From ce317dd9f809c8da9656c88761e30f0a82a8c2e6 Mon Sep 17 00:00:00 2001 From: Anirudh Venkataramanan Date: Fri, 26 Oct 2018 10:40:50 -0700 Subject: ice: Set carrier state and start/stop queues in rebuild Set the carrier state post rebuild by querying the link status. Also start/stop queues based on link status. Signed-off-by: Anirudh Venkataramanan Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/ice/ice_main.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) (limited to 'drivers/net/ethernet/intel/ice/ice_main.c') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 05993451147a..6d31ffb64940 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -3296,7 +3296,7 @@ static void ice_rebuild(struct ice_pf *pf) struct device *dev = &pf->pdev->dev; struct ice_hw *hw = &pf->hw; enum ice_status ret; - int err; + int err, i; if (test_bit(__ICE_DOWN, pf->state)) goto clear_recovery; @@ -3370,6 +3370,22 @@ static void ice_rebuild(struct ice_pf *pf) } ice_reset_all_vfs(pf, true); + + for (i = 0; i < pf->num_alloc_vsi; i++) { + bool link_up; + + if (!pf->vsi[i] || pf->vsi[i]->type != ICE_VSI_PF) + continue; + ice_get_link_status(pf->vsi[i]->port_info, &link_up); + if (link_up) { + netif_carrier_on(pf->vsi[i]->netdev); + netif_tx_wake_all_queues(pf->vsi[i]->netdev); + } else { + netif_carrier_off(pf->vsi[i]->netdev); + netif_tx_stop_all_queues(pf->vsi[i]->netdev); + } + } + /* if we get here, reset flow is successful */ clear_bit(__ICE_RESET_FAILED, pf->state); return; -- cgit From afd9d4ab58db20029a75cf82f23b6a5641cd7d6f Mon Sep 17 00:00:00 2001 From: Anirudh Venkataramanan Date: Fri, 26 Oct 2018 10:40:51 -0700 Subject: ice: Check for reset in progress during remove The remove path does not currently check to see if a reset is in progress before proceeding. This can cause a resource collision resulting in various types of errors. Check for reset in progress and wait for a reasonable amount of time before allowing the remove to progress. Signed-off-by: Dave Ertman Signed-off-by: Anirudh Venkataramanan Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/ice/ice_main.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'drivers/net/ethernet/intel/ice/ice_main.c') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 6d31ffb64940..aee22f11a41a 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -2182,6 +2182,12 @@ static void ice_remove(struct pci_dev *pdev) if (!pf) return; + for (i = 0; i < ICE_MAX_RESET_WAIT; i++) { + if (!ice_is_reset_in_progress(pf->state)) + break; + msleep(100); + } + set_bit(__ICE_DOWN, pf->state); ice_service_task_stop(pf); -- cgit From 9ecd25c26810a61b9c3abc6c73de32dca6da96e1 Mon Sep 17 00:00:00 2001 From: Anirudh Venkataramanan Date: Fri, 26 Oct 2018 10:40:54 -0700 Subject: ice: Remove duplicate addition of VLANs in replay path ice_restore_vlan and active_vlans were originally put in place to reprogram VLAN filters in the replay path. This is now done as part of the much broader VSI rebuild/replay framework. So remove both ice_restore_vlan and active_vlans Signed-off-by: Anirudh Venkataramanan Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/ice/ice_main.c | 42 +++++-------------------------- 1 file changed, 6 insertions(+), 36 deletions(-) (limited to 'drivers/net/ethernet/intel/ice/ice_main.c') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index aee22f11a41a..338abb1b9233 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -1622,7 +1622,6 @@ static int ice_vlan_rx_add_vid(struct net_device *netdev, { struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_vsi *vsi = np->vsi; - int ret; if (vid >= VLAN_N_VID) { netdev_err(netdev, "VLAN id requested %d is out of range %d\n", @@ -1635,7 +1634,8 @@ static int ice_vlan_rx_add_vid(struct net_device *netdev, /* Enable VLAN pruning when VLAN 0 is added */ if (unlikely(!vid)) { - ret = ice_cfg_vlan_pruning(vsi, true); + int ret = ice_cfg_vlan_pruning(vsi, true); + if (ret) return ret; } @@ -1644,12 +1644,7 @@ static int ice_vlan_rx_add_vid(struct net_device *netdev, * needed to continue allowing all untagged packets since VLAN prune * list is applied to all packets by the switch */ - ret = ice_vsi_add_vlan(vsi, vid); - - if (!ret) - set_bit(vid, vsi->active_vlans); - - return ret; + return ice_vsi_add_vlan(vsi, vid); } /** @@ -1677,8 +1672,6 @@ static int ice_vlan_rx_kill_vid(struct net_device *netdev, if (status) return status; - clear_bit(vid, vsi->active_vlans); - /* Disable VLAN pruning when VLAN 0 is removed */ if (unlikely(!vid)) status = ice_cfg_vlan_pruning(vsi, false); @@ -2515,31 +2508,6 @@ static int ice_vsi_vlan_setup(struct ice_vsi *vsi) return ret; } -/** - * ice_restore_vlan - Reinstate VLANs when vsi/netdev comes back up - * @vsi: the VSI being brought back up - */ -static int ice_restore_vlan(struct ice_vsi *vsi) -{ - int err; - u16 vid; - - if (!vsi->netdev) - return -EINVAL; - - err = ice_vsi_vlan_setup(vsi); - if (err) - return err; - - for_each_set_bit(vid, vsi->active_vlans, VLAN_N_VID) { - err = ice_vlan_rx_add_vid(vsi->netdev, htons(ETH_P_8021Q), vid); - if (err) - break; - } - - return err; -} - /** * ice_vsi_cfg - Setup the VSI * @vsi: the VSI being configured @@ -2552,7 +2520,9 @@ static int ice_vsi_cfg(struct ice_vsi *vsi) if (vsi->netdev) { ice_set_rx_mode(vsi->netdev); - err = ice_restore_vlan(vsi); + + err = ice_vsi_vlan_setup(vsi); + if (err) return err; } -- cgit From 25525b69bb44a628841492f44a5a8e74f34724f4 Mon Sep 17 00:00:00 2001 From: Dave Ertman Date: Fri, 26 Oct 2018 10:40:57 -0700 Subject: ice: Fix napi delete calls for remove In the remove path, the vsi->netdev is being set to NULL before the call to free vectors. This is causing the netif_napi_del call to never be made. Add a call to ice_napi_del to the same location as the calls to unregister_netdev and just prior to them. This will use the reverse flow as the register and netif_napi_add calls. Signed-off-by: Dave Ertman Signed-off-by: Anirudh Venkataramanan Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/ice/ice_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/net/ethernet/intel/ice/ice_main.c') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 338abb1b9233..82f49dbd762c 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -1465,7 +1465,7 @@ skip_req_irq: * ice_napi_del - Remove NAPI handler for the VSI * @vsi: VSI for which NAPI handler is to be removed */ -static void ice_napi_del(struct ice_vsi *vsi) +void ice_napi_del(struct ice_vsi *vsi) { int v_idx; -- cgit From c585ea42ec75e8d3afa278b7095d9f0dd6ee515b Mon Sep 17 00:00:00 2001 From: Brett Creeley Date: Fri, 26 Oct 2018 10:40:58 -0700 Subject: ice: Fix tx_timeout in PF driver Prior to this commit the driver was running into tx_timeouts when a queue was stressed enough. This was happening because the HW tail and SW tail (NTU) were incorrectly out of sync. Consequently this was causing the HW head to collide with the HW tail, which to the hardware means that all descriptors posted for Tx have been processed. Due to the Tx logic used in the driver SW tail and HW tail are allowed to be out of sync. This is done as an optimization because it allows the driver to write HW tail as infrequently as possible, while still updating the SW tail index to keep track. However, there are situations where this results in the tail never getting updated, resulting in Tx timeouts. Tx HW tail write condition: if (netif_xmit_stopped(txring_txq(tx_ring) || !skb->xmit_more) writel(sw_tail, tx_ring->tail); An issue was found in the Tx logic that was causing the afore mentioned condition for updating HW tail to never happen, causing tx_timeouts. In ice_xmit_frame_ring we calculate how many descriptors we need for the Tx transaction based on the skb the kernel hands us. This is then passed into ice_maybe_stop_tx along with some extra padding to determine if we have enough descriptors available for this transaction. If we don't then we return -EBUSY to the stack, otherwise we move on and eventually prepare the Tx descriptors accordingly in ice_tx_map and set next_to_watch. In ice_tx_map we make another call to ice_maybe_stop_tx with a value of MAX_SKB_FRAGS + 4. The key here is that this value is possibly less than the value we sent in the first call to ice_maybe_stop_tx in ice_xmit_frame_ring. Now, if the number of unused descriptors is between MAX_SKB_FRAGS + 4 and the value used in the first call to ice_maybe_stop_tx in ice_xmit_frame_ring then we do not update the HW tail because of the "Tx HW tail write condition" above. This is because in ice_maybe_stop_tx we return success from ice_maybe_stop_tx instead of calling __ice_maybe_stop_tx and subsequently calling netif_stop_subqueue, which sets the __QUEUE_STATE_DEV_XOFF bit. This bit is then checked in the "Tx HW tail write condition" by calling netif_xmit_stopped and subsequently updating HW tail if the afore mentioned bit is set. In ice_clean_tx_irq, if next_to_watch is not NULL, we end up cleaning the descriptors that HW sets the DD bit on and we have the budget. The HW head will eventually run into the HW tail in response to the description in the paragraph above. The next time through ice_xmit_frame_ring we make the initial call to ice_maybe_stop_tx with another skb from the stack. This time we do not have enough descriptors available and we return NETDEV_TX_BUSY to the stack and end up setting next_to_watch to NULL. This is where we are stuck. In ice_clean_tx_irq we never clean anything because next_to_watch is always NULL and in ice_xmit_frame_ring we never update HW tail because we already return NETDEV_TX_BUSY to the stack and eventually we hit a tx_timeout. This issue was fixed by making sure that the second call to ice_maybe_stop_tx in ice_tx_map is passed a value that is >= the value that was used on the initial call to ice_maybe_stop_tx in ice_xmit_frame_ring. This was done by adding the following defines to make the logic more clear and to reduce the chance of mucking this up again: ICE_CACHE_LINE_BYTES 64 ICE_DESCS_PER_CACHE_LINE (ICE_CACHE_LINE_BYTES / \ sizeof(struct ice_tx_desc)) ICE_DESCS_FOR_CTX_DESC 1 ICE_DESCS_FOR_SKB_DATA_PTR 1 The ICE_CACHE_LINE_BYTES being 64 is an assumption being made so we don't have to figure this out on every pass through the Tx path. Instead I added a sanity check in ice_probe to verify cache line size and print a message if it's not 64 Bytes. This will make it easier to file issues if they are seen when the cache line size is not 64 Bytes when reading from the GLPCI_CNF2 register. Signed-off-by: Brett Creeley Signed-off-by: Anirudh Venkataramanan Tested-by: Andrew Bowers Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/ice/ice_main.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'drivers/net/ethernet/intel/ice/ice_main.c') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 82f49dbd762c..333312a1d595 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -1994,6 +1994,22 @@ static int ice_init_interrupt_scheme(struct ice_pf *pf) return 0; } +/** + * ice_verify_cacheline_size - verify driver's assumption of 64 Byte cache lines + * @pf: pointer to the PF structure + * + * There is no error returned here because the driver should be able to handle + * 128 Byte cache lines, so we only print a warning in case issues are seen, + * specifically with Tx. + */ +static void ice_verify_cacheline_size(struct ice_pf *pf) +{ + if (rd32(&pf->hw, GLPCI_CNF2) & GLPCI_CNF2_CACHELINE_SIZE_M) + dev_warn(&pf->pdev->dev, + "%d Byte cache line assumption is invalid, driver may have Tx timeouts!\n", + ICE_CACHE_LINE_BYTES); +} + /** * ice_probe - Device initialization routine * @pdev: PCI device information struct @@ -2144,6 +2160,8 @@ static int ice_probe(struct pci_dev *pdev, /* since everything is good, start the service timer */ mod_timer(&pf->serv_tmr, round_jiffies(jiffies + pf->serv_tmr_period)); + ice_verify_cacheline_size(pf); + return 0; err_alloc_sw_unroll: -- cgit