diff mbox

[net-next-2.6,2/5] ixgbe: Make descriptor ring allocations NUMA-aware

Message ID 20100107044845.28605.42794.stgit@localhost.localdomain
State Changes Requested, archived
Delegated to: David Miller
Headers show

Commit Message

Kirsher, Jeffrey T Jan. 7, 2010, 4:48 a.m. UTC 
From: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com>

This patch allocates the ring structures themselves on each
NUMA node along with the buffer_info structures.  This way we
don't allocate the entire ring memory on a single node in one
big block, thus reducing NUMA node memory crosstalk.

Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---

 drivers/net/ixgbe/ixgbe.h         |    7 +
 drivers/net/ixgbe/ixgbe_ethtool.c |   71 +++++-----
 drivers/net/ixgbe/ixgbe_fcoe.c    |    4 -
 drivers/net/ixgbe/ixgbe_main.c    |  270 +++++++++++++++++++------------------
 4 files changed, 185 insertions(+), 167 deletions(-)


--
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

Comments

David Miller Jan. 8, 2010, 8:21 a.m. UTC | #1 
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Wed, 06 Jan 2010 20:48:46 -0800

> @@ -147,7 +147,7 @@ struct ixgbe_ring {
>  
>  #ifdef CONFIG_IXGBE_DCA
>  	/* cpu for tx queue */
> -	int cpu;
> +	u8 cpu;
>  #endif
>  
>  	u16 work_limit;			/* max work per interrupt */

Is truncating cpu and node numbers to 8-bits ok?  I really don't
see how it can be fine, even for DCA.

This is especially the case since dca3_get_tag() and the
DCA ->get_tag() callback explicitly take an 'int' argument
too.
--
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
Waskiewicz Jr, Peter P Jan. 8, 2010, 8:25 a.m. UTC | #2 
On Fri, 8 Jan 2010, David Miller wrote:

> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> Date: Wed, 06 Jan 2010 20:48:46 -0800
> 
> > @@ -147,7 +147,7 @@ struct ixgbe_ring {
> >  
> >  #ifdef CONFIG_IXGBE_DCA
> >  	/* cpu for tx queue */
> > -	int cpu;
> > +	u8 cpu;
> >  #endif
> >  
> >  	u16 work_limit;			/* max work per interrupt */
> 
> Is truncating cpu and node numbers to 8-bits ok?  I really don't
> see how it can be fine, even for DCA.

In our hardware (82598 and 82599), the CPU field in the DCA registers is 8 
bits.

The reason I truncated this was to fit these values into the 
first cacheline of the struct.  I also didn't figure we'd see a system 
that would have 255 NUMA nodes before 10 GbE was something that was on the 
shelf collecting dust.  :-)

> This is especially the case since dca3_get_tag() and the
> DCA ->get_tag() callback explicitly take an 'int' argument
> too.

Cheers,
-PJ
--
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 Jan. 8, 2010, 8:31 a.m. UTC | #3 
From: "Waskiewicz Jr, Peter P" <peter.p.waskiewicz.jr@intel.com>
Date: Fri, 8 Jan 2010 00:25:26 -0800 (Pacific Standard Time)

> In our hardware (82598 and 82599), the CPU field in the DCA registers is 8 
> bits.
> 
> The reason I truncated this was to fit these values into the 
> first cacheline of the struct.  I also didn't figure we'd see a system 
> that would have 255 NUMA nodes before 10 GbE was something that was on the 
> shelf collecting dust.  :-)

The X86 port supports 9 for "NODES_SHIFT", and up to 4096 cpus.

Even when !MAXSMP, the cpu limit is 512.

Don't make the type ungeneric just because some current piece of
hardware has that limitation.  This is how subtle bugs slip into the
tree, and it makes auditing for such bugs a nightmare.

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
Waskiewicz Jr, Peter P Jan. 8, 2010, 8:36 a.m. UTC | #4 
On Fri, 8 Jan 2010, David Miller wrote:

> From: "Waskiewicz Jr, Peter P" <peter.p.waskiewicz.jr@intel.com>
> Date: Fri, 8 Jan 2010 00:25:26 -0800 (Pacific Standard Time)
> 
> > In our hardware (82598 and 82599), the CPU field in the DCA registers is 8 
> > bits.
> > 
> > The reason I truncated this was to fit these values into the 
> > first cacheline of the struct.  I also didn't figure we'd see a system 
> > that would have 255 NUMA nodes before 10 GbE was something that was on the 
> > shelf collecting dust.  :-)
> 
> The X86 port supports 9 for "NODES_SHIFT", and up to 4096 cpus.
> 
> Even when !MAXSMP, the cpu limit is 512.
> 
> Don't make the type ungeneric just because some current piece of
> hardware has that limitation.  This is how subtle bugs slip into the
> tree, and it makes auditing for such bugs a nightmare.
> 
> Thanks.

Understood.  Let me see if I can shuffle things around in the struct to 
keep my cacheline packed, and not screw with the data type sizes.  I'll 
respin and have Jeff push a new patch once I get it sorted out.

Thanks for the review Dave,

-PJ
--
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
diff mbox

Patch

diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 998b8d9..5e60358 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -147,7 +147,7 @@  struct ixgbe_ring {
 
 #ifdef CONFIG_IXGBE_DCA
 	/* cpu for tx queue */
-	int cpu;
+	u8 cpu;
 #endif
 
 	u16 work_limit;			/* max work per interrupt */
@@ -156,6 +156,7 @@  struct ixgbe_ring {
 					 * associated with this ring, which is
 					 * different for DCB and RSS modes
 					 */
+	u8 numa_node;
 
 	struct ixgbe_queue_stats stats;
 	unsigned long reinit_state;
@@ -277,7 +278,7 @@  struct ixgbe_adapter {
 	u16 eitr_high;
 
 	/* TX */
-	struct ixgbe_ring *tx_ring ____cacheline_aligned_in_smp; /* One per active queue */
+	struct ixgbe_ring *tx_ring[MAX_TX_QUEUES] ____cacheline_aligned_in_smp;
 	int num_tx_queues;
 	u32 tx_timeout_count;
 	bool detect_tx_hung;
@@ -286,7 +287,7 @@  struct ixgbe_adapter {
 	u64 lsc_int;
 
