diff options
Diffstat (limited to 'drivers/net/ethernet/cisco/enic/enic_main.c')
-rw-r--r-- | drivers/net/ethernet/cisco/enic/enic_main.c | 555 |
1 files changed, 384 insertions, 171 deletions
diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index d266a87297a5..49f6cab01ed5 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -46,6 +46,7 @@ #include <linux/crash_dump.h> #include <net/busy_poll.h> #include <net/vxlan.h> +#include <net/netdev_queues.h> #include "cq_enet_desc.h" #include "vnic_dev.h" @@ -108,7 +109,7 @@ static struct enic_intr_mod_table mod_table[ENIC_MAX_COALESCE_TIMERS + 1] = { static struct enic_intr_mod_range mod_range[ENIC_MAX_LINK_SPEEDS] = { {0, 0}, /* 0 - 4 Gbps */ {0, 3}, /* 4 - 10 Gbps */ - {3, 6}, /* 10 - 40 Gbps */ + {3, 6}, /* 10+ Gbps */ }; static void enic_init_affinity_hint(struct enic *enic) @@ -339,6 +340,10 @@ static void enic_free_wq_buf(struct vnic_wq *wq, struct vnic_wq_buf *buf) static void enic_wq_free_buf(struct vnic_wq *wq, struct cq_desc *cq_desc, struct vnic_wq_buf *buf, void *opaque) { + struct enic *enic = vnic_dev_priv(wq->vdev); + + enic->wq[wq->index].stats.cq_work++; + enic->wq[wq->index].stats.cq_bytes += buf->len; enic_free_wq_buf(wq, buf); } @@ -347,18 +352,20 @@ static int enic_wq_service(struct vnic_dev *vdev, struct cq_desc *cq_desc, { struct enic *enic = vnic_dev_priv(vdev); - spin_lock(&enic->wq_lock[q_number]); + spin_lock(&enic->wq[q_number].lock); - vnic_wq_service(&enic->wq[q_number], cq_desc, + vnic_wq_service(&enic->wq[q_number].vwq, cq_desc, completed_index, enic_wq_free_buf, opaque); if (netif_tx_queue_stopped(netdev_get_tx_queue(enic->netdev, q_number)) && - vnic_wq_desc_avail(&enic->wq[q_number]) >= - (MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS)) + vnic_wq_desc_avail(&enic->wq[q_number].vwq) >= + (MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS)) { netif_wake_subqueue(enic->netdev, q_number); + enic->wq[q_number].stats.wake++; + } - spin_unlock(&enic->wq_lock[q_number]); + spin_unlock(&enic->wq[q_number].lock); return 0; } @@ -370,7 +377,7 @@ static bool enic_log_q_error(struct enic *enic) bool err = false; for (i = 0; i < enic->wq_count; i++) { - error_status = vnic_wq_error_status(&enic->wq[i]); + error_status = vnic_wq_error_status(&enic->wq[i].vwq); err |= error_status; if (error_status) netdev_err(enic->netdev, "WQ[%d] error_status %d\n", @@ -378,7 +385,7 @@ static bool enic_log_q_error(struct enic *enic) } for (i = 0; i < enic->rq_count; i++) { - error_status = vnic_rq_error_status(&enic->rq[i]); + error_status = vnic_rq_error_status(&enic->rq[i].vrq); err |= error_status; if (error_status) netdev_err(enic->netdev, "RQ[%d] error_status %d\n", @@ -421,6 +428,36 @@ static void enic_mtu_check(struct enic *enic) } } +static void enic_set_rx_coal_setting(struct enic *enic) +{ + unsigned int speed; + int index = -1; + struct enic_rx_coal *rx_coal = &enic->rx_coalesce_setting; + + /* 1. Read the link speed from fw + * 2. Pick the default range for the speed + * 3. Update it in enic->rx_coalesce_setting + */ + speed = vnic_dev_port_speed(enic->vdev); + if (speed > ENIC_LINK_SPEED_10G) + index = ENIC_LINK_40G_INDEX; + else if (speed > ENIC_LINK_SPEED_4G) + index = ENIC_LINK_10G_INDEX; + else + index = ENIC_LINK_4G_INDEX; + + rx_coal->small_pkt_range_start = mod_range[index].small_pkt_range_start; + rx_coal->large_pkt_range_start = mod_range[index].large_pkt_range_start; + rx_coal->range_end = ENIC_RX_COALESCE_RANGE_END; + + /* Start with the value provided by UCSM */ + for (index = 0; index < enic->rq_count; index++) + enic->cq[index].cur_rx_coal_timeval = + enic->config.intr_timer_usec; + + rx_coal->use_adaptive_rx_coalesce = 1; +} + static void enic_link_check(struct enic *enic) { int link_status = vnic_dev_link_status(enic->vdev); @@ -429,6 +466,7 @@ static void enic_link_check(struct enic *enic) if (link_status && !carrier_ok) { netdev_info(enic->netdev, "Link UP\n"); netif_carrier_on(enic->netdev); + enic_set_rx_coal_setting(enic); } else if (!link_status && carrier_ok) { netdev_info(enic->netdev, "Link DOWN\n"); netif_carrier_off(enic->netdev); @@ -590,6 +628,11 @@ static int enic_queue_wq_skb_vlan(struct enic *enic, struct vnic_wq *wq, if (!eop) err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + /* The enic_queue_wq_desc() above does not do HW checksum */ + enic->wq[wq->index].stats.csum_none++; + enic->wq[wq->index].stats.packets++; + enic->wq[wq->index].stats.bytes += skb->len; + return err; } @@ -622,6 +665,10 @@ static int enic_queue_wq_skb_csum_l4(struct enic *enic, struct vnic_wq *wq, if (!eop) err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + enic->wq[wq->index].stats.csum_partial++; + enic->wq[wq->index].stats.packets++; + enic->wq[wq->index].stats.bytes += skb->len; + return err; } @@ -676,15 +723,18 @@ static int enic_queue_wq_skb_tso(struct enic *enic, struct vnic_wq *wq, unsigned int offset = 0; unsigned int hdr_len; dma_addr_t dma_addr; + unsigned int pkts; unsigned int len; skb_frag_t *frag; if (skb->encapsulation) { hdr_len = skb_inner_tcp_all_headers(skb); enic_preload_tcp_csum_encap(skb); + enic->wq[wq->index].stats.encap_tso++; } else { hdr_len = skb_tcp_all_headers(skb); enic_preload_tcp_csum(skb); + enic->wq[wq->index].stats.tso++; } /* Queue WQ_ENET_MAX_DESC_LEN length descriptors @@ -705,7 +755,7 @@ static int enic_queue_wq_skb_tso(struct enic *enic, struct vnic_wq *wq, } if (eop) - return 0; + goto tso_out_stats; /* Queue WQ_ENET_MAX_DESC_LEN length descriptors * for additional data fragments @@ -732,6 +782,15 @@ static int enic_queue_wq_skb_tso(struct enic *enic, struct vnic_wq *wq, } } +tso_out_stats: + /* calculate how many packets tso sent */ + len = skb->len - hdr_len; + pkts = len / mss; + if ((len % mss) > 0) + pkts++; + enic->wq[wq->index].stats.packets += pkts; + enic->wq[wq->index].stats.bytes += (len + (pkts * hdr_len)); + return 0; } @@ -764,6 +823,10 @@ static inline int enic_queue_wq_skb_encap(struct enic *enic, struct vnic_wq *wq, if (!eop) err = enic_queue_wq_skb_cont(enic, wq, skb, len_left, loopback); + enic->wq[wq->index].stats.encap_csum++; + enic->wq[wq->index].stats.packets++; + enic->wq[wq->index].stats.bytes += skb->len; + return err; } @@ -780,6 +843,7 @@ static inline int enic_queue_wq_skb(struct enic *enic, /* VLAN tag from trunking driver */ vlan_tag_insert = 1; vlan_tag = skb_vlan_tag_get(skb); + enic->wq[wq->index].stats.add_vlan++; } else if (enic->loop_enable) { vlan_tag = enic->loop_tag; loopback = 1; @@ -792,7 +856,7 @@ static inline int enic_queue_wq_skb(struct enic *enic, else if (skb->encapsulation) err = enic_queue_wq_skb_encap(enic, wq, skb, vlan_tag_insert, vlan_tag, loopback); - else if (skb->ip_summed == CHECKSUM_PARTIAL) + else if (skb->ip_summed == CHECKSUM_PARTIAL) err = enic_queue_wq_skb_csum_l4(enic, wq, skb, vlan_tag_insert, vlan_tag, loopback); else @@ -825,13 +889,15 @@ static netdev_tx_t enic_hard_start_xmit(struct sk_buff *skb, unsigned int txq_map; struct netdev_queue *txq; + txq_map = skb_get_queue_mapping(skb) % enic->wq_count; + wq = &enic->wq[txq_map].vwq; + if (skb->len <= 0) { dev_kfree_skb_any(skb); + enic->wq[wq->index].stats.null_pkt++; return NETDEV_TX_OK; } - txq_map = skb_get_queue_mapping(skb) % enic->wq_count; - wq = &enic->wq[txq_map]; txq = netdev_get_tx_queue(netdev, txq_map); /* Non-TSO sends must fit within ENIC_NON_TSO_MAX_DESC descs, @@ -843,31 +909,35 @@ static netdev_tx_t enic_hard_start_xmit(struct sk_buff *skb, skb_shinfo(skb)->nr_frags + 1 > ENIC_NON_TSO_MAX_DESC && skb_linearize(skb)) { dev_kfree_skb_any(skb); + enic->wq[wq->index].stats.skb_linear_fail++; return NETDEV_TX_OK; } - spin_lock(&enic->wq_lock[txq_map]); + spin_lock(&enic->wq[txq_map].lock); if (vnic_wq_desc_avail(wq) < skb_shinfo(skb)->nr_frags + ENIC_DESC_MAX_SPLITS) { netif_tx_stop_queue(txq); /* This is a hard error, log it */ netdev_err(netdev, "BUG! Tx ring full when queue awake!\n"); - spin_unlock(&enic->wq_lock[txq_map]); + spin_unlock(&enic->wq[txq_map].lock); + enic->wq[wq->index].stats.desc_full_awake++; return NETDEV_TX_BUSY; } if (enic_queue_wq_skb(enic, wq, skb)) goto error; - if (vnic_wq_desc_avail(wq) < MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS) + if (vnic_wq_desc_avail(wq) < MAX_SKB_FRAGS + ENIC_DESC_MAX_SPLITS) { netif_tx_stop_queue(txq); + enic->wq[wq->index].stats.stopped++; + } skb_tx_timestamp(skb); if (!netdev_xmit_more() || netif_xmit_stopped(txq)) vnic_wq_doorbell(wq); error: - spin_unlock(&enic->wq_lock[txq_map]); + spin_unlock(&enic->wq[txq_map].lock); return NETDEV_TX_OK; } @@ -878,7 +948,10 @@ static void enic_get_stats(struct net_device *netdev, { struct enic *enic = netdev_priv(netdev); struct vnic_stats *stats; + u64 pkt_truncated = 0; + u64 bad_fcs = 0; int err; + int i; err = enic_dev_stats_dump(enic, &stats); /* return only when dma_alloc_coherent fails in vnic_dev_stats_dump @@ -897,8 +970,17 @@ static void enic_get_stats(struct net_device *netdev, net_stats->rx_bytes = stats->rx.rx_bytes_ok; net_stats->rx_errors = stats->rx.rx_errors; net_stats->multicast = stats->rx.rx_multicast_frames_ok; - net_stats->rx_over_errors = enic->rq_truncated_pkts; - net_stats->rx_crc_errors = enic->rq_bad_fcs; + + for (i = 0; i < enic->rq_count; i++) { + struct enic_rq_stats *rqs = &enic->rq[i].stats; + + if (!enic->rq[i].vrq.ctrl) + break; + pkt_truncated += rqs->pkt_truncated; + bad_fcs += rqs->bad_fcs; + } + net_stats->rx_over_errors = pkt_truncated; + net_stats->rx_crc_errors = bad_fcs; net_stats->rx_dropped = stats->rx.rx_no_bufs + stats->rx.rx_drop; } @@ -1117,18 +1199,30 @@ static int enic_set_vf_port(struct net_device *netdev, int vf, pp->request = nla_get_u8(port[IFLA_PORT_REQUEST]); if (port[IFLA_PORT_PROFILE]) { + if (nla_len(port[IFLA_PORT_PROFILE]) != PORT_PROFILE_MAX) { + memcpy(pp, &prev_pp, sizeof(*pp)); + return -EINVAL; + } pp->set |= ENIC_SET_NAME; memcpy(pp->name, nla_data(port[IFLA_PORT_PROFILE]), PORT_PROFILE_MAX); } if (port[IFLA_PORT_INSTANCE_UUID]) { + if (nla_len(port[IFLA_PORT_INSTANCE_UUID]) != PORT_UUID_MAX) { + memcpy(pp, &prev_pp, sizeof(*pp)); + return -EINVAL; + } pp->set |= ENIC_SET_INSTANCE; memcpy(pp->instance_uuid, nla_data(port[IFLA_PORT_INSTANCE_UUID]), PORT_UUID_MAX); } if (port[IFLA_PORT_HOST_UUID]) { + if (nla_len(port[IFLA_PORT_HOST_UUID]) != PORT_UUID_MAX) { + memcpy(pp, &prev_pp, sizeof(*pp)); + return -EINVAL; + } pp->set |= ENIC_SET_HOST; memcpy(pp->host_uuid, nla_data(port[IFLA_PORT_HOST_UUID]), PORT_UUID_MAX); @@ -1249,8 +1343,10 @@ static int enic_rq_alloc_buf(struct vnic_rq *rq) return 0; } skb = netdev_alloc_skb_ip_align(netdev, len); - if (!skb) + if (!skb) { + enic->rq[rq->index].stats.no_skb++; return -ENOMEM; + } dma_addr = dma_map_single(&enic->pdev->dev, skb->data, len, DMA_FROM_DEVICE); @@ -1301,6 +1397,7 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, struct net_device *netdev = enic->netdev; struct sk_buff *skb; struct vnic_cq *cq = &enic->cq[enic_cq_rq(enic, rq->index)]; + struct enic_rq_stats *rqstats = &enic->rq[rq->index].stats; u8 type, color, eop, sop, ingress_port, vlan_stripped; u8 fcoe, fcoe_sof, fcoe_fc_crc_ok, fcoe_enc_error, fcoe_eof; @@ -1311,8 +1408,11 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, u32 rss_hash; bool outer_csum_ok = true, encap = false; - if (skipped) + rqstats->packets++; + if (skipped) { + rqstats->desc_skip++; return; + } skb = buf->os_buf; @@ -1330,9 +1430,9 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, if (!fcs_ok) { if (bytes_written > 0) - enic->rq_bad_fcs++; + rqstats->bad_fcs++; else if (bytes_written == 0) - enic->rq_truncated_pkts++; + rqstats->pkt_truncated++; } dma_unmap_single(&enic->pdev->dev, buf->dma_addr, buf->len, @@ -1347,7 +1447,7 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, /* Good receive */ - + rqstats->bytes += bytes_written; if (!enic_rxcopybreak(netdev, &skb, buf, bytes_written)) { buf->os_buf = NULL; dma_unmap_single(&enic->pdev->dev, buf->dma_addr, @@ -1365,11 +1465,13 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, case CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6: case CQ_ENET_RQ_DESC_RSS_TYPE_TCP_IPv6_EX: skb_set_hash(skb, rss_hash, PKT_HASH_TYPE_L4); + rqstats->l4_rss_hash++; break; case CQ_ENET_RQ_DESC_RSS_TYPE_IPv4: case CQ_ENET_RQ_DESC_RSS_TYPE_IPv6: case CQ_ENET_RQ_DESC_RSS_TYPE_IPv6_EX: skb_set_hash(skb, rss_hash, PKT_HASH_TYPE_L3); + rqstats->l3_rss_hash++; break; } } @@ -1406,11 +1508,16 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, (ipv4_csum_ok || ipv6)) { skb->ip_summed = CHECKSUM_UNNECESSARY; skb->csum_level = encap; + if (encap) + rqstats->csum_unnecessary_encap++; + else + rqstats->csum_unnecessary++; } - if (vlan_stripped) + if (vlan_stripped) { __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tci); - + rqstats->vlan_stripped++; + } skb_mark_napi_id(skb, &enic->napi[rq->index]); if (!(netdev->features & NETIF_F_GRO)) netif_receive_skb(skb); @@ -1423,7 +1530,7 @@ static void enic_rq_indicate_buf(struct vnic_rq *rq, /* Buffer overflow */ - + rqstats->pkt_truncated++; dma_unmap_single(&enic->pdev->dev, buf->dma_addr, buf->len, DMA_FROM_DEVICE); dev_kfree_skb_any(skb); @@ -1436,7 +1543,7 @@ static int enic_rq_service(struct vnic_dev *vdev, struct cq_desc *cq_desc, { struct enic *enic = vnic_dev_priv(vdev); - vnic_rq_service(&enic->rq[q_number], cq_desc, + vnic_rq_service(&enic->rq[q_number].vrq, cq_desc, completed_index, VNIC_RQ_RETURN_DESC, enic_rq_indicate_buf, opaque); @@ -1533,7 +1640,7 @@ static int enic_poll(struct napi_struct *napi, int budget) 0 /* don't unmask intr */, 0 /* don't reset intr timer */); - err = vnic_rq_fill(&enic->rq[0], enic_rq_alloc_buf); + err = vnic_rq_fill(&enic->rq[0].vrq, enic_rq_alloc_buf); /* Buffer allocation failed. Stay in polling * mode so we can try to fill the ring again. @@ -1545,7 +1652,7 @@ static int enic_poll(struct napi_struct *napi, int budget) /* Call the function which refreshes the intr coalescing timer * value based on the traffic. */ - enic_calc_int_moderation(enic, &enic->rq[0]); + enic_calc_int_moderation(enic, &enic->rq[0].vrq); if ((rq_work_done < budget) && napi_complete_done(napi, rq_work_done)) { @@ -1554,8 +1661,11 @@ static int enic_poll(struct napi_struct *napi, int budget) */ if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) - enic_set_int_moderation(enic, &enic->rq[0]); + enic_set_int_moderation(enic, &enic->rq[0].vrq); vnic_intr_unmask(&enic->intr[intr]); + enic->rq[0].stats.napi_complete++; + } else { + enic->rq[0].stats.napi_repoll++; } return rq_work_done; @@ -1604,7 +1714,7 @@ static int enic_poll_msix_wq(struct napi_struct *napi, int budget) struct net_device *netdev = napi->dev; struct enic *enic = netdev_priv(netdev); unsigned int wq_index = (napi - &enic->napi[0]) - enic->rq_count; - struct vnic_wq *wq = &enic->wq[wq_index]; + struct vnic_wq *wq = &enic->wq[wq_index].vwq; unsigned int cq; unsigned int intr; unsigned int wq_work_to_do = ENIC_WQ_NAPI_BUDGET; @@ -1658,7 +1768,7 @@ static int enic_poll_msix_rq(struct napi_struct *napi, int budget) 0 /* don't unmask intr */, 0 /* don't reset intr timer */); - err = vnic_rq_fill(&enic->rq[rq], enic_rq_alloc_buf); + err = vnic_rq_fill(&enic->rq[rq].vrq, enic_rq_alloc_buf); /* Buffer allocation failed. Stay in polling mode * so we can try to fill the ring again. @@ -1670,7 +1780,7 @@ static int enic_poll_msix_rq(struct napi_struct *napi, int budget) /* Call the function which refreshes the intr coalescing timer * value based on the traffic. */ - enic_calc_int_moderation(enic, &enic->rq[rq]); + enic_calc_int_moderation(enic, &enic->rq[rq].vrq); if ((work_done < budget) && napi_complete_done(napi, work_done)) { @@ -1679,8 +1789,11 @@ static int enic_poll_msix_rq(struct napi_struct *napi, int budget) */ if (enic->rx_coalesce_setting.use_adaptive_rx_coalesce) - enic_set_int_moderation(enic, &enic->rq[rq]); + enic_set_int_moderation(enic, &enic->rq[rq].vrq); vnic_intr_unmask(&enic->intr[intr]); + enic->rq[rq].stats.napi_complete++; + } else { + enic->rq[rq].stats.napi_repoll++; } return work_done; @@ -1710,7 +1823,7 @@ static void enic_free_intr(struct enic *enic) free_irq(enic->pdev->irq, enic); break; case VNIC_DEV_INTR_MODE_MSIX: - for (i = 0; i < ARRAY_SIZE(enic->msix); i++) + for (i = 0; i < enic->intr_count; i++) if (enic->msix[i].requested) free_irq(enic->msix_entry[i].vector, enic->msix[i].devid); @@ -1777,7 +1890,7 @@ static int enic_request_intr(struct enic *enic) enic->msix[intr].isr = enic_isr_msix_notify; enic->msix[intr].devid = enic; - for (i = 0; i < ARRAY_SIZE(enic->msix); i++) + for (i = 0; i < enic->intr_count; i++) enic->msix[i].requested = 0; for (i = 0; i < enic->intr_count; i++) { @@ -1819,36 +1932,6 @@ static void enic_synchronize_irqs(struct enic *enic) } } -static void enic_set_rx_coal_setting(struct enic *enic) -{ - unsigned int speed; - int index = -1; - struct enic_rx_coal *rx_coal = &enic->rx_coalesce_setting; - - /* 1. Read the link speed from fw - * 2. Pick the default range for the speed - * 3. Update it in enic->rx_coalesce_setting - */ - speed = vnic_dev_port_speed(enic->vdev); - if (ENIC_LINK_SPEED_10G < speed) - index = ENIC_LINK_40G_INDEX; - else if (ENIC_LINK_SPEED_4G < speed) - index = ENIC_LINK_10G_INDEX; - else - index = ENIC_LINK_4G_INDEX; - - rx_coal->small_pkt_range_start = mod_range[index].small_pkt_range_start; - rx_coal->large_pkt_range_start = mod_range[index].large_pkt_range_start; - rx_coal->range_end = ENIC_RX_COALESCE_RANGE_END; - - /* Start with the value provided by UCSM */ - for (index = 0; index < enic->rq_count; index++) - enic->cq[index].cur_rx_coal_timeval = - enic->config.intr_timer_usec; - - rx_coal->use_adaptive_rx_coalesce = 1; -} - static int enic_dev_notify_set(struct enic *enic) { int err; @@ -1907,10 +1990,10 @@ static int enic_open(struct net_device *netdev) for (i = 0; i < enic->rq_count; i++) { /* enable rq before updating rq desc */ - vnic_rq_enable(&enic->rq[i]); - vnic_rq_fill(&enic->rq[i], enic_rq_alloc_buf); + vnic_rq_enable(&enic->rq[i].vrq); + vnic_rq_fill(&enic->rq[i].vrq, enic_rq_alloc_buf); /* Need at least one buffer on ring to get going */ - if (vnic_rq_desc_used(&enic->rq[i]) == 0) { + if (vnic_rq_desc_used(&enic->rq[i].vrq) == 0) { netdev_err(netdev, "Unable to alloc receive buffers\n"); err = -ENOMEM; goto err_out_free_rq; @@ -1918,7 +2001,7 @@ static int enic_open(struct net_device *netdev) } for (i = 0; i < enic->wq_count; i++) - vnic_wq_enable(&enic->wq[i]); + vnic_wq_enable(&enic->wq[i].vwq); if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) enic_dev_add_station_addr(enic); @@ -1945,9 +2028,9 @@ static int enic_open(struct net_device *netdev) err_out_free_rq: for (i = 0; i < enic->rq_count; i++) { - ret = vnic_rq_disable(&enic->rq[i]); + ret = vnic_rq_disable(&enic->rq[i].vrq); if (!ret) - vnic_rq_clean(&enic->rq[i], enic_free_rq_buf); + vnic_rq_clean(&enic->rq[i].vrq, enic_free_rq_buf); } enic_dev_notify_unset(enic); err_out_free_intr: @@ -1989,12 +2072,12 @@ static int enic_stop(struct net_device *netdev) enic_dev_del_station_addr(enic); for (i = 0; i < enic->wq_count; i++) { - err = vnic_wq_disable(&enic->wq[i]); + err = vnic_wq_disable(&enic->wq[i].vwq); if (err) return err; } for (i = 0; i < enic->rq_count; i++) { - err = vnic_rq_disable(&enic->rq[i]); + err = vnic_rq_disable(&enic->rq[i].vrq); if (err) return err; } @@ -2004,9 +2087,9 @@ static int enic_stop(struct net_device *netdev) enic_free_intr(enic); for (i = 0; i < enic->wq_count; i++) - vnic_wq_clean(&enic->wq[i], enic_free_wq_buf); + vnic_wq_clean(&enic->wq[i].vwq, enic_free_wq_buf); for (i = 0; i < enic->rq_count; i++) - vnic_rq_clean(&enic->rq[i], enic_free_rq_buf); + vnic_rq_clean(&enic->rq[i].vrq, enic_free_rq_buf); for (i = 0; i < enic->cq_count; i++) vnic_cq_clean(&enic->cq[i]); for (i = 0; i < enic->intr_count; i++) @@ -2027,7 +2110,7 @@ static int _enic_change_mtu(struct net_device *netdev, int new_mtu) return err; } - netdev->mtu = new_mtu; + WRITE_ONCE(netdev->mtu, new_mtu); if (running) { err = enic_open(netdev); @@ -2360,112 +2443,56 @@ static void enic_tx_hang_reset(struct work_struct *work) static int enic_set_intr_mode(struct enic *enic) { - unsigned int n = min_t(unsigned int, enic->rq_count, ENIC_RQ_MAX); - unsigned int m = min_t(unsigned int, enic->wq_count, ENIC_WQ_MAX); unsigned int i; + int num_intr; /* Set interrupt mode (INTx, MSI, MSI-X) depending * on system capabilities. * * Try MSI-X first - * - * We need n RQs, m WQs, n+m CQs, and n+m+2 INTRs - * (the second to last INTR is used for WQ/RQ errors) - * (the last INTR is used for notifications) - */ - - BUG_ON(ARRAY_SIZE(enic->msix_entry) < n + m + 2); - for (i = 0; i < n + m + 2; i++) - enic->msix_entry[i].entry = i; - - /* Use multiple RQs if RSS is enabled */ - if (ENIC_SETTING(enic, RSS) && - enic->config.intr_mode < 1 && - enic->rq_count >= n && - enic->wq_count >= m && - enic->cq_count >= n + m && - enic->intr_count >= n + m + 2) { - - if (pci_enable_msix_range(enic->pdev, enic->msix_entry, - n + m + 2, n + m + 2) > 0) { - - enic->rq_count = n; - enic->wq_count = m; - enic->cq_count = n + m; - enic->intr_count = n + m + 2; - - vnic_dev_set_intr_mode(enic->vdev, - VNIC_DEV_INTR_MODE_MSIX); - - return 0; - } - } - if (enic->config.intr_mode < 1 && - enic->rq_count >= 1 && - enic->wq_count >= m && - enic->cq_count >= 1 + m && - enic->intr_count >= 1 + m + 2) { - if (pci_enable_msix_range(enic->pdev, enic->msix_entry, - 1 + m + 2, 1 + m + 2) > 0) { - - enic->rq_count = 1; - enic->wq_count = m; - enic->cq_count = 1 + m; - enic->intr_count = 1 + m + 2; - + enic->intr_avail >= ENIC_MSIX_MIN_INTR) { + for (i = 0; i < enic->intr_avail; i++) + enic->msix_entry[i].entry = i; + + num_intr = pci_enable_msix_range(enic->pdev, enic->msix_entry, + ENIC_MSIX_MIN_INTR, + enic->intr_avail); + if (num_intr > 0) { vnic_dev_set_intr_mode(enic->vdev, - VNIC_DEV_INTR_MODE_MSIX); - + VNIC_DEV_INTR_MODE_MSIX); + enic->intr_avail = num_intr; return 0; } } /* Next try MSI * - * We need 1 RQ, 1 WQ, 2 CQs, and 1 INTR + * We need 1 INTR */ if (enic->config.intr_mode < 2 && - enic->rq_count >= 1 && - enic->wq_count >= 1 && - enic->cq_count >= 2 && - enic->intr_count >= 1 && + enic->intr_avail >= 1 && !pci_enable_msi(enic->pdev)) { - - enic->rq_count = 1; - enic->wq_count = 1; - enic->cq_count = 2; - enic->intr_count = 1; - + enic->intr_avail = 1; vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_MSI); - return 0; } /* Next try INTx * - * We need 1 RQ, 1 WQ, 2 CQs, and 3 INTRs + * We need 3 INTRs * (the first INTR is used for WQ/RQ) * (the second INTR is used for WQ/RQ errors) * (the last INTR is used for notifications) */ if (enic->config.intr_mode < 3 && - enic->rq_count >= 1 && - enic->wq_count >= 1 && - enic->cq_count >= 2 && - enic->intr_count >= 3) { - - enic->rq_count = 1; - enic->wq_count = 1; - enic->cq_count = 2; - enic->intr_count = 3; - + enic->intr_avail >= 3) { + enic->intr_avail = 3; vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_INTX); - return 0; } @@ -2490,6 +2517,124 @@ static void enic_clear_intr_mode(struct enic *enic) vnic_dev_set_intr_mode(enic->vdev, VNIC_DEV_INTR_MODE_UNKNOWN); } +static int enic_adjust_resources(struct enic *enic) +{ + unsigned int max_queues; + unsigned int rq_default; + unsigned int rq_avail; + unsigned int wq_avail; + + if (enic->rq_avail < 1 || enic->wq_avail < 1 || enic->cq_avail < 2) { + dev_err(enic_get_dev(enic), + "Not enough resources available rq: %d wq: %d cq: %d\n", + enic->rq_avail, enic->wq_avail, + enic->cq_avail); + return -ENOSPC; + } + + if (is_kdump_kernel()) { + dev_info(enic_get_dev(enic), "Running from within kdump kernel. Using minimal resources\n"); + enic->rq_avail = 1; + enic->wq_avail = 1; + enic->config.rq_desc_count = ENIC_MIN_RQ_DESCS; + enic->config.wq_desc_count = ENIC_MIN_WQ_DESCS; + enic->config.mtu = min_t(u16, 1500, enic->config.mtu); + } + + /* if RSS isn't set, then we can only use one RQ */ + if (!ENIC_SETTING(enic, RSS)) + enic->rq_avail = 1; + + switch (vnic_dev_get_intr_mode(enic->vdev)) { + case VNIC_DEV_INTR_MODE_INTX: + case VNIC_DEV_INTR_MODE_MSI: + enic->rq_count = 1; + enic->wq_count = 1; + enic->cq_count = 2; + enic->intr_count = enic->intr_avail; + break; + case VNIC_DEV_INTR_MODE_MSIX: + /* Adjust the number of wqs/rqs/cqs/interrupts that will be + * used based on which resource is the most constrained + */ + wq_avail = min(enic->wq_avail, ENIC_WQ_MAX); + rq_default = netif_get_num_default_rss_queues(); + rq_avail = min3(enic->rq_avail, ENIC_RQ_MAX, rq_default); + max_queues = min(enic->cq_avail, + enic->intr_avail - ENIC_MSIX_RESERVED_INTR); + if (wq_avail + rq_avail <= max_queues) { + enic->rq_count = rq_avail; + enic->wq_count = wq_avail; + } else { + /* recalculate wq/rq count */ + if (rq_avail < wq_avail) { + enic->rq_count = min(rq_avail, max_queues / 2); + enic->wq_count = max_queues - enic->rq_count; + } else { + enic->wq_count = min(wq_avail, max_queues / 2); + enic->rq_count = max_queues - enic->wq_count; + } + } + enic->cq_count = enic->rq_count + enic->wq_count; + enic->intr_count = enic->cq_count + ENIC_MSIX_RESERVED_INTR; + + break; + default: + dev_err(enic_get_dev(enic), "Unknown interrupt mode\n"); + return -EINVAL; + } + + return 0; +} + +static void enic_get_queue_stats_rx(struct net_device *dev, int idx, + struct netdev_queue_stats_rx *rxs) +{ + struct enic *enic = netdev_priv(dev); + struct enic_rq_stats *rqstats = &enic->rq[idx].stats; + + rxs->bytes = rqstats->bytes; + rxs->packets = rqstats->packets; + rxs->hw_drops = rqstats->bad_fcs + rqstats->pkt_truncated; + rxs->hw_drop_overruns = rqstats->pkt_truncated; + rxs->csum_unnecessary = rqstats->csum_unnecessary + + rqstats->csum_unnecessary_encap; +} + +static void enic_get_queue_stats_tx(struct net_device *dev, int idx, + struct netdev_queue_stats_tx *txs) +{ + struct enic *enic = netdev_priv(dev); + struct enic_wq_stats *wqstats = &enic->wq[idx].stats; + + txs->bytes = wqstats->bytes; + txs->packets = wqstats->packets; + txs->csum_none = wqstats->csum_none; + txs->needs_csum = wqstats->csum_partial + wqstats->encap_csum + + wqstats->tso; + txs->hw_gso_packets = wqstats->tso; + txs->stop = wqstats->stopped; + txs->wake = wqstats->wake; +} + +static void enic_get_base_stats(struct net_device *dev, + struct netdev_queue_stats_rx *rxs, + struct netdev_queue_stats_tx *txs) +{ + rxs->bytes = 0; + rxs->packets = 0; + rxs->hw_drops = 0; + rxs->hw_drop_overruns = 0; + rxs->csum_unnecessary = 0; + txs->bytes = 0; + txs->packets = 0; + txs->csum_none = 0; + txs->needs_csum = 0; + txs->hw_gso_packets = 0; + txs->stop = 0; + txs->wake = 0; +} + static const struct net_device_ops enic_netdev_dynamic_ops = { .ndo_open = enic_open, .ndo_stop = enic_stop, @@ -2538,6 +2683,77 @@ static const struct net_device_ops enic_netdev_ops = { .ndo_features_check = enic_features_check, }; +static const struct netdev_stat_ops enic_netdev_stat_ops = { + .get_queue_stats_rx = enic_get_queue_stats_rx, + .get_queue_stats_tx = enic_get_queue_stats_tx, + .get_base_stats = enic_get_base_stats, +}; + +static void enic_free_enic_resources(struct enic *enic) +{ + kfree(enic->wq); + enic->wq = NULL; + + kfree(enic->rq); + enic->rq = NULL; + + kfree(enic->cq); + enic->cq = NULL; + + kfree(enic->napi); + enic->napi = NULL; + + kfree(enic->msix_entry); + enic->msix_entry = NULL; + + kfree(enic->msix); + enic->msix = NULL; + + kfree(enic->intr); + enic->intr = NULL; +} + +static int enic_alloc_enic_resources(struct enic *enic) +{ + enic->wq = kcalloc(enic->wq_avail, sizeof(struct enic_wq), GFP_KERNEL); + if (!enic->wq) + goto free_queues; + + enic->rq = kcalloc(enic->rq_avail, sizeof(struct enic_rq), GFP_KERNEL); + if (!enic->rq) + goto free_queues; + + enic->cq = kcalloc(enic->cq_avail, sizeof(struct vnic_cq), GFP_KERNEL); + if (!enic->cq) + goto free_queues; + + enic->napi = kcalloc(enic->wq_avail + enic->rq_avail, + sizeof(struct napi_struct), GFP_KERNEL); + if (!enic->napi) + goto free_queues; + + enic->msix_entry = kcalloc(enic->intr_avail, sizeof(struct msix_entry), + GFP_KERNEL); + if (!enic->msix_entry) + goto free_queues; + + enic->msix = kcalloc(enic->intr_avail, sizeof(struct enic_msix_entry), + GFP_KERNEL); + if (!enic->msix) + goto free_queues; + + enic->intr = kcalloc(enic->intr_avail, sizeof(struct vnic_intr), + GFP_KERNEL); + if (!enic->intr) + goto free_queues; + + return 0; + +free_queues: + enic_free_enic_resources(enic); + return -ENOMEM; +} + static void enic_dev_deinit(struct enic *enic) { unsigned int i; @@ -2555,18 +2771,7 @@ static void enic_dev_deinit(struct enic *enic) enic_free_vnic_resources(enic); enic_clear_intr_mode(enic); enic_free_affinity_hint(enic); -} - -static void enic_kdump_kernel_config(struct enic *enic) -{ - if (is_kdump_kernel()) { - dev_info(enic_get_dev(enic), "Running from within kdump kernel. Using minimal resources\n"); - enic->rq_count = 1; - enic->wq_count = 1; - enic->config.rq_desc_count = ENIC_MIN_RQ_DESCS; - enic->config.wq_desc_count = ENIC_MIN_WQ_DESCS; - enic->config.mtu = min_t(u16, 1500, enic->config.mtu); - } + enic_free_enic_resources(enic); } static int enic_dev_init(struct enic *enic) @@ -2598,19 +2803,26 @@ static int enic_dev_init(struct enic *enic) enic_get_res_counts(enic); - /* modify resource count if we are in kdump_kernel - */ - enic_kdump_kernel_config(enic); + err = enic_alloc_enic_resources(enic); + if (err) { + dev_err(dev, "Failed to allocate enic resources\n"); + return err; + } - /* Set interrupt mode based on resource counts and system - * capabilities - */ + /* Set interrupt mode based on system capabilities */ err = enic_set_intr_mode(enic); if (err) { dev_err(dev, "Failed to set intr mode based on resource " "counts and system capabilities, aborting\n"); - return err; + goto err_out_free_vnic_resources; + } + + /* Adjust resource counts based on most constrained resources */ + err = enic_adjust_resources(enic); + if (err) { + dev_err(dev, "Failed to adjust resources\n"); + goto err_out_free_vnic_resources; } /* Allocate and configure vNIC resources @@ -2652,6 +2864,7 @@ err_out_free_vnic_resources: enic_free_affinity_hint(enic); enic_clear_intr_mode(enic); enic_free_vnic_resources(enic); + enic_free_enic_resources(enic); return err; } @@ -2851,13 +3064,12 @@ static int enic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) timer_setup(&enic->notify_timer, enic_notify_timer, 0); enic_rfs_flw_tbl_init(enic); - enic_set_rx_coal_setting(enic); INIT_WORK(&enic->reset, enic_reset); INIT_WORK(&enic->tx_hang_reset, enic_tx_hang_reset); INIT_WORK(&enic->change_mtu_work, enic_change_mtu_work); for (i = 0; i < enic->wq_count; i++) - spin_lock_init(&enic->wq_lock[i]); + spin_lock_init(&enic->wq[i].lock); /* Register net device */ @@ -2880,6 +3092,7 @@ static int enic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) netdev->netdev_ops = &enic_netdev_dynamic_ops; else netdev->netdev_ops = &enic_netdev_ops; + netdev->stat_ops = &enic_netdev_stat_ops; netdev->watchdog_timeo = 2 * HZ; enic_set_ethtool_ops(netdev); |