How Xavier capture image?

Hi All,

I’m curious that how xavier capture image.
1.) I can capture by v4l2 command, but I don’t know how it work?
2.) If I capture ov5693 image(2592x1944), xavier how to know start to capture and capture whole image( didn’t loss any pixel)?

Can someone show the work flow?

Thanks.

hello syan,

may I know what’s the actual purpose for your question?

if you using v4l command, it’ll use the VI driver for request capture buffer, and process the frames sending to vb2.
please check VI-5 driver for reference,
for example,
$public_sources/kernel_src/kernel/nvidia/drivers/media/platform/tegra/camera/vi/vi5_fops.c

Hi Jerry,
Thanks your reference.

I want to know where did xavier do packet analysis.
how to sure packet didn’t loss?

hello syan,

from driver side, it’s waitqueue for starting frame capture when buffer is available.
it’ll report timeout if VI engine cannot wait SOF.

besides, it’ll check the capture status before sending the capture buffer to user-space. if there’s an error. it’ll dropping the frame and reporting corr_err (i.e. correctable error) with discarding frame messages.

from the low-level side, it’s RCE (Real-time CameraEngine) camera firmware to handle MIPI signaling. you may see-also Xavier TRM. please refer to [Figure 2.1 Xavier Processor Block Diagram].

so,
you should expect packet didn’t loss if you not seeing failure reported.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.