netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Zilvinas Valinskas <zilvinas@gemtek.lt>
To: netdev@oss.sgi.com
Subject: Crypto tests via tcrypt.o modules failes
Date: Thu, 16 Sep 2004 21:11:22 +0300	[thread overview]
Message-ID: <1095358282.3359.83.camel@swoop.gemtek.lt> (raw)

Hello, 

I've got here Xscale IXP425, vanilla linux 2.4.27 + ipsec + Xscale arch
patch (I use IPsec patch backported from 2.6 kernel by Herbert Xu).  
I've successfully compiled kernel, and booted on Xscale and Intel CPUs.

Loaded encryption modules and ran 'insmod tcrypt.o' to verify if
encryption is fine. Not all tests went through. 

Details are here,http://www.gemtek.lt/~zilvinas/crypto/tcrypt-tests

Further testing reveals :
1. AH with 'Null' encryption and w/o IPComp  works just fine. 

2. I've tried ESP with AES-CBC.
2.1  <Xscale> <-> <Xscale> no luck, computers can't ping each other.
  (the same linux 2.4.27 + ipsec kernel on both boards).

2.2  <Xscale> <-> <Intel> , the same 2.4.27 + ipsec kernel was used,
  compiled for a different architectures though. No luck either.

2.3  <Xscale> <-> <Intel 2.6.9-rc1> - no luch either.

2.4  <Intel - 2.4.27 + ipsec> <-> <Intel - 2.6.9-rc1> success, 
   with ESP/AES-CBC, AH  ... 


Does it mean, crypto is not big-endian safe/clean ? 
 
Any pointers where to start or ideas how to track and fix - are welcome.
(I don't have much expertise here though .. Never too late to start
learning).

I would say Xscale IXP425 has a rather interesting 32, 16 bits
read/write results if address is not aligned properly. In the same
directory - http://www.gemtek.lt/~zilvinas/crypto/ i've put *.pdf
describing what exactly is happening. 

Perhaps that is a culprit ?

Thank you, I appreciate any help.

ps.
I am not subscribed to netdev@oss.sgi.com, please keep me on CC:

             reply	other threads:[~2004-09-16 18:11 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-09-16 18:11 Zilvinas Valinskas [this message]
2004-09-16 19:25 ` Crypto tests via tcrypt.o modules failes James Morris
2004-09-17  7:49   ` Zilvinas Valinskas
2004-09-16 19:29 ` James Morris
2004-09-17  7:43   ` Zilvinas Valinskas
2004-09-17  7:56   ` Zilvinas Valinskas
2004-09-17 14:58 ` James Morris
2004-09-17 15:21   ` Zilvinas Valinskas
2004-09-17 16:09     ` James Morris
2004-09-28  9:37       ` Zilvinas Valinskas

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=1095358282.3359.83.camel@swoop.gemtek.lt \
    --to=zilvinas@gemtek.lt \
    --cc=netdev@oss.sgi.com \
    /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).