Add SVE implementation of av1_compute_stats
Add SVE implementation of av1_compute_stats as well as the corresponding
tests. This gives between 30% and 45% uplift over the Neon
implementation.
Change-Id: I45ba993e85d9b43ba1beefcc3f75a6b1e44ba280
diff --git a/av1/av1.cmake b/av1/av1.cmake
index 32645f6..c1206e9 100644
--- a/av1/av1.cmake
+++ b/av1/av1.cmake
@@ -375,6 +375,7 @@
list(APPEND AOM_AV1_ENCODER_INTRIN_SVE
"${AOM_ROOT}/av1/encoder/arm/neon/av1_error_sve.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/pickrst_sve.c"
"${AOM_ROOT}/av1/encoder/arm/neon/wedge_utils_sve.c")
list(APPEND AOM_AV1_ENCODER_INTRIN_ARM_CRC32
diff --git a/av1/common/av1_rtcd_defs.pl b/av1/common/av1_rtcd_defs.pl
index c083133..8bf4b07 100644
--- a/av1/common/av1_rtcd_defs.pl
+++ b/av1/common/av1_rtcd_defs.pl
@@ -458,7 +458,7 @@
if (aom_config("CONFIG_REALTIME_ONLY") ne "yes") {
add_proto qw/void av1_compute_stats/, "int wiener_win, const uint8_t *dgd8, const uint8_t *src8, int16_t *dgd_avg, int16_t *src_avg, int h_start, int h_end, int v_start, int v_end, int dgd_stride, int src_stride, int64_t *M, int64_t *H, int use_downsampled_wiener_stats";
- specialize qw/av1_compute_stats sse4_1 avx2 neon/;
+ specialize qw/av1_compute_stats sse4_1 avx2 neon sve/;
add_proto qw/void av1_calc_proj_params/, "const uint8_t *src8, int width, int height, int src_stride, const uint8_t *dat8, int dat_stride, int32_t *flt0, int flt0_stride, int32_t *flt1, int flt1_stride, int64_t H[2][2], int64_t C[2], const sgr_params_type *params";
specialize qw/av1_calc_proj_params sse4_1 avx2 neon/;
add_proto qw/int64_t av1_lowbd_pixel_proj_error/, "const uint8_t *src8, int width, int height, int src_stride, const uint8_t *dat8, int dat_stride, int32_t *flt0, int flt0_stride, int32_t *flt1, int flt1_stride, int xq[2], const sgr_params_type *params";
diff --git a/av1/encoder/arm/neon/pickrst_sve.c b/av1/encoder/arm/neon/pickrst_sve.c
new file mode 100644
index 0000000..a519ecc
--- /dev/null
+++ b/av1/encoder/arm/neon/pickrst_sve.c
@@ -0,0 +1,590 @@
+/*
+ * Copyright (c) 2024, Alliance for Open Media. All rights reserved
+ *
+ * This source code is subject to the terms of the BSD 2 Clause License and
+ * the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
+ * was not distributed with this source code in the LICENSE file, you can
+ * obtain it at www.aomedia.org/license/software. If the Alliance for Open
+ * Media Patent License 1.0 was not distributed with this source code in the
+ * PATENTS file, you can obtain it at www.aomedia.org/license/patent.
+ */
+
+#include <arm_neon.h>
+#include <arm_sve.h>
+#include <string.h>
+
+#include "config/aom_config.h"
+#include "config/av1_rtcd.h"
+
+#include "aom_dsp/arm/aom_neon_sve_bridge.h"
+#include "aom_dsp/arm/mem_neon.h"
+#include "aom_dsp/arm/sum_neon.h"
+#include "aom_dsp/arm/transpose_neon.h"
+#include "av1/common/restoration.h"
+#include "av1/encoder/pickrst.h"
+
+static INLINE uint8_t find_average_sve(const uint8_t *src, int src_stride,
+ int width, int height) {
+ uint32x4_t avg_u32 = vdupq_n_u32(0);
+ uint8x16_t ones = vdupq_n_u8(1);
+
+ // Use a predicate to compute the last columns.
+ svbool_t pattern = svwhilelt_b8_u32(0, width % 16);
+
+ int h = height;
+ do {
+ int j = width;
+ const uint8_t *src_ptr = src;
+ while (j >= 16) {
+ uint8x16_t s = vld1q_u8(src_ptr);
+ avg_u32 = vdotq_u32(avg_u32, s, ones);
+
+ j -= 16;
+ src_ptr += 16;
+ }
+ uint8x16_t s_end = svget_neonq_u8(svld1_u8(pattern, src_ptr));
+ avg_u32 = vdotq_u32(avg_u32, s_end, ones);
+
+ src += src_stride;
+ } while (--h != 0);
+ return (uint8_t)(vaddlvq_u32(avg_u32) / (width * height));
+}
+
+static INLINE void compute_sub_avg(const uint8_t *buf, int buf_stride, int avg,
+ int16_t *buf_avg, int buf_avg_stride,
+ int width, int height,
+ int downsample_factor) {
+ uint8x8_t avg_u8 = vdup_n_u8(avg);
+
+ // Use a predicate to compute the last columns.
+ svbool_t pattern = svwhilelt_b8_u32(0, width % 8);
+
+ uint8x8_t avg_end = vget_low_u8(svget_neonq_u8(svdup_n_u8_z(pattern, avg)));
+
+ do {
+ int j = width;
+ const uint8_t *buf_ptr = buf;
+ int16_t *buf_avg_ptr = buf_avg;
+ while (j >= 8) {
+ uint8x8_t d = vld1_u8(buf_ptr);
+ vst1q_s16(buf_avg_ptr, vreinterpretq_s16_u16(vsubl_u8(d, avg_u8)));
+
+ j -= 8;
+ buf_ptr += 8;
+ buf_avg_ptr += 8;
+ }
+ uint8x8_t d_end = vget_low_u8(svget_neonq_u8(svld1_u8(pattern, buf_ptr)));
+ vst1q_s16(buf_avg_ptr, vreinterpretq_s16_u16(vsubl_u8(d_end, avg_end)));
+
+ buf += buf_stride;
+ buf_avg += buf_avg_stride;
+ height -= downsample_factor;
+ } while (height > 0);
+}
+
+static INLINE void copy_upper_triangle(int64_t *H, int64_t *H_tmp,
+ const int wiener_win2, const int scale) {
+ for (int i = 0; i < wiener_win2 - 2; i = i + 2) {
+ // Transpose the first 2x2 square. It needs a special case as the element
+ // of the bottom left is on the diagonal.
+ int64x2_t row0 = vld1q_s64(H_tmp + i * wiener_win2 + i + 1);
+ int64x2_t row1 = vld1q_s64(H_tmp + (i + 1) * wiener_win2 + i + 1);
+
+ int64x2_t tr_row = aom_vtrn2q_s64(row0, row1);
+
+ vst1_s64(H_tmp + (i + 1) * wiener_win2 + i, vget_low_s64(row0));
+ vst1q_s64(H_tmp + (i + 2) * wiener_win2 + i, tr_row);
+
+ // Transpose and store all the remaining 2x2 squares of the line.
+ for (int j = i + 3; j < wiener_win2; j = j + 2) {
+ row0 = vld1q_s64(H_tmp + i * wiener_win2 + j);
+ row1 = vld1q_s64(H_tmp + (i + 1) * wiener_win2 + j);
+
+ int64x2_t tr_row0 = aom_vtrn1q_s64(row0, row1);
+ int64x2_t tr_row1 = aom_vtrn2q_s64(row0, row1);
+
+ vst1q_s64(H_tmp + j * wiener_win2 + i, tr_row0);
+ vst1q_s64(H_tmp + (j + 1) * wiener_win2 + i, tr_row1);
+ }
+ }
+ for (int i = 0; i < wiener_win2 * wiener_win2; i++) {
+ H[i] += H_tmp[i] * scale;
+ }
+}
+
+// Transpose the matrix that has just been computed and accumulate it in M.
+static INLINE void acc_transpose_M(int64_t *M, const int64_t *M_trn,
+ const int wiener_win, int scale) {
+ for (int i = 0; i < wiener_win; ++i) {
+ for (int j = 0; j < wiener_win; ++j) {
+ int tr_idx = j * wiener_win + i;
+ *M++ += (int64_t)(M_trn[tr_idx] * scale);
+ }
+ }
+}
+
+// Swap each half of the dgd vectors so that we can accumulate the result of
+// the dot-products directly in the destination matrix.
+static INLINE int16x8x2_t transpose_dgd(int16x8_t dgd0, int16x8_t dgd1) {
+ int16x8_t dgd_trn0 = vreinterpretq_s16_s64(
+ vzip1q_s64(vreinterpretq_s64_s16(dgd0), vreinterpretq_s64_s16(dgd1)));
+ int16x8_t dgd_trn1 = vreinterpretq_s16_s64(
+ vzip2q_s64(vreinterpretq_s64_s16(dgd0), vreinterpretq_s64_s16(dgd1)));
+
+ return (struct int16x8x2_t){ dgd_trn0, dgd_trn1 };
+}
+
+static INLINE void compute_M_one_row_win5(int16x8_t src, int16x8_t dgd[5],
+ int64_t *M, int row) {
+ const int wiener_win = 5;
+
+ int64x2_t m01 = vld1q_s64(M + row * wiener_win + 0);
+ int16x8x2_t dgd01 = transpose_dgd(dgd[0], dgd[1]);
+
+ int64x2_t cross_corr01 = aom_svdot_lane_s16(m01, dgd01.val[0], src, 0);
+ cross_corr01 = aom_svdot_lane_s16(cross_corr01, dgd01.val[1], src, 1);
+ vst1q_s64(M + row * wiener_win + 0, cross_corr01);
+
+ int64x2_t m23 = vld1q_s64(M + row * wiener_win + 2);
+ int16x8x2_t dgd23 = transpose_dgd(dgd[2], dgd[3]);
+
+ int64x2_t cross_corr23 = aom_svdot_lane_s16(m23, dgd23.val[0], src, 0);
+ cross_corr23 = aom_svdot_lane_s16(cross_corr23, dgd23.val[1], src, 1);
+ vst1q_s64(M + row * wiener_win + 2, cross_corr23);
+
+ int64x2_t m4 = aom_sdotq_s16(vdupq_n_s64(0), src, dgd[4]);
+ M[row * wiener_win + 4] += vaddvq_s64(m4);
+}
+
+static INLINE void compute_M_one_row_win7(int16x8_t src, int16x8_t dgd[7],
+ int64_t *M, int row) {
+ const int wiener_win = 7;
+
+ int64x2_t m01 = vld1q_s64(M + row * wiener_win + 0);
+ int16x8x2_t dgd01 = transpose_dgd(dgd[0], dgd[1]);
+
+ int64x2_t cross_corr01 = aom_svdot_lane_s16(m01, dgd01.val[0], src, 0);
+ cross_corr01 = aom_svdot_lane_s16(cross_corr01, dgd01.val[1], src, 1);
+ vst1q_s64(M + row * wiener_win + 0, cross_corr01);
+
+ int64x2_t m23 = vld1q_s64(M + row * wiener_win + 2);
+ int16x8x2_t dgd23 = transpose_dgd(dgd[2], dgd[3]);
+
+ int64x2_t cross_corr23 = aom_svdot_lane_s16(m23, dgd23.val[0], src, 0);
+ cross_corr23 = aom_svdot_lane_s16(cross_corr23, dgd23.val[1], src, 1);
+ vst1q_s64(M + row * wiener_win + 2, cross_corr23);
+
+ int64x2_t m45 = vld1q_s64(M + row * wiener_win + 4);
+ int16x8x2_t dgd45 = transpose_dgd(dgd[4], dgd[5]);
+
+ int64x2_t cross_corr45 = aom_svdot_lane_s16(m45, dgd45.val[0], src, 0);
+ cross_corr45 = aom_svdot_lane_s16(cross_corr45, dgd45.val[1], src, 1);
+ vst1q_s64(M + row * wiener_win + 4, cross_corr45);
+
+ int64x2_t m6 = aom_sdotq_s16(vdupq_n_s64(0), src, dgd[6]);
+ M[row * wiener_win + 6] += vaddvq_s64(m6);
+}
+
+static INLINE void compute_H_one_col(int16x8_t *dgd, int col, int64_t *H,
+ const int wiener_win,
+ const int wiener_win2) {
+ for (int row0 = 0; row0 < wiener_win; row0++) {
+ for (int row1 = row0; row1 < wiener_win; row1++) {
