Re: [PATCH net-next v3 1/5] devlink: Reduce struct devlink exposure

From: Leon Romanovsky <leon@kernel.org>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: "David S . Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Ido Schimmel <idosch@nvidia.com>, Ingo Molnar <mingo@redhat.com>,
	Jiri Pirko <jiri@nvidia.com>,
	linux-kernel@vger.kernel.org, linux-rdma@vger.kernel.org,
	mlxsw@nvidia.com, Moshe Shemesh <moshe@nvidia.com>,
	netdev@vger.kernel.org, Saeed Mahameed <saeedm@nvidia.com>,
	Salil Mehta <salil.mehta@huawei.com>,
	Shay Drory <shayd@nvidia.com>, Tariq Toukan <tariqt@nvidia.com>,
	Yisen Zhuang <yisen.zhuang@huawei.com>
Subject: Re: [PATCH net-next v3 1/5] devlink: Reduce struct devlink exposure
Date: Fri, 8 Oct 2021 03:44:48 +0300	[thread overview]
Message-ID: <YV+UgAEYz2GH8LIW@unreal> (raw)
In-Reply-To: <20211007155800.1ff26948@gandalf.local.home>

On Thu, Oct 07, 2021 at 03:58:00PM -0400, Steven Rostedt wrote:
> On Thu,  7 Oct 2021 09:55:15 +0300
> Leon Romanovsky <leon@kernel.org> wrote:
> 
> > +void *devlink_priv(struct devlink *devlink)
> > +{
> > +	BUG_ON(!devlink);
> 
> Do we really want to bring down the kernel in this case?

It was before.

> 
> Can't we just have:
> 
> 	if (WARN_ON(!devlink))
> 		return NULL;
> ?

Callers of devlink_priv() are not prepared to have NULL here, they don't
check return value at all,and this BUG_ON() can't happen at all.

> 
> Same for the below as well.

I can send followup patch.

Thanks

> 
> -- Steve
> 
> > +	return &devlink->priv;
> > +}
> > +EXPORT_SYMBOL_GPL(devlink_priv);
> > +
> > +struct devlink *priv_to_devlink(void *priv)
> > +{
> > +	BUG_ON(!priv);
> > +	return container_of(priv, struct devlink, priv);
> > +}
> > +EXPORT_SYMBOL_GPL(priv_to_devlink);
> > +
> > +struct device *devlink_to_dev(const struct devlink *devlink)
> > +{
> > +	return devlink->dev;
> > +}
> > +EXPORT_SYMBOL_GPL(devlink_to_dev);
> > +