]> asedeno.scripts.mit.edu Git - linux.git/commitdiff
ibmvnic: Allocate number of rx/tx buffers agreed on by firmware
authorThomas Falcon <tlfalcon@linux.vnet.ibm.com>
Sun, 5 Mar 2017 18:18:42 +0000 (12:18 -0600)
committerDavid S. Miller <davem@davemloft.net>
Tue, 7 Mar 2017 22:14:30 +0000 (14:14 -0800)
The amount of TX/RX buffers that the vNIC driver currently allocates
is different from the amount agreed upon in negotiation with firmware.
Correct that by allocating the requested number of buffers confirmed
by firmware.

Signed-off-by: Thomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ethernet/ibm/ibmvnic.c

index 0e87b83f25e2c3993dd404d548de30e5eced0c42..5f11b4dc95d2d1b271c9a47b3e890ef4662af2e4 100644 (file)
@@ -404,7 +404,7 @@ static int ibmvnic_open(struct net_device *netdev)
        send_map_query(adapter);
        for (i = 0; i < rxadd_subcrqs; i++) {
                init_rx_pool(adapter, &adapter->rx_pool[i],
-                            IBMVNIC_BUFFS_PER_POOL, i,
+                            adapter->req_rx_add_entries_per_subcrq, i,
                             be64_to_cpu(size_array[i]), 1);
                if (alloc_rx_pool(adapter, &adapter->rx_pool[i])) {
                        dev_err(dev, "Couldn't alloc rx pool\n");
@@ -419,23 +419,23 @@ static int ibmvnic_open(struct net_device *netdev)
        for (i = 0; i < tx_subcrqs; i++) {
                tx_pool = &adapter->tx_pool[i];
                tx_pool->tx_buff =
-                   kcalloc(adapter->max_tx_entries_per_subcrq,
+                   kcalloc(adapter->req_tx_entries_per_subcrq,
                            sizeof(struct ibmvnic_tx_buff), GFP_KERNEL);
                if (!tx_pool->tx_buff)
                        goto tx_pool_alloc_failed;
 
                if (alloc_long_term_buff(adapter, &tx_pool->long_term_buff,
-                                        adapter->max_tx_entries_per_subcrq *
+                                        adapter->req_tx_entries_per_subcrq *
                                         adapter->req_mtu))
                        goto tx_ltb_alloc_failed;
 
                tx_pool->free_map =
-                   kcalloc(adapter->max_tx_entries_per_subcrq,
+                   kcalloc(adapter->req_tx_entries_per_subcrq,
                            sizeof(int), GFP_KERNEL);
                if (!tx_pool->free_map)
                        goto tx_fm_alloc_failed;
 
-               for (j = 0; j < adapter->max_tx_entries_per_subcrq; j++)
+               for (j = 0; j < adapter->req_tx_entries_per_subcrq; j++)
                        tx_pool->free_map[j] = j;
 
                tx_pool->consumer_index = 0;
@@ -746,7 +746,7 @@ static int ibmvnic_xmit(struct sk_buff *skb, struct net_device *netdev)
 
        tx_pool->consumer_index =
            (tx_pool->consumer_index + 1) %
-               adapter->max_tx_entries_per_subcrq;
+               adapter->req_tx_entries_per_subcrq;
 
        tx_buff = &tx_pool->tx_buff[index];
        tx_buff->skb = skb;
@@ -819,7 +819,7 @@ static int ibmvnic_xmit(struct sk_buff *skb, struct net_device *netdev)
 
                if (tx_pool->consumer_index == 0)
                        tx_pool->consumer_index =
-                               adapter->max_tx_entries_per_subcrq - 1;
+                               adapter->req_tx_entries_per_subcrq - 1;
                else
                        tx_pool->consumer_index--;
 
@@ -1387,7 +1387,7 @@ static int ibmvnic_complete_tx(struct ibmvnic_adapter *adapter,
                                                     producer_index] = index;
                        adapter->tx_pool[pool].producer_index =
                            (adapter->tx_pool[pool].producer_index + 1) %
-                           adapter->max_tx_entries_per_subcrq;
+                           adapter->req_tx_entries_per_subcrq;
                }
                /* remove tx_comp scrq*/
                next->tx_comp.first = 0;