claude-noether/libva-v4l2-request-fourier
1
0
Fork 0
forked from marfrit/libva-v4l2-request-fourier
Code Pull Requests Activity

h264: bit-parse slice_header to populate DECODE_PARAMS bit-size fields

Browse Source 
The load-bearing fix from diff_against_ffmpeg.md (campaign repo).

Adds src/h264_slice_header.{c,h} — a minimal H.264 slice_header()
bit-parser per ITU-T H.264 (08/2024) §7.3.3. Parses just enough of
the slice header to populate the V4L2 DECODE_PARAMS fields VAAPI
doesn't carry and that hantro G1 hardware reads directly out of
DECODE_PARAMS into MMIO registers:

  dec_param->dec_ref_pic_marking_bit_size  -> G1_REG_DEC_CTRL5_REFPIC_MK_LEN
  dec_param->idr_pic_id                    -> G1_REG_DEC_CTRL5_IDR_PIC_ID
  dec_param->pic_order_cnt_bit_size        -> G1_REG_DEC_CTRL6_POC_LENGTH
  dec_param->pic_order_cnt_lsb             -> hantro reflist builder (poc_type=0)
  dec_param->delta_pic_order_cnt_bottom    -> same
  dec_param->delta_pic_order_cnt0/1        -> hantro reflist builder (poc_type=1)

Without these set correctly, hantro's hardware bitstream parser
walks past zero bits in the slice header, lands on garbage, decodes
zero pixels — the all-zero CAPTURE output observed across both mpv
and Firefox during 2026-05-04 Phase 0 (see libva-multiplanar campaign
phase0_evidence/2026-05-04-kernel-trace/findings.md).

Implementation:
- Minimal RBSP bit reader (br_read_u/_ue/_se), MSB-first, fault-flag
  on overrun.
- Emulation-prevention unescape (strips 0x03 after 0x00 0x00) on
  the first 64 bytes of the slice — slice headers fit comfortably.
- Walks slice_header() up to and including dec_ref_pic_marking(),
  measuring bit positions for the *_bit_size fields.
- Skips ref_pic_list_modification() and pred_weight_table() —
  needed only to advance the bit position to dec_ref_pic_marking().
- Returns a struct with the V4L2 fields plus diagnostics
  (first_mb_in_slice, slice_type, pps_id, frame_num).

Wired into h264_va_picture_to_v4l2 (src/h264.c) right after the
nal_ref_idc/nal_unit_type extraction. SPS/PPS context is built from
VAPicture's seq_fields and pic_fields; num_ref_idx_l0/l1_active
defaults come from VASlice (best available substitute for the
parsed PPS values). On parse success, populates decode_params with
the recovered values + emits a request_log with the decoded fields
for cross-validation against VAAPI's pre-parsed values.

src/meson.build: adds h264_slice_header.{c,h} to sources.

Cross-references:
- FFmpeg libavcodec/h264_slice.c (Kwiboo v4l2-request-n8.1) — populates
  H264SliceContext::ref_pic_marking_bit_size / pic_order_cnt_bit_size
  by the same bit-precise parse, then v4l2_request_h264.c forwards
  to V4L2.
- Linux drivers/media/platform/verisilicon/hantro_g1_h264_dec.c
  set_params() — the register-write code that reads these fields.

