[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <3093393.es1Za2YUDY@jernej-laptop>
Date: Sun, 27 Oct 2019 22:56:45 +0100
From: Jernej Škrabec <jernej.skrabec@...l.net>
To: linux-sunxi@...glegroups.com
Cc: mripard@...nel.org, paul.kocialkowski@...tlin.com,
mchehab@...nel.org, hverkuil-cisco@...all.nl,
gregkh@...uxfoundation.org, wens@...e.org, p.zabel@...gutronix.de,
ezequiel@...labora.com, boris.brezillon@...labora.com,
linux-media@...r.kernel.org, devel@...verdev.osuosl.org,
linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [linux-sunxi] [PATCH 4/4] media: cedrus: hevc: Add support for multiple slices
Dne sobota, 26. oktober 2019 ob 19:47:03 CET je Jernej Skrabec napisal(a):
> Now that segment address is available, support for multi-slice frames
> can be easily added.
>
> Signed-off-by: Jernej Skrabec <jernej.skrabec@...l.net>
> ---
> .../staging/media/sunxi/cedrus/cedrus_h265.c | 21 +++++++++++++++----
> .../staging/media/sunxi/cedrus/cedrus_video.c | 1 +
> 2 files changed, 18 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> b/drivers/staging/media/sunxi/cedrus/cedrus_h265.c index
> 888bfd5ca224..e909adf6f30f 100644
> --- a/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> +++ b/drivers/staging/media/sunxi/cedrus/cedrus_h265.c
> @@ -366,15 +366,28 @@ static void cedrus_h265_setup(struct cedrus_ctx *ctx,
> reg = VE_DEC_H265_BITS_END_ADDR_BASE(src_buf_end_addr);
> cedrus_write(dev, VE_DEC_H265_BITS_END_ADDR, reg);
>
> - /* Coding tree block address: start at the beginning. */
> + /* Coding tree block address */
> reg = VE_DEC_H265_DEC_CTB_ADDR_X(0) |
VE_DEC_H265_DEC_CTB_ADDR_Y(0);
> + if (!ctx->fh.m2m_ctx->new_frame) {
There is no reason why would this block be guarded by above if clause. I'll
remove it in next revision. However, it uncovered a bug where new_frame is set
to true for every slice. I have to debug this further. At this point I can't
say for sure if it is in kernel or in ffmpeg.
Best regards,
Jernej
> + unsigned int log2_max_luma_coding_block_size =
> + sps->log2_min_luma_coding_block_size_minus3 +
3 +
> + sps->log2_diff_max_min_luma_coding_block_size;
> + unsigned int ctb_size_luma =
> + 1UL << log2_max_luma_coding_block_size;
> + unsigned int width_in_ctb_luma =
> + DIV_ROUND_UP(sps->pic_width_in_luma_samples,
ctb_size_luma);
> +
> + reg = VE_DEC_H265_DEC_CTB_ADDR_X(slice_params-
>slice_segment_addr %
> width_in_ctb_luma); + reg |=
> VE_DEC_H265_DEC_CTB_ADDR_Y(slice_params->slice_segment_addr /
> width_in_ctb_luma); + }
> cedrus_write(dev, VE_DEC_H265_DEC_CTB_ADDR, reg);
>
> cedrus_write(dev, VE_DEC_H265_TILE_START_CTB, 0);
> cedrus_write(dev, VE_DEC_H265_TILE_END_CTB, 0);
>
> /* Clear the number of correctly-decoded coding tree blocks. */
> - cedrus_write(dev, VE_DEC_H265_DEC_CTB_NUM, 0);
> + if (ctx->fh.m2m_ctx->new_frame)
> + cedrus_write(dev, VE_DEC_H265_DEC_CTB_NUM, 0);
>
> /* Initialize bitstream access. */
> cedrus_write(dev, VE_DEC_H265_TRIGGER,
VE_DEC_H265_TRIGGER_INIT_SWDEC);
> @@ -523,8 +536,8 @@ static void cedrus_h265_setup(struct cedrus_ctx *ctx,
>
V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT,
> pps->flags);
>
> - /* FIXME: For multi-slice support. */
> - reg |=
VE_DEC_H265_DEC_SLICE_HDR_INFO0_FLAG_FIRST_SLICE_SEGMENT_IN_PIC;
> + if (ctx->fh.m2m_ctx->new_frame)
> + reg |=
VE_DEC_H265_DEC_SLICE_HDR_INFO0_FLAG_FIRST_SLICE_SEGMENT_IN_PIC;
>
> cedrus_write(dev, VE_DEC_H265_DEC_SLICE_HDR_INFO0, reg);
>
> diff --git a/drivers/staging/media/sunxi/cedrus/cedrus_video.c
> b/drivers/staging/media/sunxi/cedrus/cedrus_video.c index
> 15cf1f10221b..497b1199d3fe 100644
> --- a/drivers/staging/media/sunxi/cedrus/cedrus_video.c
> +++ b/drivers/staging/media/sunxi/cedrus/cedrus_video.c
> @@ -311,6 +311,7 @@ static int cedrus_s_fmt_vid_out(struct file *file, void
> *priv,
>
> switch (ctx->src_fmt.pixelformat) {
> case V4L2_PIX_FMT_H264_SLICE:
> + case V4L2_PIX_FMT_HEVC_SLICE:
> vq->subsystem_flags |=
> VB2_V4L2_FL_SUPPORTS_M2M_HOLD_CAPTURE_BUF;
> break;
Powered by blists - more mailing lists