Merge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue
Jeff Kirsher says:
====================
100GbE Intel Wired LAN Driver Updates 2015-12-05
This series contains updates to fm10k only.
Jacob provides the remaining fm10k patches in the series. First change
ensures that all the logic regarding the setting of netdev features is
consolidated in one place of the driver. Fixed an issue where an assumption
was being made on how many queues are available, especially when init_hw_vf()
errors out. Fixed up an number of issues with init_hw() where failures
were not being handled properly or at all, so update the driver to check
returned error codes and respond appropriately. Fixed up typecasting
issues found where either the incorrect typecast size was used or
explicitly typecast values. Added additional debugging statistics and
rename statistic to better reflect its true value. Added support for
ITR scaling based on PCIe link speed for fm10k. Fixed up code comment
where "hardware" was misspelled.
v2: Dropped patches #1 and #10 from original submission, patch #1 was from
Nick Krause and due to his past kernel interactions, dropping his patch.
Patch #10 had questions and concerns from Tom Herbert which cannot be
addressed at this time since the author (Jacob Keller) is currently on
sabbatical, so dropping this patch for now until we can properly address
Tom's questions and concerns.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k.h b/drivers/net/ethernet/intel/fm10k/fm10k.h
index 48809e5..fa26e20 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k.h
+++ b/drivers/net/ethernet/intel/fm10k/fm10k.h
@@ -164,14 +164,20 @@
unsigned int total_packets; /* total packets processed this int */
u16 work_limit; /* total work allowed per interrupt */
u16 itr; /* interrupt throttle rate value */
+ u8 itr_scale; /* ITR adjustment scaler based on PCI speed */
u8 count; /* total number of rings in vector */
};
#define FM10K_ITR_MAX 0x0FFF /* maximum value for ITR */
#define FM10K_ITR_10K 100 /* 100us */
#define FM10K_ITR_20K 50 /* 50us */
+#define FM10K_ITR_40K 25 /* 25us */
#define FM10K_ITR_ADAPTIVE 0x8000 /* adaptive interrupt moderation flag */
+#define ITR_IS_ADAPTIVE(itr) (!!(itr & FM10K_ITR_ADAPTIVE))
+
+#define FM10K_TX_ITR_DEFAULT FM10K_ITR_40K
+#define FM10K_RX_ITR_DEFAULT FM10K_ITR_20K
#define FM10K_ITR_ENABLE (FM10K_ITR_AUTOMASK | FM10K_ITR_MASK_CLEAR)
static inline struct netdev_queue *txring_txq(const struct fm10k_ring *ring)
@@ -484,7 +490,7 @@
#endif
/* Netdev */
-struct net_device *fm10k_alloc_netdev(void);
+struct net_device *fm10k_alloc_netdev(const struct fm10k_info *info);
int fm10k_setup_rx_resources(struct fm10k_ring *);
int fm10k_setup_tx_resources(struct fm10k_ring *);
void fm10k_free_rx_resources(struct fm10k_ring *);
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
index 2ce0eba..109e211 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
@@ -111,12 +111,14 @@
static const struct fm10k_stats fm10k_gstrings_mbx_stats[] = {
FM10K_MBX_STAT("mbx_tx_busy", tx_busy),
- FM10K_MBX_STAT("mbx_tx_oversized", tx_dropped),
+ FM10K_MBX_STAT("mbx_tx_dropped", tx_dropped),
FM10K_MBX_STAT("mbx_tx_messages", tx_messages),
FM10K_MBX_STAT("mbx_tx_dwords", tx_dwords),
+ FM10K_MBX_STAT("mbx_tx_mbmem_pulled", tx_mbmem_pulled),
FM10K_MBX_STAT("mbx_rx_messages", rx_messages),
FM10K_MBX_STAT("mbx_rx_dwords", rx_dwords),
FM10K_MBX_STAT("mbx_rx_parse_err", rx_parse_err),
+ FM10K_MBX_STAT("mbx_rx_mbmem_pushed", rx_mbmem_pushed),
};
#define FM10K_GLOBAL_STATS_LEN ARRAY_SIZE(fm10k_gstrings_global_stats)
@@ -699,12 +701,10 @@
{
struct fm10k_intfc *interface = netdev_priv(dev);
- ec->use_adaptive_tx_coalesce =
- !!(interface->tx_itr & FM10K_ITR_ADAPTIVE);
+ ec->use_adaptive_tx_coalesce = ITR_IS_ADAPTIVE(interface->tx_itr);
ec->tx_coalesce_usecs = interface->tx_itr & ~FM10K_ITR_ADAPTIVE;
- ec->use_adaptive_rx_coalesce =
- !!(interface->rx_itr & FM10K_ITR_ADAPTIVE);
+ ec->use_adaptive_rx_coalesce = ITR_IS_ADAPTIVE(interface->rx_itr);
ec->rx_coalesce_usecs = interface->rx_itr & ~FM10K_ITR_ADAPTIVE;
return 0;
@@ -729,10 +729,10 @@
/* set initial values for adaptive ITR */
if (ec->use_adaptive_tx_coalesce)
- tx_itr = FM10K_ITR_ADAPTIVE | FM10K_ITR_10K;
+ tx_itr = FM10K_ITR_ADAPTIVE | FM10K_TX_ITR_DEFAULT;
if (ec->use_adaptive_rx_coalesce)
- rx_itr = FM10K_ITR_ADAPTIVE | FM10K_ITR_20K;
+ rx_itr = FM10K_ITR_ADAPTIVE | FM10K_RX_ITR_DEFAULT;
/* update interface */
interface->tx_itr = tx_itr;
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_main.c b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
index 746a198..1c17b62 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_main.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
@@ -1094,11 +1094,11 @@
netdev_tx_t fm10k_xmit_frame_ring(struct sk_buff *skb,
struct fm10k_ring *tx_ring)
{
- struct fm10k_tx_buffer *first;
- int tso;
- u32 tx_flags = 0;
- unsigned short f;
u16 count = TXD_USE_COUNT(skb_headlen(skb));
+ struct fm10k_tx_buffer *first;
+ unsigned short f;
+ u32 tx_flags = 0;
+ int tso;
/* need: 1 descriptor per page * PAGE_SIZE/FM10K_MAX_DATA_PER_TXD,
* + 1 desc for skb_headlen/FM10K_MAX_DATA_PER_TXD,
@@ -1363,10 +1363,10 @@
**/
static void fm10k_update_itr(struct fm10k_ring_container *ring_container)
{
- unsigned int avg_wire_size, packets;
+ unsigned int avg_wire_size, packets, itr_round;
/* Only update ITR if we are using adaptive setting */
- if (!(ring_container->itr & FM10K_ITR_ADAPTIVE))
+ if (!ITR_IS_ADAPTIVE(ring_container->itr))
goto clear_counts;
packets = ring_container->total_packets;
@@ -1375,18 +1375,44 @@
avg_wire_size = ring_container->total_bytes / packets;
- /* Add 24 bytes to size to account for CRC, preamble, and gap */
- avg_wire_size += 24;
+ /* The following is a crude approximation of:
+ * wmem_default / (size + overhead) = desired_pkts_per_int
+ * rate / bits_per_byte / (size + ethernet overhead) = pkt_rate
+ * (desired_pkt_rate / pkt_rate) * usecs_per_sec = ITR value
+ *
+ * Assuming wmem_default is 212992 and overhead is 640 bytes per
+ * packet, (256 skb, 64 headroom, 320 shared info), we can reduce the
+ * formula down to
+ *
+ * (34 * (size + 24)) / (size + 640) = ITR
+ *
+ * We first do some math on the packet size and then finally bitshift
+ * by 8 after rounding up. We also have to account for PCIe link speed
+ * difference as ITR scales based on this.
+ */
+ if (avg_wire_size <= 360) {
+ /* Start at 250K ints/sec and gradually drop to 77K ints/sec */
+ avg_wire_size *= 8;
+ avg_wire_size += 376;
+ } else if (avg_wire_size <= 1152) {
+ /* 77K ints/sec to 45K ints/sec */
+ avg_wire_size *= 3;
+ avg_wire_size += 2176;
+ } else if (avg_wire_size <= 1920) {
+ /* 45K ints/sec to 38K ints/sec */
+ avg_wire_size += 4480;
+ } else {
+ /* plateau at a limit of 38K ints/sec */
+ avg_wire_size = 6656;
+ }
- /* Don't starve jumbo frames */
- if (avg_wire_size > 3000)
- avg_wire_size = 3000;
-
- /* Give a little boost to mid-size frames */
- if ((avg_wire_size > 300) && (avg_wire_size < 1200))
- avg_wire_size /= 3;
- else
- avg_wire_size /= 2;
+ /* Perform final bitshift for division after rounding up to ensure
+ * that the calculation will never get below a 1. The bit shift
+ * accounts for changes in the ITR due to PCIe link speed.
