I am trying to create an image for my Jetson Nano by building a container from the docker file found at Docker build for L4T Jetpack image. Initially I ran the docker build command myself, but then I realized the makefile is there to do that. Either way the build fails and here is the output
And so you can see that I do in fact have the makefile in the same folder
My questions are:
Will the docker container/image created from this build allow me to write my code using opencv, cuda, etc.?
I chose this template docker file because it seems like it should. I’m planning to develop in C++ and I just wanted an empty container with all of the necessary dependencies so I can just start coding, then build the container and have it just run on my nano. I have gotten a container to work before, but it was a basic “Hello World” container that wouldn’t support opencv and cuda.
So yes I am building these on an x86 machine, but I do so in an emulator environment using qemu, so it would be as if I tried it on the Nano.
So I realized that one reason that I got the above error was because I had forgotten to start the emulator. I then corrected that mistake and tried building the image again and got a new error:
“E: Unable to locate package nvidia-cuda-dev”
Then I did try building it on the nano and received the same error as above:
“E: Unable to locate package nvidia-cuda-dev”
I will note that I changed the tag in the makefile to be r32.7.1 instead of the default r35.1.0, since that’s the version of, I wanna say jetpack or L4T, that is on my Nano.
I’m going to take a look at the page you sent me as well and see if I can get one of those to work.
P.S. Sorry about the late reply. I try to be quicker than this.