Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Madhuparna Bhowmik <madhuparnabhowmik10@gmail.com>
Date:   Fri, 21 Feb 2020 23:05:34 +0530
To:     Jiri Pirko <jiri@resnulli.us>
Cc:     madhuparnabhowmik10@gmail.com, jiri@mellanox.com,
        davem@davemloft.net, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, joel@joelfernandes.org,
        frextrite@gmail.com,
        linux-kernel-mentees@lists.linuxfoundation.org, paulmck@kernel.org
Subject: Re: [PATCH] net: core: devlink.c: Use built-in RCU list checking
Message-ID: <20200221173533.GA13198@madhuparna-HP-Notebook>
References: <20200221165141.24630-1-madhuparnabhowmik10@gmail.com>
 <20200221172008.GA2181@nanopsycho>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200221172008.GA2181@nanopsycho>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Feb 21, 2020 at 06:20:08PM +0100, Jiri Pirko wrote:
> Fri, Feb 21, 2020 at 05:51:41PM CET, madhuparnabhowmik10@gmail.com wrote:
> >From: Madhuparna Bhowmik <madhuparnabhowmik10@gmail.com>
> >
> >list_for_each_entry_rcu() has built-in RCU and lock checking.
> >
> >Pass cond argument to list_for_each_entry_rcu() to silence
> >false lockdep warning when CONFIG_PROVE_RCU_LIST is enabled
> >by default.
> >
> >Signed-off-by: Madhuparna Bhowmik <madhuparnabhowmik10@gmail.com>
> 
> Reviewed-by: Jiri Pirko <jiri@mellanox.com>
> 
> Thanks.
> 
> However, there is a callpath where not devlink lock neither rcu read is
> taken:
> devlink_dpipe_table_register()->devlink_dpipe_table_find()
>
Hi,

Yes I had noticed this, but I was not sure if there is some other lock
which is being used.

If yes, then can you please tell me which lock is held in this case,
and I can add that condition as well to list_for_each_entry_rcu() usage.

And if no lock or rcu_read_lock is held then may be we should
use rcu_read_lock/unlock here.

Let me know what you think about this.

Thank you,
Madhuparna

> I guess that was not the trace you were seeing, right?
> 
> 
> >---
> > net/core/devlink.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> >diff --git a/net/core/devlink.c b/net/core/devlink.c
> >index 4c63c9a4c09e..3e8c94155d93 100644
> >--- a/net/core/devlink.c
> >+++ b/net/core/devlink.c
> >@@ -2107,7 +2107,8 @@ devlink_dpipe_table_find(struct list_head *dpipe_tables,
> > {
> > 	struct devlink_dpipe_table *table;
> > 
> >-	list_for_each_entry_rcu(table, dpipe_tables, list) {
> >+	list_for_each_entry_rcu(table, dpipe_tables, list,
> >+				lockdep_is_held(&devlink->lock)) {
> > 		if (!strcmp(table->name, table_name))
> > 			return table;
> > 	}
> >-- 
> >2.17.1
> >