Return the correct number of bytes #435
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This is needed for Firefox to play more than one frame. This applies the fix from umlaeute#190 (comment).
|
i'm still unsure whether this isn't a bug in gstreamer. the kernel docs say:
and further on i think clarifies what "number of bytes occupied" is supposed to mean (although this is then in the docs for the
i'd argue that the payload doesn't include any page-padding. @hverkuil could you shed some light on this? |
|
On 13/07/2021 12:21, umläute wrote:
i'm still unsure whether this isn't a bug in gstreamer.
the kernel docs <https://www.kernel.org/doc/html/v5.11/userspace-api/media/v4l/buffer.html#struct-v4l2-buffer> say:
The number of bytes occupied by the data in the buffer. It depends on the negotiated data format and may change with each buffer for compressed variable size data like JPEG images. [...]
Applications [must set it when type] refers to an output stream. If the application sets this to 0 for an output stream, then |bytesused| will be set to the size of the buffer (see the length
field of this struct) by the driver.
and further on <https://www.kernel.org/doc/html/v5.11/userspace-api/media/v4l/buffer.html#struct-v4l2-plane> i think clarifies what "number of bytes occupied" is supposed to mean:
The number of bytes occupied by data in the plane (its payload).
i'd argue that the payload doesn't include any page-padding.
@hverkuil <https://github.com/hverkuil> could you shed some light on this?
I don't see a lot of context here, so I am not entirely sure what you mean with
page-padding. I assume you refer to the case where the allocated buffer is larger
than is necessary to align with e.g. the page size. In that case bytesused is
set to the actual image data, not including any trailing padding.
However, horizontal padding (i.e. each line is rounded up to a certain multiple)
IS included. So typically for non-compressed images bytesused is equal to
height * bytesperline (ignoring 4:2:2 or 4:2:0 complexities for the moment).
Regards,
Hans
|
|
for me the underlying issue is now fixed on HEAD 845cea0 |
jwrdegoede
added a commit
to jwrdegoede/v4l2loopback
that referenced
this pull request
May 8, 2023
Gstreamer's v4l2sink is known to submit buffers with bytesused set to the length of the buffer instead of the size of the actual image-data within the buffer which is typically smaller due to buffer sizes being rounded op to a page-size: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/2532 Despite this being a long standing issue and their being 2 merge-reqs: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/3713 https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/4527 to try and fix this it looks like this is not going to get fixed in gst's v4l2sink anytime soon and even if once it is fixed many users will likely be using an older v4l2sink which still has this bug. These buffers with a too large bytes used value are causing issues with various apps which reject these buffers when reading from the v4l2loopback device, such as e.g. ffmpeg and firefox. Add a pix_format_has_valid_sizeimage flag which gets set from vidioc_s_fmt_out() if dev->pix_format.sizeimage is known to have just been set to a valid, fixed size (so this e.g. won't be set for MJPG). And then fix this issue by making vidioc_qbuf() truncate V4L2_BUF_TYPE_VIDEO_OUTPUT buffer's bytes_used value to dev->pix_format.sizeimage when this flag is set. Closes umlaeute#190 Closes umlaeute#448 Obsoletes umlaeute#435
|
closing as obsolete |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is needed for Firefox to play more than one frame. This applies
the fix from #190 (comment).