Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/gi…

…t/jkirsher/net-next

Jeff Kirsher says:

====================
Intel Wired LAN Driver Updates 2015-03-03

This series contains updates to fm10k, i40e and i40evf.

Matthew updates the fm10k driver by cleaning up code comments and whitespace
issues.  Also modifies the tunnel length header check, to make it more robust
by calculating the inner L4 header length based on whether it is TCP or UDP.
Implemented ndo_features_check() that allows drivers to report their offload
capabilities per-skb.

Neerav updates the i40e driver to skip over priority tagging if DCB is not
enabled.  Fixes an issue where the driver is not flushing out the
DCBNL app table for applications that are not present in the local DCBX
application configuration TLVs.  Fixed i40e where, in the case of MFP
mode, the driver was returning the incorrect number of traffic classes
for partitions that are not enabled for iSCSI.  Even though the driver
was not configuring these traffic classes in the transmit scheduler for
the NIC partitions, it does use this map to setup the queue mappings.

Shannon updates i40e/i40evf to include the firmware build number in the
formatted firmware version string.

Akeem adds a safety net (by adding a 'default' case) for the possible
unmatched switch calls.

Mitch updates i40e to not automatically disable PF loopback at runtime,
now that we have the functionality to enable and disable PF loopback.  This
fix cleans up a bogus error message when removing the PF module with VFs
enabled.  Adds a extra check to make sure that the indirection table
pointer is valid before dereferencing it.

Anjali enables i40e to enable more than the max RSS qps when running in a
single TC mode for the main VSI.  It is possible to enable as many as
num_online_cpus().  Adds a firmware check to ensure that DCB is disabled for
firmware versions older than 4.33.  Updates i40e/i40evf to add missing
packet types for VXLAN offload.  Updated i40e to be able to handle varying
RSS table size for each VSI, since all VSI's do not have the same RSS table
size.

v2: Dropped previous patch #9 "i40e/i40evf: Add capability to gather VEB
    per TC stats" since the stats should be in ethtool and not debugfs.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/intel/fm10k/fm10k.h

@@ -439,6 +439,7 @@ extern char fm10k_driver_name[];
 extern const char fm10k_driver_version[];
 int fm10k_init_queueing_scheme(struct fm10k_intfc *interface);
 void fm10k_clear_queueing_scheme(struct fm10k_intfc *interface);
+__be16 fm10k_tx_encap_offload(struct sk_buff *skb);
 netdev_tx_t fm10k_xmit_frame_ring(struct sk_buff *skb,
 				  struct fm10k_ring *tx_ring);
 void fm10k_tx_timeout_reset(struct fm10k_intfc *interface);

drivers/net/ethernet/intel/fm10k/fm10k_common.c

@@ -398,7 +398,7 @@ static void fm10k_update_hw_stats_rx_q(struct fm10k_hw *hw,
 	/* Retrieve RX Owner Data */
 	id_rx = fm10k_read_reg(hw, FM10K_RXQCTL(idx));
 
-	/* Process RX Ring*/
+	/* Process RX Ring */
 	do {
 		rx_drops = fm10k_read_hw_stats_32b(hw, FM10K_QPRDC(idx),
 						   &q->rx_drops);
@@ -466,7 +466,6 @@ void fm10k_update_hw_stats_q(struct fm10k_hw *hw, struct fm10k_hw_stats_q *q,
  *  Function invalidates the index values for the queues so any updates that
  *  may have happened are ignored and the base for the queue stats is reset.
  **/
-
 void fm10k_unbind_hw_stats_q(struct fm10k_hw_stats_q *q, u32 idx, u32 count)
 {
 	u32 i;

drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c

@@ -1019,7 +1019,7 @@ static int fm10k_set_channels(struct net_device *dev,
 }
 
 static int fm10k_get_ts_info(struct net_device *dev,
-			   struct ethtool_ts_info *info)
+			     struct ethtool_ts_info *info)
 {
 	struct fm10k_intfc *interface = netdev_priv(dev);
 

drivers/net/ethernet/intel/fm10k/fm10k_iov.c

@@ -275,7 +275,7 @@ s32 fm10k_iov_update_pvid(struct fm10k_intfc *interface, u16 glort, u16 pvid)
 	if (vf_idx >= iov_data->num_vfs)
 		return FM10K_ERR_PARAM;
 
-	/* determine if an update has occured and if so notify the VF */
+	/* determine if an update has occurred and if so notify the VF */
 	vf_info = &iov_data->vf_info[vf_idx];
 	if (vf_info->sw_vid != pvid) {
 		vf_info->sw_vid = pvid;

drivers/net/ethernet/intel/fm10k/fm10k_main.c

@@ -711,26 +711,20 @@ static struct ethhdr *fm10k_gre_is_nvgre(struct sk_buff *skb)
 	if (nvgre_hdr->flags & FM10K_NVGRE_RESERVED0_FLAGS)
 		return NULL;
 
-	/* verify protocol is transparent Ethernet bridging */
-	if (nvgre_hdr->proto != htons(ETH_P_TEB))
-		return NULL;
-
 	/* report start of ethernet header */
 	if (nvgre_hdr->flags & NVGRE_TNI)
 		return (struct ethhdr *)(nvgre_hdr + 1);
 
 	return (struct ethhdr *)(&nvgre_hdr->tni);
 }
 
-static __be16 fm10k_tx_encap_offload(struct sk_buff *skb)
+__be16 fm10k_tx_encap_offload(struct sk_buff *skb)
 {
+	u8 l4_hdr = 0, inner_l4_hdr = 0, inner_l4_hlen;
 	struct ethhdr *eth_hdr;
-	u8 l4_hdr = 0;
 
-/* fm10k supports 184 octets of outer+inner headers. Minus 20 for inner L4. */
-#define FM10K_MAX_ENCAP_TRANSPORT_OFFSET	164
-	if (skb_inner_transport_header(skb) - skb_mac_header(skb) >
-	    FM10K_MAX_ENCAP_TRANSPORT_OFFSET)
+	if (skb->inner_protocol_type != ENCAP_TYPE_ETHER ||
+	    skb->inner_protocol != htons(ETH_P_TEB))
 		return 0;
 
 	switch (vlan_get_protocol(skb)) {
@@ -760,12 +754,33 @@ static __be16 fm10k_tx_encap_offload(struct sk_buff *skb)
 
 	switch (eth_hdr->h_proto) {
 	case htons(ETH_P_IP):
+		inner_l4_hdr = inner_ip_hdr(skb)->protocol;
+		break;
 	case htons(ETH_P_IPV6):
+		inner_l4_hdr = inner_ipv6_hdr(skb)->nexthdr;
 		break;
 	default:
 		return 0;
 	}
 
+	switch (inner_l4_hdr) {
+	case IPPROTO_TCP:
+		inner_l4_hlen = inner_tcp_hdrlen(skb);
+		break;
+	case IPPROTO_UDP:
+		inner_l4_hlen = 8;
+		break;
+	default:
+		return 0;
+	}
+
+	/* The hardware allows tunnel offloads only if the combined inner and
+	 * outer header is 184 bytes or less
+	 */
+	if (skb_inner_transport_header(skb) + inner_l4_hlen -
+	    skb_mac_header(skb) > FM10K_TUNNEL_HEADER_LENGTH)
+		return 0;
+
 	return eth_hdr->h_proto;
 }
 
@@ -934,10 +949,10 @@ static int __fm10k_maybe_stop_tx(struct fm10k_ring *tx_ring, u16 size)
 {
 	netif_stop_subqueue(tx_ring->netdev, tx_ring->queue_index);
 
+	/* Memory barrier before checking head and tail */
 	smp_mb();
 
-	/* We need to check again in a case another CPU has just
-	 * made room available. */
+	/* Check again in a case another CPU has just made room available */
 	if (likely(fm10k_desc_unused(tx_ring) < size))
 		return -EBUSY;
 

drivers/net/ethernet/intel/fm10k/fm10k_mbx.c

@@ -72,7 +72,7 @@ static bool fm10k_fifo_empty(struct fm10k_mbx_fifo *fifo)
  *  @fifo: pointer to FIFO
  *  @offset: offset to add to head
  *
- *  This function returns the indicies into the fifo based on head + offset
+ *  This function returns the indices into the fifo based on head + offset
  **/
 static u16 fm10k_fifo_head_offset(struct fm10k_mbx_fifo *fifo, u16 offset)
 {
@@ -84,7 +84,7 @@ static u16 fm10k_fifo_head_offset(struct fm10k_mbx_fifo *fifo, u16 offset)
  *  @fifo: pointer to FIFO
  *  @offset: offset to add to tail
  *
- *  This function returns the indicies into the fifo based on tail + offset
+ *  This function returns the indices into the fifo based on tail + offset
  **/
 static u16 fm10k_fifo_tail_offset(struct fm10k_mbx_fifo *fifo, u16 offset)
 {
@@ -326,7 +326,7 @@ static u16 fm10k_mbx_validate_msg_size(struct fm10k_mbx_info *mbx, u16 len)
  *  fm10k_mbx_write_copy - pulls data off of Tx FIFO and places it in mbmem
  *  @mbx: pointer to mailbox
  *
- *  This function will take a seciton of the Rx FIFO and copy it into the
+ *  This function will take a section of the Rx FIFO and copy it into the
 		mbx->tail--;
  *  mailbox memory.  The offset in mbmem is based on the lower bits of the
  *  tail and len determines the length to copy.
@@ -418,7 +418,7 @@ static void fm10k_mbx_pull_head(struct fm10k_hw *hw,
  *  @hw: pointer to hardware structure
  *  @mbx: pointer to mailbox
  *
- *  This function will take a seciton of the mailbox memory and copy it
+ *  This function will take a section of the mailbox memory and copy it
  *  into the Rx FIFO.  The offset is based on the lower bits of the
  *  head and len determines the length to copy.
  **/
@@ -464,7 +464,7 @@ static void fm10k_mbx_read_copy(struct fm10k_hw *hw,
  *  @tail: tail index of message
  *
  *  This function will first validate the tail index and size for the
- *  incoming message.  It then updates the acknowlegment number and
+ *  incoming message.  It then updates the acknowledgment number and
  *  copies the data into the FIFO.  It will return the number of messages
  *  dequeued on success and a negative value on error.
  **/
@@ -761,7 +761,7 @@ static s32 fm10k_mbx_enqueue_tx(struct fm10k_hw *hw,
 		err = fm10k_fifo_enqueue(&mbx->tx, msg);
 	}
 
-	/* if we failed trhead the error */
+	/* if we failed treat the error */
 	if (err) {
 		mbx->timeout = 0;
 		mbx->tx_busy++;
@@ -815,7 +815,7 @@ static void fm10k_mbx_write(struct fm10k_hw *hw, struct fm10k_mbx_info *mbx)
 {
 	u32 mbmem = mbx->mbmem_reg;
 
-	/* write new msg header to notify recepient of change */
+	/* write new msg header to notify recipient of change */
 	fm10k_write_reg(hw, mbmem, mbx->mbx_hdr);
 
 	/* write mailbox to sent interrupt */
@@ -1251,7 +1251,7 @@ static s32 fm10k_mbx_process_error(struct fm10k_hw *hw,
 	/* we will need to pull all of the fields for verification */
 	head = FM10K_MSG_HDR_FIELD_GET(*hdr, HEAD);
 
-	/* we only have lower 10 bits of error number os add upper bits */
+	/* we only have lower 10 bits of error number so add upper bits */
 	err_no = FM10K_MSG_HDR_FIELD_GET(*hdr, ERR_NO);
 	err_no |= ~FM10K_MSG_HDR_MASK(ERR_NO);
 
@@ -1548,7 +1548,7 @@ s32 fm10k_pfvf_mbx_init(struct fm10k_hw *hw, struct fm10k_mbx_info *mbx,
 	mbx->timeout = 0;
 	mbx->udelay = FM10K_MBX_INIT_DELAY;
 
-	/* initalize tail and head */
+	/* initialize tail and head */
 	mbx->tail = 1;
 	mbx->head = 1;
 
@@ -1627,7 +1627,7 @@ static void fm10k_sm_mbx_connect_reset(struct fm10k_mbx_info *mbx)
 	mbx->local = FM10K_SM_MBX_VERSION;
 	mbx->remote = 0;
 
-	/* initalize tail and head */
+	/* initialize tail and head */
 	mbx->tail = 1;
 	mbx->head = 1;
 

drivers/net/ethernet/intel/fm10k/fm10k_netdev.c

@@ -356,7 +356,7 @@ static void fm10k_free_all_rx_resources(struct fm10k_intfc *interface)
  * fm10k_request_glort_range - Request GLORTs for use in configuring rules
  * @interface: board private structure
  *
- * This function allocates a range of glorts for this inteface to use.
+ * This function allocates a range of glorts for this interface to use.
  **/
 static void fm10k_request_glort_range(struct fm10k_intfc *interface)
 {
@@ -781,7 +781,7 @@ static int fm10k_update_vid(struct net_device *netdev, u16 vid, bool set)
 
 	fm10k_mbx_lock(interface);
 
-	/* only need to update the VLAN if not in promiscous mode */
+	/* only need to update the VLAN if not in promiscuous mode */
 	if (!(netdev->flags & IFF_PROMISC)) {
 		err = hw->mac.ops.update_vlan(hw, vid, 0, set);
 		if (err)
@@ -970,7 +970,7 @@ static void fm10k_set_rx_mode(struct net_device *dev)
 
 	fm10k_mbx_lock(interface);
 
