Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Fri, 29 Jan 2021 11:19:37 -0800
From:   Jakub Kicinski <kuba@kernel.org>
To:     Oleksandr Mazur <oleksandr.mazur@plvision.eu>
Cc:     "netdev@vger.kernel.org" <netdev@vger.kernel.org>,
        "jiri@nvidia.com" <jiri@nvidia.com>,
        "davem@davemloft.net" <davem@davemloft.net>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [RFC v3 net-next] net: core: devlink: add 'dropped' stats field
 for DROP trap action
Message-ID: <20210129111937.4e7e17d0@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
In-Reply-To: <AM0P190MB0738FC4657CCB0E435C40B24E4B99@AM0P190MB0738.EURP190.PROD.OUTLOOK.COM>
References: <20210125123856.1746-1-oleksandr.mazur@plvision.eu>
        <20210125132317.418a4e35@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
        <AM0P190MB0738FC4657CCB0E435C40B24E4B99@AM0P190MB0738.EURP190.PROD.OUTLOOK.COM>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Fri, 29 Jan 2021 11:15:43 +0000 Oleksandr Mazur wrote:
> > Thinking about it again - if the action can be changed wouldn't it 
> > be best for the user to actually get a "HW condition hit" counter,
> > which would increment regardless of SW config (incl. policers)?  
> 
> > Otherwise if admin logs onto the box and temporarily enables a trap 
> > for debug this count would disappear.  
> 
> But still this counter makes sense only for 'drop' action.

Okay, well, "dropped while trap was disabled" seems a lot less useful
of a definition than "number of times this trap would trigger" but if
that's all the HW can provide then it is what it is.

Does the HW also count packets dropped because of overload / overflow
or some other event, or purely dropped because disabled?