 	/* RX */
-	struct ixgbe_ring *rx_ring ____cacheline_aligned_in_smp; /* One per active queue */
+	struct ixgbe_ring *rx_ring[MAX_RX_QUEUES] ____cacheline_aligned_in_smp;
 	int num_rx_queues;
 	u64 hw_csum_rx_error;
 	u64 hw_rx_no_dma_resources;
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index 0bd49d3..d468c03 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -834,8 +834,8 @@  static void ixgbe_get_ringparam(struct net_device *netdev,
                                 struct ethtool_ringparam *ring)
 {
 	struct ixgbe_adapter *adapter = netdev_priv(netdev);
-	struct ixgbe_ring *tx_ring = adapter->tx_ring;
-	struct ixgbe_ring *rx_ring = adapter->rx_ring;
+	struct ixgbe_ring *tx_ring = adapter->tx_ring[0];
+	struct ixgbe_ring *rx_ring = adapter->rx_ring[0];
 
 	ring->rx_max_pending = IXGBE_MAX_RXD;
 	ring->tx_max_pending = IXGBE_MAX_TXD;
@@ -867,8 +867,8 @@  static int ixgbe_set_ringparam(struct net_device *netdev,
 	new_tx_count = min(new_tx_count, (u32)IXGBE_MAX_TXD);
 	new_tx_count = ALIGN(new_tx_count, IXGBE_REQ_TX_DESCRIPTOR_MULTIPLE);
 
-	if ((new_tx_count == adapter->tx_ring->count) &&
-	    (new_rx_count == adapter->rx_ring->count)) {
+	if ((new_tx_count == adapter->tx_ring[0]->count) &&
+	    (new_rx_count == adapter->rx_ring[0]->count)) {
 		/* nothing to do */
 		return 0;
 	}
@@ -878,25 +878,24 @@  static int ixgbe_set_ringparam(struct net_device *netdev,
 
 	if (!netif_running(adapter->netdev)) {
 		for (i = 0; i < adapter->num_tx_queues; i++)
-			adapter->tx_ring[i].count = new_tx_count;
+			adapter->tx_ring[i]->count = new_tx_count;
 		for (i = 0; i < adapter->num_rx_queues; i++)
-			adapter->rx_ring[i].count = new_rx_count;
+			adapter->rx_ring[i]->count = new_rx_count;
 		adapter->tx_ring_count = new_tx_count;
 		adapter->rx_ring_count = new_rx_count;
-		goto err_setup;
+		goto clear_reset;
 	}
 
-	temp_tx_ring = kcalloc(adapter->num_tx_queues,
-	                       sizeof(struct ixgbe_ring), GFP_KERNEL);
+	temp_tx_ring = vmalloc(adapter->num_tx_queues * sizeof(struct ixgbe_ring));
 	if (!temp_tx_ring) {
 		err = -ENOMEM;
-		goto err_setup;
+		goto clear_reset;
 	}
 
 	if (new_tx_count != adapter->tx_ring_count) {
-		memcpy(temp_tx_ring, adapter->tx_ring,
-		       adapter->num_tx_queues * sizeof(struct ixgbe_ring));
 		for (i = 0; i < adapter->num_tx_queues; i++) {
+			memcpy(&temp_tx_ring[i], adapter->tx_ring[i],
+			       sizeof(struct ixgbe_ring));
 			temp_tx_ring[i].count = new_tx_count;
 			err = ixgbe_setup_tx_resources(adapter,
 			                               &temp_tx_ring[i]);
@@ -904,28 +903,24 @@  static int ixgbe_set_ringparam(struct net_device *netdev,
 				while (i) {
 					i--;
 					ixgbe_free_tx_resources(adapter,
-					                        &temp_tx_ring[i]);
+					                      &temp_tx_ring[i]);
 				}
-				goto err_setup;
+				goto clear_reset;
 			}
 		}
 		need_update = true;
 	}
 
-	temp_rx_ring = kcalloc(adapter->num_rx_queues,
-	                       sizeof(struct ixgbe_ring), GFP_KERNEL);
-	if ((!temp_rx_ring) && (need_update)) {
-		for (i = 0; i < adapter->num_tx_queues; i++)
-			ixgbe_free_tx_resources(adapter, &temp_tx_ring[i]);
-		kfree(temp_tx_ring);
+	temp_rx_ring = vmalloc(adapter->num_rx_queues * sizeof(struct ixgbe_ring));
+	if (!temp_rx_ring) {
 		err = -ENOMEM;
 		goto err_setup;
 	}
 
 	if (new_rx_count != adapter->rx_ring_count) {
-		memcpy(temp_rx_ring, adapter->rx_ring,
-		       adapter->num_rx_queues * sizeof(struct ixgbe_ring));
 		for (i = 0; i < adapter->num_rx_queues; i++) {
+			memcpy(&temp_rx_ring[i], adapter->rx_ring[i],
+			       sizeof(struct ixgbe_ring));
 			temp_rx_ring[i].count = new_rx_count;
 			err = ixgbe_setup_rx_resources(adapter,
 			                               &temp_rx_ring[i]);
@@ -947,22 +942,32 @@  static int ixgbe_set_ringparam(struct net_device *netdev,
 
 		/* tx */
 		if (new_tx_count != adapter->tx_ring_count) {
-			kfree(adapter->tx_ring);
-			adapter->tx_ring = temp_tx_ring;
-			temp_tx_ring = NULL;
+			for (i = 0; i < adapter->num_tx_queues; i++) {
+				ixgbe_free_tx_resources(adapter,
+				                        adapter->tx_ring[i]);
+				memcpy(adapter->tx_ring[i], &temp_tx_ring[i],
+				       sizeof(struct ixgbe_ring));
+			}
 			adapter->tx_ring_count = new_tx_count;
 		}
 
 		/* rx */
 		if (new_rx_count != adapter->rx_ring_count) {
-			kfree(adapter->rx_ring);
-			adapter->rx_ring = temp_rx_ring;
-			temp_rx_ring = NULL;
+			for (i = 0; i < adapter->num_rx_queues; i++) {
+				ixgbe_free_rx_resources(adapter,
+				                        adapter->rx_ring[i]);
+				memcpy(adapter->rx_ring[i], &temp_rx_ring[i],
+				       sizeof(struct ixgbe_ring));
+			}
 			adapter->rx_ring_count = new_rx_count;
 		}
 		ixgbe_up(adapter);
 	}
+
+	vfree(temp_rx_ring);
 err_setup:
+	vfree(temp_tx_ring);
+clear_reset:
 	clear_bit(__IXGBE_RESETTING, &adapter->state);
 	return err;
 }
@@ -1007,13 +1012,13 @@  static void ixgbe_get_ethtool_stats(struct net_device *netdev,
 		           sizeof(u64)) ? *(u64 *)p : *(u32 *)p;
 	}
 	for (j = 0; j < adapter->num_tx_queues; j++) {
-		queue_stat = (u64 *)&adapter->tx_ring[j].stats;
+		queue_stat = (u64 *)&adapter->tx_ring[j]->stats;
 		for (k = 0; k < stat_count; k++)
 			data[i + k] = queue_stat[k];
 		i += k;
 	}
 	for (j = 0; j < adapter->num_rx_queues; j++) {
-		queue_stat = (u64 *)&adapter->rx_ring[j].stats;
+		queue_stat = (u64 *)&adapter->rx_ring[j]->stats;
 		for (k = 0; k < stat_count; k++)
 			data[i + k] = queue_stat[k];
 		i += k;
@@ -1627,7 +1632,7 @@  static int ixgbe_setup_desc_rings(struct ixgbe_adapter *adapter)
 	reg_data |= IXGBE_RXDCTL_ENABLE;
 	IXGBE_WRITE_REG(&adapter->hw, IXGBE_RXDCTL(0), reg_data);
 	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
-		int j = adapter->rx_ring[0].reg_idx;
+		int j = adapter->rx_ring[0]->reg_idx;
 		u32 k;
 		for (k = 0; k < 10; k++) {
 			if (IXGBE_READ_REG(&adapter->hw,
@@ -2000,7 +2005,7 @@  static int ixgbe_get_coalesce(struct net_device *netdev,
 {
 	struct ixgbe_adapter *adapter = netdev_priv(netdev);
 
-	ec->tx_max_coalesced_frames_irq = adapter->tx_ring[0].work_limit;
+	ec->tx_max_coalesced_frames_irq = adapter->tx_ring[0]->work_limit;
 
 	/* only valid if in constant ITR mode */
 	switch (adapter->rx_itr_setting) {
@@ -2053,7 +2058,7 @@  static int ixgbe_set_coalesce(struct net_device *netdev,
 		return -EINVAL;
 
 	if (ec->tx_max_coalesced_frames_irq)
-		adapter->tx_ring[0].work_limit = ec->tx_max_coalesced_frames_irq;
+		adapter->tx_ring[0]->work_limit = ec->tx_max_coalesced_frames_irq;
 
 	if (ec->rx_coalesce_usecs > 1) {
 		/* check the limits */
diff --git a/drivers/net/ixgbe/ixgbe_fcoe.c b/drivers/net/ixgbe/ixgbe_fcoe.c
index da32a10..ff09706 100644
--- a/drivers/net/ixgbe/ixgbe_fcoe.c
+++ b/drivers/net/ixgbe/ixgbe_fcoe.c
@@ -525,7 +525,7 @@  void ixgbe_configure_fcoe(struct ixgbe_adapter *adapter)
 		for (i = 0; i < IXGBE_FCRETA_SIZE; i++) {
 			fcoe_i = f->mask + i % f->indices;
 			fcoe_i &= IXGBE_FCRETA_ENTRY_MASK;
-			fcoe_q = adapter->rx_ring[fcoe_i].reg_idx;
+			fcoe_q = adapter->rx_ring[fcoe_i]->reg_idx;
 			IXGBE_WRITE_REG(hw, IXGBE_FCRETA(i), fcoe_q);
 		}
 		IXGBE_WRITE_REG(hw, IXGBE_FCRECTL, IXGBE_FCRECTL_ENA);
@@ -533,7 +533,7 @@  void ixgbe_configure_fcoe(struct ixgbe_adapter *adapter)
 	} else  {
 		/* Use single rx queue for FCoE */
 		fcoe_i = f->mask;
-		fcoe_q = adapter->rx_ring[fcoe_i].reg_idx;
+		fcoe_q = adapter->rx_ring[fcoe_i]->reg_idx;
 		IXGBE_WRITE_REG(hw, IXGBE_FCRECTL, 0);
 		IXGBE_WRITE_REG(hw, IXGBE_ETQS(IXGBE_ETQF_FILTER_FCOE),
 				IXGBE_ETQS_QUEUE_EN |
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 6895de7..ceeef52 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -451,7 +451,7 @@  static void ixgbe_update_rx_dca(struct ixgbe_adapter *adapter,
 {
 	u32 rxctrl;
 	int cpu = get_cpu();
-	int q = rx_ring - adapter->rx_ring;
+	int q = rx_ring->reg_idx;
 
 	if (rx_ring->cpu != cpu) {
 		rxctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_DCA_RXCTRL(q));
@@ -479,7 +479,7 @@  static void ixgbe_update_tx_dca(struct ixgbe_adapter *adapter,
 {
 	u32 txctrl;
 	int cpu = get_cpu();
-	int q = tx_ring - adapter->tx_ring;
+	int q = tx_ring->reg_idx;
 	struct ixgbe_hw *hw = &adapter->hw;
 
 	if (tx_ring->cpu != cpu) {
@@ -513,12 +513,12 @@  static void ixgbe_setup_dca(struct ixgbe_adapter *adapter)
 	IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_CTRL, 2);
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		adapter->tx_ring[i].cpu = -1;
-		ixgbe_update_tx_dca(adapter, &adapter->tx_ring[i]);
+		adapter->tx_ring[i]->cpu = -1;
+		ixgbe_update_tx_dca(adapter, adapter->tx_ring[i]);
 	}
 	for (i = 0; i < adapter->num_rx_queues; i++) {
-		adapter->rx_ring[i].cpu = -1;
-		ixgbe_update_rx_dca(adapter, &adapter->rx_ring[i]);
+		adapter->rx_ring[i]->cpu = -1;
+		ixgbe_update_rx_dca(adapter, adapter->rx_ring[i]);
 	}
 }
 
@@ -989,7 +989,7 @@  static void ixgbe_configure_msix(struct ixgbe_adapter *adapter)
 		                       adapter->num_rx_queues);
 
 		for (i = 0; i < q_vector->rxr_count; i++) {
-			j = adapter->rx_ring[r_idx].reg_idx;
+			j = adapter->rx_ring[r_idx]->reg_idx;
 			ixgbe_set_ivar(adapter, 0, j, v_idx);
 			r_idx = find_next_bit(q_vector->rxr_idx,
 			                      adapter->num_rx_queues,
@@ -999,7 +999,7 @@  static void ixgbe_configure_msix(struct ixgbe_adapter *adapter)
 		                       adapter->num_tx_queues);
 
 		for (i = 0; i < q_vector->txr_count; i++) {
-			j = adapter->tx_ring[r_idx].reg_idx;
+			j = adapter->tx_ring[r_idx]->reg_idx;
 			ixgbe_set_ivar(adapter, 1, j, v_idx);
 			r_idx = find_next_bit(q_vector->txr_idx,
 			                      adapter->num_tx_queues,
@@ -1134,7 +1134,7 @@  static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
 
 	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
 	for (i = 0; i < q_vector->txr_count; i++) {
-		tx_ring = &(adapter->tx_ring[r_idx]);
+		tx_ring = adapter->tx_ring[r_idx];
 		ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
 		                           q_vector->tx_itr,
 		                           tx_ring->total_packets,
@@ -1149,7 +1149,7 @@  static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
 
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
 	for (i = 0; i < q_vector->rxr_count; i++) {
-		rx_ring = &(adapter->rx_ring[r_idx]);
+		rx_ring = adapter->rx_ring[r_idx];
 		ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
 		                           q_vector->rx_itr,
 		                           rx_ring->total_packets,
@@ -1268,7 +1268,7 @@  static irqreturn_t ixgbe_msix_lsc(int irq, void *data)
 			netif_tx_stop_all_queues(netdev);
 			for (i = 0; i < adapter->num_tx_queues; i++) {
 				struct ixgbe_ring *tx_ring =
-				                           &adapter->tx_ring[i];
+				                            adapter->tx_ring[i];
 				if (test_and_clear_bit(__IXGBE_FDIR_INIT_DONE,
 				                       &tx_ring->reinit_state))
 					schedule_work(&adapter->fdir_reinit_task);
@@ -1327,7 +1327,7 @@  static irqreturn_t ixgbe_msix_clean_tx(int irq, void *data)
 
 	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
 	for (i = 0; i < q_vector->txr_count; i++) {
-		tx_ring = &(adapter->tx_ring[r_idx]);
+		tx_ring = adapter->tx_ring[r_idx];
 		tx_ring->total_bytes = 0;
 		tx_ring->total_packets = 0;
 		r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
@@ -1355,7 +1355,7 @@  static irqreturn_t ixgbe_msix_clean_rx(int irq, void *data)
 
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
 	for (i = 0;  i < q_vector->rxr_count; i++) {
-		rx_ring = &(adapter->rx_ring[r_idx]);
+		rx_ring = adapter->rx_ring[r_idx];
 		rx_ring->total_bytes = 0;
 		rx_ring->total_packets = 0;
 		r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
@@ -1385,7 +1385,7 @@  static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
 
 	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
 	for (i = 0; i < q_vector->txr_count; i++) {
-		ring = &(adapter->tx_ring[r_idx]);
+		ring = adapter->tx_ring[r_idx];
 		ring->total_bytes = 0;
 		ring->total_packets = 0;
 		r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
@@ -1394,7 +1394,7 @@  static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
 
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
 	for (i = 0; i < q_vector->rxr_count; i++) {
-		ring = &(adapter->rx_ring[r_idx]);
+		ring = adapter->rx_ring[r_idx];
 		ring->total_bytes = 0;
 		ring->total_packets = 0;
 		r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
@@ -1425,7 +1425,7 @@  static int ixgbe_clean_rxonly(struct napi_struct *napi, int budget)
 	long r_idx;
 
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
-	rx_ring = &(adapter->rx_ring[r_idx]);
+	rx_ring = adapter->rx_ring[r_idx];
 #ifdef CONFIG_IXGBE_DCA
 	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
 		ixgbe_update_rx_dca(adapter, rx_ring);
@@ -1466,7 +1466,7 @@  static int ixgbe_clean_rxtx_many(struct napi_struct *napi, int budget)
 
 	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
 	for (i = 0; i < q_vector->txr_count; i++) {
-		ring = &(adapter->tx_ring[r_idx]);
+		ring = adapter->tx_ring[r_idx];
 #ifdef CONFIG_IXGBE_DCA
 		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
 			ixgbe_update_tx_dca(adapter, ring);
@@ -1482,7 +1482,7 @@  static int ixgbe_clean_rxtx_many(struct napi_struct *napi, int budget)
 	budget = max(budget, 1);
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
 	for (i = 0; i < q_vector->rxr_count; i++) {
-		ring = &(adapter->rx_ring[r_idx]);
+		ring = adapter->rx_ring[r_idx];
 #ifdef CONFIG_IXGBE_DCA
 		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
 			ixgbe_update_rx_dca(adapter, ring);
@@ -1493,7 +1493,7 @@  static int ixgbe_clean_rxtx_many(struct napi_struct *napi, int budget)
 	}
 
 	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
-	ring = &(adapter->rx_ring[r_idx]);
+	ring = adapter->rx_ring[r_idx];
 	/* If all Rx work done, exit the polling mode */
 	if (work_done < budget) {
 		napi_complete(napi);
@@ -1526,7 +1526,7 @@  static int ixgbe_clean_txonly(struct napi_struct *napi, int budget)
 	long r_idx;
 
 	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
-	tx_ring = &(adapter->tx_ring[r_idx]);
+	tx_ring = adapter->tx_ring[r_idx];
 #ifdef CONFIG_IXGBE_DCA
 	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
 		ixgbe_update_tx_dca(adapter, tx_ring);
@@ -1711,8 +1711,8 @@  static void ixgbe_set_itr(struct ixgbe_adapter *adapter)
 	struct ixgbe_q_vector *q_vector = adapter->q_vector[0];
 	u8 current_itr;
 	u32 new_itr = q_vector->eitr;
-	struct ixgbe_ring *rx_ring = &adapter->rx_ring[0];
-	struct ixgbe_ring *tx_ring = &adapter->tx_ring[0];
+	struct ixgbe_ring *rx_ring = adapter->rx_ring[0];
+	struct ixgbe_ring *tx_ring = adapter->tx_ring[0];
 
 	q_vector->tx_itr = ixgbe_update_itr(adapter, new_itr,
 	                                    q_vector->tx_itr,
@@ -1817,10 +1817,10 @@  static irqreturn_t ixgbe_intr(int irq, void *data)
 	ixgbe_check_fan_failure(adapter, eicr);
 
 	if (napi_schedule_prep(&(q_vector->napi))) {
-		adapter->tx_ring[0].total_packets = 0;
-		adapter->tx_ring[0].total_bytes = 0;
-		adapter->rx_ring[0].total_packets = 0;
-		adapter->rx_ring[0].total_bytes = 0;
+		adapter->tx_ring[0]->total_packets = 0;
+		adapter->tx_ring[0]->total_bytes = 0;
+		adapter->rx_ring[0]->total_packets = 0;
+		adapter->rx_ring[0]->total_bytes = 0;
 		/* would disable interrupts here but EIAM disabled it */
 		__napi_schedule(&(q_vector->napi));
 	}
@@ -1950,7 +1950,7 @@  static void ixgbe_configure_tx(struct ixgbe_adapter *adapter)
 
 	/* Setup the HW Tx Head and Tail descriptor pointers */
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		struct ixgbe_ring *ring = &adapter->tx_ring[i];
+		struct ixgbe_ring *ring = adapter->tx_ring[i];
 		j = ring->reg_idx;
 		tdba = ring->dma;
 		tdlen = ring->count * sizeof(union ixgbe_adv_tx_desc);
@@ -1960,8 +1960,8 @@  static void ixgbe_configure_tx(struct ixgbe_adapter *adapter)
 		IXGBE_WRITE_REG(hw, IXGBE_TDLEN(j), tdlen);
 		IXGBE_WRITE_REG(hw, IXGBE_TDH(j), 0);
 		IXGBE_WRITE_REG(hw, IXGBE_TDT(j), 0);
-		adapter->tx_ring[i].head = IXGBE_TDH(j);
-		adapter->tx_ring[i].tail = IXGBE_TDT(j);
+		adapter->tx_ring[i]->head = IXGBE_TDH(j);
+		adapter->tx_ring[i]->tail = IXGBE_TDT(j);
 		/*
 		 * Disable Tx Head Writeback RO bit, since this hoses
 		 * bookkeeping if things aren't delivered in order.
@@ -2090,7 +2090,7 @@  static void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter, int index)
 	u32 rscctrl;
 	int rx_buf_len;
 
-	rx_ring = &adapter->rx_ring[index];
+	rx_ring = adapter->rx_ring[index];
 	j = rx_ring->reg_idx;
 	rx_buf_len = rx_ring->rx_buf_len;
 	rscctrl = IXGBE_READ_REG(hw, IXGBE_RSCCTL(j));
@@ -2184,7 +2184,7 @@  static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
 #endif
 	IXGBE_WRITE_REG(hw, IXGBE_HLREG0, hlreg0);
 
-	rdlen = adapter->rx_ring[0].count * sizeof(union ixgbe_adv_rx_desc);
+	rdlen = adapter->rx_ring[0]->count * sizeof(union ixgbe_adv_rx_desc);
 	/* disable receives while setting up the descriptors */
 	rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
 	IXGBE_WRITE_REG(hw, IXGBE_RXCTRL, rxctrl & ~IXGBE_RXCTRL_RXEN);
@@ -2194,7 +2194,7 @@  static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
 	 * the Base and Length of the Rx Descriptor Ring
 	 */
 	for (i = 0; i < adapter->num_rx_queues; i++) {
-		rx_ring = &adapter->rx_ring[i];
+		rx_ring = adapter->rx_ring[i];
 		rdba = rx_ring->dma;
 		j = rx_ring->reg_idx;
 		IXGBE_WRITE_REG(hw, IXGBE_RDBAL(j), (rdba & DMA_BIT_MASK(32)));
@@ -2361,7 +2361,7 @@  static void ixgbe_vlan_rx_register(struct net_device *netdev,
 	} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
 		for (i = 0; i < adapter->num_rx_queues; i++) {
 			u32 ctrl;
-			j = adapter->rx_ring[i].reg_idx;
+			j = adapter->rx_ring[i]->reg_idx;
 			ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_RXDCTL(j));
 			ctrl |= IXGBE_RXDCTL_VME;
 			IXGBE_WRITE_REG(&adapter->hw, IXGBE_RXDCTL(j), ctrl);
@@ -2522,7 +2522,7 @@  static void ixgbe_configure_dcb(struct ixgbe_adapter *adapter)
 	ixgbe_dcb_hw_config(&adapter->hw, &adapter->dcb_cfg);
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		j = adapter->tx_ring[i].reg_idx;
+		j = adapter->tx_ring[i]->reg_idx;
 		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
 		/* PThresh workaround for Tx hang with DFP enabled. */
 		txdctl |= 32;
@@ -2539,7 +2539,7 @@  static void ixgbe_configure_dcb(struct ixgbe_adapter *adapter)
 		vlnctrl &= ~IXGBE_VLNCTRL_CFIEN;
 		IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
 		for (i = 0; i < adapter->num_rx_queues; i++) {
-			j = adapter->rx_ring[i].reg_idx;
+			j = adapter->rx_ring[i]->reg_idx;
 			vlnctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
 			vlnctrl |= IXGBE_RXDCTL_VME;
 			IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(j), vlnctrl);
@@ -2579,7 +2579,7 @@  static void ixgbe_configure(struct ixgbe_adapter *adapter)
 #endif /* IXGBE_FCOE */
 	if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) {
 		for (i = 0; i < adapter->num_tx_queues; i++)
-			adapter->tx_ring[i].atr_sample_rate =
+			adapter->tx_ring[i]->atr_sample_rate =
 			                               adapter->atr_sample_rate;
 		ixgbe_init_fdir_signature_82599(hw, adapter->fdir_pballoc);
 	} else if (adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE) {
@@ -2589,8 +2589,8 @@  static void ixgbe_configure(struct ixgbe_adapter *adapter)
 	ixgbe_configure_tx(adapter);
 	ixgbe_configure_rx(adapter);
 	for (i = 0; i < adapter->num_rx_queues; i++)
-		ixgbe_alloc_rx_buffers(adapter, &adapter->rx_ring[i],
-		                       (adapter->rx_ring[i].count - 1));
+		ixgbe_alloc_rx_buffers(adapter, adapter->rx_ring[i],
+		                       (adapter->rx_ring[i]->count - 1));
 }
 
 static inline bool ixgbe_is_sfp(struct ixgbe_hw *hw)
@@ -2673,7 +2673,7 @@  link_cfg_out:
 static inline void ixgbe_rx_desc_queue_enable(struct ixgbe_adapter *adapter,
 	                                      int rxr)
 {
-	int j = adapter->rx_ring[rxr].reg_idx;
+	int j = adapter->rx_ring[rxr]->reg_idx;
 	int k;
 
 	for (k = 0; k < IXGBE_MAX_RX_DESC_POLL; k++) {
@@ -2687,8 +2687,8 @@  static inline void ixgbe_rx_desc_queue_enable(struct ixgbe_adapter *adapter,
 		DPRINTK(DRV, ERR, "RXDCTL.ENABLE on Rx queue %d "
 		        "not set within the polling period\n", rxr);
 	}
-	ixgbe_release_rx_desc(&adapter->hw, &adapter->rx_ring[rxr],
-	                      (adapter->rx_ring[rxr].count - 1));
+	ixgbe_release_rx_desc(&adapter->hw, adapter->rx_ring[rxr],
+	                      (adapter->rx_ring[rxr]->count - 1));
 }
 
 static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
@@ -2770,7 +2770,7 @@  static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
 	}
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		j = adapter->tx_ring[i].reg_idx;
+		j = adapter->tx_ring[i]->reg_idx;
 		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
 		/* enable WTHRESH=8 descriptors, to encourage burst writeback */
 		txdctl |= (8 << 16);
@@ -2784,14 +2784,14 @@  static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
 		IXGBE_WRITE_REG(hw, IXGBE_DMATXCTL, dmatxctl);
 	}
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		j = adapter->tx_ring[i].reg_idx;
+		j = adapter->tx_ring[i]->reg_idx;
 		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
 		txdctl |= IXGBE_TXDCTL_ENABLE;
 		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j), txdctl);
 	}
 
 	for (i = 0; i < num_rx_rings; i++) {
-		j = adapter->rx_ring[i].reg_idx;
+		j = adapter->rx_ring[i]->reg_idx;
 		rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
 		/* enable PTHRESH=32 descriptors (half the internal cache)
 		 * and HTHRESH=0 descriptors (to minimize latency on fetch),
@@ -2865,7 +2865,7 @@  static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
 
 	for (i = 0; i < adapter->num_tx_queues; i++)
 		set_bit(__IXGBE_FDIR_INIT_DONE,
-		        &(adapter->tx_ring[i].reinit_state));
+		        &(adapter->tx_ring[i]->reinit_state));
 
 	/* enable transmits */
 	netif_tx_start_all_queues(netdev);
@@ -3029,7 +3029,7 @@  static void ixgbe_clean_all_rx_rings(struct ixgbe_adapter *adapter)
 	int i;
 
 	for (i = 0; i < adapter->num_rx_queues; i++)
-		ixgbe_clean_rx_ring(adapter, &adapter->rx_ring[i]);
+		ixgbe_clean_rx_ring(adapter, adapter->rx_ring[i]);
 }
 
 /**
@@ -3041,7 +3041,7 @@  static void ixgbe_clean_all_tx_rings(struct ixgbe_adapter *adapter)
 	int i;
 
 	for (i = 0; i < adapter->num_tx_queues; i++)
-		ixgbe_clean_tx_ring(adapter, &adapter->tx_ring[i]);
+		ixgbe_clean_tx_ring(adapter, adapter->tx_ring[i]);
 }
 
 void ixgbe_down(struct ixgbe_adapter *adapter)
@@ -3081,7 +3081,7 @@  void ixgbe_down(struct ixgbe_adapter *adapter)
 
 	/* disable transmits in the hardware now that interrupts are off */
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		j = adapter->tx_ring[i].reg_idx;
+		j = adapter->tx_ring[i]->reg_idx;
 		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
 		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j),
 		                (txdctl & ~IXGBE_TXDCTL_ENABLE));
@@ -3121,13 +3121,13 @@  static int ixgbe_poll(struct napi_struct *napi, int budget)
 
 #ifdef CONFIG_IXGBE_DCA
 	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {
-		ixgbe_update_tx_dca(adapter, adapter->tx_ring);
-		ixgbe_update_rx_dca(adapter, adapter->rx_ring);
+		ixgbe_update_tx_dca(adapter, adapter->tx_ring[0]);
+		ixgbe_update_rx_dca(adapter, adapter->rx_ring[0]);
 	}
 #endif
 
-	tx_clean_complete = ixgbe_clean_tx_irq(q_vector, adapter->tx_ring);
-	ixgbe_clean_rx_irq(q_vector, adapter->rx_ring, &work_done, budget);
+	tx_clean_complete = ixgbe_clean_tx_irq(q_vector, adapter->tx_ring[0]);
+	ixgbe_clean_rx_irq(q_vector, adapter->rx_ring[0], &work_done, budget);
 
 	if (!tx_clean_complete)
 		work_done = budget;
@@ -3393,9 +3393,9 @@  static inline bool ixgbe_cache_ring_rss(struct ixgbe_adapter *adapter)
 
 	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) {
 		for (i = 0; i < adapter->num_rx_queues; i++)
-			adapter->rx_ring[i].reg_idx = i;
+			adapter->rx_ring[i]->reg_idx = i;
 		for (i = 0; i < adapter->num_tx_queues; i++)
-			adapter->tx_ring[i].reg_idx = i;
+			adapter->tx_ring[i]->reg_idx = i;
 		ret = true;
 	} else {
 		ret = false;
@@ -3422,8 +3422,8 @@  static inline bool ixgbe_cache_ring_dcb(struct ixgbe_adapter *adapter)
 		if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
 			/* the number of queues is assumed to be symmetric */
 			for (i = 0; i < dcb_i; i++) {
-				adapter->rx_ring[i].reg_idx = i << 3;
-				adapter->tx_ring[i].reg_idx = i << 2;
+				adapter->rx_ring[i]->reg_idx = i << 3;
+				adapter->tx_ring[i]->reg_idx = i << 2;
 			}
 			ret = true;
 		} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
@@ -3441,18 +3441,18 @@  static inline bool ixgbe_cache_ring_dcb(struct ixgbe_adapter *adapter)
 				 * Rx TC0-TC7 are offset by 16 queues each
 				 */
 				for (i = 0; i < 3; i++) {
-					adapter->tx_ring[i].reg_idx = i << 5;
-					adapter->rx_ring[i].reg_idx = i << 4;
+					adapter->tx_ring[i]->reg_idx = i << 5;
+					adapter->rx_ring[i]->reg_idx = i << 4;
 				}
 				for ( ; i < 5; i++) {
-					adapter->tx_ring[i].reg_idx =
+					adapter->tx_ring[i]->reg_idx =
 					                         ((i + 2) << 4);
-					adapter->rx_ring[i].reg_idx = i << 4;
+					adapter->rx_ring[i]->reg_idx = i << 4;
 				}
 				for ( ; i < dcb_i; i++) {
-					adapter->tx_ring[i].reg_idx =
+					adapter->tx_ring[i]->reg_idx =
 					                         ((i + 8) << 3);
-					adapter->rx_ring[i].reg_idx = i << 4;
+					adapter->rx_ring[i]->reg_idx = i << 4;
 				}
 
 				ret = true;
@@ -3465,12 +3465,12 @@  static inline bool ixgbe_cache_ring_dcb(struct ixgbe_adapter *adapter)
 				 *
 				 * Rx TC0-TC3 are offset by 32 queues each
 				 */
-				adapter->tx_ring[0].reg_idx = 0;
-				adapter->tx_ring[1].reg_idx = 64;
-				adapter->tx_ring[2].reg_idx = 96;
-				adapter->tx_ring[3].reg_idx = 112;
+				adapter->tx_ring[0]->reg_idx = 0;
+				adapter->tx_ring[1]->reg_idx = 64;
+				adapter->tx_ring[2]->reg_idx = 96;
+				adapter->tx_ring[3]->reg_idx = 112;
 				for (i = 0 ; i < dcb_i; i++)
-					adapter->rx_ring[i].reg_idx = i << 5;
+					adapter->rx_ring[i]->reg_idx = i << 5;
 
 				ret = true;
 			} else {
@@ -3503,9 +3503,9 @@  static bool inline ixgbe_cache_ring_fdir(struct ixgbe_adapter *adapter)
 	    ((adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) ||
 	     (adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE))) {
 		for (i = 0; i < adapter->num_rx_queues; i++)
-			adapter->rx_ring[i].reg_idx = i;
+			adapter->rx_ring[i]->reg_idx = i;
 		for (i = 0; i < adapter->num_tx_queues; i++)
-			adapter->tx_ring[i].reg_idx = i;
+			adapter->tx_ring[i]->reg_idx = i;
 		ret = true;
 	}
 
@@ -3533,8 +3533,8 @@  static inline bool ixgbe_cache_ring_fcoe(struct ixgbe_adapter *adapter)
 
 			ixgbe_cache_ring_dcb(adapter);
 			/* find out queues in TC for FCoE */
-			fcoe_rx_i = adapter->rx_ring[fcoe->tc].reg_idx + 1;
-			fcoe_tx_i = adapter->tx_ring[fcoe->tc].reg_idx + 1;
+			fcoe_rx_i = adapter->rx_ring[fcoe->tc]->reg_idx + 1;
+			fcoe_tx_i = adapter->tx_ring[fcoe->tc]->reg_idx + 1;
 			/*
 			 * In 82599, the number of Tx queues for each traffic
 			 * class for both 8-TC and 4-TC modes are:
@@ -3565,8 +3565,8 @@  static inline bool ixgbe_cache_ring_fcoe(struct ixgbe_adapter *adapter)
 			fcoe_tx_i = f->mask;
 		}
 		for (i = 0; i < f->indices; i++, fcoe_rx_i++, fcoe_tx_i++) {
-			adapter->rx_ring[f->mask + i].reg_idx = fcoe_rx_i;
-			adapter->tx_ring[f->mask + i].reg_idx = fcoe_tx_i;
+			adapter->rx_ring[f->mask + i]->reg_idx = fcoe_rx_i;
+			adapter->tx_ring[f->mask + i]->reg_idx = fcoe_tx_i;
 		}
 		ret = true;
 	}
@@ -3588,8 +3588,8 @@  static inline bool ixgbe_cache_ring_fcoe(struct ixgbe_adapter *adapter)
 static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
 {
 	/* start with default case */
-	adapter->rx_ring[0].reg_idx = 0;
-	adapter->tx_ring[0].reg_idx = 0;
+	adapter->rx_ring[0]->reg_idx = 0;
+	adapter->tx_ring[0]->reg_idx = 0;
 
 #ifdef IXGBE_FCOE
 	if (ixgbe_cache_ring_fcoe(adapter))
@@ -3619,33 +3619,59 @@  static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
 static int ixgbe_alloc_queues(struct ixgbe_adapter *adapter)
 {
 	int i;
-
-	adapter->tx_ring = kcalloc(adapter->num_tx_queues,
-	                           sizeof(struct ixgbe_ring), GFP_KERNEL);
-	if (!adapter->tx_ring)
-		goto err_tx_ring_allocation;
-
-	adapter->rx_ring = kcalloc(adapter->num_rx_queues,
-	                           sizeof(struct ixgbe_ring), GFP_KERNEL);
-	if (!adapter->rx_ring)
-		goto err_rx_ring_allocation;
+	int orig_node = adapter->node;
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		adapter->tx_ring[i].count = adapter->tx_ring_count;
-		adapter->tx_ring[i].queue_index = i;
+		struct ixgbe_ring *ring = adapter->tx_ring[i];
+		if (orig_node == -1) {
+			int cur_node = next_online_node(adapter->node);
+			if (cur_node == MAX_NUMNODES)
+				cur_node = first_online_node;
+			adapter->node = cur_node;
+		}
+		ring = kzalloc_node(sizeof(struct ixgbe_ring), GFP_KERNEL,
+		                    adapter->node);
+		if (!ring)
+			goto err_tx_ring_allocation;
+		ring->count = adapter->tx_ring_count;
+		ring->queue_index = i;
+		ring->numa_node = adapter->node;
+
+		adapter->tx_ring[i] = ring;
 	}
 
+	/* Restore the adapter's original node */
+	adapter->node = orig_node;
+
 	for (i = 0; i < adapter->num_rx_queues; i++) {
-		adapter->rx_ring[i].count = adapter->rx_ring_count;
-		adapter->rx_ring[i].queue_index = i;
+		struct ixgbe_ring *ring = adapter->rx_ring[i];
+		if (orig_node == -1) {
+			int cur_node = next_online_node(adapter->node);
+			if (cur_node == MAX_NUMNODES)
+				cur_node = first_online_node;
+			adapter->node = cur_node;
+		}
+		ring = kzalloc_node(sizeof(struct ixgbe_ring), GFP_KERNEL,
+		                    adapter->node);
+		if (!ring)
+			goto err_rx_ring_allocation;
+		ring->count = adapter->rx_ring_count;
+		ring->queue_index = i;
+		ring->numa_node = adapter->node;
+
+		adapter->rx_ring[i] = ring;
 	}
 
+	/* Restore the adapter's original node */
+	adapter->node = orig_node;
+
 	ixgbe_cache_ring_register(adapter);
 
 	return 0;
 
 err_rx_ring_allocation:
-	kfree(adapter->tx_ring);
+	for (i = 0; i < adapter->num_tx_queues; i++)
+		kfree(adapter->tx_ring[i]);
 err_tx_ring_allocation:
 	return -ENOMEM;
 }
@@ -3869,10 +3895,16 @@  err_set_interrupt:
  **/
 void ixgbe_clear_interrupt_scheme(struct ixgbe_adapter *adapter)
 {
-	kfree(adapter->tx_ring);
-	kfree(adapter->rx_ring);
-	adapter->tx_ring = NULL;
-	adapter->rx_ring = NULL;
+	int i;
+
+	for (i = 0; i < adapter->num_tx_queues; i++) {
+		kfree(adapter->tx_ring[i]);
+		adapter->tx_ring[i] = NULL;
+	}
+	for (i = 0; i < adapter->num_rx_queues; i++) {
+		kfree(adapter->rx_ring[i]);
+		adapter->rx_ring[i] = NULL;
+	}
 
 	ixgbe_free_q_vectors(adapter);
 	ixgbe_reset_interrupt_capability(adapter);
@@ -4064,7 +4096,7 @@  int ixgbe_setup_tx_resources(struct ixgbe_adapter *adapter,
 	int size;
 
 	size = sizeof(struct ixgbe_tx_buffer) * tx_ring->count;
-	tx_ring->tx_buffer_info = vmalloc_node(size, adapter->node);
+	tx_ring->tx_buffer_info = vmalloc_node(size, tx_ring->numa_node);
 	if (!tx_ring->tx_buffer_info)
 		goto err;
 	memset(tx_ring->tx_buffer_info, 0, size);
@@ -4104,25 +4136,15 @@  err:
 static int ixgbe_setup_all_tx_resources(struct ixgbe_adapter *adapter)
 {
 	int i, err = 0;
-	int orig_node = adapter->node;
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
-		if (orig_node == -1) {
-			int cur_node = next_online_node(adapter->node);
-			if (cur_node == MAX_NUMNODES)
-				cur_node = first_online_node;
-			adapter->node = cur_node;
-		}
-		err = ixgbe_setup_tx_resources(adapter, &adapter->tx_ring[i]);
+		err = ixgbe_setup_tx_resources(adapter, adapter->tx_ring[i]);
 		if (!err)
 			continue;
 		DPRINTK(PROBE, ERR, "Allocation for Tx Queue %u failed\n", i);
 		break;
 	}
 
-	/* reset the node back to its starting value */
-	adapter->node = orig_node;
-
 	return err;
 }
 
@@ -4140,7 +4162,7 @@  int ixgbe_setup_rx_resources(struct ixgbe_adapter *adapter,
 	int size;
 
 	size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count;
-	rx_ring->rx_buffer_info = vmalloc_node(size, adapter->node);
+	rx_ring->rx_buffer_info = vmalloc_node(size, rx_ring->numa_node);
 	if (!rx_ring->rx_buffer_info) {
 		DPRINTK(PROBE, ERR,
 		        "vmalloc allocation failed for the rx desc ring\n");
@@ -4184,25 +4206,15 @@  alloc_failed:
 static int ixgbe_setup_all_rx_resources(struct ixgbe_adapter *adapter)
 {
 	int i, err = 0;
-	int orig_node = adapter->node;
 
 	for (i = 0; i < adapter->num_rx_queues; i++) {
-		if (orig_node == -1) {
-			int cur_node = next_online_node(adapter->node);
-			if (cur_node == MAX_NUMNODES)
-				cur_node = first_online_node;
-			adapter->node = cur_node;
-		}
-		err = ixgbe_setup_rx_resources(adapter, &adapter->rx_ring[i]);
+		err = ixgbe_setup_rx_resources(adapter, adapter->rx_ring[i]);
 		if (!err)
 			continue;
 		DPRINTK(PROBE, ERR, "Allocation for Rx Queue %u failed\n", i);
 		break;
 	}
 
-	/* reset the node back to its starting value */
-	adapter->node = orig_node;
-
 	return err;
 }
 
@@ -4239,8 +4251,8 @@  static void ixgbe_free_all_tx_resources(struct ixgbe_adapter *adapter)
 	int i;
 
 	for (i = 0; i < adapter->num_tx_queues; i++)
-		if (adapter->tx_ring[i].desc)
-			ixgbe_free_tx_resources(adapter, &adapter->tx_ring[i]);
+		if (adapter->tx_ring[i]->desc)
+			ixgbe_free_tx_resources(adapter, adapter->tx_ring[i]);
 }
 
 /**
@@ -4276,8 +4288,8 @@  static void ixgbe_free_all_rx_resources(struct ixgbe_adapter *adapter)
 	int i;
 
 	for (i = 0; i < adapter->num_rx_queues; i++)
-		if (adapter->rx_ring[i].desc)
-			ixgbe_free_rx_resources(adapter, &adapter->rx_ring[i]);
+		if (adapter->rx_ring[i]->desc)
+			ixgbe_free_rx_resources(adapter, adapter->rx_ring[i]);
 }
 
 /**
@@ -4554,8 +4566,8 @@  void ixgbe_update_stats(struct ixgbe_adapter *adapter)
 			adapter->hw_rx_no_dma_resources +=
 			                     IXGBE_READ_REG(hw, IXGBE_QPRDC(i));
 		for (i = 0; i < adapter->num_rx_queues; i++) {
-			rsc_count += adapter->rx_ring[i].rsc_count;
-			rsc_flush += adapter->rx_ring[i].rsc_flush;
+			rsc_count += adapter->rx_ring[i]->rsc_count;
+			rsc_flush += adapter->rx_ring[i]->rsc_flush;
 		}
 		adapter->rsc_total_count = rsc_count;
 		adapter->rsc_total_flush = rsc_flush;
@@ -4563,11 +4575,11 @@  void ixgbe_update_stats(struct ixgbe_adapter *adapter)
 
 	/* gather some stats to the adapter struct that are per queue */
 	for (i = 0; i < adapter->num_tx_queues; i++)
-		restart_queue += adapter->tx_ring[i].restart_queue;
+		restart_queue += adapter->tx_ring[i]->restart_queue;
 	adapter->restart_queue = restart_queue;
 
 	for (i = 0; i < adapter->num_rx_queues; i++)
-		non_eop_descs += adapter->rx_ring[i].non_eop_descs;
+		non_eop_descs += adapter->rx_ring[i]->non_eop_descs;
 	adapter->non_eop_descs = non_eop_descs;
 
 	adapter->stats.crcerrs += IXGBE_READ_REG(hw, IXGBE_CRCERRS);
@@ -4806,7 +4818,7 @@  static void ixgbe_fdir_reinit_task(struct work_struct *work)
 	if (ixgbe_reinit_fdir_tables_82599(hw) == 0) {
 		for (i = 0; i < adapter->num_tx_queues; i++)
 			set_bit(__IXGBE_FDIR_INIT_DONE,
-			        &(adapter->tx_ring[i].reinit_state));
+			        &(adapter->tx_ring[i]->reinit_state));
 	} else {
 		DPRINTK(PROBE, ERR, "failed to finish FDIR re-initialization, "
 		        "ignored adding FDIR ATR filters \n");
@@ -4903,7 +4915,7 @@  static void ixgbe_watchdog_task(struct work_struct *work)
 
 	if (!netif_carrier_ok(netdev)) {
 		for (i = 0; i < adapter->num_tx_queues; i++) {
-			tx_ring = &adapter->tx_ring[i];
+			tx_ring = adapter->tx_ring[i];
 			if (tx_ring->next_to_use != tx_ring->next_to_clean) {
 				some_tx_pending = 1;
 				break;
@@ -5402,7 +5414,7 @@  static netdev_tx_t ixgbe_xmit_frame(struct sk_buff *skb,
 		}
 	}
 
-	tx_ring = &adapter->tx_ring[skb->queue_mapping];
+	tx_ring = adapter->tx_ring[skb->queue_mapping];
 
 	if ((adapter->flags & IXGBE_FLAG_FCOE_ENABLED) &&
 	    (skb->protocol == htons(ETH_P_FCOE))) {