Received-SPF: pass (google.com: domain of linux-kernel+bounces-186394-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Message-ID: <0c12c9ea10aa97e246230fc33e6b35c571102b48.camel@gmail.com>
Subject: Re: [PATCH 1/3] tpm: Disable TCG_TPM2_HMAC by default
From: Vitor Soares <ivitro@gmail.com>
To: Jarkko Sakkinen <jarkko@kernel.org>, James Bottomley
	 <James.Bottomley@HansenPartnership.com>, linux-integrity@vger.kernel.org
Cc: keyrings@vger.kernel.org, Peter Huewe <peterhuewe@gmx.de>, Jason
 Gunthorpe <jgg@ziepe.ca>, Mimi Zohar <zohar@linux.ibm.com>, David Howells
 <dhowells@redhat.com>, Paul Moore <paul@paul-moore.com>, James Morris
 <jmorris@namei.org>, "Serge E. Hallyn" <serge@hallyn.com>, 
 linux-kernel@vger.kernel.org, linux-security-module@vger.kernel.org
Date: Wed, 22 May 2024 15:58:20 +0100
In-Reply-To: <D1G8HOCIDWTC.2ERVA0CYHLY0B@kernel.org>
References: <20240519235122.3380-1-jarkko@kernel.org>
	 <20240519235122.3380-2-jarkko@kernel.org>
	 <850862655008f84ef0b6ecd99750e8dc395304d1.camel@gmail.com>
	 <D1F4V8NMSUNZ.2VCTEKHZZ0LB@kernel.org>
	 <17dc838120b56ce342c34611596c7b46dcd9ab5a.camel@HansenPartnership.com>
	 <2dd8d49516ec9c7cb8c1182b5b8537b1e82d7067.camel@gmail.com>
	 <17a5dcd7aceb356587ef7c8f45b0f6359b2d2a91.camel@HansenPartnership.com>
	 <D1G8HOCIDWTC.2ERVA0CYHLY0B@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu2 
Precedence: bulk
MIME-Version: 1.0

On Wed, 2024-05-22 at 17:13 +0300, Jarkko Sakkinen wrote:
> On Wed May 22, 2024 at 4:35 PM EEST, James Bottomley wrote:
> > On Wed, 2024-05-22 at 09:18 +0100, Vitor Soares wrote:
> > > On Tue, 2024-05-21 at 08:33 -0400, James Bottomley wrote:
> > > > On Tue, 2024-05-21 at 10:10 +0300, Jarkko Sakkinen wrote:
> > > > > This benchmark could be done in user space using /dev/tpm0.
> > > >=20
> > > > Let's actually try that.=C2=A0 If you have the ibmtss installed, th=
e
> > > > command to time primary key generation from userspace on your tpm
> > > > is
> > > >=20
> > > > time tsscreateprimary -hi n -ecc nistp256
> > > >=20
> > > >=20
> > > > And just for chuckles and grins, try it in the owner hierarchy as
> > > > well (sometimes slow TPMs cache this)
> > > >=20
> > > > time tsscreateprimary -hi o -ecc nistp256
> > > >=20
> > > > And if you have tpm2 tools, the above commands should be:
> > > >=20
> > > > time tpm2_createprimary -C n -G ecc256
> > > > time tpm2_createprimary -C o -G ecc256
> > > >=20
> > > > James
> > > >=20
> > > >=20
> > >=20
> > > Testing on an arm64 platform I get the following results.
> > >=20
> > > hmac disabled:
> > > =C2=A0 time modprobe tpm_tis_spi
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 0m2.776s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.006s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.015s
> > >=20
> > > =C2=A0 time tpm2_createprimary -C n -G ecc256
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 0m0.686s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.044s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.025s
> > >=20
> > > =C2=A0 time tpm2_createprimary -C o -G ecc256
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 0m0.638s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.048s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.009s
> > >=20
> > >=20
> > > hmac enabled:
> > > =C2=A0 time modprobe tpm_tis_spi
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 8m5.840s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.005s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.018s
> > >=20
> > >=20
> > > =C2=A0 time tpm2_createprimary -C n -G ecc256
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 5m27.678s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.059s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.009s
> > >=20
> > > =C2=A0 (after first command)
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 0m0.395s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.040s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.015s
> > >=20
> > > =C2=A0 time tpm2_createprimary -C o -G ecc256
> > > =C2=A0 real=C2=A0=C2=A0=C2=A0 0m0.418s
> > > =C2=A0 user=C2=A0=C2=A0=C2=A0 0m0.049s
> > > =C2=A0 sys=C2=A0=C2=A0=C2=A0=C2=A0 0m0.009s
> >=20
> > That's interesting: it suggests the create primary is fast (as
> > expected) but that the TPM is blocked for some reason.=C2=A0 Is there
> > anything else in dmesg if you do
> >=20
> > dmesg|grep -i tpm
> >=20
> > ?
> >=20
> > Unfortunately we don't really do timeouts on our end (we have the TPM
> > do it instead), but we could instrument your kernel with command and
> > time sent and returned.=C2=A0 That may tell us where the problem lies.
>=20
> If there was possibility to use bpftrace it is trivial to get histogram
> of time used where. I can bake a script but I need to know first if it
> is available in the first place before going through that trouble.
>=20
> BR, Jarkko

I did run with ftrace, but need some more time to go through it.

Here the step I did:
kernel config:
  CONFIG_FUNCTION_TRACER
  CONFIG_FUNCTION_GRAPH_TRACER

ftrace:
  # set filters
  echo tpm* > set_ftrace_filter

  # set tracer
  echo function_graph > current_tracer

  # take the sample
  echo 1 > tracing_on; time modprobe tpm_tis_spi; echo 0 > tracing_on

regards,
Vitor Soares