/*
* Copyright (C) 2025 Linux Studio Plugins Project
* (C) 2025 Vladimir Sadovnikov
*
* This file is part of lsp-dsp-lib
* Created on: 13 апр. 2025 г.
*
* lsp-dsp-lib 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 3 of the License, or
* any later version.
*
* lsp-dsp-lib 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 lsp-dsp-lib. If not, see .
*/
#ifndef PRIVATE_DSP_ARCH_AARCH64_ASIMD_PMATH_SIN_H_
#define PRIVATE_DSP_ARCH_AARCH64_ASIMD_PMATH_SIN_H_
#ifndef PRIVATE_DSP_ARCH_AARCH64_ASIMD_IMPL
#error "This header should not be included directly"
#endif /* PRIVATE_DSP_ARCH_AARCH64_ASIMD_IMPL */
namespace lsp
{
namespace asimd
{
IF_ARCH_AARCH64(
static const float kp_gen_const[] __lsp_aligned16 =
{
0.0f, 1.0f, 2.0f, 3.0f, // +0x00: Initial values
4.0f, 5.0f, 6.0f, 7.0f,
LSP_DSP_VEC4(8.0), // +0x20: Step x8
LSP_DSP_VEC4(4.0), // +0x30: Step x4
};
static const uint32_t sinf_const[] __lsp_aligned16 =
{
LSP_DSP_VEC4(0x3fc90fdb), // +0x000: PI/2
LSP_DSP_VEC4(0x40490fdb), // +0x010: PI
LSP_DSP_VEC4(0x3e22f983), // +0x020: 1/(2*PI)
LSP_DSP_VEC4(0xb2d7322b), // +0x030: C0 = -1/11! = -2.50521083854e-08
LSP_DSP_VEC4(0x3638ef1d), // +0x040: C1 = 1/9! = 2.7557319224e-06
LSP_DSP_VEC4(0xb9500d01), // +0x050: C2 = -1/7! = -0.000198412698413
LSP_DSP_VEC4(0x3c088889), // +0x060: C3 = 1/5! = 0.00833333333333
LSP_DSP_VEC4(0xbe2aaaab), // +0x070: C4 = -1/3! = -0.166666666667
};
)
#define SINF_X_PLUS_PI_2_CORE_X16 \
/* v0 = X0 = x0 + PI/2 */ \
/* v1 = X1 = x1 + PI/2 */ \
/* v2 = X2 = x2 + PI/2 */ \
/* v3 = X3 = x3 + PI/2 */ \
/* v16 = PI/2 */ \
/* v17 = PI */ \
/* v18 = 1/(2*PI) */ \
/* v19 = C0 */ \
/* v20 = C1 */ \
/* v21 = C2 */ \
/* v22 = C3 */ \
/* v23 = C4 */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v18.4s") /* v4 = x / (2*PI) */ \
__ASM_EMIT("fmul v5.4s, v1.4s, v18.4s") \
__ASM_EMIT("fmul v6.4s, v2.4s, v18.4s") \
__ASM_EMIT("fmul v7.4s, v3.4s, v18.4s") \
__ASM_EMIT("fcvtms v8.4s, v4.4s") /* v8 = int(X / (2*PI)) + [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("fcvtms v9.4s, v5.4s") \
__ASM_EMIT("fcvtms v10.4s, v6.4s") \
__ASM_EMIT("fcvtms v11.4s, v7.4s") \
__ASM_EMIT("fadd v12.4s, v17.4s, v17.4s") /* v12 = 2*PI */ \
__ASM_EMIT("scvtf v8.4s, v8.4s") /* v8 = period = int(X / (2*PI)) [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("scvtf v9.4s, v9.4s") \
__ASM_EMIT("scvtf v10.4s, v10.4s") \
__ASM_EMIT("scvtf v11.4s, v11.4s") \
__ASM_EMIT("fmls v0.4s, v8.4s, v12.4s") /* v0 = Y = X - period * 2 * PI */ \
__ASM_EMIT("fmls v1.4s, v9.4s, v12.4s") \
__ASM_EMIT("fmls v2.4s, v10.4s, v12.4s") \
__ASM_EMIT("fmls v3.4s, v11.4s, v12.4s") \
__ASM_EMIT("fadd v12.4s, v16.4s, v17.4s") /* v12 = 3*PI/2 */ \
__ASM_EMIT("fcmge v4.4s, v0.4s, v17.4s") /* v4 = [ Y >= PI ] */ \
__ASM_EMIT("fcmge v5.4s, v1.4s, v17.4s") \
__ASM_EMIT("fcmge v6.4s, v2.4s, v17.4s") \
__ASM_EMIT("fcmge v7.4s, v3.4s, v17.4s") \
__ASM_EMIT("fsub v8.4s, v12.4s, v0.4s") /* v8 = 3*PI/2 - Y */ \
__ASM_EMIT("fsub v9.4s, v12.4s, v1.4s") \
__ASM_EMIT("fsub v10.4s, v12.4s, v2.4s") \
__ASM_EMIT("fsub v11.4s, v12.4s, v3.4s") \
__ASM_EMIT("fsub v0.4s, v0.4s, v16.4s") /* v0 = Y - PI/2 */ \
__ASM_EMIT("fsub v1.4s, v1.4s, v16.4s") \
__ASM_EMIT("fsub v2.4s, v2.4s, v16.4s") \
__ASM_EMIT("fsub v3.4s, v3.4s, v16.4s") \
__ASM_EMIT("bit v0.16b, v8.16b, v4.16b") /* v0 = XX = [ Y < PI ] ? (Y - PI/2) : (3*PI/2 - Y) */ \
__ASM_EMIT("bit v1.16b, v9.16b, v5.16b") \
__ASM_EMIT("bit v2.16b, v10.16b, v6.16b") \
__ASM_EMIT("bit v3.16b, v11.16b, v7.16b") \
/* v0 = XX */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v0.4s") /* v4 = X2 = XX*XX */ \
__ASM_EMIT("fmul v5.4s, v1.4s, v1.4s") \
__ASM_EMIT("fmul v6.4s, v2.4s, v2.4s") \
__ASM_EMIT("fmul v7.4s, v3.4s, v3.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v19.4s") /* v8 = X2*C0 */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v19.4s") \
__ASM_EMIT("fmul v10.4s, v6.4s, v19.4s") \
__ASM_EMIT("fmul v11.4s, v7.4s, v19.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v20.4s") /* v8 = C1 + X2*C0 */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v20.4s") \
__ASM_EMIT("fadd v10.4s, v10.4s, v20.4s") \
__ASM_EMIT("fadd v11.4s, v11.4s, v20.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C1 + X2*C0) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fmul v10.4s, v6.4s, v10.4s") \
__ASM_EMIT("fmul v11.4s, v7.4s, v11.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v21.4s") /* v8 = C2 + X2*(C1 + X2*C0) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v21.4s") \
__ASM_EMIT("fadd v10.4s, v10.4s, v21.4s") \
__ASM_EMIT("fadd v11.4s, v11.4s, v21.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fmul v10.4s, v6.4s, v10.4s") \
__ASM_EMIT("fmul v11.4s, v7.4s, v11.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v22.4s") /* v8 = C3 + X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v22.4s") \
__ASM_EMIT("fadd v10.4s, v10.4s, v22.4s") \
__ASM_EMIT("fadd v11.4s, v11.4s, v22.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fmul v10.4s, v6.4s, v10.4s") \
__ASM_EMIT("fmul v11.4s, v7.4s, v11.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v23.4s") /* v8 = C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v23.4s") \
__ASM_EMIT("fadd v10.4s, v10.4s, v23.4s") \
__ASM_EMIT("fadd v11.4s, v11.4s, v23.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0)))) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fmul v10.4s, v6.4s, v10.4s") \
__ASM_EMIT("fmul v11.4s, v7.4s, v11.4s") \
__ASM_EMIT("fmla v0.4s, v0.4s, v8.4s") /* v0 = XX*(1.0 + X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))))) */ \
__ASM_EMIT("fmla v1.4s, v1.4s, v9.4s") \
__ASM_EMIT("fmla v2.4s, v2.4s, v10.4s") \
__ASM_EMIT("fmla v3.4s, v3.4s, v11.4s")
#define SINF_X_PLUS_PI_2_CORE_X8 \
/* v0 = X0 = x0 + PI/2 */ \
/* v1 = X1 = x1 + PI/2 */ \
/* v16 = PI/2 */ \
/* v17 = PI */ \
/* v18 = 1/(2*PI) */ \
/* v19 = C0 */ \
/* v20 = C1 */ \
/* v21 = C2 */ \
/* v22 = C3 */ \
/* v23 = C4 */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v18.4s") /* v4 = x / (2*PI) */ \
__ASM_EMIT("fmul v5.4s, v1.4s, v18.4s") \
__ASM_EMIT("fcvtms v8.4s, v4.4s") /* v8 = int(X / (2*PI)) + [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("fcvtms v9.4s, v5.4s") \
__ASM_EMIT("fadd v12.4s, v17.4s, v17.4s") /* v12 = 2*PI */ \
__ASM_EMIT("scvtf v8.4s, v8.4s") /* v8 = period = int(X / (2*PI)) [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("scvtf v9.4s, v9.4s") \
__ASM_EMIT("fmls v0.4s, v8.4s, v12.4s") /* v0 = Y = X - period * 2 * PI */ \
__ASM_EMIT("fmls v1.4s, v9.4s, v12.4s") \
__ASM_EMIT("fadd v12.4s, v16.4s, v17.4s") /* v12 = 3*PI/2 */ \
__ASM_EMIT("fcmge v4.4s, v0.4s, v17.4s") /* v4 = [ Y >= PI ] */ \
__ASM_EMIT("fcmge v5.4s, v1.4s, v17.4s") \
__ASM_EMIT("fsub v8.4s, v12.4s, v0.4s") /* v8 = 3*PI/2 - Y */ \
__ASM_EMIT("fsub v9.4s, v12.4s, v1.4s") \
__ASM_EMIT("fsub v0.4s, v0.4s, v16.4s") /* v0 = Y - PI/2 */ \
__ASM_EMIT("fsub v1.4s, v1.4s, v16.4s") \
__ASM_EMIT("bit v0.16b, v8.16b, v4.16b") /* v0 = XX = [ Y < PI ] ? (Y - PI/2) : (3*PI/2 - Y) */ \
__ASM_EMIT("bit v1.16b, v9.16b, v5.16b") \
/* v0 = XX */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v0.4s") /* v4 = X2 = XX*XX */ \
__ASM_EMIT("fmul v5.4s, v1.4s, v1.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v19.4s") /* v8 = X2*C0 */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v19.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v20.4s") /* v8 = C1 + X2*C0 */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v20.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C1 + X2*C0) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v21.4s") /* v8 = C2 + X2*(C1 + X2*C0) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v21.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v22.4s") /* v8 = C3 + X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v22.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fadd v8.4s, v8.4s, v23.4s") /* v8 = C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fadd v9.4s, v9.4s, v23.4s") \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0)))) */ \
__ASM_EMIT("fmul v9.4s, v5.4s, v9.4s") \
__ASM_EMIT("fmla v0.4s, v0.4s, v8.4s") /* v0 = XX*(1.0 + X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))))) */ \
__ASM_EMIT("fmla v1.4s, v1.4s, v9.4s")
#define SINF_X_PLUS_PI_2_CORE_X4 \
/* v0 = X = x + PI/2 */ \
/* v16 = PI/2 */ \
/* v17 = PI */ \
/* v18 = 1/(2*PI) */ \
/* v19 = C0 */ \
/* v20 = C1 */ \
/* v21 = C2 */ \
/* v22 = C3 */ \
/* v23 = C4 */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v18.4s") /* v4 = x / (2*PI) */ \
__ASM_EMIT("fcvtms v8.4s, v4.4s") /* v8 = int(X / (2*PI)) + [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("fadd v12.4s, v17.4s, v17.4s") /* v12 = 2*PI */ \
__ASM_EMIT("scvtf v8.4s, v8.4s") /* v8 = period = int(X / (2*PI)) [ X < 0.0 ] ? -1 : 0 */ \
__ASM_EMIT("fmls v0.4s, v8.4s, v12.4s") /* v0 = Y = X - period * 2 * PI */ \
__ASM_EMIT("fadd v12.4s, v16.4s, v17.4s") /* v12 = 3*PI/2 */ \
__ASM_EMIT("fcmge v4.4s, v0.4s, v17.4s") /* v4 = [ Y >= PI ] */ \
__ASM_EMIT("fsub v8.4s, v12.4s, v0.4s") /* v8 = 3*PI/2 - Y */ \
__ASM_EMIT("fsub v0.4s, v0.4s, v16.4s") /* v0 = Y - PI/2 */ \
__ASM_EMIT("bit v0.16b, v8.16b, v4.16b") /* v0 = XX = [ Y < PI ] ? (Y - PI/2) : (3*PI/2 - Y) */ \
/* v0 = XX */ \
__ASM_EMIT("fmul v4.4s, v0.4s, v0.4s") /* v4 = X2 = XX*XX */ \
__ASM_EMIT("fmul v8.4s, v4.4s, v19.4s") /* v8 = X2*C0 */ \
__ASM_EMIT("fadd v8.4s, v8.4s, v20.4s") /* v8 = C1 + X2*C0 */ \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C1 + X2*C0) */ \
__ASM_EMIT("fadd v8.4s, v8.4s, v21.4s") /* v8 = C2 + X2*(C1 + X2*C0) */ \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fadd v8.4s, v8.4s, v22.4s") /* v8 = C3 + X2*(C2 + X2*(C1 + X2*C0)) */ \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fadd v8.4s, v8.4s, v23.4s") /* v8 = C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))) */ \
__ASM_EMIT("fmul v8.4s, v4.4s, v8.4s") /* v8 = X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0)))) */ \
__ASM_EMIT("fmla v0.4s, v0.4s, v8.4s") /* v0 = XX*(1.0 + X2*(C4 + X2*(C3 + X2*(C2 + X2*(C1 + X2*C0))))) */
void sinf1(float *dst, size_t count)
{
ARCH_AARCH64_ASM(
__ASM_EMIT("ldp q16, q17, [%[S2C], #0x00]") /* v16 = PI/2, v17 = PI */
__ASM_EMIT("ldp q18, q19, [%[S2C], #0x20]") /* v18 = 1/(2*PI), v19 = C0 */
__ASM_EMIT("ldp q20, q21, [%[S2C], #0x40]") /* v20 = C1, v21 = C2 */
__ASM_EMIT("ldp q22, q23, [%[S2C], #0x60]") /* v22 = C3, v23 = C4 */
// x16 blocks
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("b.lo 2f")
__ASM_EMIT("1:")
__ASM_EMIT("ldp q0, q1, [%[dst], #0x00]") // v0 = x0, v1 = x1
__ASM_EMIT("ldp q2, q3, [%[dst], #0x20]") // v2 = x2, v3 = x3
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v16.4s") // v1 = x1 + PI/2
__ASM_EMIT("fadd v2.4s, v2.4s, v16.4s") // v2 = x2 + PI/2
__ASM_EMIT("fadd v3.4s, v3.4s, v16.4s") // v3 = x3 + PI/2
SINF_X_PLUS_PI_2_CORE_X16
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("stp q2, q3, [%[dst], #0x20]")
