Add Neon implementation for av1_calc_frame_error
Add Neon implementation for av1_calc_frame_error and the
corresponding tests as well.
Change-Id: I6ccbd0b3f40de32d8df6c0213b118389630c1f83
diff --git a/av1/av1.cmake b/av1/av1.cmake
index 4399942..4f7cdda 100644
--- a/av1/av1.cmake
+++ b/av1/av1.cmake
@@ -353,21 +353,21 @@
"${AOM_ROOT}/av1/encoder/x86/ml_avx2.c")
list(APPEND AOM_AV1_ENCODER_INTRIN_NEON
- "${AOM_ROOT}/av1/encoder/arm/neon/quantize_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/av1_error_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/av1_fwd_txfm2d_neon.c"
"${AOM_ROOT}/av1/encoder/arm/neon/av1_highbd_quantize_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/av1_k_means_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/encodetxb_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/frame_error_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/highbd_fwd_txfm_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/hybrid_fwd_txfm_neon.c"
"${AOM_ROOT}/av1/encoder/arm/neon/ml_neon.c"
"${AOM_ROOT}/av1/encoder/arm/neon/pickrst_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/pickrst_neon.h"
+ "${AOM_ROOT}/av1/encoder/arm/neon/quantize_neon.c"
"${AOM_ROOT}/av1/encoder/arm/neon/rdopt_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/av1_error_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/encodetxb_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/hybrid_fwd_txfm_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/av1_k_means_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/av1_fwd_txfm2d_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/highbd_fwd_txfm_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/wedge_utils_neon.c"
"${AOM_ROOT}/av1/encoder/arm/neon/reconinter_enc_neon.c"
- "${AOM_ROOT}/av1/encoder/arm/neon/temporal_filter_neon.c")
+ "${AOM_ROOT}/av1/encoder/arm/neon/temporal_filter_neon.c"
+ "${AOM_ROOT}/av1/encoder/arm/neon/wedge_utils_neon.c")
list(APPEND AOM_AV1_ENCODER_INTRIN_NEON_DOTPROD
"${AOM_ROOT}/av1/encoder/arm/neon/temporal_filter_neon_dotprod.c")
diff --git a/av1/common/av1_rtcd_defs.pl b/av1/common/av1_rtcd_defs.pl
index bf784f0..997777f 100644
--- a/av1/common/av1_rtcd_defs.pl
+++ b/av1/common/av1_rtcd_defs.pl
@@ -498,7 +498,7 @@
# Global motion
if (aom_config("CONFIG_REALTIME_ONLY") ne "yes") {
add_proto qw/int64_t av1_calc_frame_error/, "const uint8_t *const ref, int ref_stride, const uint8_t *const dst, int dst_stride, int p_width, int p_height";
- specialize qw/av1_calc_frame_error sse2 avx2/;
+ specialize qw/av1_calc_frame_error sse2 avx2 neon/;
if (aom_config("CONFIG_AV1_HIGHBITDEPTH") eq "yes") {
add_proto qw/int64_t av1_calc_highbd_frame_error/, "const uint16_t *const ref, int ref_stride, const uint16_t *const dst, int dst_stride, int p_width, int p_height, int bd";
diff --git a/av1/encoder/arm/neon/frame_error_neon.c b/av1/encoder/arm/neon/frame_error_neon.c
new file mode 100644
index 0000000..a00d821
--- /dev/null
+++ b/av1/encoder/arm/neon/frame_error_neon.c
@@ -0,0 +1,138 @@
+/*
+ * Copyright (c) 2023, 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 <stdlib.h>
+
+#include "config/aom_config.h"
+#include "config/aom_dsp_rtcd.h"
+
+/* clang-format off */
+// Error metric used for global motion evaluation.
+static const uint16_t error_measure_lut[256] = {
+ 0, 339, 550, 731, 894, 1045, 1187, 1323,
+ 1452, 1577, 1698, 1815, 1929, 2040, 2148, 2255,
+ 2359, 2461, 2562, 2661, 2758, 2854, 2948, 3041,
+ 3133, 3224, 3314, 3402, 3490, 3577, 3663, 3748,
+ 3832, 3916, 3998, 4080, 4162, 4242, 4322, 4401,
+ 4480, 4558, 4636, 4713, 4789, 4865, 4941, 5015,
+ 5090, 5164, 5237, 5311, 5383, 5456, 5527, 5599,
+ 5670, 5741, 5811, 5881, 5950, 6020, 6089, 6157,
+ 6225, 6293, 6361, 6428, 6495, 6562, 6628, 6695,
+ 6760, 6826, 6891, 6956, 7021, 7086, 7150, 7214,
+ 7278, 7341, 7405, 7468, 7531, 7593, 7656, 7718,
+ 7780, 7842, 7903, 7965, 8026, 8087, 8148, 8208,
+ 8269, 8329, 8389, 8449, 8508, 8568, 8627, 8686,
+ 8745, 8804, 8862, 8921, 8979, 9037, 9095, 9153,
+ 9211, 9268, 9326, 9383, 9440, 9497, 9553, 9610,
+ 9666, 9723, 9779, 9835, 9891, 9947, 10002, 10058,
+ 10113, 10168, 10224, 10279, 10333, 10388, 10443, 10497,
+ 10552, 10606, 10660, 10714, 10768, 10822, 10875, 10929,
+ 10982, 11036, 11089, 11142, 11195, 11248, 11301, 11353,
+ 11406, 11458, 11511, 11563, 11615, 11667, 11719, 11771,
+ 11823, 11875, 11926, 11978, 12029, 12080, 12132, 12183,
+ 12234, 12285, 12335, 12386, 12437, 12487, 12538, 12588,
