phc-discussions - Re: [PHC] Re: Argon2 PHC release...

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 for Android: free password hash cracker in your pocket

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

Message-Id: <32E656FD-EFED-4121-A489-E2F8BADD44E5@gmail.com>
Date: Sat, 17 Oct 2015 15:14:28 +0200
From: Dmitry Khovratovich <khovratovich@...il.com>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] Re: Argon2 PHC release...

At some point the ref version yielded better performance than optimized:)

Will check how it works now. Also a lot of Blake code is not used, so we think of trimming/commenting it

> On 17 Oct 2015, at 14:25, Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
> 
> Another question (sorry): why just using blake2 ref code rather than the version supporting SIMD instructions? for consistency of benchmarks?
> 
> Thanks!
> 
>> On Sat, Oct 17, 2015 at 8:05 AM Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
>> Recommended parameters section says 
>> "If you do not know the difference between them or you consider side-channel attacks as viable threat, choose Argon2i."
>> and I think it's a good recommendation; safer by default.
>> 
>>> On Sat, Oct 17, 2015 at 1:29 AM Dmitry Khovratovich <khovratovich@...il.com> wrote:
>>> The spec lists 2d and 2i as primary ones, any suggestions to modify?
>>> 
>>> Sent from my iPhone
>>> 
>>>> On 17 Oct 2015, at 00:30, Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
>>>> 
>>>> The simplified API hash() (formerly PHS()) now defaults to Argon2i, as used in
>>>> https://github.com/P-H-C/phc-winner-argon2/blob/master/examples/hash.c
>>>> 
>>>> 4 instances in the final version is too much, we should have only 2: i and d, or some variants thereof. 
>>>> 
>>>>> On Fri, Oct 16, 2015 at 11:04 PM Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
>>>>> Samuel and I did a number of fixes to https://github.com/P-H-C/phc-winner-argon2. If you clone this repo and run ./test.sh you should get OKs for all the tests. Or not; please let us know then.
>>>>> 
>>>>> I've started a basic README at https://github.com/P-H-C/phc-winner-argon2, which will contain usage instructions (copying from https://github.com/khovratovich/Argon2). Any suggestion or contribution welcome :-)
>>>>> 
>>>>> 
>>>>> 
>>>>>> On Fri, Oct 16, 2015 at 4:09 PM Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
>>>>>> Did a couple more tweaks, including a usage() function https://github.com/P-H-C/phc-winner-argon2/blob/master/src/argon2-test.c printed when no argument is given. Still a lot to do here, and ideally we should use getopt for this rather than manual parsing.
>>>>>> 
>>>>>>> On Fri, Oct 16, 2015 at 10:58 AM Jean-Philippe Aumasson <jeanphilippe.aumasson@...il.com> wrote:
>>>>>>> Working on a simpler code package, current version posted at
>>>>>>> https://github.com/P-H-C/phc-winner-argon2
>>>>>>> 
>>>>>>> So far reorganized and cleaned up the whole structure, keeping C only and not C++, ported a minor change to the blake2b().
>>>>>>> 
>>>>>>> With the Argon2 designers we'll improve the test suite and will make the argon2 command-line interface more user-friendly.
>>>>>>> 
>>>>>>> Comments and pull requests welcome :-)

Content of type "text/html" skipped