Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue

Jeff Kirsher says:

====================
Intel Wired LAN Driver Updates 2018-12-20

This series contains updates to e100, igb, ixgbe, i40e and ice drivers.

I replaced spinlocks for mutex locks to reduce the latency on CPU0 for
igb when updating the statistics. This work was based off a patch
provided by Jan Jablonsky, which was against an older version of the igb
driver.

Jesus adjusts the receive packet buffer size from 32K to 30K when
running in QAV mode, to stay within 60K for total packet buffer size for
igb.

Vinicius adds igb kernel documentation regarding the CBS algorithm and
its implementation in the i210 family of NICs.

YueHaibing from Huawei fixed the e100 driver that was potentially
passing a NULL pointer, so use the kernel macro IS_ERR_OR_NULL()
instead.

Konstantin Khorenko fixes i40e where we were not setting up the
neigh_priv_len in our net_device, which caused the driver to read beyond
the neighbor entry allocated memory.

Miroslav Lichvar extends the PTP gettime() to read the system clock by
adding support for PTP_SYS_OFFSET_EXTENDED ioctl in i40e.

Young Xiao fixed the ice driver to only enable NAPI on q_vectors that
actually have transmit and receive rings.

Kai-Heng Feng fixes an igb issue that when placed in suspend mode, the
NIC does not wake up when a cable is plugged in. This was due to the
driver not setting PME during runtime suspend.

Stephen Douthit enables the ixgbe driver allow DSA devices to use the
MII interface to talk to switches.
====================

Signed-off-by: David S. Miller <[e-mail]>

Documentation/networking/device_drivers/intel/igb.rst

···
 IProute2 tool from Sourceforge if your version does not have all the features
 you require.
 
+Credit Based Shaper (Qav Mode)
+------------------------------
+When enabling the CBS qdisc in the hardware offload mode, traffic shaping using
+the CBS (described in the IEEE 802.1Q-2018 Section 8.6.8.2 and discussed in the
+Annex L) algorithm will run in the i210 controller, so it's more accurate and
+uses less CPU.
+
+When using offloaded CBS, and the traffic rate obeys the configured rate
+(doesn't go above it), CBS should have little to no effect in the latency.
+
+The offloaded version of the algorithm has some limits, caused by how the idle
+slope is expressed in the adapter's registers. It can only represent idle slopes
+in 16.38431 kbps units, which means that if a idle slope of 2576kbps is
+requested, the controller will be configured to use a idle slope of ~2589 kbps,
+because the driver rounds the value up. For more details, see the comments on
+:c:func:`igb_config_tx_modes()`.
+
+NOTE: This feature is exclusive to i210 models.
+
 
 Support
 =======

drivers/net/ethernet/intel/Kconfig

···
 	tristate "Intel(R) 10GbE PCI Express adapters support"
 	depends on PCI
 	select MDIO
+	select MDIO_DEVICE
 	imply PTP_1588_CLOCK
 	---help---
 	  This driver supports Intel(R) 10GbE PCI Express family of

drivers/net/ethernet/intel/e100.c

···
 
 	fw = e100_request_firmware(nic);
 	/* If it's NULL, then no ucode is required */
-	if (!fw || IS_ERR(fw))
-		return PTR_ERR(fw);
+	if (IS_ERR_OR_NULL(fw))
+		return PTR_ERR_OR_ZERO(fw);
 
 	if ((err = e100_exec_cb(nic, (void *)fw, e100_setup_ucode)))
 		netif_err(nic, probe, nic->netdev,

drivers/net/ethernet/intel/i40e/i40e_main.c

···
 	ether_addr_copy(netdev->dev_addr, mac_addr);
 	ether_addr_copy(netdev->perm_addr, mac_addr);
 
+	/* i40iw_net_event() reads 16 bytes from neigh->primary_key */
+	netdev->neigh_priv_len = sizeof(u32) * 4;
+
 	netdev->priv_flags |= IFF_UNICAST_FLT;
 	netdev->priv_flags |= IFF_SUPP_NOFCS;
 	/* Setup netdev TC information */

drivers/net/ethernet/intel/i40e/i40e_ptp.c

···
  * i40e_ptp_read - Read the PHC time from the device
  * @pf: Board private structure
  * @ts: timespec structure to hold the current time value
+ * @sts: structure to hold the system time before and after reading the PHC
  *
  * This function reads the PRTTSYN_TIME registers and stores them in a
  * timespec. However, since the registers are 64 bits of nanoseconds, we must
  * convert the result to a timespec before we can return.
  **/
-static void i40e_ptp_read(struct i40e_pf *pf, struct timespec64 *ts)
+static void i40e_ptp_read(struct i40e_pf *pf, struct timespec64 *ts,
+			  struct ptp_system_timestamp *sts)
 {
 	struct i40e_hw *hw = &pf->hw;
 	u32 hi, lo;
 	u64 ns;
 
 	/* The timer latches on the lowest register read. */
+	ptp_read_system_prets(sts);
 	lo = rd32(hw, I40E_PRTTSYN_TIME_L);
+	ptp_read_system_postts(sts);
 	hi = rd32(hw, I40E_PRTTSYN_TIME_H);
 
 	ns = (((u64)hi) << 32) | lo;
···
 
 	mutex_lock(&pf->tmreg_lock);
 
-	i40e_ptp_read(pf, &now);
+	i40e_ptp_read(pf, &now, NULL);
 	timespec64_add_ns(&now, delta);
 	i40e_ptp_write(pf, (const struct timespec64 *)&now);
 
···
 }
 
 /**
- * i40e_ptp_gettime - Get the time of the PHC
+ * i40e_ptp_gettimex - Get the time of the PHC
  * @ptp: The PTP clock structure
  * @ts: timespec structure to hold the current time value
+ * @sts: structure to hold the system time before and after reading the PHC
  *
  * Read the device clock and return the correct value on ns, after converting it
  * into a timespec struct.
  **/
-static int i40e_ptp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts)
+static int i40e_ptp_gettimex(struct ptp_clock_info *ptp, struct timespec64 *ts,
+			     struct ptp_system_timestamp *sts)
 {
 	struct i40e_pf *pf = container_of(ptp, struct i40e_pf, ptp_caps);
 
 	mutex_lock(&pf->tmreg_lock);
-	i40e_ptp_read(pf, ts);
+	i40e_ptp_read(pf, ts, sts);
 	mutex_unlock(&pf->tmreg_lock);
 
 	return 0;
···
 	pf->ptp_caps.pps = 0;
 	pf->ptp_caps.adjfreq = i40e_ptp_adjfreq;
 	pf->ptp_caps.adjtime = i40e_ptp_adjtime;
-	pf->ptp_caps.gettime64 = i40e_ptp_gettime;
+	pf->ptp_caps.gettimex64 = i40e_ptp_gettimex;
 	pf->ptp_caps.settime64 = i40e_ptp_settime;
 	pf->ptp_caps.enable = i40e_ptp_feature_enable;
 

drivers/net/ethernet/intel/ice/ice_main.c

···
 	if (!vsi->netdev)
 		return;
 
-	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++)
-		napi_enable(&vsi->q_vectors[q_idx]->napi);
+	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++) {
+		struct ice_q_vector *q_vector = vsi->q_vectors[q_idx];
+
+		if (q_vector->rx.ring || q_vector->tx.ring)
+			napi_enable(&q_vector->napi);
+	}
 }
 
 /**
···
 	if (!vsi->netdev)
 		return;
 
-	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++)
-		napi_disable(&vsi->q_vectors[q_idx]->napi);
+	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++) {
+		struct ice_q_vector *q_vector = vsi->q_vectors[q_idx];
+
+		if (q_vector->rx.ring || q_vector->tx.ring)
+			napi_disable(&q_vector->napi);
+	}
 }
 
