[net-2.6 PATCH 1/2] ixgbe: fix tx queue index

[net-2.6 PATCH 1/2] ixgbe: fix tx queue index

[Jeff Kirsher]

From: Wu Fengguang <fengguang.wu@intel.com>

Don't do the num_tx_queues based masking on calculating tx queue
index.

 1) num_tx_queues is not always power-of-2, because it also depends on
    the online cpu numbers. So the masking could be a performance bug
    on a 6 cpu system.
 2) queue_mapping will be limited by real_num_tx_queues=num_tx_queues
    in the generic netdev function set_cur_queue_map(). So the bound
    limiting here is not necessary.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---

 drivers/net/ixgbe/ixgbe_main.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 9ef128a..862dd34 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -4342,7 +4342,7 @@ static int ixgbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 	int count = 0;
 	unsigned int f;
 
-	r_idx = (adapter->num_tx_queues - 1) & skb->queue_mapping;
+	r_idx = skb->queue_mapping;
 	tx_ring = &adapter->tx_ring[r_idx];
 
 	if (adapter->vlgrp && vlan_tx_tag_present(skb)) {

[net-2.6 PATCH 2/2] ixgbe: update real_num_tx_queues on changing num_rx_queues

[Jeff Kirsher]

From: Wu Fengguang <fengguang.wu@intel.com>

Move the update of real_num_tx_queues from
ixgbe_acquire_msix_vectors() to ixgbe_set_num_queues(), to ensure it
be always in sync with num_tx_queues.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---

 drivers/net/ixgbe/ixgbe_main.c |   19 ++++++++++---------
 1 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 862dd34..11fd153 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -2723,17 +2723,21 @@ static inline bool ixgbe_set_rss_queues(struct ixgbe_adapter *adapter)
  **/
 static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
 {
-	/* Start with base case */
-	adapter->num_rx_queues = 1;
-	adapter->num_tx_queues = 1;
-
 #ifdef CONFIG_IXGBE_DCB
 	if (ixgbe_set_dcb_queues(adapter))
-		return;
+		goto done;
 
 #endif
 	if (ixgbe_set_rss_queues(adapter))
-		return;
+		goto done;
+
+	/* fallback to base case */
+	adapter->num_rx_queues = 1;
+	adapter->num_tx_queues = 1;
+
+done:
+	/* Notify the stack of the (possibly) reduced Tx Queue count. */
+	adapter->netdev->real_num_tx_queues = adapter->num_tx_queues;
 }
 
 static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter,
@@ -2992,9 +2996,6 @@ try_msi:
 	}
 
 out:
-	/* Notify the stack of the (possibly) reduced Tx Queue count. */
-	adapter->netdev->real_num_tx_queues = adapter->num_tx_queues;
-
 	return err;
 }
 

Re: [net-2.6 PATCH 1/2] ixgbe: fix tx queue index

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 14 Apr 2009 20:51:58 -0700

> Don't do the num_tx_queues based masking on calculating tx queue
> index.
> 
>  1) num_tx_queues is not always power-of-2, because it also depends on
>     the online cpu numbers. So the masking could be a performance bug
>     on a 6 cpu system.
>  2) queue_mapping will be limited by real_num_tx_queues=num_tx_queues
>     in the generic netdev function set_cur_queue_map(). So the bound
>     limiting here is not necessary.
> 
> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Applied.

Re: [net-2.6 PATCH 2/2] ixgbe: update real_num_tx_queues on changing num_rx_queues

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 14 Apr 2009 20:52:17 -0700

> Move the update of real_num_tx_queues from
> ixgbe_acquire_msix_vectors() to ixgbe_set_num_queues(), to ensure it
> be always in sync with num_tx_queues.
> 
> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Applied.