V4L2 Codec/camera fixes #4765
Closed
V4L2 Codec/camera fixes #4765
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
pelwell
reviewed
Dec 13, 2021
6by9
force-pushed
the
rpi-5.10.y-codecs
branch
from
53fd070 to
20e3d07
Compare
|
Hold fire - messed up in last changes. |
MMAL error values were being returned to userspace by some of the functions. Map them to appropriate Linux ones. Signed-off-by: Dave Stevenson <[email protected]>
The mmal-vchiq layer allows submission of buffers before the port is enabled, and it queues them to be sent on port enable. There isn't a way of returning those buffers, which leads to videobuf2 complaining of buffers not being returned by drivers. Add such a call, and reuse the same code in port_disable. Signed-off-by: Dave Stevenson <[email protected]>
Buffers submitted before VIDIOC_STREAMON would be passed to the mmal-vchiq layer, but weren't returned to videobuf2 if STREAMON failed as there wasn't a call into mmal-vchiq to do so. As there is now a suitable call, ensure buffers are returned. Signed-off-by: Dave Stevenson <[email protected]>
…ailure If STREAMON failed, then buffers weren't being returned to videobuf2, leading to it producing WARNs in the kernel logs. Return all buffers stored with the V4L2 M2M framework and with mmal-vchiq should STREAMON fail. Signed-off-by: Dave Stevenson <[email protected]>
If the encode tunnel failed to enable, then the camera was left enabled and that caused a failure on a subsequent request to change format. Disable the camera on failure. Signed-off-by: Dave Stevenson <[email protected]>
Replace several instances of if (foo == 0) with if (!foo) and if (foo != 0) with if (foo) Signed-off-by: Dave Stevenson <[email protected]>
6by9
force-pushed
the
rpi-5.10.y-codecs
branch
from
20e3d07 to
4a8ff7e
Compare
|
OK, that should work now. |
|
I'll wait for some feedback from @jc-kynesim, but LGTM. |
|
Ok, I have tested this code, and the result is no HW decoding for video files. There are no errors reported, but I see only a black square instead of a rendered picture. Don't you know whether a newer GStreamer is required for this to work? I am on 1.16. |
|
I've tried a newer GStreamer 1.18.4, and it is the same. When rendering a web page with an mp4 file included, the only black square is present on the screen. Load and unload of that element is fine without any error or warning. When I play using gst-play, I could see some warnings: Here is the output of the gst-inspect | grep v4l |
|
@timemaster5 GStreamer issue may well be the same as raspberrypi/firmware#1673 I'll pull these patches forward to 5.15 as 5.10 is effectively frozen now. |
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.
@jc-kynesim Fixes for the kernel WARN due to us not returning any buffers that have been queued with mmal-vchiq before the port is enabled.