[dpdk-dev] [RFC] lib/crypto: mark asym session-buffer non-reuseable

Ayuj Verma ayverma at marvell.com
Thu Jul 18 11:56:18 CEST 2019


Hi Arek, Fiona


Please see in-line.


Thanks and regards

Ayuj Verma


________________________________
From: Kusztal, ArkadiuszX <arkadiuszx.kusztal at intel.com>
Sent: 17 July 2019 19:40
To: Trahe, Fiona; Ayuj Verma; akhil.goyal at nxp.com
Cc: Shally Verma; Sunila Sahu; Kanaka Durga Kotamarthy; dev at dpdk.org
Subject: RE: [RFC] lib/crypto: mark asym session-buffer non-reuseable



> -----Original Message-----
> From: Trahe, Fiona
> Sent: Wednesday, July 17, 2019 4:04 PM
> To: Ayuj Verma <ayverma at marvell.com>; akhil.goyal at nxp.com
> Cc: Kusztal, ArkadiuszX <arkadiuszx.kusztal at intel.com>;
> shallyv at marvell.com; ssahu at marvell.com; kkotamarthy at marvell.com;
> dev at dpdk.org; Trahe, Fiona <fiona.trahe at intel.com>
> Subject: RE: [RFC] lib/crypto: mark asym session-buffer non-reuseable
>
> Hi Ayuj,
>
> > -----Original Message-----
> > From: Ayuj Verma [mailto:ayverma at marvell.com]
> > Sent: Wednesday, July 17, 2019 1:34 PM
> > To: akhil.goyal at nxp.com
> > Cc: Kusztal, ArkadiuszX <arkadiuszx.kusztal at intel.com>; Trahe, Fiona
> > <fiona.trahe at intel.com>; shallyv at marvell.com; ssahu at marvell.com;
> > kkotamarthy at marvell.com; dev at dpdk.org; Ayuj Verma
> > <ayverma at marvell.com>
> > Subject: [RFC] lib/crypto: mark asym session-buffer non-reuseable
> >
> > Update asym xform usage in cryptodev documentation.
> >
> > Xform are immutable, non-reuseable entity till life time of session.
> > This allow some PMD to optimize session setup time.
> >
> > Signed-off-by: Ayuj Verma <ayverma at marvell.com>
> > Signed-off-by: Shally Verma <shallyv at marvell.com>
> > ---
> >  doc/guides/prog_guide/cryptodev_lib.rst  | 6 ++++++
> > lib/librte_cryptodev/rte_cryptodev_pmd.h | 2 +-
> >  2 files changed, 7 insertions(+), 1 deletion(-)
> >
> > diff --git a/doc/guides/prog_guide/cryptodev_lib.rst
> > b/doc/guides/prog_guide/cryptodev_lib.rst
> > index 9719944..c97a239 100644
> > --- a/doc/guides/prog_guide/cryptodev_lib.rst
> > +++ b/doc/guides/prog_guide/cryptodev_lib.rst
> > @@ -894,6 +894,12 @@ asymmetric crypto chaining is Diffie-Hellman
> > private key generation followed by  public generation. Also, currently
> > API does not support chaining of symmetric and  asymmetric crypto
> > xforms.
> >
> > +Transform is attached to session during asym session initialization
> > +and can't be modified during session configuration. It remains
> > +constant till the end of life
> [Fiona] I agree with the intent of the RFC. And suggest being more explicit
> and saying "It and all the data buffers it points to should remain constant..."
[AK] I agree with RFC and Fiona :).
[Ayuj] ACK. will submit a v2.
>
> > +span of a session. It should be used as it is in PMD, PMDs which
> > +requires modification of these immutable data should internally do
> > +memcpy of data and perform required operations.
> [Fiona] for security I suggest adding here "It's the PMDs responsibility to
> ensure that any private data copied to local PMD storage is not stored by the
> PMD for longer than the session lifetime.
>
> >  Each xform defines specific asymmetric crypto algo. Currently supported
> are:
> >  * RSA
> >  * Modular operations (Exponentiation and Inverse) diff --git
> > a/lib/librte_cryptodev/rte_cryptodev_pmd.h
> > b/lib/librte_cryptodev/rte_cryptodev_pmd.h
> > index defe05e..1f083ea 100644
> > --- a/lib/librte_cryptodev/rte_cryptodev_pmd.h
> > +++ b/lib/librte_cryptodev/rte_cryptodev_pmd.h
> > @@ -290,7 +290,7 @@ typedef int
> (*cryptodev_sym_configure_session_t)(struct rte_cryptodev *dev,
> >   *  - Returns -ENOMEM if the private session could not be allocated.
> >   */
> >  typedef int (*cryptodev_asym_configure_session_t)(struct rte_cryptodev
> *dev,
> > -           struct rte_crypto_asym_xform *xform,
> > +           const struct rte_crypto_asym_xform *xform,
> >              struct rte_cryptodev_asym_session *session,
> >              struct rte_mempool *mp);
> >  /**
> > --
> > 1.8.3.1



More information about the dev mailing list