Re: [PATCH 14/16 v3] [v2] ibmveth: enable driver for CMO

[Robert Jennings <rcj@linux.vnet.ibm.com>]

I removed references to 'entitlement' after having changed the function
'get_io_entitlement' to 'get_desired_dma' to correctly indicate what the
function was doing.  Also, this function does not need to page align the
return value, the VIO bus is responsible for this.

(We would like to take this patch through linuxppc-dev with the full
 change set for this feature.  We are copying netdev for review and ack)

Enable ibmveth for Cooperative Memory Overcommitment (CMO).  For this driver
it means calculating a desired amount of IO memory based on the current MTU
and updating this value with the bus when MTU changes occur.  Because DMA
mappings can fail, we have added a bounce buffer for temporary cases where
the driver can not map IO memory for the buffer pool.

The following changes are made to enable the driver for CMO:
 * DMA mapping errors will not result in error messages if entitlement has
   been exceeded and resources were not available.
 * DMA mapping errors are handled gracefully, ibmveth_replenish_buffer_pool()
   is corrected to check the return from dma_map_single and fail gracefully.
 * The driver will have a get_desired_dma function defined to function
   in a CMO environment.
 * When the MTU is changed, the driver will update the device IO entitlement

Signed-off-by: Robert Jennings <rcj@linux.vnet.ibm.com>
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: Santiago Leon <santil@us.ibm.com>

---

 drivers/net/ibmveth.c |  169 ++++++++++++++++++++++++++++++++++++++++----------
 drivers/net/ibmveth.h |    5 +
 2 files changed, 140 insertions(+), 34 deletions(-)

Index: b/drivers/net/ibmveth.c
===================================================================
--- a/drivers/net/ibmveth.c
+++ b/drivers/net/ibmveth.c
@@ -33,6 +33,7 @@
 */
 
 #include <linux/module.h>
+#include <linux/moduleparam.h>
 #include <linux/types.h>
 #include <linux/errno.h>
 #include <linux/ioport.h>
@@ -52,7 +53,9 @@
 #include <asm/hvcall.h>
 #include <asm/atomic.h>
 #include <asm/vio.h>
+#include <asm/iommu.h>
 #include <asm/uaccess.h>
+#include <asm/firmware.h>
 #include <linux/seq_file.h>
 
 #include "ibmveth.h"
@@ -94,8 +97,10 @@ static void ibmveth_proc_register_adapte
 static void ibmveth_proc_unregister_adapter(struct ibmveth_adapter *adapter);
 static irqreturn_t ibmveth_interrupt(int irq, void *dev_instance);
 static void ibmveth_rxq_harvest_buffer(struct ibmveth_adapter *adapter);
+static unsigned long ibmveth_get_desired_dma(struct vio_dev *vdev);
 static struct kobj_type ktype_veth_pool;
 
+
 #ifdef CONFIG_PROC_FS
 #define IBMVETH_PROC_DIR "ibmveth"
 static struct proc_dir_entry *ibmveth_proc_dir;
@@ -226,16 +231,16 @@ static void ibmveth_replenish_buffer_poo
 	u32 i;
 	u32 count = pool->size - atomic_read(&pool->available);
 	u32 buffers_added = 0;
+	struct sk_buff *skb;
+	unsigned int free_index, index;
+	u64 correlator;
+	unsigned long lpar_rc;
+	dma_addr_t dma_addr;
 
 	mb();
 
 	for(i = 0; i < count; ++i) {
-		struct sk_buff *skb;
-		unsigned int free_index, index;
-		u64 correlator;
 		union ibmveth_buf_desc desc;
-		unsigned long lpar_rc;
-		dma_addr_t dma_addr;
 
 		skb = alloc_skb(pool->buff_size, GFP_ATOMIC);
 
@@ -255,6 +260,9 @@ static void ibmveth_replenish_buffer_poo
 		dma_addr = dma_map_single(&adapter->vdev->dev, skb->data,
 				pool->buff_size, DMA_FROM_DEVICE);
 
