If unprivileged userns are *believed* to be disabled (such as with
"kernel.unprivileged_userns_clone = 0"), Lix would previously *give up*
on trying to use a user namespace before actually trying it, even if, in
cases such as unprivileged_userns_clone, it would actually be allowed
since Nix has CAP_SYS_ADMIN when running as daemon.
(see, e.g. https://github.com/archlinux/linux/commit/25d4709a4fc0e4f3b432c24b60dd508fb84f0cb2)
We changed it to actually try it first, and then diagnose possible
causes, and also to be more loud about the whole thing, using warnings
instead of debugs. These warnings will only print on the first build run
by the daemon, which is, tbh, eh, shrug.
This is what led to us realizing that no-userns was a poorly exercised
condition.
Change-Id: I8e4f21afc89c574020dc7e89a560cc740ce6573a
In unprivileged podman containers, /proc is not fully visible (there
are other filesystems mounted on subdirectories of /proc). Therefore
we can't mount a new /proc in the sandbox that matches the PID
namespace of the sandbox. So this commit automatically disables
sandboxing if /proc is not fully visible.
This didn't work because sandboxing doesn't work in Docker. However,
the sandboxing check is done lazily - after clone(CLONE_NEWNS) fails,
we retry with sandboxing disabled. But at that point, we've already
done UID allocation under the assumption that user namespaces are
enabled.
So let's get rid of the "goto fallback" logic and just detect early
whether user / mount namespaces are enabled.
This commit also gets rid of a compatibility hack for some ancient
Linux kernels (<2.13).
Jade Lovelace