From 77972b55fb9d35d4a6b0abca99abffaa4ec6a85b Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Tue, 22 Sep 2020 09:29:31 +0200 Subject: Revert "ravb: Fixed to be able to unload modules" This reverts commit 1838d6c62f57836639bd3d83e7855e0ee4f6defc. This commit moved the ravb_mdio_init() call (and thus the of_mdiobus_register() call) from the ravb_probe() to the ravb_open() call. This causes a regression during system resume (s2idle/s2ram), as new PHY devices cannot be bound while suspended. During boot, the Micrel PHY is detected like this: Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=228) ravb e6800000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off During system suspend, (A) defer_all_probes is set to true, and (B) usermodehelper_disabled is set to UMH_DISABLED, to avoid drivers being probed while suspended. A. If CONFIG_MODULES=n, phy_device_register() calling device_add() merely adds the device, but does not probe it yet, as really_probe() returns early due to defer_all_probes being set: dpm_resume+0x128/0x4f8 device_resume+0xcc/0x1b0 dpm_run_callback+0x74/0x340 ravb_resume+0x190/0x1b8 ravb_open+0x84/0x770 of_mdiobus_register+0x1e0/0x468 of_mdiobus_register_phy+0x1b8/0x250 of_mdiobus_phy_device_register+0x178/0x1e8 phy_device_register+0x114/0x1b8 device_add+0x3d4/0x798 bus_probe_device+0x98/0xa0 device_initial_probe+0x10/0x18 __device_attach+0xe4/0x140 bus_for_each_drv+0x64/0xc8 __device_attach_driver+0xb8/0xe0 driver_probe_device.part.11+0xc4/0xd8 really_probe+0x32c/0x3b8 Later, phy_attach_direct() notices no PHY driver has been bound, and falls back to the Generic PHY, leading to degraded operation: Generic PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Generic PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=POLL) ravb e6800000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off B. If CONFIG_MODULES=y, request_module() returns early with -EBUSY due to UMH_DISABLED, and MDIO initialization fails completely: mdio_bus e6800000.ethernet-ffffffff:00: error -16 loading PHY driver module for ID 0x00221622 ravb e6800000.ethernet eth0: failed to initialize MDIO PM: dpm_run_callback(): ravb_resume+0x0/0x1b8 returns -16 PM: Device e6800000.ethernet failed to resume: error -16 Ignoring -EBUSY in phy_request_driver_module(), like was done for -ENOENT in commit 21e194425abd65b5 ("net: phy: fix issue with loading PHY driver w/o initramfs"), would makes it fall back to the Generic PHY, like in the CONFIG_MODULES=n case. Signed-off-by: Geert Uytterhoeven Cc: stable@vger.kernel.org Reviewed-by: Sergei Shtylyov Signed-off-by: David S. Miller --- drivers/net/ethernet/renesas/ravb_main.c | 110 +++++++++++++++---------------- 1 file changed, 55 insertions(+), 55 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index df89d09b253e..99f7aae102ce 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -1342,51 +1342,6 @@ static inline int ravb_hook_irq(unsigned int irq, irq_handler_t handler, return error; } -/* MDIO bus init function */ -static int ravb_mdio_init(struct ravb_private *priv) -{ - struct platform_device *pdev = priv->pdev; - struct device *dev = &pdev->dev; - int error; - - /* Bitbang init */ - priv->mdiobb.ops = &bb_ops; - - /* MII controller setting */ - priv->mii_bus = alloc_mdio_bitbang(&priv->mdiobb); - if (!priv->mii_bus) - return -ENOMEM; - - /* Hook up MII support for ethtool */ - priv->mii_bus->name = "ravb_mii"; - priv->mii_bus->parent = dev; - snprintf(priv->mii_bus->id, MII_BUS_ID_SIZE, "%s-%x", - pdev->name, pdev->id); - - /* Register MDIO bus */ - error = of_mdiobus_register(priv->mii_bus, dev->of_node); - if (error) - goto out_free_bus; - - return 0; - -out_free_bus: - free_mdio_bitbang(priv->mii_bus); - return error; -} - -/* MDIO bus release function */ -static int ravb_mdio_release(struct ravb_private *priv) -{ - /* Unregister mdio bus */ - mdiobus_unregister(priv->mii_bus); - - /* Free bitbang info */ - free_mdio_bitbang(priv->mii_bus); - - return 0; -} - /* Network device open function for Ethernet AVB */ static int ravb_open(struct net_device *ndev) { @@ -1395,13 +1350,6 @@ static int ravb_open(struct net_device *ndev) struct device *dev = &pdev->dev; int error; - /* MDIO bus init */ - error = ravb_mdio_init(priv); - if (error) { - netdev_err(ndev, "failed to initialize MDIO\n"); - return error; - } - napi_enable(&priv->napi[RAVB_BE]); napi_enable(&priv->napi[RAVB_NC]); @@ -1479,7 +1427,6 @@ out_free_irq: out_napi_off: napi_disable(&priv->napi[RAVB_NC]); napi_disable(&priv->napi[RAVB_BE]); - ravb_mdio_release(priv); return error; } @@ -1789,8 +1736,6 @@ static int ravb_close(struct net_device *ndev) ravb_ring_free(ndev, RAVB_BE); ravb_ring_free(ndev, RAVB_NC); - ravb_mdio_release(priv); - return 0; } @@ -1942,6 +1887,51 @@ static const struct net_device_ops ravb_netdev_ops = { .ndo_set_features = ravb_set_features, }; +/* MDIO bus init function */ +static int ravb_mdio_init(struct ravb_private *priv) +{ + struct platform_device *pdev = priv->pdev; + struct device *dev = &pdev->dev; + int error; + + /* Bitbang init */ + priv->mdiobb.ops = &bb_ops; + + /* MII controller setting */ + priv->mii_bus = alloc_mdio_bitbang(&priv->mdiobb); + if (!priv->mii_bus) + return -ENOMEM; + + /* Hook up MII support for ethtool */ + priv->mii_bus->name = "ravb_mii"; + priv->mii_bus->parent = dev; + snprintf(priv->mii_bus->id, MII_BUS_ID_SIZE, "%s-%x", + pdev->name, pdev->id); + + /* Register MDIO bus */ + error = of_mdiobus_register(priv->mii_bus, dev->of_node); + if (error) + goto out_free_bus; + + return 0; + +out_free_bus: + free_mdio_bitbang(priv->mii_bus); + return error; +} + +/* MDIO bus release function */ +static int ravb_mdio_release(struct ravb_private *priv) +{ + /* Unregister mdio bus */ + mdiobus_unregister(priv->mii_bus); + + /* Free bitbang info */ + free_mdio_bitbang(priv->mii_bus); + + return 0; +} + static const struct of_device_id ravb_match_table[] = { { .compatible = "renesas,etheravb-r8a7790", .data = (void *)RCAR_GEN2 }, { .compatible = "renesas,etheravb-r8a7794", .data = (void *)RCAR_GEN2 }, @@ -2184,6 +2174,13 @@ static int ravb_probe(struct platform_device *pdev) eth_hw_addr_random(ndev); } + /* MDIO bus init */ + error = ravb_mdio_init(priv); + if (error) { + dev_err(&pdev->dev, "failed to initialize MDIO\n"); + goto out_dma_free; + } + netif_napi_add(ndev, &priv->napi[RAVB_BE], ravb_poll, 64); netif_napi_add(ndev, &priv->napi[RAVB_NC], ravb_poll, 64); @@ -2205,6 +2202,8 @@ static int ravb_probe(struct platform_device *pdev) out_napi_del: netif_napi_del(&priv->napi[RAVB_NC]); netif_napi_del(&priv->napi[RAVB_BE]); + ravb_mdio_release(priv); +out_dma_free: dma_free_coherent(ndev->dev.parent, priv->desc_bat_size, priv->desc_bat, priv->desc_bat_dma); @@ -2236,6 +2235,7 @@ static int ravb_remove(struct platform_device *pdev) unregister_netdev(ndev); netif_napi_del(&priv->napi[RAVB_NC]); netif_napi_del(&priv->napi[RAVB_BE]); + ravb_mdio_release(priv); pm_runtime_disable(&pdev->dev); free_netdev(ndev); platform_set_drvdata(pdev, NULL); -- cgit From f9317ae5523f99999fb54c513ebabbb2bc887ddf Mon Sep 17 00:00:00 2001 From: Hauke Mehrtens Date: Tue, 22 Sep 2020 23:41:12 +0200 Subject: net: lantiq: Add locking for TX DMA channel The TX DMA channel data is accessed by the xrx200_start_xmit() and the xrx200_tx_housekeeping() function from different threads. Make sure the accesses are synchronized by acquiring the netif_tx_lock() in the xrx200_tx_housekeeping() function too. This lock is acquired by the kernel before calling xrx200_start_xmit(). Signed-off-by: Hauke Mehrtens Signed-off-by: David S. Miller --- drivers/net/ethernet/lantiq_xrx200.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c index 635ff3a5dcfb..51ed8a54d380 100644 --- a/drivers/net/ethernet/lantiq_xrx200.c +++ b/drivers/net/ethernet/lantiq_xrx200.c @@ -245,6 +245,7 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget) int pkts = 0; int bytes = 0; + netif_tx_lock(net_dev); while (pkts < budget) { struct ltq_dma_desc *desc = &ch->dma.desc_base[ch->tx_free]; @@ -268,6 +269,7 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget) net_dev->stats.tx_bytes += bytes; netdev_completed_queue(ch->priv->net_dev, pkts, bytes); + netif_tx_unlock(net_dev); if (netif_queue_stopped(net_dev)) netif_wake_queue(net_dev); -- cgit From 7241c5a697479c7d0c5a96595822cdab750d41ae Mon Sep 17 00:00:00 2001 From: Voon Weifeng Date: Wed, 23 Sep 2020 16:56:14 +0800 Subject: net: stmmac: removed enabling eee in EEE set callback EEE should be only be enabled during stmmac_mac_link_up() when the link are up and being set up properly. set_eee should only do settings configuration and disabling the eee. Without this fix, turning on EEE using ethtool will return "Operation not supported". This is due to the driver is in a dead loop waiting for eee to be advertised in the for eee to be activated but the driver will only configure the EEE advertisement after the eee is activated. Ethtool should only return "Operation not supported" if there is no EEE capbility in the MAC controller. Fixes: 8a7493e58ad6 ("net: stmmac: Fix a race in EEE enable callback") Signed-off-by: Voon Weifeng Acked-by: Mark Gross Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 15 ++++----------- 1 file changed, 4 insertions(+), 11 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c index ac5e8cc5fb9f..430a4b32ec1e 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c @@ -675,23 +675,16 @@ static int stmmac_ethtool_op_set_eee(struct net_device *dev, struct stmmac_priv *priv = netdev_priv(dev); int ret; - if (!edata->eee_enabled) { + if (!priv->dma_cap.eee) + return -EOPNOTSUPP; + + if (!edata->eee_enabled) stmmac_disable_eee_mode(priv); - } else { - /* We are asking for enabling the EEE but it is safe - * to verify all by invoking the eee_init function. - * In case of failure it will return an error. - */ - edata->eee_enabled = stmmac_eee_init(priv); - if (!edata->eee_enabled) - return -EOPNOTSUPP; - } ret = phylink_ethtool_set_eee(priv->phylink, edata); if (ret) return ret; - priv->eee_enabled = edata->eee_enabled; priv->tx_lpi_timer = edata->tx_lpi_timer; return 0; } -- cgit From f68910a8056f9451ee9fe7e1b962f7d90d326ad3 Mon Sep 17 00:00:00 2001 From: Luo bin Date: Thu, 24 Sep 2020 09:31:51 +0800 Subject: hinic: fix wrong return value of mac-set cmd It should also be regarded as an error when hw return status=4 for PF's setting mac cmd. Only if PF return status=4 to VF should this cmd be taken special treatment. Fixes: 7dd29ee12865 ("hinic: add sriov feature support") Signed-off-by: Luo bin Signed-off-by: David S. Miller --- drivers/net/ethernet/huawei/hinic/hinic_port.c | 6 +++--- drivers/net/ethernet/huawei/hinic/hinic_sriov.c | 12 ++---------- 2 files changed, 5 insertions(+), 13 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/huawei/hinic/hinic_port.c b/drivers/net/ethernet/huawei/hinic/hinic_port.c index 02cd635d6914..eb97f2d6b1ad 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_port.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_port.c @@ -58,9 +58,9 @@ static int change_mac(struct hinic_dev *nic_dev, const u8 *addr, sizeof(port_mac_cmd), &port_mac_cmd, &out_size); if (err || out_size != sizeof(port_mac_cmd) || - (port_mac_cmd.status && - port_mac_cmd.status != HINIC_PF_SET_VF_ALREADY && - port_mac_cmd.status != HINIC_MGMT_STATUS_EXIST)) { + (port_mac_cmd.status && + (port_mac_cmd.status != HINIC_PF_SET_VF_ALREADY || !HINIC_IS_VF(hwif)) && + port_mac_cmd.status != HINIC_MGMT_STATUS_EXIST)) { dev_err(&pdev->dev, "Failed to change MAC, err: %d, status: 0x%x, out size: 0x%x\n", err, port_mac_cmd.status, out_size); return -EFAULT; diff --git a/drivers/net/ethernet/huawei/hinic/hinic_sriov.c b/drivers/net/ethernet/huawei/hinic/hinic_sriov.c index 4d63680f2143..f8a26459ff65 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_sriov.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_sriov.c @@ -38,8 +38,7 @@ static int hinic_set_mac(struct hinic_hwdev *hwdev, const u8 *mac_addr, err = hinic_port_msg_cmd(hwdev, HINIC_PORT_CMD_SET_MAC, &mac_info, sizeof(mac_info), &mac_info, &out_size); if (err || out_size != sizeof(mac_info) || - (mac_info.status && mac_info.status != HINIC_PF_SET_VF_ALREADY && - mac_info.status != HINIC_MGMT_STATUS_EXIST)) { + (mac_info.status && mac_info.status != HINIC_MGMT_STATUS_EXIST)) { dev_err(&hwdev->func_to_io.hwif->pdev->dev, "Failed to set MAC, err: %d, status: 0x%x, out size: 0x%x\n", err, mac_info.status, out_size); return -EIO; @@ -503,8 +502,7 @@ struct hinic_sriov_info *hinic_get_sriov_info_by_pcidev(struct pci_dev *pdev) static int hinic_check_mac_info(u8 status, u16 vlan_id) { - if ((status && status != HINIC_MGMT_STATUS_EXIST && - status != HINIC_PF_SET_VF_ALREADY) || + if ((status && status != HINIC_MGMT_STATUS_EXIST) || (vlan_id & CHECK_IPSU_15BIT && status == HINIC_MGMT_STATUS_EXIST)) return -EINVAL; @@ -546,12 +544,6 @@ static int hinic_update_mac(struct hinic_hwdev *hwdev, u8 *old_mac, return -EINVAL; } - if (mac_info.status == HINIC_PF_SET_VF_ALREADY) { - dev_warn(&hwdev->hwif->pdev->dev, - "PF has already set VF MAC. Ignore update operation\n"); - return HINIC_PF_SET_VF_ALREADY; - } - if (mac_info.status == HINIC_MGMT_STATUS_EXIST) dev_warn(&hwdev->hwif->pdev->dev, "MAC is repeated. Ignore update operation\n"); -- cgit From 0eb11dfe224f9bf0f7d66a68500a43a9b9c83b02 Mon Sep 17 00:00:00 2001 From: Wang Qing Date: Thu, 24 Sep 2020 14:50:24 +0800 Subject: net/ethernet/broadcom: fix spelling typo Modify the comment typo: "compliment" -> "complement". Signed-off-by: Wang Qing Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h index bfc0e45d4a2b..5caa75b41b73 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h @@ -284,12 +284,12 @@ #define CCM_REG_GR_ARB_TYPE 0xd015c /* [RW 2] Load (FIC0) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed; that the Store channel priority is - the compliment to 4 of the rest priorities - Aggregation channel; Load + the complement to 4 of the rest priorities - Aggregation channel; Load (FIC0) channel and Load (FIC1). */ #define CCM_REG_GR_LD0_PR 0xd0164 /* [RW 2] Load (FIC1) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed; that the Store channel priority is - the compliment to 4 of the rest priorities - Aggregation channel; Load + the complement to 4 of the rest priorities - Aggregation channel; Load (FIC0) channel and Load (FIC1). */ #define CCM_REG_GR_LD1_PR 0xd0168 /* [RW 2] General flags index. */ @@ -4489,11 +4489,11 @@ #define TCM_REG_GR_ARB_TYPE 0x50114 /* [RW 2] Load (FIC0) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Store channel is the - compliment of the other 3 groups. */ + complement of the other 3 groups. */ #define TCM_REG_GR_LD0_PR 0x5011c /* [RW 2] Load (FIC1) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Store channel is the - compliment of the other 3 groups. */ + complement of the other 3 groups. */ #define TCM_REG_GR_LD1_PR 0x50120 /* [RW 4] The number of double REG-pairs; loaded from the STORM context and sent to STORM; for a specific connection type. The double REG-pairs are @@ -5020,11 +5020,11 @@ #define UCM_REG_GR_ARB_TYPE 0xe0144 /* [RW 2] Load (FIC0) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Store channel group is - compliment to the others. */ + complement to the others. */ #define UCM_REG_GR_LD0_PR 0xe014c /* [RW 2] Load (FIC1) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Store channel group is - compliment to the others. */ + complement to the others. */ #define UCM_REG_GR_LD1_PR 0xe0150 /* [RW 2] The queue index for invalidate counter flag decision. */ #define UCM_REG_INV_CFLG_Q 0xe00e4 @@ -5523,11 +5523,11 @@ #define XCM_REG_GR_ARB_TYPE 0x2020c /* [RW 2] Load (FIC0) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Channel group is the - compliment of the other 3 groups. */ + complement of the other 3 groups. */ #define XCM_REG_GR_LD0_PR 0x20214 /* [RW 2] Load (FIC1) channel group priority. The lowest priority is 0; the highest priority is 3. It is supposed that the Channel group is the - compliment of the other 3 groups. */ + complement of the other 3 groups. */ #define XCM_REG_GR_LD1_PR 0x20218 /* [RW 1] Input nig0 Interface enable. If 0 - the valid input is disregarded; acknowledge output is deasserted; all other signals are -- cgit From 1ec8e74855588cecb2620b28b877c08f45765374 Mon Sep 17 00:00:00 2001 From: Jamie Iles Date: Thu, 24 Sep 2020 15:56:45 +0100 Subject: net/fsl: quieten expected MDIO access failures MDIO reads can happen during PHY probing, and printing an error with dev_err can result in a large number of error messages during device probe. On a platform with a serial console this can result in excessively long boot times in a way that looks like an infinite loop when multiple busses are present. Since 0f183fd151c (net/fsl: enable extended scanning in xgmac_mdio) we perform more scanning so there are potentially more failures. Reduce the logging level to dev_dbg which is consistent with the Freescale enetc driver. Cc: Jeremy Linton Signed-off-by: Jamie Iles Reviewed-by: Andrew Lunn Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/xgmac_mdio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/freescale/xgmac_mdio.c b/drivers/net/ethernet/freescale/xgmac_mdio.c index 98be51d8b08c..bfa2826c5545 100644 --- a/drivers/net/ethernet/freescale/xgmac_mdio.c +++ b/drivers/net/ethernet/freescale/xgmac_mdio.c @@ -229,7 +229,7 @@ static int xgmac_mdio_read(struct mii_bus *bus, int phy_id, int regnum) /* Return all Fs if nothing was there */ if ((xgmac_read32(®s->mdio_stat, endian) & MDIO_STAT_RD_ER) && !priv->has_a011043) { - dev_err(&bus->dev, + dev_dbg(&bus->dev, "Error while reading PHY%d reg at %d.%hhu\n", phy_id, dev_addr, regnum); return 0xffff; -- cgit From 75598a8fc0e0dff2aa5d46c62531b36a595f1d4f Mon Sep 17 00:00:00 2001 From: Sylwester Dziedziuch Date: Wed, 2 Sep 2020 12:54:59 +0000 Subject: iavf: Fix incorrect adapter get in iavf_resume When calling iavf_resume there was a crash because wrong function was used to get iavf_adapter and net_device pointers. Changed how iavf_resume is getting iavf_adapter and net_device pointers from pci_dev. Fixes: 5eae00c57f5e ("i40evf: main driver core") Signed-off-by: Sylwester Dziedziuch Reviewed-by: Aleksandr Loktionov Tested-by: Aaron Brown Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index d870343cf689..cf539db79af9 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -3806,8 +3806,8 @@ static int __maybe_unused iavf_suspend(struct device *dev_d) static int __maybe_unused iavf_resume(struct device *dev_d) { struct pci_dev *pdev = to_pci_dev(dev_d); - struct iavf_adapter *adapter = pci_get_drvdata(pdev); - struct net_device *netdev = adapter->netdev; + struct net_device *netdev = pci_get_drvdata(pdev); + struct iavf_adapter *adapter = netdev_priv(netdev); u32 err; pci_set_master(pdev); -- cgit From 466e4392920083c58c7ee8905795453d62c4add7 Mon Sep 17 00:00:00 2001 From: Anirudh Venkataramanan Date: Wed, 2 Sep 2020 08:53:45 -0700 Subject: ice: Fix call trace on suspend It appears that the ice_suspend flow is missing a call to pci_save_state and this is triggering the message "State of device not saved by ice_suspend" and a call trace. Fix it. Fixes: 769c500dcc1e ("ice: Add advanced power mgmt for WoL") Signed-off-by: Anirudh Venkataramanan Tested-by: Aaron Brown Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ice/ice_main.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 4634b48949bb..954f11f86f50 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -4522,6 +4522,7 @@ static int __maybe_unused ice_suspend(struct device *dev) } ice_clear_interrupt_scheme(pf); + pci_save_state(pdev); pci_wake_from_d3(pdev, pf->wol_ena); pci_set_power_state(pdev, PCI_D3hot); return 0; -- cgit From 135f4b9e9340dadb78e9737bb4eb9817b9c89dac Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 2 Sep 2020 08:53:46 -0700 Subject: ice: fix memory leak if register_netdev_fails The ice_setup_pf_sw function can cause a memory leak if register_netdev fails, due to accidentally failing to free the VSI rings. Fix the memory leak by using ice_vsi_release, ensuring we actually go through the full teardown process. This should be safe even if the netdevice is not registered because we will have set the netdev pointer to NULL, ensuring ice_vsi_release won't call unregister_netdev. An alternative fix would be moving management of the PF VSI netdev into the main VSI setup code. This is complicated and likely requires significant refactor in how we manage VSIs Fixes: 3a858ba392c3 ("ice: Add support for VSI allocation and deallocation") Signed-off-by: Jacob Keller Tested-by: Aaron Brown Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ice/ice_lib.c | 6 +++--- drivers/net/ethernet/intel/ice/ice_lib.h | 6 ------ drivers/net/ethernet/intel/ice/ice_main.c | 13 +++---------- 3 files changed, 6 insertions(+), 19 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index f2682776f8c8..65cc78279501 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -246,7 +246,7 @@ static int ice_get_free_slot(void *array, int size, int curr) * ice_vsi_delete - delete a VSI from the switch * @vsi: pointer to VSI being removed */ -void ice_vsi_delete(struct ice_vsi *vsi) +static void ice_vsi_delete(struct ice_vsi *vsi) { struct ice_pf *pf = vsi->back; struct ice_vsi_ctx *ctxt; @@ -313,7 +313,7 @@ static void ice_vsi_free_arrays(struct ice_vsi *vsi) * * Returns 0 on success, negative on failure */ -int ice_vsi_clear(struct ice_vsi *vsi) +static int ice_vsi_clear(struct ice_vsi *vsi) { struct ice_pf *pf = NULL; struct device *dev; @@ -563,7 +563,7 @@ static int ice_vsi_get_qs(struct ice_vsi *vsi) * ice_vsi_put_qs - Release queues from VSI to PF * @vsi: the VSI that is going to release queues */ -void ice_vsi_put_qs(struct ice_vsi *vsi) +static void ice_vsi_put_qs(struct ice_vsi *vsi) { struct ice_pf *pf = vsi->back; int i; diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h index 981f3a156c24..3da17895a2b1 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.h +++ b/drivers/net/ethernet/intel/ice/ice_lib.h @@ -45,10 +45,6 @@ int ice_cfg_vlan_pruning(struct ice_vsi *vsi, bool ena, bool vlan_promisc); void ice_cfg_sw_lldp(struct ice_vsi *vsi, bool tx, bool create); -void ice_vsi_delete(struct ice_vsi *vsi); - -int ice_vsi_clear(struct ice_vsi *vsi); - #ifdef CONFIG_DCB int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc); #endif /* CONFIG_DCB */ @@ -79,8 +75,6 @@ bool ice_is_reset_in_progress(unsigned long *state); void ice_write_qrxflxp_cntxt(struct ice_hw *hw, u16 pf_q, u32 rxdid, u32 prio); -void ice_vsi_put_qs(struct ice_vsi *vsi); - void ice_vsi_dis_irq(struct ice_vsi *vsi); void ice_vsi_free_irq(struct ice_vsi *vsi); diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 954f11f86f50..54a7f55eb8c1 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -3169,10 +3169,8 @@ static int ice_setup_pf_sw(struct ice_pf *pf) return -EBUSY; vsi = ice_pf_vsi_setup(pf, pf->hw.port_info); - if (!vsi) { - status = -ENOMEM; - goto unroll_vsi_setup; - } + if (!vsi) + return -ENOMEM; status = ice_cfg_netdev(vsi); if (status) { @@ -3219,12 +3217,7 @@ unroll_napi_add: } unroll_vsi_setup: - if (vsi) { - ice_vsi_free_q_vectors(vsi); - ice_vsi_delete(vsi); - ice_vsi_put_qs(vsi); - ice_vsi_clear(vsi); - } + ice_vsi_release(vsi); return status; } -- cgit From f6a07271bb1535d9549380461437cc48d9e19958 Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 2 Sep 2020 08:53:47 -0700 Subject: ice: fix memory leak in ice_vsi_setup During ice_vsi_setup, if ice_cfg_vsi_lan fails, it does not properly release memory associated with the VSI rings. If we had used devres allocations for the rings, this would be ok. However, we use kzalloc and kfree_rcu for these ring structures. Using the correct label to cleanup the rings during ice_vsi_setup highlights an issue in the ice_vsi_clear_rings function: it can leave behind stale ring pointers in the q_vectors structure. When releasing rings, we must also ensure that no q_vector associated with the VSI will point to this ring again. To resolve this, loop over all q_vectors and release their ring mapping. Because we are about to free all rings, no q_vector should remain pointing to any of the rings in this VSI. Fixes: 5513b920a4f7 ("ice: Update Tx scheduler tree for VSI multi-Tx queue support") Signed-off-by: Jacob Keller Tested-by: Aaron Brown Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ice/ice_lib.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index 65cc78279501..ebbb8f54871c 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -1196,6 +1196,18 @@ static void ice_vsi_clear_rings(struct ice_vsi *vsi) { int i; + /* Avoid stale references by clearing map from vector to ring */ + if (vsi->q_vectors) { + ice_for_each_q_vector(vsi, i) { + struct ice_q_vector *q_vector = vsi->q_vectors[i]; + + if (q_vector) { + q_vector->tx.ring = NULL; + q_vector->rx.ring = NULL; + } + } + } + if (vsi->tx_rings) { for (i = 0; i < vsi->alloc_txq; i++) { if (vsi->tx_rings[i]) { @@ -2291,7 +2303,7 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi, if (status) { dev_err(dev, "VSI %d failed lan queue config, error %s\n", vsi->vsi_num, ice_stat_str(status)); - goto unroll_vector_base; + goto unroll_clear_rings; } /* Add switch rule to drop all Tx Flow Control Frames, of look up -- cgit From ac322f86b56cb99d1c4224c209095aa67647c967 Mon Sep 17 00:00:00 2001 From: Wong Vee Khee Date: Fri, 25 Sep 2020 17:54:06 +0800 Subject: net: stmmac: Fix clock handling on remove path While unloading the dwmac-intel driver, clk_disable_unprepare() is being called twice in stmmac_dvr_remove() and intel_eth_pci_remove(). This causes kernel panic on the second call. Removing the second call of clk_disable_unprepare() in intel_eth_pci_remove(). Fixes: 09f012e64e4b ("stmmac: intel: Fix clock handling on error and remove paths") Cc: Andy Shevchenko Reviewed-by: Voon Weifeng Signed-off-by: Wong Vee Khee Reviewed-by: Andy Shevchenko Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c index 2ac9dfb3462c..9e6d60e75f85 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c @@ -653,7 +653,6 @@ static void intel_eth_pci_remove(struct pci_dev *pdev) pci_free_irq_vectors(pdev); - clk_disable_unprepare(priv->plat->stmmac_clk); clk_unregister_fixed_rate(priv->plat->stmmac_clk); pcim_iounmap_regions(pdev, BIT(0)); -- cgit From 4663ff60257aec4ee1e2e969a7c046f0aff35ab8 Mon Sep 17 00:00:00 2001 From: Ivan Khoronzhuk Date: Fri, 25 Sep 2020 15:44:39 +0300 Subject: net: ethernet: cavium: octeon_mgmt: use phy_start and phy_stop To start also "phy state machine", with UP state as it should be, the phy_start() has to be used, in another case machine even is not triggered. After this change negotiation is supposed to be triggered by SM workqueue. It's not correct usage, but it appears after the following patch, so add it as a fix. Fixes: 74a992b3598a ("net: phy: add phy_check_link_status") Signed-off-by: Ivan Khoronzhuk Reviewed-by: Andrew Lunn Signed-off-by: David S. Miller --- drivers/net/ethernet/cavium/octeon/octeon_mgmt.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c index 3e17ce0d2314..6cb2162a75d4 100644 --- a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c +++ b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c @@ -1219,7 +1219,7 @@ static int octeon_mgmt_open(struct net_device *netdev) */ if (netdev->phydev) { netif_carrier_off(netdev); - phy_start_aneg(netdev->phydev); + phy_start(netdev->phydev); } netif_wake_queue(netdev); @@ -1247,8 +1247,10 @@ static int octeon_mgmt_stop(struct net_device *netdev) napi_disable(&p->napi); netif_stop_queue(netdev); - if (netdev->phydev) + if (netdev->phydev) { + phy_stop(netdev->phydev); phy_disconnect(netdev->phydev); + } netif_carrier_off(netdev); -- cgit From 72e27c38abb3dc8da7c05f829d44b3f3c46d06da Mon Sep 17 00:00:00 2001 From: Ioana Ciornei Date: Fri, 25 Sep 2020 17:35:30 +0300 Subject: dpaa2-eth: fix command version for Tx shaping When adding the support for TBF offload, the improper command version was added even though the command format is for the V2 of dpni_set_tx_shaping(). This does not affect the functionality of TBF since the only change between these two versions is the addition of the exceeded parameters which are not used in TBF. Still, fix the bug so that we keep things in sync. Fixes: 39344a89623d ("dpaa2-eth: add API for Tx shaping") Signed-off-by: Ioana Ciornei Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/dpaa2/dpni-cmd.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/freescale/dpaa2/dpni-cmd.h b/drivers/net/ethernet/freescale/dpaa2/dpni-cmd.h index 593e3812af93..3c06f5fb5759 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpni-cmd.h +++ b/drivers/net/ethernet/freescale/dpaa2/dpni-cmd.h @@ -11,9 +11,11 @@ #define DPNI_VER_MAJOR 7 #define DPNI_VER_MINOR 0 #define DPNI_CMD_BASE_VERSION 1 +#define DPNI_CMD_2ND_VERSION 2 #define DPNI_CMD_ID_OFFSET 4 #define DPNI_CMD(id) (((id) << DPNI_CMD_ID_OFFSET) | DPNI_CMD_BASE_VERSION) +#define DPNI_CMD_V2(id) (((id) << DPNI_CMD_ID_OFFSET) | DPNI_CMD_2ND_VERSION) #define DPNI_CMDID_OPEN DPNI_CMD(0x801) #define DPNI_CMDID_CLOSE DPNI_CMD(0x800) @@ -45,7 +47,7 @@ #define DPNI_CMDID_SET_MAX_FRAME_LENGTH DPNI_CMD(0x216) #define DPNI_CMDID_GET_MAX_FRAME_LENGTH DPNI_CMD(0x217) #define DPNI_CMDID_SET_LINK_CFG DPNI_CMD(0x21A) -#define DPNI_CMDID_SET_TX_SHAPING DPNI_CMD(0x21B) +#define DPNI_CMDID_SET_TX_SHAPING DPNI_CMD_V2(0x21B) #define DPNI_CMDID_SET_MCAST_PROMISC DPNI_CMD(0x220) #define DPNI_CMDID_GET_MCAST_PROMISC DPNI_CMD(0x221) -- cgit From 059432495e209279bae12db3d2b0bc8c8fe987bb Mon Sep 17 00:00:00 2001 From: Igor Russkikh Date: Fri, 25 Sep 2020 23:27:35 +0300 Subject: net: atlantic: fix build when object tree is separate Driver subfolder files refer parent folder includes in an absolute manner. Makefile contains a -I for this, but apparently that does not work if object tree is separated. Adding srctree to fix that. Signed-off-by: Igor Russkikh Signed-off-by: David S. Miller --- drivers/net/ethernet/aquantia/atlantic/Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/aquantia/atlantic/Makefile b/drivers/net/ethernet/aquantia/atlantic/Makefile index 130a105d03f3..8ebcc68e807f 100644 --- a/drivers/net/ethernet/aquantia/atlantic/Makefile +++ b/drivers/net/ethernet/aquantia/atlantic/Makefile @@ -8,7 +8,7 @@ obj-$(CONFIG_AQTION) += atlantic.o -ccflags-y += -I$(src) +ccflags-y += -I$(srctree)/$(src) atlantic-objs := aq_main.o \ aq_nic.o \ @@ -33,4 +33,4 @@ atlantic-objs := aq_main.o \ atlantic-$(CONFIG_MACSEC) += aq_macsec.o -atlantic-$(CONFIG_PTP_1588_CLOCK) += aq_ptp.o \ No newline at end of file +atlantic-$(CONFIG_PTP_1588_CLOCK) += aq_ptp.o -- cgit From 72865028582a678be1e05240e55d452e5c258eca Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Sun, 27 Sep 2020 09:42:11 +0300 Subject: mlxsw: spectrum_acl: Fix mlxsw_sp_acl_tcam_group_add()'s error path If mlxsw_sp_acl_tcam_group_id_get() fails, the mutex initialized earlier is not destroyed. Fix this by initializing the mutex after calling the function. This is symmetric to mlxsw_sp_acl_tcam_group_del(). Fixes: 5ec2ee28d27b ("mlxsw: spectrum_acl: Introduce a mutex to guard region list updates") Signed-off-by: Ido Schimmel Reviewed-by: Jiri Pirko Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 5c020403342f..7cccc41dd69c 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -292,13 +292,14 @@ mlxsw_sp_acl_tcam_group_add(struct mlxsw_sp_acl_tcam *tcam, int err; group->tcam = tcam; - mutex_init(&group->lock); INIT_LIST_HEAD(&group->region_list); err = mlxsw_sp_acl_tcam_group_id_get(tcam, &group->id); if (err) return err; + mutex_init(&group->lock); + return 0; } -- cgit From 709a16be0593c08190982cfbdca6df95e6d5823b Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Sun, 27 Sep 2020 19:44:29 +0200 Subject: r8169: fix RTL8168f/RTL8411 EPHY config Mistakenly bit 2 was set instead of bit 3 as in the vendor driver. Fixes: a7a92cf81589 ("r8169: sync PCIe PHY init with vendor driver 8.047.01") Signed-off-by: Heiner Kallweit Signed-off-by: David S. Miller --- drivers/net/ethernet/realtek/r8169_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index fc9e6626db55..5e867da1aad3 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -2904,7 +2904,7 @@ static void rtl_hw_start_8168f_1(struct rtl8169_private *tp) { 0x08, 0x0001, 0x0002 }, { 0x09, 0x0000, 0x0080 }, { 0x19, 0x0000, 0x0224 }, - { 0x00, 0x0000, 0x0004 }, + { 0x00, 0x0000, 0x0008 }, { 0x0c, 0x3df0, 0x0200 }, }; @@ -2921,7 +2921,7 @@ static void rtl_hw_start_8411(struct rtl8169_private *tp) { 0x06, 0x00c0, 0x0020 }, { 0x0f, 0xffff, 0x5200 }, { 0x19, 0x0000, 0x0224 }, - { 0x00, 0x0000, 0x0004 }, + { 0x00, 0x0000, 0x0008 }, { 0x0c, 0x3df0, 0x0200 }, }; -- cgit From eff7423365a6938d2d34dbce989febed2ae1f957 Mon Sep 17 00:00:00 2001 From: Taehee Yoo Date: Fri, 25 Sep 2020 18:13:12 +0000 Subject: net: core: introduce struct netdev_nested_priv for nested interface infrastructure Functions related to nested interface infrastructure such as netdev_walk_all_{ upper | lower }_dev() pass both private functions and "data" pointer to handle their own things. At this point, the data pointer type is void *. In order to make it easier to expand common variables and functions, this new netdev_nested_priv structure is added. In the following patch, a new member variable will be added into this struct to fix the lockdep issue. Signed-off-by: Taehee Yoo Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 37 +++++++++++++++------- drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 24 +++++++------- .../net/ethernet/mellanox/mlxsw/spectrum_router.c | 11 +++++-- .../ethernet/mellanox/mlxsw/spectrum_switchdev.c | 10 ++++-- drivers/net/ethernet/rocker/rocker_main.c | 9 ++++-- 5 files changed, 60 insertions(+), 31 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 2f8a4cfc5fa1..86ca8b9ea1b8 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -5396,9 +5396,10 @@ static int ixgbe_fwd_ring_up(struct ixgbe_adapter *adapter, return err; } -static int ixgbe_macvlan_up(struct net_device *vdev, void *data) +static