x86-64: Add vector log1p/log1pf implementation to libmvec

Implement vectorized log1p/log1pf containing SSE, AVX, AVX2 and
AVX512 versions for libmvec as per vector ABI.  It also contains
accuracy and ABI tests for vector log1p/log1pf with regenerated ulps.

Reviewed-by: H.J. Lu <hjl.tools@gmail.com>

1 files changed, 271 insertions, 0 deletions

diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_log1pf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_log1pf16_core_avx512.S
new file mode 100644
index 0000000000..78b2fe417f
--- /dev/null
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_log1pf16_core_avx512.S
@@ -0,0 +1,271 @@
+/* Function log1pf vectorized with AVX-512.
+   Copyright (C) 2021 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
+   https://www.gnu.org/licenses/.  */
+
+/*
+ * ALGORITHM DESCRIPTION:
+ *
+ *    1+x = 2^k*(xh + xl) is computed in high-low parts; xh in [1,2)
+ *    Get short reciprocal approximation Rcp ~ 1/xh
+ *    R = (Rcp*xh - 1.0) + Rcp*xl
+ *    log1p(x) = k*log(2.0) - log(Rcp) + poly(R)
+ *       log(Rcp) is tabulated
+ *
+ *
+ */
+
+/* Offsets for data table __svml_slog1p_data_internal
+ */
+#define SgnMask                       	0
+#define sOne                          	64
+#define sPoly_1                       	128
+#define sPoly_2                       	192
+#define sPoly_3                       	256
+#define sPoly_4                       	320
+#define sPoly_5                       	384
+#define sPoly_6                       	448
+#define sPoly_7                       	512
+#define sPoly_8                       	576
+#define iHiDelta                      	640
+#define iLoRange                      	704
+#define iBrkValue                     	768
+#define iOffExpoMask                  	832
+#define sLn2                          	896
+
+#include <sysdep.h>
+
+        .text
+	.section .text.exex512,"ax",@progbits
+ENTRY(_ZGVeN16v_log1pf_skx)
+        pushq     %rbp
+        cfi_def_cfa_offset(16)
+        movq      %rsp, %rbp
+        cfi_def_cfa(6, 16)
+        cfi_offset(6, -16)
+        andq      $-64, %rsp
+        subq      $192, %rsp
+        vmovups   sOne+__svml_slog1p_data_internal(%rip), %zmm2
+
+/* reduction: compute r,n */
+        vmovups   iBrkValue+__svml_slog1p_data_internal(%rip), %zmm12
+        vmovups   SgnMask+__svml_slog1p_data_internal(%rip), %zmm4
+        vmovaps   %zmm0, %zmm3
+
+/* compute 1+x as high, low parts */
+        vmaxps    {sae}, %zmm3, %zmm2, %zmm5
+        vminps    {sae}, %zmm3, %zmm2, %zmm7
+        vandnps   %zmm3, %zmm4, %zmm1
+        vpternlogd $255, %zmm4, %zmm4, %zmm4
+        vaddps    {rn-sae}, %zmm7, %zmm5, %zmm9
+        vpsubd    %zmm12, %zmm9, %zmm10
+        vsubps    {rn-sae}, %zmm9, %zmm5, %zmm6
+
+/* check argument value ranges */
+        vpaddd    iHiDelta+__svml_slog1p_data_internal(%rip), %zmm9, %zmm8
+        vpsrad    $23, %zmm10, %zmm13
+        vmovups   sPoly_5+__svml_slog1p_data_internal(%rip), %zmm9
+        vpcmpd    $5, iLoRange+__svml_slog1p_data_internal(%rip), %zmm8, %k1
+        vpslld    $23, %zmm13, %zmm14
+        vaddps    {rn-sae}, %zmm7, %zmm6, %zmm15
+        vcvtdq2ps {rn-sae}, %zmm13, %zmm0
+        vpsubd    %zmm14, %zmm2, %zmm13
+        vmovups   sPoly_8+__svml_slog1p_data_internal(%rip), %zmm7
+        vmovups   sPoly_1+__svml_slog1p_data_internal(%rip), %zmm14
+        vmulps    {rn-sae}, %zmm13, %zmm15, %zmm6
+        vpandd    iOffExpoMask+__svml_slog1p_data_internal(%rip), %zmm10, %zmm11
