Add a script that measures DAD behavior (Dictionary Attack Defense)

Add a script that measures DAD behavior (Dictionary Attack Defense)

Change-Id: I303bb68c366c382caff20c1ee8dbfb97ed5e1c2d

BUG=none
TEST=ran the script

Committed: http://chrome-svn/viewvc/chromeos?view=rev&revision=7c6a69f

[Luigi Semenzato, 2010-09-24 21:30:16]

I figure we may need this to test changes to DAD policy.

[fes, 2010-09-27 22:20:08]

So, this basically repeatedly hammers the TPM until it goes into defend lock;
and then measures the amount of time spent in defend lock?  On some vendors,
would the failure of tpm_clear not extend the time in DAD?

http://codereview.chromium.org/3492011/diff/1/2
File utility/tpm-dad-lock (right):

http://codereview.chromium.org/3492011/diff/1/2#newcode26
utility/tpm-dad-lock:26: before=$(date +%s)
nit: space

http://codereview.chromium.org/3492011/diff/1/2#newcode37
utility/tpm-dad-lock:37: after=$(date +%s)
nit: space

[Luigi Semenzato, 2010-09-27 22:45:44]

Indeed different vendors have different ways of managing DAD and may have
different locking strategies for different types of attacks.  This is just one
attempt which happens to work with all of the vendors I have tried so far (and
which I will not name here).

The space nit, you mean $( date +%s ) right?

Thanks!

On 2010/09/27 22:20:08, fes wrote:
> So, this basically repeatedly hammers the TPM until it goes into defend lock;
> and then measures the amount of time spent in defend lock?  On some vendors,
> would the failure of tpm_clear not extend the time in DAD?
> 
> http://codereview.chromium.org/3492011/diff/1/2
> File utility/tpm-dad-lock (right):
> 
> http://codereview.chromium.org/3492011/diff/1/2#newcode26
> utility/tpm-dad-lock:26: before=$(date +%s)
> nit: space
> 
> http://codereview.chromium.org/3492011/diff/1/2#newcode37
> utility/tpm-dad-lock:37: after=$(date +%s)
> nit: space

[fes, 2010-09-27 22:55:21]

LGTM then.

No, in my haste, I read it wrong, ignore that.

On 2010/09/27 22:45:44, Luigi Semenzato wrote:
> Indeed different vendors have different ways of managing DAD and may have
> different locking strategies for different types of attacks.  This is just one
> attempt which happens to work with all of the vendors I have tried so far (and
> which I will not name here).
> 
> The space nit, you mean $( date +%s ) right?
> 
> Thanks!
> 
> On 2010/09/27 22:20:08, fes wrote:
> > So, this basically repeatedly hammers the TPM until it goes into defend
lock;
> > and then measures the amount of time spent in defend lock?  On some vendors,
> > would the failure of tpm_clear not extend the time in DAD?
> > 
> > http://codereview.chromium.org/3492011/diff/1/2
> > File utility/tpm-dad-lock (right):
> > 
> > http://codereview.chromium.org/3492011/diff/1/2#newcode26
> > utility/tpm-dad-lock:26: before=$(date +%s)
> > nit: space
> > 
> > http://codereview.chromium.org/3492011/diff/1/2#newcode37
> > utility/tpm-dad-lock:37: after=$(date +%s)
> > nit: space

[Luigi Semenzato, 2010-09-28 17:31:47]

PTAL!  I changed tpm_clear to tpm_getpubek.  Same authentication, but tpm_clear
has unwanted side effects when it succeeds.  (It's not supposed to ever succeed
in the script, but for extra testing I want to issue the same command
successfully outside the script.)

On 2010/09/27 22:55:21, fes wrote:
> LGTM then.
> 
> No, in my haste, I read it wrong, ignore that.
> 
> On 2010/09/27 22:45:44, Luigi Semenzato wrote:
> > Indeed different vendors have different ways of managing DAD and may have
> > different locking strategies for different types of attacks.  This is just
one
> > attempt which happens to work with all of the vendors I have tried so far
(and
> > which I will not name here).
> > 
> > The space nit, you mean $( date +%s ) right?
> > 
> > Thanks!
> > 
> > On 2010/09/27 22:20:08, fes wrote:
> > > So, this basically repeatedly hammers the TPM until it goes into defend
> lock;
> > > and then measures the amount of time spent in defend lock?  On some
vendors,
> > > would the failure of tpm_clear not extend the time in DAD?
> > > 
> > > http://codereview.chromium.org/3492011/diff/1/2
> > > File utility/tpm-dad-lock (right):
> > > 
> > > http://codereview.chromium.org/3492011/diff/1/2#newcode26
> > > utility/tpm-dad-lock:26: before=$(date +%s)
> > > nit: space
> > > 
> > > http://codereview.chromium.org/3492011/diff/1/2#newcode37
> > > utility/tpm-dad-lock:37: after=$(date +%s)
> > > nit: space

[fes, 2010-09-30 17:02:21]

LGTM