[sheepdog] [PATCH] test: Implentment fast random on AES

[Liu Yuan]

On 05/06/2013 03:45 PM, Joseph Glanville wrote:
> Turns out this doesn't increase the test performance any as it is
> actually I/O bound on my system.

I have tried to modify '/dev/urandom' as '/dev/zero' for 008 and the
result is

vanilla: 94s, modified: 80s

Looks it is a good boost if we can have a faster generator as fast as
'dd /dev/zero'.

Thanks,
Yuan