From 7829451c695e5b795fef95d72cd258e78d52f62d Mon Sep 17 00:00:00 2001 From: Hariprasad Shenai Date: Thu, 11 Aug 2016 21:06:23 +0530 Subject: [PATCH] cxgb4: Add control net_device for configuring PCIe VF Issue: For instance, the current APIs assume a 1-to-1 mapping of Network Ports, Physical Functions and the SR-IOV Virtual Functions of those Physical Functions. This is not the case with our cards where any Virtual Function can be hooked up to any Port -- or any number of Ports the current Linux APIs also assume only 1 Network Interface/Port can be accessed per Virtual Function. Another issue is that these APIs assume that the Administrative Driver is attached to the Physical Function Associated with a Virtual Function. This is not the case with our card where all administration is performed by a Driver which is not attached to any of the Physical Functions which have SR-IOV PCI Capabilities. Another consequence of these assumptions is the inability to utilize all of the cards SR-IOV resources. For instance, our cards have SR-IOV Capabilities on Physical Functions 0..3 and the administrative Driver attaches to Physical Function 4. Each of the Physical Functions 0..3 can support up to 16 Virtual Functions. With the current Linux APIs, a 2-Port card would only be able to use the Virtual Functions on Physical Function 0..1 and not allow the Virtual Functions on Physical Functions 2..3 to be used since there are no Ports 2..3 on a 2-Port card. Fix: Since the control node is always the netdevice for all VF ACL commands. Created a dummy netdevice for each Physical Function from 0 to 3 through which one could control their VFs. The device won't be associated with any port, since it doesn't need to transmit/receive. Its purely used for VF management purpose only. The device will be registered only when VF for a particular PF is configured using PCI sysfs interface and unregistered while pci_disable_sriov() for the PF is called. Signed-off-by: Hariprasad Shenai Signed-off-by: David S. Miller --- .../net/ethernet/chelsio/cxgb4/cxgb4_main.c | 121 +++++++++++++++--- 1 file changed, 103 insertions(+), 18 deletions(-) diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c index c45de49dc963..b7e932ec6894 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c @@ -3139,6 +3139,24 @@ static const struct net_device_ops cxgb4_netdev_ops = { }; +static const struct net_device_ops cxgb4_mgmt_netdev_ops = { +}; + +static void get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) +{ + struct adapter *adapter = netdev2adap(dev); + + strlcpy(info->driver, cxgb4_driver_name, sizeof(info->driver)); + strlcpy(info->version, cxgb4_driver_version, + sizeof(info->version)); + strlcpy(info->bus_info, pci_name(adapter->pdev), + sizeof(info->bus_info)); +} + +static const struct ethtool_ops cxgb4_mgmt_ethtool_ops = { + .get_drvinfo = get_drvinfo, +}; + void t4_fatal_err(struct adapter *adap) { t4_set_reg_field(adap, SGE_CONTROL_A, GLOBALENABLE_F, 0); @@ -4836,19 +4854,12 @@ static int get_chip_type(struct pci_dev *pdev, u32 pl_rev) #ifdef CONFIG_PCI_IOV static int cxgb4_iov_configure(struct pci_dev *pdev, int num_vfs) { + struct adapter *adap = pci_get_drvdata(pdev); int err = 0; int current_vfs = pci_num_vf(pdev); u32 pcie_fw; - void __iomem *regs; - regs = pci_ioremap_bar(pdev, 0); - if (!regs) { - dev_err(&pdev->dev, "cannot map device registers\n"); - return -ENOMEM; - } - - pcie_fw = readl(regs + PCIE_FW_A); - iounmap(regs); + pcie_fw = readl(adap->regs + PCIE_FW_A); /* Check if cxgb4 is the MASTER and fw is initialized */ if (!(pcie_fw & PCIE_FW_INIT_F) || !(pcie_fw & PCIE_FW_MASTER_VLD_F) || @@ -4875,6 +4886,8 @@ static int cxgb4_iov_configure(struct pci_dev *pdev, int num_vfs) */ if (!num_vfs) { pci_disable_sriov(pdev); + if (adap->port[0]->reg_state == NETREG_REGISTERED) + unregister_netdev(adap->port[0]); return num_vfs; } @@ -4882,6 +4895,12 @@ static int cxgb4_iov_configure(struct pci_dev *pdev, int num_vfs) err = pci_enable_sriov(pdev, num_vfs); if (err) return err; + + if (adap->port[0]->reg_state == NETREG_UNINITIALIZED) { + err = register_netdev(adap->port[0]); + if (err < 0) + pr_info("Unable to register VF mgmt netdev\n"); + } } return num_vfs; } @@ -4893,9 +4912,14 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) struct port_info *pi; bool highdma = false; struct adapter *adapter = NULL; + struct net_device *netdev; +#ifdef CONFIG_PCI_IOV + char name[IFNAMSIZ]; +#endif void __iomem *regs; u32 whoami, pl_rev; enum chip_type chip; + static int adap_idx = 1; printk_once(KERN_INFO "%s - version %s\n", DRV_DESC, DRV_VERSION); @@ -4930,7 +4954,9 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) func = CHELSIO_CHIP_VERSION(chip) <= CHELSIO_T5 ? SOURCEPF_G(whoami) : T6_SOURCEPF_G(whoami); if (func != ent->driver_data) { +#ifndef CONFIG_PCI_IOV iounmap(regs); +#endif pci_disable_device(pdev); pci_save_state(pdev); /* to restore SR-IOV later */ goto sriov; @@ -4962,6 +4988,7 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) err = -ENOMEM; goto out_unmap_bar0; } + adap_idx++; adapter->workq = create_singlethread_workqueue("cxgb4"); if (!adapter->workq) { @@ -5048,8 +5075,6 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) T6_STATMODE_V(0))); for_each_port(adapter, i) { - struct net_device *netdev; - netdev = alloc_etherdev_mq(sizeof(struct port_info), MAX_ETH_QSETS); if (!netdev) { @@ -5217,6 +5242,7 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) attach_ulds(adapter); print_adapter_info(adapter); + return 0; sriov: #ifdef CONFIG_PCI_IOV @@ -5230,9 +5256,58 @@ sriov: "instantiated %u virtual functions\n", num_vf[func]); } -#endif + + adapter = kzalloc(sizeof(*adapter), GFP_KERNEL); + if (!adapter) { + err = -ENOMEM; + goto free_pci_region; + } + + snprintf(name, IFNAMSIZ, "mgmtpf%d%d", adap_idx, func); + netdev = alloc_netdev(0, name, NET_NAME_UNKNOWN, ether_setup); + if (!netdev) { + err = -ENOMEM; + goto free_adapter; + } + + adapter->pdev = pdev; + adapter->pdev_dev = &pdev->dev; + adapter->name = pci_name(pdev); + adapter->mbox = func; + adapter->pf = func; + adapter->regs = regs; + adapter->mbox_log = kzalloc(sizeof(*adapter->mbox_log) + + (sizeof(struct mbox_cmd) * + T4_OS_LOG_MBOX_CMDS), + GFP_KERNEL); + if (!adapter->mbox_log) { + err = -ENOMEM; + goto free_netdevice; + } + pi = netdev_priv(netdev); + pi->adapter = adapter; + SET_NETDEV_DEV(netdev, &pdev->dev); + pci_set_drvdata(pdev, adapter); + + adapter->port[0] = netdev; + netdev->netdev_ops = &cxgb4_mgmt_netdev_ops; + netdev->ethtool_ops = &cxgb4_mgmt_ethtool_ops; + return 0; + free_netdevice: + free_netdev(adapter->port[0]); + free_adapter: + kfree(adapter); + free_pci_region: + iounmap(regs); + pci_disable_sriov(pdev); + pci_release_regions(pdev); + return err; +#else + return 0; +#endif + out_free_dev: free_some_resources(adapter); out_unmap_bar: @@ -5258,12 +5333,12 @@ static void remove_one(struct pci_dev *pdev) { struct adapter *adapter = pci_get_drvdata(pdev); -#ifdef CONFIG_PCI_IOV - pci_disable_sriov(pdev); + if (!adapter) { + pci_release_regions(pdev); + return; + } -#endif - - if (adapter) { + if (adapter->pf == 4) { int i; /* Tear down per-adapter Work Queue first since it can contain @@ -5312,8 +5387,18 @@ static void remove_one(struct pci_dev *pdev) kfree(adapter->mbox_log); synchronize_rcu(); kfree(adapter); - } else + } +#ifdef CONFIG_PCI_IOV + else { + if (adapter->port[0]->reg_state == NETREG_REGISTERED) + unregister_netdev(adapter->port[0]); + free_netdev(adapter->port[0]); + iounmap(adapter->regs); + kfree(adapter); + pci_disable_sriov(pdev); pci_release_regions(pdev); + } +#endif } static struct pci_driver cxgb4_driver = {