bdd071fd711e9f56df149a7 (patch) | |
tree | 21b047c3aadb643a0eb86e277e496146bd9b5f84 /sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S | |
parent | 24a2718f595bc11dc6abb31303ceb8fdcb664f2f (diff) | |
download | glibc-04f496d6025753058bdd071fd711e9f56df149a7.zip glibc-04f496d6025753058bdd071fd711e9f56df149a7.tar.gz glibc-04f496d6025753058bdd071fd711e9f56df149a7.tar.bz2 |
Vector cosf for x86_64.
Here is implementation of vectorized cosf containing SSE, AVX,
AVX2 and AVX512 versions according to Vector ABI
<https://groups.google.com/forum/#!topic/x86-64-abi/LmppCfN1rZ4>.
* sysdeps/x86_64/fpu/Makefile (libmvec-support): Added new files.
* sysdeps/x86_64/fpu/Versions: New versions added.
* sysdeps/x86_64/fpu/svml_s_cosf4_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S: New file.
* sysdeps/x86_64/fpu/svml_s_cosf8_core_avx.S: New file.
* sysdeps/x86_64/fpu/svml_s_cosf8_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf8_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf8_core_avx2.S: New file.
* sysdeps/x86_64/fpu/svml_s_cosf16_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core.S: New file.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S: New file.
* sysdeps/x86_64/fpu/svml_s_wrapper_impl.h: New file.
* sysdeps/x86_64/fpu/svml_s_cosf_data.S: New file.
* sysdeps/x86_64/fpu/svml_s_cosf_data.h: New file.
* sysdeps/x86_64/fpu/multiarch/Makefile (libmvec-sysdep_routines): Added
build of SSE, AVX2 and AVX512 IFUNC versions.
* sysdeps/unix/sysv/linux/x86_64/libmvec.abilist: New versions added.
* sysdeps/x86/fpu/bits/math-vector.h: Added SIMD declaration for cosf.
* NEWS: Mention addition of x86_64 vector cosf.
Diffstat (limited to 'sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S')
-rw-r--r-- | sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S | 227 |
1 files changed, 227 insertions, 0 deletions
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S b/sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S new file mode 100644 index 0000000..f231ba2 --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_cosf4_core_sse4.S @@ -0,0 +1,227 @@ +/* Function cosf vectorized with SSE4. + Copyright (C) 2014-2015 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 + <http://www.gnu.org/licenses/>. */ + +#include <sysdep.h> +#include "svml_s_cosf_data.h" + + .text +ENTRY (_ZGVbN4v_cosf_sse4) +/* + ALGORITHM DESCRIPTION: + + 1) Range reduction to [-Pi/2; +Pi/2] interval + a) We remove sign using AND operation + b) Add Pi/2 value to argument X for Cos to Sin transformation + c) Getting octant Y by 1/Pi multiplication + d) Add "Right Shifter" value + e) Treat obtained value as integer for destination sign setting. + Shift first bit of this value to the last (sign) position + f) Subtract "Right Shifter" value + g) Subtract 0.5 from result for octant correction + h) Subtract Y*PI from X argument, where PI divided to 4 parts: + X = X - Y*PI1 - Y*PI2 - Y*PI3 - Y*PI4; + 2) Polynomial (minimax for sin within [-Pi/2; +Pi/2] interval) + a) Calculate X^2 = X * X + b) Calculate polynomial: + R = X + X * X^2 * (A3 + x^2 * (A5 + ..... + 3) Destination sign setting + a) Set shifted destination sign using XOR operation: + R = XOR( R, S ); + */ + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $320, %rsp + movaps %xmm0, %xmm4 + movq __svml_scos_data@GOTPCREL(%rip), %rax + movups __sHalfPI(%rax), %xmm1 + movups __sRShifter(%rax), %xmm5 + +/* b) Add Pi/2 value to argument X for Cos to Sin transformation */ + addps %xmm4, %xmm1 + +/* + 1) Range reduction to [-Pi/2; +Pi/2] interval + c) Getting octant Y by 1/Pi multiplication + d) Add "Right Shifter" (0x4B000000) value + */ + mulps __sInvPI(%rax), %xmm1 + movups __sPI1(%rax), %xmm6 + addps %xmm5, %xmm1 + +/* + e) Treat obtained value as integer for destination sign setting. + Shift first bit of this value to the last (sign) position (S << 31) + */ + movaps %xmm1, %xmm2 + +/* f) Subtract "Right Shifter" (0x4B000000) value */ + subps %xmm5, %xmm1 + movups __sPI2(%rax), %xmm7 + pslld $31, %xmm2 + movups __sPI3(%rax), %xmm5 + movups __sAbsMask(%rax), %xmm3 + +/* Check for large and special arguments */ + andps %xmm4, %xmm3 + +/* g) Subtract 0.5 from result for octant correction */ + subps __sOneHalf(%rax), %xmm1 + cmpnleps __sRangeReductionVal(%rax), %xmm3 + +/* + h) Subtract Y*PI from X argument, where PI divided to 4 parts: + X = X - Y*PI1 - Y*PI2 - Y*PI3 - Y*PI4; + */ + mulps %xmm1, %xmm6 + mulps %xmm1, %xmm7 + mulps %xmm1, %xmm5 + subps %xmm6, %xmm0 + movmskps %xmm3, %ecx + movups __sPI4(%rax), %xmm6 + subps %xmm7, %xmm0 + mulps %xmm6, %xmm1 + subps %xmm5, %xmm0 + subps %xmm1, %xmm0 + +/* a) Calculate X^2 = X * X */ + movaps %xmm0, %xmm1 + mulps %xmm0, %xmm1 + +/* + 3) Destination sign setting + a) Set shifted destination sign using XOR operation: + R = XOR( R, S ); + */ + xorps %xmm2, %xmm0 + movups __sA9(%rax), %xmm2 + +/* + b) Calculate polynomial: + R = X + X * X^2 * (A3 + x^2 * (A5 + x^2 * (A7 + x^2 * (A9)))); + */ + mulps %xmm1, %xmm2 + addps __sA7(%rax), %xmm2 + mulps %xmm1, %xmm2 + addps __sA5(%rax), %xmm2 + mulps %xmm1, %xmm2 + addps __sA3(%rax), %xmm2 + mulps %xmm2, %xmm1 + mulps %xmm0, %xmm1 + addps %xmm1, %xmm0 + testl %ecx, %ecx + jne .LBL_1_3 + +.LBL_1_2: + cfi_remember_state + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret + +.LBL_1_3: + cfi_restore_state + movups %xmm4, 192(%rsp) + movups %xmm0, 256(%rsp) + je .LBL_1_2 + + xorb %dl, %dl + xorl %eax, %eax + movups %xmm8, 112(%rsp) + movups %xmm9, 96(%rsp) + movups %xmm10, 80(%rsp) + movups %xmm11, 64(%rsp) + movups %xmm12, 48(%rsp) + movups %xmm13, 32(%rsp) + movups %xmm14, 16(%rsp) + movups %xmm15, (%rsp) + movq %rsi, 136(%rsp) + movq %rdi, 128(%rsp) + movq %r12, 168(%rsp) + cfi_offset_rel_rsp (12, 168) + movb %dl, %r12b + movq %r13, 160(%rsp) + cfi_offset_rel_rsp (13, 160) + movl %ecx, %r13d + movq %r14, 152(%rsp) + cfi_offset_rel_rsp (14, 152) + movl %eax, %r14d + movq %r15, 144(%rsp) + cfi_offset_rel_rsp (15, 144) + cfi_remember_state + +.LBL_1_6: + btl %r14d, %r13d + jc .LBL_1_12 + +.LBL_1_7: + lea 1(%r14), %esi + btl %esi, %r13d + jc .LBL_1_10 + +.LBL_1_8: + incb %r12b + addl $2, %r14d + cmpb $16, %r12b + jb .LBL_1_6 + + movups 112(%rsp), %xmm8 + movups 96(%rsp), %xmm9 + movups 80(%rsp), %xmm10 + movups 64(%rsp), %xmm11 + movups 48(%rsp), %xmm12 + movups 32(%rsp), %xmm13 + movups 16(%rsp), %xmm14 + movups (%rsp), %xmm15 + movq 136(%rsp), %rsi + movq 128(%rsp), %rdi + movq 168(%rsp), %r12 + cfi_restore (%r12) + movq 160(%rsp), %r13 + cfi_restore (%r13) + movq 152(%rsp), %r14 + cfi_restore (%r14) + movq 144(%rsp), %r15 + cfi_restore (%r15) + movups 256(%rsp), %xmm0 + jmp .LBL_1_2 + +.LBL_1_10: + cfi_restore_state + movzbl %r12b, %r15d + movss 196(%rsp,%r15,8), %xmm0 + + call cosf@PLT + + movss %xmm0, 260(%rsp,%r15,8) + jmp .LBL_1_8 + +.LBL_1_12: + movzbl %r12b, %r15d + movss 192(%rsp,%r15,8), %xmm0 + + call cosf@PLT + + movss %xmm0, 256(%rsp,%r15,8) + jmp .LBL_1_7 +END (_ZGVbN4v_cosf_sse4) |