breeze/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2026-01-12 00:06:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

vulkan: merge ProRes and ProRes RAW iDCTs

Browse Source 
This cleans up the code a bit, and reduces binary size.
This commit is contained in:
Lynne
2025-12-18 13:30:30 +01:00
parent ea14f8a28f
commit 6eced88188
6 changed files with 130 additions and 168 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
libavcodec/vulkan/Makefile
View File

@@ -14,11 +14,11 @@ OBJS-$(CONFIG_FFV1_VULKAN_HWACCEL) += vulkan/common.o \
vulkan/ffv1_common.o vulkan/ffv1_reset.o \
vulkan/ffv1_dec_setup.o vulkan/ffv1_dec.o
OBJS-$(CONFIG_PRORES_RAW_VULKAN_HWACCEL) += vulkan/common.o \
OBJS-$(CONFIG_PRORES_RAW_VULKAN_HWACCEL) += vulkan/common.o vulkan/dct.o \
vulkan/prores_raw_decode.o \
vulkan/prores_raw_idct.o
OBJS-$(CONFIG_PRORES_VULKAN_HWACCEL) += vulkan/common.o \
OBJS-$(CONFIG_PRORES_VULKAN_HWACCEL) += vulkan/common.o vulkan/dct.o \
vulkan/prores_vld.o \
vulkan/prores_idct.o

119
libavcodec/vulkan/dct.comp Normal file
View File

@@ -0,0 +1,119 @@
/*
* Copyright (c) 2025 Lynne <dev@lynne.ee>
* Copyright (c) 2016 Nathan Egge <unlord@xiph.org>
*
* This file is part of FFmpeg.
*
* FFmpeg 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 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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 FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* Orthonormal inverse 8-point Type-II DCT based on the Chen factorization[1].
* 1D with scale factors moved up front.
* This computes an n-point Type-II DCT by first computing an n/2-point Type-II DCT
* of the even indexed inputs and an n/2-point Type-IV DST of the odd indexed inputs,
* and then combining them using a "butterfly" operation.
*
* [1] W.H. Chen, C. Smith, and S. Fralick,
* "A Fast Computational Algorithm for the Discrete Cosine Transform",
* IEEE Transactions on Communications, Vol. 25, No. 9, pp 1004-1009, Sept. 1977
*/
#ifndef NB_COMPONENTS
#define NB_COMPONENTS 1
#endif
/* Padded by 1 row to avoid bank conflicts */
shared float blocks[NB_BLOCKS][NB_COMPONENTS*8*(8 + 1)];
const float idct_scale[64] = {
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.1733799806652684, 0.2404849415639108, 0.2265318615882219, 0.2038732892122293,
0.1733799806652684, 0.1362237766939547, 0.0938325693794663, 0.0478354290456362,
0.1633203706095471, 0.2265318615882219, 0.2133883476483184, 0.1920444391778541,
0.1633203706095471, 0.1283199917898342, 0.0883883476483185, 0.0450599888754343,
0.1469844503024199, 0.2038732892122293, 0.1920444391778541, 0.1728354290456362,
0.1469844503024199, 0.1154849415639109, 0.0795474112858021, 0.0405529186026822,
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.0982118697983878, 0.1362237766939547, 0.1283199917898342, 0.1154849415639109,
0.0982118697983878, 0.0771645709543638, 0.0531518809229535, 0.0270965939155924,
0.0676495125182746, 0.0938325693794663, 0.0883883476483185, 0.0795474112858021,
0.0676495125182746, 0.0531518809229535, 0.0366116523516816, 0.0186644585125857,
0.0344874224103679, 0.0478354290456362, 0.0450599888754343, 0.0405529186026822,
0.0344874224103679, 0.0270965939155924, 0.0186644585125857, 0.0095150584360892,
};
void idct8(uint block, uint offset, uint stride)
{
float t0, t1, t2, t3, t4, t5, t6, t7, u8;
float u0, u1, u2, u3, u4, u5, u6, u7;
/* Input */
t0 = blocks[block][0*stride + offset];
u4 = blocks[block][1*stride + offset];
t2 = blocks[block][2*stride + offset];
u6 = blocks[block][3*stride + offset];
t1 = blocks[block][4*stride + offset];
u5 = blocks[block][5*stride + offset];
t3 = blocks[block][6*stride + offset];
u7 = blocks[block][7*stride + offset];
/* Embedded scaled inverse 4-point Type-II DCT */
u0 = t0 + t1;
u1 = t0 - t1;
u3 = t2 + t3;
u2 = (t2 - t3)*(1.4142135623730950488016887242097f) - u3;
t0 = u0 + u3;
t3 = u0 - u3;
t1 = u1 + u2;
t2 = u1 - u2;
/* Embedded scaled inverse 4-point Type-IV DST */
t5 = u5 + u6;
t6 = u5 - u6;
t7 = u4 + u7;
t4 = u4 - u7;
u7 = t7 + t5;
u5 = (t7 - t5)*(1.4142135623730950488016887242097f);
u8 = (t4 + t6)*(1.8477590650225735122563663787936f);
u4 = u8 - t4*(1.0823922002923939687994464107328f);
u6 = u8 - t6*(2.6131259297527530557132863468544f);
t7 = u7;
t6 = t7 - u6;
t5 = t6 + u5;
t4 = t5 - u4;
/* Butterflies */
u0 = t0 + t7;
u7 = t0 - t7;
u6 = t1 + t6;
u1 = t1 - t6;
u2 = t2 + t5;
u5 = t2 - t5;
u4 = t3 + t4;
u3 = t3 - t4;
/* Output */
blocks[block][0*stride + offset] = u0;
blocks[block][1*stride + offset] = u1;
blocks[block][2*stride + offset] = u2;
blocks[block][3*stride + offset] = u3;
blocks[block][4*stride + offset] = u4;
blocks[block][5*stride + offset] = u5;
blocks[block][6*stride + offset] = u6;
blocks[block][7*stride + offset] = u7;
}

84
libavcodec/vulkan/prores_idct.comp
View File

@@ -16,9 +16,6 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/* Two macroblocks, padded to avoid bank conflicts */
shared float blocks[4*2][8*(8+1)];
uint get_px(uint tex_idx, ivec2 pos)
{
#ifndef INTERLACED
@@ -37,87 +34,6 @@ void put_px(uint tex_idx, ivec2 pos, uint v)
#endif
}
const float idct_scale[64] = {
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.1733799806652684, 0.2404849415639108, 0.2265318615882219, 0.2038732892122293,
0.1733799806652684, 0.1362237766939547, 0.0938325693794663, 0.0478354290456362,
0.1633203706095471, 0.2265318615882219, 0.2133883476483184, 0.1920444391778541,
0.1633203706095471, 0.1283199917898342, 0.0883883476483185, 0.0450599888754343,
0.1469844503024199, 0.2038732892122293, 0.1920444391778541, 0.1728354290456362,
0.1469844503024199, 0.1154849415639109, 0.0795474112858021, 0.0405529186026822,
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.0982118697983878, 0.1362237766939547, 0.1283199917898342, 0.1154849415639109,
0.0982118697983878, 0.0771645709543638, 0.0531518809229535, 0.0270965939155924,
0.0676495125182746, 0.0938325693794663, 0.0883883476483185, 0.0795474112858021,
0.0676495125182746, 0.0531518809229535, 0.0366116523516816, 0.0186644585125857,
0.0344874224103679, 0.0478354290456362, 0.0450599888754343, 0.0405529186026822,
0.0344874224103679, 0.0270965939155924, 0.0186644585125857, 0.0095150584360892,
};
/* 7.4 Inverse Transform */
void idct8(uint block, uint offset, uint stride)
{
float t0, t1, t2, t3, t4, t5, t6, t7, u8;
float u0, u1, u2, u3, u4, u5, u6, u7;
/* Input */
t0 = blocks[block][0*stride + offset];
u4 = blocks[block][1*stride + offset];
t2 = blocks[block][2*stride + offset];
u6 = blocks[block][3*stride + offset];
t1 = blocks[block][4*stride + offset];
u5 = blocks[block][5*stride + offset];
t3 = blocks[block][6*stride + offset];
u7 = blocks[block][7*stride + offset];
/* Embedded scaled inverse 4-point Type-II DCT */
u0 = t0 + t1;
u1 = t0 - t1;
u3 = t2 + t3;
u2 = (t2 - t3)*(1.4142135623730950488016887242097f) - u3;
t0 = u0 + u3;
t3 = u0 - u3;
t1 = u1 + u2;
t2 = u1 - u2;
/* Embedded scaled inverse 4-point Type-IV DST */
t5 = u5 + u6;
t6 = u5 - u6;
t7 = u4 + u7;
t4 = u4 - u7;
u7 = t7 + t5;
u5 = (t7 - t5)*(1.4142135623730950488016887242097f);
u8 = (t4 + t6)*(1.8477590650225735122563663787936f);
u4 = u8 - t4*(1.0823922002923939687994464107328f);
u6 = u8 - t6*(2.6131259297527530557132863468544f);
t7 = u7;
t6 = t7 - u6;
t5 = t6 + u5;
t4 = t5 - u4;
/* Butterflies */
u0 = t0 + t7;
u7 = t0 - t7;
u6 = t1 + t6;
u1 = t1 - t6;
u2 = t2 + t5;
u5 = t2 - t5;
u4 = t3 + t4;
u3 = t3 - t4;
/* Output */
blocks[block][0*stride + offset] = u0;
blocks[block][1*stride + offset] = u1;
blocks[block][2*stride + offset] = u2;
blocks[block][3*stride + offset] = u3;
blocks[block][4*stride + offset] = u4;
blocks[block][5*stride + offset] = u5;
blocks[block][6*stride + offset] = u6;
blocks[block][7*stride + offset] = u7;
}
void main(void)
{
uvec3 gid = gl_GlobalInvocationID, lid = gl_LocalInvocationID;

82
libavcodec/vulkan/prores_raw_idct.comp
View File

@@ -24,8 +24,6 @@
#define BLOCK_ID (gl_LocalInvocationID.y)
#define ROW_ID (gl_LocalInvocationID.x)
shared float blocks[16][4*64];
const ivec2 scan[64] = {
ivec2( 0, 0), ivec2( 4, 0), ivec2( 0, 2), ivec2( 4, 2),
ivec2( 0, 8), ivec2( 4, 8), ivec2( 6, 8), ivec2( 2, 10),
@@ -45,86 +43,6 @@ const ivec2 scan[64] = {
ivec2(12, 12), ivec2( 8, 14), ivec2(10, 14), ivec2(14, 14),
};
const float idct_scale[64] = {
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.1733799806652684, 0.2404849415639108, 0.2265318615882219, 0.2038732892122293,
0.1733799806652684, 0.1362237766939547, 0.0938325693794663, 0.0478354290456362,
0.1633203706095471, 0.2265318615882219, 0.2133883476483184, 0.1920444391778541,
0.1633203706095471, 0.1283199917898342, 0.0883883476483185, 0.0450599888754343,
0.1469844503024199, 0.2038732892122293, 0.1920444391778541, 0.1728354290456362,
0.1469844503024199, 0.1154849415639109, 0.0795474112858021, 0.0405529186026822,
0.1250000000000000, 0.1733799806652684, 0.1633203706095471, 0.1469844503024199,
0.1250000000000000, 0.0982118697983878, 0.0676495125182746, 0.0344874224103679,
0.0982118697983878, 0.1362237766939547, 0.1283199917898342, 0.1154849415639109,
0.0982118697983878, 0.0771645709543638, 0.0531518809229535, 0.0270965939155924,
0.0676495125182746, 0.0938325693794663, 0.0883883476483185, 0.0795474112858021,
0.0676495125182746, 0.0531518809229535, 0.0366116523516816, 0.0186644585125857,
0.0344874224103679, 0.0478354290456362, 0.0450599888754343, 0.0405529186026822,
0.0344874224103679, 0.0270965939155924, 0.0186644585125857, 0.0095150584360892,
};
void idct8(uint block, uint offset, uint stride)
{
float t0, t1, t2, t3, t4, t5, t6, t7, u8;
float u0, u1, u2, u3, u4, u5, u6, u7;
/* Input */
t0 = blocks[block][0*stride + offset];
u4 = blocks[block][1*stride + offset];
t2 = blocks[block][2*stride + offset];
u6 = blocks[block][3*stride + offset];
t1 = blocks[block][4*stride + offset];
u5 = blocks[block][5*stride + offset];
t3 = blocks[block][6*stride + offset];
u7 = blocks[block][7*stride + offset];
/* Embedded scaled inverse 4-point Type-II DCT */
u0 = t0 + t1;
u1 = t0 - t1;
u3 = t2 + t3;
