Please confirm below REG by my previous comment.
15a10344: NVCSI_STREAM_0_ERR_INTR_MASK_NOVC_0
15a10270: NVCSI_STREAM_0_INTR_MASK_NOVC_0
15a18344: NVCSI_STREAM_1_ERR_INTR_MASK_NOVC_0
15a18270: NVCSI_STREAM_1_INTR_MASK_NOVC_0
15a20344: NVCSI_STREAM_2_ERR_INTR_MASK_NOVC_0
15a20270: NVCSI_STREAM_2_INTR_MASK_NOVC_0
15a28344: NVCSI_STREAM_3_ERR_INTR_MASK_NOVC_0
15a28270: NVCSI_STREAM_3_INTR_MASK_NOVC_0
15a30344: NVCSI_STREAM_4_ERR_INTR_MASK_NOVC_0
15a30270: NVCSI_STREAM_4_INTR_MASK_NOVC_0
15a38344: NVCSI_STREAM_5_ERR_INTR_MASK_NOVC_0
15a38270: NVCSI_STREAM_5_INTR_MASK_NOVC_0
Hi @ShaneCCC,
Accessing above registers gave same error : Bus error
nvidia@ubuntu:~$ sudo devmem2 15a10344 w
/dev/mem opened.
Memory mapped at address 0xffff9783a000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a10270 w
/dev/mem opened.
Memory mapped at address 0xffff8ea21000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a18344 w
/dev/mem opened.
Memory mapped at address 0xffffb2edb000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a20344 w
/dev/mem opened.
Memory mapped at address 0xffff896ff000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a20270 w
/dev/mem opened.
Memory mapped at address 0xffff86f6b000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a28344 w
/dev/mem opened.
Memory mapped at address 0xffff92ae3000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a28270 w
/dev/mem opened.
Memory mapped at address 0xffff92fc6000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a30344 w
/dev/mem opened.
Memory mapped at address 0xffffaf7df000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a30270 w
/dev/mem opened.
Memory mapped at address 0xffff89860000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a38344 w
/dev/mem opened.
Memory mapped at address 0xffff8ca73000.
Bus error
nvidia@ubuntu:~$ sudo devmem2 15a38270 w
/dev/mem opened.
Memory mapped at address 0xffff93ee8000.
Bus error
Please note that I am working with Nvidia Jetson Xavier NX.
Do you run the camera before access the REG?
Also you need to add “0x” before the address like 0x15a10344
Hi @ShaneCCC,
Yes I did open the camera with v4l2-ctl command (as you can see in the terminal on the right),
With prefix 0x, some addresses (ended with xy344) reported Bus error, others (ended with xy270) reported 0x0
Try access by b like
sudo devmem2 0x15a10344 b
Hi @ShaneCCC,
REGs ended with 344 are readable but gave 0x0 :
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a10344 b
/dev/mem opened.
Memory mapped at address 0xffffb1d9f000.
Value at address 0x15A10344 (0xffffb1d9f344): 0x0
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a18344 b
/dev/mem opened.
Memory mapped at address 0xffffb9c97000.
Value at address 0x15A18344 (0xffffb9c97344): 0x0
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a20344 b
/dev/mem opened.
Memory mapped at address 0xffffb206f000.
Value at address 0x15A20344 (0xffffb206f344): 0x0
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a28344 b
/dev/mem opened.
Memory mapped at address 0xffffaf98c000.
Value at address 0x15A28344 (0xffffaf98c344): 0x0
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a30344 b
/dev/mem opened.
Memory mapped at address 0xffff8fe35000.
Value at address 0x15A30344 (0xffff8fe35344): 0x0
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a38344 b
/dev/mem opened.
Memory mapped at address 0xffff9d3b4000.
Value at address 0x15A38344 (0xffff9d3b4344): 0x0
Looks like different with default configuration 0x3
1 Like
Hi @ShaneCCC,
I am sorry, I was testing the access to the REGs in the setup that I haven’t disabled the CRC yet.
-
I will re-do the test on the board with which I used the RTCPU firmware that you provided and the patch for the CSI driver as in the initial comment and report to you later.
-
However, I also need to ensure that accessing the NVCSI REGs (the NVCSI_STREAM_0_INTR_STATUS_VC0_0) is possible in order to report to the FPGA designer about the status of the MIPI-CSI Rx so that he could review/tune his design in other discussion. Unfortunately, the reading to NVCSI_STREAM_0_INTR_STATUS_VC0_0 (0x15a10228 b/w) and others always gives 0x0.
Hi @ShaneCCC,
To update for the 2) in previous comment, I could read the NVCSI_STREAM_0_PF_CRC_0 (0x15a101d8) as below :
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a101d8 b
/dev/mem opened.
Memory mapped at address 0xffff89ddd000.
Value at address 0x15A101D8 (0xffff89ddd1d8): 0xE9
nvidia@ubuntu:~/Workspace$ sudo devmem2 0x15a101d8 w
/dev/mem opened.
Memory mapped at address 0xffffbbb92000.
Value at address 0x15A101D8 (0xffffbbb921d8): 0x61006100
At least, I could see that the calculated CRC and the received CRC are matched each other.
Could you give the trace log for enable and disable CRC by the sensor to check.
1 Like