+ */
+ itr_round = ACCESS_ONCE(ring_container->itr_scale) + 8;
+ avg_wire_size += (1 << itr_round) - 1;
+ avg_wire_size >>= itr_round;
/* write back value and retain adaptive flag */
ring_container->itr = avg_wire_size | FM10K_ITR_ADAPTIVE;
@@ -1604,6 +1630,7 @@
q_vector->tx.ring = ring;
q_vector->tx.work_limit = FM10K_DEFAULT_TX_WORK;
q_vector->tx.itr = interface->tx_itr;
+ q_vector->tx.itr_scale = interface->hw.mac.itr_scale;
q_vector->tx.count = txr_count;
while (txr_count) {
@@ -1632,6 +1659,7 @@
/* save Rx ring container info */
q_vector->rx.ring = ring;
q_vector->rx.itr = interface->rx_itr;
+ q_vector->rx.itr_scale = interface->hw.mac.itr_scale;
q_vector->rx.count = rxr_count;
while (rxr_count) {
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_mbx.c b/drivers/net/ethernet/intel/fm10k/fm10k_mbx.c
index af09a1b..2bce474 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_mbx.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_mbx.c
@@ -375,6 +375,8 @@
if (!tail)
tail++;
+ mbx->tx_mbmem_pulled++;
+
/* write message to hardware FIFO */
fm10k_write_reg(hw, mbmem + tail++, *(head++));
} while (--len && --end);
@@ -459,6 +461,8 @@
if (!head)
head++;
+ mbx->rx_mbmem_pushed++;
+
/* read message from hardware FIFO */
*(tail++) = fm10k_read_reg(hw, mbmem + head++);
} while (--len && --end);
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_mbx.h b/drivers/net/ethernet/intel/fm10k/fm10k_mbx.h
index 0419a7f..c4f18a8 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_mbx.h
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_mbx.h
@@ -1,5 +1,5 @@
/* Intel Ethernet Switch Host Interface Driver
- * Copyright(c) 2013 - 2014 Intel Corporation.
+ * Copyright(c) 2013 - 2015 Intel Corporation.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms and conditions of the GNU General Public License,
@@ -291,8 +291,10 @@
u64 tx_dropped;
u64 tx_messages;
u64 tx_dwords;
+ u64 tx_mbmem_pulled;
u64 rx_messages;
u64 rx_dwords;
+ u64 rx_mbmem_pushed;
u64 rx_parse_err;
/* Buffer to store messages */
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c
index 7781e80..79f6b7d 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c
@@ -1388,8 +1388,9 @@
#define DEFAULT_DEBUG_LEVEL_SHIFT 3
-struct net_device *fm10k_alloc_netdev(void)
+struct net_device *fm10k_alloc_netdev(const struct fm10k_info *info)
{
+ netdev_features_t hw_features;
struct fm10k_intfc *interface;
struct net_device *dev;
@@ -1412,27 +1413,31 @@
NETIF_F_TSO |
NETIF_F_TSO6 |
NETIF_F_TSO_ECN |
- NETIF_F_GSO_UDP_TUNNEL |
NETIF_F_RXHASH |
NETIF_F_RXCSUM;
+ /* Only the PF can support VXLAN and NVGRE tunnel offloads */
+ if (info->mac == fm10k_mac_pf) {
+ dev->hw_enc_features = NETIF_F_IP_CSUM |
+ NETIF_F_TSO |
+ NETIF_F_TSO6 |
+ NETIF_F_TSO_ECN |
+ NETIF_F_GSO_UDP_TUNNEL |
+ NETIF_F_IPV6_CSUM |
+ NETIF_F_SG;
+
+ dev->features |= NETIF_F_GSO_UDP_TUNNEL;
+ }
+
/* all features defined to this point should be changeable */
- dev->hw_features |= dev->features;
+ hw_features = dev->features;
/* allow user to enable L2 forwarding acceleration */
- dev->hw_features |= NETIF_F_HW_L2FW_DOFFLOAD;
+ hw_features |= NETIF_F_HW_L2FW_DOFFLOAD;
/* configure VLAN features */
dev->vlan_features |= dev->features;
- /* configure tunnel offloads */
- dev->hw_enc_features |= NETIF_F_IP_CSUM |
- NETIF_F_TSO |
- NETIF_F_TSO6 |
- NETIF_F_TSO_ECN |
- NETIF_F_GSO_UDP_TUNNEL |
- NETIF_F_IPV6_CSUM;
-
/* we want to leave these both on as we cannot disable VLAN tag
* insertion or stripping on the hardware since it is contained
* in the FTAG and not in the frame itself.
@@ -1443,5 +1448,7 @@
dev->priv_flags |= IFF_UNICAST_FLT;
+ dev->hw_features |= hw_features;
+
return dev;
}
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_pci.c b/drivers/net/ethernet/intel/fm10k/fm10k_pci.c
index 5fbffba..15d8e10 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_pci.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_pci.c
@@ -159,13 +159,30 @@
fm10k_mbx_free_irq(interface);
+ /* free interrupts */
+ fm10k_clear_queueing_scheme(interface);
+
/* delay any future reset requests */
interface->last_reset = jiffies + (10 * HZ);
/* reset and initialize the hardware so it is in a known state */
- err = hw->mac.ops.reset_hw(hw) ? : hw->mac.ops.init_hw(hw);
- if (err)
+ err = hw->mac.ops.reset_hw(hw);
+ if (err) {
+ dev_err(&interface->pdev->dev, "reset_hw failed: %d\n", err);
+ goto reinit_err;
+ }
+
+ err = hw->mac.ops.init_hw(hw);
+ if (err) {
dev_err(&interface->pdev->dev, "init_hw failed: %d\n", err);
+ goto reinit_err;
+ }
+
+ err = fm10k_init_queueing_scheme(interface);
+ if (err) {
+ dev_err(&interface->pdev->dev, "init_queueing_scheme failed: %d\n", err);
+ goto reinit_err;
+ }
/* reassociate interrupts */
fm10k_mbx_request_irq(interface);
@@ -193,6 +210,10 @@
fm10k_iov_resume(interface->pdev);
+reinit_err:
+ if (err)
+ netif_device_detach(netdev);
+
rtnl_unlock();
clear_bit(__FM10K_RESETTING, &interface->state);
@@ -563,7 +584,7 @@
/* store tail pointer */
ring->tail = &interface->uc_addr[FM10K_TDT(reg_idx)];
- /* reset ntu and ntc to place SW in sync with hardwdare */
+ /* reset ntu and ntc to place SW in sync with hardware */
ring->next_to_clean = 0;
ring->next_to_use = 0;
@@ -669,7 +690,7 @@
/* store tail pointer */
ring->tail = &interface->uc_addr[FM10K_RDT(reg_idx)];
- /* reset ntu and ntc to place SW in sync with hardwdare */
+ /* reset ntu and ntc to place SW in sync with hardware */
ring->next_to_clean = 0;
ring->next_to_use = 0;
ring->next_to_alloc = 0;
@@ -859,7 +880,8 @@
/* re-enable mailbox interrupt and indicate 20us delay */
fm10k_write_reg(hw, FM10K_VFITR(FM10K_MBX_VECTOR),
- FM10K_ITR_ENABLE | FM10K_MBX_INT_DELAY);
+ FM10K_ITR_ENABLE | (FM10K_MBX_INT_DELAY >>
+ hw->mac.itr_scale));
/* service upstream mailbox */
if (fm10k_mbx_trylock(interface)) {
@@ -1090,7 +1112,8 @@
/* re-enable mailbox interrupt and indicate 20us delay */
fm10k_write_reg(hw, FM10K_ITR(FM10K_MBX_VECTOR),
- FM10K_ITR_ENABLE | FM10K_MBX_INT_DELAY);
+ FM10K_ITR_ENABLE | (FM10K_MBX_INT_DELAY >>
+ hw->mac.itr_scale));
return IRQ_HANDLED;
}
@@ -1684,7 +1707,13 @@
interface->last_reset = jiffies + (10 * HZ);
/* reset and initialize the hardware so it is in a known state */
