Re: [PATCH 1/3] crypto: inside-secure - add support for authenc(hmac(sha1),cbc(des3_ede))

[Antoine Tenart <antoine.tenart@bootlin.com>]

On Tue, Jul 30, 2019 at 02:01:46PM +0000, Pascal Van Leeuwen wrote:
> > From: linux-crypto-owner@vger.kernel.org <linux-crypto-owner@vger.kernel.org> On Behalf Of
> > Pascal Van Leeuwen
> > Sent: Friday, July 26, 2019 2:57 PM
> > To: Antoine Tenart <antoine.tenart@bootlin.com>; Pascal van Leeuwen <pascalvanl@gmail.com>
> > Cc: linux-crypto@vger.kernel.org; herbert@gondor.apana.org.au; davem@davemloft.net
> > Subject: RE: [PATCH 1/3] crypto: inside-secure - add support for
> > authenc(hmac(sha1),cbc(des3_ede))
> > 
> > > > +	.alg.aead = {
> > > > +		.setkey = safexcel_aead_setkey,
> > > > +		.encrypt = safexcel_aead_encrypt_3des,
> > > > +		.decrypt = safexcel_aead_decrypt_3des,
> > > > +		.ivsize = DES3_EDE_BLOCK_SIZE,
> > > > +		.maxauthsize = SHA1_DIGEST_SIZE,
> > > > +		.base = {
> > > > +			.cra_name = "authenc(hmac(sha1),cbc(des3_ede))",
> > > > +			.cra_driver_name = "safexcel-authenc-hmac-sha1-cbc-des3_ede",
> > >
> > > You could drop "_ede" here, or s/_/-/.
> > >
> > Agree the underscore should not be there.
> > Our HW does not support any other form of 3DES so EDE doesn't
> > really add much here, therefore I will just remove "_ede" entirely.
> >
> Actually, while looking into fixing this, I noticed that this 
> naming style is actually consistent with the already existing
> 3des ecb and cbc ciphersuites, e.g.: "safexcel-cbc-des3_ebe",
> so for consistency I would then suggest keeping it (or 
> change the other 2 3des references at the same time, but I
> don't know if that would break any legacy dependency).

Good catch :) As you said, you should keep it this way then.

Thanks,
Antoine

-- 
Antoine Ténart, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com