linux-kernel - Re: [PATCH v4 1/2] crypto: engine

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <AM6PR0402MB370147736BEAD099AAF061558CCC0@AM6PR0402MB3701.eurprd04.prod.outlook.com>
Date:   Fri, 27 Mar 2020 10:44:02 +0000
From:   Iuliana Prodan <iuliana.prodan@....com>
To:     Herbert Xu <herbert@...dor.apana.org.au>
CC:     Baolin Wang <baolin.wang@...aro.org>,
        Ard Biesheuvel <ard.biesheuvel@...aro.org>,
        Corentin Labbe <clabbe.montjoie@...il.com>,
        Horia Geanta <horia.geanta@....com>,
        Maxime Coquelin <mcoquelin.stm32@...il.com>,
        Alexandre Torgue <alexandre.torgue@...com>,
        Maxime Ripard <mripard@...nel.org>,
        Aymen Sghaier <aymen.sghaier@....com>,
        "David S. Miller" <davem@...emloft.net>,
        Silvano Di Ninno <silvano.dininno@....com>,
        Franck Lenormand <franck.lenormand@....com>,
        "linux-crypto@...r.kernel.org" <linux-crypto@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        dl-linux-imx <linux-imx@....com>
Subject: Re: [PATCH v4 1/2] crypto: engine - support for parallel requests

On 3/27/2020 6:44 AM, Herbert Xu wrote:
> On Tue, Mar 17, 2020 at 01:08:18PM +0000, Iuliana Prodan wrote:
>>
>> This case can happen right now, also. I can't guarantee that all drivers
>> send all requests via crypto-engine.
> 
> I think this is something that we should address.  


I agree we should not have failed requests, _with_ MAY_BACKLOG, that 
pass through crypto-engine. All these requests must me executed. For 
this case I have a solution and I'll work on it.

> If a driver
> is going to use crypto_engine then really all requests should go
> through that.  IOW, we should have a one-to-one relationship between
> engines and hardware.
> 
This cannot happen right now.
For non-crypto API requests (like split key or rng) I cannot pass them 
through crypto-engine which only supports aead, skcipher, hash and 
akcipher requests.

So, I believe that if the first problem is solved (to have all crypto 
API requests, that have backlog flag executed - retry mechanism instead 
of reporting failure) the non-crypto API request doesn't have to be pass 
through crypto-engine - these can be handled by driver.

Thanks,
Iulia