On Fri, Dec 11, 2020 at 04:00:04PM +0000, Andre Przywara wrote:
> static inline bool __must_check arch_get_random_seed_long(unsigned long *v)
> {
> + struct arm_smccc_res res;
> +
> + /*
> + * We prefer the SMCCC call, since its semantics (return actual
> + * hardware backed entropy) is closer to the idea behind this
> + * function here than what even the RNDRSS register provides
> + * (the output of a pseudo RNG freshly seeded by a TRNG).
> + */
This logic...
> @@ -77,10 +117,20 @@ arch_get_random_seed_long_early(unsigned long *v)
> {
> WARN_ON(system_state != SYSTEM_BOOTING);
>
> - if (!__early_cpu_has_rndr())
> - return false;
> + if (__early_cpu_has_rndr())
> + return __arm64_rndr(v);
> +
> + if (smccc_trng_available) {
> + struct arm_smccc_res res;
>
> - return __arm64_rndr(v);
> + arm_smccc_1_1_invoke(ARM_SMCCC_TRNG_RND64, 64, &res);
> + if ((int)res.a0 >= 0) {
> + *v = res.a3;
> + return true;
> + }
> + }
> +
> + return false;
...seems to also apply here but we prefer the RNDR instead of the SMCC.
We probably want to either do the same thing or add a comment saying
what's going on.
Attachment:
signature.asc
Description: PGP signature
_______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
