Thanks for the update dusty. I have a few questions though. I cannot find the “NVIDIA Tegra Linux Driver Package Development Guide” in the documentation tar ball provided with this release. Link is [url]http://developer.nvidia.com/embedded/dlc/l4t-documentation-23-2[/url]. It was present in the documentation for L4T_R23.1
I have a TX1 board running L4T_R23.1 and I wish to update it to the latest one without flashing all over again. There was a section for updating the L4T version in the Jetson TK1’s development guide. I would like to do the same with this board as well. Could you please mention the steps to be done or atleast point me to the updated development guide containing this procedure.
The directory structure of the Tegra_Linux_Driver_Package_Documents linked to above was modified a little bit since last release, but still contains the info from before with 23.1. The PDF version of the driver documents which used to be located in the baggage/ subdirectory was removed in lieu of the HTML documentation for synchronization. The same info is included in the HTML as the previous 23.1, with the addition of the V4L2 section.
After untarring the document package, open Start_L4T_Docs.html or index.html to bring up the L4T driver docs.
I recall that back in the early days of TK1, there may have been option to upgrade one of the L4T releases in-place (R21.1 to R21.2 I think), however in general (due to bootloader & kernel updates/patches/ect.) this approach is not recommended for stability. To my knowledge it hasn’t been included in the docs since early TK1 release and no TX1 releases, sorry. If you wish, you can attempt using the old TK1 procedure and report the results. For backup, you can clone your old setup following these steps: https://devtalk.nvidia.com/default/topic/898999/jetson-tx1/tx1-r23-1-new-flash-structure-how-to-clone-/post/4784149/#4784149
Thanks dusty. I missed the HTML documentation. Let me check it out.
I didnt upgrade the entire board to L4T_R23.2. I just built the kernel, modules, dtbs and just replaced them with the ones already present on the board. All the other libraries are from L4T_R23.1. The board still works fine for now. The boot arguments that were present in the new release were the same as before. So I didnt bother updating u-boot.
Two questions… How do you upgrade the JetPack…? I have JetPack 2.0 installed and want to upgrade to JetPack 2.1… Remove 2.0 then install 2.1, or just install 2.1 on top of 2.0, or is there another way…
Next, how do you update the Jetson TX1 to L4T R23.1 to R23.2… Complete reflash…?
Thanks,
-Michael
Jetpack is self-contained within the directory you ran Jetpack from on your host x86_64 machine. So all you have to do is save/run Jetpack 2.1 from a different directory than you saved/ran Jetpack 2.0 from (or rename/delete Jetpack 2.0 directory).
Cool, Thank you for your promptly response, so when I check the release document on R23.2, it seems NVidia updated a lot drives related to V4L2 interface, include the hdmi to csi- 2 bridge chip Tc358840xbg, so my question is that if NVidia validate the hardware design already, if so, is that possible to access those design information, or it just use for internal testing? Thank you.
Hi,
I get the following error when trying to install L4T R23.2:
[b]Installing CUDA Toolkit for Ubuntu 14.04 7.0.73 failed.
Return Code: 1
Error installing cuda to host, please check /home/mkanyix/Documents/_installer/logs/TX1/cuda_host_tx1.log for more details[/b]
I looked at the cuda_host_tx1.log and it shows the following errors: Err http://extras.ubuntu.com trusty/main armhf Packages
404 Not Found
Err http://security.ubuntu.com trusty-security/main armhf Packages
404 Not Foun
My host is Fedora, so I can’t verify anything with JetPack, but it sounds like the setup is installing to your desktop host, but trying to install the armhf which belongs on the Jetson. Was there further information in the logs which could identify where the CUDA was attempting to be installed to…host versus Jetson? If for some reason JetPack was looking for armhf on x86_64 servers, this would account for not finding the package.
@linuxdev.Correct, the problem is that the setup is trying to download the armhf packages on my ubuntu desktop host first and that is why it fails.
I captured the installer command that runs on the desktop host.
I’ve tried three times to do the upgrade without success. I’ve a 14.04 host who already did the upgrade with 2.0 without problem.
The installation script hang after copying my SSH key :
[ 129.3366 ] tegradevflash --write BCT P2180_A00_LP4_DSC_204Mhz.bct
[ 129.3393 ] Cboot version 00.01.0000
[ 129.5381 ] Writing partition BCT with P2180_A00_LP4_DSC_204Mhz.bct
[ 129.5391 ] [................................................] 100%
[ 129.6463 ]
[ 129.6463 ] Flashing completed
[ 129.6464 ] Coldbooting the device
[ 129.6491 ] tegradevflash --reboot coldboot
[ 129.6523 ] Cboot version 00.01.0000
[ 129.8511 ]
*** The target t210ref has been flashed successfully. ***
Reset the board to boot from internal eMMC.
1
Finished Flashing OS
Determining the IP address of target...
192.169.1.10
Waiting 30 seconds to make sure target is fully up
# Host 192.169.1.10 found: line 9 type RSA
# Host 192.169.1.10 found: line 10 type ECDSA
/home/ben/.ssh/known_hosts updated.
Original contents retained as /home/ben/.ssh/known_hosts.old
# 192.169.1.10 SSH-2.0-OpenSSH_6.6.1p1 Ubuntu-2ubuntu2
# 192.169.1.10 SSH-2.0-OpenSSH_6.6.1p1 Ubuntu-2ubuntu2
/home/ben
Identity added: /home/ben/.ssh/id_rsa (/home/ben/.ssh/id_rsa)
scp -o PubkeyAuthentication=no -o ConnectTimeout=30 -o StrictHostKeyChecking=no /home/ben/.ssh/id_rsa.pub ubuntu@192.169.1.10:/home/ubuntu/tmp.pub
.1.1
id_rsa.pub 100% 393 0.4KB/s 00:00
ubuntu@tegra-ubuntu:~$ b /home/ubuntu/.ssh/authorized_keys
ubuntu@tegra-ubuntu:~$ exit
logout
Connection to 192.169.1.10 closed.
The TX1 boot and I can login but there is many problems :
No sudo (error : sudo bit id not set for user ubuntu)
No Wifi
User management app don’t open
And so on…
So I beleive it has not finish install.
When I try to continue without flashing it stop after copying two files.
What can I do ?
Thanks for your help.
I had the same issue because I was flashing from a non root permitive partition from my host.
Try to copy all in your home location and flash again and let me know.
Thank you Ale,
You’re right.
I was running short on disk space so I’ve run the installation on an external NTFS drive. As the rootfs is build locally it can’t work.
Now my TX1 is working fine.