On SELinux-enabled systems, /etc/hosts has a different type `net_conf_t`
than the other files in /etc, so the temporary file that overwrites it
ends up with the wrong context, resulting in many system services
becoming unable to access the file. To fix this, manually look up the
context /etc/hosts has and copy it to the temporary file before
the rename.
In order to avoid depending on libselinux on systems that don't use it,
this support is gated behind the new "selinux" feature. It *is*
installed and enabled in the Dockerfile, however, in order to ensure
that it still builds.
The recommended netlink buffer size is based on the system's page size,
which means that the current size is far too small for systems with 16k
or 64k pages, such as Asahi Linux or RHEL's kernel-64k for ARM64. On
these systems, the server fails to start with errors like this:
Error: Decode error occurred: invalid netlink buffer: length field says 1444 the buffer is 1260 bytes long
Instead, follow the kernel's own netlink docs to compute the buffer
size. The approach here matches the approach merged into Chromium
recently:
https://chromium-review.googlesource.com/c/chromium/src/+/4312885
Should fix#242.
We depend on `getrandom` feature of `rand_core`, but we weren't specifying that.
It worked because `curve25519-dalek 4.0.0-pre.2` was pulling the `getrandom` feature of `rand_core`, but after updating it to `4.0.0-pre.5` it stopped pulling `rand_core` altogether.
After the cargo-update, all out dependencies are on their newest version, sans clap 3.2, which could be updated to 4.0. It has quite some breaking API changes, so I'll leave that for a separate PR.
Ryan Gonzalez