Pi Camera module v2 not working in Chromium/Electron with kernel v4.4 #1498
Labels
Waiting for internal comment
Waiting for comment from a member of the Raspberry Pi engineering team
Comments
|
Why duplicate the issue - #1499 ? Unrelated to #849 - GStreamer is objecting to particular formatting of an H264 elementary stream. |
|
@6by9 sorry for the duplicate (and pinging an old unrelated issue), I had connectivity issues. Did not have the chance to test with a V1 camera. Here is the full log with Also tried with latest electron |
|
Your app is asking for a 0x0 YU12 image - that sounds a fairly unhelpful resolution, even if it is relying on V4L2 to clip it to a valid range (16x16 is the minimum supposedly supported on Pi) Admittedly something odd is going on on the GPU side - 16x16 doesn't return any buffers. 24x16 and 32x16 do. 18x16 doesn't. I haven't run any extra tests yet to find exactly the limit. I would however suggest you investigate why your application is asking for such a low resolution. That sounds like a bug on that side, possibly down to not supporting drivers that advertise stepwise resolution support from VIDIOC_ENUM_FRAMESIZES (see http://hverkuil.home.xs4all.nl/spec/media.html#vidioc-enum-framesizes) (edited to correct the requested resolution) |
|
I don't think there is anything specific to the app itself, here is the JavaScript source code, and it is basically just Chrome's navigator.getUserMedia, that does not look like it allows any format/size customization. |
|
From the docs section "Examples/Width and height" So you can ask for specific width and height. Note too the main banner heading:
I can fix the underlying issue, but you'll get a 16x16 image streamed into your browser - it will look terrible, and that is a result of the app asking for something silly. |
|
Indeed, I refered/linked to an incomplete translated page, will try the navigator.mediaDevices.getUserMedia with different sizes and report back. Thanks! |
|
Still no luck tonight with this code (working thanks to this polyfill): Got the following output: So it looks like width/height contraints won't impact the format. Also, getUserMedia does usually looks good on OSX and Linux, is this 16x16 issue ARM specific? |
|
The app is still asking for 0x0, so it's still resetting to the default minimum of 16x16 which we know has issues. I am still trying to pin down why it doesn't work. Most USB webcams only support specific resolutions. TBH My simplest fix is going to be reset the minimum from 16x16 to 32x16 by bumping MIN_WIDTH at https://github.com/raspberrypi/linux/blob/rpi-4.4.y/drivers/media/platform/bcm2835/bcm2835-camera.c#L39. It'll work and I'll close this issue, but you'll get rubbish pictures - you need to work out how and why the framework you're using asks for such a silly resolution as that is not a kernel issue. |
|
The exact same code with a Logitech C920 webcam plugged on the Pi (USB) does work and produce a very good looking image. Is there any chance that it only tries to grab a 0x0/16x16 frame to test if the camera actually works before switching to a better resolution? Is there a way to get similar debug output than EDIT: Just created an issue on Chromium bug tracker: https://bugs.chromium.org/p/chromium/issues/detail?id=616007 EDIT 2: or maybe it just tries to grab the default resolution, the C920 first resolution in the list is 640x480, probably acting as default? Maybe a fix would bt to add a similar default/native discrete resolution for the bcm2835_v4l2 module? EDIT 3: Also tried https://webrtchacks.github.io/WebRTC-Camera-Resolution/ (from this blog post), that polls several resolution, and it fails for every resolution using the camera module with a With the C920 plugged in: With the Camera module v2.1 plugged in: |
|
There is a default resolution set on the Pi - 1024x768 JPEG.
Your app asked for 0x0. We checked against our capabilities and returned the closest match - 16x16. It looks like the UVC (USB Video Class - almost all USB webcams) driver selects the default resolution advertised by the webcam - https://github.com/raspberrypi/linux/blob/rpi-4.4.y/drivers/media/usb/uvc/uvc_v4l2.c#L176 I think you must be missing something in your config. I doubt it's the latest code as it hasn't been updated in almost 2 years, but https://src.chromium.org/viewvc/chrome/trunk/src/media/video/capture/linux/video_capture_device_linux.cc line 258 is the call setting up the camera. Note that there are width & height parameters passed in to there. Google for "chromium getUserMedia resolution" seems to throw up a bundle of hits, including things like https://bugs.chromium.org/p/chromium/issues/detail?id=312928 where someone else was having issues trying to request a resolution. Have you followed all of the recommendations there? I'm going to create a patch to shift the minimum res to being 32x32, otherwise I'm going to waste too much time on a calling app issue. |
6by9
added a commit
to 6by9/linux
that referenced
this issue
May 31, 2016
raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
|
@6by9 Thank you for your time and help, I know it is precious. We will see what the chromium folks say on the ticket I've created. I still think it's not an issue from my side (but indeed it may be a Chromium issue) as https://webrtchacks.github.io/WebRTC-Camera-Resolution/ works and properly set constraints for every combo (device/os+camera) I can try it with, it only fails with the Pi + Camera module. For instance, here is the related output on the Pi for the C920: And here it is for the Camera module: https://bugs.chromium.org/p/chromium/issues/detail?id=312928 is related to multiple tab locking, I'm only starting a single electron app on a single url (nothing else is running on the device). Thanks! |
No, the first 11 comments are all related to setting the resolution. Multiple tabs are only mentioned in comment 14 causing the issue in comment 13. Old commit, but may explain some things: https://chromium.googlesource.com/chromium/src/+/1577a17cdb0fe92f3e2dcd1ea2327ea814ca6549%5E!/ That code was removed in the patch, but that implies STEPWISE is not supported. The bug report is still open, but that may be an oversight. PR to increase minimum resolution to 32x32 has been merged. Either rebuild your own version of the kernel based on top of tree, or wait for an update to https://github.com/Hexxeh/rpi-firmware that can be collected with |
|
rpi-update kernel contains @6by9's increase of resolution to 32x32. It may be worth re-testing with this update. |
|
Ok so I got it working with https://simpl.info/getusermedia/sources, but as predicted, I get a terrible image quality. Still no luck with custom resolutions: https://webrtchacks.github.io/WebRTC-Camera-Resolution/ I'm running an up-to-date Electron EDIT 1: Just found that I can simply add the https://simpl.info/getusermedia/sources outputs: Which links to this file: That confirms your suspicion: So http://crbug.com/249953 it is. Any chance that the Raspberry Pi could provide both discrete and stepwise resolutions? |
No, they're mutually exclusive, reported back via a union - see http://hverkuil.home.xs4all.nl/spec/media.html#vidioc-enum-framesizes and http://hverkuil.home.xs4all.nl/spec/media.html#v4l2-frmsizeenum. It looks like they have already fixed it to some degree with the updated code I linked to, and it makes a call over which resolutions it wants to query for. Are you feeling brave and fancy compiling top of tree?! Unless your version has long term support, I suspect that is all the Chromium guys are going to direct you to do anyway. |
|
BTW You're going to hit an unsupported at https://chromium.googlesource.com/chromium/src/+/51.0.2704.54/media/capture/video/linux/video_capture_device_factory_linux.cc#71 too, for frame intervals (aka frame rate). |
popcornmix
pushed a commit
that referenced
this issue
May 31, 2016
#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
Being excited to play with webRTC on the Pi, I obviously really like this idea. I guess we could advertise relatively standard resolutions in the That would fix 99% of the webRTC use cases.
