Shoot for 256 bits minimum entropy in the seed pool, gives us
some safety margin over our required 128.
note that the pool only gets more entropy over time, as every
call to the pool folds more entropy in (roughly 2 bits per call).
compression helps us reduce the amount of data we have to
shuttle between the hardware interface and the rest of the world.
It takes more samples from the TRNG and represents them in a compressed
form that effectively increases the bits per byte of entropy. It
does not increase the *rate* of entropy generation -- while each
byte has more entropy, the bytes come proportionally slower. However,
this is a reasonable trade-off because kernel space calls to fetch
the entropy feed are expensive, and what this does is effectively
amortize the price of a syscall: the time per syscall is longer,
but you get more entropy out per syscall, thus effectively reducing
the needed overhead to get a certain amount of entropy.
This SysCall is introduced because not all hardware targets can
afford the luxury of a private TRNG port for the kernel.
In this case, an API is needed to exfiltrate random numbers from
the kernel so we can inspect them and check them for health.
If a TRNG is operating correctly, cherry-picking some values
out of it to QC the operation of the TRNG should be harmless.
Of course, nothing is truly stateless, and there is always a chance
that someone could construct a gadget using this to reduce the
potential state-space of the final random number generator. Therefore,
this API is (a) gated behind a feature flag and (b) any implementation
that plans to expose this should be coded with some countermeasures
to reduce data leakage after sampling.