Re: [PATCH net-next v2] rps: selective flow shedding during softnet overflow

From: Eric Dumazet <eric.dumazet@gmail.com>
To: Willem de Bruijn <willemb@google.com>
Cc: netdev@vger.kernel.org, davem@davemloft.net, stephen@networkplumber.org
Subject: Re: [PATCH net-next v2] rps: selective flow shedding during softnet overflow
Date: Mon, 22 Apr 2013 15:30:08 -0700	[thread overview]
Message-ID: <1366669808.16391.90.camel@edumazet-glaptop> (raw)
In-Reply-To: <1366663561-12305-1-git-send-email-willemb@google.com>

On Mon, 2013-04-22 at 16:46 -0400, Willem de Bruijn wrote:

> +		len = sizeof(*cur) + netdev_flow_limit_table_len;
[1]
> +		mutex_lock(&flow_limit_update_mutex);
[2]
> +		for_each_possible_cpu(i) {
> +			sd = &per_cpu(softnet_data, i);
> +			cur = rcu_dereference_protected(sd->flow_limit,
> +				      lockdep_is_held(flow_limit_update_mutex));
> +			if (cur && !cpumask_test_cpu(i, mask)) {
> +				RCU_INIT_POINTER(sd->flow_limit, NULL);
> +				synchronize_rcu();
> +				kfree(cur);
> +			} else if (!cur && cpumask_test_cpu(i, mask)) {
> +				cur = kzalloc(len, GFP_KERNEL);
> +				cur->num_buckets = netdev_flow_limit_table_len;
[3]

Its a bit tricky, but the value of netdev_flow_limit_table_len could
change between [1] and [3]

So you should read its value once, or protect the whole thing using
mutex_lock(&flow_limit_update_mutex) in sysctl code ( and move [1] after
[2])