Re: [PATCH net-next RFC 09/13] devlink: Add enable_remote_dev_reset generic parameter

[Moshe Shemesh]

On 7/29/2020 11:57 PM, Jakub Kicinski wrote:
> On Wed, 29 Jul 2020 17:42:12 +0300 Moshe Shemesh wrote:
> > On 7/28/2020 3:59 AM, Jakub Kicinski wrote:
> > > On Mon, 27 Jul 2020 14:02:29 +0300 Moshe Shemesh wrote:
> > > > The enable_remote_dev_reset devlink param flags that the host admin
> > > > allows device resets that can be initiated by other hosts. This
> > > > parameter is useful for setups where a device is shared by different
> > > > hosts, such as multi-host setup. Once the user set this parameter to
> > > > false, the driver should NACK any attempt to reset the device while the
> > > > driver is loaded.
> > > >
> > > > Signed-off-by: Moshe Shemesh <moshe@xxxxxxxxxxxx>
> > > There needs to be a devlink event generated when reset is triggered
> > > (remotely or not).
> > >
> > > You're also missing failure reasons. Users need to know if the reset
> > > request was clearly nacked by some host, not supported, etc. vs
> > > unexpected failure.
> > I will fix and send extack message to the user accordingly.
> I'd suggest the reason codes to be somewhat standard.
>
> The groups I can think of:
>   - timeout - device did not respond to the reset request
>   - device reject - FW or else has nacked the activation req
>   - host incapable - one of the participating hosts (in MH) is not
>     capable of handling live activation
>   - host denied - one of the participating hosts has NACKed
>   - host timeout - one of the p. hosts did not ack or done the procedure
>     in time (e.g. has not toggled the link)
>   - failed reset - the activation itself had failed
>   - failed reinit - one of p. hosts was not able to cleanly come back up


Sounds good, that seems to cover all options of fw_reset process to fail.