x/dav1d
1
0
Fork 0
mirror of https://code.videolan.org/videolan/dav1d synced 2026-06-11 04:03:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
3a2a874994809a51bc4f9ec5bac51bd6af69767e
dav1d/tests/checkasm/loopfilter.c
T
Niklas Haas 3a2a874994 tests/checkasm: switch to external checkasm 
There are a number of benefits tied to the upstream / third-party checkasm
version, including:

- Improved long-term maintainability, code reuse with other projects, etc.

- Vastly improved overall performance / runtime for benchmarking, due
  primarily to the ability to scale the runtime of each test to that test's
  complexity.

- Much more robust statistical analysis of benchmarking results; including
  robust outlier rejection, an estimation of the histogram, and the ability
  to report the variance / stddev in addition to the (trimmed) mean.

- Interactive HTML and JSON output formats in addition to CSV/TSV.

- More readable and user-friendly output across the board, especially for
  failures and data dumps (e.g. also showing errors inside padding bytes).

- Better cross-platform support, including dynamic fallback of timer
  implementations on ARM platforms, a better RISC-V harness, and more.

There are multiple approaches to how we can solve the problem of integrating
this third party checkasm into dav1d, but I think the hybrid approach of
loading it as an external dependency, falling back to a meson wrap file,
provides the best overall compromise. This avoids the messiness of git e.g.
git submodules, while still allowing us to pin individual tags.
2026-01-01 17:33:55 +01:00

