Re: [dpdk-dev] [PATCH] doc: announce ABI change for crypto info struct

["De Lara Guarch, Pablo" <pablo.de.lara.guarch@intel.com>]

Hi Shally,

> -----Original Message-----
> From: Verma, Shally [mailto:Shally.Verma@cavium.com]
> Sent: Tuesday, January 30, 2018 11:54 AM
> To: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>; Akhil Goyal
> <akhil.goyal@nxp.com>; Trahe, Fiona <fiona.trahe@intel.com>;
> hemant.agrawal@nxp.com; Doherty, Declan <declan.doherty@intel.com>;
> Griffin, John <john.griffin@intel.com>; Jain, Deepak K
> <deepak.k.jain@intel.com>; jck@semihalf.com; tdu@semihalf.com;
> dima@marvell.com; nsamsono@marvell.com; jianbo.liu@arm.com; Jacob,
> Jerin <Jerin.JacobKollanukkaran@cavium.com>; Athreya, Narayana Prasad
> <NarayanaPrasad.Athreya@cavium.com>; Murthy, Nidadavolu
> <Nidadavolu.Murthy@cavium.com>
> Cc: dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH] doc: announce ABI change for crypto info
> struct
> 
> 
> 
> >-----Original Message-----
> >From: De Lara Guarch, Pablo [mailto:pablo.de.lara.guarch@intel.com]
> >Sent: 30 January 2018 16:51
> >To: Verma, Shally <Shally.Verma@cavium.com>; Akhil Goyal
> ><akhil.goyal@nxp.com>; Trahe, Fiona <fiona.trahe@intel.com>;
> >hemant.agrawal@nxp.com; Doherty, Declan <declan.doherty@intel.com>;
> >Griffin, John <john.griffin@intel.com>; Jain, Deepak K
> ><deepak.k.jain@intel.com>; jck@semihalf.com; tdu@semihalf.com;
> >dima@marvell.com; nsamsono@marvell.com; jianbo.liu@arm.com; Jacob,
> >Jerin <Jerin.JacobKollanukkaran@cavium.com>; Athreya, Narayana Prasad
> ><NarayanaPrasad.Athreya@cavium.com>; Murthy, Nidadavolu
> ><Nidadavolu.Murthy@cavium.com>
> >Cc: dev@dpdk.org
> >Subject: RE: [dpdk-dev] [PATCH] doc: announce ABI change for crypto
> >info struct
> >
> >Hi Shally/Ahkil,
> >
> >> -----Original Message-----
> >> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Verma, Shally
> >> Sent: Tuesday, January 30, 2018 7:56 AM
> >> To: Akhil Goyal <akhil.goyal@nxp.com>; De Lara Guarch, Pablo
> >> <pablo.de.lara.guarch@intel.com>; Trahe, Fiona
> >> <fiona.trahe@intel.com>; hemant.agrawal@nxp.com; Doherty, Declan
> >> <declan.doherty@intel.com>; Griffin, John <john.griffin@intel.com>;
> >> Jain, Deepak K <deepak.k.jain@intel.com>; jck@semihalf.com;
> >> tdu@semihalf.com; dima@marvell.com; nsamsono@marvell.com;
> >> jianbo.liu@arm.com; Jacob, Jerin
> >> <Jerin.JacobKollanukkaran@cavium.com>; Athreya, Narayana Prasad
> >> <NarayanaPrasad.Athreya@cavium.com>; Murthy, Nidadavolu
> >> <Nidadavolu.Murthy@cavium.com>
> >> Cc: dev@dpdk.org
> >> Subject: Re: [dpdk-dev] [PATCH] doc: announce ABI change for crypto
> >> info struct
> >>
> >> I do see current cryptodev unit testcase (inside \test dir) uses
> >> info.sym.max_nb_sessions param for session mempool_create. So, such
> >> testcases change are also in proposal?
> >
> >Yes, for these tests, we can just define a macro in the tests, instead of
> using the info structure.
> 
> [Shally] Ok, then you mean applications will choose any random number
> during mempool_create and not dependent on device max_nb_sessions?

Yes, actually for the unit tests, even one session is enough.

> 
> >>
> >> Another point, we recently submitted an RFC patch on lib/cryptodev
> >> with asymmetric crypto support
> >> (https://dpdk.org/dev/patchwork/patch/34308/) which is awaiting
> >> review and these fields have role to play there.
> >> So, could this change be please viewed in conjunction with asym RFC?
> >
> >Do you need it for asymmetric? Anyway, this would remove the
> symmetric function and structures, not applicable for you.
> 
> [Shally] I would say addition of asym in lib/cryptodev is not entirely
> standalone, specifically for PMDs that can support both.
> My key concern are max_nb_sessions_per_qp and related
> qp_attach_sym/asym APIs which enable management of queue distribution
> among sym and asym in current proposal, specifically, for PMDs that can
> support both but have dedicated qp for each. Right now proposal is open
> for feedback and would prefer to be covered before sym related changes
> could be applied.

Actually, I have been thinking about this. Given the time we have until 18.02 is out,
and that this is not urgent to be applied (this is just code cleanup),
I am postponing this until next release. 

My other reason is that the info structure has a rte_pci_device pointer which should be removed.
However, I believe it is better to leave it for next release and discuss it with other libraries which has this, like ethdev.

Thanks,
Pablo