+ int auto_cov_idx =
+ (col * wiener_win + row0) * wiener_win2 + (col * wiener_win) + row1;
+
+ int64x2_t auto_cov = aom_sdotq_s16(vdupq_n_s64(0), dgd[row0], dgd[row1]);
+ H[auto_cov_idx] += vaddvq_s64(auto_cov);
+ }
+ }
+}
+
+static INLINE void compute_H_two_rows_win5(int16x8_t *dgd0, int16x8_t *dgd1,
+ int row0, int row1, int64_t *H) {
+ for (int col0 = 0; col0 < 5; col0++) {
+ int auto_cov_idx = (row0 * 5 + col0) * 25 + (row1 * 5);
+
+ int64x2_t h01 = vld1q_s64(H + auto_cov_idx);
+ int16x8x2_t dgd01 = transpose_dgd(dgd1[0], dgd1[1]);
+
+ int64x2_t auto_cov01 = aom_svdot_lane_s16(h01, dgd01.val[0], dgd0[col0], 0);
+ auto_cov01 = aom_svdot_lane_s16(auto_cov01, dgd01.val[1], dgd0[col0], 1);
+ vst1q_s64(H + auto_cov_idx, auto_cov01);
+
+ int64x2_t h23 = vld1q_s64(H + auto_cov_idx + 2);
+ int16x8x2_t dgd23 = transpose_dgd(dgd1[2], dgd1[3]);
+
+ int64x2_t auto_cov23 = aom_svdot_lane_s16(h23, dgd23.val[0], dgd0[col0], 0);
+ auto_cov23 = aom_svdot_lane_s16(auto_cov23, dgd23.val[1], dgd0[col0], 1);
+ vst1q_s64(H + auto_cov_idx + 2, auto_cov23);
+
+ int64x2_t auto_cov4 = aom_sdotq_s16(vdupq_n_s64(0), dgd0[col0], dgd1[4]);
+ H[auto_cov_idx + 4] += vaddvq_s64(auto_cov4);
+ }
+}
+
+static INLINE void compute_H_two_rows_win7(int16x8_t *dgd0, int16x8_t *dgd1,
+ int row0, int row1, int64_t *H) {
+ for (int col0 = 0; col0 < 7; col0++) {
+ int auto_cov_idx = (row0 * 7 + col0) * 49 + (row1 * 7);
+
+ int64x2_t h01 = vld1q_s64(H + auto_cov_idx);
+ int16x8x2_t dgd01 = transpose_dgd(dgd1[0], dgd1[1]);
+
+ int64x2_t auto_cov01 = aom_svdot_lane_s16(h01, dgd01.val[0], dgd0[col0], 0);
+ auto_cov01 = aom_svdot_lane_s16(auto_cov01, dgd01.val[1], dgd0[col0], 1);
+ vst1q_s64(H + auto_cov_idx, auto_cov01);
+
+ int64x2_t h23 = vld1q_s64(H + auto_cov_idx + 2);
+ int16x8x2_t dgd23 = transpose_dgd(dgd1[2], dgd1[3]);
+
+ int64x2_t auto_cov23 = aom_svdot_lane_s16(h23, dgd23.val[0], dgd0[col0], 0);
+ auto_cov23 = aom_svdot_lane_s16(auto_cov23, dgd23.val[1], dgd0[col0], 1);
+ vst1q_s64(H + auto_cov_idx + 2, auto_cov23);
+
+ int64x2_t h45 = vld1q_s64(H + auto_cov_idx + 4);
+ int16x8x2_t dgd45 = transpose_dgd(dgd1[4], dgd1[5]);
+
+ int64x2_t auto_cov45 = aom_svdot_lane_s16(h45, dgd45.val[0], dgd0[col0], 0);
+ auto_cov45 = aom_svdot_lane_s16(auto_cov45, dgd45.val[1], dgd0[col0], 1);
+ vst1q_s64(H + auto_cov_idx + 4, auto_cov45);
+
+ int64x2_t auto_cov6 = aom_sdotq_s16(vdupq_n_s64(0), dgd0[col0], dgd1[6]);
+ H[auto_cov_idx + 6] += vaddvq_s64(auto_cov6);
+ }
+}
+
+// This function computes two matrices: the cross-correlation between the src
+// buffer and dgd buffer (M), and the auto-covariance of the dgd buffer (H).
+//
+// M is of size 7 * 7. It needs to be filled such that multiplying one element
+// from src with each element of a row of the wiener window will fill one
+// column of M. However this is not very convenient in terms of memory
+// accesses, as it means we do contiguous loads of dgd but strided stores to M.
+// As a result, we use an intermediate matrix M_trn which is instead filled
+// such that one row of the wiener window gives one row of M_trn. Once fully
+// computed, M_trn is then transposed to return M.
+//
+// H is of size 49 * 49. It is filled by multiplying every pair of elements of
+// the wiener window together. Since it is a symmetric matrix, we only compute
+// the upper triangle, and then copy it down to the lower one. Here we fill it
+// by taking each different pair of columns, and multiplying all the elements of
+// the first one with all the elements of the second one, with a special case
+// when multiplying a column by itself.
+static INLINE void compute_stats_win7_sve(int16_t *dgd_avg, int dgd_avg_stride,
+ int16_t *src_avg, int src_avg_stride,
+ int width, int height, int64_t *M,
+ int64_t *H, int downsample_factor) {
+ const int wiener_win = 7;
+ const int wiener_win2 = wiener_win * wiener_win;
+
+ // Use a predicate to compute the last columns of the block for H.
+ svbool_t pattern = svwhilelt_b16_u32(0, width % 8);
+
+ // Use intermediate matrices for H and M to perform the computation, they
+ // will be accumulated into the original H and M at the end.
+ int64_t M_trn[49];
+ memset(M_trn, 0, sizeof(M_trn));
+
+ int64_t H_tmp[49 * 49];
+ memset(H_tmp, 0, sizeof(H_tmp));
+
+ do {
+ // Cross-correlation (M).
+ for (int row = 0; row < wiener_win; row++) {
+ int j = 0;
+ while (j < width) {
+ int16x8_t dgd[7];
+ load_s16_8x7(dgd_avg + row * dgd_avg_stride + j, 1, &dgd[0], &dgd[1],
+ &dgd[2], &dgd[3], &dgd[4], &dgd[5], &dgd[6]);
+ int16x8_t s = vld1q_s16(src_avg + j);
+
+ // Compute all the elements of one row of M.
+ compute_M_one_row_win7(s, dgd, M_trn, row);
+
+ j += 8;
+ }
+ }
+
+ // Auto-covariance (H).
+ int j = 0;
+ while (j <= width - 8) {
+ for (int col0 = 0; col0 < wiener_win; col0++) {
+ int16x8_t dgd0[7];
+ load_s16_8x7(dgd_avg + j + col0, dgd_avg_stride, &dgd0[0], &dgd0[1],
+ &dgd0[2], &dgd0[3], &dgd0[4], &dgd0[5], &dgd0[6]);
+
+ // Perform computation of the first column with itself (28 elements).
+ // For the first column this will fill the upper triangle of the 7x7
+ // matrix at the top left of the H matrix. For the next columns this
+ // will fill the upper triangle of the other 7x7 matrices around H's
+ // diagonal.
+ compute_H_one_col(dgd0, col0, H_tmp, wiener_win, wiener_win2);
+
+ // All computation next to the matrix diagonal has already been done.
+ for (int col1 = col0 + 1; col1 < wiener_win; col1++) {
+ // Load second column and scale based on downsampling factor.
+ int16x8_t dgd1[7];
+ load_s16_8x7(dgd_avg + j + col1, dgd_avg_stride, &dgd1[0], &dgd1[1],
+ &dgd1[2], &dgd1[3], &dgd1[4], &dgd1[5], &dgd1[6]);
+
+ // Compute all elements from the combination of both columns (49
+ // elements).
+ compute_H_two_rows_win7(dgd0, dgd1, col0, col1, H_tmp);
+ }
+ }
+ j += 8;
+ }
+
+ if (j < width) {
+ // Process remaining columns using a predicate to discard excess elements.
+ for (int col0 = 0; col0 < wiener_win; col0++) {
+ // Load first column.