+		if (dma_mapping_error(dma_addr))
+			goto failure;
+
 		pool->free_map[free_index] = IBM_VETH_INVALID_MAP;
 		pool->dma_addr[index] = dma_addr;
 		pool->skbuff[index] = skb;
@@ -267,20 +275,9 @@ static void ibmveth_replenish_buffer_poo
 
 		lpar_rc = h_add_logical_lan_buffer(adapter->vdev->unit_address, desc.desc);
 
-		if(lpar_rc != H_SUCCESS) {
-			pool->free_map[free_index] = index;
-			pool->skbuff[index] = NULL;
-			if (pool->consumer_index == 0)
-				pool->consumer_index = pool->size - 1;
-			else
-				pool->consumer_index--;
-			dma_unmap_single(&adapter->vdev->dev,
-					pool->dma_addr[index], pool->buff_size,
-					DMA_FROM_DEVICE);
-			dev_kfree_skb_any(skb);
-			adapter->replenish_add_buff_failure++;
-			break;
-		} else {
+		if (lpar_rc != H_SUCCESS)
+			goto failure;
+		else {
 			buffers_added++;
 			adapter->replenish_add_buff_success++;
 		}
@@ -288,6 +285,24 @@ static void ibmveth_replenish_buffer_poo
 
 	mb();
 	atomic_add(buffers_added, &(pool->available));
+	return;
+
+failure:
+	pool->free_map[free_index] = index;
+	pool->skbuff[index] = NULL;
+	if (pool->consumer_index == 0)
+		pool->consumer_index = pool->size - 1;
+	else
+		pool->consumer_index--;
+	if (!dma_mapping_error(dma_addr))
+		dma_unmap_single(&adapter->vdev->dev,
+		                 pool->dma_addr[index], pool->buff_size,
+		                 DMA_FROM_DEVICE);
+	dev_kfree_skb_any(skb);
+	adapter->replenish_add_buff_failure++;
+
+	mb();
+	atomic_add(buffers_added, &(pool->available));
 }
 
 /* replenish routine */
@@ -297,7 +312,7 @@ static void ibmveth_replenish_task(struc
 
 	adapter->replenish_task_cycles++;
 
-	for(i = 0; i < IbmVethNumBufferPools; i++)
+	for (i = (IbmVethNumBufferPools - 1); i >= 0; i--)
 		if(adapter->rx_buff_pool[i].active)
 			ibmveth_replenish_buffer_pool(adapter,
 						     &adapter->rx_buff_pool[i]);
@@ -472,6 +487,18 @@ static void ibmveth_cleanup(struct ibmve
 		if (adapter->rx_buff_pool[i].active)
 			ibmveth_free_buffer_pool(adapter,
 						 &adapter->rx_buff_pool[i]);
+
+	if (adapter->bounce_buffer != NULL) {
+		if (!dma_mapping_error(adapter->bounce_buffer_dma)) {
+			dma_unmap_single(&adapter->vdev->dev,
+					adapter->bounce_buffer_dma,
+					adapter->netdev->mtu + IBMVETH_BUFF_OH,
+					DMA_BIDIRECTIONAL);
+			adapter->bounce_buffer_dma = DMA_ERROR_CODE;
+		}
+		kfree(adapter->bounce_buffer);
+		adapter->bounce_buffer = NULL;
+	}
 }
 
 static int ibmveth_register_logical_lan(struct ibmveth_adapter *adapter,
@@ -607,6 +634,24 @@ static int ibmveth_open(struct net_devic
 		return rc;
 	}
 
+	adapter->bounce_buffer =
+	    kmalloc(netdev->mtu + IBMVETH_BUFF_OH, GFP_KERNEL);
+	if (!adapter->bounce_buffer) {
+		ibmveth_error_printk("unable to allocate bounce buffer\n");
+		ibmveth_cleanup(adapter);
+		napi_disable(&adapter->napi);
+		return -ENOMEM;
+	}
+	adapter->bounce_buffer_dma =
+	    dma_map_single(&adapter->vdev->dev, adapter->bounce_buffer,
+			   netdev->mtu + IBMVETH_BUFF_OH, DMA_BIDIRECTIONAL);
+	if (dma_mapping_error(adapter->bounce_buffer_dma)) {
+		ibmveth_error_printk("unable to map bounce buffer\n");
+		ibmveth_cleanup(adapter);
+		napi_disable(&adapter->napi);
+		return -ENOMEM;
+	}
+
 	ibmveth_debug_printk("initial replenish cycle\n");
 	ibmveth_interrupt(netdev->irq, netdev);
 
