From 774488f88aeed6b838fe29c3c7561433c242a3c9 Mon Sep 17 00:00:00 2001 From: Andrew Senkevich Date: Wed, 17 Jun 2015 15:53:00 +0300 Subject: [PATCH] Vector logf for x86_64 and tests. Here is implementation of vectorized logf containing SSE, AVX, AVX2 and AVX512 versions according to Vector ABI . * sysdeps/unix/sysv/linux/x86_64/libmvec.abilist: New symbols added. * sysdeps/x86/fpu/bits/math-vector.h: Added SIMD declaration and asm redirections for logf. * sysdeps/x86_64/fpu/Makefile (libmvec-support): Added new files. * sysdeps/x86_64/fpu/Versions: New versions added. * sysdeps/x86_64/fpu/libm-test-ulps: Regenerated. * sysdeps/x86_64/fpu/multiarch/Makefile (libmvec-sysdep_routines): Added build of SSE, AVX2 and AVX512 IFUNC versions. * sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S: New file. * sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S: New file. * sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S: New file. * sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S: New file. * sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S: New file. * sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S: New file. * sysdeps/x86_64/fpu/svml_s_logf16_core.S: New file. * sysdeps/x86_64/fpu/svml_s_logf4_core.S: New file. * sysdeps/x86_64/fpu/svml_s_logf8_core.S: New file. * sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S: New file. * sysdeps/x86_64/fpu/svml_s_logf_data.S: New file. * sysdeps/x86_64/fpu/svml_s_logf_data.h: New file. * sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c: Vector logf tests. * sysdeps/x86_64/fpu/test-float-vlen16.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen4.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen8-avx2.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c: Likewise. * sysdeps/x86_64/fpu/test-float-vlen8.c: Likewise. * NEWS: Mention addition of x86_64 vector logf. --- ChangeLog | 30 ++ NEWS | 2 +- .../unix/sysv/linux/x86_64/libmvec.abilist | 4 + sysdeps/x86/fpu/bits/math-vector.h | 6 + sysdeps/x86_64/fpu/Makefile | 4 +- sysdeps/x86_64/fpu/Versions | 1 + sysdeps/x86_64/fpu/libm-test-ulps | 8 + sysdeps/x86_64/fpu/multiarch/Makefile | 4 +- .../x86_64/fpu/multiarch/svml_s_logf16_core.S | 39 ++ .../fpu/multiarch/svml_s_logf16_core_avx512.S | 416 ++++++++++++++++++ .../x86_64/fpu/multiarch/svml_s_logf4_core.S | 38 ++ .../fpu/multiarch/svml_s_logf4_core_sse4.S | 194 ++++++++ .../x86_64/fpu/multiarch/svml_s_logf8_core.S | 38 ++ .../fpu/multiarch/svml_s_logf8_core_avx2.S | 184 ++++++++ sysdeps/x86_64/fpu/svml_s_logf16_core.S | 25 ++ sysdeps/x86_64/fpu/svml_s_logf4_core.S | 30 ++ sysdeps/x86_64/fpu/svml_s_logf8_core.S | 29 ++ sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S | 25 ++ sysdeps/x86_64/fpu/svml_s_logf_data.S | 102 +++++ sysdeps/x86_64/fpu/svml_s_logf_data.h | 48 ++ .../x86_64/fpu/test-float-vlen16-wrappers.c | 1 + sysdeps/x86_64/fpu/test-float-vlen16.c | 1 + .../x86_64/fpu/test-float-vlen4-wrappers.c | 1 + sysdeps/x86_64/fpu/test-float-vlen4.c | 1 + .../fpu/test-float-vlen8-avx2-wrappers.c | 1 + sysdeps/x86_64/fpu/test-float-vlen8-avx2.c | 1 + .../x86_64/fpu/test-float-vlen8-wrappers.c | 1 + sysdeps/x86_64/fpu/test-float-vlen8.c | 1 + 28 files changed, 1232 insertions(+), 3 deletions(-) create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S create mode 100644 sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf16_core.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf4_core.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf8_core.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf_data.S create mode 100644 sysdeps/x86_64/fpu/svml_s_logf_data.h diff --git a/ChangeLog b/ChangeLog index bad022e036..a0c2a9a0fb 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,35 @@ 2015-06-17 Andrew Senkevich + * sysdeps/unix/sysv/linux/x86_64/libmvec.abilist: New symbols added. + * sysdeps/x86/fpu/bits/math-vector.h: Added SIMD declaration and asm + redirections for logf. + * sysdeps/x86_64/fpu/Makefile (libmvec-support): Added new files. + * sysdeps/x86_64/fpu/Versions: New versions added. + * sysdeps/x86_64/fpu/libm-test-ulps: Regenerated. + * sysdeps/x86_64/fpu/multiarch/Makefile (libmvec-sysdep_routines): Added + build of SSE, AVX2 and AVX512 IFUNC versions. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S: New file. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S: New file. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S: New file. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S: New file. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S: New file. + * sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf16_core.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf4_core.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf8_core.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf_data.S: New file. + * sysdeps/x86_64/fpu/svml_s_logf_data.h: New file. + * sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c: Vector logf tests. + * sysdeps/x86_64/fpu/test-float-vlen16.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen4.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen8-avx2.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c: Likewise. + * sysdeps/x86_64/fpu/test-float-vlen8.c: Likewise. + * NEWS: Mention addition of x86_64 vector logf. + * bits/libm-simd-decl-stubs.h: Added stubs for log. * math/bits/mathcalls.h: Added log declaration with __MATHCALL_VEC. * sysdeps/unix/sysv/linux/x86_64/libmvec.abilist: New versions added. diff --git a/NEWS b/NEWS index 4c666878dd..c361309b4c 100644 --- a/NEWS +++ b/NEWS @@ -53,7 +53,7 @@ Version 2.22 condition in some applications. * Added vector math library named libmvec with the following vectorized x86_64 - implementations: cos, cosf, sin, sinf, log. + implementations: cos, cosf, sin, sinf, log, logf. The library can be disabled with --disable-mathvec. Use of the functions is enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0. The library is linked in as needed when using -lm (no need to specify -lmvec diff --git a/sysdeps/unix/sysv/linux/x86_64/libmvec.abilist b/sysdeps/unix/sysv/linux/x86_64/libmvec.abilist index 3357957d3d..3593edcbfa 100644 --- a/sysdeps/unix/sysv/linux/x86_64/libmvec.abilist +++ b/sysdeps/unix/sysv/linux/x86_64/libmvec.abilist @@ -4,18 +4,22 @@ GLIBC_2.22 _ZGVbN2v_log F _ZGVbN2v_sin F _ZGVbN4v_cosf F + _ZGVbN4v_logf F _ZGVbN4v_sinf F _ZGVcN4v_cos F _ZGVcN4v_log F _ZGVcN4v_sin F _ZGVcN8v_cosf F + _ZGVcN8v_logf F _ZGVcN8v_sinf F _ZGVdN4v_cos F _ZGVdN4v_log F _ZGVdN4v_sin F _ZGVdN8v_cosf F + _ZGVdN8v_logf F _ZGVdN8v_sinf F _ZGVeN16v_cosf F + _ZGVeN16v_logf F _ZGVeN16v_sinf F _ZGVeN8v_cos F _ZGVeN8v_log F diff --git a/sysdeps/x86/fpu/bits/math-vector.h b/sysdeps/x86/fpu/bits/math-vector.h index ed85622560..5c3e492ef9 100644 --- a/sysdeps/x86/fpu/bits/math-vector.h +++ b/sysdeps/x86/fpu/bits/math-vector.h @@ -38,6 +38,8 @@ # define __DECL_SIMD_sinf __DECL_SIMD_x86_64 # undef __DECL_SIMD_log # define __DECL_SIMD_log __DECL_SIMD_x86_64 +# undef __DECL_SIMD_logf +# define __DECL_SIMD_logf __DECL_SIMD_x86_64 /* Workaround to exclude unnecessary symbol aliases in libmvec while GCC creates the vector names based on scalar asm name. @@ -47,6 +49,10 @@ __asm__ ("_ZGVbN2v___log_finite = _ZGVbN2v_log"); __asm__ ("_ZGVcN4v___log_finite = _ZGVcN4v_log"); __asm__ ("_ZGVdN4v___log_finite = _ZGVdN4v_log"); __asm__ ("_ZGVeN8v___log_finite = _ZGVeN8v_log"); +__asm__ ("_ZGVbN4v___logf_finite = _ZGVbN4v_logf"); +__asm__ ("_ZGVcN8v___logf_finite = _ZGVcN8v_logf"); +__asm__ ("_ZGVdN8v___logf_finite = _ZGVdN8v_logf"); +__asm__ ("_ZGVeN16v___logf_finite = _ZGVeN16v_logf"); # endif #endif diff --git a/sysdeps/x86_64/fpu/Makefile b/sysdeps/x86_64/fpu/Makefile index a5097468ee..b610e3faf5 100644 --- a/sysdeps/x86_64/fpu/Makefile +++ b/sysdeps/x86_64/fpu/Makefile @@ -8,7 +8,9 @@ libmvec-support += svml_d_cos2_core svml_d_cos4_core_avx \ svml_s_sinf4_core svml_s_sinf8_core_avx \ svml_s_sinf8_core svml_s_sinf16_core svml_s_sinf_data \ svml_d_log2_core svml_d_log4_core_avx svml_d_log4_core \ - svml_d_log8_core svml_d_log_data \ + svml_d_log8_core svml_d_log_data svml_s_logf4_core \ + svml_s_logf8_core_avx svml_s_logf8_core svml_s_logf16_core \ + svml_s_logf_data \ init-arch endif diff --git a/sysdeps/x86_64/fpu/Versions b/sysdeps/x86_64/fpu/Versions index 7bda47f4af..ecd1b7023e 100644 --- a/sysdeps/x86_64/fpu/Versions +++ b/sysdeps/x86_64/fpu/Versions @@ -5,5 +5,6 @@ libmvec { _ZGVbN2v_log; _ZGVcN4v_log; _ZGVdN4v_log; _ZGVeN8v_log; _ZGVbN4v_cosf; _ZGVcN8v_cosf; _ZGVdN8v_cosf; _ZGVeN16v_cosf; _ZGVbN4v_sinf; _ZGVcN8v_sinf; _ZGVdN8v_sinf; _ZGVeN16v_sinf; + _ZGVbN4v_logf; _ZGVcN8v_logf; _ZGVdN8v_logf; _ZGVeN16v_logf; } } diff --git a/sysdeps/x86_64/fpu/libm-test-ulps b/sysdeps/x86_64/fpu/libm-test-ulps index 949a099090..1812370ecc 100644 --- a/sysdeps/x86_64/fpu/libm-test-ulps +++ b/sysdeps/x86_64/fpu/libm-test-ulps @@ -1847,17 +1847,25 @@ ifloat: 2 ildouble: 1 ldouble: 1 +Function: "log_vlen16": +float: 3 + Function: "log_vlen2": double: 1 Function: "log_vlen4": double: 1 +float: 3 Function: "log_vlen4_avx2": double: 1 Function: "log_vlen8": double: 1 +float: 3 + +Function: "log_vlen8_avx2": +float: 2 Function: "pow": float: 3 diff --git a/sysdeps/x86_64/fpu/multiarch/Makefile b/sysdeps/x86_64/fpu/multiarch/Makefile index 16d93caf57..5fc6ea3d23 100644 --- a/sysdeps/x86_64/fpu/multiarch/Makefile +++ b/sysdeps/x86_64/fpu/multiarch/Makefile @@ -60,5 +60,7 @@ libmvec-sysdep_routines += svml_d_cos2_core_sse4 svml_d_cos4_core_avx2 \ svml_d_log8_core_avx512 \ svml_s_cosf4_core_sse4 svml_s_cosf8_core_avx2 \ svml_s_cosf16_core_avx512 svml_s_sinf4_core_sse4 \ - svml_s_sinf8_core_avx2 svml_s_sinf16_core_avx512 + svml_s_sinf8_core_avx2 svml_s_sinf16_core_avx512 \ + svml_s_logf4_core_sse4 svml_s_logf8_core_avx2 \ + svml_s_logf16_core_avx512 endif diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S new file mode 100644 index 0000000000..8756750c86 --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core.S @@ -0,0 +1,39 @@ +/* Multiple versions of vectorized logf. + 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 + . */ + +#include +#include + + .text +ENTRY (_ZGVeN16v_logf) + .type _ZGVeN16v_logf, @gnu_indirect_function + cmpl $0, KIND_OFFSET+__cpu_features(%rip) + jne 1 + call __init_cpu_features +1: leaq _ZGVeN16v_logf_skx(%rip), %rax + testl $bit_AVX512DQ_Usable, __cpu_features+FEATURE_OFFSET+index_AVX512DQ_Usable(%rip) + jnz 3 +2: leaq _ZGVeN16v_logf_knl(%rip), %rax + testl $bit_AVX512F_Usable, __cpu_features+FEATURE_OFFSET+index_AVX512F_Usable(%rip) + jnz 3 + leaq _ZGVeN16v_logf_avx2_wrapper(%rip), %rax +3: ret +END (_ZGVeN16v_logf) + +#define _ZGVeN16v_logf _ZGVeN16v_logf_avx2_wrapper +#include "../svml_s_logf16_core.S" diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S new file mode 100644 index 0000000000..86fcab6e63 --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S @@ -0,0 +1,416 @@ +/* Function logf vectorized with AVX-512. KNL and SKX versions. + 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 + . */ + +#include +#include "svml_s_logf_data.h" +#include "svml_s_wrapper_impl.h" + + .text +ENTRY (_ZGVeN16v_logf_knl) +#ifndef HAVE_AVX512_ASM_SUPPORT +WRAPPER_IMPL_AVX512 _ZGVdN8v_logf +#else +/* + ALGORITHM DESCRIPTION: + + log(x) = exponent_x*log(2) + log(mantissa_x), if mantissa_x<4/3 + log(x) = (exponent_x+1)*log(2) + log(0.5*mantissa_x), if mantissa_x>4/3 + + R = mantissa_x - 1, if mantissa_x<4/3 + R = 0.5*mantissa_x - 1, if mantissa_x>4/3 + |R|< 1/3 + + log(1+R) is approximated as a polynomial: degree 9 for 1-ulp, + degree 7 for 4-ulp, degree 3 for half-precision. */ + + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $1280, %rsp + movq __svml_slog_data@GOTPCREL(%rip), %rax + movl $-1, %ecx + +/* reduction: compute r,n */ + vpsubd _iBrkValue(%rax), %zmm0, %zmm2 + vmovups _sPoly_7(%rax), %zmm7 + vpandd _iOffExpoMask(%rax), %zmm2, %zmm3 + +/* exponent_x (mantissa_x<4/3) or exponent_x+1 (mantissa_x>4/3) */ + vpsrad $23, %zmm2, %zmm4 + +/* check for working range, + set special argument mask (denormals/zero/Inf/NaN) + */ + vpaddd _iHiDelta(%rax), %zmm0, %zmm1 + +/* mantissa_x (mantissa_x<4/3), or 0.5*mantissa_x (mantissa_x>4/3) */ + vpaddd _iBrkValue(%rax), %zmm3, %zmm6 + vpcmpd $1, _iLoRange(%rax), %zmm1, %k1 + vcvtdq2ps {rn-sae}, %zmm4, %zmm1 + +/* reduced argument R */ + vsubps _sOne(%rax), %zmm6, %zmm8 + vpbroadcastd %ecx, %zmm5{%k1}{z} + +/* polynomial evaluation starts here */ + vfmadd213ps _sPoly_6(%rax), %zmm8, %zmm7 + vptestmd %zmm5, %zmm5, %k0 + kmovw %k0, %ecx + vfmadd213ps _sPoly_5(%rax), %zmm8, %zmm7 + vfmadd213ps _sPoly_4(%rax), %zmm8, %zmm7 + vfmadd213ps _sPoly_3(%rax), %zmm8, %zmm7 + vfmadd213ps _sPoly_2(%rax), %zmm8, %zmm7 + vfmadd213ps _sPoly_1(%rax), %zmm8, %zmm7 + vmulps %zmm8, %zmm7, %zmm9 + +/* polynomial evaluation end */ + vfmadd213ps %zmm8, %zmm8, %zmm9 + +/* + final reconstruction: + add exponent_value*log2 to polynomial result + */ + vfmadd132ps _sLn2(%rax), %zmm9, %zmm1 + testl %ecx, %ecx + jne .LBL_1_3 + +.LBL_1_2: + cfi_remember_state + vmovaps %zmm1, %zmm0 + 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 + vmovups %zmm0, 1152(%rsp) + vmovups %zmm1, 1216(%rsp) + je .LBL_1_2 + + xorb %dl, %dl + kmovw %k4, 1048(%rsp) + xorl %eax, %eax + kmovw %k5, 1040(%rsp) + kmovw %k6, 1032(%rsp) + kmovw %k7, 1024(%rsp) + vmovups %zmm16, 960(%rsp) + vmovups %zmm17, 896(%rsp) + vmovups %zmm18, 832(%rsp) + vmovups %zmm19, 768(%rsp) + vmovups %zmm20, 704(%rsp) + vmovups %zmm21, 640(%rsp) + vmovups %zmm22, 576(%rsp) + vmovups %zmm23, 512(%rsp) + vmovups %zmm24, 448(%rsp) + vmovups %zmm25, 384(%rsp) + vmovups %zmm26, 320(%rsp) + vmovups %zmm27, 256(%rsp) + vmovups %zmm28, 192(%rsp) + vmovups %zmm29, 128(%rsp) + vmovups %zmm30, 64(%rsp) + vmovups %zmm31, (%rsp) + movq %rsi, 1064(%rsp) + movq %rdi, 1056(%rsp) + movq %r12, 1096(%rsp) + cfi_offset_rel_rsp (12, 1096) + movb %dl, %r12b + movq %r13, 1088(%rsp) + cfi_offset_rel_rsp (13, 1088) + movl %ecx, %r13d + movq %r14, 1080(%rsp) + cfi_offset_rel_rsp (14, 1080) + movl %eax, %r14d + movq %r15, 1072(%rsp) + cfi_offset_rel_rsp (15, 1072) + 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: + addb $1, %r12b + addl $2, %r14d + cmpb $16, %r12b + jb .LBL_1_6 + + kmovw 1048(%rsp), %k4 + movq 1064(%rsp), %rsi + kmovw 1040(%rsp), %k5 + movq 1056(%rsp), %rdi + kmovw 1032(%rsp), %k6 + movq 1096(%rsp), %r12 + cfi_restore (%r12) + movq 1088(%rsp), %r13 + cfi_restore (%r13) + kmovw 1024(%rsp), %k7 + vmovups 960(%rsp), %zmm16 + vmovups 896(%rsp), %zmm17 + vmovups 832(%rsp), %zmm18 + vmovups 768(%rsp), %zmm19 + vmovups 704(%rsp), %zmm20 + vmovups 640(%rsp), %zmm21 + vmovups 576(%rsp), %zmm22 + vmovups 512(%rsp), %zmm23 + vmovups 448(%rsp), %zmm24 + vmovups 384(%rsp), %zmm25 + vmovups 320(%rsp), %zmm26 + vmovups 256(%rsp), %zmm27 + vmovups 192(%rsp), %zmm28 + vmovups 128(%rsp), %zmm29 + vmovups 64(%rsp), %zmm30 + vmovups (%rsp), %zmm31 + movq 1080(%rsp), %r14 + cfi_restore (%r14) + movq 1072(%rsp), %r15 + cfi_restore (%r15) + vmovups 1216(%rsp), %zmm1 + jmp .LBL_1_2 + +.LBL_1_10: + cfi_restore_state + movzbl %r12b, %r15d + vmovss 1156(%rsp,%r15,8), %xmm0 + call logf@PLT + vmovss %xmm0, 1220(%rsp,%r15,8) + jmp .LBL_1_8 + +.LBL_1_12: + movzbl %r12b, %r15d + vmovss 1152(%rsp,%r15,8), %xmm0 + call logf@PLT + vmovss %xmm0, 1216(%rsp,%r15,8) + jmp .LBL_1_7 +#endif +END (_ZGVeN16v_logf_knl) + +ENTRY (_ZGVeN16v_logf_skx) +#ifndef HAVE_AVX512_ASM_SUPPORT +WRAPPER_IMPL_AVX512 _ZGVdN8v_logf +#else +/* + ALGORITHM DESCRIPTION: + + log(x) = exponent_x*log(2) + log(mantissa_x), if mantissa_x<4/3 + log(x) = (exponent_x+1)*log(2) + log(0.5*mantissa_x), if mantissa_x>4/3 + + R = mantissa_x - 1, if mantissa_x<4/3 + R = 0.5*mantissa_x - 1, if mantissa_x>4/3 + |R|< 1/3 + + log(1+R) is approximated as a polynomial: degree 9 for 1-ulp, + degree 7 for 4-ulp, degree 3 for half-precision. */ + + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $1280, %rsp + movq __svml_slog_data@GOTPCREL(%rip), %rax + vmovups .L_2il0floatpacket.7(%rip), %zmm6 + vmovups _iBrkValue(%rax), %zmm4 + vmovups _sPoly_7(%rax), %zmm8 + +/* + check for working range, + set special argument mask (denormals/zero/Inf/NaN) + */ + vpaddd _iHiDelta(%rax), %zmm0, %zmm1 + +/* reduction: compute r,n */ + vpsubd %zmm4, %zmm0, %zmm2 + vpcmpd $5, _iLoRange(%rax), %zmm1, %k1 + +/* exponent_x (mantissa_x<4/3) or exponent_x+1 (mantissa_x>4/3) */ + vpsrad $23, %zmm2, %zmm5 + vpandd _iOffExpoMask(%rax), %zmm2, %zmm3 + +/* mantissa_x (mantissa_x<4/3), or 0.5*mantissa_x (mantissa_x>4/3) */ + vpaddd %zmm4, %zmm3, %zmm7 + +/* reduced argument R */ + vsubps _sOne(%rax), %zmm7, %zmm9 + +/* polynomial evaluation starts here */ + vfmadd213ps _sPoly_6(%rax), %zmm9, %zmm8 + vfmadd213ps _sPoly_5(%rax), %zmm9, %zmm8 + vfmadd213ps _sPoly_4(%rax), %zmm9, %zmm8 + vfmadd213ps _sPoly_3(%rax), %zmm9, %zmm8 + vfmadd213ps _sPoly_2(%rax), %zmm9, %zmm8 + vfmadd213ps _sPoly_1(%rax), %zmm9, %zmm8 + vmulps %zmm9, %zmm8, %zmm10 + +/* polynomial evaluation end */ + vfmadd213ps %zmm9, %zmm9, %zmm10 + vpandnd %zmm1, %zmm1, %zmm6{%k1} + vptestmd %zmm6, %zmm6, %k0 + vcvtdq2ps {rn-sae}, %zmm5, %zmm1 + kmovw %k0, %ecx + +/* + final reconstruction: + add exponent_value*log2 to polynomial result + */ + vfmadd132ps _sLn2(%rax), %zmm10, %zmm1 + testl %ecx, %ecx + jne .LBL_2_3 + +.LBL_2_2: + cfi_remember_state + vmovaps %zmm1, %zmm0 + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret + +.LBL_2_3: + cfi_restore_state + vmovups %zmm0, 1152(%rsp) + vmovups %zmm1, 1216(%rsp) + je .LBL_2_2 + + xorb %dl, %dl + xorl %eax, %eax + kmovw %k4, 1048(%rsp) + kmovw %k5, 1040(%rsp) + kmovw %k6, 1032(%rsp) + kmovw %k7, 1024(%rsp) + vmovups %zmm16, 960(%rsp) + vmovups %zmm17, 896(%rsp) + vmovups %zmm18, 832(%rsp) + vmovups %zmm19, 768(%rsp) + vmovups %zmm20, 704(%rsp) + vmovups %zmm21, 640(%rsp) + vmovups %zmm22, 576(%rsp) + vmovups %zmm23, 512(%rsp) + vmovups %zmm24, 448(%rsp) + vmovups %zmm25, 384(%rsp) + vmovups %zmm26, 320(%rsp) + vmovups %zmm27, 256(%rsp) + vmovups %zmm28, 192(%rsp) + vmovups %zmm29, 128(%rsp) + vmovups %zmm30, 64(%rsp) + vmovups %zmm31, (%rsp) + movq %rsi, 1064(%rsp) + movq %rdi, 