 /**

drivers/net/ethernet/intel/igb/e1000_defines.h

···
 
 #define I210_RXPBSIZE_DEFAULT		0x000000A2 /* RXPBSIZE default */
 #define I210_RXPBSIZE_MASK		0x0000003F
+#define I210_RXPBSIZE_PB_30KB		0x0000001E
 #define I210_RXPBSIZE_PB_32KB		0x00000020
 #define I210_TXPBSIZE_DEFAULT		0x04000014 /* TXPBSIZE default */
 #define I210_TXPBSIZE_MASK		0xC0FFFFFF

drivers/net/ethernet/intel/igb/igb.h

···
 	/* OS defined structs */
 	struct pci_dev *pdev;
 
-	spinlock_t stats64_lock;
+	struct mutex stats64_lock;
 	struct rtnl_link_stats64 stats64;
 
 	/* structs defined in e1000_hw.h */

drivers/net/ethernet/intel/igb/igb_ethtool.c

···
 	int i, j;
 	char *p;
 
-	spin_lock(&adapter->stats64_lock);
+	mutex_lock(&adapter->stats64_lock);
 	igb_update_stats(adapter);
 
 	for (i = 0; i < IGB_GLOBAL_STATS_LEN; i++) {
···
 		} while (u64_stats_fetch_retry_irq(&ring->rx_syncp, start));
 		i += IGB_RX_QUEUE_STATS_LEN;
 	}
-	spin_unlock(&adapter->stats64_lock);
+	mutex_unlock(&adapter->stats64_lock);
 }
 
 static void igb_get_strings(struct net_device *netdev, u32 stringset, u8 *data)

drivers/net/ethernet/intel/igb/igb_main.c

···
 
 		val = rd32(E1000_RXPBS);
 		val &= ~I210_RXPBSIZE_MASK;
-		val |= I210_RXPBSIZE_PB_32KB;
+		val |= I210_RXPBSIZE_PB_30KB;
 		wr32(E1000_RXPBS, val);
 
 		/* Section 8.12.9 states that MAX_TPKT_SIZE from DTXMXPKTSZ
···
 	del_timer_sync(&adapter->phy_info_timer);
 
 	/* record the stats before reset*/
-	spin_lock(&adapter->stats64_lock);
+	mutex_lock(&adapter->stats64_lock);
 	igb_update_stats(adapter);
-	spin_unlock(&adapter->stats64_lock);
+	mutex_unlock(&adapter->stats64_lock);
 
 	adapter->link_speed = 0;
 	adapter->link_duplex = 0;
···
 	adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
 
 	spin_lock_init(&adapter->nfc_lock);
-	spin_lock_init(&adapter->stats64_lock);
+	mutex_init(&adapter->stats64_lock);
 #ifdef CONFIG_PCI_IOV
 	switch (hw->mac.type) {
 	case e1000_82576:
···
 		}
 	}
 
-	spin_lock(&adapter->stats64_lock);
+	mutex_lock(&adapter->stats64_lock);
 	igb_update_stats(adapter);
-	spin_unlock(&adapter->stats64_lock);
+	mutex_unlock(&adapter->stats64_lock);
 
 	for (i = 0; i < adapter->num_tx_queues; i++) {
 		struct igb_ring *tx_ring = adapter->tx_ring[i];
···
 {
 	struct igb_adapter *adapter = netdev_priv(netdev);
 
-	spin_lock(&adapter->stats64_lock);
+	mutex_lock(&adapter->stats64_lock);
 	igb_update_stats(adapter);
 	memcpy(stats, &adapter->stats64, sizeof(*stats));
-	spin_unlock(&adapter->stats64_lock);
+	mutex_unlock(&adapter->stats64_lock);
 }
 
 /**
···
 	rtnl_unlock();
 
 #ifdef CONFIG_PM
-	retval = pci_save_state(pdev);
-	if (retval)
-		return retval;
+	if (!runtime) {
+		retval = pci_save_state(pdev);
+		if (retval)
+			return retval;
+	}
 #endif
 
 	status = rd32(E1000_STATUS);

drivers/net/ethernet/intel/ixgbe/ixgbe.h

···
 #include <linux/aer.h>
 #include <linux/if_vlan.h>
 #include <linux/jiffies.h>
+#include <linux/phy.h>
 