+        vpaddd    %zmm12, %zmm11, %zmm5
+        vmovups   sPoly_4+__svml_slog1p_data_internal(%rip), %zmm10
+        vmovups   sPoly_3+__svml_slog1p_data_internal(%rip), %zmm11
+        vmovups   sPoly_2+__svml_slog1p_data_internal(%rip), %zmm12
+
+/* polynomial evaluation */
+        vsubps    {rn-sae}, %zmm2, %zmm5, %zmm2
+        vaddps    {rn-sae}, %zmm6, %zmm2, %zmm15
+        vmovups   sPoly_7+__svml_slog1p_data_internal(%rip), %zmm2
+        vfmadd231ps {rn-sae}, %zmm15, %zmm7, %zmm2
+        vpandnd   %zmm8, %zmm8, %zmm4{%k1}
+        vmovups   sPoly_6+__svml_slog1p_data_internal(%rip), %zmm8
+
+/* combine and get argument value range mask */
+        vptestmd  %zmm4, %zmm4, %k0
+        vfmadd213ps {rn-sae}, %zmm8, %zmm15, %zmm2
+        kmovw     %k0, %edx
+        vfmadd213ps {rn-sae}, %zmm9, %zmm15, %zmm2
+        vfmadd213ps {rn-sae}, %zmm10, %zmm15, %zmm2
+        vfmadd213ps {rn-sae}, %zmm11, %zmm15, %zmm2
+        vfmadd213ps {rn-sae}, %zmm12, %zmm15, %zmm2
+        vfmadd213ps {rn-sae}, %zmm14, %zmm15, %zmm2
+        vmulps    {rn-sae}, %zmm15, %zmm2, %zmm4
+        vfmadd213ps {rn-sae}, %zmm15, %zmm15, %zmm4
+
+/* final reconstruction */
+        vmovups   sLn2+__svml_slog1p_data_internal(%rip), %zmm15
+        vfmadd213ps {rn-sae}, %zmm4, %zmm15, %zmm0
+        vorps     %zmm1, %zmm0, %zmm0
+        testl     %edx, %edx
+
+/* Go to special inputs processing branch */
+        jne       L(SPECIAL_VALUES_BRANCH)
+                                # LOE rbx r12 r13 r14 r15 edx zmm0 zmm3
+
+/* Restore registers
+ * and exit the function
+ */
+
+L(EXIT):
+        movq      %rbp, %rsp
+        popq      %rbp
+        cfi_def_cfa(7, 8)
+        cfi_restore(6)
+        ret
+        cfi_def_cfa(6, 16)
+        cfi_offset(6, -16)
+
+/* Branch to process
+ * special inputs
+ */
+
+L(SPECIAL_VALUES_BRANCH):
+        vmovups   %zmm3, 64(%rsp)
+        vmovups   %zmm0, 128(%rsp)
+                                # LOE rbx r12 r13 r14 r15 edx zmm0
+
+        xorl      %eax, %eax
+                                # LOE rbx r12 r13 r14 r15 eax edx
+
+        vzeroupper
+        movq      %r12, 16(%rsp)
+        /*  DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22
+        movl      %eax, %r12d
+        movq      %r13, 8(%rsp)
+        /*  DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22
+        movl      %edx, %r13d
+        movq      %r14, (%rsp)
+        /*  DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22
+                                # LOE rbx r15 r12d r13d
+
+/* Range mask
+ * bits check
+ */
+
+L(RANGEMASK_CHECK):
+        btl       %r12d, %r13d
+
+/* Call scalar math function */
+        jc        L(SCALAR_MATH_CALL)
+                                # LOE rbx r15 r12d r13d
+
+/* Special inputs
+ * processing loop
+ */
+
+L(SPECIAL_VALUES_LOOP):
+        incl      %r12d
+        cmpl      $16, %r12d
+
+/* Check bits in range mask */
+        jl        L(RANGEMASK_CHECK)
+                                # LOE rbx r15 r12d r13d
+
+        movq      16(%rsp), %r12
+        cfi_restore(12)
+        movq      8(%rsp), %r13
+        cfi_restore(13)
+        movq      (%rsp), %r14
+        cfi_restore(14)
+        vmovups   128(%rsp), %zmm0
+
+/* Go to exit */
+        jmp       L(EXIT)
+        /*  DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22
+        /*  DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22
+        /*  DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus)  */
+        .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22
+                                # LOE rbx r12 r13 r14 r15 zmm0
+
+/* Scalar math fucntion call
+ * to process special input
+ */
+
+L(SCALAR_MATH_CALL):
+        movl      %r12d, %r14d
+        movss     64(%rsp,%r14,4), %xmm0
+        call      log1pf@PLT
+                                # LOE rbx r14 r15 r12d r13d xmm0
+
+        movss     %xmm0, 128(%rsp,%r14,4)
+
+/* Process special inputs in loop */
+        jmp       L(SPECIAL_VALUES_LOOP)
+                                # LOE rbx r15 r12d r13d
+END(_ZGVeN16v_log1pf_skx)
+
+        .section .rodata, "a"
+        .align 64
+
+#ifdef __svml_slog1p_data_internal_typedef
+typedef unsigned int VUINT32;
+typedef struct {
+        __declspec(align(64)) VUINT32 SgnMask[16][1];
+        __declspec(align(64)) VUINT32 sOne[16][1];
+        __declspec(align(64)) VUINT32 sPoly[8][16][1];
+        __declspec(align(64)) VUINT32 iHiDelta[16][1];
+        __declspec(align(64)) VUINT32 iLoRange[16][1];
+        __declspec(align(64)) VUINT32 iBrkValue[16][1];
+        __declspec(align(64)) VUINT32 iOffExpoMask[16][1];
+        __declspec(align(64)) VUINT32 sLn2[16][1];
+} __svml_slog1p_data_internal;
+#endif
+__svml_slog1p_data_internal:
+        /*== SgnMask ==*/
+        .long 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff
+        /*== sOne = SP 1.0 ==*/
+        .align 64
+        .long 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000
+        /*== sPoly[] = SP polynomial ==*/
+        .align 64
+        .long 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000 /* -5.0000000000000000000000000e-01 P0 */
+        .long 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94, 0x3eaaaa94 /*  3.3333265781402587890625000e-01 P1 */
+        .long 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e, 0xbe80058e /* -2.5004237890243530273437500e-01 P2 */
+        .long 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190, 0x3e4ce190 /*  2.0007920265197753906250000e-01 P3 */
+        .long 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37, 0xbe28ad37 /* -1.6472326219081878662109375e-01 P4 */
+        .long 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12, 0x3e0fcb12 /*  1.4042308926582336425781250e-01 P5 */
+        .long 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3, 0xbe1ad9e3 /* -1.5122179687023162841796875e-01 P6 */
+        .long 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed, 0x3e0d84ed /*  1.3820238411426544189453125e-01 P7 */
+        /*== iHiDelta = SP 80000000-7f000000 ==*/
+        .align 64
+        .long 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000, 0x01000000
+        /*== iLoRange = SP 00800000+iHiDelta ==*/
+        .align 64
+        .long 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000, 0x01800000
+        /*== iBrkValue = SP 2/3 ==*/
+        .align 64
+        .long 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab, 0x3f2aaaab
+        /*== iOffExpoMask = SP significand mask ==*/
+        .align 64
+        .long 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff, 0x007fffff
+        /*== sLn2 = SP ln(2) ==*/
+        .align 64
+        .long 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218, 0x3f317218
+        .align 64
+        .type	__svml_slog1p_data_internal,@object
+        .size	__svml_slog1p_data_internal,.-__svml_slog1p_data_internal