-	/* syncronize all of the addresses */
+	/* synchronize all of the addresses */
 	if (xcast_mode != FM10K_XCAST_MODE_PROMISC) {
 		__dev_uc_sync(dev, fm10k_uc_sync, fm10k_uc_unsync);
 		if (xcast_mode != FM10K_XCAST_MODE_ALLMULTI)
@@ -1051,7 +1051,7 @@ void fm10k_restore_rx_state(struct fm10k_intfc *interface)
 					   vid, true, 0);
 	}
 
-	/* syncronize all of the addresses */
+	/* synchronize all of the addresses */
 	if (xcast_mode != FM10K_XCAST_MODE_PROMISC) {
 		__dev_uc_sync(netdev, fm10k_uc_sync, fm10k_uc_unsync);
 		if (xcast_mode != FM10K_XCAST_MODE_ALLMULTI)
@@ -1350,6 +1350,16 @@ static void fm10k_dfwd_del_station(struct net_device *dev, void *priv)
 	}
 }
 
+static netdev_features_t fm10k_features_check(struct sk_buff *skb,
+					      struct net_device *dev,
+					      netdev_features_t features)
+{
+	if (!skb->encapsulation || fm10k_tx_encap_offload(skb))
+		return features;
+
+	return features & ~(NETIF_F_ALL_CSUM | NETIF_F_GSO_MASK);
+}
+
 static const struct net_device_ops fm10k_netdev_ops = {
 	.ndo_open		= fm10k_open,
 	.ndo_stop		= fm10k_close,
@@ -1372,6 +1382,7 @@ static const struct net_device_ops fm10k_netdev_ops = {
 	.ndo_do_ioctl		= fm10k_ioctl,
 	.ndo_dfwd_add_station	= fm10k_dfwd_add_station,
 	.ndo_dfwd_del_station	= fm10k_dfwd_del_station,
+	.ndo_features_check	= fm10k_features_check,
 };
 
 #define DEFAULT_DEBUG_LEVEL_SHIFT 3

drivers/net/ethernet/intel/fm10k/fm10k_pci.c

@@ -648,7 +648,7 @@ static void fm10k_configure_rx_ring(struct fm10k_intfc *interface,
 	/* Configure the Rx buffer size for one buff without split */
 	srrctl |= FM10K_RX_BUFSZ >> FM10K_SRRCTL_BSIZEPKT_SHIFT;
 
-	/* Configure the Rx ring to supress loopback packets */
+	/* Configure the Rx ring to suppress loopback packets */
 	srrctl |= FM10K_SRRCTL_LOOPBACK_SUPPRESS;
 	fm10k_write_reg(hw, FM10K_SRRCTL(reg_idx), srrctl);
 

drivers/net/ethernet/intel/fm10k/fm10k_pf.c

@@ -234,8 +234,7 @@ static s32 fm10k_update_vlan_pf(struct fm10k_hw *hw, u32 vid, u8 vsi, bool set)
 	vid = (vid << 17) >> 17;
 
 	/* verify the reserved 0 fields are 0 */
-	if (len >= FM10K_VLAN_TABLE_VID_MAX ||
-	    vid >= FM10K_VLAN_TABLE_VID_MAX)
+	if (len >= FM10K_VLAN_TABLE_VID_MAX || vid >= FM10K_VLAN_TABLE_VID_MAX)
 		return FM10K_ERR_PARAM;
 
 	/* Loop through the table updating all required VLANs */
@@ -312,7 +311,7 @@ bool fm10k_glort_valid_pf(struct fm10k_hw *hw, u16 glort)
 }
 
 /**
- *  fm10k_update_uc_addr_pf - Update device unicast addresss
+ *  fm10k_update_xc_addr_pf - Update device addresses
  *  @hw: pointer to the HW structure
  *  @glort: base resource tag for this request
  *  @mac: MAC address to add/remove from table
@@ -356,7 +355,7 @@ static s32 fm10k_update_xc_addr_pf(struct fm10k_hw *hw, u16 glort,
 }
 
 /**
- *  fm10k_update_uc_addr_pf - Update device unicast addresss
+ *  fm10k_update_uc_addr_pf - Update device unicast addresses
  *  @hw: pointer to the HW structure
  *  @glort: base resource tag for this request
  *  @mac: MAC address to add/remove from table
@@ -454,7 +453,7 @@ static void fm10k_update_int_moderator_pf(struct fm10k_hw *hw)
 			break;
 	}
 
-	/* always reset VFITR2[0] to point to last enabled PF vector*/
+	/* always reset VFITR2[0] to point to last enabled PF vector */
 	fm10k_write_reg(hw, FM10K_ITR2(FM10K_ITR_REG_COUNT_PF), i);
 
 	/* reset ITR2[0] to point to last enabled PF vector */
@@ -812,7 +811,7 @@ static s32 fm10k_iov_assign_int_moderator_pf(struct fm10k_hw *hw, u16 vf_idx)
 	if (vf_idx >= hw->iov.num_vfs)
 		return FM10K_ERR_PARAM;
 
-	/* determine vector offset and count*/
+	/* determine vector offset and count */
 	vf_v_idx = fm10k_vf_vector_index(hw, vf_idx);
 	vf_v_limit = vf_v_idx + fm10k_vectors_per_pool(hw);
 
@@ -951,7 +950,7 @@ static s32 fm10k_iov_reset_resources_pf(struct fm10k_hw *hw,
 	if (vf_info->mbx.ops.disconnect)
 		vf_info->mbx.ops.disconnect(hw, &vf_info->mbx);
 
-	/* determine vector offset and count*/
+	/* determine vector offset and count */
 	vf_v_idx = fm10k_vf_vector_index(hw, vf_idx);
 	vf_v_limit = vf_v_idx + fm10k_vectors_per_pool(hw);
 
@@ -1035,7 +1034,7 @@ static s32 fm10k_iov_reset_resources_pf(struct fm10k_hw *hw,
 			((u32)vf_info->mac[2]);
 	}
 
-	/* map queue pairs back to VF from last to first*/
+	/* map queue pairs back to VF from last to first */
 	for (i = queues_per_pool; i--;) {
 		fm10k_write_reg(hw, FM10K_TDBAL(vf_q_idx + i), tdbal);
 		fm10k_write_reg(hw, FM10K_TDBAH(vf_q_idx + i), tdbah);
@@ -1141,7 +1140,7 @@ static s32 fm10k_iov_report_timestamp_pf(struct fm10k_hw *hw,
  *
  *  This function is a default handler for MSI-X requests from the VF.  The
  *  assumption is that in this case it is acceptable to just directly
- *  hand off the message form the VF to the underlying shared code.
+ *  hand off the message from the VF to the underlying shared code.
  **/
 s32 fm10k_iov_msg_msix_pf(struct fm10k_hw *hw, u32 **results,
 			  struct fm10k_mbx_info *mbx)
@@ -1160,7 +1159,7 @@ s32 fm10k_iov_msg_msix_pf(struct fm10k_hw *hw, u32 **results,
  *
  *  This function is a default handler for MAC/VLAN requests from the VF.
  *  The assumption is that in this case it is acceptable to just directly
- *  hand off the message form the VF to the underlying shared code.
+ *  hand off the message from the VF to the underlying shared code.
  **/
 s32 fm10k_iov_msg_mac_vlan_pf(struct fm10k_hw *hw, u32 **results,
 			      struct fm10k_mbx_info *mbx)
@@ -1404,7 +1403,7 @@ static void fm10k_update_hw_stats_pf(struct fm10k_hw *hw,
 						    &stats->vlan_drop);
 		loopback_drop = fm10k_read_hw_stats_32b(hw,
 							FM10K_STATS_LOOPBACK_DROP,
-						     &stats->loopback_drop);
+							&stats->loopback_drop);
 		nodesc_drop = fm10k_read_hw_stats_32b(hw,
 						      FM10K_STATS_NODESC_DROP,
 						      &stats->nodesc_drop);
@@ -1573,7 +1572,7 @@ static s32 fm10k_get_host_state_pf(struct fm10k_hw *hw, bool *switch_ready)
 	s32 ret_val = 0;
 	u32 dma_ctrl2;
 
-	/* verify the switch is ready for interraction */
+	/* verify the switch is ready for interaction */
 	dma_ctrl2 = fm10k_read_reg(hw, FM10K_DMA_CTRL2);
 	if (!(dma_ctrl2 & FM10K_DMA_CTRL2_SWITCH_READY))
 		goto out;

drivers/net/ethernet/intel/fm10k/fm10k_tlv.c

@@ -710,7 +710,7 @@ void fm10k_tlv_msg_test_create(u32 *msg, u32 attr_flags)
 /**
  *  fm10k_tlv_msg_test - Validate all results on test message receive
  *  @hw: Pointer to hardware structure
- *  @results: Pointer array to attributes in the mesage
+ *  @results: Pointer array to attributes in the message
  *  @mbx: Pointer to mailbox information structure
  *
  *  This function does a check to verify all attributes match what the test