MVC nal_unit_type 20/21 unhandled (this fork strips MVC alongside
HEVC). Multi-slice non-IDR streams parse the first slice's header
only; for FRAME_BASED mode that's fine — kernel sees the whole
bitstream and parses subsequent slices itself.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Markus Fritsche
2026-05-04 12:34:47 +00:00
parent d41a4b96b3
commit 9de1be34ef
4 changed files with 543 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/h264.c
+85
View File
@@ -39,6 +39,7 @@
#include "utils.h"
#include "surface.h"
#include "v4l2.h"
#include "h264_slice_header.h"
enum h264_slice_type {
H264_SLICE_P = 0,
@@ -321,6 +322,90 @@ static void h264_va_picture_to_v4l2(struct request_data *driver_data,
nal_ref_idc = (b[0] >> 5) & 0x3;
nal_unit_type = b[0] & 0x1f;
/*
* Bit-parse the slice_header() to recover fields VAAPI doesn't
* forward and that hantro G1 hardware reads out of DECODE_PARAMS:
*
* - dec_ref_pic_marking_bit_size -> G1_REG_DEC_CTRL5_REFPIC_MK_LEN
* - idr_pic_id -> G1_REG_DEC_CTRL5_IDR_PIC_ID
* - pic_order_cnt_bit_size -> G1_REG_DEC_CTRL6_POC_LENGTH
* - pic_order_cnt_lsb / delta_pic_order_cnt_* (used by hantro
* reference-list builder for poc_type=0/1 inter prediction)
*
* Without these set correctly, hantro's hardware bitstream parser
* walks past zero bits, lands on garbage, decodes zero pixels —
* the all-zero CAPTURE output observed during 2026-05-04 Phase 0.
*
* Spec: ITU-T H.264 §7.3.3 slice_header. Cross-reference (proven
* working): FFmpeg libavcodec/h264_slice.c populates
* H264SliceContext::ref_pic_marking_bit_size and
* pic_order_cnt_bit_size by the same bit-precise parse.
*/
{
const struct h264_slice_header_context sh_ctx = {
.separate_colour_plane_flag =
(VAPicture->seq_fields.bits.residual_colour_transform_flag != 0),
.log2_max_frame_num_minus4 =
VAPicture->seq_fields.bits.log2_max_frame_num_minus4,
.frame_mbs_only_flag =
(VAPicture->seq_fields.bits.frame_mbs_only_flag != 0),
.pic_order_cnt_type =
VAPicture->seq_fields.bits.pic_order_cnt_type,
.log2_max_pic_order_cnt_lsb_minus4 =
VAPicture->seq_fields.bits.log2_max_pic_order_cnt_lsb_minus4,
.delta_pic_order_always_zero_flag =
(VAPicture->seq_fields.bits.delta_pic_order_always_zero_flag != 0),
.bottom_field_pic_order_in_frame_present_flag =
(VAPicture->pic_fields.bits.pic_order_present_flag != 0),
.redundant_pic_cnt_present_flag =
(VAPicture->pic_fields.bits.redundant_pic_cnt_present_flag != 0),
.weighted_pred_flag =
(VAPicture->pic_fields.bits.weighted_pred_flag != 0),
.weighted_bipred_idc =
VAPicture->pic_fields.bits.weighted_bipred_idc,
.num_ref_idx_l0_default_active_minus1 =
surface->params.h264.slice.num_ref_idx_l0_active_minus1,
.num_ref_idx_l1_default_active_minus1 =
surface->params.h264.slice.num_ref_idx_l1_active_minus1,
.chroma_format_idc =
VAPicture->seq_fields.bits.chroma_format_idc,
.bit_depth_luma_minus8 =
VAPicture->bit_depth_luma_minus8,
.bit_depth_chroma_minus8 =
VAPicture->bit_depth_chroma_minus8,
.nal_unit_type = nal_unit_type,
.nal_ref_idc = nal_ref_idc,
};
struct h264_slice_header_info sh = { 0 };
unsigned char *nal_payload = b + 1; /* past NAL header byte */
size_t nal_payload_len = surface->slices_size -
(size_t)((nal_payload) - (unsigned char *)surface->source_data);
int sh_rc = h264_parse_slice_header(nal_payload, nal_payload_len,
&sh_ctx, &sh);
if (sh_rc == 0) {
decode->idr_pic_id = sh.idr_pic_id;
decode->pic_order_cnt_lsb = sh.pic_order_cnt_lsb;
decode->delta_pic_order_cnt_bottom = sh.delta_pic_order_cnt_bottom;
decode->delta_pic_order_cnt0 = sh.delta_pic_order_cnt0;
decode->delta_pic_order_cnt1 = sh.delta_pic_order_cnt1;
decode->pic_order_cnt_bit_size = sh.pic_order_cnt_bit_size;
decode->dec_ref_pic_marking_bit_size = sh.dec_ref_pic_marking_bit_size;
request_log("slice_header parse: idr_pic_id=%u "
"poc_lsb=%u poc_bits=%u refmark_bits=%u "
"frame_num=%u slice_type=%u pps_id=%u\n",
sh.idr_pic_id, sh.pic_order_cnt_lsb,
sh.pic_order_cnt_bit_size,
sh.dec_ref_pic_marking_bit_size,
sh.frame_num, sh.slice_type,
sh.pic_parameter_set_id);
} else {
request_log("slice_header parse FAILED rc=%d "
"(payload_len=%zu) — DECODE_PARAMS bit_size "
"fields left zero, hantro will likely produce zeros\n",
sh_rc, nal_payload_len);
}
}
h264_fill_dpb(driver_data, context, VAPicture, decode);
/*