DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas@monjalon.net>
To: "Dumitrescu, Cristian" <cristian.dumitrescu@intel.com>
Cc: dev@dpdk.org,
	"jerin.jacob@caviumnetworks.com" <jerin.jacob@caviumnetworks.com>,
	"hemant.agrawal@nxp.com" <hemant.agrawal@nxp.com>,
	"Singh, Jasvinder" <jasvinder.singh@intel.com>,
	"Lu, Wenzhuo" <wenzhuo.lu@intel.com>,
	"Yigit, Ferruh" <ferruh.yigit@intel.com>
Subject: Re: [dpdk-dev] [pull-request] next-tm 17.08 pre-rc1
Date: Tue, 04 Jul 2017 22:21:42 +0200	[thread overview]
Message-ID: <1793823.SEqWim7Qm5@xps> (raw)
In-Reply-To: <3EB4FA525960D640B5BDFFD6A3D891267BA78E7C@IRSMSX108.ger.corp.intel.com>

04/07/2017 18:52, Dumitrescu, Cristian:
> > Hi Cristian,
> > 
> > > Dumitrescu, Cristian (2):
> > >       ethdev: add traffic management ops get API
> > >       ethdev: add traffic management API
> > 
> > The original request was to split this huge patch.
> > It is too messy to bring a whole new API area in one patch.
> > We have nothing to refer in case of bug, and it is hard to dive in.
> > 
> > Please, could you try to split it, bringing features one by one?
> 
> Hi Thomas,
> 
> Technically, it can be done, but IMO it should not be done this way for the following reasons:
> 
> 1. None of the new APIs recently introduced in DPDK follow this approach. The rte_flow [1] and the eventdev [2] API are of the same order of magnitude with the TM API, and both were introduced as a single patch header file. Why do things differently for TM API?

Yes you're right, same magnitude (but 2 times bigger).
I would have preffered eventdev and rte_flow be better introduced.

> 2. Breaking an API header file into multiple patches usually does not make sense because the sub-components are inter-connected and cross-referenced. When evaluating an API, it needs to be evaluated as a whole for consistency reasons rather than piece by piece. On TM API for example, the capability API is inter-connected with congestion management, shaping, scheduling and marking features; cman and shaping are connected to the nodes that make up the scheduling tree, etc. IMO the end result is adding more confusion than clarity.

For me it's simpler to start with basic stuff and add more features.
But it may be just a taste.

> This request also comes very late in our preparation to hit RC1. I know you made this mention previously, but I regarded it as a comment/suggestion rather than a hard requirement (sorry for not explaining it my rationale better at the time). You also had several other comments and requests that we fulfilled, as described in the revision history.

I had not seen any news about this patchset and the tree was empty
during a long time so I thought you were working on it.

> So, what do you want me to do?
> If you still want to go ahead with this request, I will do my best to do it and still meet RC1.

No, I do not want to insist.
I understand you have a different taste than mine :)

I will check for pulling your tree in following days.
Please try to be available on IRC, in case I catch a last minute detail to fix.
Thanks

  reply	other threads:[~2017-07-04 20:21 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-07-04 15:38 Cristian Dumitrescu
2017-07-04 15:47 ` Thomas Monjalon
2017-07-04 16:52   ` Dumitrescu, Cristian
2017-07-04 20:21     ` Thomas Monjalon [this message]
2017-07-05 10:36       ` Dumitrescu, Cristian
2017-07-09 20:01 ` Thomas Monjalon
2017-07-10  7:43   ` Adrien Mazarguil
2017-07-10  7:51     ` Thomas Monjalon
2017-07-10 10:55   ` Dumitrescu, Cristian
2017-07-10 12:57     ` Thomas Monjalon
2017-07-10 13:21       ` Dumitrescu, Cristian
2017-07-10 13:49         ` Thomas Monjalon
2017-07-10 15:46           ` Dumitrescu, Cristian
2017-07-10 15:54             ` Thomas Monjalon
2017-07-10 16:47               ` Dumitrescu, Cristian
2017-07-10 16:58                 ` Thomas Monjalon
2017-07-11 18:20                   ` Dumitrescu, Cristian
2017-07-12 17:33                     ` Thomas Monjalon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1793823.SEqWim7Qm5@xps \
    --to=thomas@monjalon.net \
    --cc=cristian.dumitrescu@intel.com \
    --cc=dev@dpdk.org \
    --cc=ferruh.yigit@intel.com \
    --cc=hemant.agrawal@nxp.com \
    --cc=jasvinder.singh@intel.com \
    --cc=jerin.jacob@caviumnetworks.com \
    --cc=wenzhuo.lu@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).