diff options
Diffstat (limited to 'drivers/net/ethernet/intel/ixgbe/ixgbe_main.c')
| -rw-r--r-- | drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2991 |
1 files changed, 1203 insertions, 1788 deletions
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 3dc6cef58107..398fc223cab9 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -55,8 +55,13 @@ char ixgbe_driver_name[] = "ixgbe"; static const char ixgbe_driver_string[] = "Intel(R) 10 Gigabit PCI Express Network Driver"; +#ifdef IXGBE_FCOE char ixgbe_default_device_descr[] = "Intel(R) 10 Gigabit Network Connection"; +#else +static char ixgbe_default_device_descr[] = + "Intel(R) 10 Gigabit Network Connection"; +#endif #define MAJ 3 #define MIN 6 #define BUILD 7 @@ -131,6 +136,11 @@ MODULE_PARM_DESC(max_vfs, "Maximum number of virtual functions to allocate per physical function"); #endif /* CONFIG_PCI_IOV */ +static unsigned int allow_unsupported_sfp; +module_param(allow_unsupported_sfp, uint, 0); +MODULE_PARM_DESC(allow_unsupported_sfp, + "Allow unsupported and untested SFP+ modules on 82599-based adapters"); + MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>"); MODULE_DESCRIPTION("Intel(R) 10 Gigabit PCI Express Network Driver"); MODULE_LICENSE("GPL"); @@ -284,7 +294,7 @@ static void ixgbe_dump(struct ixgbe_adapter *adapter) struct ixgbe_reg_info *reginfo; int n = 0; struct ixgbe_ring *tx_ring; - struct ixgbe_tx_buffer *tx_buffer_info; + struct ixgbe_tx_buffer *tx_buffer; union ixgbe_adv_tx_desc *tx_desc; struct my_u0 { u64 a; u64 b; } *u0; struct ixgbe_ring *rx_ring; @@ -324,14 +334,13 @@ static void ixgbe_dump(struct ixgbe_adapter *adapter) pr_info("Queue [NTU] [NTC] [bi(ntc)->dma ] leng ntw timestamp\n"); for (n = 0; n < adapter->num_tx_queues; n++) { tx_ring = adapter->tx_ring[n]; - tx_buffer_info = - &tx_ring->tx_buffer_info[tx_ring->next_to_clean]; + tx_buffer = &tx_ring->tx_buffer_info[tx_ring->next_to_clean]; pr_info(" %5d %5X %5X %016llX %04X %p %016llX\n", n, tx_ring->next_to_use, tx_ring->next_to_clean, - (u64)tx_buffer_info->dma, - tx_buffer_info->length, - tx_buffer_info->next_to_watch, - (u64)tx_buffer_info->time_stamp); + (u64)dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + tx_buffer->next_to_watch, + (u64)tx_buffer->time_stamp); } /* Print TX Rings */ @@ -361,18 +370,18 @@ static void ixgbe_dump(struct ixgbe_adapter *adapter) "leng ntw timestamp bi->skb\n"); for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) { - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, i); - tx_buffer_info = &tx_ring->tx_buffer_info[i]; + tx_desc = IXGBE_TX_DESC(tx_ring, i); + tx_buffer = &tx_ring->tx_buffer_info[i]; u0 = (struct my_u0 *)tx_desc; pr_info("T [0x%03X] %016llX %016llX %016llX" " %04X %p %016llX %p", i, le64_to_cpu(u0->a), le64_to_cpu(u0->b), - (u64)tx_buffer_info->dma, - tx_buffer_info->length, - tx_buffer_info->next_to_watch, - (u64)tx_buffer_info->time_stamp, - tx_buffer_info->skb); + (u64)dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + tx_buffer->next_to_watch, + (u64)tx_buffer->time_stamp, + tx_buffer->skb); if (i == tx_ring->next_to_use && i == tx_ring->next_to_clean) pr_cont(" NTC/U\n"); @@ -384,11 +393,13 @@ static void ixgbe_dump(struct ixgbe_adapter *adapter) pr_cont("\n"); if (netif_msg_pktdata(adapter) && - tx_buffer_info->dma != 0) + dma_unmap_len(tx_buffer, len) != 0) print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS, 16, 1, - phys_to_virt(tx_buffer_info->dma), - tx_buffer_info->length, true); + phys_to_virt(dma_unmap_addr(tx_buffer, + dma)), + dma_unmap_len(tx_buffer, len), + true); } } @@ -442,7 +453,7 @@ rx_ring_summary: for (i = 0; i < rx_ring->count; i++) { rx_buffer_info = &rx_ring->rx_buffer_info[i]; - rx_desc = IXGBE_RX_DESC_ADV(rx_ring, i); + rx_desc = IXGBE_RX_DESC(rx_ring, i); u0 = (struct my_u0 *)rx_desc; staterr = le32_to_cpu(rx_desc->wb.upper.status_error); if (staterr & IXGBE_RXD_STAT_DD) { @@ -464,17 +475,7 @@ rx_ring_summary: print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS, 16, 1, phys_to_virt(rx_buffer_info->dma), - rx_ring->rx_buf_len, true); - - if (rx_ring->rx_buf_len - < IXGBE_RXBUFFER_2K) - print_hex_dump(KERN_INFO, "", - DUMP_PREFIX_ADDRESS, 16, 1, - phys_to_virt( - rx_buffer_info->page_dma + - rx_buffer_info->page_offset - ), - PAGE_SIZE/2, true); + ixgbe_rx_bufsz(rx_ring), true); } } @@ -584,32 +585,26 @@ static inline void ixgbe_irq_rearm_queues(struct ixgbe_adapter *adapter, } } -static inline void ixgbe_unmap_tx_resource(struct ixgbe_ring *ring, - struct ixgbe_tx_buffer *tx_buffer) +void ixgbe_unmap_and_free_tx_resource(struct ixgbe_ring *ring, + struct ixgbe_tx_buffer *tx_buffer) { - if (tx_buffer->dma) { - if (tx_buffer->tx_flags & IXGBE_TX_FLAGS_MAPPED_AS_PAGE) - dma_unmap_page(ring->dev, - tx_buffer->dma, - tx_buffer->length, - DMA_TO_DEVICE); - else + if (tx_buffer->skb) { + dev_kfree_skb_any(tx_buffer->skb); + if (dma_unmap_len(tx_buffer, len)) dma_unmap_single(ring->dev, - tx_buffer->dma, - tx_buffer->length, - DMA_TO_DEVICE); + dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + DMA_TO_DEVICE); + } else if (dma_unmap_len(tx_buffer, len)) { + dma_unmap_page(ring->dev, + dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + DMA_TO_DEVICE); } - tx_buffer->dma = 0; -} - -void ixgbe_unmap_and_free_tx_resource(struct ixgbe_ring *tx_ring, - struct ixgbe_tx_buffer *tx_buffer_info) -{ - ixgbe_unmap_tx_resource(tx_ring, tx_buffer_info); - if (tx_buffer_info->skb) - dev_kfree_skb_any(tx_buffer_info->skb); - tx_buffer_info->skb = NULL; - /* tx_buffer_info must be completely set up in the transmit path */ + tx_buffer->next_to_watch = NULL; + tx_buffer->skb = NULL; + dma_unmap_len_set(tx_buffer, len, 0); + /* tx_buffer must be completely set up in the transmit path */ } static void ixgbe_update_xoff_received(struct ixgbe_adapter *adapter) @@ -666,7 +661,7 @@ static void ixgbe_update_xoff_received(struct ixgbe_adapter *adapter) static u64 ixgbe_get_tx_completed(struct ixgbe_ring *ring) { - return ring->tx_stats.completed; + return ring->stats.packets; } static u64 ixgbe_get_tx_pending(struct ixgbe_ring *ring) @@ -746,56 +741,88 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector, union ixgbe_adv_tx_desc *tx_desc; unsigned int total_bytes = 0, total_packets = 0; unsigned int budget = q_vector->tx.work_limit; - u16 i = tx_ring->next_to_clean; + unsigned int i = tx_ring->next_to_clean; + + if (test_bit(__IXGBE_DOWN, &adapter->state)) + return true; tx_buffer = &tx_ring->tx_buffer_info[i]; - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, i); + tx_desc = IXGBE_TX_DESC(tx_ring, i); + i -= tx_ring->count; - for (; budget; budget--) { + do { union ixgbe_adv_tx_desc *eop_desc = tx_buffer->next_to_watch; /* if next_to_watch is not set then there is no work pending */ if (!eop_desc) break; + /* prevent any other reads prior to eop_desc */ + rmb(); + /* if DD is not set pending work has not been completed */ if (!(eop_desc->wb.status & cpu_to_le32(IXGBE_TXD_STAT_DD))) break; - /* count the packet as being completed */ - tx_ring->tx_stats.completed++; - /* clear next_to_watch to prevent false hangs */ tx_buffer->next_to_watch = NULL; - /* prevent any other reads prior to eop_desc being verified */ - rmb(); + /* update the statistics for this packet */ + total_bytes += tx_buffer->bytecount; + total_packets += tx_buffer->gso_segs; - do { - ixgbe_unmap_tx_resource(tx_ring, tx_buffer); - tx_desc->wb.status = 0; - if (likely(tx_desc == eop_desc)) { - eop_desc = NULL; - dev_kfree_skb_any(tx_buffer->skb); - tx_buffer->skb = NULL; + /* free the skb */ + dev_kfree_skb_any(tx_buffer->skb); - total_bytes += tx_buffer->bytecount; - total_packets += tx_buffer->gso_segs; - } + /* unmap skb header data */ + dma_unmap_single(tx_ring->dev, + dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + DMA_TO_DEVICE); + + /* clear tx_buffer data */ + tx_buffer->skb = NULL; + dma_unmap_len_set(tx_buffer, len, 0); + /* unmap remaining buffers */ + while (tx_desc != eop_desc) { tx_buffer++; tx_desc++; i++; - if (unlikely(i == tx_ring->count)) { - i = 0; - + if (unlikely(!i)) { + i -= tx_ring->count; tx_buffer = tx_ring->tx_buffer_info; - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, 0); + tx_desc = IXGBE_TX_DESC(tx_ring, 0); } - } while (eop_desc); - } + /* unmap any remaining paged data */ + if (dma_unmap_len(tx_buffer, len)) { + dma_unmap_page(tx_ring->dev, + dma_unmap_addr(tx_buffer, dma), + dma_unmap_len(tx_buffer, len), + DMA_TO_DEVICE); + dma_unmap_len_set(tx_buffer, len, 0); + } + } + /* move us one more past the eop_desc for start of next pkt */ + tx_buffer++; + tx_desc++; + i++; + if (unlikely(!i)) { + i -= tx_ring->count; + tx_buffer = tx_ring->tx_buffer_info; + tx_desc = IXGBE_TX_DESC(tx_ring, 0); + } + + /* issue prefetch for next Tx descriptor */ + prefetch(tx_desc); + + /* update budget accounting */ + budget--; + } while (likely(budget)); + + i += tx_ring->count; tx_ring->next_to_clean = i; u64_stats_update_begin(&tx_ring->syncp); tx_ring->stats.bytes += total_bytes; @@ -807,7 +834,6 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector, if (check_for_tx_hang(tx_ring) && ixgbe_check_tx_hang(tx_ring)) { /* schedule immediate reset if we believe we hung */ struct ixgbe_hw *hw = &adapter->hw; - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, i); e_err(drv, "Detected Tx Unit Hang\n" " Tx Queue <%d>\n" " TDH, TDT <%x>, <%x>\n" @@ -835,6 +861,9 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector, return true; } + netdev_tx_completed_queue(txring_txq(tx_ring), + total_packets, total_bytes); + #define TX_WAKE_THRESHOLD (DESC_NEEDED * 2) if (unlikely(total_packets && netif_carrier_ok(tx_ring->netdev) && (ixgbe_desc_unused(tx_ring) >= TX_WAKE_THRESHOLD))) { @@ -842,9 +871,11 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector, * sees the new next_to_clean. */ smp_mb(); - if (__netif_subqueue_stopped(tx_ring->netdev, tx_ring->queue_index) && - !test_bit(__IXGBE_DOWN, &adapter->state)) { - netif_wake_subqueue(tx_ring->netdev, tx_ring->queue_index); + if (__netif_subqueue_stopped(tx_ring->netdev, + tx_ring->queue_index) + && !test_bit(__IXGBE_DOWN, &adapter->state)) { + netif_wake_subqueue(tx_ring->netdev, + tx_ring->queue_index); ++tx_ring->tx_stats.restart_queue; } } @@ -853,63 +884,68 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector, } #ifdef CONFIG_IXGBE_DCA -static void ixgbe_update_rx_dca(struct ixgbe_adapter *adapter, - struct ixgbe_ring *rx_ring, +static void ixgbe_update_tx_dca(struct ixgbe_adapter *adapter, + struct ixgbe_ring *tx_ring, int cpu) { struct ixgbe_hw *hw = &adapter->hw; - u32 rxctrl; - u8 reg_idx = rx_ring->reg_idx; + u32 txctrl = dca3_get_tag(tx_ring->dev, cpu); + u16 reg_offset; - rxctrl = IXGBE_READ_REG(hw, IXGBE_DCA_RXCTRL(reg_idx)); switch (hw->mac.type) { case ixgbe_mac_82598EB: - rxctrl &= ~IXGBE_DCA_RXCTRL_CPUID_MASK; - rxctrl |= dca3_get_tag(rx_ring->dev, cpu); + reg_offset = IXGBE_DCA_TXCTRL(tx_ring->reg_idx); break; case ixgbe_mac_82599EB: case ixgbe_mac_X540: - rxctrl &= ~IXGBE_DCA_RXCTRL_CPUID_MASK_82599; - rxctrl |= (dca3_get_tag(rx_ring->dev, cpu) << - IXGBE_DCA_RXCTRL_CPUID_SHIFT_82599); + reg_offset = IXGBE_DCA_TXCTRL_82599(tx_ring->reg_idx); + txctrl <<= IXGBE_DCA_TXCTRL_CPUID_SHIFT_82599; break; default: - break; + /* for unknown hardware do not write register */ + return; } - rxctrl |= IXGBE_DCA_RXCTRL_DESC_DCA_EN; - rxctrl |= IXGBE_DCA_RXCTRL_HEAD_DCA_EN; - rxctrl &= ~(IXGBE_DCA_RXCTRL_DESC_RRO_EN); - IXGBE_WRITE_REG(hw, IXGBE_DCA_RXCTRL(reg_idx), rxctrl); + + /* + * We can enable relaxed ordering for reads, but not writes when + * DCA is enabled. This is due to a known issue in some chipsets + * which will cause the DCA tag to be cleared. + */ + txctrl |= IXGBE_DCA_TXCTRL_DESC_RRO_EN | + IXGBE_DCA_TXCTRL_DATA_RRO_EN | + IXGBE_DCA_TXCTRL_DESC_DCA_EN; + + IXGBE_WRITE_REG(hw, reg_offset, txctrl); } -static void ixgbe_update_tx_dca(struct ixgbe_adapter *adapter, - struct ixgbe_ring *tx_ring, +static void ixgbe_update_rx_dca(struct ixgbe_adapter *adapter, + struct ixgbe_ring *rx_ring, int cpu) { struct ixgbe_hw *hw = &adapter->hw; - u32 txctrl; - u8 reg_idx = tx_ring->reg_idx; + u32 rxctrl = dca3_get_tag(rx_ring->dev, cpu); + u8 reg_idx = rx_ring->reg_idx; + switch (hw->mac.type) { - case ixgbe_mac_82598EB: - txctrl = IXGBE_READ_REG(hw, IXGBE_DCA_TXCTRL(reg_idx)); - txctrl &= ~IXGBE_DCA_TXCTRL_CPUID_MASK; - txctrl |= dca3_get_tag(tx_ring->dev, cpu); - txctrl |= IXGBE_DCA_TXCTRL_DESC_DCA_EN; - IXGBE_WRITE_REG(hw, IXGBE_DCA_TXCTRL(reg_idx), txctrl); - break; case ixgbe_mac_82599EB: case ixgbe_mac_X540: - txctrl = IXGBE_READ_REG(hw, IXGBE_DCA_TXCTRL_82599(reg_idx)); - txctrl &= ~IXGBE_DCA_TXCTRL_CPUID_MASK_82599; - txctrl |= (dca3_get_tag(tx_ring->dev, cpu) << - IXGBE_DCA_TXCTRL_CPUID_SHIFT_82599); - txctrl |= IXGBE_DCA_TXCTRL_DESC_DCA_EN; - IXGBE_WRITE_REG(hw, IXGBE_DCA_TXCTRL_82599(reg_idx), txctrl); + rxctrl <<= IXGBE_DCA_RXCTRL_CPUID_SHIFT_82599; break; default: break; } + + /* + * We can enable relaxed ordering for reads, but not writes when + * DCA is enabled. This is due to a known issue in some chipsets + * which will cause the DCA tag to be cleared. + */ + rxctrl |= IXGBE_DCA_RXCTRL_DESC_RRO_EN | + IXGBE_DCA_RXCTRL_DATA_DCA_EN | + IXGBE_DCA_RXCTRL_DESC_DCA_EN; + + IXGBE_WRITE_REG(hw, IXGBE_DCA_RXCTRL(reg_idx), rxctrl); } static void ixgbe_update_dca(struct ixgbe_q_vector *q_vector) @@ -921,10 +957,10 @@ static void ixgbe_update_dca(struct ixgbe_q_vector *q_vector) if (q_vector->cpu == cpu) goto out_no_update; - for (ring = q_vector->tx.ring; ring != NULL; ring = ring->next) + ixgbe_for_each_ring(ring, q_vector->tx) ixgbe_update_tx_dca(adapter, ring, cpu); - for (ring = q_vector->rx.ring; ring != NULL; ring = ring->next) + ixgbe_for_each_ring(ring, q_vector->rx) ixgbe_update_rx_dca(adapter, ring, cpu); q_vector->cpu = cpu; @@ -984,14 +1020,17 @@ static int __ixgbe_notify_dca(struct device *dev, void *data) return 0; } -#endif /* CONFIG_IXGBE_DCA */ -static inline void ixgbe_rx_hash(union ixgbe_adv_rx_desc *rx_desc, +#endif /* CONFIG_IXGBE_DCA */ +static inline void ixgbe_rx_hash(struct ixgbe_ring *ring, + union ixgbe_adv_rx_desc *rx_desc, struct sk_buff *skb) { - skb->rxhash = le32_to_cpu(rx_desc->wb.lower.hi_dword.rss); + if (ring->netdev->features & NETIF_F_RXHASH) + skb->rxhash = le32_to_cpu(rx_desc->wb.lower.hi_dword.rss); } +#ifdef IXGBE_FCOE /** * ixgbe_rx_is_fcoe - check the rx desc for incoming pkt type * @adapter: address of board private structure @@ -1010,73 +1049,45 @@ static inline bool ixgbe_rx_is_fcoe(struct ixgbe_adapter *adapter, IXGBE_RXDADV_PKTTYPE_ETQF_SHIFT))); } -/** - * ixgbe_receive_skb - Send a completed packet up the stack - * @adapter: board private structure - * @skb: packet to send up - * @status: hardware indication of status of receive - * @rx_ring: rx descriptor ring (for a specific queue) to setup - * @rx_desc: rx descriptor - **/ -static void ixgbe_receive_skb(struct ixgbe_q_vector *q_vector, - struct sk_buff *skb, u8 status, - struct ixgbe_ring *ring, - union ixgbe_adv_rx_desc *rx_desc) -{ - struct ixgbe_adapter *adapter = q_vector->adapter; - struct napi_struct *napi = &q_vector->napi; - bool is_vlan = (status & IXGBE_RXD_STAT_VP); - u16 tag = le16_to_cpu(rx_desc->wb.upper.vlan); - - if (is_vlan && (tag & VLAN_VID_MASK)) - __vlan_hwaccel_put_tag(skb, tag); - - if (!(adapter->flags & IXGBE_FLAG_IN_NETPOLL)) - napi_gro_receive(napi, skb); - else - netif_rx(skb); -} - +#endif /* IXGBE_FCOE */ /** * ixgbe_rx_checksum - indicate in skb if hw indicated a good cksum - * @adapter: address of board private structure - * @status_err: hardware indication of status of receive + * @ring: structure containing ring specific data + * @rx_desc: current Rx descriptor being processed * @skb: skb currently being received and modified - * @status_err: status error value of last descriptor in packet **/ -static inline void ixgbe_rx_checksum(struct ixgbe_adapter *adapter, +static inline void ixgbe_rx_checksum(struct ixgbe_ring *ring, union ixgbe_adv_rx_desc *rx_desc, - struct sk_buff *skb, - u32 status_err) + struct sk_buff *skb) { - skb->ip_summed = CHECKSUM_NONE; + skb_checksum_none_assert(skb); /* Rx csum disabled */ - if (!(adapter->flags & IXGBE_FLAG_RX_CSUM_ENABLED)) + if (!(ring->netdev->features & NETIF_F_RXCSUM)) return; /* if IP and error */ - if ((status_err & IXGBE_RXD_STAT_IPCS) && - (status_err & IXGBE_RXDADV_ERR_IPE)) { - adapter->hw_csum_rx_error++; + if (ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_IPCS) && + ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_ERR_IPE)) { + ring->rx_stats.csum_err++; return; } - if (!(status_err & IXGBE_RXD_STAT_L4CS)) + if (!ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_L4CS)) return; - if (status_err & IXGBE_RXDADV_ERR_TCPE) { - u16 pkt_info = rx_desc->wb.lower.lo_dword.hs_rss.pkt_info; + if (ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_ERR_TCPE)) { + __le16 pkt_info = rx_desc->wb.lower.lo_dword.hs_rss.pkt_info; /* * 82599 errata, UDP frames with a 0 checksum can be marked as * checksum errors. */ - if ((pkt_info & IXGBE_RXDADV_PKTTYPE_UDP) && - (adapter->hw.mac.type == ixgbe_mac_82599EB)) + if ((pkt_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_UDP)) && + test_bit(__IXGBE_RX_CSUM_UDP_ZERO_ERR, &ring->state)) return; - adapter->hw_csum_rx_error++; + ring->rx_stats.csum_err++; return; } @@ -1086,6 +1097,10 @@ static inline void ixgbe_rx_checksum(struct ixgbe_adapter *adapter, static inline void ixgbe_release_rx_desc(struct ixgbe_ring *rx_ring, u32 val) { + rx_ring->next_to_use = val; + + /* update next to alloc since we have filled the ring */ + rx_ring->next_to_alloc = val; /* * Force memory writes to complete before letting h/w * know there are new descriptors to fetch. (Only @@ -1096,8 +1111,51 @@ static inline void ixgbe_release_rx_desc(struct ixgbe_ring *rx_ring, u32 val) writel(val, rx_ring->tail); } +static bool ixgbe_alloc_mapped_page(struct ixgbe_ring *rx_ring, + struct ixgbe_rx_buffer *bi) +{ + struct page *page = bi->page; + dma_addr_t dma = bi->dma; + + /* since we are recycling buffers we should seldom need to alloc */ + if (likely(dma)) + return true; + + /* alloc new page for storage */ + if (likely(!page)) { + page = alloc_pages(GFP_ATOMIC | __GFP_COLD, + ixgbe_rx_pg_order(rx_ring)); + if (unlikely(!page)) { + rx_ring->rx_stats.alloc_rx_page_failed++; + return false; + } + bi->page = page; + } + + /* map page for use */ + dma = dma_map_page(rx_ring->dev, page, 0, + ixgbe_rx_pg_size(rx_ring), DMA_FROM_DEVICE); + + /* + * if mapping failed free memory back to system since + * there isn't much point in holding memory we can't use + */ + if (dma_mapping_error(rx_ring->dev, dma)) { + put_page(page); + bi->page = NULL; + + rx_ring->rx_stats.alloc_rx_page_failed++; + return false; + } + + bi->dma = dma; + bi->page_offset ^= ixgbe_rx_bufsz(rx_ring); + + return true; +} + /** - * ixgbe_alloc_rx_buffers - Replace used receive buffers; packet split + * ixgbe_alloc_rx_buffers - Replace used receive buffers * @rx_ring: ring to place buffers on * @cleaned_count: number of buffers to replace **/ @@ -1105,344 +1163,599 @@ void ixgbe_alloc_rx_buffers(struct ixgbe_ring *rx_ring, u16 cleaned_count) { union ixgbe_adv_rx_desc *rx_desc; struct ixgbe_rx_buffer *bi; - struct sk_buff *skb; u16 i = rx_ring->next_to_use; - /* do nothing if no valid netdev defined */ - if (!rx_ring->netdev) + /* nothing to do */ + if (!cleaned_count) return; - while (cleaned_count--) { - rx_desc = IXGBE_RX_DESC_ADV(rx_ring, i); - bi = &rx_ring->rx_buffer_info[i]; - skb = bi->skb; + rx_desc = IXGBE_RX_DESC(rx_ring, i); + bi = &rx_ring->rx_buffer_info[i]; + i -= rx_ring->count; - if (!skb) { - skb = netdev_alloc_skb_ip_align(rx_ring->netdev, - rx_ring->rx_buf_len); - if (!skb) { - rx_ring->rx_stats.alloc_rx_buff_failed++; - goto no_buffers; - } - /* initialize queue mapping */ - skb_record_rx_queue(skb, rx_ring->queue_index); - bi->skb = skb; - } + do { + if (!ixgbe_alloc_mapped_page(rx_ring, bi)) + break; - if (!bi->dma) { - bi->dma = dma_map_single(rx_ring->dev, - skb->data, - rx_ring->rx_buf_len, - DMA_FROM_DEVICE); - if (dma_mapping_error(rx_ring->dev, bi->dma)) { - rx_ring->rx_stats.alloc_rx_buff_failed++; - bi->dma = 0; - goto no_buffers; - } + /* + * Refresh the desc even if buffer_addrs didn't change + * because each write-back erases this info. + */ + rx_desc->read.pkt_addr = cpu_to_le64(bi->dma + bi->page_offset); + + rx_desc++; + bi++; + i++; + if (unlikely(!i)) { + rx_desc = IXGBE_RX_DESC(rx_ring, 0); + bi = rx_ring->rx_buffer_info; + i -= rx_ring->count; } - if (ring_is_ps_enabled(rx_ring)) { - if (!bi->page) { - bi->page = alloc_page(GFP_ATOMIC | __GFP_COLD); - if (!bi->page) { - rx_ring->rx_stats.alloc_rx_page_failed++; - goto no_buffers; - } - } + /* clear the hdr_addr for the next_to_use descriptor */ + rx_desc->read.hdr_addr = 0; - if (!bi->page_dma) { - /* use a half page if we're re-using */ - bi->page_offset ^= PAGE_SIZE / 2; - bi->page_dma = dma_map_page(rx_ring->dev, - bi->page, - bi->page_offset, - PAGE_SIZE / 2, - DMA_FROM_DEVICE); - if (dma_mapping_error(rx_ring->dev, - bi->page_dma)) { - rx_ring->rx_stats.alloc_rx_page_failed++; - bi->page_dma = 0; - goto no_buffers; - } - } + cleaned_count--; + } while (cleaned_count); - /* Refresh the desc even if buffer_addrs didn't change - * because each write-back erases this info. */ - rx_desc->read.pkt_addr = cpu_to_le64(bi->page_dma); - rx_desc->read.hdr_addr = cpu_to_le64(bi->dma); - } else { - rx_desc->read.pkt_addr = cpu_to_le64(bi->dma); - rx_desc->read.hdr_addr = 0; - } + i += rx_ring->count; - i++; - if (i == rx_ring->count) - i = 0; + if (rx_ring->next_to_use != i) + ixgbe_release_rx_desc(rx_ring, i); +} + +/** + * ixgbe_get_headlen - determine size of header for RSC/LRO/GRO/FCOE + * @data: pointer to the start of the headers + * @max_len: total length of section to find headers in + * + * This function is meant to determine the length of headers that will + * be recognized by hardware for LRO, GRO, and RSC offloads. The main + * motivation of doing this is to only perform one pull for IPv4 TCP + * packets so that we can do basic things like calculating the gso_size + * based on the average data per packet. + **/ +static unsigned int ixgbe_get_headlen(unsigned char *data, + unsigned int max_len) +{ + union { + unsigned char *network; + /* l2 headers */ + struct ethhdr *eth; + struct vlan_hdr *vlan; + /* l3 headers */ + struct iphdr *ipv4; + } hdr; + __be16 protocol; + u8 nexthdr = 0; /* default to not TCP */ + u8 hlen; + + /* this should never happen, but better safe than sorry */ + if (max_len < ETH_HLEN) + return max_len; + + /* initialize network frame pointer */ + hdr.network = data; + + /* set first protocol and move network header forward */ + protocol = hdr.eth->h_proto; + hdr.network += ETH_HLEN; + + /* handle any vlan tag if present */ + if (protocol == __constant_htons(ETH_P_8021Q)) { + if ((hdr.network - data) > (max_len - VLAN_HLEN)) + return max_len; + + protocol = hdr.vlan->h_vlan_encapsulated_proto; + hdr.network += VLAN_HLEN; } -no_buffers: - if (rx_ring->next_to_use != i) { - rx_ring->next_to_use = i; - ixgbe_release_rx_desc(rx_ring, i); + /* handle L3 protocols */ + if (protocol == __constant_htons(ETH_P_IP)) { + if ((hdr.network - data) > (max_len - sizeof(struct iphdr))) + return max_len; + + /* access ihl as a u8 to avoid unaligned access on ia64 */ + hlen = (hdr.network[0] & 0x0F) << 2; + + /* verify hlen meets minimum size requirements */ + if (hlen < sizeof(struct iphdr)) + return hdr.network - data; + + /* record next protocol */ + nexthdr = hdr.ipv4->protocol; + hdr.network += hlen; +#ifdef IXGBE_FCOE + } else if (protocol == __constant_htons(ETH_P_FCOE)) { + if ((hdr.network - data) > (max_len - FCOE_HEADER_LEN)) + return max_len; + hdr.network += FCOE_HEADER_LEN; +#endif + } else { + return hdr.network - data; } + + /* finally sort out TCP */ + if (nexthdr == IPPROTO_TCP) { + if ((hdr.network - data) > (max_len - sizeof(struct tcphdr))) + return max_len; + + /* access doff as a u8 to avoid unaligned access on ia64 */ + hlen = (hdr.network[12] & 0xF0) >> 2; + + /* verify hlen meets minimum size requirements */ + if (hlen < sizeof(struct tcphdr)) + return hdr.network - data; + + hdr.network += hlen; + } + + /* + * If everything has gone correctly hdr.network should be the + * data section of the packet and will be the end of the header. + * If not then it probably represents the end of the last recognized + * header. + */ + if ((hdr.network - data) < max_len) + return hdr.network - data; + else + return max_len; } -static inline u16 ixgbe_get_hlen(union ixgbe_adv_rx_desc *rx_desc) +static void ixgbe_get_rsc_cnt(struct ixgbe_ring *rx_ring, + union ixgbe_adv_rx_desc *rx_desc, + struct sk_buff *skb) { - /* HW will not DMA in data larger than the given buffer, even if it - * parses the (NFS, of course) header to be larger. In that case, it - * fills the header buffer and spills the rest into the page. - */ - u16 hdr_info = le16_to_cpu(rx_desc->wb.lower.lo_dword.hs_rss.hdr_info); - u16 hlen = (hdr_info & IXGBE_RXDADV_HDRBUFLEN_MASK) >> - IXGBE_RXDADV_HDRBUFLEN_SHIFT; - if (hlen > IXGBE_RX_HDR_SIZE) - hlen = IXGBE_RX_HDR_SIZE; - return hlen; + __le32 rsc_enabled; + u32 rsc_cnt; + + if (!ring_is_rsc_enabled(rx_ring)) + return; + + rsc_enabled = rx_desc->wb.lower.lo_dword.data & + cpu_to_le32(IXGBE_RXDADV_RSCCNT_MASK); + + /* If this is an RSC frame rsc_cnt should be non-zero */ + if (!rsc_enabled) + return; + + rsc_cnt = le32_to_cpu(rsc_enabled); + rsc_cnt >>= IXGBE_RXDADV_RSCCNT_SHIFT; + + IXGBE_CB(skb)->append_cnt += rsc_cnt - 1; +} + +static void ixgbe_set_rsc_gso_size(struct ixgbe_ring *ring, + struct sk_buff *skb) +{ + u16 hdr_len = skb_headlen(skb); + + /* set gso_size to avoid messing up TCP MSS */ + skb_shinfo(skb)->gso_size = DIV_ROUND_UP((skb->len - hdr_len), + IXGBE_CB(skb)->append_cnt); +} + +static void ixgbe_update_rsc_stats(struct ixgbe_ring *rx_ring, + struct sk_buff *skb) +{ + /* if append_cnt is 0 then frame is not RSC */ + if (!IXGBE_CB(skb)->append_cnt) + return; + + rx_ring->rx_stats.rsc_count += IXGBE_CB(skb)->append_cnt; + rx_ring->rx_stats.rsc_flush++; + + ixgbe_set_rsc_gso_size(rx_ring, skb); + + /* gso_size is computed using append_cnt so always clear it last */ + IXGBE_CB(skb)->append_cnt = 0; +} + +/** + * ixgbe_process_skb_fields - Populate skb header fields from Rx descriptor + * @rx_ring: rx descriptor ring packet is being transacted on + * @rx_desc: pointer to the EOP Rx descriptor + * @skb: pointer to current skb being populated + * + * This function checks the ring, descriptor, and packet information in + * order to populate the hash, checksum, VLAN, timestamp, protocol, and + * other fields within the skb. + **/ +static void ixgbe_process_skb_fields(struct ixgbe_ring *rx_ring, + union ixgbe_adv_rx_desc *rx_desc, + struct sk_buff *skb) +{ + ixgbe_update_rsc_stats(rx_ring, skb); + + ixgbe_rx_hash(rx_ring, rx_desc, skb); + + ixgbe_rx_checksum(rx_ring, rx_desc, skb); + + if (ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_VP)) { + u16 vid = le16_to_cpu(rx_desc->wb.upper.vlan); + __vlan_hwaccel_put_tag(skb, vid); + } + + skb_record_rx_queue(skb, rx_ring->queue_index); + + skb->protocol = eth_type_trans(skb, rx_ring->netdev); +} + +static void ixgbe_rx_skb(struct ixgbe_q_vector *q_vector, + struct sk_buff *skb) +{ + struct ixgbe_adapter *adapter = q_vector->adapter; + + if (!(adapter->flags & IXGBE_FLAG_IN_NETPOLL)) + napi_gro_receive(&q_vector->napi, skb); + else + netif_rx(skb); } /** - * ixgbe_transform_rsc_queue - change rsc queue into a full packet - * @skb: pointer to the last skb in the rsc queue + * ixgbe_is_non_eop - process handling of non-EOP buffers + * @rx_ring: Rx ring being processed + * @rx_desc: Rx descriptor for current buffer + * @skb: Current socket buffer containing buffer in progress * - * This function changes a queue full of hw rsc buffers into a completed - * packet. It uses the ->prev pointers to find the first packet and then - * turns it into the frag list owner. + * This function updates next to clean. If the buffer is an EOP buffer + * this function exits returning false, otherwise it will place the + * sk_buff in the next buffer to be chained and return true indicating + * that this is in fact a non-EOP buffer. **/ -static inline struct sk_buff *ixgbe_transform_rsc_queue(struct sk_buff *skb) +static bool ixgbe_is_non_eop(struct ixgbe_ring *rx_ring, + union ixgbe_adv_rx_desc *rx_desc, + struct sk_buff *skb) { - unsigned int frag_list_size = 0; - unsigned int skb_cnt = 1; + u32 ntc = rx_ring->next_to_clean + 1; + + /* fetch, update, and store next to clean */ + ntc = (ntc < rx_ring->count) ? ntc : 0; + rx_ring->next_to_clean = ntc; + + prefetch(IXGBE_RX_DESC(rx_ring, ntc)); + + if (likely(ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_EOP))) + return false; + + /* append_cnt indicates packet is RSC, if so fetch nextp */ + if (IXGBE_CB(skb)->append_cnt) { + ntc = le32_to_cpu(rx_desc->wb.upper.status_error); + ntc &= IXGBE_RXDADV_NEXTP_MASK; + ntc >>= IXGBE_RXDADV_NEXTP_SHIFT; + } + + /* place skb in next buffer to be received */ + rx_ring->rx_buffer_info[ntc].skb = skb; + rx_ring->rx_stats.non_eop_descs++; + + return true; +} + +/** + * ixgbe_cleanup_headers - Correct corrupted or empty headers + * @rx_ring: rx descriptor ring packet is being transacted on + * @rx_desc: pointer to the EOP Rx descriptor + * @skb: pointer to current skb being fixed + * + * Check for corrupted packet headers caused by senders on the local L2 + * embedded NIC switch not setting up their Tx Descriptors right. These + * should be very rare. + * + * Also address the case where we are pulling data in on pages only + * and as such no data is present in the skb header. + * + * In addition if skb is not at least 60 bytes we need to pad it so that + * it is large enough to qualify as a valid Ethernet frame. + * + * Returns true if an error was encountered and skb was freed. + **/ +static bool ixgbe_cleanup_headers(struct ixgbe_ring *rx_ring, + union ixgbe_adv_rx_desc *rx_desc, + struct sk_buff *skb) +{ + struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[0]; + struct net_device *netdev = rx_ring->netdev; + unsigned char *va; + unsigned int pull_len; + + /* if the page was released unmap it, else just sync our portion */ + if (unlikely(IXGBE_CB(skb)->page_released)) { + dma_unmap_page(rx_ring->dev, IXGBE_CB(skb)->dma, + ixgbe_rx_pg_size(rx_ring), DMA_FROM_DEVICE); + IXGBE_CB(skb)->page_released = false; + } else { + dma_sync_single_range_for_cpu(rx_ring->dev, + IXGBE_CB(skb)->dma, + frag->page_offset, + ixgbe_rx_bufsz(rx_ring), + DMA_FROM_DEVICE); + } + IXGBE_CB(skb)->dma = 0; + + /* verify that the packet does not have any known errors */ + if (unlikely(ixgbe_test_staterr(rx_desc, + IXGBE_RXDADV_ERR_FRAME_ERR_MASK) && + !(netdev->features & NETIF_F_RXALL))) { + dev_kfree_skb_any(skb); + return true; + } + + /* + * it is valid to use page_address instead of kmap since we are + * working with pages allocated out of the lomem pool per + * alloc_page(GFP_ATOMIC) + */ + va = skb_frag_address(frag); + + /* + * we need the header to contain the greater of either ETH_HLEN or + * 60 bytes if the skb->len is less than 60 for skb_pad. + */ + pull_len = skb_frag_size(frag); + if (pull_len > 256) + pull_len = ixgbe_get_headlen(va, pull_len); + + /* align pull length to size of long to optimize memcpy performance */ + skb_copy_to_linear_data(skb, va, ALIGN(pull_len, sizeof(long))); + + /* update all of the pointers */ + skb_frag_size_sub(frag, pull_len); + frag->page_offset += pull_len; + skb->data_len -= pull_len; + skb->tail += pull_len; + + /* + * if we sucked the frag empty then we should free it, + * if there are other frags here something is screwed up in hardware + */ + if (skb_frag_size(frag) == 0) { + BUG_ON(skb_shinfo(skb)->nr_frags != 1); + skb_shinfo(skb)->nr_frags = 0; + __skb_frag_unref(frag); + skb->truesize -= ixgbe_rx_bufsz(rx_ring); + } + + /* if skb_pad returns an error the skb was freed */ + if (unlikely(skb->len < 60)) { + int pad_len = 60 - skb->len; - while (skb->prev) { - struct sk_buff *prev = skb->prev; - frag_list_size += skb->len; - skb->prev = NULL; - skb = prev; - skb_cnt++; + if (skb_pad(skb, pad_len)) + return true; + __skb_put(skb, pad_len); } - skb_shinfo(skb)->frag_list = skb->next; - skb->next = NULL; - skb->len += frag_list_size; - skb->data_len += frag_list_size; - skb->truesize += frag_list_size; - IXGBE_RSC_CB(skb)->skb_cnt = skb_cnt; + return false; +} + +/** + * ixgbe_can_reuse_page - determine if we can reuse a page + * @rx_buffer: pointer to rx_buffer containing the page we want to reuse + * + * Returns true if page can be reused in another Rx buffer + **/ +static inline bool ixgbe_can_reuse_page(struct ixgbe_rx_buffer *rx_buffer) +{ + struct page *page = rx_buffer->page; + + /* if we are only owner of page and it is local we can reuse it */ + return likely(page_count(page) == 1) && + likely(page_to_nid(page) == numa_node_id()); +} + +/** + * ixgbe_reuse_rx_page - page flip buffer and store it back on the ring + * @rx_ring: rx descriptor ring to store buffers on + * @old_buff: donor buffer to have page reused + * + * Syncronizes page for reuse by the adapter + **/ +static void ixgbe_reuse_rx_page(struct ixgbe_ring *rx_ring, + struct ixgbe_rx_buffer *old_buff) +{ + struct ixgbe_rx_buffer *new_buff; + u16 nta = rx_ring->next_to_alloc; + u16 bufsz = ixgbe_rx_bufsz(rx_ring); + + new_buff = &rx_ring->rx_buffer_info[nta]; + + /* update, and store next to alloc */ + nta++; + rx_ring->next_to_alloc = (nta < rx_ring->count) ? nta : 0; + + /* transfer page from old buffer to new buffer */ + new_buff->page = old_buff->page; + new_buff->dma = old_buff->dma; - return skb; + /* flip page offset to other buffer and store to new_buff */ + new_buff->page_offset = old_buff->page_offset ^ bufsz; + + /* sync the buffer for use by the device */ + dma_sync_single_range_for_device(rx_ring->dev, new_buff->dma, + new_buff->page_offset, bufsz, + DMA_FROM_DEVICE); + + /* bump ref count on page before it is given to the stack */ + get_page(new_buff->page); } -static inline bool ixgbe_get_rsc_state(union ixgbe_adv_rx_desc *rx_desc) +/** + * ixgbe_add_rx_frag - Add contents of Rx buffer to sk_buff + * @rx_ring: rx descriptor ring to transact packets on + * @rx_buffer: buffer containing page to add + * @rx_desc: descriptor containing length of buffer written by hardware + * @skb: sk_buff to place the data into + * + * This function is based on skb_add_rx_frag. I would have used that + * function however it doesn't handle the truesize case correctly since we + * are allocating more memory than might be used for a single receive. + **/ +static void ixgbe_add_rx_frag(struct ixgbe_ring *rx_ring, + struct ixgbe_rx_buffer *rx_buffer, + struct sk_buff *skb, int size) { - return !!(le32_to_cpu(rx_desc->wb.lower.lo_dword.data) & - IXGBE_RXDADV_RSCCNT_MASK); + skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, + rx_buffer->page, rx_buffer->page_offset, + size); + skb->len += size; + skb->data_len += size; + skb->truesize += ixgbe_rx_bufsz(rx_ring); } +/** + * ixgbe_clean_rx_irq - Clean completed descriptors from Rx ring - bounce buf + * @q_vector: structure containing interrupt and ring information + * @rx_ring: rx descriptor ring to transact packets on + * @budget: Total limit on number of packets to process + * + * This function provides a "bounce buffer" approach to Rx interrupt + * processing. The advantage to this is that on systems that have + * expensive overhead for IOMMU access this provides a means of avoiding + * it by maintaining the mapping of the page to the syste. + * + * Returns true if all work is completed without reaching budget + **/ static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector, struct ixgbe_ring *rx_ring, int budget) { - struct ixgbe_adapter *adapter = q_vector->adapter; - union ixgbe_adv_rx_desc *rx_desc, *next_rxd; - struct ixgbe_rx_buffer *rx_buffer_info, *next_buffer; - struct sk_buff *skb; unsigned int total_rx_bytes = 0, total_rx_packets = 0; - const int current_node = numa_node_id(); #ifdef IXGBE_FCOE + struct ixgbe_adapter *adapter = q_vector->adapter; int ddp_bytes = 0; #endif /* IXGBE_FCOE */ - u32 staterr; - u16 i; - u16 cleaned_count = 0; - bool pkt_is_rsc = false; - - i = rx_ring->next_to_clean; - rx_desc = IXGBE_RX_DESC_ADV(rx_ring, i); - staterr = le32_to_cpu(rx_desc->wb.upper.status_error); - - while (staterr & IXGBE_RXD_STAT_DD) { - u32 upper_len = 0; - - rmb(); /* read descriptor and rx_buffer_info after status DD */ - - rx_buffer_info = &rx_ring->rx_buffer_info[i]; - - skb = rx_buffer_info->skb; - rx_buffer_info->skb = NULL; - prefetch(skb->data); - - if (ring_is_rsc_enabled(rx_ring)) - pkt_is_rsc = ixgbe_get_rsc_state(rx_desc); - - /* linear means we are building an skb from multiple pages */ - if (!skb_is_nonlinear(skb)) { - u16 hlen; - if (pkt_is_rsc && - !(staterr & IXGBE_RXD_STAT_EOP) && - !skb->prev) { - /* - * When HWRSC is enabled, delay unmapping - * of the first packet. It carries the - * header information, HW may still - * access the header after the writeback. - * Only unmap it when EOP is reached - */ - IXGBE_RSC_CB(skb)->delay_unmap = true; - IXGBE_RSC_CB(skb)->dma = rx_buffer_info->dma; - } else { - dma_unmap_single(rx_ring->dev, - rx_buffer_info->dma, - rx_ring->rx_buf_len, - DMA_FROM_DEVICE); - } - rx_buffer_info->dma = 0; + u16 cleaned_count = ixgbe_desc_unused(rx_ring); - if (ring_is_ps_enabled(rx_ring)) { - hlen = ixgbe_get_hlen(rx_desc); - upper_len = le16_to_cpu(rx_desc->wb.upper.length); - } else { - hlen = le16_to_cpu(rx_desc->wb.upper.length); - } + do { + struct ixgbe_rx_buffer *rx_buffer; + union ixgbe_adv_rx_desc *rx_desc; + struct sk_buff *skb; + struct page *page; + u16 ntc; - skb_put(skb, hlen); - } else { - /* assume packet split since header is unmapped */ - upper_len = le16_to_cpu(rx_desc->wb.upper.length); + /* return some buffers to hardware, one at a time is too slow */ + if (cleaned_count >= IXGBE_RX_BUFFER_WRITE) { + ixgbe_alloc_rx_buffers(rx_ring, cleaned_count); + cleaned_count = 0; } - if (upper_len) { - dma_unmap_page(rx_ring->dev, - rx_buffer_info->page_dma, - PAGE_SIZE / 2, - DMA_FROM_DEVICE); - rx_buffer_info->page_dma = 0; - skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, - rx_buffer_info->page, - rx_buffer_info->page_offset, - upper_len); - - if ((page_count(rx_buffer_info->page) == 1) && - (page_to_nid(rx_buffer_info->page) == current_node)) - get_page(rx_buffer_info->page); - else - rx_buffer_info->page = NULL; + ntc = rx_ring->next_to_clean; + rx_desc = IXGBE_RX_DESC(rx_ring, ntc); + rx_buffer = &rx_ring->rx_buffer_info[ntc]; - skb->len += upper_len; - skb->data_len += upper_len; - skb->truesize += PAGE_SIZE / 2; - } + if (!ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_DD)) + break; - i++; - if (i == rx_ring->count) - i = 0; + /* + * This memory barrier is needed to keep us from reading + * any other fields out of the rx_desc until we know the + * RXD_STAT_DD bit is set + */ + rmb(); - next_rxd = IXGBE_RX_DESC_ADV(rx_ring, i); - prefetch(next_rxd); - cleaned_count++; + page = rx_buffer->page; + prefetchw(page); - if (pkt_is_rsc) { - u32 nextp = (staterr & IXGBE_RXDADV_NEXTP_MASK) >> - IXGBE_RXDADV_NEXTP_SHIFT; - next_buffer = &rx_ring->rx_buffer_info[nextp]; - } else { - next_buffer = &rx_ring->rx_buffer_info[i]; - } + skb = rx_buffer->skb; - if (!(staterr & IXGBE_RXD_STAT_EOP)) { - if (ring_is_ps_enabled(rx_ring)) { - rx_buffer_info->skb = next_buffer->skb; - rx_buffer_info->dma = next_buffer->dma; - next_buffer->skb = skb; - next_buffer->dma = 0; - } else { - skb->next = next_buffer->skb; - skb->next->prev = skb; - } - rx_ring->rx_stats.non_eop_descs++; - goto next_desc; - } + if (likely(!skb)) { + void *page_addr = page_address(page) + + rx_buffer->page_offset; - if (skb->prev) { - skb = ixgbe_transform_rsc_queue(skb); - /* if we got here without RSC the packet is invalid */ - if (!pkt_is_rsc) { - __pskb_trim(skb, 0); - rx_buffer_info->skb = skb; - goto next_desc; - } - } + /* prefetch first cache line of first page */ + prefetch(page_addr); +#if L1_CACHE_BYTES < 128 + prefetch(page_addr + L1_CACHE_BYTES); +#endif - if (ring_is_rsc_enabled(rx_ring)) { - if (IXGBE_RSC_CB(skb)->delay_unmap) { - dma_unmap_single(rx_ring->dev, - IXGBE_RSC_CB(skb)->dma, - rx_ring->rx_buf_len, - DMA_FROM_DEVICE); - IXGBE_RSC_CB(skb)->dma = 0; - IXGBE_RSC_CB(skb)->delay_unmap = false; + /* allocate a skb to store the frags */ + skb = netdev_alloc_skb_ip_align(rx_ring->netdev, + IXGBE_RX_HDR_SIZE); + if (unlikely(!skb)) { + rx_ring->rx_stats.alloc_rx_buff_failed++; + break; } + + /* + * we will be copying header into skb->data in + * pskb_may_pull so it is in our interest to prefetch + * it now to avoid a possible cache miss + */ + prefetchw(skb->data); + + /* + * Delay unmapping of the first packet. It carries the + * header information, HW may still access the header + * after the writeback. Only unmap it when EOP is + * reached + */ + IXGBE_CB(skb)->dma = rx_buffer->dma; + } else { + /* we are reusing so sync this buffer for CPU use */ + dma_sync_single_range_for_cpu(rx_ring->dev, + rx_buffer->dma, + rx_buffer->page_offset, + ixgbe_rx_bufsz(rx_ring), + DMA_FROM_DEVICE); } - if (pkt_is_rsc) { - if (ring_is_ps_enabled(rx_ring)) - rx_ring->rx_stats.rsc_count += - skb_shinfo(skb)->nr_frags; - else - rx_ring->rx_stats.rsc_count += - IXGBE_RSC_CB(skb)->skb_cnt; - rx_ring->rx_stats.rsc_flush++; - } - /* ERR_MASK will only have valid bits if EOP set */ - if (unlikely(staterr & IXGBE_RXDADV_ERR_FRAME_ERR_MASK)) { - dev_kfree_skb_any(skb); - goto next_desc; + /* pull page into skb */ + ixgbe_add_rx_frag(rx_ring, rx_buffer, skb, + le16_to_cpu(rx_desc->wb.upper.length)); + + if (ixgbe_can_reuse_page(rx_buffer)) { + /* hand second half of page back to the ring */ + ixgbe_reuse_rx_page(rx_ring, rx_buffer); + } else if (IXGBE_CB(skb)->dma == rx_buffer->dma) { + /* the page has been released from the ring */ + IXGBE_CB(skb)->page_released = true; + } else { + /* we are not reusing the buffer so unmap it */ + dma_unmap_page(rx_ring->dev, rx_buffer->dma, + ixgbe_rx_pg_size(rx_ring), + DMA_FROM_DEVICE); } - ixgbe_rx_checksum(adapter, rx_desc, skb, staterr); - if (adapter->netdev->features & NETIF_F_RXHASH) - ixgbe_rx_hash(rx_desc, skb); + /* clear contents of buffer_info */ + rx_buffer->skb = NULL; + rx_buffer->dma = 0; + rx_buffer->page = NULL; + + ixgbe_get_rsc_cnt(rx_ring, rx_desc, skb); + + cleaned_count++; + + /* place incomplete frames back on ring for completion */ + if (ixgbe_is_non_eop(rx_ring, rx_desc, skb)) + continue; + + /* verify the packet layout is correct */ + if (ixgbe_cleanup_headers(rx_ring, rx_desc, skb)) + continue; /* probably a little skewed due to removing CRC */ total_rx_bytes += skb->len; total_rx_packets++; - skb->protocol = eth_type_trans(skb, rx_ring->netdev); + /* populate checksum, timestamp, VLAN, and protocol */ + ixgbe_process_skb_fields(rx_ring, rx_desc, skb); + #ifdef IXGBE_FCOE /* if ddp, not passing to ULD unless for FCP_RSP or error */ if (ixgbe_rx_is_fcoe(adapter, rx_desc)) { - ddp_bytes = ixgbe_fcoe_ddp(adapter, rx_desc, skb, - staterr); + ddp_bytes = ixgbe_fcoe_ddp(adapter, rx_desc, skb); if (!ddp_bytes) { dev_kfree_skb_any(skb); - goto next_desc; + continue; } } + #endif /* IXGBE_FCOE */ - ixgbe_receive_skb(q_vector, skb, staterr, rx_ring, rx_desc); + ixgbe_rx_skb(q_vector, skb); + /* update budget accounting */ budget--; -next_desc: - rx_desc->wb.upper.status_error = 0; - - if (!budget) - break; - - /* return some buffers to hardware, one at a time is too slow */ - if (cleaned_count >= IXGBE_RX_BUFFER_WRITE) { - ixgbe_alloc_rx_buffers(rx_ring, cleaned_count); - cleaned_count = 0; - } - - /* use prefetched values */ - rx_desc = next_rxd; - staterr = le32_to_cpu(rx_desc->wb.upper.status_error); - } - - rx_ring->next_to_clean = i; - cleaned_count = ixgbe_desc_unused(rx_ring); - - if (cleaned_count) - ixgbe_alloc_rx_buffers(rx_ring, cleaned_count); + } while (likely(budget)); #ifdef IXGBE_FCOE /* include DDPed FCoE data */ @@ -1457,8 +1770,8 @@ next_desc: total_rx_bytes += ddp_bytes; total_rx_packets += DIV_ROUND_UP(ddp_bytes, mss); } -#endif /* IXGBE_FCOE */ +#endif /* IXGBE_FCOE */ u64_stats_update_begin(&rx_ring->syncp); rx_ring->stats.packets += total_rx_packets; rx_ring->stats.bytes += total_rx_bytes; @@ -1466,6 +1779,9 @@ next_desc: q_vector->rx.total_packets += total_rx_packets; q_vector->rx.total_bytes += total_rx_bytes; + if (cleaned_count) + ixgbe_alloc_rx_buffers(rx_ring, cleaned_count); + return !!budget; } @@ -1498,10 +1814,10 @@ static void ixgbe_configure_msix(struct ixgbe_adapter *adapter) struct ixgbe_ring *ring; q_vector = adapter->q_vector[v_idx]; - for (ring = q_vector->rx.ring; ring != NULL; ring = ring->next) + ixgbe_for_each_ring(ring, q_vector->rx) ixgbe_set_ivar(adapter, 0, ring->reg_idx, v_idx); - for (ring = q_vector->tx.ring; ring != NULL; ring = ring->next) + ixgbe_for_each_ring(ring, q_vector->tx) ixgbe_set_ivar(adapter, 1, ring->reg_idx, v_idx); if (q_vector->tx.ring && !q_vector->rx.ring) { @@ -1569,20 +1885,19 @@ enum latency_range { static void ixgbe_update_itr(struct ixgbe_q_vector *q_vector, struct ixgbe_ring_container *ring_container) { - u64 bytes_perint; - struct ixgbe_adapter *adapter = q_vector->adapter; int bytes = ring_container->total_bytes; int packets = ring_container->total_packets; u32 timepassed_us; + u64 bytes_perint; u8 itr_setting = ring_container->itr; if (packets == 0) return; /* simple throttlerate management - * 0-20MB/s lowest (100000 ints/s) - * 20-100MB/s low (20000 ints/s) - * 100-1249MB/s bulk (8000 ints/s) + * 0-10MB/s lowest (100000 ints/s) + * 10-20MB/s low (20000 ints/s) + * 20-1249MB/s bulk (8000 ints/s) */ /* what was last interrupt timeslice? */ timepassed_us = q_vector->itr >> 2; @@ -1590,17 +1905,17 @@ static void ixgbe_update_itr(struct ixgbe_q_vector *q_vector, switch (itr_setting) { case lowest_latency: - if (bytes_perint > adapter->eitr_low) + if (bytes_perint > 10) itr_setting = low_latency; break; case low_latency: - if (bytes_perint > adapter->eitr_high) + if (bytes_perint > 20) itr_setting = bulk_latency; - else if (bytes_perint <= adapter->eitr_low) + else if (bytes_perint <= 10) itr_setting = lowest_latency; break; case bulk_latency: - if (bytes_perint <= adapter->eitr_high) + if (bytes_perint <= 20) itr_setting = low_latency; break; } @@ -1626,7 +1941,7 @@ void ixgbe_write_eitr(struct ixgbe_q_vector *q_vector) struct ixgbe_adapter *adapter = q_vector->adapter; struct ixgbe_hw *hw = &adapter->hw; int v_idx = q_vector->v_idx; - u32 itr_reg = q_vector->itr; + u32 itr_reg = q_vector->itr & IXGBE_MAX_EITR; switch (adapter->hw.mac.type) { case ixgbe_mac_82598EB: @@ -1678,14 +1993,14 @@ static void ixgbe_set_itr(struct ixgbe_q_vector *q_vector) ((9 * new_itr) + q_vector->itr); /* save the algorithm value here */ - q_vector->itr = new_itr & IXGBE_MAX_EITR; + q_vector->itr = new_itr; ixgbe_write_eitr(q_vector); } } /** - * ixgbe_check_overtemp_subtask - check for over tempurature + * ixgbe_check_overtemp_subtask - check for over temperature * @adapter: pointer to adapter **/ static void ixgbe_check_overtemp_subtask(struct ixgbe_adapter *adapter) @@ -1997,76 +2312,53 @@ static irqreturn_t ixgbe_msix_clean_rings(int irq, void *data) return IRQ_HANDLED; } -static inline void map_vector_to_rxq(struct ixgbe_adapter *a, int v_idx, - int r_idx) -{ - struct ixgbe_q_vector *q_vector = a->q_vector[v_idx]; - struct ixgbe_ring *rx_ring = a->rx_ring[r_idx]; - - rx_ring->q_vector = q_vector; - rx_ring->next = q_vector->rx.ring; - q_vector->rx.ring = rx_ring; - q_vector->rx.count++; -} - -static inline void map_vector_to_txq(struct ixgbe_adapter *a, int v_idx, - int t_idx) -{ - struct ixgbe_q_vector *q_vector = a->q_vector[v_idx]; - struct ixgbe_ring *tx_ring = a->tx_ring[t_idx]; - - tx_ring->q_vector = q_vector; - tx_ring->next = q_vector->tx.ring; - q_vector->tx.ring = tx_ring; - q_vector->tx.count++; - q_vector->tx.work_limit = a->tx_work_limit; -} - /** - * ixgbe_map_rings_to_vectors - Maps descriptor rings to vectors - * @adapter: board private structure to initialize + * ixgbe_poll - NAPI Rx polling callback + * @napi: structure for representing this polling device + * @budget: how many packets driver is allowed to clean * - * This function maps descriptor rings to the queue-specific vectors - * we were allotted through the MSI-X enabling code. Ideally, we'd have - * one vector per ring/queue, but on a constrained vector budget, we - * group the rings as "efficiently" as possible. You would add new - * mapping configurations in here. + * This function is used for legacy and MSI, NAPI mode **/ -static void ixgbe_map_rings_to_vectors(struct ixgbe_adapter *adapter) +int ixgbe_poll(struct napi_struct *napi, int budget) { - int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS; - int rxr_remaining = adapter->num_rx_queues, rxr_idx = 0; - int txr_remaining = adapter->num_tx_queues, txr_idx = 0; - int v_start = 0; + struct ixgbe_q_vector *q_vector = + container_of(napi, struct ixgbe_q_vector, napi); + struct ixgbe_adapter *adapter = q_vector->adapter; + struct ixgbe_ring *ring; + int per_ring_budget; + bool clean_complete = true; - /* only one q_vector if MSI-X is disabled. */ - if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED)) - q_vectors = 1; +#ifdef CONFIG_IXGBE_DCA + if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) + ixgbe_update_dca(q_vector); +#endif - /* - * If we don't have enough vectors for a 1-to-1 mapping, we'll have to - * group them so there are multiple queues per vector. - * - * Re-adjusting *qpv takes care of the remainder. - */ - for (; v_start < q_vectors && rxr_remaining; v_start++) { - int rqpv = DIV_ROUND_UP(rxr_remaining, q_vectors - v_start); - for (; rqpv; rqpv--, rxr_idx++, rxr_remaining--) - map_vector_to_rxq(adapter, v_start, rxr_idx); - } + ixgbe_for_each_ring(ring, q_vector->tx) + clean_complete &= !!ixgbe_clean_tx_irq(q_vector, ring); - /* - * If there are not enough q_vectors for each ring to have it's own - * vector then we must pair up Rx/Tx on a each vector - */ - if ((v_start + txr_remaining) > q_vectors) - v_start = 0; + /* attempt to distribute budget to each queue fairly, but don't allow + * the budget to go below 1 because we'll exit polling */ + if (q_vector->rx.count > 1) + per_ring_budget = max(budget/q_vector->rx.count, 1); + else + per_ring_budget = budget; - for (; v_start < q_vectors && txr_remaining; v_start++) { - int tqpv = DIV_ROUND_UP(txr_remaining, q_vectors - v_start); - for (; tqpv; tqpv--, txr_idx++, txr_remaining--) - map_vector_to_txq(adapter, v_start, txr_idx); - } + ixgbe_for_each_ring(ring, q_vector->rx) + clean_complete &= ixgbe_clean_rx_irq(q_vector, ring, + per_ring_budget); + + /* If all work not completed, return budget and keep polling */ + if (!clean_complete) + return budget; + + /* all work done, exit the polling mode */ + napi_complete(napi); + if (adapter->rx_itr_setting & 1) + ixgbe_set_itr(q_vector); + if (!test_bit(__IXGBE_DOWN, &adapter->state)) + ixgbe_irq_enable_queues(adapter, ((u64)1 << q_vector->v_idx)); + + return 0; } /** @@ -2112,14 +2404,14 @@ static int ixgbe_request_msix_irqs(struct ixgbe_adapter *adapter) if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) { /* assign the mask for this irq */ irq_set_affinity_hint(entry->vector, - q_vector->affinity_mask); + &q_vector->affinity_mask); } } err = request_irq(adapter->msix_entries[vector].vector, ixgbe_msix_other, 0, netdev->name, adapter); if (err) { - e_err(probe, "request_irq for msix_lsc failed: %d\n", err); + e_err(probe, "request_irq for msix_other failed: %d\n", err); goto free_queue_irqs; } @@ -2153,7 +2445,7 @@ static irqreturn_t ixgbe_intr(int irq, void *data) u32 eicr; /* - * Workaround for silicon errata on 82598. Mask the interrupts + * Workaround for silicon errata #26 on 82598. Mask the interrupt * before the read of EICR. */ IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_IRQ_CLEAR_MASK); @@ -2193,47 +2485,19 @@ static irqreturn_t ixgbe_intr(int irq, void *data) ixgbe_check_fan_failure(adapter, eicr); - if (napi_schedule_prep(&(q_vector->napi))) { - /* would disable interrupts here but EIAM disabled it */ - __napi_schedule(&(q_vector->napi)); - } + /* would disable interrupts here but EIAM disabled it */ + napi_schedule(&q_vector->napi); /* * re-enable link(maybe) and non-queue interrupts, no flush. * ixgbe_poll will re-enable the queue interrupts */ - if (!test_bit(__IXGBE_DOWN, &adapter->state)) ixgbe_irq_enable(adapter, false, false); return IRQ_HANDLED; } -static inline void ixgbe_reset_q_vectors(struct ixgbe_adapter *adapter) -{ - int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS; - int i; - - /* legacy and MSI only use one vector */ - if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED)) - q_vectors = 1; - - for (i = 0; i < adapter->num_rx_queues; i++) { - adapter->rx_ring[i]->q_vector = NULL; - adapter->rx_ring[i]->next = NULL; - } - for (i = 0; i < adapter->num_tx_queues; i++) { - adapter->tx_ring[i]->q_vector = NULL; - adapter->tx_ring[i]->next = NULL; - } - - for (i = 0; i < q_vectors; i++) { - struct ixgbe_q_vector *q_vector = adapter->q_vector[i]; - memset(&q_vector->rx, 0, sizeof(struct ixgbe_ring_container)); - memset(&q_vector->tx, 0, sizeof(struct ixgbe_ring_container)); - } -} - /** * ixgbe_request_irq - initialize interrupts * @adapter: board private structure @@ -2246,9 +2510,6 @@ static int ixgbe_request_irq(struct ixgbe_adapter *adapter) struct net_device *netdev = adapter->netdev; int err; - /* map all of the rings to the q_vectors */ - ixgbe_map_rings_to_vectors(adapter); - if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) err = ixgbe_request_msix_irqs(adapter); else if (adapter->flags & IXGBE_FLAG_MSI_ENABLED) @@ -2258,13 +2519,9 @@ static int ixgbe_request_irq(struct ixgbe_adapter *adapter) err = request_irq(adapter->pdev->irq, ixgbe_intr, IRQF_SHARED, netdev->name, adapter); - if (err) { + if (err) e_err(probe, "request_irq failed, Error %d\n", err); - /* place q_vectors and rings back into a known good state */ - ixgbe_reset_q_vectors(adapter); - } - return err; } @@ -2294,9 +2551,6 @@ static void ixgbe_free_irq(struct ixgbe_adapter *adapter) } else { free_irq(adapter->pdev->irq, adapter); } - - /* clear q_vector state information */ - ixgbe_reset_q_vectors(adapter); } /** @@ -2387,12 +2641,15 @@ void ixgbe_configure_tx_ring(struct ixgbe_adapter *adapter, * to or less than the number of on chip descriptors, which is * currently 40. */ - if (!adapter->tx_itr_setting || !adapter->rx_itr_setting) + if (!ring->q_vector || (ring->q_vector->itr < 8)) txdctl |= (1 << 16); /* WTHRESH = 1 */ else txdctl |= (8 << 16); /* WTHRESH = 8 */ - /* PTHRESH=32 is needed to avoid a Tx hang with DFP enabled. */ + /* + * Setting PTHRESH to 32 both improves performance + * and avoids a TX hang with DFP enabled + */ txdctl |= (1 << 8) | /* HTHRESH = 1 */ 32; /* PTHRESH = 32 */ @@ -2411,6 +2668,8 @@ void ixgbe_configure_tx_ring(struct ixgbe_adapter *adapter, /* enable queue */ IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(reg_idx), txdctl); + netdev_tx_reset_queue(txring_txq(ring)); + /* TXDCTL.EN will return 0 on 82598 if link is down, so skip it */ if (hw->mac.type == ixgbe_mac_82598EB && !(IXGBE_READ_REG(hw, IXGBE_LINKS) & IXGBE_LINKS_UP)) @@ -2527,18 +2786,12 @@ static void ixgbe_configure_srrctl(struct ixgbe_adapter *adapter, srrctl |= (IXGBE_RX_HDR_SIZE << IXGBE_SRRCTL_BSIZEHDRSIZE_SHIFT) & IXGBE_SRRCTL_BSIZEHDR_MASK; - if (ring_is_ps_enabled(rx_ring)) { -#if (PAGE_SIZE / 2) > IXGBE_MAX_RXBUFFER - srrctl |= IXGBE_MAX_RXBUFFER >> IXGBE_SRRCTL_BSIZEPKT_SHIFT; +#if PAGE_SIZE > IXGBE_MAX_RXBUFFER + srrctl |= IXGBE_MAX_RXBUFFER >> IXGBE_SRRCTL_BSIZEPKT_SHIFT; #else - srrctl |= (PAGE_SIZE / 2) >> IXGBE_SRRCTL_BSIZEPKT_SHIFT; + srrctl |= ixgbe_rx_bufsz(rx_ring) >> IXGBE_SRRCTL_BSIZEPKT_SHIFT; #endif - srrctl |= IXGBE_SRRCTL_DESCTYPE_HDR_SPLIT_ALWAYS; - } else { - srrctl |= ALIGN(rx_ring->rx_buf_len, 1024) >> - IXGBE_SRRCTL_BSIZEPKT_SHIFT; - srrctl |= IXGBE_SRRCTL_DESCTYPE_ADV_ONEBUF; - } + srrctl |= IXGBE_SRRCTL_DESCTYPE_ADV_ONEBUF; IXGBE_WRITE_REG(&adapter->hw, IXGBE_SRRCTL(reg_idx), srrctl); } @@ -2608,6 +2861,11 @@ static void ixgbe_setup_mrqc(struct ixgbe_adapter *adapter) | IXGBE_MRQC_RSS_FIELD_IPV6 | IXGBE_MRQC_RSS_FIELD_IPV6_TCP; + if (adapter->flags2 & IXGBE_FLAG2_RSS_FIELD_IPV4_UDP) + mrqc |= IXGBE_MRQC_RSS_FIELD_IPV4_UDP; + if (adapter->flags2 & IXGBE_FLAG2_RSS_FIELD_IPV6_UDP) + mrqc |= IXGBE_MRQC_RSS_FIELD_IPV6_UDP; + IXGBE_WRITE_REG(hw, IXGBE_MRQC, mrqc); } @@ -2621,13 +2879,11 @@ static void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter, { struct ixgbe_hw *hw = &adapter->hw; u32 rscctrl; - int rx_buf_len; u8 reg_idx = ring->reg_idx; if (!ring_is_rsc_enabled(ring)) return; - rx_buf_len = ring->rx_buf_len; rscctrl = IXGBE_READ_REG(hw, IXGBE_RSCCTL(reg_idx)); rscctrl |= IXGBE_RSCCTL_RSCEN; /* @@ -2635,24 +2891,13 @@ static void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter, * total size of max desc * buf_len is not greater * than 65536 */ - if (ring_is_ps_enabled(ring)) { -#if (PAGE_SIZE < 8192) - rscctrl |= IXGBE_RSCCTL_MAXDESC_16; -#elif (PAGE_SIZE < 16384) - rscctrl |= IXGBE_RSCCTL_MAXDESC_8; -#elif (PAGE_SIZE < 32768) - rscctrl |= IXGBE_RSCCTL_MAXDESC_4; +#if (PAGE_SIZE <= 8192) + rscctrl |= IXGBE_RSCCTL_MAXDESC_16; +#elif (PAGE_SIZE <= 16384) + rscctrl |= IXGBE_RSCCTL_MAXDESC_8; #else - rscctrl |= IXGBE_RSCCTL_MAXDESC_1; + rscctrl |= IXGBE_RSCCTL_MAXDESC_4; #endif - } else { - if (rx_buf_len <= IXGBE_RXBUFFER_4K) - rscctrl |= IXGBE_RSCCTL_MAXDESC_16; - else if (rx_buf_len <= IXGBE_RXBUFFER_8K) - rscctrl |= IXGBE_RSCCTL_MAXDESC_8; - else - rscctrl |= IXGBE_RSCCTL_MAXDESC_4; - } IXGBE_WRITE_REG(hw, IXGBE_RSCCTL(reg_idx), rscctrl); } @@ -2869,23 +3114,10 @@ static void ixgbe_set_rx_buffer_len(struct ixgbe_adapter *adapter) struct ixgbe_hw *hw = &adapter->hw; struct net_device *netdev = adapter->netdev; int max_frame = netdev->mtu + ETH_HLEN + ETH_FCS_LEN; - int rx_buf_len; struct ixgbe_ring *rx_ring; int i; u32 mhadd, hlreg0; - /* Decide whether to use packet split mode or not */ - /* On by default */ - adapter->flags |= IXGBE_FLAG_RX_PS_ENABLED; - - /* Do not use packet split if we're in SR-IOV Mode */ - if (adapter->num_vfs) - adapter->flags &= ~IXGBE_FLAG_RX_PS_ENABLED; - - /* Disable packet split due to 82599 erratum #45 */ - if (hw->mac.type == ixgbe_mac_82599EB) - adapter->flags &= ~IXGBE_FLAG_RX_PS_ENABLED; - #ifdef IXGBE_FCOE /* adjust max frame to be able to do baby jumbo for FCoE */ if ((adapter->flags & IXGBE_FLAG_FCOE_ENABLED) && @@ -2904,27 +3136,6 @@ static void ixgbe_set_rx_buffer_len(struct ixgbe_adapter *adapter) /* MHADD will allow an extra 4 bytes past for vlan tagged frames */ max_frame += VLAN_HLEN; - /* Set the RX buffer length according to the mode */ - if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) { - rx_buf_len = IXGBE_RX_HDR_SIZE; - } else { - if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) && - (netdev->mtu <= ETH_DATA_LEN)) - rx_buf_len = MAXIMUM_ETHERNET_VLAN_SIZE; - /* - * Make best use of allocation by using all but 1K of a - * power of 2 allocation that will be used for skb->head. - */ - else if (max_frame <= IXGBE_RXBUFFER_3K) - rx_buf_len = IXGBE_RXBUFFER_3K; - else if (max_frame <= IXGBE_RXBUFFER_7K) - rx_buf_len = IXGBE_RXBUFFER_7K; - else if (max_frame <= IXGBE_RXBUFFER_15K) - rx_buf_len = IXGBE_RXBUFFER_15K; - else - rx_buf_len = IXGBE_MAX_RXBUFFER; - } - hlreg0 = IXGBE_READ_REG(hw, IXGBE_HLREG0); /* set jumbo enable since MHADD.MFS is keeping size locked at max_frame */ hlreg0 |= IXGBE_HLREG0_JUMBOEN; @@ -2936,32 +3147,16 @@ static void ixgbe_set_rx_buffer_len(struct ixgbe_adapter *adapter) */ for (i = 0; i < adapter->num_rx_queues; i++) { rx_ring = adapter->rx_ring[i]; - rx_ring->rx_buf_len = rx_buf_len; - - if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) - set_ring_ps_enabled(rx_ring); - else - clear_ring_ps_enabled(rx_ring); - if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) set_ring_rsc_enabled(rx_ring); else clear_ring_rsc_enabled(rx_ring); - #ifdef IXGBE_FCOE if (netdev->features & NETIF_F_FCOE_MTU) { struct ixgbe_ring_feature *f; f = &adapter->ring_feature[RING_F_FCOE]; - if ((i >= f->mask) && (i < f->mask + f->indices)) { - clear_ring_ps_enabled(rx_ring); - if (rx_buf_len < IXGBE_FCOE_JUMBO_FRAME_SIZE) - rx_ring->rx_buf_len = - IXGBE_FCOE_JUMBO_FRAME_SIZE; - } else if (!ring_is_rsc_enabled(rx_ring) && - !ring_is_ps_enabled(rx_ring)) { - rx_ring->rx_buf_len = - IXGBE_FCOE_JUMBO_FRAME_SIZE; - } + if ((i >= f->mask) && (i < f->mask + f->indices)) + set_bit(__IXGBE_RX_FCOE_BUFSZ, &rx_ring->state); } #endif /* IXGBE_FCOE */ } @@ -3235,6 +3430,7 @@ void ixgbe_set_rx_mode(struct net_device *netdev) fctrl = IXGBE_READ_REG(hw, IXGBE_FCTRL); /* set all bits that we expect to always be set */ + fctrl &= ~IXGBE_FCTRL_SBP; /* disable store-bad-packets */ fctrl |= IXGBE_FCTRL_BAM; fctrl |= IXGBE_FCTRL_DPF; /* discard pause frames when FC enabled */ fctrl |= IXGBE_FCTRL_PMCF; @@ -3283,6 +3479,18 @@ void ixgbe_set_rx_mode(struct net_device *netdev) IXGBE_WRITE_REG(hw, IXGBE_VMOLR(adapter->num_vfs), vmolr); } + /* This is useful for sniffing bad packets. */ + if (adapter->netdev->features & NETIF_F_RXALL) { + /* UPE and MPE will be handled by normal PROMISC logic + * in e1000e_set_rx_mode */ + fctrl |= (IXGBE_FCTRL_SBP | /* Receive bad packets */ + IXGBE_FCTRL_BAM | /* RX All Bcast Pkts */ + IXGBE_FCTRL_PMCF); /* RX All MAC Ctrl Pkts */ + + fctrl &= ~(IXGBE_FCTRL_DPF); + /* NOTE: VLAN filtering is disabled by setting PROMISC */ + } + IXGBE_WRITE_REG(hw, IXGBE_FCTRL, fctrl); if (netdev->features & NETIF_F_HW_VLAN_RX) @@ -3554,6 +3762,8 @@ static void ixgbe_fdir_filter_restore(struct ixgbe_adapter *adapter) static void ixgbe_configure(struct ixgbe_adapter *adapter) { + struct ixgbe_hw *hw = &adapter->hw; + ixgbe_configure_pb(adapter); #ifdef CONFIG_IXGBE_DCB ixgbe_configure_dcb(adapter); @@ -3567,6 +3777,16 @@ static void ixgbe_configure(struct ixgbe_adapter *adapter) ixgbe_configure_fcoe(adapter); #endif /* IXGBE_FCOE */ + + switch (hw->mac.type) { + case ixgbe_mac_82599EB: + case ixgbe_mac_X540: + hw->mac.ops.disable_rx_buff(hw); + break; + default: + break; + } + if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) { ixgbe_init_fdir_signature_82599(&adapter->hw, adapter->fdir_pballoc); @@ -3576,6 +3796,15 @@ static void ixgbe_configure(struct ixgbe_adapter *adapter) ixgbe_fdir_filter_restore(adapter); } + switch (hw->mac.type) { + case ixgbe_mac_82599EB: + case ixgbe_mac_X540: + hw->mac.ops.enable_rx_buff(hw); + break; + default: + break; + } + ixgbe_configure_virtualization(adapter); ixgbe_configure_tx(adapter); @@ -3849,6 +4078,27 @@ void ixgbe_reset(struct ixgbe_adapter *adapter) } /** + * ixgbe_init_rx_page_offset - initialize page offset values for Rx buffers + * @rx_ring: ring to setup + * + * On many IA platforms the L1 cache has a critical stride of 4K, this + * results in each receive buffer starting in the same cache set. To help + * reduce the pressure on this cache set we can interleave the offsets so + * that only every other buffer will be in the same cache set. + **/ +static void ixgbe_init_rx_page_offset(struct ixgbe_ring *rx_ring) +{ + struct ixgbe_rx_buffer *rx_buffer = rx_ring->rx_buffer_info; + u16 i; + + for (i = 0; i < rx_ring->count; i += 2) { + rx_buffer[0].page_offset = 0; + rx_buffer[1].page_offset = ixgbe_rx_bufsz(rx_ring); + rx_buffer = &rx_buffer[2]; + } +} + +/** * ixgbe_clean_rx_ring - Free Rx Buffers per Queue * @rx_ring: ring to free buffers from **/ @@ -3864,50 +4114,40 @@ static void ixgbe_clean_rx_ring(struct ixgbe_ring *rx_ring) /* Free all the Rx ring sk_buffs */ for (i = 0; i < rx_ring->count; i++) { - struct ixgbe_rx_buffer *rx_buffer_info; - - rx_buffer_info = &rx_ring->rx_buffer_info[i]; - if (rx_buffer_info->dma) { - dma_unmap_single(rx_ring->dev, rx_buffer_info->dma, - rx_ring->rx_buf_len, - DMA_FROM_DEVICE); - rx_buffer_info->dma = 0; - } - if (rx_buffer_info->skb) { - struct sk_buff *skb = rx_buffer_info->skb; - rx_buffer_info->skb = NULL; - do { - struct sk_buff *this = skb; - if (IXGBE_RSC_CB(this)->delay_unmap) { - dma_unmap_single(dev, - IXGBE_RSC_CB(this)->dma, - rx_ring->rx_buf_len, - DMA_FROM_DEVICE); - IXGBE_RSC_CB(this)->dma = 0; - IXGBE_RSC_CB(skb)->delay_unmap = false; - } - skb = skb->prev; - dev_kfree_skb(this); - } while (skb); - } - if (!rx_buffer_info->page) - continue; - if (rx_buffer_info->page_dma) { - dma_unmap_page(dev, rx_buffer_info->page_dma, - PAGE_SIZE / 2, DMA_FROM_DEVICE); - rx_buffer_info->page_dma = 0; + struct ixgbe_rx_buffer *rx_buffer; + + rx_buffer = &rx_ring->rx_buffer_info[i]; + if (rx_buffer->skb) { + struct sk_buff *skb = rx_buffer->skb; + if (IXGBE_CB(skb)->page_released) { + dma_unmap_page(dev, + IXGBE_CB(skb)->dma, + ixgbe_rx_bufsz(rx_ring), + DMA_FROM_DEVICE); + IXGBE_CB(skb)->page_released = false; + } + dev_kfree_skb(skb); } - put_page(rx_buffer_info->page); - rx_buffer_info->page = NULL; - rx_buffer_info->page_offset = 0; + rx_buffer->skb = NULL; + if (rx_buffer->dma) + dma_unmap_page(dev, rx_buffer->dma, + ixgbe_rx_pg_size(rx_ring), + DMA_FROM_DEVICE); + rx_buffer->dma = 0; + if (rx_buffer->page) + put_page(rx_buffer->page); + rx_buffer->page = NULL; } size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count; memset(rx_ring->rx_buffer_info, 0, size); + ixgbe_init_rx_page_offset(rx_ring); + /* Zero out the descriptor ring */ memset(rx_ring->desc, 0, rx_ring->size); + rx_ring->next_to_alloc = 0; rx_ring->next_to_clean = 0; rx_ring->next_to_use = 0; } @@ -4073,55 +4313,6 @@ void ixgbe_down(struct ixgbe_adapter *adapter) } /** - * ixgbe_poll - NAPI Rx polling callback - * @napi: structure for representing this polling device - * @budget: how many packets driver is allowed to clean - * - * This function is used for legacy and MSI, NAPI mode - **/ -static int ixgbe_poll(struct napi_struct *napi, int budget) -{ - struct ixgbe_q_vector *q_vector = - container_of(napi, struct ixgbe_q_vector, napi); - struct ixgbe_adapter *adapter = q_vector->adapter; - struct ixgbe_ring *ring; - int per_ring_budget; - bool clean_complete = true; - -#ifdef CONFIG_IXGBE_DCA - if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) - ixgbe_update_dca(q_vector); -#endif - - for (ring = q_vector->tx.ring; ring != NULL; ring = ring->next) - clean_complete &= !!ixgbe_clean_tx_irq(q_vector, ring); - - /* attempt to distribute budget to each queue fairly, but don't allow - * the budget to go below 1 because we'll exit polling */ - if (q_vector->rx.count > 1) - per_ring_budget = max(budget/q_vector->rx.count, 1); - else - per_ring_budget = budget; - - for (ring = q_vector->rx.ring; ring != NULL; ring = ring->next) - clean_complete &= ixgbe_clean_rx_irq(q_vector, ring, - per_ring_budget); - - /* If all work not completed, return budget and keep polling */ - if (!clean_complete) - return budget; - - /* all work done, exit the polling mode */ - napi_complete(napi); - if (adapter->rx_itr_setting & 1) - ixgbe_set_itr(q_vector); - if (!test_bit(__IXGBE_DOWN, &adapter->state)) - ixgbe_irq_enable_queues(adapter, ((u64)1 << q_vector->v_idx)); - - return 0; -} - -/** * ixgbe_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ @@ -4134,802 +4325,6 @@ static void ixgbe_tx_timeout(struct net_device *netdev) } /** - * ixgbe_set_rss_queues: Allocate queues for RSS - * @adapter: board private structure to initialize - * - * This is our "base" multiqueue mode. RSS (Receive Side Scaling) will try - * to allocate one Rx queue per CPU, and if available, one Tx queue per CPU. - * - **/ -static inline bool ixgbe_set_rss_queues(struct ixgbe_adapter *adapter) -{ - bool ret = false; - struct ixgbe_ring_feature *f = &adapter->ring_feature[RING_F_RSS]; - - if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) { - f->mask = 0xF; - adapter->num_rx_queues = f->indices; - adapter->num_tx_queues = f->indices; - ret = true; - } else { - ret = false; - } - - return ret; -} - -/** - * ixgbe_set_fdir_queues: Allocate queues for Flow Director - * @adapter: board private structure to initialize - * - * Flow Director is an advanced Rx filter, attempting to get Rx flows back - * to the original CPU that initiated the Tx session. This runs in addition - * to RSS, so if a packet doesn't match an FDIR filter, we can still spread the - * Rx load across CPUs using RSS. - * - **/ -static inline bool ixgbe_set_fdir_queues(struct ixgbe_adapter *adapter) -{ - bool ret = false; - struct ixgbe_ring_feature *f_fdir = &adapter->ring_feature[RING_F_FDIR]; - - f_fdir->indices = min((int)num_online_cpus(), f_fdir->indices); - f_fdir->mask = 0; - - /* Flow Director must have RSS enabled */ - if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) && - (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE)) { - adapter->num_tx_queues = f_fdir->indices; - adapter->num_rx_queues = f_fdir->indices; - ret = true; - } else { - adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE; - } - return ret; -} - -#ifdef IXGBE_FCOE -/** - * ixgbe_set_fcoe_queues: Allocate queues for Fiber Channel over Ethernet (FCoE) - * @adapter: board private structure to initialize - * - * FCoE RX FCRETA can use up to 8 rx queues for up to 8 different exchanges. - * The ring feature mask is not used as a mask for FCoE, as it can take any 8 - * rx queues out of the max number of rx queues, instead, it is used as the - * index of the first rx queue used by FCoE. - * - **/ -static inline bool ixgbe_set_fcoe_queues(struct ixgbe_adapter *adapter) -{ - struct ixgbe_ring_feature *f = &adapter->ring_feature[RING_F_FCOE]; - - if (!