Message-ID: <65c90674808fc0a93c7d329067bf3e80736a003a.camel@gmail.com>
Subject: Re: [PATCH] mctp: Remove only static neighbour on RTM_DELNEIGH
From:   Gagan Kumar <gagan1kumar.cs@gmail.com>
To:     Jeremy Kerr <jk@codeconstruct.com.au>,
        Jakub Kicinski <kuba@kernel.org>
Cc:     matt@codeconstruct.com.au, davem@davemloft.net,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Date:   Fri, 31 Dec 2021 17:44:15 +0530
In-Reply-To: <e20e47833649b5141fa327aa8113e34d4b1bbe15.camel@codeconstruct.com.au>
References: <20211228130956.8553-1-gagan1kumar.cs@gmail.com>
         <20211230175112.7daeb74e@kicinski-fedora-PC1C0HJN.hsd1.ca.comcast.net>
         <e20e47833649b5141fa327aa8113e34d4b1bbe15.camel@codeconstruct.com.au>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.40.4-1 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi Jeremy and Jakub,

Thanks for the response.

On Fri, 2021-12-31 at 11:33 +0800, Jeremy Kerr wrote:

> Hi Jakub & Gagan,
> 
> > > Add neighbour source flag in mctp_neigh_remove(...) to allow
> > > removal of only static neighbours.
> > 
> > Which are the only ones that exist today right?
> 
> That's correct. There may be a future facility for the kernel to
> perform
> neighbour discovery itself (somewhat analogous to ARP), but only the
> static entries are possible at the moment.
> 
> > Can you clarify the motivation and practical impact of the change 
> > in the commit message to make it clear? AFAICT this is a no-op / prep
> > for some later changes, right Jeremy?
> 
> Yes, it'll be a no-op now; I'm not aware of any changes coming that
> require parameterisation of the neighbour type yet.
> 
> Gagan - can you provide any context on this change?

I was exploring the repository and wanted to get familiar with the
patching process. During that, I was looking for some TODOs in /net for
my first patch and came across mctp.

I thought `TODO: add a "source" flag so netlink can only delete static
neighbours?` might be of some use in the future. So, thought of sending
a patch for the same.

If I were to think like a critic, "You aren't gonna need it" principle
can be applied here.

If you think it's ok to proceed I can update the commit message to
include the motivation and impact as a no-op.

> 
> Cheers,
> 
> 
> Jeremy

Thanks,
Gagan