From: romain.perier@free-electrons.com (Romain Perier)
To: linux-arm-kernel@lists.infradead.org
Subject: [crypto] [marvell-cesa] Possible regression after Linux 4.7
Date: Thu, 8 Sep 2016 16:52:16 +0200 [thread overview]
Message-ID: <57D17B20.8070606@free-electrons.com> (raw)
In-Reply-To: <57ce53d6.0a28190a.3d6d2.44b2@mx.google.com>
Hello,
Le 06/09/2016 07:27, radioconfusion at gmail.com a ?crit :
> Hello,
>
> I'm testing the marvell-cesa -driver on Armada 385 board and I think I've found
> a regression between Linux 4.7 and 4.8-rc4.
> I want to accelerate my curl connections. So I compiled the 4.8-rc4 with
> marvell-cesa enabled and HEAD revision of cryptodev-linux.
>
> Here is my output:
>
> ~# uname -r
> 4.8.0-rc4
> ~# modprobe cryptodev
> ~# modprobe marvell-cesa
> ~# curl -k --ciphers AES256-SHA256 https://myserver/myfile >/dev/null
> % Total % Received % Xferd Average Speed Time Time Time Current
> Dload Upload Total Spent Left Speed
> 19 200M 19 39.7M 0 0 16.2M 0 0:00:12 0:00:02 0:00:10 16.2M
> curl: (56) SSL read: error:1408F119:SSL routines:SSL3_GET_RECORD:decryption
> failed or bad record mac, errno 0
>
> This error is quite easy to reproduce. Especially if you have some other
> connections open at same time or two simultaneous curl-connections.
> I had a ssh-connection running top(1).
>
> I can not reproduce this with Linux 4.7 or without modprobing marvell-cesa.
>
> I'm not using any special kernel parameter, patches, custom dts, etc.
>
> Best Regards,
> Jussi
>
Thank you for your feedbacks.
I can reproduce the issue with and without multiple engines. With the
following steps:
1. open an ssh connection on the target, and start the following command:
# while true; do curl -k -v --ciphers AES256-SHA256 https://foobar
>/dev/null > /dev/null; done
2. From the uart :
# curl -k --ciphers AES256-SHA256 https://foobar >/dev/null
Then, the second command will fail easily.
However, I have found an interesting thing, If I use only --ciphers
AES256, it works perfectly fine, could you confirm this ?
In any case, I will investigate.
Thanks!
Romain
--
Romain Perier, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
next prev parent reply other threads:[~2016-09-08 14:52 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-06 5:27 [crypto] [marvell-cesa] Possible regression after Linux 4.7 radioconfusion at gmail.com
2016-09-08 14:52 ` Romain Perier [this message]
-- strict thread matches above, loose matches on Subject: below --
2016-09-09 7:24 radioconfusion at gmail.com
2016-10-31 13:59 radioconfusion at gmail.com
2016-11-03 13:57 ` Romain Perier
2016-12-05 9:19 ` Romain Perier
2016-11-07 8:53 radioconfusion at gmail.com
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=57D17B20.8070606@free-electrons.com \
--to=romain.perier@free-electrons.com \
--cc=linux-arm-kernel@lists.infradead.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