1056(%rsp) + movq %r12, 1096(%rsp) + cfi_offset_rel_rsp (12, 1096) + movb %dl, %r12b + movq %r13, 1088(%rsp) + cfi_offset_rel_rsp (13, 1088) + movl %ecx, %r13d + movq %r14, 1080(%rsp) + cfi_offset_rel_rsp (14, 1080) + movl %eax, %r14d + movq %r15, 1072(%rsp) + cfi_offset_rel_rsp (15, 1072) + cfi_remember_state + +.LBL_2_6: + btl %r14d, %r13d + jc .LBL_2_12 + +.LBL_2_7: + lea 1(%r14), %esi + btl %esi, %r13d + jc .LBL_2_10 + +.LBL_2_8: + incb %r12b + addl $2, %r14d + cmpb $16, %r12b + jb .LBL_2_6 + + kmovw 1048(%rsp), %k4 + kmovw 1040(%rsp), %k5 + kmovw 1032(%rsp), %k6 + kmovw 1024(%rsp), %k7 + vmovups 960(%rsp), %zmm16 + vmovups 896(%rsp), %zmm17 + vmovups 832(%rsp), %zmm18 + vmovups 768(%rsp), %zmm19 + vmovups 704(%rsp), %zmm20 + vmovups 640(%rsp), %zmm21 + vmovups 576(%rsp), %zmm22 + vmovups 512(%rsp), %zmm23 + vmovups 448(%rsp), %zmm24 + vmovups 384(%rsp), %zmm25 + vmovups 320(%rsp), %zmm26 + vmovups 256(%rsp), %zmm27 + vmovups 192(%rsp), %zmm28 + vmovups 128(%rsp), %zmm29 + vmovups 64(%rsp), %zmm30 + vmovups (%rsp), %zmm31 + vmovups 1216(%rsp), %zmm1 + movq 1064(%rsp), %rsi + movq 1056(%rsp), %rdi + movq 1096(%rsp), %r12 + cfi_restore (%r12) + movq 1088(%rsp), %r13 + cfi_restore (%r13) + movq 1080(%rsp), %r14 + cfi_restore (%r14) + movq 1072(%rsp), %r15 + cfi_restore (%r15) + jmp .LBL_2_2 + +.LBL_2_10: + cfi_restore_state + movzbl %r12b, %r15d + vmovss 1156(%rsp,%r15,8), %xmm0 + vzeroupper + vmovss 1156(%rsp,%r15,8), %xmm0 + + call logf@PLT + + vmovss %xmm0, 1220(%rsp,%r15,8) + jmp .LBL_2_8 + +.LBL_2_12: + movzbl %r12b, %r15d + vmovss 1152(%rsp,%r15,8), %xmm0 + vzeroupper + vmovss 1152(%rsp,%r15,8), %xmm0 + + call logf@PLT + + vmovss %xmm0, 1216(%rsp,%r15,8) + jmp .LBL_2_7 + +#endif +END (_ZGVeN16v_logf_skx) + + .section .rodata, "a" +.L_2il0floatpacket.7: + .long 0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff + .type .L_2il0floatpacket.7,@object diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S new file mode 100644 index 0000000000..153ed8ebc2 --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core.S @@ -0,0 +1,38 @@ +/* Multiple versions of vectorized logf. + 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 + . */ + +#include +#include + + .text +ENTRY (_ZGVbN4v_logf) + .type _ZGVbN4v_logf, @gnu_indirect_function + cmpl $0, KIND_OFFSET+__cpu_features(%rip) + jne 1f + call __init_cpu_features +1: leaq _ZGVbN4v_logf_sse4(%rip), %rax + testl $bit_SSE4_1, __cpu_features+CPUID_OFFSET+index_SSE4_1(%rip) + jz 2f + ret +2: leaq _ZGVbN4v_logf_sse2(%rip), %rax + ret +END (_ZGVbN4v_logf) +libmvec_hidden_def (_ZGVbN4v_logf) + +#define _ZGVbN4v_logf _ZGVbN4v_logf_sse2 +#include "../svml_s_logf4_core.S" diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S new file mode 100644 index 0000000000..68f11033d9 --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf4_core_sse4.S @@ -0,0 +1,194 @@ +/* Function logf 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 + . */ + +#include +#include "svml_s_logf_data.h" + + .text +ENTRY (_ZGVbN4v_logf_sse4) +/* + ALGORITHM DESCRIPTION: + + log(x) = exponent_x*log(2) + log(mantissa_x), if mantissa_x<4/3 + log(x) = (exponent_x+1)*log(2) + log(0.5*mantissa_x), if mantissa_x>4/3 + + R = mantissa_x - 1, if mantissa_x<4/3 + R = 0.5*mantissa_x - 1, if mantissa_x>4/3 + |R|< 1/3 + + log(1+R) is approximated as a polynomial: degree 9 for 1-ulp, + degree 7 for 4-ulp, degree 3 for half-precision. */ + + 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 + +/* reduction: compute r,n */ + movaps %xmm0, %xmm2 + +/* check for working range, + set special argument mask (denormals/zero/Inf/NaN) */ + movq __svml_slog_data@GOTPCREL(%rip), %rax + movdqu _iHiDelta(%rax), %xmm1 + movdqu _iLoRange(%rax), %xmm4 + paddd %xmm0, %xmm1 + movdqu _iBrkValue(%rax), %xmm3 + pcmpgtd %xmm1, %xmm4 + movdqu _iOffExpoMask(%rax), %xmm1 + psubd %xmm3, %xmm2 + pand %xmm2, %xmm1 + +/* exponent_x (mantissa_x<4/3) or exponent_x+1 (mantissa_x>4/3) */ + psrad $23, %xmm2 + paddd %xmm3, %xmm1 + movups _sPoly_7(%rax), %xmm5 + +/* mantissa_x (mantissa_x<4/3), or 0.5*mantissa_x (mantissa_x>4/3) */ + cvtdq2ps %xmm2, %xmm6 + +/* reduced argument R */ + subps _sOne(%rax), %xmm1 + movmskps %xmm4, %ecx + +/* final reconstruction: + add exponent_value*log2 to polynomial result */ + mulps _sLn2(%rax), %xmm6 + +/* polynomial evaluation starts here */ + mulps %xmm1, %xmm5 + addps _sPoly_6(%rax), %xmm5 + mulps %xmm1, %xmm5 + addps _sPoly_5(%rax), %xmm5 + mulps %xmm1, %xmm5 + addps _sPoly_4(%rax), %xmm5 + mulps %xmm1, %xmm5 + addps _sPoly_3(%rax), %xmm5 + mulps %xmm1, %xmm5 + addps _sPoly_2(%rax), %xmm5 + mulps %xmm1, %xmm5 + addps _sPoly_1(%rax), %xmm5 + mulps %xmm1, %xmm5 + +/* polynomial evaluation end */ + mulps %xmm1, %xmm5 + addps %xmm5, %xmm1 + addps %xmm6, %xmm1 + testl %ecx, %ecx + jne .LBL_1_3 + +.LBL_1_2: + cfi_remember_state + movdqa %xmm1, %xmm0 + 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 %xmm0, 192(%rsp) + movups %xmm1, 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), %xmm1 + jmp .LBL_1_2 + +.LBL_1_10: + cfi_restore_state + movzbl %r12b, %r15d + movss 196(%rsp,%r15,8), %xmm0 + + call logf@PLT + + movss %xmm0, 260(%rsp,%r15,8) + jmp .LBL_1_8 + +.LBL_1_12: + movzbl %r12b, %r15d + movss 192(%rsp,%r15,8), %xmm0 + + call logf@PLT + + movss %xmm0, 256(%rsp,%r15,8) + jmp .LBL_1_7 + +END (_ZGVbN4v_logf_sse4) diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S new file mode 100644 index 0000000000..6f50bf6bdb --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core.S @@ -0,0 +1,38 @@ +/* Multiple versions of vectorized logf. + 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 + . */ + +#include +#include + + .text +ENTRY (_ZGVdN8v_logf) + .type _ZGVdN8v_logf, @gnu_indirect_function + cmpl $0, KIND_OFFSET+__cpu_features(%rip) + jne 1f + call __init_cpu_features +1: leaq _ZGVdN8v_logf_avx2(%rip), %rax + testl $bit_AVX2_Usable, __cpu_features+FEATURE_OFFSET+index_AVX2_Usable(%rip) + jz 2f + ret +2: leaq _ZGVdN8v_logf_sse_wrapper(%rip), %rax + ret +END (_ZGVdN8v_logf) +libmvec_hidden_def (_ZGVdN8v_logf) + +#define _ZGVdN8v_logf _ZGVdN8v_logf_sse_wrapper +#include "../svml_s_logf8_core.S" diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S new file mode 100644 index 0000000000..1f08b4218a --- /dev/null +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf8_core_avx2.S @@ -0,0 +1,184 @@ +/* Function logf vectorized with AVX2. + 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 + . */ + +#include +#include "svml_s_logf_data.h" + + .text +ENTRY(_ZGVdN8v_logf_avx2) +/* + ALGORITHM DESCRIPTION: + + log(x) = exponent_x*log(2) + log(mantissa_x), if mantissa_x<4/3 + log(x) = (exponent_x+1)*log(2) + log(0.5*mantissa_x), if mantissa_x>4/3 + + R = mantissa_x - 1, if mantissa_x<4/3 + R = 0.5*mantissa_x - 1, if mantissa_x>4/3 + |R|< 1/3 + + log(1+R) is approximated as a polynomial: degree 9 for 1-ulp, + degree 7 for 4-ulp, degree 3 for half-precision. */ + + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $448, %rsp + movq __svml_slog_data@GOTPCREL(%rip), %rax + vmovaps %ymm0, %ymm2 + vmovups _iBrkValue(%rax), %ymm6 + vmovups _iLoRange(%rax), %ymm1 +/* check for working range, + set special argument mask (denormals/zero/Inf/NaN) */ + vpaddd _iHiDelta(%rax), %ymm2, %ymm7 + +/* reduction: compute r,n */ + vpsubd %ymm6, %ymm2, %ymm4 + +/* exponent_x (mantissa_x<4/3) or exponent_x+1 (mantissa_x>4/3) */ + vpsrad $23, %ymm4, %ymm3 + vpand _iOffExpoMask(%rax), %ymm4, %ymm5 + vmovups _sPoly_7(%rax), %ymm4 + vcvtdq2ps %ymm3, %ymm0 + +/* mantissa_x (mantissa_x<4/3), or 0.5*mantissa_x (mantissa_x>4/3) */ + vpaddd %ymm6, %ymm5, %ymm3 + +/* reduced argument R */ + vsubps _sOne(%rax), %ymm3, %ymm5 + +/* polynomial evaluation starts here */ + vfmadd213ps _sPoly_6(%rax), %ymm5, %ymm4 + vfmadd213ps _sPoly_5(%rax), %ymm5, %ymm4 + vfmadd213ps _sPoly_4(%rax), %ymm5, %ymm4 + vfmadd213ps _sPoly_3(%rax), %ymm5, %ymm4 + vfmadd213ps _sPoly_2(%rax), %ymm5, %ymm4 + vfmadd213ps _sPoly_1(%rax), %ymm5, %ymm4 + vmulps %ymm5, %ymm4, %ymm6 + +/* polynomial evaluation end */ + vfmadd213ps %ymm5, %ymm5, %ymm6 + vpcmpgtd %ymm7, %ymm1, %ymm1 + vmovmskps %ymm1, %ecx + +/* final reconstruction: + add exponent_value*log2 to polynomial result */ + vfmadd132ps _sLn2(%rax), %ymm6, %ymm0 + 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 + vmovups %ymm2, 320(%rsp) + vmovups %ymm0, 384(%rsp) + je .LBL_1_2 + + xorb %dl, %dl + xorl %eax, %eax + vmovups %ymm8, 224(%rsp) + vmovups %ymm9, 192(%rsp) + vmovups %ymm10, 160(%rsp) + vmovups %ymm11, 128(%rsp) + vmovups %ymm12, 96(%rsp) + vmovups %ymm13, 64(%rsp) + vmovups %ymm14, 32(%rsp) + vmovups %ymm15, (%rsp) + movq %rsi, 264(%rsp) + movq %rdi, 256(%rsp) + movq %r12, 296(%rsp) + cfi_offset_rel_rsp (12, 296) + movb %dl, %r12b + movq %r13, 288(%rsp) + cfi_offset_rel_rsp (13, 288) + movl %ecx, %r13d + movq %r14, 280(%rsp) + cfi_offset_rel_rsp (14, 280) + movl %eax, %r14d + movq %r15, 272(%rsp) + cfi_offset_rel_rsp (15, 272) + 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 + + vmovups 224(%rsp), %ymm8 + vmovups 192(%rsp), %ymm9 + vmovups 160(%rsp), %ymm10 + vmovups 128(%rsp), %ymm11 + vmovups 96(%rsp), %ymm12 + vmovups 64(%rsp), %ymm13 + vmovups 32(%rsp), %ymm14 + vmovups (%rsp), %ymm15 + vmovups 384(%rsp), %ymm0 + movq 264(%rsp), %rsi + movq 256(%rsp), %rdi + movq 296(%rsp), %r12 + cfi_restore (%r12) + movq 288(%rsp), %r13 + cfi_restore (%r13) + movq 280(%rsp), %r14 + cfi_restore (%r14) + movq 272(%rsp), %r15 + cfi_restore (%r15) + jmp .LBL_1_2 + +.LBL_1_10: + cfi_restore_state + movzbl %r12b, %r15d + vmovss 324(%rsp,%r15,8), %xmm0 + vzeroupper + + call logf@PLT + + vmovss %xmm0, 388(%rsp,%r15,8) + jmp .LBL_1_8 + +.LBL_1_12: + movzbl %r12b, %r15d + vmovss 320(%rsp,%r15,8), %xmm0 + vzeroupper + + call logf@PLT + + vmovss %xmm0, 384(%rsp,%r15,8) + jmp .LBL_1_7 + +END(_ZGVdN8v_logf_avx2) diff --git a/sysdeps/x86_64/fpu/svml_s_logf16_core.S b/sysdeps/x86_64/fpu/svml_s_logf16_core.S new file mode 100644 index 0000000000..47ae7855a3 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf16_core.S @@ -0,0 +1,25 @@ +/* Function logf vectorized with AVX-512. Wrapper to AVX2 version. + 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 + . */ + +#include +#include "svml_s_wrapper_impl.h" + + .text +ENTRY (_ZGVeN16v_logf) +WRAPPER_IMPL_AVX512 _ZGVdN8v_logf +END (_ZGVeN16v_logf) diff --git a/sysdeps/x86_64/fpu/svml_s_logf4_core.S b/sysdeps/x86_64/fpu/svml_s_logf4_core.S new file mode 100644 index 0000000000..09be406d3c --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf4_core.S @@ -0,0 +1,30 @@ +/* Function logf vectorized with SSE2. + 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 + . */ + + +#include +#include "svml_s_wrapper_impl.h" + + .text +ENTRY (_ZGVbN4v_logf) +WRAPPER_IMPL_SSE2 logf +END (_ZGVbN4v_logf) + +#ifndef USE_MULTIARCH + libmvec_hidden_def (_ZGVbN4v_logf) +#endif diff --git a/sysdeps/x86_64/fpu/svml_s_logf8_core.S b/sysdeps/x86_64/fpu/svml_s_logf8_core.S new file mode 100644 index 0000000000..cf4e9be537 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf8_core.S @@ -0,0 +1,29 @@ +/* Function logf vectorized with AVX2, wrapper version. + 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 + . */ + +#include +#include "svml_s_wrapper_impl.h" + + .text +ENTRY (_ZGVdN8v_logf) +WRAPPER_IMPL_AVX _ZGVbN4v_logf +END (_ZGVdN8v_logf) + +#ifndef USE_MULTIARCH + libmvec_hidden_def (_ZGVdN8v_logf) +#endif diff --git a/sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S b/sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S new file mode 100644 index 0000000000..7ab572bb30 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf8_core_avx.S @@ -0,0 +1,25 @@ +/* Function logf vectorized in AVX ISA as wrapper to SSE4 ISA version. + 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 + . */ + +#include +#include "svml_s_wrapper_impl.h" + + .text +ENTRY(_ZGVcN8v_logf) +WRAPPER_IMPL_AVX _ZGVbN4v_logf +END(_ZGVcN8v_logf) diff --git a/sysdeps/x86_64/fpu/svml_s_logf_data.S b/sysdeps/x86_64/fpu/svml_s_logf_data.S new file mode 100644 index 0000000000..1e7f7015d3 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf_data.S @@ -0,0 +1,102 @@ +/* Data for vector function logf. + 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 + . */ + +#include "svml_s_logf_data.h" + + .section .rodata, "a" + .align 64 + +/* Data table for vector implementations of function logf. + The table may contain polynomial, reduction, lookup coefficients and + other coefficients obtained through different methods of research and + experimental work. */ + + .globl __svml_slog_data +__svml_slog_data: + +/* Polynomial sPoly[] coefficients: + * -5.0000000000000000000000000e-01 */ +float_vector _sPoly_1 0xbf000000 + +/* 3.3336564898490905761718750e-01 */ +float_vector _sPoly_2 0x3eaaaee7 + +/* -2.5004664063453674316406250e-01 */ +float_vector _sPoly_3 0xbe80061d + +/* 1.9822503626346588134765625e-01 */ +float_vector _sPoly_4 0x3e4afb81 + +/* -1.6462457180023193359375000e-01 */ +float_vector _sPoly_5 0xbe289358 + +/* 1.6964881122112274169921875e-01 */ +float_vector _sPoly_6 0x3e2db86b + +/* -1.5177205204963684082031250e-01 */ +float_vector _sPoly_7 0xbe1b6a22 + +/* Constant for work range check: Delta 80000000-7f800000 */ +float_vector _iHiDelta 0x00800000 + +/* Constant for work range check: 00800000 + Delta */ +float_vector _iLoRange 0x01000000 + +/* Mantissa break point SP 2/3 */ +float_vector _iBrkValue 0x3f2aaaab + +/* SP significand mask */ +float_vector _iOffExpoMask 0x007fffff + +/* 1.0f */ +float_vector _sOne 0x3f800000 + +/* SP log(2) */ +float_vector _sLn2 0x3f317218 + +/* SP infinity, +/- */ +.if .