MIME-Version: 1.0
In-Reply-To: <d7e206a5a03d46a69c0be3b8ed651518@AcuMS.aculab.com>
References: <CAHmME9qbnYmhvsuarButi6s=58=FPiti0Z-QnGMJ=OsMzy1eOg@mail.gmail.com>
 <20220111134934.324663-1-Jason@zx2c4.com> <20220111134934.324663-2-Jason@zx2c4.com>
 <Yd8enQTocuCSQVkT@gmail.com> <CAHmME9qGs8yfYy0GVcV8XaUt9cjCqQF2D79RvrsQE+CNLCeojA@mail.gmail.com>
 <d7e206a5a03d46a69c0be3b8ed651518@AcuMS.aculab.com>
From:   "Jason A. Donenfeld" <Jason@zx2c4.com>
Date:   Wed, 12 Jan 2022 23:00:48 +0100
Message-ID: <CAHmME9p-JZS480-e3MvPsa1Q07rqX79h8oNEZsF970GjT-_nqA@mail.gmail.com>
Subject: Re: [PATCH crypto 1/2] lib/crypto: blake2s-generic: reduce code size
 on small systems
To:     David Laight <David.Laight@aculab.com>
Cc:     Eric Biggers <ebiggers@kernel.org>,
        Linux Crypto Mailing List <linux-crypto@vger.kernel.org>,
        Netdev <netdev@vger.kernel.org>,
        WireGuard mailing list <wireguard@lists.zx2c4.com>,
        LKML <linux-kernel@vger.kernel.org>, bpf <bpf@vger.kernel.org>,
        Geert Uytterhoeven <geert@linux-m68k.org>,
        "Theodore Ts'o" <tytso@mit.edu>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Jean-Philippe Aumasson <jeanphilippe.aumasson@gmail.com>,
        Ard Biesheuvel <ardb@kernel.org>,
        Herbert Xu <herbert@gondor.apana.org.au>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi David,

On 1/12/22, David Laight <David.Laight@aculab.com> wrote:
> I think you mentioned in another thread that the buffers (eg for IPv6
> addresses) are actually often quite short.
>
> For short buffers the 'rolled-up' loop may be of similar performance
> to the unrolled one because of the time taken to read all the instructions
> into the I-cache and decode them.
> If the loop ends up small enough it will fit into the 'decoded loop
> buffer' of modern Intel x86 cpu and won't even need decoding on
> each iteration.
>
> I really suspect that the heavily unrolled loop is only really fast
> for big buffers and/or when it is already in the I-cache.
> In real life I wonder how often that actually happens?
> Especially for the uses the kernel is making of the code.
>
> You need to benchmark single executions of the function
> (doable on x86 with the performance monitor cycle counter)
> to get typical/best clocks/byte figures rather than a
> big average for repeated operation on a long buffer.
>
> 	David

This patch has been dropped entirely from future revisions. The latest
as of writing is at:

https://lore.kernel.org/linux-crypto/20220111220506.742067-1-Jason@zx2c4.com/

If you'd like to do something with blake2s, by all means submit a
patch and include various rationale and metrics and benchmarks. I do
not intend to do that myself and do not think my particular patch here
should be merged. But if you'd like to do something, feel free to CC
me for a review. However, as mentioned, I don't think much needs to be
done here.

Again, v3 is here:
https://lore.kernel.org/linux-crypto/20220111220506.742067-1-Jason@zx2c4.com/

Thanks,
Jason