Date:   Fri, 25 Jan 2019 21:49:02 +0800
From:   Herbert Xu <herbert@gondor.apana.org.au>
To:     Roland Hieber <rhi@pengutronix.de>
Cc:     Horia =?utf-8?Q?Geant=C4=83?= <horia.geanta@nxp.com>,
        Aymen Sghaier <aymen.sghaier@nxp.com>,
        linux-crypto@vger.kernel.org,
        Roland Hieber <r.hieber@pengutronix.de>,
        "David S. Miller" <davem@davemloft.net>, kernel@pengutronix.de
Subject: Re: [PATCH 2/2] crypto: caam - fix DMA mapping of stack memory
Message-ID: <20190125134902.yp7kjja6aecce6vm@gondor.apana.org.au>
References: <20190122152651.1150-1-rhi@pengutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20190122152651.1150-1-rhi@pengutronix.de>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-crypto-owner@vger.kernel.org
Precedence: bulk

On Tue, Jan 22, 2019 at 04:26:51PM +0100, Roland Hieber wrote:
>
> The cause is that the struct ahash_request *req is created as a
> stack-local variable up in the stack (presumably somewhere in the IMA
> implementation), then passed down into the CAAM driver, which tries to
> dma_single_map the req->result (indirectly via map_seq_out_ptr_result)
> in order to make that buffer available for the CAAM to store the result
> of the following hash operation.

Then IMA needs to be fixed.  Asynchronous crypto requests must be
allocated with kmalloc so this is an API violation.

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt