V4L2 sensor shows low FPS with CHANSEL_NOMATCH

I have 3840x2160, 14 bit sensor and I can see the sensor output image from gstreamer. In this case, 40 fps sensor setting has no problem.

I want to receive ‘3840x4320, 14bit, 40 fps image’ but I the FPS shows about 23.
3840x4320x14x40 = 9.3G. It is within CSI brick bandwidth 10G.

I tried clock speed up but the results is not changed. I checked vi trace and I found that CHANSEL_NOMATCH occured quite often.

 vi-output, sami-288454  [000] .... 18576.809184: tegra_channel_capture_frame: sof:18599.958663264
 vi-output, sami-288454  [000] .... 18576.809185: tegra_channel_capture_frame: eof:18599.984958848
     kworker/0:1-285061  [000] .... 18576.864733: rtcpu_vinotify_error: tstamp:581249589571 cch:0 vi:0 tag:CHANSEL_NOMATCH channel:0x01 frame:4 vi_tstamp:18599986795648 data:0x00000000000005a9
     kworker/0:1-285061  [000] .... 18576.864737: rtcpu_vinotify_event: tstamp:581249909584 cch:0 vi:0 tag:CHANSEL_PXL_EOF channel:0x23 frame:3 vi_tstamp:18599984958144 data:0x00000000087f0002
     kworker/0:1-285061  [000] .... 18576.864737: rtcpu_vinotify_event: tstamp:581249909736 cch:0 vi:0 tag:FE channel:0x00 frame:3 vi_tstamp:18599984958624 data:0x0000000300000020
     kworker/0:1-285061  [000] .... 18576.864737: rtcpu_vinotify_event: tstamp:581249909872 cch:0 vi:0 tag:ATOMP_FE channel:0x00 frame:3 vi_tstamp:18599984958656 data:0x0000000800000000
     kworker/0:1-285061  [000] .... 18576.864737: rtcpu_vinotify_event: tstamp:581249910025 cch:0 vi:0 tag:ATOMP_FRAME_DONE channel:0x23 frame:3 vi_tstamp:18599984958848 data:0x0000000000000000
     kworker/0:1-285061  [000] .... 18576.864738: rtcpu_vinotify_event: tstamp:581249910153 cch:0 vi:0 tag:VIFALC_ACTIONLST channel:0x23 frame:3 vi_tstamp:18599984984800 data:0x0000000002021994
     kworker/0:1-285061  [000] .... 18576.864738: rtcpu_vinotify_event: tstamp:581249910302 cch:0 vi:0 tag:VIFALC_ACTIONLST channel:0x23 frame:3 vi_tstamp:18599984993024 data:0x0000000000021994
     kworker/0:1-285061  [000] .... 18576.864738: rtcpu_vinotify_event: tstamp:581249910432 cch:0 vi:0 tag:FS channel:0x00 frame:4 vi_tstamp:18599986215712 data:0x0000000400000010
     kworker/0:1-285061  [000] .... 18576.864738: rtcpu_vinotify_event: tstamp:581249910578 cch:0 vi:0 tag:CHANSEL_NOMATCH channel:0x01 frame:4 vi_tstamp:18599986795648 data:0x00000000000005a9
     kworker/0:1-285061  [000] .... 18576.864739: rtcpu_vinotify_event: tstamp:581250316812 cch:0 vi:0 tag:VIFALC_TDSTATE channel:0x23 frame:0 vi_tstamp:18600002755136 data:0x379d580010000000
     kworker/0:1-285061  [000] .... 18576.864739: rtcpu_vinotify_event: tstamp:581250316965 cch:0 vi:0 tag:VIFALC_TDSTATE channel:0x23 frame:0 vi_tstamp:18600002776512 data:0x0000000031001995
     kworker/0:1-285061  [000] .... 18576.864740: rtcpu_vinotify_event: tstamp:581250450285 cch:0 vi:0 tag:FE channel:0x00 frame:4 vi_tstamp:18600013090976 data:0x0000000400000020
     kworker/0:1-285061  [000] .... 18576.864740: rtcpu_vinotify_event: tstamp:581250450433 cch:0 vi:0 tag:FS channel:0x00 frame:1 vi_tstamp:18600014348096 data:0x0000000100000010
     kworker/0:1-285061  [000] .... 18576.864740: rtcpu_vinotify_event: tstamp:581250450575 cch:0 vi:0 tag:ATOMP_FS channel:0x00 frame:1 vi_tstamp:18600014348128 data:0x0000000800000000
     kworker/0:1-285061  [000] .... 18576.864741: rtcpu_vinotify_event: tstamp:581250845992 cch:0 vi:0 tag:CHANSEL_PXL_SOF channel:0x23 frame:1 vi_tstamp:18600014928000 data:0x0000000000000001
     kworker/0:1-285061  [000] .... 18576.864741: rtcpu_vinotify_event: tstamp:581250846122 cch:0 vi:0 tag:VIFALC_ACTIONLST channel:0x23 frame:1 vi_tstamp:18600014933472 data:0x0000000008021995
 vi-output, sami-288454  [002] .... 18576.865438: tegra_channel_capture_frame: sof:18600.14928000
 vi-output, sami-288454  [002] .... 18576.865439: tegra_channel_capture_frame: eof:18600.41223616

Could you explain why this CHANSEL_NOMATCH happened?
I think low fps is caused by these CHANSEL_NOMATCHs.

‘v4l2-ctl --all’ command shows Pixel Format is ‘GR14’ (14-bit Bayer GRGR/BGBG). I don’t know why CHANSEL_NOMATCH occured.

Could be incorrect embedded_metadata_height

https://elinux.org/Jetson/l4t/Camera_BringUp

Thanks for your reply.
I checked sensor setting but there is no embedded data lines.

Did you check 3840x2160 have the same message?

In case of 3840x2160, there is no error and FPS is not dropped.

Could you try lower frame rate like?

3840x4320x14x20

We checked lower fps setting and found that CHANSEL_NOMATCH does not occured in this case. FPS is not dropped.

Do you have any idea?

Boost the clocks to try.
If still please narrow down which fps cause the problem.
It could be the bandwidth unable to support it.

sudo su
echo 1 > /sys/kernel/debug/bpmp/debug/clk/vi/mrq_rate_locked
echo 1 > /sys/kernel/debug/bpmp/debug/clk/isp/mrq_rate_locked
echo 1 > /sys/kernel/debug/bpmp/debug/clk/nvcsi/mrq_rate_locked
echo 1 > /sys/kernel/debug/bpmp/debug/clk/emc/mrq_rate_locked
cat /sys/kernel/debug/bpmp/debug/clk/vi/max_rate |tee /sys/kernel/debug/bpmp/debug/clk/vi/rate
cat /sys/kernel/debug/bpmp/debug/clk/isp/max_rate | tee  /sys/kernel/debug/bpmp/debug/clk/isp/rate
cat /sys/kernel/debug/bpmp/debug/clk/nvcsi/max_rate | tee /sys/kernel/debug/bpmp/debug/clk/nvcsi/rate
cat /sys/kernel/debug/bpmp/debug/clk/emc/max_rate | tee /sys/kernel/debug/bpmp/debug/clk/emc/rate

I’m using LI-JXAV-MIPI-ADPT-4CAM - Leopard Imaging Inc. board. One camera is conntected to one port of this board.

Does this environment make bandwidth limit?

It’s doesn’t matter without the board. I mean the sensor output data like huge data through put “WidthHeightFPS” but VI clocks configure unable to handle it.