u2 = (t2 - t3)*(1.4142135623730950488016887242097f) - u3;
t0 = u0 + u3;
t3 = u0 - u3;
t1 = u1 + u2;
t2 = u1 - u2;
/* Embedded scaled inverse 4-point Type-IV DST */
t5 = u5 + u6;
t6 = u5 - u6;
t7 = u4 + u7;
t4 = u4 - u7;
u7 = t7 + t5;
u5 = (t7 - t5)*(1.4142135623730950488016887242097f);
u8 = (t4 + t6)*(1.8477590650225735122563663787936f);
u4 = u8 - t4*(1.0823922002923939687994464107328f);
u6 = u8 - t6*(2.6131259297527530557132863468544f);
t7 = u7;
t6 = t7 - u6;
t5 = t6 + u5;
t4 = t5 - u4;
/* Butterflies */
u0 = t0 + t7;
u7 = t0 - t7;
u6 = t1 + t6;
u1 = t1 - t6;
u2 = t2 + t5;
u5 = t2 - t5;
u4 = t3 + t4;
u3 = t3 - t4;
/* Output */
blocks[block][0*stride + offset] = u0;
blocks[block][1*stride + offset] = u1;
blocks[block][2*stride + offset] = u2;
blocks[block][3*stride + offset] = u3;
blocks[block][4*stride + offset] = u4;
blocks[block][5*stride + offset] = u5;
blocks[block][6*stride + offset] = u6;
blocks[block][7*stride + offset] = u7;
}
void main(void)
{
const uint tile_idx = gl_WorkGroupID.y*gl_NumWorkGroups.x + gl_WorkGroupID.x;

4
libavcodec/vulkan_prores.c
View File

@@ -24,6 +24,7 @@
#include "libavutil/vulkan_spirv.h"
extern const char *ff_source_common_comp;
extern const char *ff_source_dct_comp;
extern const char *ff_source_prores_vld_comp;
extern const char *ff_source_prores_idct_comp;
@@ -511,6 +512,9 @@ static int init_idct_shader(AVCodecContext *avctx, FFVulkanContext *s,
};
RET(ff_vk_shader_add_descriptor_set(s, shd, desc_set, 2, 0, 0));
GLSLC(0, #define NB_BLOCKS 4*2);
GLSLD(ff_source_dct_comp);
GLSLD(ff_source_prores_idct_comp);
RET(spv->compile_shader(s, spv, shd, &spv_data, &spv_len, "main",

5
libavcodec/vulkan_prores_raw.c
View File

@@ -26,6 +26,7 @@
#include "libavutil/mem.h"
extern const char *ff_source_common_comp;
extern const char *ff_source_dct_comp;
extern const char *ff_source_prores_raw_decode_comp;
extern const char *ff_source_prores_raw_idct_comp;
@@ -385,6 +386,10 @@ static int init_idct_shader(AVCodecContext *avctx, FFVulkanContext *s,
RET(add_common_data(avctx, s, shd, 0));
GLSLC(0, #define NB_BLOCKS 16);
GLSLC(0, #define NB_COMPONENTS 4);
GLSLD(ff_source_dct_comp);
GLSLD(ff_source_prores_raw_idct_comp);
RET(spv->compile_shader(s, spv, shd, &spv_data, &spv_len, "main",