* [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount
@ 2009-11-02 5:45 Eric Dumazet
2009-11-02 7:57 ` David Miller
2009-11-02 21:14 ` Jarek Poplawski
0 siblings, 2 replies; 5+ messages in thread
From: Eric Dumazet @ 2009-11-02 5:45 UTC (permalink / raw)
To: David S. Miller, Linux Netdev List
Avoids touching dev refcount in hotpath
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
drivers/net/ifb.c | 6 ++++--
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
index 030913f..69c2566 100644
--- a/drivers/net/ifb.c
+++ b/drivers/net/ifb.c
@@ -98,13 +98,15 @@ static void ri_tasklet(unsigned long dev)
stats->tx_packets++;
stats->tx_bytes +=skb->len;
- skb->dev = dev_get_by_index(&init_net, skb->iif);
+ rcu_read_lock();
+ skb->dev = dev_get_by_index_rcu(&init_net, skb->iif);
if (!skb->dev) {
+ rcu_read_unlock();
dev_kfree_skb(skb);
stats->tx_dropped++;
break;
}
- dev_put(skb->dev);
+ rcu_read_unlock();
skb->iif = _dev->ifindex;
if (from & AT_EGRESS) {
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount
2009-11-02 5:45 [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount Eric Dumazet
@ 2009-11-02 7:57 ` David Miller
2009-11-02 21:14 ` Jarek Poplawski
1 sibling, 0 replies; 5+ messages in thread
From: David Miller @ 2009-11-02 7:57 UTC (permalink / raw)
To: eric.dumazet; +Cc: netdev
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Mon, 02 Nov 2009 06:45:16 +0100
> Avoids touching dev refcount in hotpath
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Applied.
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount
2009-11-02 5:45 [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount Eric Dumazet
2009-11-02 7:57 ` David Miller
@ 2009-11-02 21:14 ` Jarek Poplawski
2009-11-02 21:34 ` Eric Dumazet
1 sibling, 1 reply; 5+ messages in thread
From: Jarek Poplawski @ 2009-11-02 21:14 UTC (permalink / raw)
To: Eric Dumazet; +Cc: David S. Miller, Linux Netdev List
Eric Dumazet wrote, On 11/02/2009 06:45 AM:
> Avoids touching dev refcount in hotpath
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> ---
> drivers/net/ifb.c | 6 ++++--
> 1 files changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
> index 030913f..69c2566 100644
> --- a/drivers/net/ifb.c
> +++ b/drivers/net/ifb.c
> @@ -98,13 +98,15 @@ static void ri_tasklet(unsigned long dev)
> stats->tx_packets++;
> stats->tx_bytes +=skb->len;
>
> - skb->dev = dev_get_by_index(&init_net, skb->iif);
> + rcu_read_lock();
> + skb->dev = dev_get_by_index_rcu(&init_net, skb->iif);
> if (!skb->dev) {
> + rcu_read_unlock();
> dev_kfree_skb(skb);
> stats->tx_dropped++;
> break;
> }
> - dev_put(skb->dev);
> + rcu_read_unlock();
I wonder if this rcu_read_unlock() isn't too early here. I know, it
functionally fully replaces the old method, but as a whole it looks
strange:
> rcu_read_lock();
> skb->dev = dev_get_by_index_rcu(&init_net, skb->iif);
> if (!skb->dev) {
> rcu_read_unlock();
> dev_kfree_skb(skb);
> stats->tx_dropped++;
> break;
> }
> rcu_read_unlock();
> skb->iif = _dev->ifindex;
>
> if (from & AT_EGRESS) {
> dp->st_rx_frm_egr++;
> dev_queue_xmit(skb);
> } else if (from & AT_INGRESS) {
> dp->st_rx_frm_ing++;
> skb_pull(skb, skb->dev->hard_header_len);
So, how is skb->dev protected here, above and below? It seems these
rcu read blocks need extending, don't they?
Jarek P.
> netif_rx(skb);
> } else
> BUG();
> }
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount
2009-11-02 21:14 ` Jarek Poplawski
@ 2009-11-02 21:34 ` Eric Dumazet
2009-11-02 21:40 ` Jarek Poplawski
0 siblings, 1 reply; 5+ messages in thread
From: Eric Dumazet @ 2009-11-02 21:34 UTC (permalink / raw)
To: Jarek Poplawski; +Cc: David S. Miller, Linux Netdev List
Jarek Poplawski a écrit :
> Eric Dumazet wrote, On 11/02/2009 06:45 AM:
>
>> Avoids touching dev refcount in hotpath
>>
>> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
>> ---
>> drivers/net/ifb.c | 6 ++++--
>> 1 files changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
>> index 030913f..69c2566 100644
>> --- a/drivers/net/ifb.c
>> +++ b/drivers/net/ifb.c
>> @@ -98,13 +98,15 @@ static void ri_tasklet(unsigned long dev)
>> stats->tx_packets++;
>> stats->tx_bytes +=skb->len;
>>
>> - skb->dev = dev_get_by_index(&init_net, skb->iif);
>> + rcu_read_lock();
>> + skb->dev = dev_get_by_index_rcu(&init_net, skb->iif);
>> if (!skb->dev) {
>> + rcu_read_unlock();
>> dev_kfree_skb(skb);
>> stats->tx_dropped++;
>> break;
>> }
>> - dev_put(skb->dev);
>> + rcu_read_unlock();
>
> I wonder if this rcu_read_unlock() isn't too early here. I know, it
> functionally fully replaces the old method, but as a whole it looks
> strange:
>
>> rcu_read_lock();
>> skb->dev = dev_get_by_index_rcu(&init_net, skb->iif);
>> if (!skb->dev) {
>> rcu_read_unlock();
>> dev_kfree_skb(skb);
>> stats->tx_dropped++;
>> break;
>> }
>> rcu_read_unlock();
>> skb->iif = _dev->ifindex;
>>
>> if (from & AT_EGRESS) {
>> dp->st_rx_frm_egr++;
>> dev_queue_xmit(skb);
>> } else if (from & AT_INGRESS) {
>> dp->st_rx_frm_ing++;
>> skb_pull(skb, skb->dev->hard_header_len);
>
>
> So, how is skb->dev protected here, above and below? It seems these
> rcu read blocks need extending, don't they?
>
Well, this might be true, but we run under tasklet (softirq) with preemption disabled.
We might move rcu_read_unlock() some lines down to not rely on this.
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount
2009-11-02 21:34 ` Eric Dumazet
@ 2009-11-02 21:40 ` Jarek Poplawski
0 siblings, 0 replies; 5+ messages in thread
From: Jarek Poplawski @ 2009-11-02 21:40 UTC (permalink / raw)
To: Eric Dumazet; +Cc: David S. Miller, Linux Netdev List
On Mon, Nov 02, 2009 at 10:34:04PM +0100, Eric Dumazet wrote:
> Jarek Poplawski a écrit :
...
> > So, how is skb->dev protected here, above and below? It seems these
> > rcu read blocks need extending, don't they?
> >
>
> Well, this might be true, but we run under tasklet (softirq) with preemption disabled.
>
> We might move rcu_read_unlock() some lines down to not rely on this.
I think it's needed now at least for readability.
Jarek P.
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2009-11-02 21:41 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-11-02 5:45 [PATCH net-next-2.6] ifb: RCU locking avoids touching dev refcount Eric Dumazet
2009-11-02 7:57 ` David Miller
2009-11-02 21:14 ` Jarek Poplawski
2009-11-02 21:34 ` Eric Dumazet
2009-11-02 21:40 ` Jarek Poplawski
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.