int ixgbe_macvlan_up(struct net_device *vdev, + struct netdev_nested_priv *priv) { - struct ixgbe_adapter *adapter = data; + struct ixgbe_adapter *adapter = (struct ixgbe_adapter *)priv->data; struct ixgbe_fwd_adapter *accel; if (!netif_is_macvlan(vdev)) @@ -5415,8 +5416,12 @@ static int ixgbe_macvlan_up(struct net_device *vdev, void *data) static void ixgbe_configure_dfwd(struct ixgbe_adapter *adapter) { + struct netdev_nested_priv priv = { + .data = (void *)adapter, + }; + netdev_walk_all_upper_dev_rcu(adapter->netdev, - ixgbe_macvlan_up, adapter); + ixgbe_macvlan_up, &priv); } static void ixgbe_configure(struct ixgbe_adapter *adapter) @@ -9023,9 +9028,10 @@ static void ixgbe_set_prio_tc_map(struct ixgbe_adapter *adapter) } #endif /* CONFIG_IXGBE_DCB */ -static int ixgbe_reassign_macvlan_pool(struct net_device *vdev, void *data) +static int ixgbe_reassign_macvlan_pool(struct net_device *vdev, + struct netdev_nested_priv *priv) { - struct ixgbe_adapter *adapter = data; + struct ixgbe_adapter *adapter = (struct ixgbe_adapter *)priv->data; struct ixgbe_fwd_adapter *accel; int pool; @@ -9062,13 +9068,16 @@ static int ixgbe_reassign_macvlan_pool(struct net_device *vdev, void *data) static void ixgbe_defrag_macvlan_pools(struct net_device *dev) { struct ixgbe_adapter *adapter = netdev_priv(dev); + struct netdev_nested_priv priv = { + .data = (void *)adapter, + }; /* flush any stale bits out of the fwd bitmask */ bitmap_clear(adapter->fwd_bitmask, 1, 63); /* walk through upper devices reassigning pools */ netdev_walk_all_upper_dev_rcu(dev, ixgbe_reassign_macvlan_pool, - adapter); + &priv); } /** @@ -9242,14 +9251,18 @@ struct upper_walk_data { u8 queue; }; -static int get_macvlan_queue(struct net_device *upper, void *_data) +static int get_macvlan_queue(struct net_device *upper, + struct netdev_nested_priv *priv) { if (netif_is_macvlan(upper)) { struct ixgbe_fwd_adapter *vadapter = macvlan_accel_priv(upper); - struct upper_walk_data *data = _data; - struct ixgbe_adapter *adapter = data->adapter; - int ifindex = data->ifindex; + struct ixgbe_adapter *adapter; + struct upper_walk_data *data; + int ifindex; + data = (struct upper_walk_data *)priv->data; + ifindex = data->ifindex; + adapter = data->adapter; if (vadapter && upper->ifindex == ifindex) { data->queue = adapter->rx_ring[vadapter->rx_base_queue]->reg_idx; data->action = data->queue; @@ -9265,6 +9278,7 @@ static int handle_redirect_action(struct ixgbe_adapter *adapter, int ifindex, { struct ixgbe_ring_feature *vmdq = &adapter->ring_feature[RING_F_VMDQ]; unsigned int num_vfs = adapter->num_vfs, vf; + struct netdev_nested_priv priv; struct upper_walk_data data; struct net_device *upper; @@ -9284,8 +9298,9 @@ static int handle_redirect_action(struct ixgbe_adapter *adapter, int ifindex, data.ifindex = ifindex; data.action = 0; data.queue = 0; + priv.data = (void *)&data; if (netdev_walk_all_upper_dev_rcu(adapter->netdev, - get_macvlan_queue, &data)) { + get_macvlan_queue, &priv)) { *action = data.action; *queue = data.queue; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 4186e29119c2..f3c0e241e1b4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -3690,13 +3690,13 @@ bool mlxsw_sp_port_dev_check(const struct net_device *dev) return dev->netdev_ops == &mlxsw_sp_port_netdev_ops; } -static int mlxsw_sp_lower_dev_walk(struct net_device *lower_dev, void *data) +static int mlxsw_sp_lower_dev_walk(struct net_device *lower_dev, + struct netdev_nested_priv *priv) { - struct mlxsw_sp_port **p_mlxsw_sp_port = data; int ret = 0; if (mlxsw_sp_port_dev_check(lower_dev)) { - *p_mlxsw_sp_port = netdev_priv(lower_dev); + priv->data = (void *)netdev_priv(lower_dev); ret = 1; } @@ -3705,15 +3705,16 @@ static int mlxsw_sp_lower_dev_walk(struct net_device *lower_dev, void *data) struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev) { - struct mlxsw_sp_port *mlxsw_sp_port; + struct netdev_nested_priv priv = { + .data = NULL, + }; if (mlxsw_sp_port_dev_check(dev)) return netdev_priv(dev); - mlxsw_sp_port = NULL; - netdev_walk_all_lower_dev(dev, mlxsw_sp_lower_dev_walk, &mlxsw_sp_port); + netdev_walk_all_lower_dev(dev, mlxsw_sp_lower_dev_walk, &priv); - return mlxsw_sp_port; + return (struct mlxsw_sp_port *)priv.data; } struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev) @@ -3726,16 +3727,17 @@ struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev) struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find_rcu(struct net_device *dev) { - struct mlxsw_sp_port *mlxsw_sp_port; + struct netdev_nested_priv priv = { + .data = NULL, + }; if (mlxsw_sp_port_dev_check(dev)) return netdev_priv(dev); - mlxsw_sp_port = NULL; netdev_walk_all_lower_dev_rcu(dev, mlxsw_sp_lower_dev_walk, - &mlxsw_sp_port); + &priv); - return mlxsw_sp_port; + return (struct mlxsw_sp_port *)priv.data; } struct mlxsw_sp_port *mlxsw_sp_port_lower_dev_hold(struct net_device *dev) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c index 24f1fd1f8d56..460cb523312f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -7351,9 +7351,10 @@ int mlxsw_sp_netdevice_vrf_event(struct net_device *l3_dev, unsigned long event, return err; } -static int __mlxsw_sp_rif_macvlan_flush(struct net_device *dev, void *data) +static int __mlxsw_sp_rif_macvlan_flush(struct net_device *dev, + struct netdev_nested_priv *priv) { - struct mlxsw_sp_rif *rif = data; + struct mlxsw_sp_rif *rif = (struct mlxsw_sp_rif *)priv->data; if (!netif_is_macvlan(dev)) return 0; @@ -7364,12 +7365,16 @@ static int __mlxsw_sp_rif_macvlan_flush(struct net_device *dev, void *data) static int mlxsw_sp_rif_macvlan_flush(struct mlxsw_sp_rif *rif) { + struct netdev_nested_priv priv = { + .data = (void *)rif, + }; + if (!netif_is_macvlan_port(rif->dev)) return 0; netdev_warn(rif->dev, "Router interface is deleted. Upper macvlans will not work\n"); return netdev_walk_all_upper_dev_rcu(rif->dev, - __mlxsw_sp_rif_macvlan_flush, rif); + __mlxsw_sp_rif_macvlan_flush, &priv); } static void mlxsw_sp_rif_subport_setup(struct mlxsw_sp_rif *rif, diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c index 72912afa6f72..6501ce94ace5 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c @@ -136,9 +136,9 @@ bool mlxsw_sp_bridge_device_is_offloaded(const struct mlxsw_sp *mlxsw_sp, } static int mlxsw_sp_bridge_device_upper_rif_destroy(struct net_device *dev, - void *data) + struct netdev_nested_priv *priv) { - struct mlxsw_sp *mlxsw_sp = data; + struct mlxsw_sp *mlxsw_sp = priv->data; mlxsw_sp_rif_destroy_by_dev(mlxsw_sp, dev); return 0; @@ -147,10 +147,14 @@ static int mlxsw_sp_bridge_device_upper_rif_destroy(struct net_device *dev, static void mlxsw_sp_bridge_device_rifs_destroy(struct mlxsw_sp *mlxsw_sp, struct net_device *dev) { + struct netdev_nested_priv priv = { + .data = (void *)mlxsw_sp, + }; + mlxsw_sp_rif_destroy_by_dev(mlxsw_sp, dev); netdev_walk_all_upper_dev_rcu(dev, mlxsw_sp_bridge_device_upper_rif_destroy, - mlxsw_sp); + &priv); } static int mlxsw_sp_bridge_device_vxlan_init(struct mlxsw_sp_bridge *bridge, diff --git a/drivers/net/ethernet/rocker/rocker_main.c b/drivers/net/ethernet/rocker/rocker_main.c index 42458a46ffaf..9cc31f7e0df1 100644 --- a/drivers/net/ethernet/rocker/rocker_main.c +++ b/drivers/net/ethernet/rocker/rocker_main.c @@ -3099,9 +3099,10 @@ struct rocker_walk_data { struct rocker_port *port; }; -static int rocker_lower_dev_walk(struct net_device *lower_dev, void *_data) +static int rocker_lower_dev_walk(struct net_device *lower_dev, + struct netdev_nested_priv *priv) { - struct rocker_walk_data *data = _data; + struct rocker_walk_data *data = (struct rocker_walk_data *)priv->data; int ret = 0; if (rocker_port_dev_check_under(lower_dev, data->rocker)) { @@ -3115,6 +3116,7 @@ static int rocker_lower_dev_walk(struct net_device *lower_dev, void *_data) struct rocker_port *rocker_port_dev_lower_find(struct net_device *dev, struct rocker *rocker) { + struct netdev_nested_priv priv; struct rocker_walk_data data; if (rocker_port_dev_check_under(dev, rocker)) @@ -3122,7 +3124,8 @@ struct rocker_port *rocker_port_dev_lower_find(struct net_device *dev, data.rocker = rocker; data.port = NULL; - netdev_walk_all_lower_dev(dev, rocker_lower_dev_walk, &data); + priv.data = (void *)&data; + netdev_walk_all_lower_dev(dev, rocker_lower_dev_walk, &priv); return data.port; } -- cgit From d120c9a81e32c43cba8017dec873b6a414898716 Mon Sep 17 00:00:00 2001 From: Kevin Brace Date: Tue, 29 Sep 2020 13:09:40 -0700 Subject: via-rhine: Fix for the hardware having a reset failure after resume In rhine_resume() and rhine_suspend(), the code calls netif_running() to see if the network interface is down or not. If it is down (i.e., netif_running() returning false), they will skip any housekeeping work within the function relating to the hardware. This becomes a problem when the hardware resumes from a standby since it is counting on rhine_resume() to map its MMIO and power up rest of the hardware. Not getting its MMIO remapped and rest of the hardware powered up lead to a soft reset failure and hardware disappearance. The solution is to map its MMIO and power up rest of the hardware inside rhine_open() before soft reset is to be performed. This solution was verified on ASUS P5V800-VM mainboard's integrated Rhine-II Ethernet MAC inside VIA Technologies VT8251 South Bridge. Signed-off-by: Kevin Brace Signed-off-by: David S. Miller --- drivers/net/ethernet/via/via-rhine.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c index 803247d51fe9..a20492da3407 100644 --- a/drivers/net/ethernet/via/via-rhine.c +++ b/drivers/net/ethernet/via/via-rhine.c @@ -1706,6 +1706,8 @@ static int rhine_open(struct net_device *dev) goto out_free_ring; alloc_tbufs(dev); + enable_mmio(rp->pioaddr, rp->quirks); + rhine_power_init(dev); rhine_chip_reset(dev); rhine_task_enable(rp); init_registers(dev); -- cgit From aa15190cf2cf25ec7e6c6d7373ae3ca563d48601 Mon Sep 17 00:00:00 2001 From: Kevin Brace Date: Tue, 29 Sep 2020 13:09:41 -0700 Subject: via-rhine: VTunknown1 device is really VT8251 South Bridge The VIA Technologies VT8251 South Bridge's integrated Rhine-II Ethernet MAC comes has a PCI revision value of 0x7c. This was verified on ASUS P5V800-VM mainboard. Signed-off-by: Kevin Brace Signed-off-by: David S. Miller --- drivers/net/ethernet/via/via-rhine.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c index a20492da3407..d3a2be2e75d0 100644 --- a/drivers/net/ethernet/via/via-rhine.c +++ b/drivers/net/ethernet/via/via-rhine.c @@ -243,7 +243,7 @@ enum rhine_revs { VT8233 = 0x60, /* Integrated MAC */ VT8235 = 0x74, /* Integrated MAC */ VT8237 = 0x78, /* Integrated MAC */ - VTunknown1 = 0x7C, + VT8251 = 0x7C, /* Integrated MAC */ VT6105 = 0x80, VT6105_B0 = 0x83, VT6105L = 0x8A, -- cgit From 9f5159e89d5d1535e3da0222805c355a86d3e2dc Mon Sep 17 00:00:00 2001 From: Kevin Brace Date: Tue, 29 Sep 2020 13:09:42 -0700 Subject: via-rhine: Eliminate version information Signed-off-by: Kevin Brace Signed-off-by: David S. Miller --- drivers/net/ethernet/via/via-rhine.c | 15 --------------- 1 file changed, 15 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c index d3a2be2e75d0..8e8cfe110d95 100644 --- a/drivers/net/ethernet/via/via-rhine.c +++ b/drivers/net/ethernet/via/via-rhine.c @@ -32,8 +32,6 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #define DRV_NAME "via-rhine" -#define DRV_VERSION "1.5.1" -#define DRV_RELDATE "2010-10-09" #include @@ -117,10 +115,6 @@ static const int multicast_filter_limit = 32; #include #include -/* These identify the driver base version and may not be removed. */ -static const char version[] = - "v1.10-LK" DRV_VERSION " " DRV_RELDATE " Written by Donald Becker"; - MODULE_AUTHOR("Donald Becker "); MODULE_DESCRIPTION("VIA Rhine PCI Fast Ethernet driver"); MODULE_LICENSE("GPL"); @@ -1051,11 +1045,6 @@ static int rhine_init_one_pci(struct pci_dev *pdev, u32 quirks = 0; #endif -/* when built into the kernel, we only print version if device is found */ -#ifndef MODULE - pr_info_once("%s\n", version); -#endif - rc = pci_enable_device(pdev); if (rc) goto err_out; @@ -2296,7 +2285,6 @@ static void netdev_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *i struct device *hwdev = dev->dev.parent; strlcpy(info->driver, DRV_NAME, sizeof(info->driver)); - strlcpy(info->version, DRV_VERSION, sizeof(info->version)); strlcpy(info->bus_info, dev_name(hwdev), sizeof(info->bus_info)); } @@ -2618,9 +2606,6 @@ static int __init rhine_init(void) int ret_pci, ret_platform; /* when a module, this is printed whether or not devices are found in probe */ -#ifdef MODULE - pr_info("%s\n", version); -#endif if (dmi_check_system(rhine_dmi_table)) { /* these BIOSes fail at PXE boot if chip is in D3 */ avoid_D3 = true; -- cgit From 2b6b78e0828948fdc8374dc6fe6b2911252333c9 Mon Sep 17 00:00:00 2001 From: Kevin Brace Date: Tue, 29 Sep 2020 13:09:43 -0700 Subject: via-rhine: New device driver maintainer Signed-off-by: Kevin Brace Signed-off-by: David S. Miller --- drivers/net/ethernet/via/via-rhine.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c index 8e8cfe110d95..55b0ddab1776 100644 --- a/drivers/net/ethernet/via/via-rhine.c +++ b/drivers/net/ethernet/via/via-rhine.c @@ -2,7 +2,7 @@ /* Written 1998-2001 by Donald Becker. - Current Maintainer: Roger Luethi + Current Maintainer: Kevin Brace This software may be used and distributed according to the terms of the GNU General Public License (GPL), incorporated herein by reference. -- cgit From 0ec86e8e82b57cfb2f83a9de14732002816d3282 Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 26 Aug 2020 11:30:34 -0700 Subject: ice: increase maximum wait time for flash write commands The ice driver needs to wait for a firmware response to each command to write a block of data to the scratch area used to update the device firmware. The driver currently waits for up to 1 second for this to be returned. It turns out that firmware might take longer than 1 second to return a completion in some cases. If this happens, the flash update will fail to complete. Fix this by increasing the maximum time that the driver will wait for both writing a block of data, and for activating the new NVM bank. The timeout for an erase command is already several minutes, as the firmware had to erase the entire bank which was already expected to take a minute or more in the worst case. In the case where firmware really won't respond, we will now take longer to fail. However, this ensures that if the firmware is simply slow to respond, the flash update can still complete. This new maximum timeout should not adversely increase the update time, as the implementation for wait_event_interruptible_timeout, and should wake very soon after we get a completion event. It is better for a flash update be slow but still succeed than to fail because we gave up too quickly. Fixes: d69ea414c9b4 ("ice: implement device flash update via devlink") Signed-off-by: Jacob Keller Tested-by: Brijesh Behera Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ice/ice_fw_update.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ice/ice_fw_update.c b/drivers/net/ethernet/intel/ice/ice_fw_update.c index deaefe00c9c0..8968fdd4816b 100644 --- a/drivers/net/ethernet/intel/ice/ice_fw_update.c +++ b/drivers/net/ethernet/intel/ice/ice_fw_update.c @@ -289,7 +289,13 @@ ice_write_one_nvm_block(struct ice_pf *pf, u16 module, u32 offset, return -EIO; } - err = ice_aq_wait_for_event(pf, ice_aqc_opc_nvm_write, HZ, &event); + /* In most cases, firmware reports a write completion within a few + * milliseconds. However, it has been observed that a completion might + * take more than a second to complete in some cases. The timeout here + * is conservative and is intended to prevent failure to update when + * firmware is slow to respond. + */ + err = ice_aq_wait_for_event(pf, ice_aqc_opc_nvm_write, 15 * HZ, &event); if (err) { dev_err(dev, "Timed out waiting for firmware write completion for module 0x%02x, err %d\n", module, err); @@ -513,7 +519,7 @@ static int ice_switch_flash_banks(struct ice_pf *pf, u8 activate_flags, return -EIO; } - err = ice_aq_wait_for_event(pf, ice_aqc_opc_nvm_write_activate, HZ, + err = ice_aq_wait_for_event(pf, ice_aqc_opc_nvm_write_activate, 30 * HZ, &event); if (err) { dev_err(dev, "Timed out waiting for firmware to switch active flash banks, err %d\n", -- cgit From be49b1ad299a638e6181aefc7bdbfd621b46a7ef Mon Sep 17 00:00:00 2001 From: Jacob Keller Date: Wed, 2 Sep 2020 08:53:42 -0700 Subject: ice: preserve NVM capabilities in safe mode If the driver initializes in safe mode, it will call ice_set_safe_mode_caps. This results in clearing the capabilities structures, in order to set them up for operating in safe mode, ensuring many features are disabled. This has a side effect of also clearing the capability bits that relate to NVM update. The result is that the device driver will not indicate support for unified update, even if the firmware is capable. Fix this by adding the relevant capability fields to the list of values we preserve. To simplify the code, use a common_cap structure instead of a handful of local variables. To reduce some duplication of the capability name, introduce a couple of macros used to restore the capabilities values from the cached copy. Fixes: de9b277ee032 ("ice: Add support for unified NVM update flow capability") Signed-off-by: Jacob Keller Tested-by: Brijesh Behera Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/ice/ice_common.c | 49 ++++++++++++++++------------- 1 file changed, 27 insertions(+), 22 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c index 34abfcea9858..7db5fd977367 100644 --- a/drivers/net/ethernet/intel/ice/ice_common.c +++ b/drivers/net/ethernet/intel/ice/ice_common.c @@ -2288,26 +2288,28 @@ void ice_set_safe_mode_caps(struct ice_hw *hw) { struct ice_hw_func_caps *func_caps = &hw->func_caps; struct ice_hw_dev_caps *dev_caps = &hw->dev_caps; - u32 valid_func, rxq_first_id, txq_first_id; - u32 msix_vector_first_id, max_mtu; + struct ice_hw_common_caps cached_caps; u32 num_funcs; /* cache some func_caps values that should be restored after memset */ - valid_func = func_caps->common_cap.valid_functions; - txq_first_id = func_caps->common_cap.txq_first_id; - rxq_first_id = func_caps->common_cap.rxq_first_id; - msix_vector_first_id = func_caps->common_cap.msix_vector_first_id; - max_mtu = func_caps->common_cap.max_mtu; + cached_caps = func_caps->common_cap; /* unset func capabilities */ memset(func_caps, 0, sizeof(*func_caps)); +#define ICE_RESTORE_FUNC_CAP(name) \ + func_caps->common_cap.name = cached_caps.name + /* restore cached values */ - func_caps->common_cap.valid_functions = valid_func; - func_caps->common_cap.txq_first_id = txq_first_id; - func_caps->common_cap.rxq_first_id = rxq_first_id; - func_caps->common_cap.msix_vector_first_id = msix_vector_first_id; - func_caps->common_cap.max_mtu = max_mtu; + ICE_RESTORE_FUNC_CAP(valid_functions); + ICE_RESTORE_FUNC_CAP(txq_first_id); + ICE_RESTORE_FUNC_CAP(rxq_first_id); + ICE_RESTORE_FUNC_CAP(msix_vector_first_id); + ICE_RESTORE_FUNC_CAP(max_mtu); + ICE_RESTORE_FUNC_CAP(nvm_unified_update); + ICE_RESTORE_FUNC_CAP(nvm_update_pending_nvm); + ICE_RESTORE_FUNC_CAP(nvm_update_pending_orom); + ICE_RESTORE_FUNC_CAP(nvm_update_pending_netlist); /* one Tx and one Rx queue in safe mode */ func_caps->common_cap.num_rxq = 1; @@ -2318,22 +2320,25 @@ void ice_set_safe_mode_caps(struct ice_hw *hw) func_caps->guar_num_vsi = 1; /* cache some dev_caps values that should be restored after memset */ - valid_func = dev_caps->common_cap.valid_functions; - txq_first_id = dev_caps->common_cap.txq_first_id; - rxq_first_id = dev_caps->common_cap.rxq_first_id; - msix_vector_first_id = dev_caps->common_cap.msix_vector_first_id; - max_mtu = dev_caps->common_cap.max_mtu; + cached_caps = dev_caps->common_cap; num_funcs = dev_caps->num_funcs; /* unset dev capabilities */ memset(dev_caps, 0, sizeof(*dev_caps)); +#define ICE_RESTORE_DEV_CAP(name) \ + dev_caps->common_cap.name = cached_caps.name + /* restore cached values */ - dev_caps->common_cap.valid_functions = valid_func; - dev_caps->common_cap.txq_first_id = txq_first_id; - dev_caps->common_cap.rxq_first_id = rxq_first_id; - dev_caps->common_cap.msix_vector_first_id = msix_vector_first_id; - dev_caps->common_cap.max_mtu = max_mtu; + ICE_RESTORE_DEV_CAP(valid_functions); + ICE_RESTORE_DEV_CAP(txq_first_id); + ICE_RESTORE_DEV_CAP(rxq_first_id); + ICE_RESTORE_DEV_CAP(msix_vector_first_id); + ICE_RESTORE_DEV_CAP(max_mtu); + ICE_RESTORE_DEV_CAP(nvm_unified_update); + ICE_RESTORE_DEV_CAP(nvm_update_pending_nvm); + ICE_RESTORE_DEV_CAP(nvm_update_pending_orom); + ICE_RESTORE_DEV_CAP(nvm_update_pending_netlist); dev_caps->num_funcs = num_funcs; /* one Tx and one Rx queue per function in safe mode */ -- cgit From e154b5b70368a84a19505a0be9b0096c66562b56 Mon Sep 17 00:00:00 2001 From: Subbaraya Sundeep Date: Wed, 30 Sep 2020 21:38:27 +0530 Subject: octeontx2-af: Fix enable/disable of default NPC entries Packet replication feature present in Octeontx2 is a hardware linked list of PF and its VF interfaces so that broadcast packets are sent to all interfaces present in the list. It is driver job to add and delete a PF/VF interface to/from the list when the interface is brought up and down. This patch fixes the npc_enadis_default_entries function to handle broadcast replication properly if packet replication feature is present. Fixes: 40df309e4166 ("octeontx2-af: Support to enable/disable default MCAM entries") Signed-off-by: Subbaraya Sundeep Signed-off-by: Geetha sowjanya Signed-off-by: Sunil Goutham Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/octeontx2/af/rvu.h | 3 ++- .../net/ethernet/marvell/octeontx2/af/rvu_nix.c | 5 ++--- .../net/ethernet/marvell/octeontx2/af/rvu_npc.c | 26 ++++++++++++++++------ 3 files changed, 23 insertions(+), 11 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h index dcf25a092008..b89dde2c8b08 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu.h @@ -463,6 +463,7 @@ void rvu_nix_freemem(struct rvu *rvu); int rvu_get_nixlf_count(struct rvu *rvu); void rvu_nix_lf_teardown(struct rvu *rvu, u16 pcifunc, int blkaddr, int npalf); int nix_get_nixlf(struct rvu *rvu, u16 pcifunc, int *nixlf, int *nix_blkaddr); +int nix_update_bcast_mce_list(struct rvu *rvu, u16 pcifunc, bool add); /* NPC APIs */ int rvu_npc_init(struct rvu *rvu); @@ -477,7 +478,7 @@ void rvu_npc_disable_promisc_entry(struct rvu *rvu, u16 pcifunc, int nixlf); void rvu_npc_enable_promisc_entry(struct rvu *rvu, u16 pcifunc, int nixlf); void rvu_npc_install_bcast_match_entry(struct rvu *rvu, u16 pcifunc, int nixlf, u64 chan); -void rvu_npc_disable_bcast_entry(struct rvu *rvu, u16 pcifunc); +void rvu_npc_enable_bcast_entry(struct rvu *rvu, u16 pcifunc, bool enable); int rvu_npc_update_rxvlan(struct rvu *rvu, u16 pcifunc, int nixlf); void rvu_npc_disable_mcam_entries(struct rvu *rvu, u16 pcifunc, int nixlf); void rvu_npc_disable_default_entries(struct rvu *rvu, u16 pcifunc, int nixlf); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c index 01a793105599..0fc70824fd6b 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c @@ -17,7 +17,6 @@ #include "npc.h" #include "cgx.h" -static int nix_update_bcast_mce_list(struct rvu *rvu, u16 pcifunc, bool add); static int rvu_nix_get_bpid(struct rvu *rvu, struct nix_bp_cfg_req *req, int type, int chan_id); @@ -2020,7 +2019,7 @@ static int nix_update_mce_list(struct nix_mce_list *mce_list, return 0; } -static int nix_update_bcast_mce_list(struct rvu *rvu, u16 pcifunc, bool add) +int nix_update_bcast_mce_list(struct rvu *rvu, u16 pcifunc, bool add) { int err = 0, idx, next_idx, last_idx; struct nix_mce_list *mce_list; @@ -2065,7 +2064,7 @@ static int nix_update_bcast_mce_list(struct rvu *rvu, u16 pcifunc, bool add) /* Disable MCAM entry in NPC */ if (!mce_list->count) { - rvu_npc_disable_bcast_entry(rvu, pcifunc); + rvu_npc_enable_bcast_entry(rvu, pcifunc, false); goto end; } diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c index 0a214084406a..fbaf9bcd83f2 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c @@ -530,7 +530,7 @@ void rvu_npc_install_bcast_match_entry(struct rvu *rvu, u16 pcifunc, NIX_INTF_RX, &entry, true); } -void rvu_npc_disable_bcast_entry(struct rvu *rvu, u16 pcifunc) +void rvu_npc_enable_bcast_entry(struct rvu *rvu, u16 pcifunc, bool enable) { struct npc_mcam *mcam = &rvu->hw->mcam; int blkaddr, index; @@ -543,7 +543,7 @@ void rvu_npc_disable_bcast_entry(struct rvu *rvu, u16 pcifunc) pcifunc = pcifunc & ~RVU_PFVF_FUNC_MASK; index = npc_get_nixlf_mcam_index(mcam, pcifunc, 0, NIXLF_BCAST_ENTRY); - npc_enable_mcam_entry(rvu, mcam, blkaddr, index, false); + npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); } void rvu_npc_update_flowkey_alg_idx(struct rvu *rvu, u16 pcifunc, int nixlf, @@ -622,23 +622,35 @@ static void npc_enadis_default_entries(struct rvu *rvu, u16 pcifunc, nixlf, NIXLF_UCAST_ENTRY); npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); - /* For PF, ena/dis promisc and bcast MCAM match entries */ - if (pcifunc & RVU_PFVF_FUNC_MASK) + /* For PF, ena/dis promisc and bcast MCAM match entries. + * For VFs add/delete from bcast list when RX multicast + * feature is present. + */ + if (pcifunc & RVU_PFVF_FUNC_MASK && !rvu->hw->cap.nix_rx_multicast) return; /* For bcast, enable/disable only if it's action is not * packet replication, incase if action is replication - * then this PF's nixlf is removed from bcast replication + * then this PF/VF's nixlf is removed from bcast replication * list. */ - index = npc_get_nixlf_mcam_index(mcam, pcifunc, + index = npc_get_nixlf_mcam_index(mcam, pcifunc & ~RVU_PFVF_FUNC_MASK, nixlf, NIXLF_BCAST_ENTRY); bank = npc_get_bank(mcam, index); *(u64 *)&action = rvu_read64(rvu, blkaddr, NPC_AF_MCAMEX_BANKX_ACTION(index & (mcam->banksize - 1), bank)); - if (action.op != NIX_RX_ACTIONOP_MCAST) + + /* VFs will not have BCAST entry */ + if (action.op != NIX_RX_ACTIONOP_MCAST && + !(pcifunc & RVU_PFVF_FUNC_MASK)) { npc_enable_mcam_entry(rvu, mcam, blkaddr, index, enable); + } else { + nix_update_bcast_mce_list(rvu, pcifunc, enable); + /* Enable PF's BCAST entry for packet replication */ + rvu_npc_enable_bcast_entry(rvu, pcifunc, enable); + } + if (enable) rvu_npc_enable_promisc_entry(rvu, pcifunc, nixlf); else -- cgit From 89eae5e87b4fa799726a3e8911c90d418cb5d2b1 Mon Sep 17 00:00:00 2001 From: Geetha sowjanya Date: Wed, 30 Sep 2020 21:38:52 +0530 Subject: octeontx2-pf: Fix TCP/UDP checksum offload for IPv6 frames For TCP/UDP checksum offload feature in Octeontx2 expects L3TYPE to be set irrespective of IP header checksum is being offloaded or not. Currently for IPv6 frames L3TYPE is not being set resulting in packet drop with checksum error. This patch fixes this issue. Fixes: 3ca6c4c88 ("octeontx2-pf: Add packet transmission support") Signed-off-by: Geetha sowjanya Signed-off-by: Sunil Goutham Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c index 3a5b34a2a7a6..e46834e043be 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c @@ -524,6 +524,7 @@ static void otx2_sqe_add_hdr(struct otx2_nic *pfvf, struct otx2_snd_queue *sq, sqe_hdr->ol3type = NIX_SENDL3TYPE_IP4_CKSUM; } else if (skb->protocol == htons(ETH_P_IPV6)) { proto = ipv6_hdr(skb)->nexthdr; + sqe_hdr->ol3type = NIX_SENDL3TYPE_IP6; } if (proto == IPPROTO_TCP) -- cgit From 1ea0166da0509e987caa42c30a6a71f2c6ca1875 Mon Sep 17 00:00:00 2001 From: Hariprasad Kelam Date: Wed, 30 Sep 2020 21:39:14 +0530 Subject: octeontx2-pf: Fix the device state on error Currently in otx2_open on failure of nix_lf_start transmit queues are not stopped which are already started in link_event. Since the tx queues are not stopped network stack still try's to send the packets leading to driver crash while access the device resources. Fixes: 50fe6c02e ("octeontx2-pf: Register and handle link notifications") Signed-off-by: Hariprasad Kelam Signed-off-by: Geetha sowjanya Signed-off-by: Sunil Goutham Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c index 75a8c407e815..5d620a39ea80 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c @@ -1560,10 +1560,13 @@ int otx2_open(struct net_device *netdev) err = otx2_rxtx_enable(pf, true); if (err) - goto err_free_cints; + goto err_tx_stop_queues; return 0; +err_tx_stop_queues: + netif_tx_stop_all_queues(netdev); + netif_carrier_off(netdev); err_free_cints: otx2_free_cints(pf, qidx); vec = pci_irq_vector(pf->pdev, -- cgit From 66a5209b53418111757716d71e52727b782eabd4 Mon Sep 17 00:00:00 2001 From: Hariprasad Kelam Date: Wed, 30 Sep 2020 21:39:35 +0530 Subject: octeontx2-pf: Fix synchnorization issue in mbox Mbox implementation in octeontx2 driver has three states alloc, send and reset in mbox response. VF allocate and sends message to PF for processing, PF ACKs them back and reset the mbox memory. In some case we see synchronization issue where after msgs_acked is incremented and before mbox_reset API is called, if current execution is scheduled out and a different thread is scheduled in which checks for msgs_acked. Since the new thread sees msgs_acked == msgs_sent it will try to allocate a new message and to send a new mbox message to PF.Now if mbox_reset is scheduled in, PF will see '0' in msgs_send. This patch fixes the issue by calling mbox_reset before incrementing msgs_acked flag for last processing message and checks for valid message size. Fixes: d424b6c02 ("octeontx2-pf: Enable SRIOV and added VF mbox handling") Signed-off-by: Hariprasad Kelam Signed-off-by: Geetha sowjanya Signed-off-by: Sunil Goutham Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/octeontx2/af/mbox.c | 12 ++++++++++-- drivers/net/ethernet/marvell/octeontx2/af/mbox.h | 1 + drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c | 11 ++++++----- drivers/net/ethernet/marvell/octeontx2/nic/otx2_vf.c | 4 ++-- 4 files changed, 19 insertions(+), 9 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/marvell/octeontx2/af/mbox.c b/drivers/net/ethernet/marvell/octeontx2/af/mbox.c index 387e33fa417a..2718fe201c14 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/mbox.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/mbox.c @@ -17,7 +17,7 @@ static const u16 msgs_offset = ALIGN(sizeof(struct mbox_hdr), MBOX_MSG_ALIGN); -void otx2_mbox_reset(struct otx2_mbox *mbox, int devid) +void __otx2_mbox_reset(struct otx2_mbox *mbox, int devid) { void *hw_mbase = mbox->hwbase + (devid * MBOX_SIZE); struct otx2_mbox_dev *mdev = &mbox->dev[devid]; @@ -26,13 +26,21 @@ void otx2_mbox_reset(struct otx2_mbox *mbox, int devid) tx_hdr = hw_mbase + mbox->tx_start; rx_hdr = hw_mbase + mbox->rx_start; - spin_lock(&mdev->mbox_lock); mdev->msg_size = 0; mdev->rsp_size = 0; tx_hdr->num_msgs = 0; tx_hdr->msg_size = 0; rx_hdr->num_msgs = 0; rx_hdr->msg_size = 0; +} +EXPORT_SYMBOL(__otx2_mbox_reset); + +void otx2_mbox_reset(struct otx2_mbox *mbox, int devid) +{ + struct otx2_mbox_dev *mdev = &mbox->dev[devid]; + + spin_lock(&mdev->mbox_lock); + __otx2_mbox_reset(mbox, devid); spin_unlock(&mdev->mbox_lock); } EXPORT_SYMBOL(otx2_mbox_reset); diff --git a/drivers/net/ethernet/marvell/octeontx2/af/mbox.h b/drivers/net/ethernet/marvell/octeontx2/af/mbox.h index 6dfd0f90cd70..ab433789d2c3 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/mbox.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/mbox.h @@ -93,6 +93,7 @@ struct mbox_msghdr { }; void otx2_mbox_reset(struct otx2_mbox *mbox, int devid); +void __otx2_mbox_reset(struct otx2_mbox *mbox, int devid); void otx2_mbox_destroy(struct otx2_mbox *mbox); int otx2_mbox_init(struct otx2_mbox *mbox, void __force *hwbase, struct pci_dev *pdev, void __force *reg_base, diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c index 5d620a39ea80..2fb45670aca4 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c @@ -370,8 +370,8 @@ static int otx2_forward_vf_mbox_msgs(struct otx2_nic *pf, dst_mbox = &pf->mbox; dst_size = dst_mbox->mbox.tx_size - ALIGN(sizeof(*mbox_hdr), MBOX_MSG_ALIGN); - /* Check if msgs fit into destination area */ - if (mbox_hdr->msg_size > dst_size) + /* Check if msgs fit into destination area and has valid size */ + if (mbox_hdr->msg_size > dst_size || !mbox_hdr->msg_size) return -EINVAL; dst_mdev = &dst_mbox->mbox.dev[0]; @@ -526,10 +526,10 @@ static void otx2_pfvf_mbox_up_handler(struct work_struct *work) end: offset = mbox->rx_start + msg->next_msgoff; + if (mdev->msgs_acked == (vf_mbox->up_num_msgs - 1)) + __otx2_mbox_reset(mbox, 0); mdev->msgs_acked++; } - - otx2_mbox_reset(mbox, vf_idx); } static irqreturn_t otx2_pfvf_mbox_intr_handler(int irq, void *pf_irq) @@ -803,10 +803,11 @@ static void otx2_pfaf_mbox_handler(struct work_struct *work) msg = (struct mbox_msghdr *)(mdev->mbase + offset); otx2_process_pfaf_mbox_msg(pf, msg); offset = mbox->rx_start + msg->next_msgoff; + if (mdev->msgs_acked == (af_mbox->num_msgs - 1)) + __otx2_mbox_reset(mbox, 0); mdev->msgs_acked++; } - otx2_mbox_reset(mbox, 0); } static void otx2_handle_link_event(struct otx2_nic *pf) diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_vf.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_vf.c index 92a3db69a6cd..2f90f1721441 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_vf.