+ int16x8_t dgd0[7];
+ dgd0[0] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 0 * dgd_avg_stride + j + col0));
+ dgd0[1] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 1 * dgd_avg_stride + j + col0));
+ dgd0[2] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 2 * dgd_avg_stride + j + col0));
+ dgd0[3] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 3 * dgd_avg_stride + j + col0));
+ dgd0[4] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 4 * dgd_avg_stride + j + col0));
+ dgd0[5] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 5 * dgd_avg_stride + j + col0));
+ dgd0[6] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 6 * dgd_avg_stride + j + col0));
+
+ // Perform computation of the first column with itself (28 elements).
+ // For the first column this will fill the upper triangle of the 7x7
+ // matrix at the top left of the H matrix. For the next columns this
+ // will fill the upper triangle of the other 7x7 matrices around H's
+ // diagonal.
+ compute_H_one_col(dgd0, col0, H_tmp, wiener_win, wiener_win2);
+
+ // All computation next to the matrix diagonal has already been done.
+ for (int col1 = col0 + 1; col1 < wiener_win; col1++) {
+ // Load second column and scale based on downsampling factor.
+ int16x8_t dgd1[7];
+ load_s16_8x7(dgd_avg + j + col1, dgd_avg_stride, &dgd1[0], &dgd1[1],
+ &dgd1[2], &dgd1[3], &dgd1[4], &dgd1[5], &dgd1[6]);
+
+ // Compute all elements from the combination of both columns (49
+ // elements).
+ compute_H_two_rows_win7(dgd0, dgd1, col0, col1, H_tmp);
+ }
+ }
+ }
+ dgd_avg += downsample_factor * dgd_avg_stride;
+ src_avg += src_avg_stride;
+ } while (--height != 0);
+
+ // Transpose M_trn.
+ acc_transpose_M(M, M_trn, 7, downsample_factor);
+
+ // Copy upper triangle of H in the lower one.
+ copy_upper_triangle(H, H_tmp, wiener_win2, downsample_factor);
+}
+
+// This function computes two matrices: the cross-correlation between the src
+// buffer and dgd buffer (M), and the auto-covariance of the dgd buffer (H).
+//
+// M is of size 5 * 5. It needs to be filled such that multiplying one element
+// from src with each element of a row of the wiener window will fill one
+// column of M. However this is not very convenient in terms of memory
+// accesses, as it means we do contiguous loads of dgd but strided stores to M.
+// As a result, we use an intermediate matrix M_trn which is instead filled
+// such that one row of the wiener window gives one row of M_trn. Once fully
+// computed, M_trn is then transposed to return M.
+//
+// H is of size 25 * 25. It is filled by multiplying every pair of elements of
+// the wiener window together. Since it is a symmetric matrix, we only compute
+// the upper triangle, and then copy it down to the lower one. Here we fill it
+// by taking each different pair of columns, and multiplying all the elements of
+// the first one with all the elements of the second one, with a special case
+// when multiplying a column by itself.
+static INLINE void compute_stats_win5_sve(int16_t *dgd_avg, int dgd_avg_stride,
+ int16_t *src_avg, int src_avg_stride,
+ int width, int height, int64_t *M,
+ int64_t *H, int downsample_factor) {
+ const int wiener_win = 5;
+ const int wiener_win2 = wiener_win * wiener_win;
+
+ // Use a predicate to compute the last columns of the block for H.
+ svbool_t pattern = svwhilelt_b16_u32(0, width % 8);
+
+ // Use intermediate matrices for H and M to perform the computation, they
+ // will be accumulated into the original H and M at the end.
+ int64_t M_trn[25];
+ memset(M_trn, 0, sizeof(M_trn));
+
+ int64_t H_tmp[25 * 25];
+ memset(H_tmp, 0, sizeof(H_tmp));
+
+ do {
+ // Cross-correlation (M).
+ for (int row = 0; row < wiener_win; row++) {
+ int j = 0;
+ while (j < width) {
+ int16x8_t dgd[5];
+ load_s16_8x5(dgd_avg + row * dgd_avg_stride + j, 1, &dgd[0], &dgd[1],
+ &dgd[2], &dgd[3], &dgd[4]);
+ int16x8_t s = vld1q_s16(src_avg + j);
+
+ // Compute all the elements of one row of M.
