Re: [PATCH v2 1/2] sev/i386: Introduce sev_add_kernel_loader_hashes for measured linux boot

[Connor Kuehl <ckuehl@redhat.com>]

On 6/21/21 2:05 PM, Dov Murik wrote:
> +static void fill_sev_hash_table_entry(SevHashTableEntry *e, const uint8_t *guid,
> +                                      const uint8_t *hash, size_t hash_len)
> +{
> +    memcpy(e->guid, guid, sizeof(e->guid));
> +    e->len = sizeof(*e);
> +    memcpy(e->hash, hash, hash_len);

Should this memcpy be constrained to MIN(sizeof(e->hash), hash_len)? Or
perhaps an assert statement since I see below that this function's
caller sets this to HASH_SIZE which is currently == sizeof(e->hash).

Actually, the assert statement would be easier to debug if the input
to this function is ever unexpected, especially since it avoids an
outcome where the input is silently truncated; which is a pitfall that
that the memcpy clamping would fall into.

Connor