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_vf.c @@ -99,10 +99,10 @@ static void otx2vf_vfaf_mbox_handler(struct work_struct *work) msg = (struct mbox_msghdr *)(mdev->mbase + offset); otx2vf_process_vfaf_mbox_msg(af_mbox->pfvf, msg); offset = mbox->rx_start + msg->next_msgoff; + if (mdev->msgs_acked == (af_mbox->num_msgs - 1)) + __otx2_mbox_reset(mbox, 0); mdev->msgs_acked++; } - - otx2_mbox_reset(mbox, 0); } static int otx2vf_process_mbox_msg_up(struct otx2_nic *vf, -- cgit From bb13a800620c2b9046187facdf7dfd45699f5a74 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Thu, 1 Oct 2020 08:44:19 +0200 Subject: r8169: fix handling ether_clk Petr reported that system freezes on r8169 driver load on a system using ether_clk. The original change was done under the assumption that the clock isn't needed for basic operations like chip register access. But obviously that was wrong. Therefore effectively revert the original change, and in addition leave the clock active when suspending and WoL is enabled. Chip may not be able to process incoming packets otherwise. Fixes: 9f0b54cd1672 ("r8169: move switching optional clock on/off to pll power functions") Reported-by: Petr Tesarik Tested-by: Petr Tesarik Signed-off-by: Heiner Kallweit Reviewed-by: Hans de Goede Signed-off-by: David S. Miller --- drivers/net/ethernet/realtek/r8169_main.c | 32 ++++++++++++++++++------------- 1 file changed, 19 insertions(+), 13 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 5e867da1aad3..4ae1a6fbf0ba 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -2239,14 +2239,10 @@ static void rtl_pll_power_down(struct rtl8169_private *tp) default: break; } - - clk_disable_unprepare(tp->clk); } static void rtl_pll_power_up(struct rtl8169_private *tp) { - clk_prepare_enable(tp->clk); - switch (tp->mac_version) { case RTL_GIGA_MAC_VER_25 ... RTL_GIGA_MAC_VER_33: case RTL_GIGA_MAC_VER_37: @@ -4826,29 +4822,39 @@ static void rtl8169_net_suspend(struct rtl8169_private *tp) #ifdef CONFIG_PM +static int rtl8169_net_resume(struct rtl8169_private *tp) +{ + rtl_rar_set(tp, tp->dev->dev_addr); + + if (tp->TxDescArray) + rtl8169_up(tp); + + netif_device_attach(tp->dev); + + return 0; +} + static int __maybe_unused rtl8169_suspend(struct device *device) { struct rtl8169_private *tp = dev_get_drvdata(device); rtnl_lock(); rtl8169_net_suspend(tp); + if (!device_may_wakeup(tp_to_dev(tp))) + clk_disable_unprepare(tp->clk); rtnl_unlock(); return 0; } -static int rtl8169_resume(struct device *device) +static int __maybe_unused rtl8169_resume(struct device *device) { struct rtl8169_private *tp = dev_get_drvdata(device); - rtl_rar_set(tp, tp->dev->dev_addr); - - if (tp->TxDescArray) - rtl8169_up(tp); + if (!device_may_wakeup(tp_to_dev(tp))) + clk_prepare_enable(tp->clk); - netif_device_attach(tp->dev); - - return 0; + return rtl8169_net_resume(tp); } static int rtl8169_runtime_suspend(struct device *device) @@ -4874,7 +4880,7 @@ static int rtl8169_runtime_resume(struct device *device) __rtl8169_set_wol(tp, tp->saved_wolopts); - return rtl8169_resume(device); + return rtl8169_net_resume(tp); } static int rtl8169_runtime_idle(struct device *device) -- cgit From ef9da46ddef071e1bbb943afbbe9b38771855554 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Thu, 1 Oct 2020 09:23:02 +0200 Subject: r8169: fix data corruption issue on RTL8402 Petr reported that after resume from suspend RTL8402 partially truncates incoming packets, and re-initializing register RxConfig before the actual chip re-initialization sequence is needed to avoid the issue. Reported-by: Petr Tesarik Proposed-by: Petr Tesarik Tested-by: Petr Tesarik Signed-off-by: Heiner Kallweit Signed-off-by: David S. Miller --- drivers/net/ethernet/realtek/r8169_main.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 4ae1a6fbf0ba..256bc4878ef1 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -4854,6 +4854,10 @@ static int __maybe_unused rtl8169_resume(struct device *device) if (!device_may_wakeup(tp_to_dev(tp))) clk_prepare_enable(tp->clk); + /* Reportedly at least Asus X453MA truncates packets otherwise */ + if (tp->mac_version == RTL_GIGA_MAC_VER_37) + rtl_init_rxcfg(tp); + return rtl8169_net_resume(tp); } -- cgit From 432161ea26d6d5e5c3f7306d9407d26ed1e1953e Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Thu, 13 Aug 2020 16:55:20 +0300 Subject: net/mlx5: Fix a race when moving command interface to polling mode As part of driver unload, it destroys the commands EQ (via FW command). As the commands EQ is destroyed, FW will not generate EQEs for any command that driver sends afterwards. Driver should poll for later commands status. Driver commands mode metadata is updated before the commands EQ is actually destroyed. This can lead for double completion handle by the driver (polling and interrupt), if a command is executed and completed by FW after the mode was changed, but before the EQ was destroyed. Fix that by using the mlx5_cmd_allowed_opcode mechanism to guarantee that only DESTROY_EQ command can be executed during this time period. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/eq.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 31ef9f8420c8..1318d774b18f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -656,8 +656,10 @@ static void destroy_async_eqs(struct mlx5_core_dev *dev) cleanup_async_eq(dev, &table->pages_eq, "pages"); cleanup_async_eq(dev, &table->async_eq, "async"); + mlx5_cmd_allowed_opcode(dev, MLX5_CMD_OP_DESTROY_EQ); mlx5_cmd_use_polling(dev); cleanup_async_eq(dev, &table->cmd_eq, "cmd"); + mlx5_cmd_allowed_opcode(dev, CMD_ALLOWED_OPCODE_ALL); mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); } -- cgit From 50b2412b7e7862c5af0cbf4b10d93bc5c712d021 Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Tue, 4 Aug 2020 10:40:21 +0300 Subject: net/mlx5: Avoid possible free of command entry while timeout comp handler Upon command completion timeout, driver simulates a forced command completion. In a rare case where real interrupt for that command arrives simultaneously, it might release the command entry while the forced handler might still access it. Fix that by adding an entry refcount, to track current amount of allowed handlers. Command entry to be released only when this refcount is decremented to zero. Command refcount is always initialized to one. For callback commands, command completion handler is the symmetric flow to decrement it. For non-callback commands, it is wait_func(). Before ringing the doorbell, increment the refcount for the real completion handler. Once the real completion handler is called, it will decrement it. For callback commands, once the delayed work is scheduled, increment the refcount. Upon callback command completion handler, we will try to cancel the timeout callback. In case of success, we need to decrement the callback refcount as it will never run. In addition, gather the entry index free and the entry free into a one flow for all command types release. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 109 +++++++++++++++++--------- 1 file changed, 71 insertions(+), 38 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index 1d91a0d0ab1d..c0055f5479ce 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -69,12 +69,10 @@ enum { MLX5_CMD_DELIVERY_STAT_CMD_DESCR_ERR = 0x10, }; -static struct mlx5_cmd_work_ent *alloc_cmd(struct mlx5_cmd *cmd, - struct mlx5_cmd_msg *in, - struct mlx5_cmd_msg *out, - void *uout, int uout_size, - mlx5_cmd_cbk_t cbk, - void *context, int page_queue) +static struct mlx5_cmd_work_ent * +cmd_alloc_ent(struct mlx5_cmd *cmd, struct mlx5_cmd_msg *in, + struct mlx5_cmd_msg *out, void *uout, int uout_size, + mlx5_cmd_cbk_t cbk, void *context, int page_queue) { gfp_t alloc_flags = cbk ? GFP_ATOMIC : GFP_KERNEL; struct mlx5_cmd_work_ent *ent; @@ -83,6 +81,7 @@ static struct mlx5_cmd_work_ent *alloc_cmd(struct mlx5_cmd *cmd, if (!ent) return ERR_PTR(-ENOMEM); + ent->idx = -EINVAL; ent->in = in; ent->out = out; ent->uout = uout; @@ -91,10 +90,16 @@ static struct mlx5_cmd_work_ent *alloc_cmd(struct mlx5_cmd *cmd, ent->context = context; ent->cmd = cmd; ent->page_queue = page_queue; + refcount_set(&ent->refcnt, 1); return ent; } +static void cmd_free_ent(struct mlx5_cmd_work_ent *ent) +{ + kfree(ent); +} + static u8 alloc_token(struct mlx5_cmd *cmd) { u8 token; @@ -109,7 +114,7 @@ static u8 alloc_token(struct mlx5_cmd *cmd) return token; } -static int alloc_ent(struct mlx5_cmd *cmd) +static int cmd_alloc_index(struct mlx5_cmd *cmd) { unsigned long flags; int ret; @@ -123,7 +128,7 @@ static int alloc_ent(struct mlx5_cmd *cmd) return ret < cmd->max_reg_cmds ? ret : -ENOMEM; } -static void free_ent(struct mlx5_cmd *cmd, int idx) +static void cmd_free_index(struct mlx5_cmd *cmd, int idx) { unsigned long flags; @@ -132,6 +137,22 @@ static void free_ent(struct mlx5_cmd *cmd, int idx) spin_unlock_irqrestore(&cmd->alloc_lock, flags); } +static void cmd_ent_get(struct mlx5_cmd_work_ent *ent) +{ + refcount_inc(&ent->refcnt); +} + +static void cmd_ent_put(struct mlx5_cmd_work_ent *ent) +{ + if (!refcount_dec_and_test(&ent->refcnt)) + return; + + if (ent->idx >= 0) + cmd_free_index(ent->cmd, ent->idx); + + cmd_free_ent(ent); +} + static struct mlx5_cmd_layout *get_inst(struct mlx5_cmd *cmd, int idx) { return cmd->cmd_buf + (idx << cmd->log_stride); @@ -219,11 +240,6 @@ static void poll_timeout(struct mlx5_cmd_work_ent *ent) ent->ret = -ETIMEDOUT; } -static void free_cmd(struct mlx5_cmd_work_ent *ent) -{ - kfree(ent); -} - static int verify_signature(struct mlx5_cmd_work_ent *ent) { struct mlx5_cmd_mailbox *next = ent->out->next; @@ -842,6 +858,7 @@ static void cb_timeout_handler(struct work_struct *work) mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); + cmd_ent_put(ent); /* for the cmd_ent_get() took on schedule delayed work */ } static void free_msg(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *msg); @@ -873,14 +890,14 @@ static void cmd_work_handler(struct work_struct *work) sem = ent->page_queue ? &cmd->pages_sem : &cmd->sem; down(sem); if (!ent->page_queue) { - alloc_ret = alloc_ent(cmd); + alloc_ret = cmd_alloc_index(cmd); if (alloc_ret < 0) { mlx5_core_err_rl(dev, "failed to allocate command entry\n"); if (ent->callback) { ent->callback(-EAGAIN, ent->context); mlx5_free_cmd_msg(dev, ent->out); free_msg(dev, ent->in); - free_cmd(ent); + cmd_ent_put(ent); } else { ent->ret = -EAGAIN; complete(&ent->done); @@ -916,8 +933,8 @@ static void cmd_work_handler(struct work_struct *work) ent->ts1 = ktime_get_ns(); cmd_mode = cmd->mode; - if (ent->callback) - schedule_delayed_work(&ent->cb_timeout_work, cb_timeout); + if (ent->callback && schedule_delayed_work(&ent->cb_timeout_work, cb_timeout)) + cmd_ent_get(ent); set_bit(MLX5_CMD_ENT_STATE_PENDING_COMP, &ent->state); /* Skip sending command to fw if internal error */ @@ -933,13 +950,10 @@ static void cmd_work_handler(struct work_struct *work) MLX5_SET(mbox_out, ent->out, syndrome, drv_synd); mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); - /* no doorbell, no need to keep the entry */ - free_ent(cmd, ent->idx); - if (ent->callback) - free_cmd(ent); return; } + cmd_ent_get(ent); /* for the _real_ FW event on completion */ /* ring doorbell after the descriptor is valid */ mlx5_core_dbg(dev, "writing 0x%x to command doorbell\n", 1 << ent->idx); wmb(); @@ -1039,11 +1053,16 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *in, if (callback && page_queue) return -EINVAL; - ent = alloc_cmd(cmd, in, out, uout, uout_size, callback, context, - page_queue); + ent = cmd_alloc_ent(cmd, in, out, uout, uout_size, + callback, context, page_queue); if (IS_ERR(ent)) return PTR_ERR(ent); + /* put for this ent is when consumed, depending on the use case + * 1) (!callback) blocking flow: by caller after wait_func completes + * 2) (callback) flow: by mlx5_cmd_comp_handler() when ent is handled + */ + ent->token = token; ent->polling = force_polling; @@ -1062,12 +1081,10 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *in, } if (callback) - goto out; + goto out; /* mlx5_cmd_comp_handler() will put(ent) */ err = wait_func(dev, ent); - if (err == -ETIMEDOUT) - goto out; - if (err == -ECANCELED) + if (err == -ETIMEDOUT || err == -ECANCELED) goto out_free; ds = ent->ts2 - ent->ts1; @@ -1085,7 +1102,7 @@ static int mlx5_cmd_invoke(struct mlx5_core_dev *dev, struct mlx5_cmd_msg *in, *status = ent->status; out_free: - free_cmd(ent); + cmd_ent_put(ent); out: return err; } @@ -1516,14 +1533,19 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force if (!forced) { mlx5_core_err(dev, "Command completion arrived after timeout (entry idx = %d).\n", ent->idx); - free_ent(cmd, ent->idx); - free_cmd(ent); + cmd_ent_put(ent); } continue; } - if (ent->callback) - cancel_delayed_work(&ent->cb_timeout_work); + if (ent->callback && cancel_delayed_work(&ent->cb_timeout_work)) + cmd_ent_put(ent); /* timeout work was canceled */ + + if (!forced || /* Real FW completion */ + pci_channel_offline(dev->pdev) || /* FW is inaccessible */ + dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) + cmd_ent_put(ent); + if (ent->page_queue) sem = &cmd->pages_sem; else @@ -1545,10 +1567,6 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force ent->ret, deliv_status_to_str(ent->status), ent->status); } - /* only real completion will free the entry slot */ - if (!forced) - free_ent(cmd, ent->idx); - if (ent->callback) { ds = ent->ts2 - ent->ts1; if (ent->op < MLX5_CMD_OP_MAX) { @@ -1576,10 +1594,13 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force free_msg(dev, ent->in); err = err ? err : ent->status; - if (!forced) - free_cmd(ent); + /* final consumer is done, release ent */ + cmd_ent_put(ent); callback(err, context); } else { + /* release wait_func() so mlx5_cmd_invoke() + * can make the final ent_put() + */ complete(&ent->done); } up(sem); @@ -1589,8 +1610,11 @@ static void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, u64 vec, bool force void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev) { + struct mlx5_cmd *cmd = &dev->cmd; + unsigned long bitmask; unsigned long flags; u64 vector; + int i; /* wait for pending handlers to complete */ mlx5_eq_synchronize_cmd_irq(dev); @@ -1599,11 +1623,20 @@ void mlx5_cmd_trigger_completions(struct mlx5_core_dev *dev) if (!vector) goto no_trig; + bitmask = vector; + /* we must increment the allocated entries refcount before triggering the completions + * to guarantee pending commands will not get freed in the meanwhile. + * For that reason, it also has to be done inside the alloc_lock. + */ + for_each_set_bit(i, &bitmask, (1 << cmd->log_sz)) + cmd_ent_get(cmd->ent_arr[i]); vector |= MLX5_TRIGGERED_CMD_COMP; spin_unlock_irqrestore(&dev->cmd.alloc_lock, flags); mlx5_core_dbg(dev, "vector 0x%llx\n", vector); mlx5_cmd_comp_handler(dev, vector, true); + for_each_set_bit(i, &bitmask, (1 << cmd->log_sz)) + cmd_ent_put(cmd->ent_arr[i]); return; no_trig: -- cgit From 1d5558b1f0de81f54ddee05f3793acc5260d107f Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Tue, 21 Jul 2020 10:25:52 +0300 Subject: net/mlx5: poll cmd EQ in case of command timeout Once driver detects a command interface command timeout, it warns the user and returns timeout error to the caller. In such case, the entry of the command is not evacuated (because only real event interrupt is allowed to clear command interface entry). If the HW event interrupt of this entry will never arrive, this entry will be left unused forever. Command interface entries are limited and eventually we can end up without the ability to post a new command. In addition, if driver will not consume the EQE of the lost interrupt and rearm the EQ, no new interrupts will arrive for other commands. Add a resiliency mechanism for manually polling the command EQ in case of a command timeout. In case resiliency mechanism will find non-handled EQE, it will consume it, and the command interface will be fully functional again. Once the resiliency flow finished, wait another 5 seconds for the command interface to complete for this command entry. Define mlx5_cmd_eq_recover() to manage the cmd EQ polling resiliency flow. Add an async EQ spinlock to avoid races between resiliency flows and real interrupts that might run simultaneously. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 53 ++++++++++++++++++++---- drivers/net/ethernet/mellanox/mlx5/core/eq.c | 40 +++++++++++++++++- drivers/net/ethernet/mellanox/mlx5/core/lib/eq.h | 2 + 3 files changed, 86 insertions(+), 9 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index c0055f5479ce..37dae95e61d5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -853,11 +853,21 @@ static void cb_timeout_handler(struct work_struct *work) struct mlx5_core_dev *dev = container_of(ent->cmd, struct mlx5_core_dev, cmd); + mlx5_cmd_eq_recover(dev); + + /* Maybe got handled by eq recover ? */ + if (!test_bit(MLX5_CMD_ENT_STATE_PENDING_COMP, &ent->state)) { + mlx5_core_warn(dev, "cmd[%d]: %s(0x%x) Async, recovered after timeout\n", ent->idx, + mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); + goto out; /* phew, already handled */ + } + ent->ret = -ETIMEDOUT; - mlx5_core_warn(dev, "%s(0x%x) timeout. Will cause a leak of a command resource\n", - mlx5_command_str(msg_to_opcode(ent->in)), - msg_to_opcode(ent->in)); + mlx5_core_warn(dev, "cmd[%d]: %s(0x%x) Async, timeout. Will cause a leak of a command resource\n", + ent->idx, mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); + +out: cmd_ent_put(ent); /* for the cmd_ent_get() took on schedule delayed work */ } @@ -997,6 +1007,35 @@ static const char *deliv_status_to_str(u8 status) } } +enum { + MLX5_CMD_TIMEOUT_RECOVER_MSEC = 5 * 1000, +}; + +static void wait_func_handle_exec_timeout(struct mlx5_core_dev *dev, + struct mlx5_cmd_work_ent *ent) +{ + unsigned long timeout = msecs_to_jiffies(MLX5_CMD_TIMEOUT_RECOVER_MSEC); + + mlx5_cmd_eq_recover(dev); + + /* Re-wait on the ent->done after executing the recovery flow. If the + * recovery flow (or any other recovery flow running simultaneously) + * has recovered an EQE, it should cause the entry to be completed by + * the command interface. + */ + if (wait_for_completion_timeout(&ent->done, timeout)) { + mlx5_core_warn(dev, "cmd[%d]: %s(0x%x) recovered after timeout\n", ent->idx, + mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); + return; + } + + mlx5_core_warn(dev, "cmd[%d]: %s(0x%x) No done completion\n", ent->idx, + mlx5_command_str(msg_to_opcode(ent->in)), msg_to_opcode(ent->in)); + + ent->ret = -ETIMEDOUT; + mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); +} + static int wait_func(struct mlx5_core_dev *dev, struct mlx5_cmd_work_ent *ent) { unsigned long timeout = msecs_to_jiffies(MLX5_CMD_TIMEOUT_MSEC); @@ -1008,12 +1047,10 @@ static int wait_func(struct mlx5_core_dev *dev, struct mlx5_cmd_work_ent *ent) ent->ret = -ECANCELED; goto out_err; } - if (cmd->mode == CMD_MODE_POLLING || ent->polling) { + if (cmd->mode == CMD_MODE_POLLING || ent->polling) wait_for_completion(&ent->done); - } else if (!wait_for_completion_timeout(&ent->done, timeout)) { - ent->ret = -ETIMEDOUT; - mlx5_cmd_comp_handler(dev, 1UL << ent->idx, true); - } + else if (!wait_for_completion_timeout(&ent->done, timeout)) + wait_func_handle_exec_timeout(dev, ent); out_err: err = ent->ret; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 1318d774b18f..22a19d391e17 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -189,6 +189,29 @@ u32 mlx5_eq_poll_irq_disabled(struct mlx5_eq_comp *eq) return count_eqe; } +static void mlx5_eq_async_int_lock(struct mlx5_eq_async *eq, unsigned long *flags) + __acquires(&eq->lock) +{ + if (in_irq()) + spin_lock(&eq->lock); + else + spin_lock_irqsave(&eq->lock, *flags); +} + +static void mlx5_eq_async_int_unlock(struct mlx5_eq_async *eq, unsigned long *flags) + __releases(&eq->lock) +{ + if (in_irq()) + spin_unlock(&eq->lock); + else + spin_unlock_irqrestore(&eq->lock, *flags); +} + +enum async_eq_nb_action { + ASYNC_EQ_IRQ_HANDLER = 0, + ASYNC_EQ_RECOVER = 1, +}; + static int mlx5_eq_async_int(struct notifier_block *nb, unsigned long action, void *data) { @@ -198,11 +221,14 @@ static int mlx5_eq_async_int(struct notifier_block *nb, struct mlx5_eq_table *eqt; struct mlx5_core_dev *dev; struct mlx5_eqe *eqe; + unsigned long flags; int num_eqes = 0; dev = eq->dev; eqt = dev->priv.eq_table; + mlx5_eq_async_int_lock(eq_async, &flags); + eqe = next_eqe_sw(eq); if (!eqe) goto out; @@ -223,8 +249,19 @@ static int mlx5_eq_async_int(struct notifier_block *nb, out: eq_update_ci(eq, 1); + mlx5_eq_async_int_unlock(eq_async, &flags); - return 0; + return unlikely(action == ASYNC_EQ_RECOVER) ? num_eqes : 0; +} + +void mlx5_cmd_eq_recover(struct mlx5_core_dev *dev) +{ + struct mlx5_eq_async *eq = &dev->priv.eq_table->cmd_eq; + int eqes; + + eqes = mlx5_eq_async_int(&eq->irq_nb, ASYNC_EQ_RECOVER, NULL); + if (eqes) + mlx5_core_warn(dev, "Recovered %d EQEs on cmd_eq\n", eqes); } static void init_eq_buf(struct mlx5_eq *eq) @@ -569,6 +606,7 @@ setup_async_eq(struct mlx5_core_dev *dev, struct mlx5_eq_async *eq, int err; eq->irq_nb.notifier_call = mlx5_eq_async_int; + spin_lock_init(&eq->lock); err = create_async_eq(dev, &eq->core, param); if (err) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/eq.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/eq.h index 4aaca7400fb2..5c681e31983b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/eq.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/eq.h @@ -37,6 +37,7 @@ struct mlx5_eq { struct mlx5_eq_async { struct mlx5_eq core; struct notifier_block irq_nb; + spinlock_t lock; /* To avoid irq EQ handle races with resiliency flows */ }; struct mlx5_eq_comp { @@ -81,6 +82,7 @@ void mlx5_cq_tasklet_cb(unsigned long data); struct cpumask *mlx5_eq_comp_cpumask(struct mlx5_core_dev *dev, int ix); u32 mlx5_eq_poll_irq_disabled(struct mlx5_eq_comp *eq); +void mlx5_cmd_eq_recover(struct mlx5_core_dev *dev); void mlx5_eq_synchronize_async_irq(struct mlx5_core_dev *dev); void mlx5_eq_synchronize_cmd_irq(struct mlx5_core_dev *dev); -- cgit From 410bd754cd73c4a2ac3856d9a03d7b08f9c906bf Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Mon, 31 Aug 2020 15:04:35 +0300 Subject: net/mlx5: Add retry mechanism to the command entry index allocation It is possible that new command entry index allocation will temporarily fail. The new command holds the semaphore, so it means that a free entry should be ready soon. Add one second retry mechanism before returning an error. Patch "net/mlx5: Avoid possible free of command entry while timeout comp handler" increase the possibility to bump into this temporarily failure as it delays the entry index release for non-callback commands. Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index 37dae95e61d5..2b597ac365f8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -883,6 +883,25 @@ static bool opcode_allowed(struct mlx5_cmd *cmd, u16 opcode) return cmd->allowed_opcode == opcode; } +static int cmd_alloc_index_retry(struct mlx5_cmd *cmd) +{ + unsigned long alloc_end = jiffies + msecs_to_jiffies(1000); + int idx; + +retry: + idx = cmd_alloc_index(cmd); + if (idx < 0 && time_before(jiffies, alloc_end)) { + /* Index allocation can fail on heavy load of commands. This is a temporary + * situation as the current command already holds the semaphore, meaning that + * another command completion is being handled and it is expected to release + * the entry index soon. + */ + cpu_relax(); + goto retry; + } + return idx; +} + static void cmd_work_handler(struct work_struct *work) { struct mlx5_cmd_work_ent *ent = container_of(work, struct mlx5_cmd_work_ent, work); @@ -900,7 +919,7 @@ static void cmd_work_handler(struct work_struct *work) sem = ent->page_queue ? &cmd->pages_sem : &cmd->sem; down(sem); if (!ent->page_queue) { - alloc_ret = cmd_alloc_index(cmd); + alloc_ret = cmd_alloc_index_retry(cmd); if (alloc_ret < 0) { mlx5_core_err_rl(dev, "failed to allocate command entry\n"); if (ent->callback) { -- cgit From b898ce7bccf13087719c021d829dab607c175246 Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Fri, 11 Sep 2020 11:48:55 -0700 Subject: net/mlx5: cmdif, Avoid skipping reclaim pages if FW is not accessible In case of pci is offline reclaim_pages_cmd() will still try to call the FW to release FW pages, cmd_exec() in this case will return a silent success without actually calling the FW. This is wrong and will cause page leaks, what we should do is to detect pci offline or command interface un-available before tying to access the FW and manually release the FW pages in the driver. In this patch we share the code to check for FW command interface availability and we call it in sensitive places e.g. reclaim_pages_cmd(). Alternative fix: 1. Remove MLX5_CMD_OP_MANAGE_PAGES form mlx5_internal_err_ret_value, command success simulation list. 2. Always Release FW pages even if cmd_exec fails in reclaim_pages_cmd(). Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 17 +++++++++-------- drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c | 2 +- 2 files changed, 10 insertions(+), 9 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index 2b597ac365f8..2d1f4b3be9bf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -902,6 +902,13 @@ retry: return idx; } +bool mlx5_cmd_is_down(struct mlx5_core_dev *dev) +{ + return pci_channel_offline(dev->pdev) || + dev->cmd.state != MLX5_CMDIF_STATE_UP || + dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR; +} + static void cmd_work_handler(struct work_struct *work) { struct mlx5_cmd_work_ent *ent = container_of(work, struct mlx5_cmd_work_ent, work); @@ -967,10 +974,7 @@ static void cmd_work_handler(struct work_struct *work) set_bit(MLX5_CMD_ENT_STATE_PENDING_COMP, &ent->state); /* Skip sending command to fw if internal error */ - if (pci_channel_offline(dev->pdev) || - dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || - cmd->state != MLX5_CMDIF_STATE_UP || - !opcode_allowed(&dev->cmd, ent->op)) { + if (mlx5_cmd_is_down(dev) || !opcode_allowed(&dev->cmd, ent->op)) { u8 status = 0; u32 drv_synd; @@ -1800,10 +1804,7 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out, u8 token; opcode = MLX5_GET(mbox_in, in, opcode); - if (pci_channel_offline(dev->pdev) || - dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR || - dev->cmd.state != MLX5_CMDIF_STATE_UP || - !opcode_allowed(&dev->cmd, opcode)) { + if (mlx5_cmd_is_down(dev) || !opcode_allowed(&dev->cmd, opcode)) { err = mlx5_internal_err_ret_value(dev, opcode, &drv_synd, &status); MLX5_SET(mbox_out, out, status, status); MLX5_SET(mbox_out, out, syndrome, drv_synd); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c index f9b798af6335..c0e18f2ade99 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c @@ -432,7 +432,7 @@ static int reclaim_pages_cmd(struct mlx5_core_dev *dev, u32 npages; u32 i = 0; - if (dev->state != MLX5_DEVICE_STATE_INTERNAL_ERROR) + if (!mlx5_cmd_is_down(dev)) return mlx5_cmd_exec(dev, in, in_size, out, out_size); /* No hard feelings, we want our pages back! */ -- cgit From 732ebfab7fe96b7ac9a3df3208f14752a4bb6db3 Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Mon, 31 Aug 2020 21:37:31 +0300 Subject: net/mlx5: Fix request_irqs error flow Fix error flow handling in request_irqs which try to free irq that we failed to request. It fixes the below trace. WARNING: CPU: 1 PID: 7587 at kernel/irq/manage.c:1684 free_irq+0x4d/0x60 CPU: 1 PID: 7587 Comm: bash Tainted: G W OE 4.15.15-1.el7MELLANOXsmp-x86_64 #1 Hardware name: Advantech SKY-6200/SKY-6200, BIOS F2.00 08/06/2020 RIP: 0010:free_irq+0x4d/0x60 RSP: 0018:ffffc9000ef47af0 EFLAGS: 00010282 RAX: ffff88001476ae00 RBX: 0000000000000655 RCX: 0000000000000000 RDX: ffff88001476ae00 RSI: ffffc9000ef47ab8 RDI: ffff8800398bb478 RBP: ffff88001476a838 R08: ffff88001476ae00 R09: 000000000000156d R10: 0000000000000000 R11: 0000000000000004 R12: ffff88001476a838 R13: 0000000000000006 R14: ffff88001476a888 R15: 00000000ffffffe4 FS: 00007efeadd32740(0000) GS:ffff88047fc40000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fc9cc010008 CR3: 00000001a2380004 CR4: 00000000007606e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 PKRU: 55555554 Call Trace: mlx5_irq_table_create+0x38d/0x400 [mlx5_core] ? atomic_notifier_chain_register+0x50/0x60 mlx5_load_one+0x7ee/0x1130 [mlx5_core] init_one+0x4c9/0x650 [mlx5_core] pci_device_probe+0xb8/0x120 driver_probe_device+0x2a1/0x470 ? driver_allows_async_probing+0x30/0x30 bus_for_each_drv+0x54/0x80 __device_attach+0xa3/0x100 pci_bus_add_device+0x4a/0x90 pci_iov_add_virtfn+0x2dc/0x2f0 pci_enable_sriov+0x32e/0x420 mlx5_core_sriov_configure+0x61/0x1b0 [mlx5_core] ? kstrtoll+0x22/0x70 num_vf_store+0x4b/0x70 [mlx5_core] kernfs_fop_write+0x102/0x180 __vfs_write+0x26/0x140 ? rcu_all_qs+0x5/0x80 ? _cond_resched+0x15/0x30 ? __sb_start_write+0x41/0x80 vfs_write+0xad/0x1a0 SyS_write+0x42/0x90 do_syscall_64+0x60/0x110 entry_SYSCALL_64_after_hwframe+0x3d/0xa2 Fixes: 24163189da48 ("net/mlx5: Separate IRQ request/free from EQ life cycle") Signed-off-by: Maor Gottlieb Reviewed-by: Eran Ben Elisha Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c index 373981a659c7..6fd974920394 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c @@ -115,7 +115,7 @@ static int request_irqs(struct mlx5_core_dev *dev, int nvec) return 0; err_request_irq: - for (; i >= 0; i--) { + while (i--) { struct mlx5_irq *irq = mlx5_irq_get(dev, i); int irqn = pci_irq_vector(dev->pdev, i); -- cgit From 08a762cecc60e19cd64e379df6580f1997408e04 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Mon, 3 Aug 2020 16:22:42 +0300 Subject: net/mlx5e: Fix error path for RQ alloc Increase granularity of the error path to avoid unneeded free/release. Fix the cleanup to be symmetric to the order of creation. Fixes: 0ddf543226ac ("xdp/mlx5: setup xdp_rxq_info") Fixes: 422d4c401edd ("net/mlx5e: RX, Split WQ objects for different RQ types") Signed-off-by: Aya Levin Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 32 ++++++++++++----------- 1 file changed, 17 insertions(+), 15 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index b3cda7b6e5e1..1fbd7a0f3ca6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -396,7 +396,7 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, rq_xdp_ix += params->num_channels * MLX5E_RQ_GROUP_XSK; err = xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq_xdp_ix); if (err < 0) - goto err_rq_wq_destroy; + goto err_rq_xdp_prog; rq->buff.map_dir = params->xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE; rq->buff.headroom = mlx5e_get_rq_headroom(mdev, params, xsk); @@ -407,7 +407,7 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, err = mlx5_wq_ll_create(mdev, &rqp->wq, rqc_wq, &rq->mpwqe.wq, &rq->wq_ctrl); if (err) - goto err_rq_wq_destroy; + goto err_rq_xdp; rq->mpwqe.wq.db = &rq->mpwqe.wq.db[MLX5_RCV_DBR]; @@ -429,13 +429,13 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, err = mlx5e_rq_alloc_mpwqe_info(rq, c); if (err) - goto err_free; + goto err_rq_mkey; break; default: /* MLX5_WQ_TYPE_CYCLIC */ err = mlx5_wq_cyc_create(mdev, &rqp->wq, rqc_wq, &rq->wqe.wq, &rq->wq_ctrl); if (err) - goto err_rq_wq_destroy; + goto err_rq_xdp; rq->wqe.wq.db = &rq->wqe.wq.db[MLX5_RCV_DBR]; @@ -450,19 +450,19 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, GFP_KERNEL, cpu_to_node(c->cpu)); if (!rq->wqe.frags) { err = -ENOMEM; - goto err_free; + goto err_rq_wq_destroy; } err = mlx5e_init_di_list(rq, wq_sz, c->cpu); if (err) - goto err_free; + goto err_rq_frags; rq->mkey_be = c->mkey_be; } err = mlx5e_rq_set_handlers(rq, params, xsk); if (err) - goto err_free; + goto err_free_by_rq_type; if (xsk) { err = xdp_rxq_info_reg_mem_model(&rq->xdp_rxq, @@ -486,13 +486,13 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, if (IS_ERR(rq->page_pool)) { err = PTR_ERR(rq->page_pool); rq->page_pool = NULL; - goto err_free; + goto err_free_by_rq_type; } err = xdp_rxq_info_reg_mem_model(&rq->xdp_rxq, MEM_TYPE_PAGE_POOL, rq->page_pool); } if (err) - goto err_free; + goto err_free_by_rq_type; for (i = 0; i < wq_sz; i++) { if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) { @@ -542,23 +542,25 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, return 0; -err_free: +err_free_by_rq_type: switch (rq->wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: kvfree(rq->mpwqe.info); +err_rq_mkey: mlx5_core_destroy_mkey(mdev, &rq->umr_mkey); break; default: /* MLX5_WQ_TYPE_CYCLIC */ - kvfree(rq->wqe.frags); mlx5e_free_di_list(rq); +err_rq_frags: + kvfree(rq->wqe.frags); } - err_rq_wq_destroy: + mlx5_wq_destroy(&rq->wq_ctrl); +err_rq_xdp: + xdp_rxq_info_unreg(&rq->xdp_rxq); +err_rq_xdp_prog: if (params->xdp_prog) bpf_prog_put(params->xdp_prog); - xdp_rxq_info_unreg(&rq->xdp_rxq); - page_pool_destroy(rq->page_pool); - mlx5_wq_destroy(&rq->wq_ctrl); return err; } -- cgit From c3c9402373fe20e2d08c04f437ce4dcd252cffb2 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Mon, 20 Jul 2020 16:53:18 +0300 Subject: net/mlx5e: Add resiliency in Striding RQ mode for packets larger than MTU Prior to this fix, in Striding RQ mode the driver was vulnerable when receiving packets in the range (stride size - headroom, stride size]. Where stride size is calculated by mtu+headroom+tailroom aligned to the closest power of 2. Usually, this filtering is performed by the HW, except for a few cases: - Between 2 VFs over the same PF with different MTUs - On bluefield, when the host physical function sets a larger MTU than the ARM has configured on its representor and uplink representor. When the HW filtering is not present, packets that are larger than MTU might be harmful for the RQ's integrity, in the following impacts: 1) Overflow from one WQE to the next, causing a memory corruption that in most cases is unharmful: as the write happens to the headroom of next packet, which will be overwritten by build_skb(). In very rare cases, high stress/load, this is harmful. When the next WQE is not yet reposted and points to existing SKB head. 2) Each oversize packet overflows to the headroom of the next WQE. On the last WQE of the WQ, where addresses wrap-around, the address of the remainder headroom does not belong to the next WQE, but it is out of the memory region range. This results in a HW CQE error that moves the RQ into an error state. Solution: Add a page buffer at the end of each WQE to absorb the leak. Actually the maximal overflow size is headroom but since all memory units must be of the same size, we use page size to comply with UMR WQEs. The increase in memory consumption is of a single page per RQ. Initialize the mkey with all MTTs pointing to a default page. When the channels are activated, UMR WQEs will redirect the RX WQEs to the actual memory from the RQ's pool, while the overflow MTTs remain mapped to the default page. Fixes: 73281b78a37a ("net/mlx5e: Derive Striding RQ size from MTU") Signed-off-by: Aya Levin Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 8 +++- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 55 +++++++++++++++++++++-- 2 files changed, 58 insertions(+), 5 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 90d5caabd6af..356f5852955f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -91,7 +91,12 @@ struct page_pool; #define MLX5_MPWRQ_PAGES_PER_WQE BIT(MLX5_MPWRQ_WQE_PAGE_ORDER) #define MLX5_MTT_OCTW(npages) (ALIGN(npages, 8) / 2) -#define MLX5E_REQUIRED_WQE_MTTS (ALIGN(MLX5_MPWRQ_PAGES_PER_WQE, 8)) +/* Add another page to MLX5E_REQUIRED_WQE_MTTS as a buffer between + * WQEs, This page will absorb write overflow by the hardware, when + * receiving packets larger than MTU. These oversize packets are + * dropped by the driver at a later stage. + */ +#define MLX5E_REQUIRED_WQE_MTTS (ALIGN(MLX5_MPWRQ_PAGES_PER_WQE + 1, 8)) #define MLX5E_LOG_ALIGNED_MPWQE_PPW (ilog2(MLX5E_REQUIRED_WQE_MTTS)) #define MLX5E_REQUIRED_MTTS(wqes) (wqes * MLX5E_REQUIRED_WQE_MTTS) #define MLX5E_MAX_RQ_NUM_MTTS \ @@ -617,6 +622,7 @@ struct mlx5e_rq { u32 rqn; struct mlx5_core_dev *mdev; struct mlx5_core_mkey umr_mkey; + struct mlx5e_dma_info wqe_overflow; /* XDP read-mostly */ struct xdp_rxq_info xdp_rxq; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 1fbd7a0f3ca6..b1a16fb9667e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -246,12 +246,17 @@ static int mlx5e_rq_alloc_mpwqe_info(struct mlx5e_rq *rq, static int mlx5e_create_umr_mkey(struct mlx5_core_dev *mdev, u64 npages, u8 page_shift, - struct mlx5_core_mkey *umr_mkey) + struct mlx5_core_mkey *umr_mkey, + dma_addr_t filler_addr) { - int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); + struct mlx5_mtt *mtt; + int inlen; void *mkc; u32 *in; int err; + int i; + + inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + sizeof(*mtt) * npages; in = kvzalloc(inlen, GFP_KERNEL); if (!in) @@ -271,6 +276,18 @@ static int mlx5e_create_umr_mkey(struct mlx5_core_dev *mdev, MLX5_SET(mkc, mkc, translations_octword_size, MLX5_MTT_OCTW(npages)); MLX5_SET(mkc, mkc, log_page_size, page_shift); + MLX5_SET(create_mkey_in, in, translations_octword_actual_size, + MLX5_MTT_OCTW(npages)); + + /* Initialize the mkey with all MTTs pointing to a default + * page (filler_addr). When the channels are activated, UMR + * WQEs will redirect the RX WQEs to the actual memory from + * the RQ's pool, while the gaps (wqe_overflow) remain mapped + * to the default page. + */ + mtt = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + for (i = 0 ; i < npages ; i++) + mtt[i].ptag = cpu_to_be64(filler_addr); err = mlx5_core_create_mkey(mdev, umr_mkey, in, inlen); @@ -282,7 +299,8 @@ static int mlx5e_create_rq_umr_mkey(struct mlx5_core_dev *mdev, struct mlx5e_rq { u64 num_mtts = MLX5E_REQUIRED_MTTS(mlx5_wq_ll_get_size(&rq->mpwqe.wq)); - return mlx5e_create_umr_mkey(mdev, num_mtts, PAGE_SHIFT, &rq->umr_mkey); + return mlx5e_create_umr_mkey(mdev, num_mtts, PAGE_SHIFT, &rq->umr_mkey, + rq->wqe_overflow.addr); } static inline u64 mlx5e_get_mpwqe_offset(struct mlx5e_rq *rq, u16 wqe_ix) @@ -350,6 +368,28 @@ static void mlx5e_rq_err_cqe_work(struct work_struct *recover_work) mlx5e_reporter_rq_cqe_err(rq); } +static int mlx5e_alloc_mpwqe_rq_drop_page(struct mlx5e_rq *rq) +{ + rq->wqe_overflow.page = alloc_page(GFP_KERNEL); + if (!rq->wqe_overflow.page) + return -ENOMEM; + + rq->wqe_overflow.addr = dma_map_page(rq->pdev, rq->wqe_overflow.page, 0, + PAGE_SIZE, rq->buff.map_dir); + if (dma_mapping_error(rq->pdev, rq->wqe_overflow.addr)) { + __free_page(rq->wqe_overflow.page); + return -ENOMEM; + } + return 0; +} + +static void mlx5e_free_mpwqe_rq_drop_page(struct mlx5e_rq *rq) +{ + dma_unmap_page(rq->pdev, rq->wqe_overflow.addr, PAGE_SIZE, + rq->buff.map_dir); + __free_page(rq->wqe_overflow.page); +} + static int mlx5e_alloc_rq(struct mlx5e_channel *c, struct mlx5e_params *params, struct mlx5e_xsk_param *xsk, @@ -409,6 +449,10 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, if (err) goto err_rq_xdp; + err = mlx5e_alloc_mpwqe_rq_drop_page(rq); + if (err) + goto err_rq_wq_destroy; + rq->mpwqe.wq.db = &rq->mpwqe.wq.db[MLX5_RCV_DBR]; wq_sz = mlx5_wq_ll_get_size(&rq->mpwqe.wq); @@ -424,7 +468,7 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, err = mlx5e_create_rq_umr_mkey(mdev, rq); if (err) - goto err_rq_wq_destroy; + goto err_rq_drop_page; rq->mkey_be = cpu_to_be32(rq->umr_mkey.key); err = mlx5e_rq_alloc_mpwqe_info(rq, c); @@ -548,6 +592,8 @@ err_free_by_rq_type: kvfree(rq->mpwqe.info); err_rq_mkey: mlx5_core_destroy_mkey(mdev, &rq->umr_mkey); +err_rq_drop_page: + mlx5e_free_mpwqe_rq_drop_page(rq); break; default: /* MLX5_WQ_TYPE_CYCLIC */ mlx5e_free_di_list(rq); @@ -582,6 +628,7 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq) case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: kvfree(rq->mpwqe.info); mlx5_core_destroy_mkey(rq->mdev, &rq->umr_mkey); + mlx5e_free_mpwqe_rq_drop_page(rq); break; default: /* MLX5_WQ_TYPE_CYCLIC */ kvfree(rq->wqe.frags); -- cgit From 2b0219898b8684075ff66027551fb02ea12f652b Mon Sep 17 00:00:00 2001 From: Maor Dickman Date: Tue, 1 Sep 2020 12:33:18 +0300 Subject: net/mlx5e: CT, Fix coverity issue The cited commit introduced the following coverity issue at function mlx5_tc_ct_rule_to_tuple_nat: - Memory - corruptions (OVERRUN) Overrunning array "tuple->ip.src_v6.in6_u.u6_addr32" of 4 4-byte elements at element index 7 (byte offset 31) using index "ip6_offset" (which evaluates to 7). In case of IPv6 destination address rewrite, ip6_offset values are between 4 to 7, which will cause memory overrun of array "tuple->ip.src_v6.in6_u.u6_addr32" to array "tuple->ip.dst_v6.in6_u.u6_addr32". Fixed by writing the value directly to array "tuple->ip.dst_v6.in6_u.u6_addr32" in case ip6_offset values are between 4 to 7. Fixes: bc562be9674b ("net/mlx5e: CT: Save ct entries tuples in hashtables") Signed-off-by: Maor Dickman Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index bc5f72ec3623..a8be40cbe325 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -246,8 +246,10 @@ mlx5_tc_ct_rule_to_tuple_nat(struct mlx5_ct_tuple *tuple, case FLOW_ACT_MANGLE_HDR_TYPE_IP6: ip6_offset = (offset - offsetof(struct ipv6hdr, saddr)); ip6_offset /= 4; - if (ip6_offset < 8) + if (ip6_offset < 4) tuple->ip.src_v6.s6_addr32[ip6_offset] = cpu_to_be32(val); + else if (ip6_offset < 8) + tuple->ip.dst_v6.s6_addr32[ip6_offset - 4] = cpu_to_be32(val); else return -EOPNOTSUPP; break; -- cgit From 3d093bc2369003b4ce6c3522d9b383e47c40045d Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Sun, 9 Aug 2020 12:34:21 +0300 Subject: net/mlx5e: Fix driver's declaration to support GRE offload Declare GRE offload support with respect to the inner protocol. Add a list of supported inner protocols on which the driver can offload checksum and GSO. For other protocols, inform the stack to do the needed operations. There is no noticeable impact on GRE performance. Fixes: 2729984149e6 ("net/mlx5e: Support TSO and TX checksum offloads for GRE tunnels") Signed-off-by: Aya Levin Reviewed-by: Moshe Shemesh Reviewed-by: Tariq Toukan Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index b1a16fb9667e..42ec28e29834 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -4226,6 +4226,21 @@ int mlx5e_get_vf_stats(struct net_device *dev, } #endif +static bool mlx5e_gre_tunnel_inner_proto_offload_supported(struct mlx5_core_dev *mdev, + struct sk_buff *skb) +{ + switch (skb->inner_protocol) { + case htons(ETH_P_IP): + case htons(ETH_P_IPV6): + case htons(ETH_P_TEB): + return true; + case htons(ETH_P_MPLS_UC): + case htons(ETH_P_MPLS_MC): + return MLX5_CAP_ETH(mdev, tunnel_stateless_mpls_over_gre); + } + return false; +} + static netdev_features_t mlx5e_tunnel_features_check(struct mlx5e_priv *priv, struct sk_buff *skb, netdev_features_t features) @@ -4248,7 +4263,9 @@ static netdev_features_t mlx5e_tunnel_features_check(struct mlx5e_priv *priv, switch (proto) { case IPPROTO_GRE: - return features; + if (mlx5e_gre_tunnel_inner_proto_offload_supported(priv->mdev, skb)) + return features; + break; case IPPROTO_IPIP: case IPPROTO_IPV6: if (mlx5e_tunnel_proto_supported(priv->mdev, IPPROTO_IPIP)) -- cgit From 2608a2f831c47dfdf18885a7289be5af97182b05 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Wed, 12 Aug 2020 10:44:36 +0300 Subject: net/mlx5e: Fix return status when setting unsupported FEC mode Verify the configured FEC mode is supported by at least a single link mode before applying the command. Otherwise fail the command and return "Operation not supported". Prior to this patch, the command was successful, yet it falsely set all link modes to FEC auto mode - like configuring FEC mode to auto. Auto mode is the default configuration if a link mode doesn't support the configured FEC mode. Fixes: b5ede32d3329 ("net/mlx5e: Add support for FEC modes based on 50G per lane links") Signed-off-by: Aya Levin Reviewed-by: Eran Ben Elisha Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en/port.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/port.c b/drivers/net/ethernet/mellanox/mlx5/core/en/port.c index 96608dbb9314..308fd279669e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/port.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/port.c @@ -569,6 +569,9 @@ int mlx5e_set_fec_mode(struct mlx5_core_dev *dev, u16 fec_policy) if (fec_policy >= (1 << MLX5E_FEC_LLRS_272_257_1) && !fec_50g_per_lane) return -EOPNOTSUPP; + if (fec_policy && !mlx5e_fec_in_caps(dev, fec_policy)) + return -EOPNOTSUPP; + MLX5_SET(pplm_reg, in, local_port, 1); err = mlx5_core_access_reg(dev, in, sz, out, sz, MLX5_REG_PPLM, 0, 0); if (err) -- cgit From 8c7353b6f716436ad0bfda2b5c5524ab2dde5894 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Sun, 13 Sep 2020 17:57:23 +0300 Subject: net/mlx5e: Fix VLAN cleanup flow Prior to this patch unloading an interface in promiscuous mode with RX VLAN filtering feature turned off - resulted in a warning. This is due to a wrong condition in the VLAN rules cleanup flow, which left the any-vid rules in the VLAN steering table. These rules prevented destroying the flow group and the flow table. The any-vid rules are removed in 2 flows, but none of them remove it in case both promiscuous is set and VLAN filtering is off. Fix the issue by changing the condition of the VLAN table cleanup flow to clean also in case of promiscuous mode. mlx5_core 0000:00:08.0: mlx5_destroy_flow_group:2123:(pid 28729): Flow group 20 wasn't destroyed, refcount > 1 mlx5_core 0000:00:08.0: mlx5_destroy_flow_group:2123:(pid 28729): Flow group 19 wasn't destroyed, refcount > 1 mlx5_core 0000:00:08.0: mlx5_destroy_flow_table:2112:(pid 28729): Flow table 262149 wasn't destroyed, refcount > 1 ... ... ------------[ cut here ]------------ FW pages counter is 11560 after reclaiming all pages WARNING: CPU: 1 PID: 28729 at drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c:660 mlx5_reclaim_startup_pages+0x178/0x230 [mlx5_core] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014 Call Trace: mlx5_function_teardown+0x2f/0x90 [mlx5_core] mlx5_unload_one+0x71/0x110 [mlx5_core] remove_one+0x44/0x80 [mlx5_core] pci_device_remove+0x3e/0xc0 device_release_driver_internal+0xfb/0x1c0 device_release_driver+0x12/0x20 pci_stop_bus_device+0x68/0x90 pci_stop_and_remove_bus_device+0x12/0x20 hv_eject_device_work+0x6f/0x170 [pci_hyperv] ? __schedule+0x349/0x790 process_one_work+0x206/0x400 worker_thread+0x34/0x3f0 ? process_one_work+0x400/0x400 kthread+0x126/0x140 ? kthread_park+0x90/0x90 ret_from_fork+0x22/0x30 ---[ end trace 6283bde8d26170dc ]--- Fixes: 9df30601c843 ("net/mlx5e: Restore vlan filter after seamless reset") Signed-off-by: Aya Levin Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index 64d002d92250..55a4c3adaa05 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -415,8 +415,12 @@ static void mlx5e_del_vlan_rules(struct mlx5e_priv *priv) for_each_set_bit(i, priv->fs.vlan.active_svlans, VLAN_N_VID) mlx5e_del_vlan_rule(priv, MLX5E_VLAN_RULE_TYPE_MATCH_STAG_VID, i); - if (priv->fs.vlan.cvlan_filter_disabled && - !(priv->netdev->flags & IFF_PROMISC)) + WARN_ON_ONCE(!(test_bit(MLX5E_STATE_DESTROYING, &priv->state))); + + /* must be called after DESTROY bit is set and + * set_rx_mode is called and flushed + */ + if (priv->fs.vlan.cvlan_filter_disabled) mlx5e_del_any_vid_rules(priv); } -- cgit From d4a16052bccdd695982f89d815ca075825115821 Mon Sep 17 00:00:00 2001 From: Aya Levin Date: Sun, 13 Sep 2020 18:05:40 +0300 Subject: net/mlx5e: Fix VLAN create flow When interface is attached while in promiscuous mode and with VLAN filtering turned off, both configurations are not respected and VLAN filtering is performed. There are 2 flows which add the any-vid rules during interface attach: VLAN creation table and set rx mode. Each is relaying on the other to add any-vid rules, eventually non of them does. Fix this by adding any-vid rules on VLAN creation regardless of promiscuous mode. Fixes: 9df30601c843 ("net/mlx5e: Restore vlan filter after seamless reset") Signed-off-by: Aya Levin Reviewed-by: Moshe Shemesh Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index 55a4c3adaa05..1f48f99c0997 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -217,6 +217,9 @@ static int __mlx5e_add_vlan_rule(struct mlx5e_priv *priv, break; } + if (WARN_ONCE(*rule_p, "VLAN rule already exists type %d", rule_type)) + return 0; + *rule_p = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1); if (IS_ERR(*rule_p)) { @@ -397,8 +400,7 @@ static void mlx5e_add_vlan_rules(struct mlx5e_priv *priv) for_each_set_bit(i, priv->fs.vlan.active_svlans, VLAN_N_VID) mlx5e_add_vlan_rule(priv, MLX5E_VLAN_RULE_TYPE_MATCH_STAG_VID, i); - if (priv->fs.vlan.cvlan_filter_disabled && - !(priv->netdev->flags & IFF_PROMISC)) + if (priv->fs.vlan.cvlan_filter_disabled) mlx5e_add_any_vid_rules(priv); } -- cgit From 1253935ad801485270194d5651acab04abc97b36 Mon Sep 17 00:00:00 2001 From: Vlad Buslov Date: Sun, 20 Sep 2020 19:59:08 +0300 Subject: net/mlx5e: Fix race condition on nhe->n pointer in neigh update Current neigh update event handler implementation takes reference to neighbour structure, assigns it to nhe->n, tries to schedule workqueue task and releases the reference if task was already enqueued. This results potentially overwriting existing nhe->n pointer with another neighbour instance, which causes double release of the instance (once in neigh update handler that failed to enqueue to workqueue and another one in neigh update workqueue task that processes updated nhe->n pointer instead of original one): [ 3376.512806] ------------[ cut here ]------------ [ 3376.513534] refcount_t: underflow; use-after-free. [ 3376.521213] Modules linked in: act_skbedit act_mirred act_tunnel_key vxlan ip6_udp_tunnel udp_tunnel nfnetlink act_gact cls_flower sch_ingress openvswitch nsh nf_conncount nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 mlx5_ib mlx5_core mlxfw pci_hyperv_intf ptp pps_core nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace fscache ib_isert iscsi_target_mod ib_srpt target_core_mod ib_srp rpcrdma rdma_ucm ib_umad ib_ipoib ib_iser rdma_cm ib_cm iw_cm rfkill ib_uverbs ib_core sunrpc kvm_intel kvm iTCO_wdt iTCO_vendor_support virtio_net irqbypass net_failover crc32_pclmul lpc_ich i2c_i801 failover pcspkr i2c_smbus mfd_core ghash_clmulni_intel sch_fq_codel drm i2c _core ip_tables crc32c_intel serio_raw [last unloaded: mlxfw] [ 3376.529468] CPU: 8 PID: 22756 Comm: kworker/u20:5 Not tainted 5.9.0-rc5+ #6 [ 3376.530399] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014 [ 3376.531975] Workqueue: mlx5e mlx5e_rep_neigh_update [mlx5_core] [ 3376.532820] RIP: 0010:refcount_warn_saturate+0xd8/0xe0 [ 3376.533589] Code: ff 48 c7 c7 e0 b8 27 82 c6 05 0b b6 09 01 01 e8 94 93 c1 ff 0f 0b c3 48 c7 c7 88 b8 27 82 c6 05 f7 b5 09 01 01 e8 7e 93 c1 ff <0f> 0b c3 0f 1f 44 00 00 8b 07 3d 00 00 00 c0 74 12 83 f8 01 74 13 [ 3376.536017] RSP: 0018:ffffc90002a97e30 EFLAGS: 00010286 [ 3376.536793] RAX: 0000000000000000 RBX: ffff8882de30d648 RCX: 0000000000000000 [ 3376.537718] RDX: ffff8882f5c28f20 RSI: ffff8882f5c18e40 RDI: ffff8882f5c18e40 [ 3376.538654] RBP: ffff8882cdf56c00 R08: 000000000000c580 R09: 0000000000001a4d [ 3376.539582] R10: 0000000000000731 R11: ffffc90002a97ccd R12: 0000000000000000 [ 3376.540519] R13: ffff8882de30d600 R14: ffff8882de30d640 R15: ffff88821e000900 [ 3376.541444] FS: 0000000000000000(0000) GS:ffff8882f5c00000(0000) knlGS:0000000000000000 [ 3376.542732] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 3376.543545] CR2: 0000556e5504b248 CR3: 00000002c6f10005 CR4: 0000000000770ee0 [ 3376.544483] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 3376.545419] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 3376.546344] PKRU: 55555554 [ 3376.546911] Call Trace: [ 3376.547479] mlx5e_rep_neigh_update.cold+0x33/0xe2 [mlx5_core] [ 3376.548299] process_one_work+0x1d8/0x390 [ 3376.548977] worker_thread+0x4d/0x3e0 [ 3376.549631] ? rescuer_thread+0x3e0/0x3e0 [ 3376.550295] kthread+0x118/0x130 [ 3376.550914] ? kthread_create_worker_on_cpu+0x70/0x70 [ 3376.551675] ret_from_fork+0x1f/0x30 [ 3376.552312] ---[ end trace d84e8f46d2a77eec ]--- Fix the bug by moving work_struct to dedicated dynamically-allocated structure. This enabled every event handler to work on its own private neighbour pointer and removes the need for handling the case when task is already enqueued. Fixes: 232c001398ae ("net/mlx5e: Add support to neighbour update flow") Signed-off-by: Vlad Buslov Reviewed-by: Roi Dayan Signed-off-by: Saeed Mahameed --- .../net/ethernet/mellanox/mlx5/core/en/rep/neigh.c | 81 +++++++++++++--------- drivers/net/ethernet/mellanox/mlx5/core/en_rep.h | 6 -- 2 files changed, 50 insertions(+), 37 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/neigh.