From c8f644ec4432089a212c192f71d30bc6addf06fe Mon Sep 17 00:00:00 2001 From: Henry Goffin Date: Wed, 29 Dec 2021 09:20:30 +0000 Subject: [PATCH] frontends/va: ignore incoming frame_num from VA picture parameters The Gallium pipe video "frame_num" variable is internally used as a counter of elapsed reference frames since the last IDR. The incoming frame_num field from VA picture parameters is not equivalent; the VA value may wrap to zero prematurely, as it is a 16-bit struct field with a documented max value of 2^(log2_max_frame_num_minus4 + 4)-1. This change improves "infinite GOP" single-client live streaming, where it is reasonable for the server to desire an endless series of P-frames without IDR. Without this change, it is difficult/impossible for an application to encode a P- or B-frame after the VA frame_num field wraps around to zero, depending on the backend encoder implementation. This change has no effect on existing applications that always signal an IDR frame and reset the VA frame_num to zero before it wraps around. For example, the FFmpeg vaapi encoder ignores the VA documentation and sends an un-wrapped VA frame_num, which results in identical computation of the internal frame_num (as long as each GOP is less than 65536 frames). Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5768 Reviewed-by: Thong Thai patch revision 3: correctly avoid incrementing frame_num when the encoded frame is not a reference, per h264 spec and ffmpeg behavior Part-of: --- src/gallium/frontends/va/picture.c | 2 ++ src/gallium/frontends/va/picture_h264_enc.c | 5 +++-- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/src/gallium/frontends/va/picture.c b/src/gallium/frontends/va/picture.c index 84cc16ccb01..ba8f9de64dc 100644 --- a/src/gallium/frontends/va/picture.c +++ b/src/gallium/frontends/va/picture.c @@ -813,6 +813,8 @@ vlVaEndPicture(VADriverContextP ctx, VAContextID context_id) context->first_single_submitted = false; surf->force_flushed = true; } + if (!context->desc.h264enc.not_referenced) + context->desc.h264enc.frame_num++; } else if (context->decoder->entrypoint == PIPE_VIDEO_ENTRYPOINT_ENCODE && u_reduce_video_profile(context->templat.profile) == PIPE_VIDEO_FORMAT_HEVC) context->desc.h265enc.frame_num++; diff --git a/src/gallium/frontends/va/picture_h264_enc.c b/src/gallium/frontends/va/picture_h264_enc.c index 3f9b3b199b7..1da11d29188 100644 --- a/src/gallium/frontends/va/picture_h264_enc.c +++ b/src/gallium/frontends/va/picture_h264_enc.c @@ -36,7 +36,8 @@ vlVaHandleVAEncPictureParameterBufferTypeH264(vlVaDriver *drv, vlVaContext *cont vlVaBuffer *coded_buf; h264 = buf->data; - context->desc.h264enc.frame_num = h264->frame_num; + if (h264->pic_fields.bits.idr_pic_flag == 1) + context->desc.h264enc.frame_num = 0; context->desc.h264enc.not_referenced = !h264->pic_fields.bits.reference_pic_flag; context->desc.h264enc.pic_order_cnt = h264->CurrPic.TopFieldOrderCnt; if (context->desc.h264enc.gop_cnt == 0) @@ -54,7 +55,7 @@ vlVaHandleVAEncPictureParameterBufferTypeH264(vlVaDriver *drv, vlVaContext *cont _mesa_hash_table_insert(context->desc.h264enc.frame_idx, UINT_TO_PTR(h264->CurrPic.picture_id + 1), - UINT_TO_PTR(h264->frame_num)); + UINT_TO_PTR(context->desc.h264enc.frame_num)); if (h264->pic_fields.bits.idr_pic_flag == 1) context->desc.h264enc.picture_type = PIPE_H2645_ENC_PICTURE_TYPE_IDR;