Received-SPF: softfail (google.com: domain of transitioning linux-crypto-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
MIME-Version: 1.0
References: <20220126145322.646723-1-p.zabel@pengutronix.de>
 <c572bf6f0b0a5d7fd3f8f0744a85eb5660a003d4.camel@pengutronix.de> <YjE5UCeoziA8f+Q4@gondor.apana.org.au>
In-Reply-To: <YjE5UCeoziA8f+Q4@gondor.apana.org.au>
From:   Ard Biesheuvel <ardb@kernel.org>
Date:   Wed, 16 Mar 2022 08:23:24 +0100
Message-ID: <CAMj1kXF-BdRCN-239cRHgSGM3K9EPSrRFEDJu+e6Gtri2pONaA@mail.gmail.com>
Subject: Re: [PATCH] crypto: algapi - Remove test larvals to fix error paths
To:     Herbert Xu <herbert@gondor.apana.org.au>
Cc:     Philipp Zabel <p.zabel@pengutronix.de>,
        Linux Crypto Mailing List <linux-crypto@vger.kernel.org>,
        kernel@pengutronix.de, "David S. Miller" <davem@davemloft.net>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, 16 Mar 2022 at 02:11, Herbert Xu <herbert@gondor.apana.org.au> wrote:
>
> On Wed, Jan 26, 2022 at 04:01:04PM +0100, Philipp Zabel wrote:
> >
> > I see this happen on ARM with CONFIG_CRYPTO_AES_ARM_BS=y since v5.16-rc1
> > because the simd_skcipher_create_compat("ecb(aes)", "ecb-aes-neonbs",
> > "__ecb-aes-neonbs") call in arch/arm/crypto/aes-neonbs-glue.c returns
> > -ENOENT. I believe that is the same issue as reported in [1].
>
> I cannot reproduce this crash with qemu.  If you can still
> reproduce this, please send me your complete kconfig file.
>

According to the bisect log in the other thread,
adad556efcdd42a1d9e060cb is the culprit, which does not seem
surprising, at is would result in the SIMD skcipher being encapsulated
to not be available yet when the SIMD helper tries to take a reference
to it.

Given that the resulting algo will be subject to testing as well,
could we simply override the test requirement for the inner algo? That
should make it available immediately, IIUC, and the outer algo test is
guaranteed to cover the inner one.