+ compute_M_one_row_win5(s, dgd, M_trn, row);
+
+ j += 8;
+ }
+ }
+
+ // Auto-covariance (H).
+ int j = 0;
+ while (j <= width - 8) {
+ for (int col0 = 0; col0 < wiener_win; col0++) {
+ // Load first column.
+ int16x8_t dgd0[5];
+ load_s16_8x5(dgd_avg + j + col0, dgd_avg_stride, &dgd0[0], &dgd0[1],
+ &dgd0[2], &dgd0[3], &dgd0[4]);
+
+ // Perform computation of the first column with itself (15 elements).
+ // For the first column this will fill the upper triangle of the 5x5
+ // matrix at the top left of the H matrix. For the next columns this
+ // will fill the upper triangle of the other 5x5 matrices around H's
+ // diagonal.
+ compute_H_one_col(dgd0, col0, H_tmp, wiener_win, wiener_win2);
+
+ // All computation next to the matrix diagonal has already been done.
+ for (int col1 = col0 + 1; col1 < wiener_win; col1++) {
+ // Load second column and scale based on downsampling factor.
+ int16x8_t dgd1[5];
+ load_s16_8x5(dgd_avg + j + col1, dgd_avg_stride, &dgd1[0], &dgd1[1],
+ &dgd1[2], &dgd1[3], &dgd1[4]);
+
+ // Compute all elements from the combination of both columns (25
+ // elements).
+ compute_H_two_rows_win5(dgd0, dgd1, col0, col1, H_tmp);
+ }
+ }
+ j += 8;
+ }
+
+ // Process remaining columns using a predicate to discard excess elements.
+ if (j < width) {
+ for (int col0 = 0; col0 < wiener_win; col0++) {
+ int16x8_t dgd0[5];
+ dgd0[0] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 0 * dgd_avg_stride + j + col0));
+ dgd0[1] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 1 * dgd_avg_stride + j + col0));
+ dgd0[2] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 2 * dgd_avg_stride + j + col0));
+ dgd0[3] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 3 * dgd_avg_stride + j + col0));
+ dgd0[4] = svget_neonq_s16(
+ svld1_s16(pattern, dgd_avg + 4 * dgd_avg_stride + j + col0));
+
+ // Perform computation of the first column with itself (15 elements).
+ // For the first column this will fill the upper triangle of the 5x5
+ // matrix at the top left of the H matrix. For the next columns this
+ // will fill the upper triangle of the other 5x5 matrices around H's
+ // diagonal.
+ compute_H_one_col(dgd0, col0, H_tmp, wiener_win, wiener_win2);
+
+ // All computation next to the matrix diagonal has already been done.
+ for (int col1 = col0 + 1; col1 < wiener_win; col1++) {
+ // Load second column and scale based on downsampling factor.
+ int16x8_t dgd1[5];
+ load_s16_8x5(dgd_avg + j + col1, dgd_avg_stride, &dgd1[0], &dgd1[1],
+ &dgd1[2], &dgd1[3], &dgd1[4]);
+
+ // Compute all elements from the combination of both columns (25
+ // elements).
+ compute_H_two_rows_win5(dgd0, dgd1, col0, col1, H_tmp);
+ }
+ }
+ }
+ dgd_avg += downsample_factor * dgd_avg_stride;
+ src_avg += src_avg_stride;
+ } while (--height != 0);
+
+ // Transpose M_trn.
+ acc_transpose_M(M, M_trn, 5, downsample_factor);
+
+ // Copy upper triangle of H in the lower one.