__ASM_EMIT("add %[dst], %[dst], #0x40")
__ASM_EMIT("b.hs 1b")
__ASM_EMIT("2:")
// x8 block
__ASM_EMIT("adds %[count], %[count], #8")
__ASM_EMIT("b.lt 4f")
__ASM_EMIT("ldp q0, q1, [%[dst], #0x00]") // v0 = x0, v1 = x1
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v16.4s") // v1 = x1 + PI/2
SINF_X_PLUS_PI_2_CORE_X8
__ASM_EMIT("sub %[count], %[count], #8")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("add %[dst], %[dst], #0x20")
__ASM_EMIT("4:")
// x4 block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.lt 6f")
__ASM_EMIT("ldr q0, [%[dst], #0x00]") // v0 = x0
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("sub %[count], %[count], #4")
__ASM_EMIT("str q0, [%[dst], #0x00]")
__ASM_EMIT("add %[dst], %[dst], #0x10")
__ASM_EMIT("6:")
// Tail: 1x-3x block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.ls 14f")
__ASM_EMIT("tst %[count], #1")
__ASM_EMIT("b.eq 8f")
__ASM_EMIT("ld1 {v0.s}[0], [%[dst]]")
__ASM_EMIT("add %[dst], %[dst], #0x04")
__ASM_EMIT("8:")
__ASM_EMIT("tst %[count], #2")
__ASM_EMIT("b.eq 10f")
__ASM_EMIT("ld1 {v0.d}[1], [%[dst]]")
__ASM_EMIT("10:")
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("tst %[count], #1")
__ASM_EMIT("b.eq 12f")
__ASM_EMIT("sub %[dst], %[dst], #0x04")
__ASM_EMIT("st1 {v0.s}[0], [%[dst]]")
__ASM_EMIT("add %[dst], %[dst], #0x04")
__ASM_EMIT("12:")
__ASM_EMIT("tst %[count], #2")
__ASM_EMIT("b.eq 14f")
__ASM_EMIT("st1 {v0.d}[1], [%[dst]]")
// End
__ASM_EMIT("14:")
: [dst] "+r" (dst), [count] "+r" (count)
: [S2C] "r" (&sinf_const[0])
: "cc", "memory",
"v0", "v1", "v2", "v3",
"v4", "v5", "v6", "v7",
"v8", "v9", "v10", "v11",
"v12", "v13", "v14", "v15",
"v16", "v17", "v18", "v19",
"v20", "v21", "v22", "v23"
);
}
void sinf2(float *dst, const float *src, size_t count)
{
ARCH_AARCH64_ASM(
__ASM_EMIT("ldp q16, q17, [%[S2C], #0x00]") /* v16 = PI/2, v17 = PI */
__ASM_EMIT("ldp q18, q19, [%[S2C], #0x20]") /* v18 = 1/(2*PI), v19 = C0 */
__ASM_EMIT("ldp q20, q21, [%[S2C], #0x40]") /* v20 = C1, v21 = C2 */
__ASM_EMIT("ldp q22, q23, [%[S2C], #0x60]") /* v22 = C3, v23 = C4 */
// x16 blocks
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("b.lo 2f")
__ASM_EMIT("1:")
__ASM_EMIT("ldp q0, q1, [%[src], #0x00]") // v0 = x0, v1 = x1
__ASM_EMIT("ldp q2, q3, [%[src], #0x20]") // v2 = x2, v3 = x3
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v16.4s") // v1 = x1 + PI/2
__ASM_EMIT("fadd v2.4s, v2.4s, v16.4s") // v2 = x2 + PI/2
__ASM_EMIT("fadd v3.4s, v3.4s, v16.4s") // v3 = x3 + PI/2
SINF_X_PLUS_PI_2_CORE_X16
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("stp q2, q3, [%[dst], #0x20]")
__ASM_EMIT("add %[src], %[src], #0x40")
__ASM_EMIT("add %[dst], %[dst], #0x40")
__ASM_EMIT("b.hs 1b")
__ASM_EMIT("2:")
// x8 block
