Re: [PATCH] venus: venc: add support for V4L2_CID_MPEG_VIDEO_H264_8X8_TRANSFORM control

From: Fritz Koenig
Date: Wed Jun 02 2021 - 13:42:09 EST

On Sun, May 30, 2021 at 10:22 PM Mansur Alisha Shaik
<mansur@xxxxxxxxxxxxxx> wrote:
>
> Add support for V4L2_CID_MPEG_VIDEO_H264_8X8_TRANSFORM control for
> H264 high profile and constrained high profile.
>
> Signed-off-by: Mansur Alisha Shaik <mansur@xxxxxxxxxxxxxx>
> ---
> drivers/media/platform/qcom/venus/core.h | 1 +
> drivers/media/platform/qcom/venus/hfi_cmds.c | 10 ++++++++++
> drivers/media/platform/qcom/venus/hfi_helper.h | 5 +++++
> drivers/media/platform/qcom/venus/venc.c | 11 +++++++++++
> drivers/media/platform/qcom/venus/venc_ctrls.c | 12 +++++++++++-
> 5 files changed, 38 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/media/platform/qcom/venus/core.h b/drivers/media/platform/qcom/venus/core.h
> index 745f226..103fbd8 100644
> --- a/drivers/media/platform/qcom/venus/core.h
> +++ b/drivers/media/platform/qcom/venus/core.h
> @@ -235,6 +235,7 @@ struct venc_controls {
> u32 h264_loop_filter_mode;
> s32 h264_loop_filter_alpha;
> s32 h264_loop_filter_beta;
> + u32 h264_8x8_transform;
>
> u32 hevc_i_qp;
> u32 hevc_p_qp;
> diff --git a/drivers/media/platform/qcom/venus/hfi_cmds.c b/drivers/media/platform/qcom/venus/hfi_cmds.c
> index 11a8347..61d04a5 100644
> --- a/drivers/media/platform/qcom/venus/hfi_cmds.c
> +++ b/drivers/media/platform/qcom/venus/hfi_cmds.c
> @@ -1178,6 +1178,7 @@ pkt_session_set_property_4xx(struct hfi_session_set_property_pkt *pkt,
> {
> void *prop_data;
>
> +
> if (!pkt || !cookie || !pdata)
> return -EINVAL;
>
> @@ -1227,6 +1228,15 @@ pkt_session_set_property_4xx(struct hfi_session_set_property_pkt *pkt,
> break;
> }
>
> + case HFI_PROPERTY_PARAM_VENC_H264_TRANSFORM_8X8: {
> + struct hfi_h264_8x8x_transform *in = pdata, *tm = prop_data;
> +
> + tm->enable_type = in->enable_type;
> + pkt->shdr.hdr.size += sizeof(u32) + sizeof(*tm);
> + break;
> +
> + }
> +
> case HFI_PROPERTY_CONFIG_VENC_MAX_BITRATE:
> case HFI_PROPERTY_CONFIG_VDEC_POST_LOOP_DEBLOCKER:
> case HFI_PROPERTY_PARAM_BUFFER_ALLOC_MODE:
> diff --git a/drivers/media/platform/qcom/venus/hfi_helper.h b/drivers/media/platform/qcom/venus/hfi_helper.h
> index 63cd347..81d0536 100644
> --- a/drivers/media/platform/qcom/venus/hfi_helper.h
> +++ b/drivers/media/platform/qcom/venus/hfi_helper.h
> @@ -510,6 +510,7 @@
> #define HFI_PROPERTY_PARAM_VENC_MAX_NUM_B_FRAMES 0x2005020
> #define HFI_PROPERTY_PARAM_VENC_H264_VUI_BITSTREAM_RESTRC 0x2005021
> #define HFI_PROPERTY_PARAM_VENC_PRESERVE_TEXT_QUALITY 0x2005023
> +#define HFI_PROPERTY_PARAM_VENC_H264_TRANSFORM_8X8 0x2005025
> #define HFI_PROPERTY_PARAM_VENC_HIER_P_MAX_NUM_ENH_LAYER 0x2005026
> #define HFI_PROPERTY_PARAM_VENC_DISABLE_RC_TIMESTAMP 0x2005027
> #define HFI_PROPERTY_PARAM_VENC_INITIAL_QP 0x2005028
> @@ -565,6 +566,10 @@ struct hfi_bitrate {
> u32 layer_id;
> };
>
> +struct hfi_h264_8x8x_transform {
> + u32 enable_type;
> +};
> +
> #define HFI_CAPABILITY_FRAME_WIDTH 0x01
> #define HFI_CAPABILITY_FRAME_HEIGHT 0x02
> #define HFI_CAPABILITY_MBS_PER_FRAME 0x03
> diff --git a/drivers/media/platform/qcom/venus/venc.c b/drivers/media/platform/qcom/venus/venc.c
> index 8dd49d4..4ecf331 100644
> --- a/drivers/media/platform/qcom/venus/venc.c
> +++ b/drivers/media/platform/qcom/venus/venc.c
> @@ -567,6 +567,7 @@ static int venc_set_properties(struct venus_inst *inst)
> struct hfi_h264_vui_timing_info info;
> struct hfi_h264_entropy_control entropy;
> struct hfi_h264_db_control deblock;
> + struct hfi_h264_8x8x_transform h264_transform;
>
> ptype = HFI_PROPERTY_PARAM_VENC_H264_VUI_TIMING_INFO;
> info.enable = 1;
> @@ -597,6 +598,16 @@ static int venc_set_properties(struct venus_inst *inst)
> ret = hfi_session_set_property(inst, ptype, &deblock);
> if (ret)
> return ret;
> +
> + ptype = HFI_PROPERTY_PARAM_VENC_H264_TRANSFORM_8X8;
> + if (ctr->profile.h264 == HFI_H264_PROFILE_HIGH ||
> + ctr->profile.h264 == HFI_H264_PROFILE_CONSTRAINED_HIGH)
> + h264_transform.enable_type = ctr->h264_8x8_transform;
> +
> + ret = hfi_session_set_property(inst, ptype, &h264_transform);
> + if (ret)
> + return ret;
> +
> }
>
> if (inst->fmt_cap->pixfmt == V4L2_PIX_FMT_H264 ||
> diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c
> index 637c92f..e3ef611 100644
> --- a/drivers/media/platform/qcom/venus/venc_ctrls.c
> +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c
> @@ -319,6 +319,13 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl)
> case V4L2_CID_COLORIMETRY_HDR10_MASTERING_DISPLAY:
> ctr->mastering = *ctrl->p_new.p_hdr10_mastering;
> break;
> + case V4L2_CID_MPEG_VIDEO_H264_8X8_TRANSFORM:
> + if (ctr->profile.h264 != HFI_H264_PROFILE_HIGH ||
> + ctr->profile.h264 != HFI_H264_PROFILE_CONSTRAINED_HIGH)

This appears to be incorrect as the comparison will always be true. I
think it should be written as:
if (ctr->profile.h264 == HFI_H264_PROFILE_HIGH ||
ctr->profile.h264 == HFI_H264_PROFILE_CONSTRAINED_HIGH)
ctr->h264_8x8_transform = ctrl->val;

> + return -EINVAL;

I'm not sure -EINVAL is appropriate here. venc_op_s_ctrl will be
called to initialize the default control values from
v4l2_ctrl_handler_setup. If the default profile isn't high or
constrained high, the driver will fail to initialize.

> +
> + ctr->h264_8x8_transform = ctrl->val;
> + break;
>
> default:
> return -EINVAL;
> }
> @@ -334,7 +341,7 @@ int venc_ctrl_init(struct venus_inst *inst)
> {
> int ret;
>
> - ret = v4l2_ctrl_handler_init(&inst->ctrl_handler, 57);
> + ret = v4l2_ctrl_handler_init(&inst->ctrl_handler, 58);
> if (ret)
> return ret;
>
> @@ -438,6 +445,9 @@ int venc_ctrl_init(struct venus_inst *inst)
> V4L2_CID_MPEG_VIDEO_H264_I_FRAME_MIN_QP, 1, 51, 1, 1);
>
> v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops,
> + V4L2_CID_MPEG_VIDEO_H264_8X8_TRANSFORM, 0, 1, 1, 0);
> +
> + v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops,
> V4L2_CID_MPEG_VIDEO_H264_P_FRAME_MIN_QP, 1, 51, 1, 1);
>
> v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops,
> --
> QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member
> of Code Aurora Forum, hosted by The Linux Foundation
>