[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170123161458.GZ3781@localhost.localdomain>
Date: Mon, 23 Jan 2017 14:14:58 -0200
From: marcelo.leitner@...il.com
To: David Miller <davem@...emloft.net>
Cc: David.Laight@...LAB.COM, lucien.xin@...il.com,
netdev@...r.kernel.org, linux-sctp@...r.kernel.org,
nhorman@...driver.com, vyasevich@...il.com
Subject: Re: [PATCHv3 net-next 3/4] sctp: add support for generating stream
reconf add incoming/outgoing streams request chunk
On Mon, Jan 23, 2017 at 11:00:47AM -0500, David Miller wrote:
> From: "'Marcelo Ricardo Leitner'" <marcelo.leitner@...il.com>
> Date: Mon, 23 Jan 2017 10:36:28 -0200
>
> > So, no padding. A field just after the other, which is what we want on a
> > network header.
>
> It isn't necessary!
>
> Show me a case where it is required when you use properly fixed sized
> types and a proper ordering of the struct members. No padding is
> going in there, go and check.
>
> Do we splatter __packed all over our ipv4/ipv6 header, TCP header, UDP
> header, etc. structures? No, we don't because it's totally unecessary.
Err, sure, right.
>
> I will not accept __packed being used unless it is absolutely, provably,
> the only way to solve a particular problem. And when that does happen,
> I am going to require a huge comment explaining in detail why this is
> the case, and why no other approach or solution solved the problem.
Would this be a candidate for checkpatch.pl?
Marcelo
Powered by blists - more mailing lists