Patchwork net: fix multiqueue selection

login
register
mail settings
Submitter Eric Dumazet
Date Sept. 7, 2013, 7:02 p.m.
Message ID <1378580577.26319.26.camel@edumazet-glaptop>
Download mbox | patch
Permalink /patch/273387/
State Accepted
Delegated to: David Miller
Headers show

Comments

Eric Dumazet - Sept. 7, 2013, 7:02 p.m.
From: Eric Dumazet <edumazet@google.com>

commit 416186fbf8c5b4e4465 ("net: Split core bits of netdev_pick_tx
into __netdev_pick_tx") added a bug that disables caching of queue
index in the socket.

This is the source of packet reorders for TCP flows, and
again this is happening more often when using FQ pacing.

Old code was doing 

if (queue_index != old_index)
	sk_tx_queue_set(sk, queue_index);

Alexander renamed the variables but forgot to change sk_tx_queue_set()
2nd parameter.

if (queue_index != new_index)
	sk_tx_queue_set(sk, queue_index);

This means we store -1 over and over in sk->sk_tx_queue_mapping

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
---
 net/core/flow_dissector.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Alexander Duyck - Sept. 8, 2013, 4:08 a.m.
On 09/07/2013 12:02 PM, Eric Dumazet wrote:
> From: Eric Dumazet <edumazet@google.com>
> 
> commit 416186fbf8c5b4e4465 ("net: Split core bits of netdev_pick_tx
> into __netdev_pick_tx") added a bug that disables caching of queue
> index in the socket.
> 
> This is the source of packet reorders for TCP flows, and
> again this is happening more often when using FQ pacing.
> 
> Old code was doing 
> 
> if (queue_index != old_index)
> 	sk_tx_queue_set(sk, queue_index);
> 
> Alexander renamed the variables but forgot to change sk_tx_queue_set()
> 2nd parameter.
> 
> if (queue_index != new_index)
> 	sk_tx_queue_set(sk, queue_index);
> 
> This means we store -1 over and over in sk->sk_tx_queue_mapping
> 
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Alexander Duyck <alexander.h.duyck@intel.com>
> ---
>  net/core/flow_dissector.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
> index 0ff42f0..1929af8 100644
> --- a/net/core/flow_dissector.c
> +++ b/net/core/flow_dissector.c
> @@ -352,7 +352,7 @@ u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb)
>  
>  		if (queue_index != new_index && sk &&
>  		    rcu_access_pointer(sk->sk_dst_cache))
> -			sk_tx_queue_set(sk, queue_index);
> +			sk_tx_queue_set(sk, new_index);
>  
>  		queue_index = new_index;
>  	}
> 
> 
> --

Ugh, my bad.  This is a nasty one too since the behaviour appeared to be
correct for most cases.

It looks like this needs to go into stable for 3.9 - 3.11.

Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - Sept. 11, 2013, 8:13 p.m.
From: Alexander Duyck <alexander.duyck@gmail.com>
Date: Sat, 07 Sep 2013 21:08:59 -0700

> On 09/07/2013 12:02 PM, Eric Dumazet wrote:
>> From: Eric Dumazet <edumazet@google.com>
>> 
>> commit 416186fbf8c5b4e4465 ("net: Split core bits of netdev_pick_tx
>> into __netdev_pick_tx") added a bug that disables caching of queue
>> index in the socket.
>> 
>> This is the source of packet reorders for TCP flows, and
>> again this is happening more often when using FQ pacing.
>> 
>> Old code was doing 
>> 
>> if (queue_index != old_index)
>> 	sk_tx_queue_set(sk, queue_index);
>> 
>> Alexander renamed the variables but forgot to change sk_tx_queue_set()
>> 2nd parameter.
>> 
>> if (queue_index != new_index)
>> 	sk_tx_queue_set(sk, queue_index);
>> 
>> This means we store -1 over and over in sk->sk_tx_queue_mapping
>> 
>> Signed-off-by: Eric Dumazet <edumazet@google.com>
 ...
> Ugh, my bad.  This is a nasty one too since the behaviour appeared to be
> correct for most cases.
> 
> It looks like this needs to go into stable for 3.9 - 3.11.
> 
> Acked-by: Alexander Duyck <alexander.h.duyck@intel.com>
> 

Applied and queued up for -stable, thanks.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index 0ff42f0..1929af8 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -352,7 +352,7 @@  u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb)
 
 		if (queue_index != new_index && sk &&
 		    rcu_access_pointer(sk->sk_dst_cache))
-			sk_tx_queue_set(sk, queue_index);
+			sk_tx_queue_set(sk, new_index);
 
 		queue_index = new_index;
 	}