marfrit/daedalus-fourier
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

h264: V3D shader for qpel mc02 (vertical half-pel)

Browse Source 
Sibling of cycle 9's v3d_h264_qpel_mc20.comp.  Same 6-tap H.264 luma
half-pel filter, transposed to vertical orientation: filter reads
rows [-2..+3] of source per output pixel instead of cols.

Shader is ~58 lines (vs mc20's 86) — same WG geometry (64 lanes /
1 block per WG / 1 lane per output pixel).  The address arithmetic
flips: row_base = src_off + r*stride + c (mc20) → col_base =
src_off + c, then col_base + (r±N)*stride (mc02).

dispatch_h264_qpel_mc02_qpu mirrors the mc20 QPU dispatch; src_max
calculation differs since the V kernel reads rows -2..+10 of source
(13 rows × stride wide) vs mc20's cols -2..+10 (8 rows × stride+11).
For 8x8 blocks: src_max = src_off + 10*stride + 8.

Recipe table flips DAEDALUS_KERNEL_H264_QPEL_MC02 from CPU to QPU.

Verified on hertz:

  $ ./build/test_api_h264 | grep qpel
    H.264 qpel mc20: 1024/1024 bytes bit-exact (100.0000%)
    H.264 qpel mc02: 2048/2048 bytes bit-exact (100.0000%)

QPU coverage for the 30 qpel positions:
  put_  mc20 ✓ (cycle 9)   mc02 ✓ (this PR)
        all 13 other put_  CPU NEON
  avg_  all 15 positions   CPU NEON

Next-priority candidates by per-frame impact (per PR #24 bench):
  mc22 (2D half-pel)  — 71.5 ns/block NEON × 32 640 blocks worst
                        case = 2.33 ms/frame at 1080p.  Most-used
                        qpel position in real H.264 streams.
  mc11/mc13/mc31/mc33 — corner ¼-pel positions, structurally similar
                        to mc20 + mc02 with L2 averaging.

