Compile flow on docker

[kthoru01]

Hello,

I'm wondering what is the general compilation flow if the vp was acquired from docker? I built the linux 4.13.3 kernel, after that do I just need to modify the kmd and umd? Do I need to copy the Image and rootft.ext4 from buildroot into /usr/local/nvdla and then run the command to mount the vp? When I do this, and run insmod drm.ko it says that drm.ko does not exist. I would greatly appreciate if someone could post the general execution flow of running nvdla. I'm working on a project for school that is heavily dependent on getting nvdla up and running. I'm trying to replicate the work in this paper "Domain-Specific Accelerator Design & Profiling for Deep Learning Applications" and build on it. Any help would be greatly appreciated!