RE: [PATCH 1/3] crypto/ipsec_mb: add GCM sgl support to aesni_mb

["Power, Ciara" <ciara.power@intel.com>]

Hi Pablo,

> -----Original Message-----
> From: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>
> Sent: Sunday 8 May 2022 15:40
> To: Power, Ciara <ciara.power@intel.com>; dev@dpdk.org
> Cc: Zhang, Roy Fan <roy.fan.zhang@intel.com>; Ji, Kai <kai.ji@intel.com>
> Subject: RE: [PATCH 1/3] crypto/ipsec_mb: add GCM sgl support to aesni_mb
> 
> Hi Ciara,
> 
> 
> > -----Original Message-----
> > From: Power, Ciara <ciara.power@intel.com>
> > Sent: Thursday, April 7, 2022 11:31 AM
> > To: dev@dpdk.org
> > Cc: Zhang, Roy Fan <roy.fan.zhang@intel.com>; Ji, Kai
> > <kai.ji@intel.com>; Power, Ciara <ciara.power@intel.com>; De Lara
> > Guarch, Pablo <pablo.de.lara.guarch@intel.com>
> > Subject: [PATCH 1/3] crypto/ipsec_mb: add GCM sgl support to aesni_mb
> >
> > Add SGL support for GCM algorithm through JOB API.
> >
> > This change supports IN-PLACE SGL, OOP SGL IN and LB OUT, and OOP SGL
> > IN and SGL OUT.
> >
> > Feature flags are not added, as the PMD does not yet support SGL for
> > all other algorithms.
> >
> > Signed-off-by: Ciara Power <ciara.power@intel.com>
> > ---
> >  drivers/crypto/ipsec_mb/pmd_aesni_mb.c      | 144
> +++++++++++++++++++-
> >  drivers/crypto/ipsec_mb/pmd_aesni_mb_priv.h |   2 +
> >  2 files changed, 142 insertions(+), 4 deletions(-)
<snip>
> 
> >--- a/drivers/crypto/ipsec_mb/pmd_aesni_mb_priv.h
> > +++ b/drivers/crypto/ipsec_mb/pmd_aesni_mb_priv.h
> > @@ -946,6 +946,8 @@ struct aesni_mb_session {
> >  	struct {
> >  		/* * AAD data length */
> >  		uint16_t aad_len;
> > +
> > +		struct gcm_context_data gcm_sgl_ctx;
> 
> I don't think it's necessary to have this context data in here, you can declare it
> inside set_mb_job_params, unless this causes a performance drop.
> 

[CP] 
It can't be declared in set_mb_job_params unfortunately, because it needs to exist after that function ends,
which is when the final job is submitted.
I have now moved it to qp_data rather than session.

Have implemented your other suggestions in a v2, just sent to ML.

Thanks,
Ciara


> Thanks,
> Pablo
> 
> >  	} aead;
> >  } __rte_cache_aligned;
> >
> > --
> > 2.25.1