My original code that mounts devices was careless and copied from the runc specification, but without enough exploration of context. It seems that I use the mkdir command in several places, where I should be mounting a device. This creates inconsistencies where core programs expect a special device and encounter a normal directory. Commented out mkdir commands for now and returning later to mount them.
It would also be good to go over the runc command again with a closer eye and compare it to the specification used in other programs like podman. The runc way may not be the only way.
My original code that mounts devices was careless and copied from the runc specification, but without enough exploration of context. It seems that I use the mkdir command in several places, where I should be mounting a device. This creates inconsistencies where core programs expect a special device and encounter a normal directory. Commented out mkdir commands for now and returning later to mount them.
It would also be good to go over the runc command again with a closer eye and compare it to the specification used in other programs like podman. The runc way may not be the only way.