/****************************************************************************** * * * Copyright (C) 2023 The Android Open Source Project * * Licensed under the Apache License, Version 2.0 (the "License"); * you may not use this file except in compliance with the License. * You may obtain a copy of the License at: * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. * ***************************************************************************** * Originally developed and contributed by Ittiam Systems Pvt. Ltd, Bangalore */ #include #include #include "ixheaac_constants.h" #include "ixheaace_constants.h" #include "iusace_basic_ops_flt.h" #include "ixheaace_common_utils.h" #include "ixheaac_fft_ifft_rom.h" #include "ixheaac_basic_ops32.h" #include "ixheaac_basic_ops40.h" #include "ixheaac_basic_ops.h" #define DIG_REV(i, m, j) \ do { \ unsigned _ = (i); \ _ = ((_ & 0x33333333) << 2) | ((_ & ~0x33333333) >> 2); \ _ = ((_ & 0x0F0F0F0F) << 4) | ((_ & ~0x0F0F0F0F) >> 4); \ _ = ((_ & 0x00FF00FF) << 8) | ((_ & ~0x00FF00FF) >> 8); \ (j) = _ >> (m); \ } while (0) #define CPLX_MPY_FFT(re, im, a, b, c, d) \ do { \ re = ((a * c) - (b * d)); \ im = ((a * d) + (b * c)); \ } while (0) #define CPLX_MPY_IFFT(re, im, a, b, c, d) \ do { \ re = ((a * c) + (b * d)); \ im = (-(a * d) + (b * c)); \ } while (0) VOID ixheaace_hbe_apply_ifft_7(FLOAT32 *ptr_inp, FLOAT32 *ptr_op) { FLOAT32 x0r, x1r, x2r, x3r, x4r, x5r, x6r, x7r, x8r; FLOAT32 x0i, x1i, x2i, x3i, x4i, x5i, x6i, x7i, x8i; FLOAT32 y0r, y1r, y2r, y3r, y4r, y5r, y6r, y7r, y8r; FLOAT32 y0i, y1i, y2i, y3i, y4i, y5i, y6i, y7i, y8i; /* * Node 1 of Winograd FFT for 7 point * * 1 0 0 0 0 0 0 * 0 1 0 0 0 0 1 * 0 1 0 0 0 0 -1 * 0 0 1 0 0 1 0 * 0 0 1 0 0 -1 0 * 0 0 0 1 1 0 0 * 0 0 0 -1 1 0 0 * */ x0r = ptr_inp[0]; x0i = ptr_inp[1]; x1r = ptr_inp[2] + ptr_inp[12]; x1i = ptr_inp[3] + ptr_inp[13]; x2r = ptr_inp[2] - ptr_inp[12]; x2i = ptr_inp[3] - ptr_inp[13]; x3r = ptr_inp[4] + ptr_inp[10]; x3i = ptr_inp[5] + ptr_inp[11]; x4r = ptr_inp[4] - ptr_inp[10]; x4i = ptr_inp[5] - ptr_inp[11]; x5r = ptr_inp[8] + ptr_inp[6]; x5i = ptr_inp[9] + ptr_inp[7]; x6r = ptr_inp[8] - ptr_inp[6]; x6i = ptr_inp[9] - ptr_inp[7]; /* * Node 2 of Winograd FFT for 7 point * * 1 0 0 0 0 0 0 * 0 1 0 1 0 1 0 * 0 1 0 -1 0 0 0 * 0 -1 0 0 0 1 0 * 0 0 0 1 0 -1 0 * 0 0 1 0 1 0 1 * 0 0 1 0 -1 0 0 * 0 0 -1 0 0 0 1 * 0 0 0 0 1 0 -1 * */ y0r = x0r; y0i = x0i; y1r = x1r + x3r + x5r; y1i = x1i + x3i + x5i; y2r = x1r - x3r; y2i = x1i - x3i; y3r = x5r - x1r; y3i = x5i - x1i; y4r = x3r - x5r; y4i = x3i - x5i; y5r = x2r + x4r + x6r; y5i = x2i + x4i + x6i; y6r = x2r - x4r; y6i = x2i - x4i; y7r = x6r - x2r; y7i = x6i - x2i; y8r = x4r - x6r; y8i = x4i - x6i; /* * Node 3 of Winograd FFT for 7 point * * 1 1 0 0 0 0 0 0 0 * 1 c70 0 0 0 0 0 0 0 * 0 0 c71 0 0 0 0 0 0 * 0 0 0 c72 0 0 0 0 0 * 0 0 0 0 c73 0 0 0 0 * 0 0 0 0 0 jc74 0 0 0 * 0 0 0 0 0 0 jc75 0 0 * 0 0 0 0 0 0 0 jc76 0 * 0 0 0 0 0 0 0 0 jc77 * */ x0r = y0r + y1r; x0i = y0i + y1i; x1r = y0r + C70 * y1r; x1i = y0i + C70 * y1i; x2r = C71 * y2r; x2i = C71 * y2i; x3r = C72 * y3r; x3i = C72 * y3i; x4r = C73 * y4r; x4i = C73 * y4i; x5r = C74 * y5i; x5i = -C74 * y5r; x6r = C75 * y6i; x6i = -C75 * y6r; x7r = C76 * y7i; x7i = -C76 * y7r; x8r = C77 * y8i; x8i = -C77 * y8r; /* * Node 4 of Winograd FFT for 7 point * * 1 0 0 0 0 0 0 0 0 * 0 1 1 0 1 0 0 0 0 * 0 1 -1 -1 0 0 0 0 0 * 0 1 0 1 -1 0 0 0 0 * 0 0 0 0 0 1 1 0 1 * 0 0 0 0 0 1 -1 -1 0 * 0 0 0 0 0 1 0 1 -1 * */ y0r = x0r; y0i = x0i; y1r = x1r + x2r + x4r; y1i = x1i + x2i + x4i; y2r = x1r - x2r - x3r; y2i = x1i - x2i - x3i; y3r = x1r + x3r - x4r; y3i = x1i + x3i - x4i; y4r = x5r + x6r + x8r; y4i = x5i + x6i + x8i; y5r = x5r - x6r - x7r; y5i = x5i - x6i - x7i; y6r = x5r + x7r - x8r; y6i = x5i + x7i - x8i; /* * Node 5 of Winograd FFT for 7 point * * 1 0 0 0 0 0 0 * 0 1 0 0 1 0 0 * 0 0 0 1 0 0 1 * 0 0 1 0 0 -1 0 * 0 0 1 0 0 1 0 * 0 0 0 1 0 0 -1 * 0 1 0 0 -1 