 #include <linux/timecounter.h>
 #include <linux/net_tstamp.h>
···
 	struct net_device *netdev;
 	struct bpf_prog *xdp_prog;
 	struct pci_dev *pdev;
+	struct mii_bus *mii_bus;
 
 	unsigned long state;
 

drivers/net/ethernet/intel/ixgbe/ixgbe_main.c

···
 #include "ixgbe.h"
 #include "ixgbe_common.h"
 #include "ixgbe_dcb_82599.h"
+#include "ixgbe_phy.h"
 #include "ixgbe_sriov.h"
 #include "ixgbe_model.h"
 #include "ixgbe_txrx_common.h"
···
 	u16 value;
 	int rc;
 
+	if (adapter->mii_bus) {
+		int regnum = addr;
+
+		if (devad != MDIO_DEVAD_NONE)
+			regnum |= (devad << 16) | MII_ADDR_C45;
+
+		return mdiobus_read(adapter->mii_bus, prtad, regnum);
+	}
+
 	if (prtad != hw->phy.mdio.prtad)
 		return -EINVAL;
 	rc = hw->phy.ops.read_reg(hw, addr, devad, &value);
···
 {
 	struct ixgbe_adapter *adapter = netdev_priv(netdev);
 	struct ixgbe_hw *hw = &adapter->hw;
+
+	if (adapter->mii_bus) {
+		int regnum = addr;
+
+		if (devad != MDIO_DEVAD_NONE)
+			regnum |= (devad << 16) | MII_ADDR_C45;
+
+		return mdiobus_write(adapter->mii_bus, prtad, regnum, value);
+	}
 
 	if (prtad != hw->phy.mdio.prtad)
 		return -EINVAL;
···
 			IXGBE_LINK_SPEED_10GB_FULL | IXGBE_LINK_SPEED_1GB_FULL,
 			true);
 
+	ixgbe_mii_bus_init(hw);
+
 	return 0;
 
 err_register:
···
 	set_bit(__IXGBE_REMOVING, &adapter->state);
 	cancel_work_sync(&adapter->service_task);
 
+	if (adapter->mii_bus)
+		mdiobus_unregister(adapter->mii_bus);
 
 #ifdef CONFIG_IXGBE_DCA
 	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {

drivers/net/ethernet/intel/ixgbe/ixgbe_phy.c

···
 
 #include <linux/pci.h>
 #include <linux/delay.h>
+#include <linux/iopoll.h>
 #include <linux/sched.h>
 
 #include "ixgbe.h"
···
 	}
 
 	return status;
+}
+
+#define IXGBE_HW_READ_REG(addr) IXGBE_READ_REG(hw, addr)
+
+/**
+ *  ixgbe_msca_cmd - Write the command register and poll for completion/timeout
+ *  @hw: pointer to hardware structure
+ *  @cmd: command register value to write
+ **/
+static s32 ixgbe_msca_cmd(struct ixgbe_hw *hw, u32 cmd)
+{
+	IXGBE_WRITE_REG(hw, IXGBE_MSCA, cmd);
+
+	return readx_poll_timeout(IXGBE_HW_READ_REG, IXGBE_MSCA, cmd,
+				  !(cmd & IXGBE_MSCA_MDI_COMMAND), 10,
+				  10 * IXGBE_MDIO_COMMAND_TIMEOUT);
+}
+
+/**
+ *  ixgbe_mii_bus_read_generic - Read a clause 22/45 register with gssr flags
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ *  @gssr: semaphore flags to acquire
+ **/
+static s32 ixgbe_mii_bus_read_generic(struct ixgbe_hw *hw, int addr,
+				      int regnum, u32 gssr)
+{
+	u32 hwaddr, cmd;
+	s32 data;
+
+	if (hw->mac.ops.acquire_swfw_sync(hw, gssr))
+		return -EBUSY;
+
+	hwaddr = addr << IXGBE_MSCA_PHY_ADDR_SHIFT;
+	if (regnum & MII_ADDR_C45) {
+		hwaddr |= regnum & GENMASK(21, 0);
+		cmd = hwaddr | IXGBE_MSCA_ADDR_CYCLE | IXGBE_MSCA_MDI_COMMAND;
+	} else {
+		hwaddr |= (regnum & GENMASK(5, 0)) << IXGBE_MSCA_DEV_TYPE_SHIFT;
+		cmd = hwaddr | IXGBE_MSCA_OLD_PROTOCOL |
+			IXGBE_MSCA_READ_AUTOINC | IXGBE_MSCA_MDI_COMMAND;
+	}
+
+	data = ixgbe_msca_cmd(hw, cmd);
+	if (data < 0)
+		goto mii_bus_read_done;
+
+	/* For a clause 45 access the address cycle just completed, we still
+	 * need to do the read command, otherwise just get the data
+	 */
+	if (!(regnum & MII_ADDR_C45))
+		goto do_mii_bus_read;
+
+	cmd = hwaddr | IXGBE_MSCA_READ | IXGBE_MSCA_MDI_COMMAND;
+	data = ixgbe_msca_cmd(hw, cmd);
+	if (data < 0)
+		goto mii_bus_read_done;
+
+do_mii_bus_read:
+	data = IXGBE_READ_REG(hw, IXGBE_MSRWD);
+	data = (data >> IXGBE_MSRWD_READ_DATA_SHIFT) & GENMASK(16, 0);
+
+mii_bus_read_done:
+	hw->mac.ops.release_swfw_sync(hw, gssr);
+	return data;
+}
+
+/**
+ *  ixgbe_mii_bus_write_generic - Write a clause 22/45 register with gssr flags
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ *  @val: value to write
+ *  @gssr: semaphore flags to acquire
+ **/
+static s32 ixgbe_mii_bus_write_generic(struct ixgbe_hw *hw, int addr,
+				       int regnum, u16 val, u32 gssr)
+{
+	u32 hwaddr, cmd;
+	s32 err;
+
+	if (hw->mac.ops.acquire_swfw_sync(hw, gssr))
+		return -EBUSY;
+
+	IXGBE_WRITE_REG(hw, IXGBE_MSRWD, (u32)val);
+
+	hwaddr = addr << IXGBE_MSCA_PHY_ADDR_SHIFT;
+	if (regnum & MII_ADDR_C45) {
+		hwaddr |= regnum & GENMASK(21, 0);
+		cmd = hwaddr | IXGBE_MSCA_ADDR_CYCLE | IXGBE_MSCA_MDI_COMMAND;
+	} else {
+		hwaddr |= (regnum & GENMASK(5, 0)) << IXGBE_MSCA_DEV_TYPE_SHIFT;
+		cmd = hwaddr | IXGBE_MSCA_OLD_PROTOCOL | IXGBE_MSCA_WRITE |
+			IXGBE_MSCA_MDI_COMMAND;
+	}
+
+	/* For clause 45 this is an address cycle, for clause 22 this is the
+	 * entire transaction
+	 */
+	err = ixgbe_msca_cmd(hw, cmd);
+	if (err < 0 || !(regnum & MII_ADDR_C45))
+		goto mii_bus_write_done;
+
+	cmd = hwaddr | IXGBE_MSCA_WRITE | IXGBE_MSCA_MDI_COMMAND;
+	err = ixgbe_msca_cmd(hw, cmd);
+
+mii_bus_write_done:
+	hw->mac.ops.release_swfw_sync(hw, gssr);
+	return err;
+}
+
+/**
+ *  ixgbe_mii_bus_read - Read a clause 22/45 register
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ **/
+static s32 ixgbe_mii_bus_read(struct mii_bus *bus, int addr, int regnum)
+{
+	struct ixgbe_adapter *adapter = bus->priv;
+	struct ixgbe_hw *hw = &adapter->hw;
+	u32 gssr = hw->phy.phy_semaphore_mask;
+
+	return ixgbe_mii_bus_read_generic(hw, addr, regnum, gssr);
+}
+
+/**
+ *  ixgbe_mii_bus_write - Write a clause 22/45 register
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ *  @val: value to write
+ **/
+static s32 ixgbe_mii_bus_write(struct mii_bus *bus, int addr, int regnum,
+			       u16 val)
+{
+	struct ixgbe_adapter *adapter = bus->priv;
+	struct ixgbe_hw *hw = &adapter->hw;
+	u32 gssr = hw->phy.phy_semaphore_mask;
+
+	return ixgbe_mii_bus_write_generic(hw, addr, regnum, val, gssr);
+}
+
+/**
+ *  ixgbe_x550em_a_mii_bus_read - Read a clause 22/45 register on x550em_a
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ **/
+static s32 ixgbe_x550em_a_mii_bus_read(struct mii_bus *bus, int addr,
+				       int regnum)
+{
+	struct ixgbe_adapter *adapter = bus->priv;
+	struct ixgbe_hw *hw = &adapter->hw;
+	u32 gssr = hw->phy.phy_semaphore_mask;
+
+	gssr |= IXGBE_GSSR_TOKEN_SM | IXGBE_GSSR_PHY0_SM;
+	return ixgbe_mii_bus_read_generic(hw, addr, regnum, gssr);
+}
+
+/**
+ *  ixgbe_x550em_a_mii_bus_write - Write a clause 22/45 register on x550em_a
+ *  @hw: pointer to hardware structure
+ *  @addr: address
+ *  @regnum: register number
+ *  @val: value to write
+ **/
+static s32 ixgbe_x550em_a_mii_bus_write(struct mii_bus *bus, int addr,
+					int regnum, u16 val)
+{
+	struct ixgbe_adapter *adapter = bus->priv;
+	struct ixgbe_hw *hw = &adapter->hw;
+	u32 gssr = hw->phy.phy_semaphore_mask;
+
+	gssr |= IXGBE_GSSR_TOKEN_SM | IXGBE_GSSR_PHY0_SM;
+	return ixgbe_mii_bus_write_generic(hw, addr, regnum, val, gssr);
+}
+
+/**
+ * ixgbe_get_first_secondary_devfn - get first device downstream of root port
+ * @devfn: PCI_DEVFN of root port on domain 0, bus 0
+ *
+ * Returns pci_dev pointer to PCI_DEVFN(0, 0) on subordinate side of root
+ * on domain 0, bus 0, devfn = 'devfn'
+ **/
+static struct pci_dev *ixgbe_get_first_secondary_devfn(unsigned int devfn)
+{
+	struct pci_dev *rp_pdev;
+	int bus;
+
+	rp_pdev = pci_get_domain_bus_and_slot(0, 0, devfn);
+	if (rp_pdev && rp_pdev->subordinate) {
+		bus = rp_pdev->subordinate->number;
+		return pci_get_domain_bus_and_slot(0, bus, 0);
+	}
+
+	return NULL;
+}
+
+/**
+ * ixgbe_x550em_a_has_mii - is this the first ixgbe x550em_a PCI function?
+ * @hw: pointer to hardware structure
+ *
+ * Returns true if hw points to lowest numbered PCI B:D.F x550_em_a device in
+ * the SoC.  There are up to 4 MACs sharing a single MDIO bus on the x550em_a,
+ * but we only want to register one MDIO bus.
+ **/
+static bool ixgbe_x550em_a_has_mii(struct ixgbe_hw *hw)
+{
+	struct ixgbe_adapter *adapter = hw->back;
+	struct pci_dev *pdev = adapter->pdev;
+	struct pci_dev *func0_pdev;
+
+	/* For the C3000 family of SoCs (x550em_a) the internal ixgbe devices
+	 * are always downstream of root ports @ 0000:00:16.0 & 0000:00:17.0
+	 * It's not valid for function 0 to be disabled and function 1 is up,
+	 * so the lowest numbered ixgbe dev will be device 0 function 0 on one
+	 * of those two root ports
+	 */
+	func0_pdev = ixgbe_get_first_secondary_devfn(PCI_DEVFN(0x16, 0));
+	if (func0_pdev) {
+		if (func0_pdev == pdev)
+			return true;
+		else
+			return false;
+	}
+	func0_pdev = ixgbe_get_first_secondary_devfn(PCI_DEVFN(0x17, 0));
+	if (func0_pdev == pdev)
+		return true;
+
+	return false;
+}
+
+/**
+ * ixgbe_mii_bus_init - mii_bus structure setup
+ * @hw: pointer to hardware structure
+ *
+ * Returns 0 on success, negative on failure
+ *
+ * ixgbe_mii_bus_init initializes a mii_bus structure in adapter
+ **/
+s32 ixgbe_mii_bus_init(struct ixgbe_hw *hw)
+{
+	struct ixgbe_adapter *adapter = hw->back;
+	struct pci_dev *pdev = adapter->pdev;
+	struct device *dev = &adapter->netdev->dev;
+	struct mii_bus *bus;
+
+	adapter->mii_bus = devm_mdiobus_alloc(dev);
+	if (!adapter->mii_bus)
+		return -ENOMEM;
+
+	bus = adapter->mii_bus;
+
+	switch (hw->device_id) {
+	/* C3000 SoCs */
+	case IXGBE_DEV_ID_X550EM_A_KR:
+	case IXGBE_DEV_ID_X550EM_A_KR_L:
+	case IXGBE_DEV_ID_X550EM_A_SFP_N:
+	case IXGBE_DEV_ID_X550EM_A_SGMII:
+	case IXGBE_DEV_ID_X550EM_A_SGMII_L:
+	case IXGBE_DEV_ID_X550EM_A_10G_T:
+	case IXGBE_DEV_ID_X550EM_A_SFP:
+	case IXGBE_DEV_ID_X550EM_A_1G_T:
+	case IXGBE_DEV_ID_X550EM_A_1G_T_L:
+		if (!ixgbe_x550em_a_has_mii(hw))
+			goto ixgbe_no_mii_bus;
+		bus->read = &ixgbe_x550em_a_mii_bus_read;
+		bus->write = &ixgbe_x550em_a_mii_bus_write;
+		break;
+	default:
+		bus->read = &ixgbe_mii_bus_read;
+		bus->write = &ixgbe_mii_bus_write;
+		break;
+	}
+
+	/* Use the position of the device in the PCI hierarchy as the id */
+	snprintf(bus->id, MII_BUS_ID_SIZE, "%s-mdio-%s", ixgbe_driver_name,
+		 pci_name(pdev));
+
+	bus->name = "ixgbe-mdio";
+	bus->priv = adapter;
+	bus->parent = dev;
+	bus->phy_mask = GENMASK(31, 0);
+
+	/* Support clause 22/45 natively.  ixgbe_probe() sets MDIO_EMULATE_C22
+	 * unfortunately that causes some clause 22 frames to be sent with
+	 * clause 45 addressing.  We don't want that.
+	 */
+	hw->phy.mdio.mode_support = MDIO_SUPPORTS_C45 | MDIO_SUPPORTS_C22;
+
+	return mdiobus_register(bus);
+
+ixgbe_no_mii_bus:
+	devm_mdiobus_free(dev, bus);
+	adapter->mii_bus = NULL;
+	return -ENODEV;
 }
 
 /**

drivers/net/ethernet/intel/ixgbe/ixgbe_phy.h

···
 /* SFP+ SFF-8472 Compliance code */
 #define IXGBE_SFF_SFF_8472_UNSUP      0x00
 
+s32 ixgbe_mii_bus_init(struct ixgbe_hw *hw);
+
 s32 ixgbe_identify_phy_generic(struct ixgbe_hw *hw);
 s32 ixgbe_reset_phy_generic(struct ixgbe_hw *hw);
 s32 ixgbe_read_phy_reg_generic(struct ixgbe_hw *hw, u32 reg_addr,