STRUCT VDEC_AV1_SLICE_MFMV(9)
π¦ struct vdec_av1_slice_mfmv(9) β the AV1 mfmv parameters glow-up π
okay so bestie, we're in the Kernel API Manual, section 9, and today's main character is struct vdec_av1_slice_mfmv. It's giving "small struct, big job" energy. Let's get into it π
β¨ NAME β¨
struct vdec_av1_slice_mfmv β literally the AV1 mfmv parameters. That's the whole name, that's the whole vibe. No cap.
πΈ SYNOPSIS πΈ
here she is in all her glory, copy-pasted with love and zero typos (do not touch, she's fragile):
struct vdec_av1_slice_mfmv {
u32 mfmv_valid_ref[3];
u32 mfmv_dir[3];
int mfmv_ref_to_cur[3];
int mfmv_ref_frame_idx[3];
int mfmv_count;
};
like. look at her. five fields. three-element arrays. u32 and int living together in harmony. we stan a compact struct π
π«Ά Members π«Ά
okay so here's the thing β and I need you to sit down for this β the documentation for every single one of these members is just the member name again. That's it. That's the whole description. The kernel-doc said "mfmv_valid_ref" and then, for the description, said "mfmv_valid_ref". Iconic honestly. Zero elaboration. Gatekeeping at its finest π
So without further ado, the full roster, exactly as the page gives it:
mfmv_valid_refβ the docs say:mfmv_valid_ref. yes queen, we know β¨mfmv_dirβ the docs say:mfmv_dir. short. mysterious. slay πmfmv_ref_to_curβ the docs say:mfmv_ref_to_cur. no notes πmfmv_ref_frame_idxβ the docs say:mfmv_ref_frame_idx. truly a legend πΈmfmv_countβ the docs say:mfmv_count. the finale. the moment. the count π
lowkey ngl, I can't tell you more than the manpage does because the manpage doesn't have more. bestie, I am not about to invent kernel behaviour β that's how you get cursed bugs. so we keep it honest: the descriptions are the names, verbatim, and that's canon π¦
π₯ SEE ALSO π₯
okay this is where it gets good because our girl does not exist in a vacuum β she's part of a whole friend group, a whole squad of vdec_av1_slice_* structs, and the manpage shouts out every single one. no member left behind fr fr π«Ά
Kernel file: ./drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_av1_req_lat_if.c
and the whole group chat:
struct vdec_av1_slice_init_vsi(9)struct vdec_av1_slice_mem(9)struct vdec_av1_slice_state(9)struct vdec_av1_slice_tile_group(9)struct vdec_av1_slice_scale_factors(9)struct vdec_av1_slice_frame_refs(9)struct vdec_av1_slice_gm(9)struct vdec_av1_slice_sm(9)struct vdec_av1_slice_seg(9)struct vdec_av1_slice_delta_q_lf(9)struct vdec_av1_slice_quantization(9)struct vdec_av1_slice_lr(9)struct vdec_av1_slice_loop_filter(9)struct vdec_av1_slice_cdef(9)struct vdec_av1_slice_tile(9)struct vdec_av1_slice_uncompressed_header(9)struct vdec_av1_slice_seq_header(9)struct vdec_av1_slice_frame(9)struct vdec_av1_slice_work_buffer(9)struct vdec_av1_slice_frame_info(9)struct vdec_av1_slice_slot(9)struct vdec_av1_slice_fb(9)struct vdec_av1_slice_vsi(9)struct vdec_av1_slice_pfc(9)struct vdec_av1_slice_instance(9)
like?? that's twenty-five structs?? the AV1 slice code has more structs than I have unread notifications π and every single one of them is cross-referenced right here. iykyk π
π the sign-off π
this little legend comes to us from the Kernel API Manual, page struct vdec_av1_slice_mfmv(9), dated October 2026, sourced from ./drivers/media/platform/media... (yes, the footer truncated the path and we are respecting the truncation, don't @ me π).
go read your structs, hydrate, and remember: if the docs just repeat the field name, that's not a you problem, that's main character documentation energy π§ π β¨
π¦ where this page comes from
- nixpkgs unstable: linux-manual 7.3.1
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms