The path is fairly long and by embedding it in the message, we get a longer
data section in our binary. By using %s, we can make the template strings
shorter, deduplicating bytes in our binary. Those are error messages, i.e. by
definition they are only used very rarely, so it's completely fine if printf
does a bit more work when generating the message.
$ strings build/systemd-random-seed.0 | rg '/var/lib/systemd/random-seed'
/var/lib/systemd/random-seed
Failed to open /var/lib/systemd/random-seed for writing: %m
Failed to open /var/lib/systemd/random-seed for reading: %m
Failed to open /var/lib/systemd/random-seed: %m
Failed to stat() seed file /var/lib/systemd/random-seed: %m
Failed to read seed from /var/lib/systemd/random-seed: %m
Seed file /var/lib/systemd/random-seed not yet initialized, proceeding.
$ strings build/systemd-random-seed | rg '/var/lib/systemd/random-seed'
/var/lib/systemd/random-seed
$ ls -l build/systemd-random-seed{,.0}
-rwxr-xr-x 1 zbyszek zbyszek 68872 May 31 13:39 build/systemd-random-seed.0
-rwxr-xr-x 1 zbyszek zbyszek 64912 May 31 13:45 build/systemd-random-seed
That's a ~6% smaller binary!
Let's ensure all of our source file names are unique without having
to take the directory into account.
This allows us to create meson targets or unit tests identified by the
the name of the source file they operate on without having to include
the full path of the source file in the target or test name to avoid
conflicts.
This commit cleans up the includes for all the small tools across
the tree.
A few cases of returning EXIT_SUCCESS are replaced with returning
0 to avoid including <stdlib.h>.
Split out of #37344.
Plus, linux/random.h never defined getrandom(), hence remove
the custom machinery for sys/random.h vs linux/random.h
in favor of single HAVE_GETRANDOM.
The ESP is simply not mounted early enough for this. We want that the
regular random seed handling runs as early as we possibly could, but we
don't want to delay this until the ESP is actually mounted.
Hence, let's remove this from random-seed.c here. A follow-up commit
will then add this back in, in a separate service which just calls
"bootctl random-seed".
Effectively reverts: f913c784adFixes: #25769
-1 was used everywhere, but -EBADF or -EBADFD started being used in various
places. Let's make things consistent in the new style.
Note that there are two candidates:
EBADF 9 Bad file descriptor
EBADFD 77 File descriptor in bad state
Since we're initializating the fd, we're just assigning a value that means
"no fd yet", so it's just a bad file descriptor, and the first errno fits
better. If instead we had a valid file descriptor that became invalid because
of some operation or state change, the other errno would fit better.
In some places, initialization is dropped if unnecessary.
Since this runs at shutdown to write a new seed, we should also keep the
bootloader's seed maximally fresh by doing the same. So we follow the
same pattern - hash some new random bytes with the old seed to make a
new seed. We let this fail without warning, because it's just an
opportunistic thing. If the user happens to have set up the random seed
with bootctl, and the RNG is initialized, then things should be fine. If
not, we create a new seed if systemd-boot is in use. And if not, then we
just don't do anything.
We want the size to lie within [/proc/sys/kernel/random/poolsize,RANDOM_POOL_SIZE_MAX]
interval. Let's make it more obvious.
Also move the logic in a dedicated function.
It's not really necessary since systemd-random-seed is an internal tool for the
moment but this might change in future (to allow system installers to
initialize a random seed file for example).
Also introducing new options will be easier.
If we're consuming an on-disk seed, we usually write out a new one after
consuming it. In that case, we might be at early boot and the randomness
could be rather poor, and the kernel doesn't guarantee that it'll use
the new randomness right away for us. In order to prevent the new
entropy from getting any worse, hash together the old seed and the new
seed, and replace the final bytes of the new seed with the hash output.
This way, entropy strictly increases and never regresses.
Fixes: https://github.com/systemd/systemd/issues/21983
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
This makes two major changes to the way systemd-random-seed operates:
1. We now optionally credit entropy if this is configured (via an env
var). Previously we never would do that, with this change we still don't
by default, but it's possible to enable this if people acknowledge that
they shouldn't replicate an image with a contained random seed to
multiple systems. Note that in this patch crediting entropy is a boolean
thing (unlike in previous attempts such as #1062), where only a relative
amount of bits was credited. The simpler scheme implemented here should
be OK though as the random seeds saved to disk are now written only with
data from the kernel's entropy pool retrieved after the pool is fully
initialized. Specifically:
2. This makes systemd-random-seed.service a synchronization point for
kernel entropy pool initialization. It was already used like this, for
example by systemd-cryptsetup-generator's /dev/urandom passphrase
handling, with this change it explicitly operates like that (at least
systems which provide getrandom(), where we can support this). This
means services that rely on an initialized random pool should now place
After=systemd-random-seed.service and everything should be fine. Note
that with this change sysinit.target (and thus early boot) is NOT
systematically delayed until the entropy pool is initialized, i.e.
regular services need to add explicit ordering deps on this service if
they require an initialized random pool.
Fixes: #4271
Replaces: #10621#4513
There's no reason why writing should work if reading and writing
doesn't. Let's simplify this hence. /dev/urandom is generally an r/w
device, and everything else would be a serious system misconfiguration.
We don't load the selinux label database anyway, hence this is not going
to do anything at all. Let's simplify this hence and drop the explicit
request.
Also note that today SELinux can take the filename into account when
automatically deducing the label for a new file, hence even if this code
actually would have done something it is redundant today.
Yu Watanabe
Zbigniew Jędrzejewski-Szmek
Daan De Meyer
Jelle van der Waa
Mike Yuan
Lennart Poettering
Frantisek Sumsal
Jason A. Donenfeld
Franck Bui
Romain Naour