Re: [RFC PATCH 4.10 1/6] crypto/sha256: Refactor the API so it can be used without shash

From: Andy Lutomirski
Date: Tue Dec 27 2016 - 14:01:05 EST

On Tue, Dec 27, 2016 at 6:16 AM, Daniel Borkmann <daniel@xxxxxxxxxxxxx> wrote:
> On 12/27/2016 10:58 AM, Herbert Xu wrote:
>> On Mon, Dec 26, 2016 at 10:08:48AM -0800, Andy Lutomirski wrote:
>>> According to Daniel, the networking folks want to let embedded systems
>>> include BPF without requiring the crypto core.
>> Last I checked the IPv4 stack depended on the crypto API so this
>> sounds bogus.
> I think there's a bit of a mixup here with what I said. To clarify,
> requirement back then from tracing folks was that bpf engine and
> therefore bpf syscall can be build w/o networking enabled for small
> devices, so dependencies preferably need to be kept on a absolute
> minimum, same counts for either making it suddenly a depend on
> CRYPTO or a select CRYPTO for just those few lines that can be
> pulled in from lib/ code instead.

Somehow I had that in my head as "networking" not "tracing", probably
because of the TCA stuff. Whoops.

Anyway, I'm rewriting the crypto part of the patch completely based on
Ard's feedback.