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/neigh.c index 906292035088..58e27038c947 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/neigh.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/neigh.c @@ -110,11 +110,25 @@ static void mlx5e_rep_neigh_stats_work(struct work_struct *work) rtnl_unlock(); } +struct neigh_update_work { + struct work_struct work; + struct neighbour *n; + struct mlx5e_neigh_hash_entry *nhe; +}; + +static void mlx5e_release_neigh_update_work(struct neigh_update_work *update_work) +{ + neigh_release(update_work->n); + mlx5e_rep_neigh_entry_release(update_work->nhe); + kfree(update_work); +} + static void mlx5e_rep_neigh_update(struct work_struct *work) { - struct mlx5e_neigh_hash_entry *nhe = - container_of(work, struct mlx5e_neigh_hash_entry, neigh_update_work); - struct neighbour *n = nhe->n; + struct neigh_update_work *update_work = container_of(work, struct neigh_update_work, + work); + struct mlx5e_neigh_hash_entry *nhe = update_work->nhe; + struct neighbour *n = update_work->n; struct mlx5e_encap_entry *e; unsigned char ha[ETH_ALEN]; struct mlx5e_priv *priv; @@ -146,30 +160,42 @@ static void mlx5e_rep_neigh_update(struct work_struct *work) mlx5e_rep_update_flows(priv, e, neigh_connected, ha); mlx5e_encap_put(priv, e); } - mlx5e_rep_neigh_entry_release(nhe); rtnl_unlock(); - neigh_release(n); + mlx5e_release_neigh_update_work(update_work); } -static void mlx5e_rep_queue_neigh_update_work(struct mlx5e_priv *priv, - struct mlx5e_neigh_hash_entry *nhe, - struct neighbour *n) +static struct neigh_update_work *mlx5e_alloc_neigh_update_work(struct mlx5e_priv *priv, + struct neighbour *n) { - /* Take a reference to ensure the neighbour and mlx5 encap - * entry won't be destructed until we drop the reference in - * delayed work. - */ - neigh_hold(n); + struct neigh_update_work *update_work; + struct mlx5e_neigh_hash_entry *nhe; + struct mlx5e_neigh m_neigh = {}; - /* This assignment is valid as long as the the neigh reference - * is taken - */ - nhe->n = n; + update_work = kzalloc(sizeof(*update_work), GFP_ATOMIC); + if (WARN_ON(!update_work)) + return NULL; - if (!queue_work(priv->wq, &nhe->neigh_update_work)) { - mlx5e_rep_neigh_entry_release(nhe); - neigh_release(n); + m_neigh.dev = n->dev; + m_neigh.family = n->ops->family; + memcpy(&m_neigh.dst_ip, n->primary_key, n->tbl->key_len); + + /* Obtain reference to nhe as last step in order not to release it in + * atomic context. + */ + rcu_read_lock(); + nhe = mlx5e_rep_neigh_entry_lookup(priv, &m_neigh); + rcu_read_unlock(); + if (!nhe) { + kfree(update_work); + return NULL; } + + INIT_WORK(&update_work->work, mlx5e_rep_neigh_update); + neigh_hold(n); + update_work->n = n; + update_work->nhe = nhe; + + return update_work; } static int mlx5e_rep_netevent_event(struct notifier_block *nb, @@ -181,7 +207,7 @@ static int mlx5e_rep_netevent_event(struct notifier_block *nb, struct net_device *netdev = rpriv->netdev; struct mlx5e_priv *priv = netdev_priv(netdev); struct mlx5e_neigh_hash_entry *nhe = NULL; - struct mlx5e_neigh m_neigh = {}; + struct neigh_update_work *update_work; struct neigh_parms *p; struct neighbour *n; bool found = false; @@ -196,17 +222,11 @@ static int mlx5e_rep_netevent_event(struct notifier_block *nb, #endif return NOTIFY_DONE; - m_neigh.dev = n->dev; - m_neigh.family = n->ops->family; - memcpy(&m_neigh.dst_ip, n->primary_key, n->tbl->key_len); - - rcu_read_lock(); - nhe = mlx5e_rep_neigh_entry_lookup(priv, &m_neigh); - rcu_read_unlock(); - if (!nhe) + update_work = mlx5e_alloc_neigh_update_work(priv, n); + if (!update_work) return NOTIFY_DONE; - mlx5e_rep_queue_neigh_update_work(priv, nhe, n); + queue_work(priv->wq, &update_work->work); break; case NETEVENT_DELAY_PROBE_TIME_UPDATE: @@ -352,7 +372,6 @@ int mlx5e_rep_neigh_entry_create(struct mlx5e_priv *priv, (*nhe)->priv = priv; memcpy(&(*nhe)->m_neigh, &e->m_neigh, sizeof(e->m_neigh)); - INIT_WORK(&(*nhe)->neigh_update_work, mlx5e_rep_neigh_update); spin_lock_init(&(*nhe)->encap_list_lock); INIT_LIST_HEAD(&(*nhe)->encap_list); refcount_set(&(*nhe)->refcnt, 1); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index 622c27ae4ac7..0d1562e20118 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -135,12 +135,6 @@ struct mlx5e_neigh_hash_entry { /* encap list sharing the same neigh */ struct list_head encap_list; - /* valid only when the neigh reference is taken during - * neigh_update_work workqueue callback. - */ - struct neighbour *n; - struct work_struct neigh_update_work; - /* neigh hash entry can be deleted only when the refcount is zero. * refcount is needed to avoid neigh hash entry removal by TC, while * it's used by the neigh notification call. -- cgit From 388e201d41fa1ed8f2dce0f0567f56f8e919ffb0 Mon Sep 17 00:00:00 2001 From: "Vineetha G. Jaya Kumaran" Date: Thu, 1 Oct 2020 23:56:09 +0800 Subject: net: stmmac: Modify configuration method of EEE timers Ethtool manual stated that the tx-timer is the "the amount of time the device should stay in idle mode prior to asserting its Tx LPI". The previous implementation for "ethtool --set-eee tx-timer" sets the LPI TW timer duration which is not correct. Hence, this patch fixes the "ethtool --set-eee tx-timer" to configure the EEE LPI timer. The LPI TW Timer will be using the defined default value instead of "ethtool --set-eee tx-timer" which follows the EEE LS timer implementation. Changelog V2 *Not removing/modifying the eee_timer. *EEE LPI timer can be configured through ethtool and also the eee_timer module param. *EEE TW Timer will be configured with default value only, not able to be configured through ethtool or module param. This follows the implementation of the EEE LS Timer. Fixes: d765955d2ae0 ("stmmac: add the Energy Efficient Ethernet support") Signed-off-by: Vineetha G. Jaya Kumaran Signed-off-by: Voon Weifeng Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac.h | 2 ++ .../net/ethernet/stmicro/stmmac/stmmac_ethtool.c | 12 ++++++++++- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 23 ++++++++++++++-------- 3 files changed, 28 insertions(+), 9 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h index 9c02fc754bf1..545696971f65 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h @@ -203,6 +203,8 @@ struct stmmac_priv { int eee_enabled; int eee_active; int tx_lpi_timer; + int tx_lpi_enabled; + int eee_tw_timer; unsigned int mode; unsigned int chain_mode; int extend_desc; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c index 430a4b32ec1e..814879f91f76 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_ethtool.c @@ -665,6 +665,7 @@ static int stmmac_ethtool_op_get_eee(struct net_device *dev, edata->eee_enabled = priv->eee_enabled; edata->eee_active = priv->eee_active; edata->tx_lpi_timer = priv->tx_lpi_timer; + edata->tx_lpi_enabled = priv->tx_lpi_enabled; return phylink_ethtool_get_eee(priv->phylink, edata); } @@ -678,6 +679,10 @@ static int stmmac_ethtool_op_set_eee(struct net_device *dev, if (!priv->dma_cap.eee) return -EOPNOTSUPP; + if (priv->tx_lpi_enabled != edata->tx_lpi_enabled) + netdev_warn(priv->dev, + "Setting EEE tx-lpi is not supported\n"); + if (!edata->eee_enabled) stmmac_disable_eee_mode(priv); @@ -685,7 +690,12 @@ static int stmmac_ethtool_op_set_eee(struct net_device *dev, if (ret) return ret; - priv->tx_lpi_timer = edata->tx_lpi_timer; + if (edata->eee_enabled && + priv->tx_lpi_timer != edata->tx_lpi_timer) { + priv->tx_lpi_timer = edata->tx_lpi_timer; + stmmac_eee_init(priv); + } + return 0; } diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 89b2b3472852..b56b13d64ab4 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -94,7 +94,7 @@ static const u32 default_msg_level = (NETIF_MSG_DRV | NETIF_MSG_PROBE | static int eee_timer = STMMAC_DEFAULT_LPI_TIMER; module_param(eee_timer, int, 0644); MODULE_PARM_DESC(eee_timer, "LPI tx expiration time in msec"); -#define STMMAC_LPI_T(x) (jiffies + msecs_to_jiffies(x)) +#define STMMAC_LPI_T(x) (jiffies + usecs_to_jiffies(x)) /* By default the driver will use the ring mode to manage tx and rx descriptors, * but allow user to force to use the chain instead of the ring @@ -370,7 +370,7 @@ static void stmmac_eee_ctrl_timer(struct timer_list *t) struct stmmac_priv *priv = from_timer(priv, t, eee_ctrl_timer); stmmac_enable_eee_mode(priv); - mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(eee_timer)); + mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(priv->tx_lpi_timer)); } /** @@ -383,7 +383,7 @@ static void stmmac_eee_ctrl_timer(struct timer_list *t) */ bool stmmac_eee_init(struct stmmac_priv *priv) { - int tx_lpi_timer = priv->tx_lpi_timer; + int eee_tw_timer = priv->eee_tw_timer; /* Using PCS we cannot dial with the phy registers at this stage * so we do not support extra feature like EEE. @@ -403,7 +403,7 @@ bool stmmac_eee_init(struct stmmac_priv *priv) if (priv->eee_enabled) { netdev_dbg(priv->dev, "disable EEE\n"); del_timer_sync(&priv->eee_ctrl_timer); - stmmac_set_eee_timer(priv, priv->hw, 0, tx_lpi_timer); + stmmac_set_eee_timer(priv, priv->hw, 0, eee_tw_timer); } mutex_unlock(&priv->lock); return false; @@ -411,11 +411,12 @@ bool stmmac_eee_init(struct stmmac_priv *priv) if (priv->eee_active && !priv->eee_enabled) { timer_setup(&priv->eee_ctrl_timer, stmmac_eee_ctrl_timer, 0); - mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(eee_timer)); stmmac_set_eee_timer(priv, priv->hw, STMMAC_DEFAULT_LIT_LS, - tx_lpi_timer); + eee_tw_timer); } + mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(priv->tx_lpi_timer)); + mutex_unlock(&priv->lock); netdev_dbg(priv->dev, "Energy-Efficient Ethernet initialized\n"); return true; @@ -930,6 +931,7 @@ static void stmmac_mac_link_down(struct phylink_config *config, stmmac_mac_set(priv, priv->ioaddr, false); priv->eee_active = false; + priv->tx_lpi_enabled = false; stmmac_eee_init(priv); stmmac_set_eee_pls(priv, priv->hw, false); } @@ -1027,6 +1029,7 @@ static void stmmac_mac_link_up(struct phylink_config *config, if (phy && priv->dma_cap.eee) { priv->eee_active = phy_init_eee(phy, 1) >= 0; priv->eee_enabled = stmmac_eee_init(priv); + priv->tx_lpi_enabled = priv->eee_enabled; stmmac_set_eee_pls(priv, priv->hw, true); } } @@ -2061,7 +2064,7 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) if ((priv->eee_enabled) && (!priv->tx_path_in_lpi_mode)) { stmmac_enable_eee_mode(priv); - mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(eee_timer)); + mod_timer(&priv->eee_ctrl_timer, STMMAC_LPI_T(priv->tx_lpi_timer)); } /* We still have pending packets, let's call for a new scheduling */ @@ -2694,7 +2697,11 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp) netdev_warn(priv->dev, "PTP init failed\n"); } - priv->tx_lpi_timer = STMMAC_DEFAULT_TWT_LS; + priv->eee_tw_timer = STMMAC_DEFAULT_TWT_LS; + + /* Convert the timer from msec to usec */ + if (!priv->tx_lpi_timer) + priv->tx_lpi_timer = eee_timer * 1000; if (priv->use_riwt) { if (!priv->rx_riwt) -- cgit From 1f7e877c20517735bceff1535e1b7fa846b2f215 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 1 Oct 2020 10:54:49 -0700 Subject: net: hinic: fix DEVLINK build errors Fix many (lots deleted here) build errors in hinic by selecting NET_DEVLINK. ld: drivers/net/ethernet/huawei/hinic/hinic_hw_dev.o: in function `mgmt_watchdog_timeout_event_handler': hinic_hw_dev.c:(.text+0x30a): undefined reference to `devlink_health_report' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_fw_reporter_dump': hinic_devlink.c:(.text+0x1c): undefined reference to `devlink_fmsg_u32_pair_put' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_fw_reporter_dump': hinic_devlink.c:(.text+0x126): undefined reference to `devlink_fmsg_binary_pair_put' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_hw_reporter_dump': hinic_devlink.c:(.text+0x1ba): undefined reference to `devlink_fmsg_string_pair_put' ld: hinic_devlink.c:(.text+0x227): undefined reference to `devlink_fmsg_u8_pair_put' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_devlink_alloc': hinic_devlink.c:(.text+0xaee): undefined reference to `devlink_alloc' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_devlink_free': hinic_devlink.c:(.text+0xb04): undefined reference to `devlink_free' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_devlink_register': hinic_devlink.c:(.text+0xb26): undefined reference to `devlink_register' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_devlink_unregister': hinic_devlink.c:(.text+0xb46): undefined reference to `devlink_unregister' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_health_reporters_create': hinic_devlink.c:(.text+0xb75): undefined reference to `devlink_health_reporter_create' ld: hinic_devlink.c:(.text+0xb95): undefined reference to `devlink_health_reporter_create' ld: hinic_devlink.c:(.text+0xbac): undefined reference to `devlink_health_reporter_destroy' ld: drivers/net/ethernet/huawei/hinic/hinic_devlink.o: in function `hinic_health_reporters_destroy': Fixes: 51ba902a16e6 ("net-next/hinic: Initialize hw interface") Signed-off-by: Randy Dunlap Cc: Bin Luo Cc: "David S. Miller" Cc: Jakub Kicinski Cc: Aviad Krawczyk Cc: Zhao Chen Signed-off-by: David S. Miller --- drivers/net/ethernet/huawei/hinic/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/huawei/hinic/Kconfig b/drivers/net/ethernet/huawei/hinic/Kconfig index 936e2dd3bb13..b47bd5440c5f 100644 --- a/drivers/net/ethernet/huawei/hinic/Kconfig +++ b/drivers/net/ethernet/huawei/hinic/Kconfig @@ -6,6 +6,7 @@ config HINIC tristate "Huawei Intelligent PCIE Network Interface Card" depends on (PCI_MSI && (X86 || ARM64)) + select NET_DEVLINK help This driver supports HiNIC PCIE Ethernet cards. To compile this driver as part of the kernel, choose Y here. -- cgit From 790ca79d3e7f76d77291540e1728f400c6ad931f Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Fri, 2 Oct 2020 21:47:43 +0200 Subject: net: typhoon: Fix a typo Typoon --> Typhoon s/Typoon/Typhoon/ Signed-off-by: Christophe JAILLET Signed-off-by: David S. Miller --- drivers/net/ethernet/3com/typhoon.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/3com/typhoon.h b/drivers/net/ethernet/3com/typhoon.h index 2f634c64d5d1..38e6dcab4e94 100644 --- a/drivers/net/ethernet/3com/typhoon.h +++ b/drivers/net/ethernet/3com/typhoon.h @@ -33,7 +33,7 @@ struct basic_ring { u32 lastWrite; }; -/* The Typoon transmit ring -- same as a basic ring, plus: +/* The Typhoon transmit ring -- same as a basic ring, plus: * lastRead: where we're at in regard to cleaning up the ring * writeRegister: register to use for writing (different for Hi & Lo rings) */ -- cgit From f4544e5361da5050ff5c0330ceea095cb5dbdd72 Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Sat, 3 Oct 2020 11:51:21 -0700 Subject: net: mvneta: fix double free of txq->buf clang static analysis reports this problem: drivers/net/ethernet/marvell/mvneta.c:3465:2: warning: Attempt to free released memory kfree(txq->buf); ^~~~~~~~~~~~~~~ When mvneta_txq_sw_init() fails to alloc txq->tso_hdrs, it frees without poisoning txq->buf. The error is caught in the mvneta_setup_txqs() caller which handles the error by cleaning up all of the txqs with a call to mvneta_txq_sw_deinit which also frees txq->buf. Since mvneta_txq_sw_deinit is a general cleaner, all of the partial cleaning in mvneta_txq_sw_deinit()'s error handling is not needed. Fixes: 2adb719d74f6 ("net: mvneta: Implement software TSO") Signed-off-by: Tom Rix Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/mvneta.c | 13 ++----------- 1 file changed, 2 insertions(+), 11 deletions(-) (limited to 'drivers/net/ethernet') diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index c4345e3d616f..5bf0409f5d42 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -3400,24 +3400,15 @@ static int mvneta_txq_sw_init(struct mvneta_port *pp, txq->last_desc = txq->size - 1; txq->buf = kmalloc_array(txq->size, sizeof(*txq->buf), GFP_KERNEL); - if (!txq->buf) { - dma_free_coherent(pp->dev->dev.parent, - txq->size * MVNETA_DESC_ALIGNED_SIZE, - txq->descs, txq->descs_phys); + if (!txq->buf) return -ENOMEM; - } /* Allocate DMA buffers for TSO MAC/IP/TCP headers */ txq->tso_hdrs = dma_alloc_coherent(pp->dev->dev.parent, txq->size * TSO_HEADER_SIZE, &txq->tso_hdrs_phys, GFP_KERNEL); - if (!txq->tso_hdrs) { - kfree(txq->buf); - dma_free_coherent(pp->dev->dev.parent, - txq->size * MVNETA_DESC_ALIGNED_SIZE, - txq->descs, txq->descs_phys); + if (!txq->tso_hdrs) return -ENOMEM; - } /* Setup XPS mapping */ if (txq_number > 1) -- cgit