__ASM_EMIT("adds %[count], %[count], #8")
__ASM_EMIT("b.lt 4f")
__ASM_EMIT("ldp q0, q1, [%[src], #0x00]") // v0 = x0, v1 = x1
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v16.4s") // v1 = x1 + PI/2
SINF_X_PLUS_PI_2_CORE_X8
__ASM_EMIT("sub %[count], %[count], #8")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("add %[src], %[src], #0x20")
__ASM_EMIT("add %[dst], %[dst], #0x20")
__ASM_EMIT("4:")
// x4 block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.lt 6f")
__ASM_EMIT("ldr q0, [%[src], #0x00]") // v0 = x0
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("sub %[count], %[count], #4")
__ASM_EMIT("str q0, [%[dst], #0x00]")
__ASM_EMIT("add %[src], %[src], #0x10")
__ASM_EMIT("add %[dst], %[dst], #0x10")
__ASM_EMIT("6:")
// Tail: 1x-3x block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.ls 14f")
__ASM_EMIT("tst %[count], #1")
__ASM_EMIT("b.eq 8f")
__ASM_EMIT("ld1 {v0.s}[0], [%[src]]")
__ASM_EMIT("add %[src], %[src], #0x04")
__ASM_EMIT("8:")
__ASM_EMIT("tst %[count], #2")
__ASM_EMIT("b.eq 10f")
__ASM_EMIT("ld1 {v0.d}[1], [%[src]]")
__ASM_EMIT("10:")
__ASM_EMIT("fadd v0.4s, v0.4s, v16.4s") // v0 = x0 + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("tst %[count], #1")
__ASM_EMIT("b.eq 12f")
__ASM_EMIT("st1 {v0.s}[0], [%[dst]]")
__ASM_EMIT("add %[dst], %[dst], #0x04")
__ASM_EMIT("12:")
__ASM_EMIT("tst %[count], #2")
__ASM_EMIT("b.eq 14f")
__ASM_EMIT("st1 {v0.d}[1], [%[dst]]")
// End
__ASM_EMIT("14:")
: [dst] "+r" (dst), [src] "+r" (src), [count] "+r" (count)
: [S2C] "r" (&sinf_const[0])
: "cc", "memory",
"v0", "v1", "v2", "v3",
"v4", "v5", "v6", "v7",
"v8", "v9", "v10", "v11",
"v12", "v13", "v14", "v15",
"v16", "v17", "v18", "v19",
"v20", "v21", "v22", "v23"
);
}
void sinf_kp1(float *dst, float k, float p, size_t count)
{
ARCH_AARCH64_ASM(
__ASM_EMIT("ldp q16, q17, [%[S2C], #0x00]") /* v16 = PI/2, v17 = PI */
__ASM_EMIT("ldp q18, q19, [%[S2C], #0x20]") /* v18 = 1/(2*PI), v19 = C0 */
__ASM_EMIT("ldp q20, q21, [%[S2C], #0x40]") /* v20 = C1, v21 = C2 */
__ASM_EMIT("ldp q22, q23, [%[S2C], #0x60]") /* v22 = C3, v23 = C4 */
__ASM_EMIT("ld1r {v28.4s}, [%[k]]") /* v28 = k k k k */
__ASM_EMIT("ld1r {v29.4s}, [%[p]]") /* v29 = p p p p */
__ASM_EMIT("ldp q24, q25, [%[S2KP], #0x00]") /* v24 = 0 1 2 3, v25 = 4 5 6 7 */
__ASM_EMIT("ldp q26, q27, [%[S2KP], #0x20]") /* v26 = 8 8 8 8, v27 = 4 4 4 4 */
__ASM_EMIT("fadd v29.4s, v29.4s, v16.4s") /* v29 = p + PI/2 ... */
// x16 blocks
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("b.lo 2f")
__ASM_EMIT("1:")
__ASM_EMIT("fmul v0.4s, v28.4s, v24.4s") // v0 = k*x0
__ASM_EMIT("fmul v1.4s, v28.4s, v25.4s") // v1 = k*x1
__ASM_EMIT("fadd v24.4s, v24.4s, v26.4s") // v24 = i0 + 8
__ASM_EMIT("fadd v25.4s, v25.4s, v26.4s") // v25 = i1 + 8
__ASM_EMIT("fadd v0.4s, v0.4s, v29.4s") // v0 = k*x0 + p + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v29.4s") // v1 = k*x1 + p + PI/2
__ASM_EMIT("fmul v2.4s, v28.4s, v24.4s") // v2 = k*x2
__ASM_EMIT("fmul v3.4s, v28.4s, v25.4s") // v3 = k*x3
__ASM_EMIT("fadd v24.4s, v24.4s, v26.4s") // v24 = i0 + 8
__ASM_EMIT("fadd v25.4s, v25.4s, v26.4s") // v25 = i1 + 8
__ASM_EMIT("fadd v2.4s, v2.4s, v29.4s") // v2 = k*x2 + p + PI/2
__ASM_EMIT("fadd v3.4s, v3.4s, v29.4s") // v3 = k*x3 + p + PI/2
SINF_X_PLUS_PI_2_CORE_X16
__ASM_EMIT("subs %[count], %[count], #16")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("stp q2, q3, [%[dst], #0x20]")
__ASM_EMIT("add %[dst], %[dst], #0x40")
__ASM_EMIT("b.hs 1b")
__ASM_EMIT("2:")
// x8 block
__ASM_EMIT("adds %[count], %[count], #8")
__ASM_EMIT("b.lt 4f")
__ASM_EMIT("fmul v0.4s, v28.4s, v24.4s") // v0 = k*x0
__ASM_EMIT("fmul v1.4s, v28.4s, v25.4s") // v1 = k*x1
__ASM_EMIT("fadd v24.4s, v24.4s, v26.4s") // v24 = i0 + 8
__ASM_EMIT("fadd v25.4s, v25.4s, v26.4s") // v25 = i1 + 8
__ASM_EMIT("fadd v0.4s, v0.4s, v29.4s") // v0 = k*x0 + p + PI/2
__ASM_EMIT("fadd v1.4s, v1.4s, v29.4s") // v1 = k*x1 + p + PI/2
SINF_X_PLUS_PI_2_CORE_X8
__ASM_EMIT("sub %[count], %[count], #8")
__ASM_EMIT("stp q0, q1, [%[dst], #0x00]")
__ASM_EMIT("add %[dst], %[dst], #0x20")
__ASM_EMIT("4:")
// x4 block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.lt 6f")
__ASM_EMIT("fmul v0.4s, v28.4s, v24.4s") // v0 = k*x0
__ASM_EMIT("fadd v24.4s, v24.4s, v27.4s") // v24 = i0 + 4
__ASM_EMIT("fadd v0.4s, v0.4s, v29.4s") // v0 = k*x0 + p + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("sub %[count], %[count], #4")
__ASM_EMIT("str q0, [%[dst], #0x00]")
__ASM_EMIT("add %[dst], %[dst], #0x10")
__ASM_EMIT("6:")
// Tail: 1x-3x block
__ASM_EMIT("adds %[count], %[count], #4")
__ASM_EMIT("b.ls 10f")
__ASM_EMIT("fmul v0.4s, v28.4s, v24.4s") // v0 = k*x0
__ASM_EMIT("fadd v0.4s, v0.4s, v29.4s") // v0 = k*x0 + p + PI/2
SINF_X_PLUS_PI_2_CORE_X4
__ASM_EMIT("tst %[count], #1")
__ASM_EMIT("b.eq 8f")
__ASM_EMIT("st1 {v0.s}[0], [%[dst]]")
__ASM_EMIT("add %[dst], %[dst], #0x04")
__ASM_EMIT("ext v0.16b, v0.16b, v0.16b, #4") // v0 = S1 S2 S3 S0
__ASM_EMIT("8:")
__ASM_EMIT("tst %[count], #2")
__ASM_EMIT("b.eq 10f")
__ASM_EMIT("st1 {v0.d}[0], [%[dst]]")
// End
__ASM_EMIT("10:")
: [dst] "+r" (dst), [count] "+r" (count)
: [k] "r" (&k),
[p] "r" (&p),
[S2C] "r" (&sinf_const[0]),
[S2KP] "r" (&kp_gen_const[0])
: "cc", "memory",
"v0", "v1", "v2", "v3",
"v4", "v5", "v6", "v7",
"v8", "v9", "v10", "v11",
"v12", "v13", "v14", "v15",
"v16", "v17", "v18", "v19",
"v20", "v21", "v22", "v23",
"v24", "v25", "v26", "v27",
"v28", "v29"
);
}
} /* namespace asimd */
} /* namespace lsp */
#endif /* PRIVATE_DSP_ARCH_AARCH64_ASIMD_PMATH_SIN_H_ */