+ 12639, 12689, 12739, 12789, 12839, 12889, 12939, 12988,
+ 13038, 13088, 13137, 13187, 13236, 13285, 13334, 13383,
+ 13432, 13481, 13530, 13579, 13628, 13676, 13725, 13773,
+ 13822, 13870, 13918, 13967, 14015, 14063, 14111, 14159,
+ 14206, 14254, 14302, 14350, 14397, 14445, 14492, 14539,
+ 14587, 14634, 14681, 14728, 14775, 14822, 14869, 14916,
+ 14963, 15010, 15056, 15103, 15149, 15196, 15242, 15289,
+ 15335, 15381, 15427, 15474, 15520, 15566, 15612, 15657,
+ 15703, 15749, 15795, 15840, 15886, 15932, 15977, 16022,
+ 16068, 16113, 16158, 16204, 16249, 16294, 16339, 16384,
+};
+/* clang-format on */
+
+int64_t av1_calc_frame_error_neon(const uint8_t *const ref, int ref_stride,
+ const uint8_t *const dst, int dst_stride,
+ int width, int height) {
+ int64_t sum_error[4] = { 0, 0, 0, 0 };
+ int r = 0;
+ int d = 0;
+
+ do {
+ int w = width;
+ int rr = r;
+ int dd = d;
+
+ do {
+ uint8x16_t dst_v = vld1q_u8(&dst[dd]);
+ uint8x16_t ref_v = vld1q_u8(&ref[rr]);
+
+#if AOM_ARCH_AARCH64
+ uint64x2_t abs_v = vreinterpretq_u64_u8(vabdq_u8(dst_v, ref_v));
+
+ uint64_t abs0 = vgetq_lane_u64(abs_v, 0);
+ uint64_t abs1 = vgetq_lane_u64(abs_v, 1);
+
+ sum_error[0] += error_measure_lut[(abs0 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs0 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs0 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs0 >> 24) & 0xFF];
+ sum_error[0] += error_measure_lut[(abs0 >> 32) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs0 >> 40) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs0 >> 48) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs0 >> 56) & 0xFF];
+
+ sum_error[0] += error_measure_lut[(abs1 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs1 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs1 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs1 >> 24) & 0xFF];
+ sum_error[0] += error_measure_lut[(abs1 >> 32) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs1 >> 40) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs1 >> 48) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs1 >> 56) & 0xFF];
+#else // !AOM_ARCH_AARCH64
+ uint32x4_t abs_v = vreinterpretq_u32_u8(vabdq_u8(dst_v, ref_v));
+
+ uint32_t abs0 = vgetq_lane_u32(abs_v, 0);
+ uint32_t abs1 = vgetq_lane_u32(abs_v, 1);
+ uint32_t abs2 = vgetq_lane_u32(abs_v, 2);
+ uint32_t abs3 = vgetq_lane_u32(abs_v, 3);
+
+ sum_error[0] += error_measure_lut[(abs0 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs0 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs0 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs0 >> 24) & 0xFF];
+ sum_error[0] += error_measure_lut[(abs1 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs1 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs1 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs1 >> 24) & 0xFF];
+
+ sum_error[0] += error_measure_lut[(abs2 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs2 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs2 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs2 >> 24) & 0xFF];
+ sum_error[0] += error_measure_lut[(abs3 >> 0) & 0xFF];
+ sum_error[1] += error_measure_lut[(abs3 >> 8) & 0xFF];
+ sum_error[2] += error_measure_lut[(abs3 >> 16) & 0xFF];
+ sum_error[3] += error_measure_lut[(abs3 >> 24) & 0xFF];
+#endif // AOM_ARCH_AARCH64
+
+ dd += 16;
+ rr += 16;
+ w -= 16;
+ } while (w >= 16);
+
+ while (w-- != 0) {
+ sum_error[0] += error_measure_lut[abs(dst[dd] - ref[rr])];
+ dd++;
+ rr++;
+ }
+
+ r += ref_stride;
+ d += dst_stride;
+ } while (--height != 0);
+
+ return sum_error[0] + sum_error[1] + sum_error[2] + sum_error[3];
+}
diff --git a/test/frame_error_test.cc b/test/frame_error_test.cc
index 8eceb4e..265fba6 100644
--- a/test/frame_error_test.cc
+++ b/test/frame_error_test.cc
@@ -27,7 +27,7 @@
typedef int64_t (*frame_error_func)(const uint8_t *const ref, int ref_stride,
const uint8_t *const dst, int dst_stride,
int p_width, int p_height);
-#if HAVE_AVX2 || HAVE_SSE2
+#if HAVE_AVX2 || HAVE_SSE2 || HAVE_NEON
const int kBlockWidth[] = {
832, 834, 640, 1280, 1920,
};
@@ -160,6 +160,14 @@
::testing::ValuesIn(kBlockWidth),
::testing::ValuesIn(kBlockHeight)));
#endif
+
+#if HAVE_NEON
+INSTANTIATE_TEST_SUITE_P(
+ NEON, AV1FrameErrorTest,
+ ::testing::Combine(::testing::Values(&av1_calc_frame_error_neon),
+ ::testing::ValuesIn(kBlockWidth),
+ ::testing::ValuesIn(kBlockHeight)));
+#endif
} // namespace frame_error_test
#if CONFIG_AV1_HIGHBITDEPTH