+ copy_upper_triangle(H, H_tmp, wiener_win2, downsample_factor);
+}
+
+void av1_compute_stats_sve(int wiener_win, const uint8_t *dgd,
+ const uint8_t *src, int16_t *dgd_avg,
+ int16_t *src_avg, int h_start, int h_end,
+ int v_start, int v_end, int dgd_stride,
+ int src_stride, int64_t *M, int64_t *H,
+ int use_downsampled_wiener_stats) {
+ assert(wiener_win == WIENER_WIN || wiener_win == WIENER_WIN_CHROMA);
+
+ const int wiener_win2 = wiener_win * wiener_win;
+ const int wiener_halfwin = wiener_win >> 1;
+ const int32_t width = h_end - h_start;
+ const int32_t height = v_end - v_start;
+ const uint8_t *dgd_start = &dgd[v_start * dgd_stride + h_start];
+ memset(H, 0, sizeof(*H) * wiener_win2 * wiener_win2);
+ memset(M, 0, sizeof(*M) * wiener_win * wiener_win);
+
+ const uint8_t avg = find_average_sve(dgd_start, dgd_stride, width, height);
+ const int downsample_factor =
+ use_downsampled_wiener_stats ? WIENER_STATS_DOWNSAMPLE_FACTOR : 1;
+
+ // dgd_avg and src_avg have been memset to zero before calling this
+ // function, so round up the stride to the next multiple of 8 so that we
+ // don't have to worry about a tail loop when computing M.
+ const int dgd_avg_stride = ((width + 2 * wiener_halfwin) & ~7) + 8;
+ const int src_avg_stride = (width & ~7) + 8;
+
+ // Compute (dgd - avg) and store it in dgd_avg.
+ // The wiener window will slide along the dgd frame, centered on each pixel.
+ // For the top left pixel and all the pixels on the side of the frame this
+ // means half of the window will be outside of the frame. As such the actual
+ // buffer that we need to subtract the avg from will be 2 * wiener_halfwin
+ // wider and 2 * wiener_halfwin higher than the original dgd buffer.
+ const int vert_offset = v_start - wiener_halfwin;
+ const int horiz_offset = h_start - wiener_halfwin;
+ const uint8_t *dgd_win = dgd + horiz_offset + vert_offset * dgd_stride;
+ compute_sub_avg(dgd_win, dgd_stride, avg, dgd_avg, dgd_avg_stride,
+ width + 2 * wiener_halfwin, height + 2 * wiener_halfwin, 1);
+
+ // Compute (src - avg), downsample if necessary and store in src-avg.
+ const uint8_t *src_start = src + h_start + v_start * src_stride;
+ compute_sub_avg(src_start, src_stride * downsample_factor, avg, src_avg,
+ src_avg_stride, width, height, downsample_factor);
+
+ const int downsample_height = height / downsample_factor;
+
+ // Since the height is not necessarily a multiple of the downsample factor,
+ // the last line of src will be scaled according to how many rows remain.
+ const int downsample_remainder = height % downsample_factor;
+
+ if (wiener_win == WIENER_WIN) {
+ compute_stats_win7_sve(dgd_avg, dgd_avg_stride, src_avg, src_avg_stride,
+ width, downsample_height, M, H, downsample_factor);
+ } else {
+ compute_stats_win5_sve(dgd_avg, dgd_avg_stride, src_avg, src_avg_stride,
+ width, downsample_height, M, H, downsample_factor);
+ }
+
+ if (downsample_remainder > 0) {
+ const int remainder_offset = height - downsample_remainder;
+ if (wiener_win == WIENER_WIN) {
+ compute_stats_win7_sve(
+ dgd_avg + remainder_offset * dgd_avg_stride, dgd_avg_stride,
+ src_avg + downsample_height * src_avg_stride, src_avg_stride, width,
+ 1, M, H, downsample_remainder);
+ } else {
+ compute_stats_win5_sve(
+ dgd_avg + remainder_offset * dgd_avg_stride, dgd_avg_stride,
+ src_avg + downsample_height * src_avg_stride, src_avg_stride, width,
+ 1, M, H, downsample_remainder);
+ }
+ }
+}
diff --git a/test/wiener_test.cc b/test/wiener_test.cc
index b995c84..2886ed7 100644
--- a/test/wiener_test.cc
+++ b/test/wiener_test.cc
@@ -397,6 +397,12 @@
::testing::Values(av1_compute_stats_neon));
#endif // HAVE_NEON
+#if HAVE_SVE
+
+INSTANTIATE_TEST_SUITE_P(SVE, WienerTest,
+ ::testing::Values(av1_compute_stats_sve));
+#endif // HAVE_SVE
+
} // namespace wiener_lowbd
#if CONFIG_AV1_HIGHBITDEPTH