204 lines
7.8 KiB
C
Raw Blame History

/*
* Copyright © 2018, VideoLAN and dav1d authors
* Copyright © 2018, Two Orioles, LLC
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* 1. Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
* ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
* (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
* ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include "tests/checkasm/internal.h"
#include <string.h>
#include "src/levels.h"
#include "src/loopfilter.h"
static void init_lpf_border(pixel *const dst, const ptrdiff_t stride,
int E, int I, const int bitdepth_max)
{
const int bitdepth_min_8 = bitdepth_from_max(bitdepth_max) - 8;
const int F = 1 << bitdepth_min_8;
E <<= bitdepth_min_8;
I <<= bitdepth_min_8;
const int filter_type = rnd() % 4;
const int edge_diff = rnd() % ((E + 2) * 4) - 2 * (E + 2);
switch (filter_type) {
case 0: // random, unfiltered
for (int i = -8; i < 8; i++)
dst[i * stride] = rnd() & bitdepth_max;
break;
case 1: // long flat
dst[-8 * stride] = rnd() & bitdepth_max;
dst[+7 * stride] = rnd() & bitdepth_max;
dst[+0 * stride] = rnd() & bitdepth_max;
dst[-1 * stride] = iclip_pixel(dst[+0 * stride] + edge_diff);
for (int i = 1; i < 7; i++) {
dst[-(1 + i) * stride] = iclip_pixel(dst[-1 * stride] +
rnd() % (2 * (F + 1)) - (F + 1));
dst[+(0 + i) * stride] = iclip_pixel(dst[+0 * stride] +
rnd() % (2 * (F + 1)) - (F + 1));
}
break;
case 2: // short flat
for (int i = 4; i < 8; i++) {
dst[-(1 + i) * stride] = rnd() & bitdepth_max;
dst[+(0 + i) * stride] = rnd() & bitdepth_max;
}
dst[+0 * stride] = rnd() & bitdepth_max;
dst[-1 * stride] = iclip_pixel(dst[+0 * stride] + edge_diff);
for (int i = 1; i < 4; i++) {
dst[-(1 + i) * stride] = iclip_pixel(dst[-1 * stride] +
rnd() % (2 * (F + 1)) - (F + 1));
dst[+(0 + i) * stride] = iclip_pixel(dst[+0 * stride] +
rnd() % (2 * (F + 1)) - (F + 1));
}
break;
case 3: // normal or hev
for (int i = 4; i < 8; i++) {
dst[-(1 + i) * stride] = rnd() & bitdepth_max;
dst[+(0 + i) * stride] = rnd() & bitdepth_max;
}
dst[+0 * stride] = rnd() & bitdepth_max;
dst[-1 * stride] = iclip_pixel(dst[+0 * stride] + edge_diff);
for (int i = 1; i < 4; i++) {
dst[-(1 + i) * stride] = iclip_pixel(dst[-(0 + i) * stride] +
rnd() % (2 * (I + 1)) - (I + 1));
dst[+(0 + i) * stride] = iclip_pixel(dst[+(i - 1) * stride] +
rnd() % (2 * (I + 1)) - (I + 1));
}
break;
}
}
static void check_lpf_sb(loopfilter_sb_fn fn, const char *const name,
const int n_blks, const int lf_idx,
const int is_chroma, const int dir)
{
ALIGN_STK_64(pixel, c_dst_mem, 128 * 16,);
ALIGN_STK_64(pixel, a_dst_mem, 128 * 16,);
declare_func(void, pixel *dst, ptrdiff_t dst_stride, const uint32_t *mask,
const uint8_t (*l)[4], ptrdiff_t b4_stride,
const Av1FilterLUT *lut, int w HIGHBD_DECL_SUFFIX);
pixel *a_dst, *c_dst;
ptrdiff_t stride, b4_stride;
int w, h;
if (dir) {
a_dst = a_dst_mem + n_blks * 4 * 8;
c_dst = c_dst_mem + n_blks * 4 * 8;
w = n_blks * 4;
h = 16;
b4_stride = 32;
} else {
a_dst = a_dst_mem + 8;
c_dst = c_dst_mem + 8;
w = 16;
h = n_blks * 4;
b4_stride = 2;
}
stride = w * sizeof(pixel);
Av1FilterLUT lut;
const int sharp = rnd() & 7;
for (int level = 0; level < 64; level++) {
int limit = level;
if (sharp > 0) {
limit >>= (sharp + 3) >> 2;
limit = imin(limit, 9 - sharp);
}
limit = imax(limit, 1);
lut.i[level] = limit;
lut.e[level] = 2 * (level + 2) + limit;
}
lut.sharp[0] = (sharp + 3) >> 2;
lut.sharp[1] = sharp ? 9 - sharp : 0xff;
const int n_strengths = is_chroma ? 2 : 3;
for (int i = 0; i < n_strengths; i++) {
if (check_func(fn, "%s_w%d_%dbpc", name,
is_chroma ? 4 + 2 * i : 4 << i, BITDEPTH))
{
uint32_t vmask[4] = { 0 };
uint8_t l[32 * 2][4];
for (int j = 0; j < n_blks; j++) {
const int idx = rnd() % (i + 2);
if (idx) vmask[idx - 1] |= 1U << j;
if (dir) {
l[j][lf_idx] = rnd() & 63;
l[j + 32][lf_idx] = rnd() & 63;
} else {
l[j * 2][lf_idx] = rnd() & 63;
l[j * 2 + 1][lf_idx] = rnd() & 63;
}
}
#if BITDEPTH == 16
const int bitdepth_max = rnd() & 1 ? 0x3ff : 0xfff;
#else
const int bitdepth_max = 0xff;
#endif
for (int i = 0; i < 4 * n_blks; i++) {
const int x = i >> 2;
int L;
if (dir) {
L = l[32 + x][lf_idx] ? l[32 + x][lf_idx] : l[x][lf_idx];
} else {
L = l[2 * x + 1][lf_idx] ? l[2 * x + 1][lf_idx] : l[2 * x][lf_idx];
}
init_lpf_border(c_dst + i * (dir ? 1 : 16), dir ? n_blks * 4 : 1,
lut.e[L], lut.i[L], bitdepth_max);
}
memcpy(a_dst_mem, c_dst_mem, 128 * sizeof(pixel) * 16);
call_ref(c_dst, stride, vmask,
(const uint8_t(*)[4]) &l[dir ? 32 : 1][lf_idx],
b4_stride, &lut, n_blks HIGHBD_TAIL_SUFFIX);
call_new(a_dst, stride, vmask,
(const uint8_t(*)[4]) &l[dir ? 32 : 1][lf_idx],
b4_stride, &lut, n_blks HIGHBD_TAIL_SUFFIX);
checkasm_check_pixel(c_dst_mem, stride, a_dst_mem, stride,
w, h, "dst");
bench_new(alternate(c_dst, a_dst), stride, vmask,
(const uint8_t(*)[4]) &l[dir ? 32 : 1][lf_idx],
b4_stride, &lut, n_blks HIGHBD_TAIL_SUFFIX);
}
}
report("%s", name);
}
void bitfn(checkasm_check_loopfilter)(void) {
Dav1dLoopFilterDSPContext c;
bitfn(dav1d_loop_filter_dsp_init)(&c);
check_lpf_sb(c.loop_filter_sb[0][0], "lpf_h_sb_y", 32, 0, 0, 0);
check_lpf_sb(c.loop_filter_sb[0][1], "lpf_v_sb_y", 32, 1, 0, 1);
check_lpf_sb(c.loop_filter_sb[1][0], "lpf_h_sb_uv", 16, 2, 1, 0);
check_lpf_sb(c.loop_filter_sb[1][1], "lpf_v_sb_uv", 16, 2, 1, 1);
}
Reference in New Issue View Git Blame Copy Permalink