- err = hw->mac.ops.reset_hw(hw) ? : hw->mac.ops.init_hw(hw);
+ err = hw->mac.ops.reset_hw(hw);
+ if (err) {
+ dev_err(&pdev->dev, "reset_hw failed: %d\n", err);
+ return err;
+ }
+
+ err = hw->mac.ops.init_hw(hw);
if (err) {
dev_err(&pdev->dev, "init_hw failed: %d\n", err);
return err;
@@ -1722,13 +1751,6 @@
pci_resource_len(pdev, 4));
hw->sw_addr = interface->sw_addr;
- /* Only the PF can support VXLAN and NVGRE offloads */
- if (hw->mac.type != fm10k_mac_pf) {
- netdev->hw_enc_features = 0;
- netdev->features &= ~NETIF_F_GSO_UDP_TUNNEL;
- netdev->hw_features &= ~NETIF_F_GSO_UDP_TUNNEL;
- }
-
/* initialize DCBNL interface */
fm10k_dcbnl_set_ops(netdev);
@@ -1749,8 +1771,8 @@
interface->rx_ring_count = FM10K_DEFAULT_RXD;
/* set default interrupt moderation */
- interface->tx_itr = FM10K_ITR_10K;
- interface->rx_itr = FM10K_ITR_ADAPTIVE | FM10K_ITR_20K;
+ interface->tx_itr = FM10K_TX_ITR_DEFAULT;
+ interface->rx_itr = FM10K_ITR_ADAPTIVE | FM10K_RX_ITR_DEFAULT;
/* initialize vxlan_port list */
INIT_LIST_HEAD(&interface->vxlan_port);
@@ -1894,7 +1916,7 @@
pci_set_master(pdev);
pci_save_state(pdev);
- netdev = fm10k_alloc_netdev();
+ netdev = fm10k_alloc_netdev(fm10k_info_tbl[ent->driver_data]);
if (!netdev) {
err = -ENOMEM;
goto err_alloc_netdev;
@@ -2071,8 +2093,10 @@
/* reset hardware to known state */
err = hw->mac.ops.init_hw(&interface->hw);
- if (err)
+ if (err) {
+ dev_err(&pdev->dev, "init_hw failed: %d\n", err);
return err;
+ }
/* reset statistics starting values */
hw->mac.ops.rebind_hw_stats(hw, &interface->stats);
@@ -2185,6 +2209,9 @@
if (netif_running(netdev))
fm10k_close(netdev);
+ /* free interrupts */
+ fm10k_clear_queueing_scheme(interface);
+
fm10k_mbx_free_irq(interface);
pci_disable_device(pdev);
@@ -2248,11 +2275,21 @@
int err = 0;
/* reset hardware to known state */
- hw->mac.ops.init_hw(&interface->hw);
+ err = hw->mac.ops.init_hw(&interface->hw);
+ if (err) {
+ dev_err(&pdev->dev, "init_hw failed: %d\n", err);
+ return;
+ }
/* reset statistics starting values */
hw->mac.ops.rebind_hw_stats(hw, &interface->stats);
+ err = fm10k_init_queueing_scheme(interface);
+ if (err) {
+ dev_err(&interface->pdev->dev, "init_queueing_scheme failed: %d\n", err);
+ return;
+ }
+
/* reassociate interrupts */
fm10k_mbx_request_irq(interface);
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_pf.c b/drivers/net/ethernet/intel/fm10k/fm10k_pf.c
index 8c0bdc4..8b9b6ba 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_pf.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_pf.c
@@ -1,5 +1,5 @@
/* Intel Ethernet Switch Host Interface Driver
- * Copyright(c) 2013 - 2014 Intel Corporation.
+ * Copyright(c) 2013 - 2015 Intel Corporation.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms and conditions of the GNU General Public License,
@@ -150,19 +150,26 @@
FM10K_TPH_RXCTRL_HDR_WROEN);
}
- /* set max hold interval to align with 1.024 usec in all modes */
+ /* set max hold interval to align with 1.024 usec in all modes and
+ * store ITR scale
+ */
switch (hw->bus.speed) {
case fm10k_bus_speed_2500:
dma_ctrl = FM10K_DMA_CTRL_MAX_HOLD_1US_GEN1;
+ hw->mac.itr_scale = FM10K_TDLEN_ITR_SCALE_GEN1;
break;
case fm10k_bus_speed_5000:
dma_ctrl = FM10K_DMA_CTRL_MAX_HOLD_1US_GEN2;
+ hw->mac.itr_scale = FM10K_TDLEN_ITR_SCALE_GEN2;
break;
case fm10k_bus_speed_8000:
dma_ctrl = FM10K_DMA_CTRL_MAX_HOLD_1US_GEN3;
+ hw->mac.itr_scale = FM10K_TDLEN_ITR_SCALE_GEN3;
break;
default:
dma_ctrl = 0;
+ /* just in case, assume Gen3 ITR scale */
+ hw->mac.itr_scale = FM10K_TDLEN_ITR_SCALE_GEN3;
break;
}
@@ -334,8 +341,8 @@
((u32)mac[3] << 16) |
((u32)mac[4] << 8) |
((u32)mac[5]));
- mac_update.mac_upper = cpu_to_le16(((u32)mac[0] << 8) |
- ((u32)mac[1]));
+ mac_update.mac_upper = cpu_to_le16(((u16)mac[0] << 8) |
+ ((u16)mac[1]));
mac_update.vlan = cpu_to_le16(vid);
mac_update.glort = cpu_to_le16(glort);
mac_update.action = add ? 0 : 1;
@@ -903,6 +910,13 @@
fm10k_write_reg(hw, FM10K_TDBAL(vf_q_idx), tdbal);
fm10k_write_reg(hw, FM10K_TDBAH(vf_q_idx), tdbah);
+ /* Provide the VF the ITR scale, using software-defined fields in TDLEN
+ * to pass the information during VF initialization. See definition of
+ * FM10K_TDLEN_ITR_SCALE_SHIFT for more details.
+ */
+ fm10k_write_reg(hw, FM10K_TDLEN(vf_q_idx), hw->mac.itr_scale <<
+ FM10K_TDLEN_ITR_SCALE_SHIFT);
+
err_out:
/* configure Queue control register */
txqctl = ((u32)vf_vid << FM10K_TXQCTL_VID_SHIFT) &
@@ -1035,6 +1049,12 @@
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);
+ /* See definition of FM10K_TDLEN_ITR_SCALE_SHIFT for an
+ * explanation of how TDLEN is used.
+ */
+ fm10k_write_reg(hw, FM10K_TDLEN(vf_q_idx + i),
+ hw->mac.itr_scale <<
+ FM10K_TDLEN_ITR_SCALE_SHIFT);
fm10k_write_reg(hw, FM10K_TQMAP(qmap_idx + i), vf_q_idx + i);
fm10k_write_reg(hw, FM10K_RQMAP(qmap_idx + i), vf_q_idx + i);
}
@@ -1212,7 +1232,7 @@
set = !(vid & FM10K_VLAN_CLEAR);
vid &= ~FM10K_VLAN_CLEAR;
- err = fm10k_iov_select_vid(vf_info, vid);
+ err = fm10k_iov_select_vid(vf_info, (u16)vid);
if (err < 0)
return err;
else
@@ -1242,7 +1262,7 @@
if (err < 0)
return err;
else
- vlan = err;
+ vlan = (u16)err;
/* notify switch of request for new unicast address */
err = hw->mac.ops.update_uc_addr(hw, vf_info->glort,
@@ -1268,7 +1288,7 @@
if (err < 0)
return err;
else
- vlan = err;
+ vlan = (u16)err;
/* notify switch of request for new multicast address */
err = hw->mac.ops.update_mc_addr(hw, vf_info->glort,
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_type.h b/drivers/net/ethernet/intel/fm10k/fm10k_type.h
index 35afd71..0272725 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_type.h
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_type.h
@@ -1,5 +1,5 @@
/* Intel Ethernet Switch Host Interface Driver
- * Copyright(c) 2013 - 2014 Intel Corporation.
+ * Copyright(c) 2013 - 2015 Intel Corporation.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms and conditions of the GNU General Public License,
@@ -272,6 +272,20 @@
#define FM10K_TDBAL(_n) ((0x40 * (_n)) + 0x8000)
#define FM10K_TDBAH(_n) ((0x40 * (_n)) + 0x8001)
#define FM10K_TDLEN(_n) ((0x40 * (_n)) + 0x8002)
+/* When fist initialized, VFs need to know the Interrupt Throttle Rate (ITR)
+ * scale which is based on the PCIe speed but the speed information in the PCI
+ * configuration space may not be accurate. The PF already knows the ITR scale
+ * but there is no defined method to pass that information from the PF to the
+ * VF. This is accomplished during VF initialization by temporarily co-opting
+ * the yet-to-be-used TDLEN register to have the PF store the ITR shift for
+ * the VF to retrieve before the VF needs to use the TDLEN register for its
+ * intended purpose, i.e. before the Tx resources are allocated.
+ */
+#define FM10K_TDLEN_ITR_SCALE_SHIFT 9
+#define FM10K_TDLEN_ITR_SCALE_MASK 0x00000E00
+#define FM10K_TDLEN_ITR_SCALE_GEN1 2
+#define FM10K_TDLEN_ITR_SCALE_GEN2 1
+#define FM10K_TDLEN_ITR_SCALE_GEN3 0
#define FM10K_TPH_TXCTRL(_n) ((0x40 * (_n)) + 0x8003)
#define FM10K_TPH_TXCTRL_DESC_TPHEN 0x00000020
#define FM10K_TPH_TXCTRL_DESC_RROEN 0x00000200
@@ -560,6 +574,7 @@
bool get_host_state;
bool tx_ready;
u32 dglort_map;
+ u8 itr_scale;
};
struct fm10k_swapi_table_info {
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_vf.c b/drivers/net/ethernet/intel/fm10k/fm10k_vf.c
index 3a18ef1..2af697d 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_vf.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_vf.c
@@ -28,7 +28,7 @@
static s32 fm10k_stop_hw_vf(struct fm10k_hw *hw)
{
u8 *perm_addr = hw->mac.perm_addr;
- u32 bal = 0, bah = 0;
+ u32 bal = 0, bah = 0, tdlen;
s32 err;
u16 i;
@@ -48,6 +48,9 @@
((u32)perm_addr[2]);
}
+ /* restore default itr_scale for next VF initialization */
+ tdlen = hw->mac.itr_scale << FM10K_TDLEN_ITR_SCALE_SHIFT;
+
/* The queues have already been disabled so we just need to
* update their base address registers
*/
@@ -56,6 +59,12 @@
fm10k_write_reg(hw, FM10K_TDBAH(i), bah);
fm10k_write_reg(hw, FM10K_RDBAL(i), bal);
fm10k_write_reg(hw, FM10K_RDBAH(i), bah);
+ /* Restore ITR scale in software-defined mechanism in TDLEN
+ * for next VF initialization. See definition of
+ * FM10K_TDLEN_ITR_SCALE_SHIFT for more details on the use of
+ * TDLEN here.
+ */
+ fm10k_write_reg(hw, FM10K_TDLEN(i), tdlen);
}
return 0;
@@ -105,8 +114,10 @@
/* verify we have at least 1 queue */
if (!~fm10k_read_reg(hw, FM10K_TXQCTL(0)) ||
- !~fm10k_read_reg(hw, FM10K_RXQCTL(0)))
- return FM10K_ERR_NO_RESOURCES;
+ !~fm10k_read_reg(hw, FM10K_RXQCTL(0))) {
+ err = FM10K_ERR_NO_RESOURCES;
+ goto reset_max_queues;
+ }
/* determine how many queues we have */
for (i = 1; tqdloc0 && (i < FM10K_MAX_QUEUES_POOL); i++) {
@@ -124,16 +135,28 @@
/* shut down queues we own and reset DMA configuration */
err = fm10k_disable_queues_generic(hw, i);
if (err)
- return err;
+ goto reset_max_queues;
/* record maximum queue count */
hw->mac.max_queues = i;
- /* fetch default VLAN */
+ /* fetch default VLAN and ITR scale */
hw->mac.default_vid = (fm10k_read_reg(hw, FM10K_TXQCTL(0)) &
FM10K_TXQCTL_VID_MASK) >> FM10K_TXQCTL_VID_SHIFT;
+ /* Read the ITR scale from TDLEN. See the definition of
+ * FM10K_TDLEN_ITR_SCALE_SHIFT for more information about how TDLEN is
+ * used here.
+ */
+ hw->mac.itr_scale = (fm10k_read_reg(hw, FM10K_TDLEN(0)) &
+ FM10K_TDLEN_ITR_SCALE_MASK) >>
+ FM10K_TDLEN_ITR_SCALE_SHIFT;
return 0;
+
+reset_max_queues:
+ hw->mac.max_queues = 0;
+
+ return err;
}
/* This structure defines the attibutes to be parsed below */