@@ -853,10 +898,12 @@ static int ibmveth_start_xmit(struct sk_
 	unsigned int tx_packets = 0;
 	unsigned int tx_send_failed = 0;
 	unsigned int tx_map_failed = 0;
+	int used_bounce = 0;
+	unsigned long data_dma_addr;
 
 	desc.fields.flags_len = IBMVETH_BUF_VALID | skb->len;
-	desc.fields.address = dma_map_single(&adapter->vdev->dev, skb->data,
-					     skb->len, DMA_TO_DEVICE);
+	data_dma_addr = dma_map_single(&adapter->vdev->dev, skb->data,
+				       skb->len, DMA_TO_DEVICE);
 
 	if (skb->ip_summed == CHECKSUM_PARTIAL &&
 	    ip_hdr(skb)->protocol != IPPROTO_TCP && skb_checksum_help(skb)) {
@@ -875,12 +922,16 @@ static int ibmveth_start_xmit(struct sk_
 		buf[1] = 0;
 	}
 
-	if (dma_mapping_error(desc.fields.address)) {
-		ibmveth_error_printk("tx: unable to map xmit buffer\n");
+	if (dma_mapping_error(data_dma_addr)) {
+		if (!firmware_has_feature(FW_FEATURE_CMO))
+			ibmveth_error_printk("tx: unable to map xmit buffer\n");
+		skb_copy_from_linear_data(skb, adapter->bounce_buffer,
+					  skb->len);
+		desc.fields.address = adapter->bounce_buffer_dma;
 		tx_map_failed++;
-		tx_dropped++;
-		goto out;
-	}
+		used_bounce = 1;
+	} else
+		desc.fields.address = data_dma_addr;
 
 	/* send the frame. Arbitrarily set retrycount to 1024 */
 	correlator = 0;
@@ -904,8 +955,9 @@ static int ibmveth_start_xmit(struct sk_
 		netdev->trans_start = jiffies;
 	}
 
-	dma_unmap_single(&adapter->vdev->dev, desc.fields.address,
-			 skb->len, DMA_TO_DEVICE);
+	if (!used_bounce)
+		dma_unmap_single(&adapter->vdev->dev, data_dma_addr,
+				 skb->len, DMA_TO_DEVICE);
 
 out:	spin_lock_irqsave(&adapter->stats_lock, flags);
 	netdev->stats.tx_dropped += tx_dropped;
@@ -1053,8 +1105,9 @@ static void ibmveth_set_multicast_list(s
 static int ibmveth_change_mtu(struct net_device *dev, int new_mtu)
 {
 	struct ibmveth_adapter *adapter = dev->priv;
+	struct vio_dev *viodev = adapter->vdev;
 	int new_mtu_oh = new_mtu + IBMVETH_BUFF_OH;
-	int i, rc;
+	int i;
 
 	if (new_mtu < IBMVETH_MAX_MTU)
 		return -EINVAL;
@@ -1085,10 +1138,15 @@ static int ibmveth_change_mtu(struct net
 				ibmveth_close(adapter->netdev);
 				adapter->pool_config = 0;
 				dev->mtu = new_mtu;
-				if ((rc = ibmveth_open(adapter->netdev)))
-					return rc;
-			} else
-				dev->mtu = new_mtu;
+				vio_cmo_set_dev_desired(viodev,
+						ibmveth_get_desired_dma
+						(viodev));
+				return ibmveth_open(adapter->netdev);
+			}
+			dev->mtu = new_mtu;
+			vio_cmo_set_dev_desired(viodev,
+						ibmveth_get_desired_dma
+						(viodev));
 			return 0;
 		}
 	}
@@ -1103,6 +1161,46 @@ static void ibmveth_poll_controller(stru
 }
 #endif
 
