From: "Horia Geantă" <horia.geanta@nxp.com>
To: Gilad Ben-Yossef <gilad@benyossef.com>
Cc: David Gstir <david@sigma-star.at>,
Dan Douglass <dan.douglass@nxp.com>,
"herbert@gondor.apana.org.au" <herbert@gondor.apana.org.au>,
"davem@davemloft.net" <davem@davemloft.net>,
"richard@sigma-star.at" <richard@sigma-star.at>,
"linux-crypto@vger.kernel.org" <linux-crypto@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"stable@vger.kernel.org" <stable@vger.kernel.org>
Subject: Re: [PATCH] crypto: caam - properly set IV after {en,de}crypt
Date: Thu, 7 Sep 2017 10:12:56 +0000 [thread overview]
Message-ID: <VI1PR0402MB3342196084129C22870CD18898940@VI1PR0402MB3342.eurprd04.prod.outlook.com> (raw)
In-Reply-To: CAOtvUMc9PzdfifN7T3cgMemr6j3qxBb3KD_uXmHYAk0HuxUgRA@mail.gmail.com
On 9/6/2017 1:14 PM, Gilad Ben-Yossef wrote:
> On Tue, Sep 5, 2017 at 6:33 PM, Horia Geantă <horia.geanta@nxp.com> wrote:
>> On 8/14/2017 10:59 AM, Gilad Ben-Yossef wrote:
>>> Hi,
>>>
>>> On Thu, Jun 29, 2017 at 1:19 PM, Horia Geantă <horia.geanta@nxp.com> wrote:
>>>> On 6/28/2017 4:42 PM, Horia Geantă wrote:
>>>>> On 6/28/2017 4:27 PM, David Gstir wrote:
>>>>>> Certain cipher modes like CTS expect the IV (req->info) of
>>>>>> ablkcipher_request (or equivalently req->iv of skcipher_request) to
>>>>>> contain the last ciphertext block when the {en,de}crypt operation is done.
>>>>>> This is currently not the case for the CAAM driver which in turn breaks
>>>>>> e.g. cts(cbc(aes)) when the CAAM driver is enabled.
>>>>>>
>>>>>> This patch fixes the CAAM driver to properly set the IV after the
>>>>>> {en,de}crypt operation of ablkcipher finishes.
>>>>>>
>>>>>> This issue was revealed by the changes in the SW CTS mode in commit
>>>>>> 0605c41cc53ca ("crypto: cts - Convert to skcipher")
>>>>>>
>>>>>> Cc: <stable@vger.kernel.org> # 4.8+
>>>>>> Signed-off-by: David Gstir <david@sigma-star.at>
>>>>> Reviewed-by: Horia Geantă <horia.geanta@nxp.com>
>>>>>
>>>> Btw, instead of updating the IV in SW, CAAM engine could be programmed
>>>> to do it - by saving the Context Register of the AES accelerator.
>>>>
>>>> Unfortunately this would require changes in quite a few places: shared
>>>> descriptor, HW S/G generation logic, IV dma (un)mapping and maybe others.
>>>>
>>>> So it's better to have this fix now (which, considering size, is
>>>> appropriate for -stable) and later, if needed, offload IV updating in HW.
>>>>
>>>
>>> My apologies for reviving this thread from the dead, but doesn't the patch fail
>>> for in-place decryption since we are copying from req->dst after
>>> the operation is done, and therefore it no longer contains the ciphertext?
>>>
>> You are right, thanks! Will follow up with a fix.
>> Though cts(cbc(aes)) in particular is working, see below.
>>
>>> I'm asking since I ran into a similar issue in the ccree driver and thought
>>> to deploy a similar fix but could not convince myself why this works.
>>>
>> IIUC cts(cbc(aes)) in-place decryption (with cbc(aes) offloaded to CAAM
>> engine) works since SW implementation of cts, when performing the
>> ciphertext stealing phase in cts_cbc_decrypt() does not use req->iv, but
>> a previously value, saved before staring decryption in crypto_cts_decrypt():
>>
>> if (cbc_blocks <= 1)
>> memcpy(space, req->iv, bsize);
>> else
>> scatterwalk_map_and_copy(space, req->src, offset - 2 * bsize,
>> bsize, 0);
>>
>
> Is that not a performance bug in software CTS than? I mean all those
> transformation
> drivers doing that extra copy and possibly malloc and free to save the
> data for the info
> and than have the CTS implementation ignore that and do its own memory copy?
>
AFAICT, in case cbc_blocks > 1 cts saves the second from last full
block, while underlying cbc subrequest saves the last full block.
Horia
next prev parent reply other threads:[~2017-09-07 10:13 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-02 12:24 [RFC PATCH 0/2] crypto: caam - fix cts(cbc(aes)) with CAAM driver David Gstir
2017-06-02 12:24 ` [RFC PATCH 1/2] crypto: caam - properly set IV after {en,de}crypt David Gstir
2017-06-19 10:31 ` Horia Geantă
2017-06-20 1:28 ` Herbert Xu
2017-06-26 5:40 ` David Gstir
2017-06-26 6:47 ` Herbert Xu
2017-06-28 8:32 ` Horia Geantă
2017-06-28 9:03 ` David Gstir
2017-06-02 12:24 ` [RFC PATCH 2/2] crypto: caam - fix k*alloc if called from own cipher callback David Gstir
2017-06-13 11:53 ` [RFC PATCH 0/2] crypto: caam - fix cts(cbc(aes)) with CAAM driver David Gstir
2017-06-15 14:57 ` Horia Geantă
2017-06-16 7:57 ` Horia Geantă
2017-06-16 7:59 ` Herbert Xu
2017-06-16 21:01 ` Horia Geantă
2017-06-17 9:05 ` David Gstir
2017-06-19 8:44 ` [PATCH 1/2] crypto: caam - fix gfp allocation flags (part I) Horia Geantă
2017-06-19 8:44 ` [PATCH 2/2] crypto: caam - fix gfp allocation flags (part II) Horia Geantă
2017-06-22 9:00 ` Herbert Xu
2017-06-22 9:00 ` [PATCH 1/2] crypto: caam - fix gfp allocation flags (part I) Herbert Xu
2017-06-28 13:27 ` [PATCH] crypto: caam - properly set IV after {en,de}crypt David Gstir
2017-06-28 13:42 ` Horia Geantă
2017-06-29 10:19 ` Horia Geantă
2017-08-14 7:59 ` Gilad Ben-Yossef
2017-09-05 15:33 ` Horia Geantă
2017-09-06 10:14 ` Gilad Ben-Yossef
2017-09-07 10:12 ` Horia Geantă [this message]
2017-07-12 10:51 ` Herbert Xu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=VI1PR0402MB3342196084129C22870CD18898940@VI1PR0402MB3342.eurprd04.prod.outlook.com \
--to=horia.geanta@nxp.com \
--cc=dan.douglass@nxp.com \
--cc=davem@davemloft.net \
--cc=david@sigma-star.at \
--cc=gilad@benyossef.com \
--cc=herbert@gondor.apana.org.au \
--cc=linux-crypto@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=richard@sigma-star.at \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).