Received-SPF: pass (google.com: best guess record for domain of linux-crypto-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <CAOtvUMcwLtwgigFE2mx7LVjhhEgcZsSS4WyR_SQ2gixTZxyBfg@mail.gmail.com>
 <CAOtvUMeVXTDvH5bxVFemYmD9rpZ=xX3MkypAGyZn5VROw6sgZg@mail.gmail.com>
 <20200207072709.GB8284@sol.localdomain> <70156395ce424f41949feb13fd9f978b@MN2PR20MB2973.namprd20.prod.outlook.com>
 <SN4PR0401MB366399E54E5B7EE0E54A7E0BC31C0@SN4PR0401MB3663.namprd04.prod.outlook.com>
In-Reply-To: <SN4PR0401MB366399E54E5B7EE0E54A7E0BC31C0@SN4PR0401MB3663.namprd04.prod.outlook.com>
From:   Gilad Ben-Yossef <gilad@benyossef.com>
Date:   Sun, 9 Feb 2020 10:09:53 +0200
Message-ID: <CAOtvUMeFZXwxxYT1hz=e09CaBrv1qBXvWcRCghA=wRGwZZ9S3g@mail.gmail.com>
Subject: Re: Possible issue with new inauthentic AEAD in extended crypto tests
To:     "Van Leeuwen, Pascal" <pvanleeuwen@rambus.com>
Cc:     Stephan Mueller <smueller@chronox.de>,
        Eric Biggers <ebiggers@kernel.org>,
        Herbert Xu <herbert@gondor.apana.org.au>,
        Linux Crypto Mailing List <linux-crypto@vger.kernel.org>,
        Geert Uytterhoeven <geert@linux-m68k.org>,
        David Miller <davem@davemloft.net>,
        Ofir Drang <Ofir.Drang@arm.com>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-crypto-owner@vger.kernel.org
Precedence: bulk

On Fri, Feb 7, 2020 at 4:07 PM Van Leeuwen, Pascal
<pvanleeuwen@rambus.com> wrote:

> The "problem" Gilad was referring to is that the _explicit_ part of the  IV appears to be
> available  from both req->iv and from the AAD scatterbuffer. Which one should you use?
> API wise I would assume req->iv but from a (our) hardware perspective, it would
> be more efficient to extract it from the datastream. But is it allowed to assume
> there is a valid IV stored there? (which implies that it has to match req->iv,
> otherwise behaviour would deviate from implementations using that)
>


No, it isn't.

The problem that I was referring to was that part of our test suites
passes different values in req->iv and as part of the AAD,
in contrast to what we document as the API requirements in the include
file, my understanding of the relevant standard and
the single users of this API in the kernel and that the driver I'm
maintaining fails these tests,

I'm all fine with getting my hands dirty and fixing the driver, I'm
just suspect fixing a driver to pass a test that misuses the API
may not actually improve the quality of the driver.

Gilad