+/**
+ * ibmveth_get_desired_dma - Calculate IO memory desired by the driver
+ *
+ * @vdev: struct vio_dev for the device whose desired IO mem is to be returned
+ *
+ * Return value:
+ *	Number of bytes of IO data the driver will need to perform well.
+ */
+static unsigned long ibmveth_get_desired_dma(struct vio_dev *vdev)
+{
+	struct net_device *netdev = dev_get_drvdata(&vdev->dev);
+	struct ibmveth_adapter *adapter;
+	unsigned long ret;
+	int i;
+	int rxqentries = 1;
+
+	/* netdev inits at probe time along with the structures we need below*/
+	if (netdev == NULL)
+		return IOMMU_PAGE_ALIGN(IBMVETH_IO_ENTITLEMENT_DEFAULT);
+
+	adapter = netdev_priv(netdev);
+
+	ret = IBMVETH_BUFF_LIST_SIZE + IBMVETH_FILT_LIST_SIZE;
+	ret += IOMMU_PAGE_ALIGN(netdev->mtu);
+
+	for (i = 0; i < IbmVethNumBufferPools; i++) {
+		/* add the size of the active receive buffers */
+		if (adapter->rx_buff_pool[i].active)
+			ret +=
+			    adapter->rx_buff_pool[i].size *
+			    IOMMU_PAGE_ALIGN(adapter->rx_buff_pool[i].
+			            buff_size);
+		rxqentries += adapter->rx_buff_pool[i].size;
+	}
+	/* add the size of the receive queue entries */
+	ret += IOMMU_PAGE_ALIGN(rxqentries * sizeof(struct ibmveth_rx_q_entry));
+
+	return ret;
+}
+
 static int __devinit ibmveth_probe(struct vio_dev *dev, const struct vio_device_id *id)
 {
 	int rc, i;
@@ -1247,6 +1345,8 @@ static int __devexit ibmveth_remove(stru
 	ibmveth_proc_unregister_adapter(adapter);
 
 	free_netdev(netdev);
+	dev_set_drvdata(&dev->dev, NULL);
+
 	return 0;
 }
 
@@ -1491,6 +1591,7 @@ static struct vio_driver ibmveth_driver 
 	.id_table	= ibmveth_device_table,
 	.probe		= ibmveth_probe,
 	.remove		= ibmveth_remove,
+	.get_desired_dma = ibmveth_get_desired_dma,
 	.driver		= {
 		.name	= ibmveth_driver_name,
 		.owner	= THIS_MODULE,
Index: b/drivers/net/ibmveth.h
===================================================================
--- a/drivers/net/ibmveth.h
+++ b/drivers/net/ibmveth.h
@@ -93,9 +93,12 @@ static inline long h_illan_attributes(un
   plpar_hcall_norets(H_CHANGE_LOGICAL_LAN_MAC, ua, mac)
 
 #define IbmVethNumBufferPools 5
+#define IBMVETH_IO_ENTITLEMENT_DEFAULT 4243456 /* MTU of 1500 needs 4.2Mb */
 #define IBMVETH_BUFF_OH 22 /* Overhead: 14 ethernet header + 8 opaque handle */
 #define IBMVETH_MAX_MTU 68
 #define IBMVETH_MAX_POOL_COUNT 4096
+#define IBMVETH_BUFF_LIST_SIZE 4096
+#define IBMVETH_FILT_LIST_SIZE 4096
 #define IBMVETH_MAX_BUF_SIZE (1024 * 128)
 
 static int pool_size[] = { 512, 1024 * 2, 1024 * 16, 1024 * 32, 1024 * 64 };
@@ -143,6 +146,8 @@ struct ibmveth_adapter {
     struct ibmveth_rx_q rx_queue;
     int pool_config;
     int rx_csum;
+    void *bounce_buffer;
+    dma_addr_t bounce_buffer_dma;
 
     /* adapter specific stats */
     u64 replenish_task_cycles;