[Help-gnutls] Re: Simple question about performance

[Simon Josefsson]

Miroslav Kratochvil <exa.exa at gmail.com> writes:

> Hi,
> big thanks for quick response, and, before all, for correcting the
> misinformation I had found. Moreover, the connecting speed issue
> doesnt really bother me.
>
>>
>> I don't recall much feedback about speed issues.  There is certainly
>> room for optimization.  If you can provide a good test setup to compare
>> gnutls vs openssl in an application, I would be interested in optimizing
>> things.  However, the first step before optimization is to do good
>> benchmarks to illustrate that there is a significant problem.  My last
>> attempt at benchmarking didn't result in any obvious problem so I didn't
>> proceed in optimizing anything.
>
> My thought was to write a libgcrypt alternative of 'openssl speed',
> which would be simple and give fair and comparable results. I will
> probably post it here when it's ready.

Did you see libgcrypt's tests/benchmark?  It outputs results like this:

             ECB             CBC             CFB             OFB             CTR             STREAM         
             --------------- --------------- --------------- --------------- --------------- ---------------
3DES            90ms   100ms   100ms    90ms   100ms    90ms   100ms    90ms   110ms   100ms                
CAST5           30ms    40ms    40ms    40ms    40ms    40ms    40ms    40ms    50ms    40ms                
...

However I agree a tool that computes more statistics, similar to openssl
speed, would be useful.  Maybe it could perform the same computations,
then you can compare the results directly.

gcrypt-speed.c?

/Simon