Mailing List Archive

[patch] Additional HMAC unit tests
The attached patch provides HMAC unit tests for a range of the more critical
hashes - MD5, SHA224, SHA256, SHA384 and SHA512, using standard test vectors.
This patch also changes around the format of the output of --verbose mode to
make it clearer what is actually being tested.
It also changes the public key tests to what I believe is the correct
behaviour - see email from earlier today.
It also enables some tests for the Tiger hash algorithm. It isn't clear why
the tests weren't enable, but they work fine.

Overall, this is a low priority patch.

This won't build unless you already have my SHA224 patch applied, and it won't
pass unless you have my SHA384/SHA512 blocksize patch applied.

Brad
Re: [patch] Additional HMAC unit tests [ In reply to ]
On Sat, 18 Feb 2006 22:15:13 +1100, Brad Hards said:

> The attached patch provides HMAC unit tests for a range of the more critical
> hashes - MD5, SHA224, SHA256, SHA384 and SHA512, using standard test vectors.

I have just commited it. However teh HMAC tests for SHA-384 and
SHA-512 are failing. Would you mind to look after that?

> This patch also changes around the format of the output of --verbose mode to
> make it clearer what is actually being tested.

Indeed, this is better.

> It also changes the public key tests to what I believe is the correct
> behaviour - see email from earlier today.

Generating all these keys takes some time now :-(.



Shalom-Salam,

Werner


_______________________________________________
Gcrypt-devel mailing list
Gcrypt-devel@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gcrypt-devel
Re: [patch] Additional HMAC unit tests [ In reply to ]
On Friday 10 March 2006 20:59 pm, Werner Koch wrote:
> On Sat, 18 Feb 2006 22:15:13 +1100, Brad Hards said:
> > The attached patch provides HMAC unit tests for a range of the more
> > critical hashes - MD5, SHA224, SHA256, SHA384 and SHA512, using standard
> > test vectors.
>
> I have just commited it. However teh HMAC tests for SHA-384 and
> SHA-512 are failing. Would you mind to look after that?
It is a problem with the blocksize (128, instead of 64) - I raised this
problem in April 2005:
http://lists.gnupg.org/pipermail/gcrypt-devel/2005-April/000778.html.

I posted a (somewhat hacky) patch to the list:
http://lists.gnupg.org/pipermail/gcrypt-devel/2005-April/000779.html
[.The actual patch is missing from the archive - I can resend it if required].

Brad
Re: [patch] Additional HMAC unit tests [ In reply to ]
On Sun, 12 Mar 2006 17:40:36 +1100, Brad Hards said:

> It is a problem with the blocksize (128, instead of 64) - I raised this
> problem in April 2005:
> http://lists.gnupg.org/pipermail/gcrypt-devel/2005-April/000778.html.

I had something like this in mind...

> I posted a (somewhat hacky) patch to the list:
> http://lists.gnupg.org/pipermail/gcrypt-devel/2005-April/000779.html
> [.The actual patch is missing from the archive - I can resend it if required].

.. but did not found the pacth. Please resend it.

Thanks,

Werner


_______________________________________________
Gcrypt-devel mailing list
Gcrypt-devel@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gcrypt-devel