I don't really get the relation between chromium's video_capture_device_factory_linux.cc and the code you linked webrtc's device_info_linux.cc Though I'd gladly spend some time digging into it, I am a bit clueless regarding the step needed to compile top-of-tree of |
|
YUYV is not the ideal format on Pi. I'd recommend you use V4L2_PIX_FMT_YUV420. My bad on directory. I'd followed https://www.chromium.org/developers/how-tos/get-the-code to grab the depot-tools and then So first step is to get ToT building, expecting it still to fail. |
popcornmix
pushed a commit
that referenced
this issue
Jun 1, 2016
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment)
popcornmix
pushed a commit
that referenced
this issue
Jun 1, 2016
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment)
popcornmix
pushed a commit
that referenced
this issue
Jun 1, 2016
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment)
popcornmix
pushed a commit
that referenced
this issue
Jun 3, 2016
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment)
popcornmix
pushed a commit
that referenced
this issue
Jun 6, 2016
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment)
amichelotti
pushed a commit
to amichelotti/ubuntu-vme-xenial
that referenced
this issue
Oct 19, 2017
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
popcornmix
pushed a commit
that referenced
this issue
Oct 29, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown #817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO #1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 #1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: #1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
raspbian-autopush
pushed a commit
to raspbian-packages/linux-4.9
that referenced
this issue
Nov 2, 2017
commit e5a33f9 Author: Vincent Sanders <vincent.sanders@collabora.co.uk> Date: Wed Jan 30 12:45:18 2013 +0000 bcm2835: add v4l2 camera device - Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi/linux#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi/linux#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi/linux#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability Gbp-Pq: Topic rpi Gbp-Pq: Name rpi_1049_e5a33f9d545dd69ce73f18b874eb9f699a3cc66f.patch
lianhaidong
pushed a commit
to lianhaidong/ubuntu-xenial
that referenced
this issue
Nov 6, 2017
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 18, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 19, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 19, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 19, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 19, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
pelwell
pushed a commit
to pelwell/linux
that referenced
this issue
Nov 19, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
ryncsn
pushed a commit
to ryncsn/linux-rasp
that referenced
this issue
Nov 21, 2017
- Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes raspberrypi#1318 V4L2 driver updates (raspberrypi#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability
liuqun
pushed a commit
to liuqun/linux-raspi2
that referenced
this issue
Dec 14, 2017
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
liuqun
pushed a commit
to liuqun/linux-raspi2
that referenced
this issue
Dec 20, 2017
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
liuqun
pushed a commit
to liuqun/linux-raspi2
that referenced
this issue
Dec 20, 2017
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
polceanum
pushed a commit
to polceanum/ubuntu-xenial
that referenced
this issue
Feb 6, 2018
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
raspbian-autopush
pushed a commit
to raspbian-packages/linux-4.9
that referenced
this issue
Apr 7, 2018
commit f2d51ef Author: Vincent Sanders <vincent.sanders@collabora.co.uk> Date: Wed Jan 30 12:45:18 2013 +0000 bcm2835: add v4l2 camera device - Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi/linux#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi/linux#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi/linux#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability Gbp-Pq: Topic rpi Gbp-Pq: Name rpi_1049_f2d51efbeac118fa3c386f34cab3d7f2be7d5261.patch
APokorny
pushed a commit
to ubports/ubuntu_kernel_xenial
that referenced
this issue
Oct 11, 2018
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
raspbian-autopush
pushed a commit
to raspbian-packages/linux-4.9
that referenced
this issue
Nov 11, 2018
commit f2d51ef Author: Vincent Sanders <vincent.sanders@collabora.co.uk> Date: Wed Jan 30 12:45:18 2013 +0000 bcm2835: add v4l2 camera device - Supports raw YUV capture, preview, JPEG and H264. - Uses videobuf2 for data transfer, using dma_buf. - Uses 3.6.10 timestamping - Camera power based on use - Uses immutable input mode on video encoder Signed-off-by: Daniel Stone <daniels@collabora.com> Signed-off-by: Luke Diamand <luked@broadcom.com> V4L2: Fixes from 6by9 V4L2: Fix EV values. Add manual shutter speed control V4L2 EV values should be in units of 1/1000. Corrected. Add support for V4L2_CID_EXPOSURE_ABSOLUTE which should give manual shutter control. Requires manual exposure mode to be selected first. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct JPEG Q-factor range Should be 1-100, not 0-100 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue of driver jamming if STREAMON failed. Fix issue where the driver was left in a partially enabled state if STREAMON failed, and would then reject many IOCTLs as it thought it was streaming. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix ISO controls. Driver was passing the index to the GPU, and not the desired ISO value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add flicker avoidance controls Add support for V4L2_CID_POWER_LINE_FREQUENCY to set flicker avoidance frequencies. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for frame rate control. Add support for frame rate (or time per frame as V4L2 inverts it) control via s_parm. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Improve G_FBUF handling so we pass conformance Return some sane numbers for get framebuffer so that we pass conformance. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix information advertised through g_vidfmt Width and height were being stored based on incorrect values. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for inline H264 headers Add support for V4L2_CID_MPEG_VIDEO_REPEAT_SEQ_HEADER to control H264 inline headers. Requires firmware fix to work correctly, otherwise format has to be set to H264 before this parameter is set. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix JPEG timestamp issue JPEG images were coming through from the GPU with timestamp of 0. Detect this and give current system time instead of some invalid value. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix issue when switching down JPEG resolution. JPEG buffer size calculation is based on input resolution. Input resolution was being configured after output port format. Caused failures if switching from one JPEG resolution to a smaller one. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable MJPEG encoding Requires GPU firmware update to support MJPEG encoder. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct flag settings for compressed formats Set flags field correctly on enum_fmt_vid_cap for compressed image formats. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: H264 profile & level ctrls, FPS control and auto exp pri Several control handling updates. H264 profile and level controls. Timeperframe/FPS reworked to add V4L2_CID_EXPOSURE_AUTO_PRIORITY to select whether AE is allowed to override the framerate specified. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Correct BGR24 to RGB24 in format table Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add additional pixel formats. Correct colourspace Adds the other flavours of YUYV, and NV12. Corrects the overlay advertised colourspace. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Drop logging msg from info to debug Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Initial pass at scene modes. Only supports exposure mode and metering modes. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add manual white balance control. Adds support for V4L2_CID_RED_BALANCE and V4L2_CID_BLUE_BALANCE. Only has an effect if V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE has V4L2_WHITE_BALANCE_MANUAL selected. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> config: Enable V4L / MMAL driver V4L2: Increase the MMAL timeout to 3sec MJPEG codec flush is now taking longer and results in a kernel panic if the driver has stopped waiting for the result when it finally completes. Increase the timeout value from 1 to 3secs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for setting H264_I_PERIOD Adds support for the parameter V4L2_CID_MPEG_VIDEO_H264_I_PERIOD to set the frequency with which I frames are produced. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Enable GPU function for removing padding from images. GPU can now support arbitrary strides, although may require additional processing to achieve it. Enable this feature so that the images delivered are the size requested. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for V4L2_PIX_FMT_BGR32 Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Set the colourspace to avoid odd YUV-RGB conversions Removes the amiguity from the conversion routines and stops them dropping back to the SD vs HD choice of coeffs. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Make video/still threshold a run-time param Move the define for at what resolution the driver switches from a video mode capture to a stills mode capture to module parameters. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Fix incorrect pool sizing Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add option to disable enum_framesizes. Gstreamer's handling of a driver that advertises V4L2_FRMSIZE_TYPE_STEPWISE to define the supported resolutions is broken. See bug https://bugzilla.gnome.org/show_bug.cgi?id=726521 Optional parameter of gst_v4l2src_is_broken added. If non-zero, the driver claims not to support that ioctl, and gstreamer should be happy again (it guesses a set of defaults for itself). Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Add support for more image formats Adds YVU420 (YV12), YVU420SP (NV21), and BGR888. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> V4L2: Extend range for V4L2_CID_MPEG_VIDEO_H264_I_PERIOD Request to extend the range from the fairly arbitrary 1000 frames (33 seconds at 30fps). Extend out to the max range supported (int32 value). Also allow 0, which is handled by the codec as only send an I-frame on the first frame and never again. There may be an exception if it detects a significant scene change, but there's no easy way around that. Signed-off-by: Dave Stevenson <dsteve@broadcom.com> bcm2835-camera: stop_streaming now has a void return BCM2835-V4L2: Fix compliance test failures VIDIOC_TRY_FMT and VIDIOC_S_FMT tests were faling due to reporting V4L2_COLORSPACE_JPEG when the colour format wasn't V4L2_PIX_FMT_JPEG. Now reports V4L2_COLORSPACE_SMPTE170M for YUV formats. bcm2835 camera planar/packed stride length Added a field to the mmal_fmt struct used to compute the bytes per line when using a particular format. This results in the correct stride being calculated even when the format is planar. Signed-off-by: Garrett Wilson <g@floft.net> bcm2835: camera: check for scene not being found static analysis by cppcheck detected some potential NULL pointer dereference issues: [drivers/media/platform/bcm2835/controls.c:854]: (error) Possible null pointer dereference: scene (and lines 858, 859 too) it is possible that scene is not found because of an invalue ctrl->val and is therefore NULL and hence causing a null pointer dereference. Signed-off-by: Colin Ian King <colin.king@canonical.com> bcm2835: memcpy port data to m rather than rmsg static analysis by cppcheck detected a memcpy to rmsg which is not actually initialized at that point. The memcpy should be copying to variable m instead. Signed-off-by: Colin Ian King <colin.king@canonical.com> BCM2835-V4L2: Return buffers to videobuf2 on shutdown raspberrypi/linux#817 Fixes the kernel warning from videobuf2 as buffers are now returned as they are being flushed on stop_streaming. squash: Fixup bcm2835-camera for changes in kernel 4.4 api v4l2: Fix up driver to upstream timestamp changes bcm2835-camera: fix a bug in computation of frame timestamp Fixes #1318 V4L2 driver updates (#1393) * BCM2835-V4L2: Correct ISO control and add V4L2_CID_ISO_SENSITIVITY_AUTO raspberrypi/linux#1251 V4L2_CID_ISO_SENSITIVITY was not advertising ISO*1000 as it should. V4L2_CID_ISO_SENSITIVITY_AUTO was not implemented, so was taking V4L2_CID_ISO_SENSITIVITY as 0 for auto mode. Still accepts 0 for auto, but also abides by the new parameter. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add a video_nr parameter. Adds a kernel parameter "video_nr" to specify the preferred /dev/videoX device node. https://www.raspberrypi.org/forums/viewtopic.php?f=38&t=136120&p=905545 Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add support for multiple cameras Ask GPU on load how many cameras have been detected, and enumerate that number of devices. Only applicable on the Compute Module as no other device exposes multiple CSI2 interfaces. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Add control of the overlay location and alpha. Actually do something useful in vidioc_s_fmt_vid_overlay and vidioc_try_fmt_vid_overlay, rather than effectively having read-only fields. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: V4L2-Compliance failure fix VIDIOC_TRY_FMT was failing due to bytesperline not being set correctly by default. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> * BCM2835-V4L2: Make all module parameters static Clean up to correct variable scope Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> V4L2: Request maximum resolution from GPU Get resolution information about the sensors from the GPU and advertise it correctly. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-V4L2: Increase minimum resolution to 32x32 raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> [media]: bcm2835-camera: fix compilation error There is an error when compiling rpi-4.6.y branch: CC [M] drivers/media/platform/bcm2835/bcm2835-camera.o drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: error: initialization from incompatible pointer type [-Werror=incompatible-pointer-types] .queue_setup = queue_setup, ^ drivers/media/platform/bcm2835/bcm2835-camera.c:639:17: note: (near initialization for 'bm2835_mmal_video_qops.queue_setup') The const void *parg in setup_queue callback is not needed since commit: df9ecb0. This commit removes it. Signed-off-by: Slawomir Stepien <sst@poczta.fm> bcm2835-camera: Fix max/min error when looping over cameras/resolutions See: raspberrypi/linux#1447 (comment) BCM2835-V4L2: Correct handling for BGR24 vs RGB24. There was a bug in the GPU firmware that had reversed these two formats. Detect the old firmware, and reverse the formats if necessary. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com> BCM2835-v4l2: Fix a conformance test failure Format ioctls: test VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS: OK warn: v4l2-test-formats.cpp(1195): S_PARM is supported but doesn't report V4L2_CAP_TIMEPERFRAME. fail: v4l2-test-formats.cpp(1118): node->has_frmintervals && !cap->capability Gbp-Pq: Topic rpi Gbp-Pq: Name rpi_1049_f2d51efbeac118fa3c386f34cab3d7f2be7d5261.patch
hisenyiu2015
pushed a commit
to hisenyiu2015/msm-4.14
that referenced
this issue
May 20, 2021
raspberrypi/linux#1498 showed up that 16x16 is failing to work on the GPU for some reason. GPU bug being tracked on raspberrypi/firmware#607 Workaround here by increasing minimum resolution via V4L2 to 32x32. Signed-off-by: Dave Stevenson <6by9@users.noreply.github.com>
|
Hey together!
