From: Jiri Pirko <jiri@resnulli.us>
To: Moshe Shemesh <moshe@nvidia.com>
Cc: Moshe Shemesh <moshe@mellanox.com>,
"David S. Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>, Jiri Pirko <jiri@mellanox.com>,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH net-next RFC v3 01/14] devlink: Add reload action option to devlink reload command
Date: Wed, 2 Sep 2020 11:46:27 +0200 [thread overview]
Message-ID: <20200902094627.GB2568@nanopsycho> (raw)
In-Reply-To: <9fffbe80-9a2a-33de-2e11-24be34648686@nvidia.com>
Tue, Sep 01, 2020 at 09:43:00PM CEST, moshe@nvidia.com wrote:
>
>On 8/31/2020 3:15 PM, Jiri Pirko wrote:
>> Sun, Aug 30, 2020 at 05:27:21PM CEST, moshe@mellanox.com wrote:
>> > Add devlink reload action to allow the user to request a specific reload
>> > action. The action parameter is optional, if not specified then devlink
>> > driver re-init action is used (backward compatible).
>> > Note that when required to do firmware activation some drivers may need
>> > to reload the driver. On the other hand some drivers may need to reset
>> > the firmware to reinitialize the driver entities. Therefore, the devlink
>> > reload command returns the actions which were actually done.
>> > However, in case fw_activate_no_reset action is selected, then no other
>> > reload action is allowed.
>> > Reload actions supported are:
>> > driver_reinit: driver entities re-initialization, applying devlink-param
>> > and devlink-resource values.
>> > fw_activate: firmware activate.
>> > fw_activate_no_reset: Activate new firmware image without any reset.
>> > (also known as: firmware live patching).
>> >
>> > command examples:
>> > $devlink dev reload pci/0000:82:00.0 action driver_reinit
>> > reload_actions_done:
>> > driver_reinit
>> >
>> > $devlink dev reload pci/0000:82:00.0 action fw_activate
>> > reload_actions_done:
>> > driver_reinit fw_activate
>> >
>> > Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
>> > ---
>> > v2 -> v3:
>> > - Replace fw_live_patch action by fw_activate_no_reset
>> > - Devlink reload returns the actions done over netlink reply
>> > v1 -> v2:
>> > - Instead of reload levels driver,fw_reset,fw_live_patch have reload
>> > actions driver_reinit,fw_activate,fw_live_patch
>> > - Remove driver default level, the action driver_reinit is the default
>> > action for all drivers
>> > ---
>> [...]
>>
>>
>> > diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c
>> > index 08d101138fbe..c42b66d88884 100644
>> > --- a/drivers/net/ethernet/mellanox/mlxsw/core.c
>> > +++ b/drivers/net/ethernet/mellanox/mlxsw/core.c
>> > @@ -1113,7 +1113,7 @@ mlxsw_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
>> >
>> > static int
>> > mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
>> > - bool netns_change,
>> > + bool netns_change, enum devlink_reload_action action,
>> > struct netlink_ext_ack *extack)
>> > {
>> > struct mlxsw_core *mlxsw_core = devlink_priv(devlink);
>> > @@ -1126,15 +1126,23 @@ mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
>> > }
>> >
>> > static int
>> > -mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink,
>> > - struct netlink_ext_ack *extack)
>> > +mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink, enum devlink_reload_action action,
>> > + struct netlink_ext_ack *extack, unsigned long *actions_done)
>> > {
>> > struct mlxsw_core *mlxsw_core = devlink_priv(devlink);
>> > + int err;
>> >
>> > - return mlxsw_core_bus_device_register(mlxsw_core->bus_info,
>> > - mlxsw_core->bus,
>> > - mlxsw_core->bus_priv, true,
>> > - devlink, extack);
>> > + err = mlxsw_core_bus_device_register(mlxsw_core->bus_info,
>> > + mlxsw_core->bus,
>> > + mlxsw_core->bus_priv, true,
>> > + devlink, extack);
>> > + if (err)
>> > + return err;
>> > + if (actions_done)
>> > + *actions_done = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
>> > + BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE);
>> > +
>> > + return 0;
>> > }
>> >
>> > static int mlxsw_devlink_flash_update(struct devlink *devlink,
>> > @@ -1268,6 +1276,8 @@ mlxsw_devlink_trap_policer_counter_get(struct devlink *devlink,
>> > }
>> >
>> > static const struct devlink_ops mlxsw_devlink_ops = {
>> > + .supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
>> > + BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE),
>> This is confusing and open to interpretation. Does this mean that the
>> driver supports:
>> 1) REINIT && FW_ACTIVATE
>> 2) REINIT || FW_ACTIVATE
>> ?
>>
>> Because mlxsw supports only 1. I guess that mlx5 supports both. This
>> needs to be distinguished.
>
>Mlxsw supports 1, so it supports fw_activation and performs also reinit and
>vice versa.
My point is, your bitfield does not exactly tell what the driver
supports or not.
