[1/2] crypto: arm/curve25519 - add arch-specific key generation function

Somehow this was forgotten when Zinc was being split into oddly shaped
pieces, resulting in linker errors. The x86_64 glue has a specific key
generation implementation, but the Arm one does not. However, it can
still receive the NEON speedups by calling the ordinary DH function
using the base point.

Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
---
 arch/arm/crypto/curve25519-glue.c | 7 +++++++
 1 file changed, 7 insertions(+)

On Wed, 11 Dec 2019 at 10:27, Jason A. Donenfeld <Jason@zx2c4.com> wrote:
>
> Somehow this was forgotten when Zinc was being split into oddly shaped
> pieces, resulting in linker errors.

Zinc has no historical significance here, so it doesn't make sense to
keep referring to it in the commit logs.

> The x86_64 glue has a specific key
> generation implementation, but the Arm one does not. However, it can
> still receive the NEON speedups by calling the ordinary DH function
> using the base point.
>
> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>

With the first sentence dropped,

Acked-by: Ard Biesheuvel <ardb@kernel.org>

> ---
>  arch/arm/crypto/curve25519-glue.c | 7 +++++++
>  1 file changed, 7 insertions(+)
>
> diff --git a/arch/arm/crypto/curve25519-glue.c b/arch/arm/crypto/curve25519-glue.c
> index f3f42cf3b893..776ae07e0469 100644
> --- a/arch/arm/crypto/curve25519-glue.c
> +++ b/arch/arm/crypto/curve25519-glue.c
> @@ -38,6 +38,13 @@ void curve25519_arch(u8 out[CURVE25519_KEY_SIZE],
>  }
>  EXPORT_SYMBOL(curve25519_arch);
>
> +void curve25519_base_arch(u8 pub[CURVE25519_KEY_SIZE],
> +                         const u8 secret[CURVE25519_KEY_SIZE])
> +{
> +       return curve25519_arch(pub, secret, curve25519_base_point);
> +}
> +EXPORT_SYMBOL(curve25519_base_arch);
> +
>  static int curve25519_set_secret(struct crypto_kpp *tfm, const void *buf,
>                                  unsigned int len)
>  {
> --
> 2.24.0
>

On Wed, Dec 11, 2019 at 10:38 AM Ard Biesheuvel
<ard.biesheuvel@linaro.org> wrote:
>
> On Wed, 11 Dec 2019 at 10:27, Jason A. Donenfeld <Jason@zx2c4.com> wrote:
> >
> > Somehow this was forgotten when Zinc was being split into oddly shaped
> > pieces, resulting in linker errors.
>
> Zinc has no historical significance here

Wow...

On Wed, Dec 11, 2019 at 10:38 AM Ard Biesheuvel
<ard.biesheuvel@linaro.org> wrote:
> > The x86_64 glue has a specific key
> > generation implementation, but the Arm one does not. However, it can
> > still receive the NEON speedups by calling the ordinary DH function
> > using the base point.
> >
> > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
>
> With the first sentence dropped,
>
> Acked-by: Ard Biesheuvel <ardb@kernel.org>

Herbert - can you pick this up for 5.5-rc2 rather than 5.6?

On Wed, Dec 11, 2019 at 10:26:39AM +0100, Jason A. Donenfeld wrote:
> Somehow this was forgotten when Zinc was being split into oddly shaped
> pieces, resulting in linker errors. The x86_64 glue has a specific key
> generation implementation, but the Arm one does not. However, it can
> still receive the NEON speedups by calling the ordinary DH function
> using the base point.
> 
> Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
> ---
>  arch/arm/crypto/curve25519-glue.c | 7 +++++++
>  1 file changed, 7 insertions(+)

Patch applied.  Thanks.

On Wed, Dec 11, 2019 at 01:07:43PM +0100, Jason A. Donenfeld wrote:
>
> Herbert - can you pick this up for 5.5-rc2 rather than 5.6?

Yes I'll be pushing this patch for 5.5.

Cheers,