The cascaded H+V structure of mc22 means it can either share the
existing mc20 + mc02 shaders' L2 (compute mc20 into tmp, then mc02
on tmp) or get a dedicated 2-stage pipeline.  Follow-up.
This commit is contained in:
Claude (noether)
2026-05-25 18:38:38 +02:00
parent 37b75b5813
commit bc5edf656d
3 changed files with 172 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CMakeLists.txt
+13 -1
View File
@@ -350,7 +350,18 @@ if (DAEDALUS_BUILD_VULKAN)
VERBATIM
)
add_custom_target(daedalus_shaders ALL DEPENDS ${NOOP_SPV} ${IDCT8_SPV} ${LPF_SPV} ${MC_SPV} ${LPF8_SPV} ${CDEF_SPV} ${H264DEBLOCK_SPV} ${H264DEBLOCK_H_SPV} ${H264DEBLOCK_CHROMA_V_SPV} ${H264DEBLOCK_CHROMA_H_SPV} ${H264_IDCT4_SPV} ${H264_IDCT8_SPV} ${H264_QPEL_MC20_SPV})
set(H264_QPEL_MC02_SPV ${CMAKE_BINARY_DIR}/v3d_h264_qpel_mc02.spv)
add_custom_command(
OUTPUT ${H264_QPEL_MC02_SPV}
COMMAND ${GLSLANG_VALIDATOR} -V --target-env vulkan1.3
-o ${H264_QPEL_MC02_SPV}
${CMAKE_SOURCE_DIR}/src/v3d_h264_qpel_mc02.comp
DEPENDS ${CMAKE_SOURCE_DIR}/src/v3d_h264_qpel_mc02.comp
COMMENT "glslang: v3d_h264_qpel_mc02.comp -> v3d_h264_qpel_mc02.spv"
VERBATIM
)
add_custom_target(daedalus_shaders ALL DEPENDS ${NOOP_SPV} ${IDCT8_SPV} ${LPF_SPV} ${MC_SPV} ${LPF8_SPV} ${CDEF_SPV} ${H264DEBLOCK_SPV} ${H264DEBLOCK_H_SPV} ${H264DEBLOCK_CHROMA_V_SPV} ${H264DEBLOCK_CHROMA_H_SPV} ${H264_IDCT4_SPV} ${H264_IDCT8_SPV} ${H264_QPEL_MC20_SPV} ${H264_QPEL_MC02_SPV})
# v3d_runner — reusable Vulkan plumbing.
add_library(v3d_runner STATIC src/v3d_runner.c)
@@ -489,6 +500,7 @@ if (DAEDALUS_BUILD_VULKAN)
${H264_IDCT4_SPV}
${H264_IDCT8_SPV}
${H264_QPEL_MC20_SPV}
${H264_QPEL_MC02_SPV}
DESTINATION ${CMAKE_INSTALL_DATADIR}/daedalus-fourier/shaders
)
endif()
src/daedalus_core.c
+89 -3
View File
@@ -52,6 +52,8 @@ struct daedalus_ctx {
v3d_pipeline h264_idct8_pipe;
int h264_qpel_mc20_pipe_ready;
v3d_pipeline h264_qpel_mc20_pipe;
int h264_qpel_mc02_pipe_ready;
v3d_pipeline h264_qpel_mc02_pipe;
};
daedalus_ctx *daedalus_ctx_create(void)
@@ -112,6 +114,7 @@ void daedalus_ctx_destroy(daedalus_ctx *ctx)
if (ctx->h264_idct4_pipe_ready) v3d_runner_destroy_pipeline(ctx->runner, &ctx->h264_idct4_pipe);
if (ctx->h264_idct8_pipe_ready) v3d_runner_destroy_pipeline(ctx->runner, &ctx->h264_idct8_pipe);
if (ctx->h264_qpel_mc20_pipe_ready) v3d_runner_destroy_pipeline(ctx->runner, &ctx->h264_qpel_mc20_pipe);
if (ctx->h264_qpel_mc02_pipe_ready) v3d_runner_destroy_pipeline(ctx->runner, &ctx->h264_qpel_mc02_pipe);
v3d_runner_destroy(ctx->runner);
}
free(ctx);
@@ -147,7 +150,7 @@ daedalus_substrate daedalus_recipe_substrate_for(daedalus_kernel k)
case DAEDALUS_KERNEL_H264_DEBLOCK_CV_INTRA: return DAEDALUS_SUBSTRATE_CPU; /* bS=4 chroma QPU pending */
case DAEDALUS_KERNEL_H264_DEBLOCK_CH_INTRA: return DAEDALUS_SUBSTRATE_CPU;
case DAEDALUS_KERNEL_H264_QPEL_MC20: return DAEDALUS_SUBSTRATE_QPU; /* v3d_h264_qpel_mc20.spv */
case DAEDALUS_KERNEL_H264_QPEL_MC02: return DAEDALUS_SUBSTRATE_CPU; /* QPU mc02 shader pending */
case DAEDALUS_KERNEL_H264_QPEL_MC02: return DAEDALUS_SUBSTRATE_QPU; /* v3d_h264_qpel_mc02.spv */
case DAEDALUS_KERNEL_H264_QPEL_MC22: return DAEDALUS_SUBSTRATE_CPU; /* QPU mc22 shader pending (hv lowpass) */
case DAEDALUS_KERNEL_H264_QPEL_MC10: return DAEDALUS_SUBSTRATE_CPU; /* ¼-H L2 */
case DAEDALUS_KERNEL_H264_QPEL_MC30: return DAEDALUS_SUBSTRATE_CPU; /* ¾-H L2 */
@@ -1457,6 +1460,89 @@ fail:
return -1;
}
static int dispatch_h264_qpel_mc02_qpu(daedalus_ctx *ctx,
uint8_t *dst, const uint8_t *src, size_t stride,
size_t n_blocks, const daedalus_h264_qpel_meta *meta)
{
/* Same shape as mc20 but with vertical access pattern. src_max
* reflects the row-wise filter window: bottom output row (r=7)
* reads up to row r+3 = 10 of the src; so src_max = src_off +
* 10*stride + 8 (last col + 1 for memcpy semantics). */
if (!ctx->h264_qpel_mc02_pipe_ready) {
if (v3d_runner_create_pipeline(ctx->runner, "v3d_h264_qpel_mc02.spv",
3, sizeof(h264_qpel_mc20_pc),
&ctx->h264_qpel_mc02_pipe) != 0)
return -1;
ctx->h264_qpel_mc02_pipe_ready = 1;
}
size_t meta_bytes = n_blocks * 4 * sizeof(uint32_t);
size_t src_max = 0, dst_max = 0;
for (size_t i = 0; i < n_blocks; i++) {
size_t s_end = meta[i].src_off + (size_t) 10 * stride + 8;
size_t d_end = meta[i].dst_off + (size_t) 7 * stride + 8;
if (s_end > src_max) src_max = s_end;
if (d_end > dst_max) dst_max = d_end;
}
v3d_buffer bs = {0}, bd = {0}, bm = {0};
if (v3d_runner_create_buffer(ctx->runner, src_max, &bs)) return -1;
if (v3d_runner_create_buffer(ctx->runner, dst_max, &bd)) {
v3d_runner_destroy_buffer(ctx->runner, &bs); return -1;
}
if (v3d_runner_create_buffer(ctx->runner, meta_bytes, &bm)) {
v3d_runner_destroy_buffer(ctx->runner, &bd);
v3d_runner_destroy_buffer(ctx->runner, &bs); return -1;
}
memcpy(bs.mapped, src, src_max);
memcpy(bd.mapped, dst, dst_max);
uint32_t *m = bm.mapped;
for (size_t i = 0; i < n_blocks; i++) {
m[4*i+0] = meta[i].dst_off;
m[4*i+1] = meta[i].src_off;
m[4*i+2] = 0;
m[4*i+3] = 0;
}
v3d_buffer binds[3] = { bs, bd, bm };
if (v3d_runner_bind_buffers(ctx->runner, &ctx->h264_qpel_mc02_pipe, binds, 3))
goto fail;
uint32_t wg_count = (uint32_t) n_blocks;
h264_qpel_mc20_pc pc = {
.n_blocks = (uint32_t) n_blocks,
.stride_u8 = (uint32_t) stride,
};
VkCommandBuffer cb = v3d_runner_alloc_cmdbuf(ctx->runner);
if (cb == VK_NULL_HANDLE) goto fail;
VkCommandBufferBeginInfo cbbi = { .sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO };
vkBeginCommandBuffer(cb, &cbbi);
vkCmdBindPipeline(cb, VK_PIPELINE_BIND_POINT_COMPUTE,
ctx->h264_qpel_mc02_pipe.pipeline);
vkCmdBindDescriptorSets(cb, VK_PIPELINE_BIND_POINT_COMPUTE,
ctx->h264_qpel_mc02_pipe.layout, 0, 1,
&ctx->h264_qpel_mc02_pipe.desc_set, 0, NULL);
vkCmdPushConstants(cb, ctx->h264_qpel_mc02_pipe.layout,
VK_SHADER_STAGE_COMPUTE_BIT, 0, sizeof(pc), &pc);
vkCmdDispatch(cb, wg_count, 1, 1);
vkEndCommandBuffer(cb);
if (v3d_runner_submit_wait(ctx->runner, cb)) goto fail;
memcpy(dst, bd.mapped, dst_max);
v3d_runner_destroy_buffer(ctx->runner, &bm);
v3d_runner_destroy_buffer(ctx->runner, &bd);
v3d_runner_destroy_buffer(ctx->runner, &bs);
return 0;
fail:
v3d_runner_destroy_buffer(ctx->runner, &bm);
v3d_runner_destroy_buffer(ctx->runner, &bd);
v3d_runner_destroy_buffer(ctx->runner, &bs);
return -1;
}
/* -------------------- Public dispatch entry points -------------- */
#define ROUTE_CPU_ONLY(_kernel, _cpu_fn, ...) \
@@ -1676,9 +1762,9 @@ int daedalus_dispatch_h264_qpel_mc02(daedalus_ctx *ctx, daedalus_substrate sub,
eff = daedalus_recipe_substrate_for(DAEDALUS_KERNEL_H264_QPEL_MC02);
if (eff == DAEDALUS_SUBSTRATE_QPU && !daedalus_ctx_has_qpu(ctx))
eff = DAEDALUS_SUBSTRATE_CPU;
if (eff == DAEDALUS_SUBSTRATE_QPU)
return -1; /* No mc02 QPU shader yet — explicit QPU fast-fails. */
if (eff == DAEDALUS_SUBSTRATE_CPU)
return dispatch_h264_qpel_mc02_cpu(ctx, dst, src, stride, n_blocks, meta);
return dispatch_h264_qpel_mc02_qpu(ctx, dst, src, stride, n_blocks, meta);
}
int daedalus_dispatch_h264_qpel_mc22(daedalus_ctx *ctx, daedalus_substrate sub,
src/v3d_h264_qpel_mc02.comp
+69
View File
@@ -0,0 +1,69 @@
// daedalus-fourier — H.264 luma qpel mc02 (8x8, vertical half-pel), V3D 7.1.
//
// Sibling of cycle 9's v3d_h264_qpel_mc20.comp. Same 6-tap filter,
// transposed to vertical direction:
//
// dst[r,c] = clip255(
// ( s[r-2,c]
// - 5 * s[r-1,c]
// + 20 * s[r, c]
// + 20 * s[r+1,c]
// - 5 * s[r+2,c]
// + s[r+3,c]
// + 16
// ) >> 5)
//
// src+src_off points at row 0 col 0 of the OUTPUT block; the filter
// reads rows -2..+3 (2 rows of top context, 3 rows of bottom).
//
// Same WG layout as mc20: 64 lanes / 1 block-per-WG / 1 lane-per-pixel.
//
// License: BSD-2-Clause.
#version 450
#extension GL_EXT_shader_8bit_storage : require
#extension GL_EXT_shader_explicit_arithmetic_types : require
layout(local_size_x = 64, local_size_y = 1, local_size_z = 1) in;
layout(binding = 0) readonly buffer Src { uint8_t src[]; } u_src;
layout(binding = 1) buffer Dst { uint8_t dst[]; } u_dst;
layout(binding = 2) readonly buffer Meta { uvec4 meta[]; } u_meta;
layout(push_constant) uniform PC {
uint n_blocks;
uint stride_u8;
uint _pad0, _pad1;
} pc;
void main()
{
uint block_idx = gl_WorkGroupID.x;
if (block_idx >= pc.n_blocks) return;
uint lane = gl_LocalInvocationID.x;
uint r = lane >> 3;
uint c = lane & 7u;
uint dst_off = u_meta.meta[block_idx].x;
uint src_off = u_meta.meta[block_idx].y;
uint stride = pc.stride_u8;
// Read the 6 rows of vertical context at col (c) of THIS output row.
// src_off+r*stride+c is at the OUTPUT pixel position; the kernel
// samples r-2..r+3 along the column. Unsigned-safe because the
// public API contract guarantees src_off >= 2*stride.
uint col_base = src_off + c;
int s_m2 = int(u_src.src[col_base + (r - 2u) * stride]);
int s_m1 = int(u_src.src[col_base + (r - 1u) * stride]);
int s_0 = int(u_src.src[col_base + r * stride]);
int s_p1 = int(u_src.src[col_base + (r + 1u) * stride]);
int s_p2 = int(u_src.src[col_base + (r + 2u) * stride]);
int s_p3 = int(u_src.src[col_base + (r + 3u) * stride]);
int v = s_m2 - 5 * s_m1 + 20 * s_0 + 20 * s_p1 - 5 * s_p2 + s_p3 + 16;
int p = clamp(v >> 5, 0, 255);
u_dst.dst[dst_off + r * stride + c] = uint8_t(p);
}