From mboxrd@z Thu Jan  1 00:00:00 1970
From: clabbe.montjoie@gmail.com (Corentin LABBE)
Date: Sat, 26 Jul 2014 16:01:26 +0200
Subject: [PATCH v4 3/3] crypto: Add Allwinner Security System crypto
 accelerator
In-Reply-To: <20140724133823.GA9638@gondor.apana.org.au>
References: <1405169953-13695-1-git-send-email-clabbe.montjoie@gmail.com>
 <1405169953-13695-4-git-send-email-clabbe.montjoie@gmail.com>
 <20140724060054.GA6545@gondor.apana.org.au> <53D0E857.8000405@gmail.com>
 <20140724133823.GA9638@gondor.apana.org.au>
Message-ID: <53D3B4B6.3000100@gmail.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

Le 24/07/2014 15:38, Herbert Xu a ?crit :
> On Thu, Jul 24, 2014 at 01:04:55PM +0200, Corentin LABBE wrote:
>> Le 24/07/2014 08:00, Herbert Xu a ?crit :
>>> On Sat, Jul 12, 2014 at 02:59:13PM +0200, LABBE Corentin wrote:
>>>>
>>>> +/* sunxi_hash_init: initialize request context
>>>> + * Activate the SS, and configure it for MD5 or SHA1
>>>> + */
>>>> +int sunxi_hash_init(struct ahash_request *areq)
>>>> +{
>>>> +	const char *hash_type;
>>>> +	struct crypto_ahash *tfm = crypto_ahash_reqtfm(areq);
>>>> +	struct sunxi_req_ctx *op = crypto_ahash_ctx(tfm);
>>>> +
>>>> +	mutex_lock(&ss->lock);
>>>> +
>>>> +	hash_type = crypto_tfm_alg_name(areq->base.tfm);
>>>> +
>>>> +	op->byte_count = 0;
>>>> +	op->nbwait = 0;
>>>> +	op->waitbuf = 0;
>>>> +
>>>> +	/* Enable and configure SS for MD5 or SHA1 */
>>>> +	if (strcmp(hash_type, "sha1") == 0)
>>>> +		op->mode = SS_OP_SHA1;
>>>> +	else
>>>> +		op->mode = SS_OP_MD5;
>>>> +
>>>> +	writel(op->mode | SS_ENABLED, ss->base + SS_CTL);
>>>> +	return 0;
>>>
>>> The hash driver is completely broken.  You are modifying tfm
>>> ctx data which is shared by all users of a single tfm.  So
>>> if two users conduct hashes in parallel they will step all
>>> over each other.
>>
>> So where can I store data for each request ?
> 
> Well, first of all you need to stop storing state in the hardware.
> After each operation the hardware may be used by some other user
> for a completely different hash request.  So leaving the hash state
> in the hardware is a no-no.
> 
> If your hardware supports exporting the hash state then you just
> have to export it after each operation and reimporting before the
> next one.

Even if it is undocumented, the hardware seems to support it.
Since crypto_ahash_ctx is for a tfm, does ahash_request_ctx is the good place to store data ?
(after a call to crypto_ahash_set_reqsize in cra_init)

I have also seen export/import function, does I need to use it ?


> 
> If your hardware is incapable of exporting partial hash state then
> you will have to use a software fallback for init/update.  If your
> hardware is incapable of importing partial hash state then you will
> also have to do finup/final using a software fallback.
> 
> Cheers,
>