0 0 * */ x0r = y0r; x0i = y0i; x1r = y1r + y4r; x1i = y1i + y4i; x2r = y3r + y6r; x2i = y3i + y6i; x3r = y2r - y5r; x3i = y2i - y5i; x4r = y2r + y5r; x4i = y2i + y5i; x5r = y3r - y6r; x5i = y3i - y6i; x6r = y1r - y4r; x6i = y1i - y4i; ptr_op[0] = x0r; ptr_op[1] = x0i; ptr_op[2] = x1r; ptr_op[3] = x1i; ptr_op[4] = x2r; ptr_op[5] = x2i; ptr_op[6] = x3r; ptr_op[7] = x3i; ptr_op[8] = x4r; ptr_op[9] = x4i; ptr_op[10] = x5r; ptr_op[11] = x5i; ptr_op[12] = x6r; ptr_op[13] = x6i; } VOID ixheaace_hbe_apply_fft_3(FLOAT32 *ptr_inp, FLOAT32 *ptr_op, WORD32 i_sign) { FLOAT32 add_r, sub_r; FLOAT32 add_i, sub_i; FLOAT32 x_01_r, x_01_i, temp; FLOAT32 p1, p2, p3, p4; /* mu = PI / 3; The cos and sin values are in Q31 cosmu is 0.5 so used >> 1 instead of multiplication */ FLOAT64 sinmu; sinmu = -0.866025403784439 * (FLOAT64)i_sign; x_01_r = ptr_inp[0] + ptr_inp[2]; x_01_i = ptr_inp[1] + ptr_inp[3]; add_r = ptr_inp[2] + ptr_inp[4]; add_i = ptr_inp[3] + ptr_inp[5]; sub_r = ptr_inp[2] - ptr_inp[4]; sub_i = ptr_inp[3] - ptr_inp[5]; p1 = add_r / (FLOAT32)2.0; p4 = add_i / (FLOAT32)2.0; p2 = (FLOAT32)((FLOAT64)sub_i * sinmu); p3 = (FLOAT32)((FLOAT64)sub_r * sinmu); temp = ptr_inp[0] - p1; ptr_op[0] = x_01_r + ptr_inp[4]; ptr_op[1] = x_01_i + ptr_inp[5]; ptr_op[2] = temp + p2; ptr_op[3] = (ptr_inp[1] - p3) - p4; ptr_op[4] = temp - p2; ptr_op[5] = (ptr_inp[1] + p3) - p4; } VOID ixheaace_hbe_apply_tw_mult_ifft(FLOAT32 *ptr_inp, FLOAT32 *ptr_op, WORD32 dim1, WORD32 dim2, const FLOAT32 *ptr_tw) { FLOAT32 accu1, accu2; WORD32 i, j; WORD32 step_val = (dim2 - 1) << 1; for (i = 0; i < (dim2); i++) { ptr_op[0] = ptr_inp[0]; ptr_op[1] = ptr_inp[1]; ptr_op += 2; ptr_inp += 2; } for (j = 0; j < (dim1 - 1); j++) { ptr_op[0] = ptr_inp[0]; ptr_op[1] = ptr_inp[1]; ptr_inp += 2; ptr_op += 2; for (i = 0; i < (dim2 - 1); i++) { CPLX_MPY_IFFT(accu1, accu2, ptr_inp[2 * i + 0], ptr_inp[2 * i + 1], ptr_tw[2 * i + 1], ptr_tw[2 * i]); ptr_op[2 * i + 0] = accu1; ptr_op[2 * i + 1] = accu2; } ptr_inp += step_val; ptr_op += step_val; ptr_tw += (dim2 - 1) * 2; } } VOID ixheaace_hbe_apply_tw_mult_fft(FLOAT32 *ptr_inp, FLOAT32 *ptr_op, WORD32 dim1, WORD32 dim2, const FLOAT32 *ptr_tw) { FLOAT32 accu1, accu2; WORD32 i, j; WORD32 step_val = (dim2 - 1) << 1; for (i = 0; i < (dim2); i++) { ptr_op[0] = ptr_inp[0]; ptr_op[1] = ptr_inp[1]; ptr_op += 2; ptr_inp += 2; } for (j = 0; j < (dim1 - 1); j++) { ptr_op[0] = ptr_inp[0]; ptr_op[1] = ptr_inp[1]; ptr_inp += 2; ptr_op += 2; for (i = 0; i < (dim2 - 1); i++) { CPLX_MPY_FFT(accu1, accu2, ptr_inp[2 * i + 0], ptr_inp[2 * i + 1], ptr_tw[2 * i + 1], ptr_tw[2 * i]); ptr_op[2 * i + 0] = accu1; ptr_op[2 * i + 1] = accu2; } ptr_inp += step_val; ptr_op += step_val; ptr_tw += (dim2 - 1) * 2; } } VOID ixheaace_hbe_apply_cfftn(FLOAT32 re[], FLOAT32 *ptr_scratch, WORD32 n_pass, WORD32 i_sign) { WORD32 i, j, k, n_stages, h2; FLOAT32 x0r, x0i, x1r, x1i, x2r, x2i, x3r, x3i; WORD32 del, nodespacing, in_loop_cnt; WORD32 not_power_4; WORD32 dig_rev_shift; WORD32 mpass = n_pass; WORD32 npoints = n_pass; const FLOAT64 *ptr_w; FLOAT32 *ptr_x = ptr_scratch; FLOAT32 *y = ptr_scratch + (2 * n_pass); FLOAT32 *ptr_y = y; dig_rev_shift = ixheaac_norm32(mpass) + 1 - 16; n_stages = 30 - ixheaac_norm32(mpass); /* log2(npoints), if npoints=2^m */ not_power_4 = n_stages & 1; n_stages = n_stages >> 1; ptr_w = ixheaac_twid_tbl_fft_double; ptr_x = re; dig_rev_shift = MAX(dig_rev_shift, 0); if (i_sign == -1) { for (i = 0; i < npoints; i += 4) { FLOAT32 *ptr_inp = ptr_x; FLOAT32 tmk; DIG_REV(i, dig_rev_shift, h2); if (not_power_4) { h2 += 1; h2 &= ~1; } ptr_inp += (h2); x0r = *ptr_inp; x0i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x1r = *ptr_inp; x1i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x2r = *ptr_inp; x2i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x3r = *ptr_inp; x3i = *(ptr_inp + 1); x0r = x0r + x2r; x0i = x0i + x2i; tmk = x0r - x2r; x2r = tmk - x2r; tmk = x0i - x2i; x2i = tmk - x2i; x1r = x1r + x3r; x1i = x1i + x3i; tmk = x1r - x3r; x3r = tmk - x3r; tmk = x1i - x3i; x3i = tmk - x3i; x0r = x0r + x1r; x0i = x0i + x1i; tmk = x0r - x1r; x1r = tmk - x1r; tmk = x0i - x1i; x1i = tmk - x1i; x2r = x2r + x3i; x2i = x2i - x3r; tmk = x2r - x3i; x3i = tmk - x3i; tmk = x2i + x3r; x3r = tmk + x3r; *ptr_y++ = x0r; *ptr_y++ = x0i; *ptr_y++ = x2r; *ptr_y++ = x2i; *ptr_y++ = x1r; *ptr_y++ = x1i; *ptr_y++ = x3i; *ptr_y++ = x3r; } ptr_y -= 2 * npoints; del = 4; nodespacing = 64; in_loop_cnt = npoints >> 4; for (i = n_stages - 1; i > 0; i--) { const FLOAT64 *ptr_twiddle = ptr_w; FLOAT32 *data = ptr_y; FLOAT64 w_1, w_2, w_3, w_4, w_5, w_6; WORD32 sec_loop_cnt; for (k = in_loop_cnt; k != 0; k--) { x0r = (*data); x0i = (*(data + 1)); data += ((SIZE_T)del << 1); x1r = (*data); x1i = (*(data + 1)); data += ((SIZE_T)del << 1); x2r = (*data); x2i = (*(data + 1)); data += ((SIZE_T)del << 1); x3r = (*data); x3i = (*(data + 1)); data -= 3 * (del << 1); x0r = x0r + x2r; x0i = x0i + x2i; x2r = x0r - (x2r * 2); x2i = x0i - (x2i * 2); x1r = x1r + x3r; x1i = x1i + x3i; x3r = x1r - (x3r * 2); x3i = x1i - (x3i * 2); x0r = x0r + x1r; x0i = x0i + x1i; x1r = x0r - (x1r * 2); x1i = x0i - (x1i * 2); x2r = x2r + x3i; x2i = x2i - x3r; x3i = x2r - (x3i * 2); x3r = x2i + (x3r * 2); *data = x0r; *(data + 1) = x0i; data += ((SIZE_T)del << 1); *data = x2r; *(data + 1) = x2i; data += ((SIZE_T)del << 1); *data = x1r; *(data + 1) = x1i; data += ((SIZE_T)del << 1); *data = x3i; *(data + 1) = x3r; data += ((SIZE_T)del << 1); } data = ptr_y + 2; sec_loop_cnt = (nodespacing * del); sec_loop_cnt = (sec_loop_cnt / 4) + (sec_loop_cnt / 8) - (sec_loop_cnt / 16) + (sec_loop_cnt / 32) - (sec_loop_cnt / 64) + (sec_loop_cnt / 128) - (sec_loop_cnt / 256); for (j = nodespacing; j <= sec_loop_cnt; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1)); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 257); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1)); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 257); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r1, w_1) - ixheaace_dmult((FLOAT64)x1i1, w_4)); x1i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x1r1, w_4), (FLOAT64)x1i1, w_1); x1r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x2r1, w_2) - ixheaace_dmult((FLOAT64)x2i1, w_5)); x2i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x2r1, w_5), (FLOAT64)x2i1, w_2); x2r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x3r1, w_3) - ixheaace_dmult((FLOAT64)x3i1, w_6)); x3i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x3r1, w_6), (FLOAT64)x3i1, w_3); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 + (x3i1); x2i1 = x2i1 - (x3r1); x3i1 = x2r1 - (x3i1 * 2); x3r1 = x2i1 + (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j <= (nodespacing * del) >> 1; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1)); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 257); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 256); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 1); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r1, w_1) - ixheaace_dmult((FLOAT64)x1i1, w_4)); x1i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x1r1, w_4), (FLOAT64)x1i1, w_1); x1r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x2r1, w_2) - ixheaace_dmult((FLOAT64)x2i1, w_5)); x2i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x2r1, w_5), (FLOAT64)x2i1, w_2); x2r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x3r1, w_6) + ixheaace_dmult((FLOAT64)x3i1, w_3)); x3i1 = (FLOAT32)(-ixheaace_dmult((FLOAT64)x3r1, w_3) + ixheaace_dmult((FLOAT64)x3i1, w_6)); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 + (x3i1); x2i1 = x2i1 - (x3r1); x3i1 = x2r1 - (x3i1 * 2); x3r1 = x2i1 + (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j <= sec_loop_cnt * 2; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1) - 256); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 1); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 256); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 1); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r1, w_1) - ixheaace_dmult((FLOAT64)x1i1, w_4)); x1i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult(x1r1, w_4), x1i1, w_1); x1r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x2r1, w_5) + ixheaace_dmult((FLOAT64)x2i1, w_2)); x2i1 = (FLOAT32)(-ixheaace_dmult(x2r1, w_2) + ixheaace_dmult(x2i1, w_5)); x2r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x3r1, w_6) + ixheaace_dmult((FLOAT64)x3i1, w_3)); x3i1 = (FLOAT32)(-ixheaace_dmult((FLOAT64)x3r1, w_3) + ixheaace_dmult((FLOAT64)x3i1, w_6)); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 + (x3i1); x2i1 = x2i1 - (x3r1); x3i1 = x2r1 - (x3i1 * 2); x3r1 = x2i1 + (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j < nodespacing * del; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1) - 256); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 1); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 512); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 512 + 257); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r1, w_1) - ixheaace_dmult((FLOAT64)x1i1, w_4)); x1i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x1r1, w_4), (FLOAT64)x1i1, w_1); x1r1 = tmp; tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x2r1, w_5) + ixheaace_dmult((FLOAT64)x2i1, w_2)); x2i1 = (FLOAT32)(-ixheaace_dmult((FLOAT64)x2r1, w_2) + ixheaace_dmult((FLOAT64)x2i1, w_5)); x2r1 = tmp; tmp = (FLOAT32)(-ixheaace_dmult((FLOAT64)x3r1, w_3) + ixheaace_dmult((FLOAT64)x3i1, w_6)); x3i1 = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x3r1, w_6), (FLOAT64)x3i1, w_3); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 - x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 + (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 + (x3i1); x2i1 = x2i1 - (x3r1); x3i1 = x2r1 - (x3i1 * 2); x3r1 = x2i1 + (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } nodespacing >>= 2; del <<= 2; in_loop_cnt >>= 2; } if (not_power_4) { const double *ptr_twiddle = ptr_w; nodespacing <<= 1; for (j = del / 2; j != 0; j--) { FLOAT64 w_1 = *ptr_twiddle; FLOAT64 w_4 = *(ptr_twiddle + 257); FLOAT32 tmp; ptr_twiddle += nodespacing; x0r = *ptr_y; x0i = *(ptr_y + 1); ptr_y += ((SIZE_T)del << 1); x1r = *ptr_y; x1i = *(ptr_y + 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r, w_1) - ixheaace_dmult((FLOAT64)x1i, w_4)); x1i = (FLOAT32)ixheaace_dmac(ixheaace_dmult((FLOAT64)x1r, w_4), (FLOAT64)x1i, w_1); x1r = tmp; *ptr_y = (x0r) - (x1r); *(ptr_y + 1) = (x0i) - (x1i); ptr_y -= ((SIZE_T)del << 1); *ptr_y = (x0r) + (x1r); *(ptr_y + 1) = (x0i) + (x1i); ptr_y += 2; } ptr_twiddle = ptr_w; for (j = del / 2; j != 0; j--) { FLOAT64 w_1 = *ptr_twiddle; FLOAT64 w_4 = *(ptr_twiddle + 257); FLOAT32 tmp; ptr_twiddle += nodespacing; x0r = *ptr_y; x0i = *(ptr_y + 1); ptr_y += ((SIZE_T)del << 1); x1r = *ptr_y; x1i = *(ptr_y + 1); tmp = (FLOAT32)(ixheaace_dmult((FLOAT64)x1r, w_4) + ixheaace_dmult((FLOAT64)x1i, w_1)); x1i = (FLOAT32)(-ixheaace_dmult((FLOAT64)x1r, w_1) + ixheaace_dmult((FLOAT64)x1i, w_4)); x1r = tmp; *ptr_y = (x0r) - (x1r); *(ptr_y + 1) = (x0i) - (x1i); ptr_y -= ((SIZE_T)del << 1); *ptr_y = (x0r) + (x1r); *(ptr_y + 1) = (x0i) + (x1i); ptr_y += 2; } } } /**********************IFFT******************************************/ else { for (i = 0; i < npoints; i += 4) { FLOAT32 *ptr_inp = ptr_x; DIG_REV(i, dig_rev_shift, h2); if (not_power_4) { h2 += 1; h2 &= ~1; } ptr_inp += (h2); x0r = *ptr_inp; x0i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x1r = *ptr_inp; x1i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x2r = *ptr_inp; x2i = *(ptr_inp + 1); ptr_inp += (npoints >> 1); x3r = *ptr_inp; x3i = *(ptr_inp + 1); x0r = x0r + x2r; x0i = x0i + x2i; x2r = x0r - (x2r * 2); x2i = x0i - (x2i * 2); x1r = x1r + x3r; x1i = x1i + x3i; x3r = x1r - (x3r * 2); x3i = x1i - (x3i * 2); x0r = x0r + x1r; x0i = x0i + x1i; x1r = x0r - (x1r * 2); x1i = x0i - (x1i * 2); x2r = x2r - x3i; x2i = x2i + x3r; x3i = x2r + (x3i * 2); x3r = x2i - (x3r * 2); *ptr_y++ = x0r; *ptr_y++ = x0i; *ptr_y++ = x2r; *ptr_y++ = x2i; *ptr_y++ = x1r; *ptr_y++ = x1i; *ptr_y++ = x3i; *ptr_y++ = x3r; } ptr_y -= 2 * npoints; del = 4; nodespacing = 64; in_loop_cnt = npoints >> 4; for (i = n_stages - 1; i > 0; i--) { const double *ptr_twiddle = ptr_w; float *data = ptr_y; double w_1, w_2, w_3, w_4, w_5, w_6; int sec_loop_cnt; for (k = in_loop_cnt; k != 0; k--) { x0r = (*data); x0i = (*(data + 1)); data += ((SIZE_T)del << 1); x1r = (*data); x1i = (*(data + 1)); data += ((SIZE_T)del << 1); x2r = (*data); x2i = (*(data + 1)); data += ((SIZE_T)del << 1); x3r = (*data); x3i = (*(data + 1)); data -= 3 * (del << 1); x0r = x0r + x2r; x0i = x0i + x2i; x2r = x0r - (x2r * 2); x2i = x0i - (x2i * 2); x1r = x1r + x3r; x1i = x1i + x3i; x3r = x1r - (x3r * 2); x3i = x1i - (x3i * 2); x0r = x0r + x1r; x0i = x0i + x1i; x1r = x0r - (x1r * 2); x1i = x0i - (x1i * 2); x2r = x2r - x3i; x2i = x2i + x3r; x3i = x2r + (x3i * 2); x3r = x2i - (x3r * 2); *data = x0r; *(data + 1) = x0i; data += ((SIZE_T)del << 1); *data = x2r; *(data + 1) = x2i; data += ((SIZE_T)del << 1); *data = x1r; *(data + 1) = x1i; data += ((SIZE_T)del << 1); *data = x3i; *(data + 1) = x3r; data += ((SIZE_T)del << 1); } data = ptr_y + 2; sec_loop_cnt = (nodespacing * del); sec_loop_cnt = (sec_loop_cnt / 4) + (sec_loop_cnt / 8) - (sec_loop_cnt / 16) + (sec_loop_cnt / 32) - (sec_loop_cnt / 64) + (sec_loop_cnt / 128) - (sec_loop_cnt / 256); for (j = nodespacing; j <= sec_loop_cnt; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1)); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 257); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1)); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 257); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(((FLOAT64)x1r1 * w_1) + ((FLOAT64)x1i1 * w_4)); x1i1 = (FLOAT32)(-((FLOAT64)x1r1 * w_4) + (FLOAT64)x1i1 * w_1); x1r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x2r1 * w_2) + ((FLOAT64)x2i1 * w_5)); x2i1 = (FLOAT32)(-((FLOAT64)x2r1 * w_5) + (FLOAT64)x2i1 * w_2); x2r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x3r1 * w_3) + ((FLOAT64)x3i1 * w_6)); x3i1 = (FLOAT32)(-((FLOAT64)x3r1 * w_6) + (FLOAT64)x3i1 * w_3); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 - (x3i1); x2i1 = x2i1 + (x3r1); x3i1 = x2r1 + (x3i1 * 2); x3r1 = x2i1 - (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j <= (nodespacing * del) >> 1; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1)); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 257); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 256); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 1); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(((FLOAT64)x1r1 * w_1) + ((FLOAT64)x1i1 * w_4)); x1i1 = (FLOAT32)(-((FLOAT64)x1r1 * w_4) + (FLOAT64)x1i1 * w_1); x1r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x2r1 * w_2) + ((FLOAT64)x2i1 * w_5)); x2i1 = (FLOAT32)(-((FLOAT64)x2r1 * w_5) + (FLOAT64)x2i1 * w_2); x2r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x3r1 * w_6) - ((FLOAT64)x3i1 * w_3)); x3i1 = (FLOAT32)(((FLOAT64)x3r1 * w_3) + ((FLOAT64)x3i1 * w_6)); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 - (x3i1); x2i1 = x2i1 + (x3r1); x3i1 = x2r1 + (x3i1 * 2); x3r1 = x2i1 - (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j <= sec_loop_cnt * 2; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1) - 256); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 1); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 256); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) + 1); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(((FLOAT64)x1r1 * w_1) + ((FLOAT64)x1i1 * w_4)); x1i1 = (FLOAT32)(-((FLOAT64)x1r1 * w_4) + (FLOAT64)x1i1 * w_1); x1r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x2r1 * w_5) - ((FLOAT64)x2i1 * w_2)); x2i1 = (FLOAT32)(((FLOAT64)x2r1 * w_2) + ((FLOAT64)x2i1 * w_5)); x2r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x3r1 * w_6) - ((FLOAT64)x3i1 * w_3)); x3i1 = (FLOAT32)(((FLOAT64)x3r1 * w_3) + ((FLOAT64)x3i1 * w_6)); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 + x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 - (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 - (x3i1); x2i1 = x2i1 + (x3r1); x3i1 = x2r1 + (x3i1 * 2); x3r1 = x2i1 - (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } for (; j < nodespacing * del; j += nodespacing) { w_1 = *(ptr_twiddle + j); w_4 = *(ptr_twiddle + j + 257); w_2 = *(ptr_twiddle + ((SIZE_T)j << 1) - 256); w_5 = *(ptr_twiddle + ((SIZE_T)j << 1) + 1); w_3 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 512); w_6 = *(ptr_twiddle + j + ((SIZE_T)j << 1) - 512 + 257); for (k = in_loop_cnt; k != 0; k--) { FLOAT32 tmp; FLOAT32 x0r1, x0i1, x1r1, x1i1, x2r1, x2i1, x3r1, x3i1; /*x0 is loaded later to avoid register crunch*/ data += ((SIZE_T)del << 1); x1r1 = *data; x1i1 = *(data + 1); data += ((SIZE_T)del << 1); x2r1 = *data; x2i1 = *(data + 1); data += ((SIZE_T)del << 1); x3r1 = *data; x3i1 = *(data + 1); data -= 3 * (del << 1); tmp = (FLOAT32)(((FLOAT64)x1r1 * w_1) + ((FLOAT64)x1i1 * w_4)); x1i1 = (FLOAT32)(-((FLOAT64)x1r1 * w_4) + (FLOAT64)x1i1 * w_1); x1r1 = tmp; tmp = (FLOAT32)(((FLOAT64)x2r1 * w_5) - ((FLOAT64)x2i1 * w_2)); x2i1 = (FLOAT32)(((FLOAT64)x2r1 * w_2) + ((FLOAT64)x2i1 * w_5)); x2r1 = tmp; tmp = (FLOAT32)(-((FLOAT64)x3r1 * w_3) - ((FLOAT64)x3i1 * w_6)); x3i1 = (FLOAT32)(-((FLOAT64)x3r1 * w_6) + (FLOAT64)x3i1 * w_3); x3r1 = tmp; x0r1 = (*data); x0i1 = (*(data + 1)); x0r1 = x0r1 + (x2r1); x0i1 = x0i1 + (x2i1); x2r1 = x0r1 - (x2r1 * 2); x2i1 = x0i1 - (x2i1 * 2); x1r1 = x1r1 + x3r1; x1i1 = x1i1 - x3i1; x3r1 = x1r1 - (x3r1 * 2); x3i1 = x1i1 + (x3i1 * 2); x0r1 = x0r1 + (x1r1); x0i1 = x0i1 + (x1i1); x1r1 = x0r1 - (x1r1 * 2); x1i1 = x0i1 - (x1i1 * 2); x2r1 = x2r1 - (x3i1); x2i1 = x2i1 + (x3r1); x3i1 = x2r1 + (x3i1 * 2); x3r1 = x2i1 - (x3r1 * 2); *data = x0r1; *(data + 1) = x0i1; data += ((SIZE_T)del << 1); *data = x2r1; *(data + 1) = x2i1; data += ((SIZE_T)del << 1); *data = x1r1; *(data + 1) = x1i1; data += ((SIZE_T)del << 1); *data = x3i1; *(data + 1) = x3r1; data += ((SIZE_T)del << 1); } data -= 2 * npoints; data += 2; } nodespacing >>= 2; del <<= 2; in_loop_cnt >>= 2; } if (not_power_4) { const FLOAT64 *ptr_twiddle = ptr_w; nodespacing <<= 1; for (j = del / 2; j != 0; j--) { FLOAT64 w_1 = *ptr_twiddle; FLOAT64 w_4 = *(ptr_twiddle + 257); FLOAT32 tmp; ptr_twiddle += nodespacing; x0r = *ptr_y; x0i = *(ptr_y + 1); ptr_y += ((SIZE_T)del << 1); x1r = *ptr_y; x1i = *(ptr_y + 1); tmp = (FLOAT32)(((FLOAT64)x1r * w_1) + ((FLOAT64)x1i * w_4)); x1i = (FLOAT32)(-((FLOAT64)x1r * w_4) + (FLOAT64)x1i * w_1); x1r = tmp; *ptr_y = (x0r) - (x1r); *(ptr_y + 1) = (x0i) - (x1i); ptr_y -= ((SIZE_T)del << 1); *ptr_y = (x0r) + (x1r); *(ptr_y + 1) = (x0i) + (x1i); ptr_y += 2; } ptr_twiddle = ptr_w; for (j = del / 2; j != 0; j--) { FLOAT64 w_1 = *ptr_twiddle; FLOAT64 w_4 = *(ptr_twiddle + 257); FLOAT32 tmp; ptr_twiddle += nodespacing; x0r = *ptr_y; x0i = *(ptr_y + 1); ptr_y += ((SIZE_T)del << 1); x1r = *ptr_y; x1i = *(ptr_y + 1); tmp = (FLOAT32)(((FLOAT64)x1r * w_4) - ((FLOAT64)x1i * w_1)); x1i = (FLOAT32)(((FLOAT64)x1r * w_1) + ((FLOAT64)x1i * w_4)); x1r = tmp; *ptr_y = (x0r) - (x1r); *(ptr_y + 1) = (x0i) - (x1i); ptr_y -= ((SIZE_T)del << 1); *ptr_y = (x0r) + (x1r); *(ptr_y + 1) = (x0i) + (x1i); ptr_y += 2; } } } for (i = 0; i < n_pass; i++) { re[2 * i + 0] = y[2 * i + 0]; re[2 * i + 1] = y[2 * i + 1]; } } VOID ixheaace_hbe_apply_cfftn_gen(FLOAT32 in[], FLOAT32 *ptr_scratch, WORD32 n_pass, WORD32 i_sign) { WORD32 i, j; WORD32 m_points = n_pass; FLOAT32 *y, *re_3; FLOAT32 *ptr_x, *ptr_y; ptr_x = ptr_scratch; ptr_scratch += 2 * m_points; ptr_y = y = ptr_scratch; ptr_scratch += 4 * m_points; re_3 = ptr_scratch; ptr_scratch += 2 * m_points; WORD32 cnfac; WORD32 mpass = n_pass; cnfac = 0; while (mpass % 3 == 0) { mpass /= 3; cnfac++; } for (i = 0; i < 3 * cnfac; i++) { for (j = 0; j < mpass; j++) { re_3[2 * j + 0] = in[6 * j + 2 * i + 0]; re_3[2 * j + 1] = in[6 * j + 2 * i + 1]; } ixheaace_hbe_apply_cfftn(re_3, ptr_scratch, mpass, i_sign); for (j = 0; j < mpass; j++) { in[6 * j + 2 * i + 0] = re_3[2 * j + 0]; in[6 * j + 2 * i + 1] = re_3[2 * j + 1]; } } { FLOAT64 *ptr_w1r, *ptr_w1i; FLOAT32 tmp; ptr_w1r = (FLOAT64 *)ixheaac_twid_tbl_fft_ntwt3r; ptr_w1i = (FLOAT64 *)ixheaac_twid_tbl_fft_ntwt3i; if (i_sign < 0) { i = 0; while (i < n_pass) { tmp = (FLOAT32)((FLOAT64)in[2 * i + 0] * (*ptr_w1r) - (FLOAT64)in[2 * i + 1] * (*ptr_w1i)); in[2 * i + 1] = (FLOAT32)((FLOAT64)in[2 * i + 0] * (*ptr_w1i) + (FLOAT64)in[2 * i + 1] * (*ptr_w1r)); in[2 * i + 0] = tmp; ptr_w1r++; ptr_w1i++; tmp = (FLOAT32)((FLOAT64)in[2 * i + 2] * (*ptr_w1r) - (FLOAT64)in[2 * i + 3] * (*ptr_w1i)); in[2 * i + 3] = (FLOAT32)((FLOAT64)in[2 * i + 2] * (*ptr_w1i) + (FLOAT64)in[2 * i + 3] * (*ptr_w1r)); in[2 * i + 2] = tmp; ptr_w1r++; ptr_w1i++; tmp = (FLOAT32)((FLOAT64)in[2 * i + 4] * (*ptr_w1r) - (FLOAT64)in[2 * i + 5] * (*ptr_w1i)); in[2 * i + 5] = (FLOAT32)((FLOAT64)in[2 * i + 4] * (*ptr_w1i) + (FLOAT64)in[2 * i + 5] * (*ptr_w1r)); in[2 * i + 4] = tmp; ptr_w1r += 3 * (128 / mpass - 1) + 1; ptr_w1i += 3 * (128 / mpass - 1) + 1; i += 3; } } else { i = 0; while (i < n_pass) { tmp = (FLOAT32)((FLOAT64)in[2 * i + 0] * (*ptr_w1r) + (FLOAT64)in[2 * i + 1] * (*ptr_w1i)); in[2 * i + 1] = (FLOAT32)(-(FLOAT64)in[2 * i + 0] * (*ptr_w1i) + (FLOAT64)in[2 * i + 1] * (*ptr_w1r)); in[2 * i + 0] = tmp; ptr_w1r++; ptr_w1i++; tmp = (FLOAT32)((FLOAT64)in[2 * i + 2] * (*ptr_w1r) + (FLOAT64)in[2 * i + 3] * (*ptr_w1i)); in[2 * i + 3] = (FLOAT32)(-(FLOAT64)in[2 * i + 2] * (*ptr_w1i) + (FLOAT64)in[2 * i + 3] * (*ptr_w1r)); in[2 * i + 2] = tmp; ptr_w1r++; ptr_w1i++; tmp = (FLOAT32)((FLOAT64)in[2 * i + 4] * (*ptr_w1r) + (FLOAT64)in[2 * i + 5] * (*ptr_w1i)); in[2 * i + 5] = (FLOAT32)(-(FLOAT64)in[2 * i + 4] * (*ptr_w1i) + (FLOAT64)in[2 * i + 5] * (*ptr_w1r)); in[2 * i + 4] = tmp; ptr_w1r += 3 * (128 / mpass - 1) + 1; ptr_w1i += 3 * (128 / mpass - 1) + 1; i += 3; } } } for (i = 0; i < n_pass; i++) { ptr_x[2 * i + 0] = in[2 * i + 0]; ptr_x[2 * i + 1] = in[2 * i + 1]; } for (i = 0; i < mpass; i++) { ixheaace_hbe_apply_fft_3(ptr_x, ptr_y, i_sign); ptr_x = ptr_x + 6; ptr_y = ptr_y + 6; } for (i = 0; i < mpass; i++) { in[2 * i + 0] = y[6 * i + 0]; in[2 * i + 1] = y[6 * i + 1]; } for (i = 0; i < mpass; i++) { in[2 * mpass + 2 * i + 0] = y[6 * i + 2]; in[2 * mpass + 2 * i + 1] = y[6 * i + 3]; } for (i = 0; i < mpass; i++) { in[4 * mpass + 2 * i + 0] = y[6 * i + 4]; in[4 * mpass + 2 * i + 1] = y[6 * i + 5]; } } VOID ixheaace_hbe_apply_fft_288(FLOAT32 *ptr_inp, FLOAT32 *ptr_scratch, WORD32 len, WORD32 i_sign) { /* Dividing the 288-point FFT into 96x3 i.e nx3*/ FLOAT32 *ptr_op = ptr_scratch; WORD32 mpoints = len / 96; WORD32 fpoints = len / 3; WORD32 ii, jj; ptr_scratch += 2 * len; for (ii = 0; ii < mpoints; ii++) { for (jj = 0; jj < fpoints; jj++) { ptr_op[2 * jj + 0] = ptr_inp[2 * mpoints * jj + 2 * ii]; ptr_op[2 * jj + 1] = ptr_inp[2 * mpoints * jj + 2 * ii + 1]; } /* 96-point (32x3-point) of FFT */ if (fpoints & (fpoints - 1)) ixheaace_hbe_apply_cfftn_gen(ptr_op, ptr_scratch, fpoints, i_sign); else ixheaace_hbe_apply_cfftn(ptr_op, ptr_scratch, fpoints, i_sign); for (jj = 0; jj < fpoints; jj++) { ptr_inp[mpoints * 2 * jj + 2 * ii + 0] = ptr_op[2 * jj + 0]; ptr_inp[mpoints * 2 * jj + 2 * ii + 1] = ptr_op[2 * jj + 1]; } } /* Multiplication FFT with twiddle table */ ixheaace_hbe_apply_tw_mult_fft(ptr_inp, ptr_op, fpoints, mpoints, ixheaac_twid_tbl_fft_288); for (ii = 0; ii < fpoints; ii++) { /* 3-point of FFT */ ixheaace_hbe_apply_fft_3(ptr_op, ptr_scratch, i_sign); ptr_op = ptr_op + (mpoints * 2); ptr_scratch = ptr_scratch + (mpoints * 2); } ptr_scratch -= fpoints * mpoints * 2; for (jj = 0; jj < fpoints; jj++) { ptr_inp[2 * jj + 0] = ptr_scratch[6 * jj]; ptr_inp[2 * jj + 1] = ptr_scratch[6 * jj + 1]; } for (jj = 0; jj < fpoints; jj++) { ptr_inp[2 * fpoints + 2 * jj + 0] = ptr_scratch[6 * jj + 2]; ptr_inp[2 * fpoints + 2 * jj + 1] = ptr_scratch[6 * jj + 3]; } for (jj = 0; jj < fpoints; jj++) { ptr_inp[4 * fpoints + 2 * jj + 0] = ptr_scratch[6 * jj + 4]; ptr_inp[4 * fpoints + 2 * jj + 1] = ptr_scratch[6 * jj + 5]; } } VOID ixheaace_hbe_apply_ifft_224(FLOAT32 *ptr_inp, FLOAT32 *ptr_scratch, WORD32 len, WORD32 i_sign) { /* Dividing 224-point IFFT into 32x7 */ WORD32 mpoints = len / 32; WORD32 fpoints = len / 7; WORD32 ii, jj; FLOAT32 *ptr_op = ptr_scratch; ptr_scratch += 2 * len; for (ii = 0; ii < mpoints; ii++) { for (jj = 0; jj < fpoints; jj++) { ptr_op[2 * jj + 0] = ptr_inp[2 * mpoints * jj + 2 * ii]; ptr_op[2 * jj + 1] = ptr_inp[2 * mpoints * jj + 2 * ii + 1]; } /* 32-point of IFFT*/ if (fpoints & (fpoints - 1)) ixheaace_hbe_apply_cfftn_gen(ptr_op, ptr_scratch, fpoints, i_sign); else ixheaace_hbe_apply_cfftn(ptr_op, ptr_scratch, fpoints, i_sign); for (jj = 0; jj < fpoints; jj++) { ptr_inp[mpoints * 2 * jj + 2 * ii + 0] = ptr_op[2 * jj + 0]; ptr_inp[mpoints * 2 * jj + 2 * ii + 1] = ptr_op[2 * jj + 1]; } } /* Multiplication IFFT with twiddle table */ ixheaace_hbe_apply_tw_mult_ifft(ptr_inp, ptr_op, fpoints, mpoints, ixheaac_twid_tbl_fft_224); for (ii = 0; ii < fpoints; ii++) { /* 7-point of IFFT */ ixheaace_hbe_apply_ifft_7(ptr_op, ptr_scratch); ptr_scratch += (mpoints * 2); ptr_op += (mpoints * 2); } ptr_scratch -= fpoints * mpoints * 2; for (jj = 0; jj < fpoints; jj++) { for (ii = 0; ii < mpoints; ii++) { ptr_inp[fpoints * ii * 2 + 2 * jj + 0] = ptr_scratch[mpoints * jj * 2 + 2 * ii + 0]; ptr_inp[fpoints * ii * 2 + 2 * jj + 1] = ptr_scratch[mpoints * jj * 2 + 2 * ii + 1]; } } } VOID ixheaace_hbe_apply_ifft_336(FLOAT32 *ptr_inp, FLOAT32 *ptr_scratch, WORD32 len, WORD32 i_sign) { WORD32 i, j; WORD32 m_points = len / 7; WORD32 n_points = len / 48; FLOAT32 *ptr_real, *ptr_imag, *ptr_real_1, *ptr_scratch_local; ptr_real = ptr_scratch; ptr_scratch += 2 * len; ptr_imag = ptr_scratch; ptr_scratch += len; ptr_scratch_local = ptr_scratch; ptr_scratch += len; ptr_real_1 = ptr_scratch; ptr_scratch += len; for (i = 0; i < len; i++) { ptr_real[i] = ptr_inp[2 * i + 0]; ptr_imag[i] = ptr_inp[2 * i + 1]; } for (i = 0; i < m_points; i++) { for (j = 0; j < n_points; j++) { ptr_real_1[2 * j + 0] = ptr_inp[m_points * 2 * j + 2 * i + 0]; ptr_real_1[2 * j + 1] = ptr_inp[m_points * 2 * j + 2 * i + 1]; } ixheaace_hbe_apply_ifft_7(ptr_real_1, ptr_scratch); for (j = 0; j < n_points; j++) { ptr_inp[m_points * 2 * j + 2 * i + 0] = ptr_scratch[2 * j + 0]; ptr_inp[m_points * 2 * j + 2 * i + 1] = ptr_scratch[2 * j + 1]; } } switch (m_points) { case 48: ixheaace_hbe_apply_tw_mult_ifft(ptr_inp, ptr_scratch_local, n_points, m_points, ixheaac_twid_tbl_fft_336); break; default: ixheaace_hbe_apply_tw_mult_ifft(ptr_inp, ptr_scratch_local, n_points, m_points, ixheaac_twid_tbl_fft_168); break; } for (i = 0; i < len; i++) { ptr_real[2 * i + 0] = ptr_scratch_local[2 * i + 0]; ptr_real[2 * i + 1] = ptr_scratch_local[2 * i + 1]; } for (i = 0; i < n_points; i++) { ixheaace_hbe_apply_cfftn_gen(ptr_real, ptr_scratch, m_points, i_sign); ptr_real += (2 * m_points); } ptr_real -= n_points * 2 * m_points; for (j = 0; j < n_points; j++) { for (i = 0; i < m_points; i++) { ptr_inp[n_points * 2 * i + 2 * j + 0] = ptr_real[2 * m_points * j + 2 * i + 0]; ptr_inp[n_points * 2 * i + 2 * j + 1] = ptr_real[2 * m_points * j + 2 * i + 1]; } } }