On 21/04/2024 15:52, Nuo Mi wrote:
> For some error bitstreams, a CTU belongs to two slices/entry points.
> If the decoder initializes and submmits the CTU task twice, it may crash the
> program
> or cause it to enter an infinite loop.
>
> Reported-by: Frank Plowman <p...@frankplowman.com>
> ---
> libavcodec/vvc/dec.c | 7 +++++--
> libavcodec/vvc/thread.c | 43 ++++++++++++++++++++++++++++-------------
> libavcodec/vvc/thread.h | 2 +-
> 3 files changed, 36 insertions(+), 16 deletions(-)
>
> diff --git a/libavcodec/vvc/dec.c b/libavcodec/vvc/dec.c
> index 6aeec27eaf..4f7d184e43 100644
> --- a/libavcodec/vvc/dec.c
> +++ b/libavcodec/vvc/dec.c
> @@ -893,10 +893,13 @@ static int wait_delayed_frame(VVCContext *s, AVFrame
> *output, int *got_output)
>
> static int submit_frame(VVCContext *s, VVCFrameContext *fc, AVFrame *output,
> int *got_output)
> {
> - int ret;
> + int ret = ff_vvc_frame_submit(s, fc);
> + if (ret < 0)
> + return ret;
> +
> s->nb_frames++;
> s->nb_delayed++;
> - ff_vvc_frame_submit(s, fc);
> +
> if (s->nb_delayed >= s->nb_fcs) {
> if ((ret = wait_delayed_frame(s, output, got_output)) < 0)
> return ret;
> diff --git a/libavcodec/vvc/thread.c b/libavcodec/vvc/thread.c
> index 01c3ff75b1..3b27811db2 100644
> --- a/libavcodec/vvc/thread.c
> +++ b/libavcodec/vvc/thread.c
> @@ -124,11 +124,17 @@ static void task_init(VVCTask *t, VVCTaskStage stage,
> VVCFrameContext *fc, const
> atomic_store(&t->target_inter_score, 0);
> }
>
> -static void task_init_parse(VVCTask *t, SliceContext *sc, EntryPoint *ep,
> const int ctu_idx)
> +static int task_init_parse(VVCTask *t, SliceContext *sc, EntryPoint *ep,
> const int ctu_idx)
> {
> + if (t->sc) {
> + // the task already inited, error bitstream
> + return AVERROR_INVALIDDATA;
> + }
> t->sc = sc;
> t->ep = ep;
> t->ctu_idx = ctu_idx;
> +
> + return 0;
> }
>
> static uint8_t task_add_score(VVCTask *t, const VVCTaskStage stage)
> @@ -758,24 +764,35 @@ static void submit_entry_point(VVCContext *s,
> VVCFrameThread *ft, SliceContext *
> frame_thread_add_score(s, ft, t->rx, t->ry, VVC_TASK_STAGE_PARSE);
> }
>
> -void ff_vvc_frame_submit(VVCContext *s, VVCFrameContext *fc)
> +int ff_vvc_frame_submit(VVCContext *s, VVCFrameContext *fc)
> {
> VVCFrameThread *ft = fc->ft;
>
> - for (int i = 0; i < fc->nb_slices; i++) {
> - SliceContext *sc = fc->slices[i];
> - for (int j = 0; j < sc->nb_eps; j++) {
> - EntryPoint *ep = sc->eps + j;
> - for (int k = ep->ctu_start; k < ep->ctu_end; k++) {
> - const int rs = sc->sh.ctb_addr_in_curr_slice[k];
> - VVCTask *t = ft->tasks + rs;
> -
> - task_init_parse(t, sc, ep, k);
> - check_colocation(s, t);
> + // We'll handle this in two passes:
> + // Pass 0 to initialize tasks with parser, this will help detect bit
> stream error
> + // Pass 1 to shedule location check and submit the entry point
> + for (int pass = 0; pass < 2; pass++) {
> + for (int i = 0; i < fc->nb_slices; i++) {
> + SliceContext *sc = fc->slices[i];
> + for (int j = 0; j < sc->nb_eps; j++) {
> + EntryPoint *ep = sc->eps + j;
> + for (int k = ep->ctu_start; k < ep->ctu_end; k++) {
> + const int rs = sc->sh.ctb_addr_in_curr_slice[k];
> + VVCTask *t = ft->tasks + rs;
> + if (pass) {
> + check_colocation(s, t);
> + } else {
> + const int ret = task_init_parse(t, sc, ep, k);
> + if (ret < 0)
> + return ret;
> + }
> + }
> + if (pass)
> + submit_entry_point(s, ft, sc, ep);
> }
> - submit_entry_point(s, ft, sc, ep);
> }
> }
> + return 0;
> }
>
> int ff_vvc_frame_wait(VVCContext *s, VVCFrameContext *fc)
> diff --git a/libavcodec/vvc/thread.h b/libavcodec/vvc/thread.h
> index 55bb4ea244..8ac59b2ecf 100644
> --- a/libavcodec/vvc/thread.h
> +++ b/libavcodec/vvc/thread.h
> @@ -30,7 +30,7 @@ void ff_vvc_executor_free(struct AVExecutor **e);
>
> int ff_vvc_frame_thread_init(VVCFrameContext *fc);
> void ff_vvc_frame_thread_free(VVCFrameContext *fc);
> -void ff_vvc_frame_submit(VVCContext *s, VVCFrameContext *fc);
> +int ff_vvc_frame_submit(VVCContext *s, VVCFrameContext *fc);
> int ff_vvc_frame_wait(VVCContext *s, VVCFrameContext *fc);
>
> #endif // AVCODEC_VVC_THREAD_H
This patch fixes most of the fuzz bitstreams I have which enter infinite
loops, but also introduces a regression, turning some other bitstreams
which were okay before into infinite loops.
--
Frank
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
