From: hverkuil@xs4all.nl (Hans Verkuil)
Subject: [Linux-kernel-mentees] [PATCH] v4l2-ctl: support for metadata output
Date: Thu, 3 Oct 2019 09:20:06 +0200 [thread overview]
Message-ID: <cb0e7771-eba5-979d-e873-d889859a3f83@xs4all.nl> (raw)
Message-ID: <20191003072006.F5I6cvYFkKvLW8qSIiG6k9L90BMG6uceUjtSXeSPzt0@z> (raw)
In-Reply-To: <20191003070615.23129-1-bnvandana@gmail.com>
Hi Vandana,
Since the videodev2.h header is copied from the header I can't apply
this yet until support for V4L2_META_FMT_VIVID is merged into the kernel.
I also have a few other comments, see below:
On 10/3/19 9:06 AM, Vandana BN wrote:
> Adds support to test metadata output format V4L2_META_FMT_VIVID.
>
> Signed-off-by: Vandana BN <bnvandana at gmail.com>
> ---
> contrib/freebsd/include/linux/videodev2.h | 1 +
> include/linux/videodev2.h | 1 +
> utils/v4l2-ctl/v4l2-ctl-meta.cpp | 31 +++++++++++++++++++++++
> utils/v4l2-ctl/v4l2-ctl-streaming.cpp | 12 ++++++---
> utils/v4l2-ctl/v4l2-ctl.h | 1 +
> 5 files changed, 43 insertions(+), 3 deletions(-)
>
> diff --git a/contrib/freebsd/include/linux/videodev2.h b/contrib/freebsd/include/linux/videodev2.h
> index 0c12d27f..d14ca0aa 100644
> --- a/contrib/freebsd/include/linux/videodev2.h
> +++ b/contrib/freebsd/include/linux/videodev2.h
> @@ -783,6 +783,7 @@ struct v4l2_pix_format {
> #define V4L2_META_FMT_VSP1_HGT v4l2_fourcc('V', 'S', 'P', 'T') /* R-Car VSP1 2-D Histogram */
> #define V4L2_META_FMT_UVC v4l2_fourcc('U', 'V', 'C', 'H') /* UVC Payload Header metadata */
> #define V4L2_META_FMT_D4XX v4l2_fourcc('D', '4', 'X', 'X') /* D4XX Payload Header metadata */
> +#define V4L2_META_FMT_VIVID v4l2_fourcc('V', 'I', 'V', 'D') /* Vivid Output Metadata */
>
> /* priv field value to indicates that subsequent fields are valid. */
> #define V4L2_PIX_FMT_PRIV_MAGIC 0xfeedcafe
> diff --git a/include/linux/videodev2.h b/include/linux/videodev2.h
> index e2847759..be4727da 100644
> --- a/include/linux/videodev2.h
> +++ b/include/linux/videodev2.h
> @@ -749,6 +749,7 @@ struct v4l2_pix_format {
> #define V4L2_META_FMT_VSP1_HGT v4l2_fourcc('V', 'S', 'P', 'T') /* R-Car VSP1 2-D Histogram */
> #define V4L2_META_FMT_UVC v4l2_fourcc('U', 'V', 'C', 'H') /* UVC Payload Header metadata */
> #define V4L2_META_FMT_D4XX v4l2_fourcc('D', '4', 'X', 'X') /* D4XX Payload Header metadata */
> +#define V4L2_META_FMT_VIVID v4l2_fourcc('V', 'I', 'V', 'D') /* Vivid Output Metadata */
>
> /* priv field value to indicates that subsequent fields are valid. */
> #define V4L2_PIX_FMT_PRIV_MAGIC 0xfeedcafe
> diff --git a/utils/v4l2-ctl/v4l2-ctl-meta.cpp b/utils/v4l2-ctl/v4l2-ctl-meta.cpp
> index eae7438f..2b5781a7 100644
> --- a/utils/v4l2-ctl/v4l2-ctl-meta.cpp
> +++ b/utils/v4l2-ctl/v4l2-ctl-meta.cpp
> @@ -139,10 +139,18 @@ struct vivid_uvc_meta_buf {
> #define UVC_STREAM_SCR (1 << 3)
> #define UVC_STREAM_PTS (1 << 2)
>
> +struct vivid_meta_out_buf {
> + __u8 brightness;
> + __u8 contrast;
> + __u8 saturation;
> + __s16 hue;
> +};
I'd change the __u8 to __u16 in the struct above.
The problem right now is that there is a hole of one byte in the struct
between saturation and hue. That's something we try to avoid.
> +
> void print_meta_buffer(FILE *f, cv4l_buffer &buf, cv4l_fmt &fmt, cv4l_queue &q)
> {
> struct vivid_uvc_meta_buf *vbuf;
> int buf_off = 0;
> + struct vivid_meta_out_buf *vbuf_out;
>
> switch (fmt.g_pixelformat()) {
> case V4L2_META_FMT_UVC:
> @@ -164,5 +172,28 @@ void print_meta_buffer(FILE *f, cv4l_buffer &buf, cv4l_fmt &fmt, cv4l_queue &q)
> le16toh(*(__u16*)(vbuf->buf + buf_off + 4)));
> fprintf(f, "\n");
> break;
> + case V4L2_META_FMT_VIVID:
> + fprintf(f, "VIVID:");
> + vbuf_out = (vivid_meta_out_buf *)q.g_dataptr(buf.g_index(), 0);
> +
> + fprintf(f, " brightness: %u", vbuf_out->brightness);
> + fprintf(f, " contrast: %u", vbuf_out->contrast);
> + fprintf(f, " saturation: %u", vbuf_out->saturation);
> + fprintf(f, " hue: %d\n", vbuf_out->hue);
I would just combine all fprintf's into a single fprintf.
> + break;
> + }
> +}
> +
> +void meta_fillbuffer(int idx, cv4l_fmt &fmt, cv4l_queue &q)
> +{
> + struct vivid_meta_out_buf *vbuf;
Add newline.
> + switch (fmt.g_pixelformat()) {
> + case V4L2_META_FMT_VIVID:
> + vbuf = (vivid_meta_out_buf *)q.g_dataptr(idx, 0);
> + vbuf->brightness = 228;
> + vbuf->contrast = 228;
> + vbuf->saturation = 228;
> + vbuf->hue = 100;
Ah, be a bit more creative :-)
I.e. change values based on the buffer sequence number. That way the image
changes over time.
> + break;
> }
> }
> diff --git a/utils/v4l2-ctl/v4l2-ctl-streaming.cpp b/utils/v4l2-ctl/v4l2-ctl-streaming.cpp
> index 47b7d3f8..fc82d483 100644
> --- a/utils/v4l2-ctl/v4l2-ctl-streaming.cpp
> +++ b/utils/v4l2-ctl/v4l2-ctl-streaming.cpp
> @@ -1146,6 +1146,7 @@ static int do_setup_out_buffers(cv4l_fd &fd, cv4l_queue &q, FILE *fin, bool qbuf
> bool can_fill = false;
> bool is_video = q.g_type() == V4L2_BUF_TYPE_VIDEO_OUTPUT ||
> q.g_type() == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE;
> + bool is_meta = q.g_type() == V4L2_BUF_TYPE_META_OUTPUT;
>
> if (q.obtain_bufs(&fd))
> return QUEUE_ERROR;
> @@ -1177,7 +1178,6 @@ static int do_setup_out_buffers(cv4l_fd &fd, cv4l_queue &q, FILE *fin, bool qbuf
> if (V4L2_FIELD_HAS_T_OR_B(field))
> output_field = (stream_out_std & V4L2_STD_525_60) ?
> V4L2_FIELD_BOTTOM : V4L2_FIELD_TOP;
> -
Spurious line deletion.
> if (is_video) {
> tpg_init(&tpg, 640, 360);
> tpg_alloc(&tpg, fmt.g_width());
> @@ -1244,6 +1244,9 @@ static int do_setup_out_buffers(cv4l_fd &fd, cv4l_queue &q, FILE *fin, bool qbuf
> if (fin && !fill_buffer_from_file(fd, q, buf, fmt, fin))
> return QUEUE_STOPPED;
>
> + if (is_meta)
> + meta_fillbuffer(i, fmt, q);
This should be moved to before the 'if (fin...' line.
Otherwise this would overwrite the data read from a file if fin != NULL.
> +
> if (fmt.g_pixelformat() == V4L2_PIX_FMT_FWHT_STATELESS) {
> int media_fd = mi_get_media_fd(fd.g_fd());
>
> @@ -1266,7 +1269,6 @@ static int do_setup_out_buffers(cv4l_fd &fd, cv4l_queue &q, FILE *fin, bool qbuf
> }
> if (qbuf) {
> fps_timestamps fps_ts;
> -
Spurious line deletion.
> set_time_stamp(buf);
> if (fd.qbuf(buf))
> return QUEUE_ERROR;
> @@ -1480,6 +1482,7 @@ static int do_handle_out(cv4l_fd &fd, cv4l_queue &q, FILE *fin, cv4l_buffer *cap
> bool stopped, bool ignore_count_skip)
> {
> cv4l_buffer buf(q);
> + bool is_meta = q.g_type() == V4L2_BUF_TYPE_META_OUTPUT;
> int ret = 0;
>
> if (cap) {
> @@ -1522,6 +1525,7 @@ static int do_handle_out(cv4l_fd &fd, cv4l_queue &q, FILE *fin, cv4l_buffer *cap
> fprintf(stderr, ", dropped buffers: %u", dropped);
> fprintf(stderr, "\n");
> }
> +
> if (stopped)
> return 0;
>
> @@ -1542,6 +1546,8 @@ static int do_handle_out(cv4l_fd &fd, cv4l_queue &q, FILE *fin, cv4l_buffer *cap
> tpg_fillbuffer(&tpg, stream_out_std, j,
> (u8 *)q.g_dataptr(buf.g_index(), j));
> }
> + if (is_meta)
> + meta_fillbuffer(buf.g_index(), fmt, q);
>
> if (fmt.g_pixelformat() == V4L2_PIX_FMT_FWHT_STATELESS) {
> if (ioctl(buf.g_request_fd(), MEDIA_REQUEST_IOC_REINIT, NULL)) {
> @@ -2039,7 +2045,7 @@ static void streaming_set_out(cv4l_fd &fd, cv4l_fd &exp_fd)
>
> if (!(capabilities & (V4L2_CAP_VIDEO_OUTPUT | V4L2_CAP_VIDEO_OUTPUT_MPLANE |
> V4L2_CAP_VBI_OUTPUT | V4L2_CAP_SLICED_VBI_OUTPUT |
> - V4L2_CAP_SDR_OUTPUT |
> + V4L2_CAP_SDR_OUTPUT | V4L2_CAP_META_OUTPUT |
> V4L2_CAP_VIDEO_M2M | V4L2_CAP_VIDEO_M2M_MPLANE))) {
> fprintf(stderr, "unsupported stream type\n");
> return;
> diff --git a/utils/v4l2-ctl/v4l2-ctl.h b/utils/v4l2-ctl/v4l2-ctl.h
> index 2c861d82..a12a2ba7 100644
> --- a/utils/v4l2-ctl/v4l2-ctl.h
> +++ b/utils/v4l2-ctl/v4l2-ctl.h
> @@ -409,6 +409,7 @@ void meta_set(cv4l_fd &fd);
> void meta_get(cv4l_fd &fd);
> void meta_list(cv4l_fd &fd);
> void print_meta_buffer(FILE *f, cv4l_buffer &buf, cv4l_fmt &fmt, cv4l_queue &q);
> +void meta_fillbuffer(int idx, cv4l_fmt &fmt, cv4l_queue &q);
>
> // v4l2-ctl-subdev.cpp
> void subdev_usage(void);
>
Regards,
Hans
next prev parent reply other threads:[~2019-10-03 7:20 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-03 7:06 [Linux-kernel-mentees] [PATCH] v4l2-ctl: support for metadata output bnvandana
2019-10-03 7:06 ` Vandana BN
2019-10-03 7:20 ` hverkuil [this message]
2019-10-03 7:20 ` Hans Verkuil
2019-10-03 10:54 ` [Linux-kernel-mentees] [PATCH v2] " bnvandana
2019-10-03 10:54 ` Vandana BN
2019-10-14 11:44 ` hverkuil
2019-10-14 11:44 ` Hans Verkuil
2019-10-14 13:46 ` hverkuil
2019-10-14 13:46 ` Hans Verkuil
2019-10-15 12:30 ` [Linux-kernel-mentees] [PATCH v3] " bnvandana
2019-10-15 12:30 ` Vandana BN
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cb0e7771-eba5-979d-e873-d889859a3f83@xs4all.nl \
--to=hverkuil@xs4all.nl \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).