Re: [PATCH] tcp: tcp_drop adds `SNMP` and `reason` parameter for tracing

From: Steven Rostedt <rostedt@goodmis.org>
To: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Zhongya Yan <yan2228598786@gmail.com>,
	Eric Dumazet <edumazet@google.com>,
	Brendan Gregg <brendan.d.gregg@gmail.com>,
	Linux Kernel Network Developers <netdev@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Jakub Kicinski <kuba@kernel.org>, Ingo Molnar <mingo@redhat.com>,
	David Miller <davem@davemloft.net>,
	Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
	David Ahern <dsahern@kernel.org>, Yonghong Song <yhs@fb.com>,
	Zhongya Yan <2228598786@qq.com>
Subject: Re: [PATCH] tcp: tcp_drop adds `SNMP` and `reason` parameter for tracing
Date: Mon, 20 Sep 2021 13:24:40 -0400	[thread overview]
Message-ID: <20210920132440.7a2febce@oasis.local.home> (raw)
In-Reply-To: <CAM_iQpVVbwN+uzxPbJabvju0CzxcFUosy0_xN=X9anA+wtiUVQ@mail.gmail.com>

On Mon, 20 Sep 2021 10:20:33 -0700
Cong Wang <xiyou.wangcong@gmail.com> wrote:

> > The strings were requested by the networking maintainers.
> >
> >   https://lore.kernel.org/all/CANn89iJO8jzjFWvJ610TPmKDE8WKi8ojTr_HWXLz5g=4pdQHEA@mail.gmail.com/  
> 
> I think you misunderstand my point. Eric's point is hex address
> vs. string, which I never disagree. With SNMP enum, user-space
> can easily interpret it to string too, so at the end you still get strings
> but not from kernel. This would at least save a handful of strings
> from vmlinux, especially if we expand it beyond TCP.

If an enum is saved in the ring buffer, it is also trivial to have it
show a string in the buffer, as there's a way to map enums to strings
on output as well. Which I discussed in that same thread.

-- Steve