-__svml_slog_data != _sInfs +.err +.endif + .long 0x7f800000 + .long 0xff800000 + .rept 56 + .byte 0 + .endr + +/* SP one, +/- */ +.if .-__svml_slog_data != _sOnes +.err +.endif + .long 0x3f800000 + .long 0xbf800000 + .rept 56 + .byte 0 + .endr + +/* SP zero +/- */ +.if .-__svml_slog_data != _sZeros +.err +.endif + .long 0x00000000 + .long 0x80000000 + .rept 56 + .byte 0 + .endr + .type __svml_slog_data,@object + .size __svml_slog_data,.-__svml_slog_data diff --git a/sysdeps/x86_64/fpu/svml_s_logf_data.h b/sysdeps/x86_64/fpu/svml_s_logf_data.h new file mode 100644 index 0000000000..d42411a849 --- /dev/null +++ b/sysdeps/x86_64/fpu/svml_s_logf_data.h @@ -0,0 +1,48 @@ +/* Offsets for data table for vectorized function logf. + 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 + . */ + +#ifndef S_LOGF_DATA_H +#define S_LOGF_DATA_H + +#define _sPoly_1 0 +#define _sPoly_2 64 +#define _sPoly_3 128 +#define _sPoly_4 192 +#define _sPoly_5 256 +#define _sPoly_6 320 +#define _sPoly_7 384 +#define _iHiDelta 448 +#define _iLoRange 512 +#define _iBrkValue 576 +#define _iOffExpoMask 640 +#define _sOne 704 +#define _sLn2 768 +#define _sInfs 832 +#define _sOnes 896 +#define _sZeros 960 + +.macro float_vector offset value +.if .-__svml_slog_data != \offset +.err +.endif +.rept 16 +.long \value +.endr +.endm + +#endif diff --git a/sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c b/sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c index 801b03cb72..72435e4b1e 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c +++ b/sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c @@ -24,3 +24,4 @@ VECTOR_WRAPPER (WRAPPER_NAME (cosf), _ZGVeN16v_cosf) VECTOR_WRAPPER (WRAPPER_NAME (sinf), _ZGVeN16v_sinf) +VECTOR_WRAPPER (WRAPPER_NAME (logf), _ZGVeN16v_logf) diff --git a/sysdeps/x86_64/fpu/test-float-vlen16.c b/sysdeps/x86_64/fpu/test-float-vlen16.c index 8988cdbc64..10da5fefa3 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen16.c +++ b/sysdeps/x86_64/fpu/test-float-vlen16.c @@ -20,6 +20,7 @@ #define TEST_VECTOR_cosf 1 #define TEST_VECTOR_sinf 1 +#define TEST_VECTOR_logf 1 #define REQUIRE_AVX512F diff --git a/sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c b/sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c index 3a0fa6a809..f51575d35e 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c +++ b/sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c @@ -24,3 +24,4 @@ VECTOR_WRAPPER (WRAPPER_NAME (cosf), _ZGVbN4v_cosf) VECTOR_WRAPPER (WRAPPER_NAME (sinf), _ZGVbN4v_sinf) +VECTOR_WRAPPER (WRAPPER_NAME (logf), _ZGVbN4v_logf) diff --git a/sysdeps/x86_64/fpu/test-float-vlen4.c b/sysdeps/x86_64/fpu/test-float-vlen4.c index 38637877cf..5cb293f485 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen4.c +++ b/sysdeps/x86_64/fpu/test-float-vlen4.c @@ -20,5 +20,6 @@ #define TEST_VECTOR_cosf 1 #define TEST_VECTOR_sinf 1 +#define TEST_VECTOR_logf 1 #include "libm-test.c" diff --git a/sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c b/sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c index a85f5887b2..7515a5981e 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c +++ b/sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c @@ -27,3 +27,4 @@ VECTOR_WRAPPER (WRAPPER_NAME (cosf), _ZGVdN8v_cosf) VECTOR_WRAPPER (WRAPPER_NAME (sinf), _ZGVdN8v_sinf) +VECTOR_WRAPPER (WRAPPER_NAME (logf), _ZGVdN8v_logf) diff --git a/sysdeps/x86_64/fpu/test-float-vlen8-avx2.c b/sysdeps/x86_64/fpu/test-float-vlen8-avx2.c index db0b5e52d1..1026d63e95 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen8-avx2.c +++ b/sysdeps/x86_64/fpu/test-float-vlen8-avx2.c @@ -23,6 +23,7 @@ #define TEST_VECTOR_cosf 1 #define TEST_VECTOR_sinf 1 +#define TEST_VECTOR_logf 1 #define REQUIRE_AVX2 diff --git a/sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c b/sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c index fb7f696aee..6dde1a2cf1 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c +++ b/sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c @@ -24,3 +24,4 @@ VECTOR_WRAPPER (WRAPPER_NAME (cosf), _ZGVcN8v_cosf) VECTOR_WRAPPER (WRAPPER_NAME (sinf), _ZGVcN8v_sinf) +VECTOR_WRAPPER (WRAPPER_NAME (logf), _ZGVcN8v_logf) diff --git a/sysdeps/x86_64/fpu/test-float-vlen8.c b/sysdeps/x86_64/fpu/test-float-vlen8.c index f893c5b80f..3fe10ad7bc 100644 --- a/sysdeps/x86_64/fpu/test-float-vlen8.c +++ b/sysdeps/x86_64/fpu/test-float-vlen8.c @@ -20,5 +20,6 @@ #define TEST_VECTOR_cosf 1 #define TEST_VECTOR_sinf 1 +#define TEST_VECTOR_logf 1 #include "libm-test.c"