Can't tell about Pi OS based on bullseye. I am going to check the dmesg once I have enough time. Maybe in the meantime someone can tell if the issue appears too for him/her? |
|
Hi! Sorry if I got lost somewhere in this longish thread, but why is this issue closed? It seems it was closed as a duplicate of #1499 in #1498 (comment), but #1499 was also closed as a duplicate of this one. In any case, the issue still exists in Bullseye (Chromium or Electron doesn't detect the Raspberry Pi v1 camera module), so it'd be good to keep this open, linking to the currently open corresponding upstream bug: https://bugs.chromium.org/p/chromium/issues/detail?id=1406830. Thanks! |
|
Heres the additions the pi foundation added to Chromium in the past, afaik it's now used from the official debian Sources https://github.com/RPi-Distro/chromium-browser/files/7829777/camera.txt |
|
See RPi-Distro/chromium#23 for reference |
|
It's closed because that issue was solved - Chromium didn't support V4L2 camera drivers that advertised the resolutions as The issue you're probably hitting now is that Chromium doesn't support libcamera, and the default camera stack for Raspberry Pi OS is now that. I believe the preferred route for supporting libcamera within Chromium is going to be via Pipewire, but I could be wrong. |
|
I'm hitting the issue even using the legacy camera stack; which shouldn't happen if Chromium is now patched as above, if I understand correctly. This is using a Raspberry Pi 3 B and a camera module v1. I've detailed my findings and workarounds in https://forums.raspberrypi.com/viewtopic.php?t=351986. |
|
@andi34 it doesn't appear that this custom patch has made its way in either Chromium or the Debian package of it; see the patches used in: http://deb.debian.org/debian/pool/main/c/chromium/chromium_113.0.5672.126-1.debian.tar.xz, which is for the Debian chromium package in Bookworm. Looking at the chromium source http://deb.debian.org/debian/pool/main/c/chromium/chromium_113.0.5672.126.orig.tar.xz shows the patch hasn't been applied there either. Unless I've missed something, it means this bug is still relevant. Could this issue please be reopened? I've also requested the upstream bug https://bugs.chromium.org/p/chromium/issues/detail?id=249953#c25 to be reopened, as it's been archived without resolution while the source code still refers to it, calling |
|
The patch was / diff was shared by @jc-kynesim while I asked for a solution. And yes, it is not added to the official chromium sources. I was told the pi foundation used it's own sources and later shwitched to the mainline package. |
|
Besides that it's also still an issue on a 32bit PiOS to use the v4l2loopback module as it fails because of the mixed arch of the kernel and Userland.... |
|
Related, see an account of my struggle/hacky workarounds using v4l2loopback (I had to switch to 32 bits kernel/userland to avoid the above issue you linked) here: https://forums.raspberrypi.com/viewtopic.php?p=2109999#p2109999 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I'm struggling to make the Pi Camera module v2 work inside Chromium/Electron (while raspistill and python PiCamera do work) on an up-to-date Ubuntu Mate 16.04.
When I start electron v1.1.0 (Chrome 50.0.2661.102) on this url: https://simpl.info/getusermedia/sources
I get a black square with the following logs inside
/var/log/kern.log:After that, the camera module seems stuck and any further
raspistillwon't work.Any ideas?
Might be related to #849?
The text was updated successfully, but these errors were encountered: