Currently, the specification requires runtimes to produce a (fatal) error if a
container configuration requests capabilities that cannot be granted (either
the capability is "unknown" to the runtime, not supported by the kernel version
in use, or not available in the environment that the runtime operates in).
This causes problems in situations where the runtime is running in a restricted
environment (for example, docker-in-docker), or if there is a mismatch between
the list of capabilities known by higher-level runtimes and the OCI runtime.
Some examples:
- Kernel 5.8 introduced CAP_PERFMON, CAP_BPF, and CAP_CHECKPOINT_RESTORE
capabilities. Docker 20.10.0 ("higher level runtime") shipped with
an updated list of capabilities, and when creating a "privileged" container,
would determine what capabilities are known by the kernel in use, and request
all those capabilities (by including them in the container config).
However, runc did not yet have an updated list of capabilities, and therefore
reject the container specification, producing an error because the new
capabilities were "unknown".
- When running nested containers, for example, when running docker-in-docker,
the "inner" container may be using a more recent version of docker than the
"outer" container. In this situation, the "outer" container may be missing
capabilities that the inner container expects to be supported (based on
kernel version). However, starting the container would fail, because the OCI
runtime could not grant those capabilities (them not being available in the
environment it's running in).
Workarounds, and motivation
-------------------------------------
In the current situation, responsibility of detection what capabilities are
supported is left to the "higher level" runtimes. As an example, containerd
recently added code to dynamically adjust the list of requested capabilities
by attempting to detect which capabilities are available in the environment
it's running. This is only a partial solution, as it will not address
mismatches between the list of capabilities _known_ by the higher-level and
lower-level runtime (which cannot be detected).
Not only does this workaround only provide a *partial* fix, it also introduces
additional complexity in every higher-level runtime.
Proposal: WARN (but otherwise ignore) capabilities that cannot be granted
-------------------------------------
This patch changes the specification to have runtimes WARN (but otherwise
ignore) capabilities that are requested in the container config, but cannot
be granted.
Moving this responsibility to the lower-level (OCI) runtime makes more sense,
as the OCI runtime _already_ is responsible for interacting with the kernel
(detecting what capabilities are supported, and performing conversion), and
only the lower-level runtime itself knows what capabilities it supports itself.
Making the lower-level runtime responsible for handling "unknown" or "unavailable"
capabilities keeps the logic central.
Impact on security
-------------------------------------
Given that `capabilities` is an "allow-list", ignoring unknown capabilities will
not impose a security risk; worst case, a container does not get all requested
capabilities granted and as a result, some actions may fail.
Backward-compatibility
-------------------------------------
Changing this behavior should be backward compatible. Higher-level runtimes that
already dynamically adjust the list of requested capabilities can continue to do
so. Runtimes that do not adjust will see an improvement (containers can start
even if some of the requested capabilities are not granted). Container processes
MAY fail (as described in "impact on security"), but users can debug this
situation either by looking at the warnings produces by the OCI runtime, or using
tools such as `capsh` / `libcap` to get the list of actual capabilities in the
container.
Signed-off-by: Sebastiaan van Stijn <[email protected]>