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:   Tue, 8 Sep 2020 16:04:09 +0200
From:   Jiri Pirko <jiri@resnulli.us>
To:     Ido Schimmel <idosch@idosch.org>
Cc:     Aya Levin <ayal@mellanox.com>,
        "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        Jiri Pirko <jiri@mellanox.com>, netdev@vger.kernel.org,
        Moshe Shemesh <moshe@mellanox.com>,
        Eran Ben Elisha <eranbe@mellanox.com>,
        Ido Schimmel <idosch@mellanox.com>,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH net-next RFC v1 2/4] devlink: Add devlink traps under
 devlink_ports context
Message-ID: <20200908140409.GN2997@nanopsycho.orion>
References: <1599060734-26617-1-git-send-email-ayal@mellanox.com>
 <1599060734-26617-3-git-send-email-ayal@mellanox.com>
 <20200906154428.GA2431016@shredder>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200906154428.GA2431016@shredder>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Sun, Sep 06, 2020 at 05:44:28PM CEST, idosch@idosch.org wrote:
>On Wed, Sep 02, 2020 at 06:32:12PM +0300, Aya Levin wrote:

[...]

>
>I understand how this struct allows you to re-use a lot of code between
>per-device and per-port traps, but it's mainly enabled by the fact that
>you use the same netlink commands for both per-device and per-port
>traps. Is this OK?
>
>I see this is already done for health reporters, but it's inconsistent
>with the devlink-param API:
>
>DEVLINK_CMD_PARAM_GET
>DEVLINK_CMD_PARAM_SET
>DEVLINK_CMD_PARAM_NEW
>DEVLINK_CMD_PARAM_DEL
>
>DEVLINK_CMD_PORT_PARAM_GET
>DEVLINK_CMD_PORT_PARAM_SET
>DEVLINK_CMD_PORT_PARAM_NEW
>DEVLINK_CMD_PORT_PARAM_DEL
>
>And also with the general device/port commands:
>
>DEVLINK_CMD_GET
>DEVLINK_CMD_SET
>DEVLINK_CMD_NEW
>DEVLINK_CMD_DEL
>
>DEVLINK_CMD_PORT_GET
>DEVLINK_CMD_PORT_SET
>DEVLINK_CMD_PORT_NEW
>DEVLINK_CMD_PORT_DEL
>
>Wouldn't it be cleaner to add new commands?
>
>DEVLINK_CMD_PORT_TRAP_GET
>DEVLINK_CMD_PORT_TRAP_SET
>DEVLINK_CMD_PORT_TRAP_NEW
>DEVLINK_CMD_PORT_TRAP_DEL
>
>I think the health API is the exception in this case and therefore might
>not be the best thing to mimic. IIUC, existing per-port health reporters
>were exposed as per-device and later moved to be exposed as per-port
>[1]:
>
>"This patchset comes to fix a design issue as some health reporters
>report on errors and run recovery on device level while the actual
>functionality is on port level. As for the current implemented devlink
>health reporters it is relevant only to Tx and Rx reporters of mlx5,
>which has only one port, so no real effect on functionality, but this
>should be fixed before more drivers will use devlink health reporters."

Yeah, this slipped trough my fingers unfortunatelly :/ But with
introduction of per-port health reporters, we could introduce new
commands, that would be no problem. Pity :/


>
>[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=ac4cd4781eacd1fd185c85522e869bd5d3254b96
>
>Since we still don't have per-port traps, we can design it better from
>the start.

I agree. Let's have a separate commands for per-port.


[...]