>
>Mlx5 supports fw_activate and performs also reinit. However, it supports
>reinit without performing fw_activate.
>
>> I think you need an array of combinations. Or perhaps rather to extend
>> the enum with combinations. You kind of have it already with
>> DEVLINK_RELOAD_ACTION_FW_ACTIVATE_NO_RESET
>>
>> Maybe we can have something like:
>> DEVLINK_RELOAD_ACTION_DRIVER_REINIT
>> DEVLINK_RELOAD_ACTION_DRIVER_REINIT_FW_ACTIVATE_RESET
>> DEVLINK_RELOAD_ACTION_FW_ACTIVATE_RESET
>> DEVLINK_RELOAD_ACTION_FW_ACTIVATE (this is the original FW_ACTIVATE_NO_RESET)
>
>The FW_ACTIVATE_NO_RESET meant also to emphasize that driver implementation
>for this one should not do any reset.
>
>So maybe we can have
>
>DEVLINK_RELOAD_ACTION_FW_ACTIVATE_RESET
>DEVLINK_RELOAD_ACTION_FW_ACTIVATE_NO_RESET
Okay.
>
>> Each has very clear meaning.
>
>
>Yes, it the driver support here is more clear.
>
>> Also, then the "actions_done" would be a simple enum, directly returned
>> to the user. No bitfield needed.
>
>
>I agree it is more clear on the driver support side, but what about the uAPI
As I said, there would be one enum value returned to the user. Clear and
simple.
>? Do we need such change there too or keep it as is, each action by itself
>and return what was performed ?
Well, I don't know. User asks for X, X should be performed, not Y or Z.
So perhaps the return value is not needed.
Just driver advertizes it supports X, Y, Z and the users says:
1) do X, driver does X
2) do Y, driver does Y
3) do Z, driver does Z
[
I think this kindof circles back to the original proposal...
>
>>
>> > .reload_down = mlxsw_devlink_core_bus_device_reload_down,
>> > .reload_up = mlxsw_devlink_core_bus_device_reload_up,
>> > .port_type_set = mlxsw_devlink_port_type_set,
>> [...]
next prev parent reply other threads:[~2020-09-02 9:46 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-30 15:27 [PATCH net-next RFC v3 00/14] Add devlink reload action option Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 01/14] devlink: Add reload action option to devlink reload command Moshe Shemesh
2020-08-31 12:15 ` Jiri Pirko
2020-09-01 19:43 ` Moshe Shemesh
2020-09-02 9:46 ` Jiri Pirko [this message]
2020-09-02 15:30 ` Jakub Kicinski
2020-09-03 5:57 ` Jiri Pirko
2020-09-03 19:47 ` Jakub Kicinski
2020-09-04 9:04 ` Jiri Pirko
2020-09-04 19:56 ` Jakub Kicinski
2020-09-07 13:46 ` Moshe Shemesh
2020-09-07 17:58 ` Jakub Kicinski
2020-09-09 13:27 ` Moshe Shemesh
2020-09-09 19:24 ` Jakub Kicinski
2020-09-10 5:16 ` Vasundhara Volam
2020-09-10 6:51 ` Jiri Pirko
2020-08-30 15:27 ` [PATCH net-next RFC v3 02/14] devlink: Add reload actions counters Moshe Shemesh
2020-08-31 10:48 ` Jiri Pirko
2020-09-01 19:05 ` Moshe Shemesh
2020-09-02 0:01 ` Jakub Kicinski
2020-09-04 5:03 ` Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 03/14] devlink: Add reload actions counters to dev get Moshe Shemesh
2020-08-31 10:44 ` Jiri Pirko
2020-09-01 19:00 ` Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 04/14] net/mlx5: Add functions to set/query MFRL register Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 05/14] net/mlx5: Set cap for pci sync for fw update event Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 06/14] net/mlx5: Handle sync reset request event Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 07/14] net/mlx5: Handle sync reset now event Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 08/14] net/mlx5: Handle sync reset abort event Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 09/14] net/mlx5: Add support for devlink reload action fw activate Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 10/14] devlink: Add enable_remote_dev_reset generic parameter Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 11/14] net/mlx5: Add devlink param enable_remote_dev_reset support Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 12/14] net/mlx5: Add support for fw live patch event Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 13/14] net/mlx5: Add support for devlink reload action fw activate no reset Moshe Shemesh
2020-08-30 15:27 ` [PATCH net-next RFC v3 14/14] devlink: Add Documentation/networking/devlink/devlink-reload.rst Moshe Shemesh
2020-08-31 10:49 ` [PATCH net-next RFC v3 00/14] Add devlink reload action option Jiri Pirko
2020-09-01 20:05 ` Moshe Shemesh
[not found] ` <36e30108-26e3-44ae-e133-48d412f7efe6@nvidia.com>
2020-09-02 7:55 ` Jiri Pirko
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200902094627.GB2568@nanopsycho \
--to=jiri@resnulli.us \
--cc=davem@davemloft.net \
--cc=jiri@mellanox.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=moshe@mellanox.com \
--cc=moshe@nvidia.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox