[dpdk-dev] Napatech pmd

Neil Horman nhorman at tuxdriver.com
Sun Apr 19 23:16:25 CEST 2020


On Fri, Apr 17, 2020 at 04:38:42AM +0000, Michael Lilja wrote:
> > -----Original Message-----
> > From: Neil Horman <nhorman at tuxdriver.com>
> > Sent: 17. april 2020 04:55
> > To: Thomas Monjalon <thomas at monjalon.net>
> > Cc: Finn Christensen <fc at napatech.com>; dev at dpdk.org; Bent Kuhre
> > <bk at napatech.com>; Michael Lilja <ml at napatech.com>; techboard at dpdk.org
> > Subject: Re: [dpdk-dev] Napatech pmd
> >
> > On Tue, Mar 31, 2020 at 10:07:12PM +0200, Thomas Monjalon wrote:
> > > 31/03/2020 21:56, Neil Horman:
> > > > On Tue, Mar 31, 2020 at 02:29:08PM +0200, Thomas Monjalon wrote:
> > > > > 31/03/2020 14:17, Neil Horman:
> > > > > > On Tue, Mar 31, 2020 at 01:25:25PM +0200, Thomas Monjalon
> > wrote:
> > > > > > > Hi,
> > > > > > >
> > > > > > > Raising this topic again.
> > > > > > >
> > > > > > > As said in the past, it is better to have this PMD inside
> > DPDK.
> > > > > > > We discussed some concerns, but I think the consensus was to
> > > > > > > integrate Napatech PMD anyway.
> > > > > > >
> > > > > > > I am sad that you did not feel welcome enough to follow up
> > > > > > > with patches during all these years.
> > > > > > > Please would you like to restart the upstreaming process?
> > > > > > >
> > > > > > Whats changed here?
> > > > >
> > > > > Nothing changed, except years.
> > > > >
> > > > > > I still don't see what the advantage is to accepting this code
> > in the DPDK tree.
> > > > > > No one will be able to use it without accepting Napatechs
> > > > > > license for their underlying library.  As such, the code can't
> > > > > > really be maintained at all by anyone other than Napatech in
> > the
> > > > > > community, and so may as well just be maintained as an out of
> > tree driver.
> > > > >
> > > > > You are the only one having this concern.
> > > > I don't think its wise to assume that silence implies acceptance.
> > > >
> > > > > Nobody from the Technical Board looks to be against the
> > acceptance.
> > > > >
> > > > > The advantage is simple: Napatech customers will be able to run
> > any DPDK version.
> > > > Why is that not possible by having napatech maintain an out-of-
> > tree
> > > > PMD?  Theres no reason that can't be done.
> > >
> > > They are maintaining an out-of-tree PMD:
> > > https://github.com/napatech/dpdk/releases
> > >
> > > I'm just trying to improve the situation, avoiding DPDK forks.
> > >
> > >
> > >
> > Apologies, I completely missed responding to this note
> >
> > I took a look at the PMD above.  Its not an open source implementation
> > of their driver, its the same thing they offered 4 years ago, a
> > skeleton pmd that still uses the same closed licensed library.
> >
> > It was my understanding that they were working on a completely open
> > sourced PMD that could be generally useful to the community.  If that
> > exists, then yes, by all means, lets take a look at it, and consider
> > merging it.  That effort deserves consideration.
> >
> > This however, is the same thing we saw last time.  Theres no benefit
> > in including that
> >
> > Neil
> I understand the confusion. The PMD in our github is still, as you correctly state, based on our closed source driver and only a skeleton. We are working on a open source version, but currently that is WIP and not pushed yet. I'll let you know when there is something to look at.
> 

So, I have to ask.  I referenced this email from 2016 earlier in this
thread:
https://mails.dpdk.org/archives/dev/2016-September/046522.html

Where a colleague of yours from Napatech noted that you were working on
an fully open source driver.  Given that you have been working on this
to some degree since then, I would presume that you could share what
code you have thus far.  Can you place the code you have written thus
far in a public repository so we can start reviewing it?

Thanks
Neil
> Michael
> 
> Disclaimer: This email and any files transmitted with it may contain confidential information intended for the addressee(s) only. The information is not to be surrendered or copied to unauthorized persons. If you have received this communication in error, please notify the sender immediately and delete this e-mail from your system.
> 


More information about the dev mailing list