On Wed, Nov 16, 2022 at 10:44:20PM +0100, Ard Biesheuvel wrote: > On Wed, 16 Nov 2022 at 21:06, Jason A. Donenfeld <Jason@xxxxxxxxx> wrote: > > > > Taking the minimum is wrong, if the bootloader or EFI stub is actually > > passing on a bunch of bytes that it expects the kernel to hash itself. > > We still need some kind of limit, just so things don't explode if > seed->size is bogus. Alright, let's just say 1k then. Will send a v2. Jason
