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: <6c4fbd8240af542f2c5e26e990825f1232009aaf.camel.ref@cs.com> <6c4fbd8240af542f2c5e26e990825f1232009aaf.camel@cs.com>
In-Reply-To: <6c4fbd8240af542f2c5e26e990825f1232009aaf.camel@cs.com>
From:   Ard Biesheuvel <ard.biesheuvel@linaro.org>
Date:   Mon, 30 Dec 2019 16:11:31 +0100
Message-ID: <CAKv+Gu91psRa=Jyug5Hfcsse-ZgbG-6eVpgVSKUszqU6oPxZXw@mail.gmail.com>
Subject: Re: Internal crypto test fail: changed 'req->iv'
To:     Richard van Schagen <vschagen@cs.com>
Cc:     "open list:HARDWARE RANDOM NUMBER GENERATOR CORE" 
        <linux-crypto@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-crypto-owner@vger.kernel.org
Precedence: bulk

On Mon, 30 Dec 2019 at 11:10, Richard van Schagen <vschagen@cs.com> wrote:
>
> I am writing a module for the EIP93 crypto engine. From what I have
> been reading on this mailing list, the driver should return the updated
> IV in order for the caller to =E2=80=9Cresume=E2=80=9D or =E2=80=9Ccontin=
ue=E2=80=9D with this IV in
> another call.
>
> A code-snippet:
>
> int ivsize =3D crypto_skcipher_ivsize(skcipher);
>
> If (ivsize)
>         memcpy(req->iv, rctx->lastiv, ivsize);
>
> Where rctx->lastiv was read from the hardware itself.
>
> The fail message I am getting is:
> [   57.290000] alg: skcipher: changed 'req->iv'
> [   57.370000] alg: skcipher: eip93-cbc-aes encryption co
> rrupted request struct on test vector 0, cfg=3D"in-place"
> [   57.380000] alg: skcipher: changed 'req->iv'
> [   57.460000] alg: skcipher: eip93-ctr-aes encryption corrupted
> request struct on test vector 0, cfg=3D"in-place"
> [   57.470000] alg: skcipher: changed 'req->iv'
> [   57.560000] alg: skcipher: eip93-rfc3686(ctr)-aes encryption
> corrupted request struct on test vector 0, cfg=3D"in-place"
> [   57.570000] alg: skcipher: changed 'req->iv=E2=80=99
>
> Where/How should I return the new/updated IV ?
>

Are you sure you are not making req->iv point to a different buffer?