(adapter->flags & IXGBE_FLAG_FCOE_ENABLED)) - return false; - - f->indices = min((int)num_online_cpus(), f->indices); - - adapter->num_rx_queues = 1; - adapter->num_tx_queues = 1; - - if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) { - e_info(probe, "FCoE enabled with RSS\n"); - if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) - ixgbe_set_fdir_queues(adapter); - else - ixgbe_set_rss_queues(adapter); - } - - /* adding FCoE rx rings to the end */ - f->mask = adapter->num_rx_queues; - adapter->num_rx_queues += f->indices; - adapter->num_tx_queues += f->indices; - - return true; -} -#endif /* IXGBE_FCOE */ - -/* Artificial max queue cap per traffic class in DCB mode */ -#define DCB_QUEUE_CAP 8 - -#ifdef CONFIG_IXGBE_DCB -static inline bool ixgbe_set_dcb_queues(struct ixgbe_adapter *adapter) -{ - int per_tc_q, q, i, offset = 0; - struct net_device *dev = adapter->netdev; - int tcs = netdev_get_num_tc(dev); - - if (!tcs) - return false; - - /* Map queue offset and counts onto allocated tx queues */ - per_tc_q = min(dev->num_tx_queues / tcs, (unsigned int)DCB_QUEUE_CAP); - q = min((int)num_online_cpus(), per_tc_q); - - for (i = 0; i < tcs; i++) { - netdev_set_tc_queue(dev, i, q, offset); - offset += q; - } - - adapter->num_tx_queues = q * tcs; - adapter->num_rx_queues = q * tcs; - -#ifdef IXGBE_FCOE - /* FCoE enabled queues require special configuration indexed - * by feature specific indices and mask. Here we map FCoE - * indices onto the DCB queue pairs allowing FCoE to own - * configuration later. - */ - if (adapter->flags & IXGBE_FLAG_FCOE_ENABLED) { - int tc; - struct ixgbe_ring_feature *f = - &adapter->ring_feature[RING_F_FCOE]; - - tc = netdev_get_prio_tc_map(dev, adapter->fcoe.up); - f->indices = dev->tc_to_txq[tc].count; - f->mask = dev->tc_to_txq[tc].offset; - } -#endif - - return true; -} -#endif - -/** - * ixgbe_set_sriov_queues: Allocate queues for IOV use - * @adapter: board private structure to initialize - * - * IOV doesn't actually use anything, so just NAK the - * request for now and let the other queue routines - * figure out what to do. - */ -static inline bool ixgbe_set_sriov_queues(struct ixgbe_adapter *adapter) -{ - return false; -} - -/* - * ixgbe_set_num_queues: Allocate queues for device, feature dependent - * @adapter: board private structure to initialize - * - * This is the top level queue allocation routine. The order here is very - * important, starting with the "most" number of features turned on at once, - * and ending with the smallest set of features. This way large combinations - * can be allocated if they're turned on, and smaller combinations are the - * fallthrough conditions. - * - **/ -static int ixgbe_set_num_queues(struct ixgbe_adapter *adapter) -{ - /* Start with base case */ - adapter->num_rx_queues = 1; - adapter->num_tx_queues = 1; - adapter->num_rx_pools = adapter->num_rx_queues; - adapter->num_rx_queues_per_pool = 1; - - if (ixgbe_set_sriov_queues(adapter)) - goto done; - -#ifdef CONFIG_IXGBE_DCB - if (ixgbe_set_dcb_queues(adapter)) - goto done; - -#endif -#ifdef IXGBE_FCOE - if (ixgbe_set_fcoe_queues(adapter)) - goto done; - -#endif /* IXGBE_FCOE */ - if (ixgbe_set_fdir_queues(adapter)) - goto done; - - if (ixgbe_set_rss_queues(adapter)) - goto done; - - /* fallback to base case */ - adapter->num_rx_queues = 1; - adapter->num_tx_queues = 1; - -done: - if ((adapter->netdev->reg_state == NETREG_UNREGISTERED) || - (adapter->netdev->reg_state == NETREG_UNREGISTERING)) - return 0; - - /* Notify the stack of the (possibly) reduced queue counts. */ - netif_set_real_num_tx_queues(adapter->netdev, adapter->num_tx_queues); - return netif_set_real_num_rx_queues(adapter->netdev, - adapter->num_rx_queues); -} - -static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter, - int vectors) -{ - int err, vector_threshold; - - /* We'll want at least 3 (vector_threshold): - * 1) TxQ[0] Cleanup - * 2) RxQ[0] Cleanup - * 3) Other (Link Status Change, etc.) - * 4) TCP Timer (optional) - */ - vector_threshold = MIN_MSIX_COUNT; - - /* The more we get, the more we will assign to Tx/Rx Cleanup - * for the separate queues...where Rx Cleanup >= Tx Cleanup. - * Right now, we simply care about how many we'll get; we'll - * set them up later while requesting irq's. - */ - while (vectors >= vector_threshold) { - err = pci_enable_msix(adapter->pdev, adapter->msix_entries, - vectors); - if (!err) /* Success in acquiring all requested vectors. */ - break; - else if (err < 0) - vectors = 0; /* Nasty failure, quit now */ - else /* err == number of vectors we should try again with */ - vectors = err; - } - - if (vectors < vector_threshold) { - /* Can't allocate enough MSI-X interrupts? Oh well. - * This just means we'll go with either a single MSI - * vector or fall back to legacy interrupts. - */ - netif_printk(adapter, hw, KERN_DEBUG, adapter->netdev, - "Unable to allocate MSI-X interrupts\n"); - adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED; - kfree(adapter->msix_entries); - adapter->msix_entries = NULL; - } else { - adapter->flags |= IXGBE_FLAG_MSIX_ENABLED; /* Woot! */ - /* - * Adjust for only the vectors we'll use, which is minimum - * of max_msix_q_vectors + NON_Q_VECTORS, or the number of - * vectors we were allocated. - */ - adapter->num_msix_vectors = min(vectors, - adapter->max_msix_q_vectors + NON_Q_VECTORS); - } -} - -/** - * ixgbe_cache_ring_rss - Descriptor ring to register mapping for RSS - * @adapter: board private structure to initialize - * - * Cache the descriptor ring offsets for RSS to the assigned rings. - * - **/ -static inline bool ixgbe_cache_ring_rss(struct ixgbe_adapter *adapter) -{ - int i; - - if (!(adapter->flags & IXGBE_FLAG_RSS_ENABLED)) - return false; - - for (i = 0; i < adapter->num_rx_queues; i++) - adapter->rx_ring[i]->reg_idx = i; - for (i = 0; i < adapter->num_tx_queues; i++) - adapter->tx_ring[i]->reg_idx = i; - - return true; -} - -#ifdef CONFIG_IXGBE_DCB - -/* ixgbe_get_first_reg_idx - Return first register index associated with ring */ -static void ixgbe_get_first_reg_idx(struct ixgbe_adapter *adapter, u8 tc, - unsigned int *tx, unsigned int *rx) -{ - struct net_device *dev = adapter->netdev; - struct ixgbe_hw *hw = &adapter->hw; - u8 num_tcs = netdev_get_num_tc(dev); - - *tx = 0; - *rx = 0; - - switch (hw->mac.type) { - case ixgbe_mac_82598EB: - *tx = tc << 2; - *rx = tc << 3; - break; - case ixgbe_mac_82599EB: - case ixgbe_mac_X540: - if (num_tcs > 4) { - if (tc < 3) { - *tx = tc << 5; - *rx = tc << 4; - } else if (tc < 5) { - *tx = ((tc + 2) << 4); - *rx = tc << 4; - } else if (tc < num_tcs) { - *tx = ((tc + 8) << 3); - *rx = tc << 4; - } - } else { - *rx = tc << 5; - switch (tc) { - case 0: - *tx = 0; - break; - case 1: - *tx = 64; - break; - case 2: - *tx = 96; - break; - case 3: - *tx = 112; - break; - default: - break; - } - } - break; - default: - break; - } -} - -/** - * ixgbe_cache_ring_dcb - Descriptor ring to register mapping for DCB - * @adapter: board private structure to initialize - * - * Cache the descriptor ring offsets for DCB to the assigned rings. - * - **/ -static inline bool ixgbe_cache_ring_dcb(struct ixgbe_adapter *adapter) -{ - struct net_device *dev = adapter->netdev; - int i, j, k; - u8 num_tcs = netdev_get_num_tc(dev); - - if (!num_tcs) - return false; - - for (i = 0, k = 0; i < num_tcs; i++) { - unsigned int tx_s, rx_s; - u16 count = dev->tc_to_txq[i].count; - - ixgbe_get_first_reg_idx(adapter, i, &tx_s, &rx_s); - for (j = 0; j < count; j++, k++) { - adapter->tx_ring[k]->reg_idx = tx_s + j; - adapter->rx_ring[k]->reg_idx = rx_s + j; - adapter->tx_ring[k]->dcb_tc = i; - adapter->rx_ring[k]->dcb_tc = i; - } - } - - return true; -} -#endif - -/** - * ixgbe_cache_ring_fdir - Descriptor ring to register mapping for Flow Director - * @adapter: board private structure to initialize - * - * Cache the descriptor ring offsets for Flow Director to the assigned rings. - * - **/ -static inline bool ixgbe_cache_ring_fdir(struct ixgbe_adapter *adapter) -{ - int i; - bool ret = false; - - if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) && - (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE)) { - for (i = 0; i < adapter->num_rx_queues; i++) - adapter->rx_ring[i]->reg_idx = i; - for (i = 0; i < adapter->num_tx_queues; i++) - adapter->tx_ring[i]->reg_idx = i; - ret = true; - } - - return ret; -} - -#ifdef IXGBE_FCOE -/** - * ixgbe_cache_ring_fcoe - Descriptor ring to register mapping for the FCoE - * @adapter: board private structure to initialize - * - * Cache the descriptor ring offsets for FCoE mode to the assigned rings. - * - */ -static inline bool ixgbe_cache_ring_fcoe(struct ixgbe_adapter *adapter) -{ - struct ixgbe_ring_feature *f = &adapter->ring_feature[RING_F_FCOE]; - int i; - u8 fcoe_rx_i = 0, fcoe_tx_i = 0; - - if (!(adapter->flags & IXGBE_FLAG_FCOE_ENABLED)) - return false; - - if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) { - if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) - ixgbe_cache_ring_fdir(adapter); - else - ixgbe_cache_ring_rss(adapter); - - fcoe_rx_i = f->mask; - fcoe_tx_i = f->mask; - } - for (i = 0; i < f->indices; i++, fcoe_rx_i++, fcoe_tx_i++) { - adapter->rx_ring[f->mask + i]->reg_idx = fcoe_rx_i; - adapter->tx_ring[f->mask + i]->reg_idx = fcoe_tx_i; - } - return true; -} - -#endif /* IXGBE_FCOE */ -/** - * ixgbe_cache_ring_sriov - Descriptor ring to register mapping for sriov - * @adapter: board private structure to initialize - * - * SR-IOV doesn't use any descriptor rings but changes the default if - * no other mapping is used. - * - */ -static inline bool ixgbe_cache_ring_sriov(struct ixgbe_adapter *adapter) -{ - adapter->rx_ring[0]->reg_idx = adapter->num_vfs * 2; - adapter->tx_ring[0]->reg_idx = adapter->num_vfs * 2; - if (adapter->num_vfs) - return true; - else - return false; -} - -/** - * ixgbe_cache_ring_register - Descriptor ring to register mapping - * @adapter: board private structure to initialize - * - * Once we know the feature-set enabled for the device, we'll cache - * the register offset the descriptor ring is assigned to. - * - * Note, the order the various feature calls is important. It must start with - * the "most" features enabled at the same time, then trickle down to the - * least amount of features turned on at once. - **/ -static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter) -{ - /* start with default case */ - adapter->rx_ring[0]->reg_idx = 0; - adapter->tx_ring[0]->reg_idx = 0; - - if (ixgbe_cache_ring_sriov(adapter)) - return; - -#ifdef CONFIG_IXGBE_DCB - if (ixgbe_cache_ring_dcb(adapter)) - return; -#endif - -#ifdef IXGBE_FCOE - if (ixgbe_cache_ring_fcoe(adapter)) - return; -#endif /* IXGBE_FCOE */ - - if (ixgbe_cache_ring_fdir(adapter)) - return; - - if (ixgbe_cache_ring_rss(adapter)) - return; -} - -/** - * ixgbe_alloc_queues - Allocate memory for all rings - * @adapter: board private structure to initialize - * - * We allocate one ring per queue at run-time since we don't know the - * number of queues at compile-time. The polling_netdev array is - * intended for Multiqueue, but should work fine with a single queue. - **/ -static int ixgbe_alloc_queues(struct ixgbe_adapter *adapter) -{ - int rx = 0, tx = 0, nid = adapter->node; - - if (nid < 0 || !node_online(nid)) - nid = first_online_node; - - for (; tx < adapter->num_tx_queues; tx++) { - struct ixgbe_ring *ring; - - ring = kzalloc_node(sizeof(*ring), GFP_KERNEL, nid); - if (!ring) - ring = kzalloc(sizeof(*ring), GFP_KERNEL); - if (!ring) - goto err_allocation; - ring->count = adapter->tx_ring_count; - ring->queue_index = tx; - ring->numa_node = nid; - ring->dev = &adapter->pdev->dev; - ring->netdev = adapter->netdev; - - adapter->tx_ring[tx] = ring; - } - - for (; rx < adapter->num_rx_queues; rx++) { - struct ixgbe_ring *ring; - - ring = kzalloc_node(sizeof(*ring), GFP_KERNEL, nid); - if (!ring) - ring = kzalloc(sizeof(*ring), GFP_KERNEL); - if (!ring) - goto err_allocation; - ring->count = adapter->rx_ring_count; - ring->queue_index = rx; - ring->numa_node = nid; - ring->dev = &adapter->pdev->dev; - ring->netdev = adapter->netdev; - - adapter->rx_ring[rx] = ring; - } - - ixgbe_cache_ring_register(adapter); - - return 0; - -err_allocation: - while (tx) - kfree(adapter->tx_ring[--tx]); - - while (rx) - kfree(adapter->rx_ring[--rx]); - return -ENOMEM; -} - -/** - * ixgbe_set_interrupt_capability - set MSI-X or MSI if supported - * @adapter: board private structure to initialize - * - * Attempt to configure the interrupts using the best available - * capabilities of the hardware and the kernel. - **/ -static int ixgbe_set_interrupt_capability(struct ixgbe_adapter *adapter) -{ - struct ixgbe_hw *hw = &adapter->hw; - int err = 0; - int vector, v_budget; - - /* - * It's easy to be greedy for MSI-X vectors, but it really - * doesn't do us much good if we have a lot more vectors - * than CPU's. So let's be conservative and only ask for - * (roughly) the same number of vectors as there are CPU's. - */ - v_budget = min(adapter->num_rx_queues + adapter->num_tx_queues, - (int)num_online_cpus()) + NON_Q_VECTORS; - - /* - * At the same time, hardware can only support a maximum of - * hw.mac->max_msix_vectors vectors. With features - * such as RSS and VMDq, we can easily surpass the number of Rx and Tx - * descriptor queues supported by our device. Thus, we cap it off in - * those rare cases where the cpu count also exceeds our vector limit. - */ - v_budget = min(v_budget, (int)hw->mac.max_msix_vectors); - - /* A failure in MSI-X entry allocation isn't fatal, but it does - * mean we disable MSI-X capabilities of the adapter. */ - adapter->msix_entries = kcalloc(v_budget, - sizeof(struct msix_entry), GFP_KERNEL); - if (adapter->msix_entries) { - for (vector = 0; vector < v_budget; vector++) - adapter->msix_entries[vector].entry = vector; - - ixgbe_acquire_msix_vectors(adapter, v_budget); - - if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) - goto out; - } - - adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED; - adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED; - if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) { - e_err(probe, - "ATR is not supported while multiple " - "queues are disabled. Disabling Flow Director\n"); - } - adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE; - adapter->atr_sample_rate = 0; - if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) - ixgbe_disable_sriov(adapter); - - err = ixgbe_set_num_queues(adapter); - if (err) - return err; - - err = pci_enable_msi(adapter->pdev); - if (!err) { - adapter->flags |= IXGBE_FLAG_MSI_ENABLED; - } else { - netif_printk(adapter, hw, KERN_DEBUG, adapter->netdev, - "Unable to allocate MSI interrupt, " - "falling back to legacy. Error: %d\n", err); - /* reset err */ - err = 0; - } - -out: - return err; -} - -/** - * ixgbe_alloc_q_vectors - Allocate memory for interrupt vectors - * @adapter: board private structure to initialize - * - * We allocate one q_vector per queue interrupt. If allocation fails we - * return -ENOMEM. - **/ -static int ixgbe_alloc_q_vectors(struct ixgbe_adapter *adapter) -{ - int v_idx, num_q_vectors; - struct ixgbe_q_vector *q_vector; - - if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) - num_q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS; - else - num_q_vectors = 1; - - for (v_idx = 0; v_idx < num_q_vectors; v_idx++) { - q_vector = kzalloc_node(sizeof(struct ixgbe_q_vector), - GFP_KERNEL, adapter->node); - if (!q_vector) - q_vector = kzalloc(sizeof(struct ixgbe_q_vector), - GFP_KERNEL); - if (!q_vector) - goto err_out; - - q_vector->adapter = adapter; - q_vector->v_idx = v_idx; - - /* Allocate the affinity_hint cpumask, configure the mask */ - if (!alloc_cpumask_var(&q_vector->affinity_mask, GFP_KERNEL)) - goto err_out; - cpumask_set_cpu(v_idx, q_vector->affinity_mask); - netif_napi_add(adapter->netdev, &q_vector->napi, - ixgbe_poll, 64); - adapter->q_vector[v_idx] = q_vector; - } - - return 0; - -err_out: - while (v_idx) { - v_idx--; - q_vector = adapter->q_vector[v_idx]; - netif_napi_del(&q_vector->napi); - free_cpumask_var(q_vector->affinity_mask); - kfree(q_vector); - adapter->q_vector[v_idx] = NULL; - } - return -ENOMEM; -} - -/** - * ixgbe_free_q_vectors - Free memory allocated for interrupt vectors - * @adapter: board private structure to initialize - * - * This function frees the memory allocated to the q_vectors. In addition if - * NAPI is enabled it will delete any references to the NAPI struct prior - * to freeing the q_vector. - **/ -static void ixgbe_free_q_vectors(struct ixgbe_adapter *adapter) -{ - int v_idx, num_q_vectors; - - if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) - num_q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS; - else - num_q_vectors = 1; - - for (v_idx = 0; v_idx < num_q_vectors; v_idx++) { - struct ixgbe_q_vector *q_vector = adapter->q_vector[v_idx]; - adapter->q_vector[v_idx] = NULL; - netif_napi_del(&q_vector->napi); - free_cpumask_var(q_vector->affinity_mask); - kfree(q_vector); - } -} - -static void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter) -{ - if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) { - adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED; - pci_disable_msix(adapter->pdev); - kfree(adapter->msix_entries); - adapter->msix_entries = NULL; - } else if (adapter->flags & IXGBE_FLAG_MSI_ENABLED) { - adapter->flags &= ~IXGBE_FLAG_MSI_ENABLED; - pci_disable_msi(adapter->pdev); - } -} - -/** - * ixgbe_init_interrupt_scheme - Determine proper interrupt scheme - * @adapter: board private structure to initialize - * - * We determine which interrupt scheme to use based on... - * - Kernel support (MSI, MSI-X) - * - which can be user-defined (via MODULE_PARAM) - * - Hardware queue count (num_*_queues) - * - defined by miscellaneous hardware support/features (RSS, etc.) - **/ -int ixgbe_init_interrupt_scheme(struct ixgbe_adapter *adapter) -{ - int err; - - /* Number of supported queues */ - err = ixgbe_set_num_queues(adapter); - if (err) - return err; - - err = ixgbe_set_interrupt_capability(adapter); - if (err) { - e_dev_err("Unable to setup interrupt capabilities\n"); - goto err_set_interrupt; - } - - err = ixgbe_alloc_q_vectors(adapter); - if (err) { - e_dev_err("Unable to allocate memory for queue vectors\n"); - goto err_alloc_q_vectors; - } - - err = ixgbe_alloc_queues(adapter); - if (err) { - e_dev_err("Unable to allocate memory for queues\n"); - goto err_alloc_queues; - } - - e_dev_info("Multiqueue %s: Rx Queue count = %u, Tx Queue count = %u\n", - (adapter->num_rx_queues > 1) ? "Enabled" : "Disabled", - adapter->num_rx_queues, adapter->num_tx_queues); - - set_bit(__IXGBE_DOWN, &adapter->state); - - return 0; - -err_alloc_queues: - ixgbe_free_q_vectors(adapter); -err_alloc_q_vectors: - ixgbe_reset_interrupt_capability(adapter); -err_set_interrupt: - return err; -} - -/** - * ixgbe_clear_interrupt_scheme - Clear the current interrupt scheme settings - * @adapter: board private structure to clear interrupt scheme on - * - * We go through and clear interrupt specific resources and reset the structure - * to pre-load conditions - **/ -void ixgbe_clear_interrupt_scheme(struct ixgbe_adapter *adapter) -{ - int i; - - for (i = 0; i < adapter->num_tx_queues; i++) { - kfree(adapter->tx_ring[i]); - adapter->tx_ring[i] = NULL; - } - for (i = 0; i < adapter->num_rx_queues; i++) { - struct ixgbe_ring *ring = adapter->rx_ring[i]; - - /* ixgbe_get_stats64() might access this ring, we must wait - * a grace period before freeing it. - */ - kfree_rcu(ring, rcu); - adapter->rx_ring[i] = NULL; - } - - adapter->num_tx_queues = 0; - adapter->num_rx_queues = 0; - - ixgbe_free_q_vectors(adapter); - ixgbe_reset_interrupt_capability(adapter); -} - -/** * ixgbe_sw_init - Initialize general software structures (struct ixgbe_adapter) * @adapter: board private structure to initialize * @@ -4956,7 +4351,7 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter) hw->subsystem_device_id = pdev->subsystem_device; /* Set capability flags */ - rss = min(IXGBE_MAX_RSS_INDICES, (int)num_online_cpus()); + rss = min_t(int, IXGBE_MAX_RSS_INDICES, num_online_cpus()); adapter->ring_feature[RING_F_RSS].indices = rss; adapter->flags |= IXGBE_FLAG_RSS_ENABLED; switch (hw->mac.type) { @@ -5048,10 +4443,6 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter) adapter->rx_itr_setting = 1; adapter->tx_itr_setting = 1; - /* set defaults for eitr in MegaBytes */ - adapter->eitr_low = 10; - adapter->eitr_high = 20; - /* set default ring sizes */ adapter->tx_ring_count = IXGBE_DEFAULT_TXD; adapter->rx_ring_count = IXGBE_DEFAULT_RXD; @@ -5065,12 +4456,6 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter) return -EIO; } - /* enable rx csum by default */ - adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED; - - /* get assigned NUMA node */ - adapter->node = dev_to_node(&pdev->dev); - set_bit(__IXGBE_DOWN, &adapter->state); return 0; @@ -5085,10 +4470,16 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter) int ixgbe_setup_tx_resources(struct ixgbe_ring *tx_ring) { struct device *dev = tx_ring->dev; + int orig_node = dev_to_node(dev); + int numa_node = -1; int size; size = sizeof(struct ixgbe_tx_buffer) * tx_ring->count; - tx_ring->tx_buffer_info = vzalloc_node(size, tx_ring->numa_node); + + if (tx_ring->q_vector) + numa_node = tx_ring->q_vector->numa_node; + + tx_ring->tx_buffer_info = vzalloc_node(size, numa_node); if (!tx_ring->tx_buffer_info) tx_ring->tx_buffer_info = vzalloc(size); if (!tx_ring->tx_buffer_info) @@ -5098,8 +4489,15 @@ int ixgbe_setup_tx_resources(struct ixgbe_ring *tx_ring) tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc); tx_ring->size = ALIGN(tx_ring->size, 4096); - tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size, - &tx_ring->dma, GFP_KERNEL); + set_dev_node(dev, numa_node); + tx_ring->desc = dma_alloc_coherent(dev, + tx_ring->size, + &tx_ring->dma, + GFP_KERNEL); + set_dev_node(dev, orig_node); + if (!tx_ring->desc) + tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size, + &tx_ring->dma, GFP_KERNEL); if (!tx_ring->desc) goto err; @@ -5148,10 +4546,16 @@ static int ixgbe_setup_all_tx_resources(struct ixgbe_adapter *adapter) int ixgbe_setup_rx_resources(struct ixgbe_ring *rx_ring) { struct device *dev = rx_ring->dev; + int orig_node = dev_to_node(dev); + int numa_node = -1; int size; size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count; - rx_ring->rx_buffer_info = vzalloc_node(size, rx_ring->numa_node); + + if (rx_ring->q_vector) + numa_node = rx_ring->q_vector->numa_node; + + rx_ring->rx_buffer_info = vzalloc_node(size, numa_node); if (!rx_ring->rx_buffer_info) rx_ring->rx_buffer_info = vzalloc(size); if (!rx_ring->rx_buffer_info) @@ -5161,15 +4565,23 @@ int ixgbe_setup_rx_resources(struct ixgbe_ring *rx_ring) rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc); rx_ring->size = ALIGN(rx_ring->size, 4096); - rx_ring->desc = dma_alloc_coherent(dev, rx_ring->size, - &rx_ring->dma, GFP_KERNEL); - + set_dev_node(dev, numa_node); + rx_ring->desc = dma_alloc_coherent(dev, + rx_ring->size, + &rx_ring->dma, + GFP_KERNEL); + set_dev_node(dev, orig_node); + if (!rx_ring->desc) + rx_ring->desc = dma_alloc_coherent(dev, rx_ring->size, + &rx_ring->dma, GFP_KERNEL); if (!rx_ring->desc) goto err; rx_ring->next_to_clean = 0; rx_ring->next_to_use = 0; + ixgbe_init_rx_page_offset(rx_ring); + return 0; err: vfree(rx_ring->rx_buffer_info); @@ -5289,20 +4701,24 @@ static void ixgbe_free_all_rx_resources(struct ixgbe_adapter *adapter) static int ixgbe_change_mtu(struct net_device *netdev, int new_mtu) { struct ixgbe_adapter *adapter = netdev_priv(netdev); - struct ixgbe_hw *hw = &adapter->hw; int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN; /* MTU < 68 is an error and causes problems on some kernels */ - if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED && - hw->mac.type != ixgbe_mac_X540) { - if ((new_mtu < 68) || (max_frame > MAXIMUM_ETHERNET_VLAN_SIZE)) - return -EINVAL; - } else { - if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE)) + if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE)) + return -EINVAL; + + /* + * For 82599EB we cannot allow PF to change MTU greater than 1500 + * in SR-IOV mode as it may cause buffer overruns in guest VFs that + * don't allocate and chain buffers correctly. + */ + if ((adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) && + (adapter->hw.mac.type == ixgbe_mac_82599EB) && + (max_frame > MAXIMUM_ETHERNET_VLAN_SIZE)) return -EINVAL; - } e_info(probe, "changing MTU from %d to %d\n", netdev->mtu, new_mtu); + /* must set new MTU before calling down or up */ netdev->mtu = new_mtu; @@ -5558,7 +4974,7 @@ void ixgbe_update_stats(struct ixgbe_adapter *adapter) u32 i, missed_rx = 0, mpc, bprc, lxon, lxoff, xon_off_tot; u64 non_eop_descs = 0, restart_queue = 0, tx_busy = 0; u64 alloc_rx_page_failed = 0, alloc_rx_buff_failed = 0; - u64 bytes = 0, packets = 0; + u64 bytes = 0, packets = 0, hw_csum_rx_error = 0; #ifdef IXGBE_FCOE struct ixgbe_fcoe *fcoe = &adapter->fcoe; unsigned int cpu; @@ -5588,12 +5004,14 @@ void ixgbe_update_stats(struct ixgbe_adapter *adapter) non_eop_descs += rx_ring->rx_stats.non_eop_descs; alloc_rx_page_failed += rx_ring->rx_stats.alloc_rx_page_failed; alloc_rx_buff_failed += rx_ring->rx_stats.alloc_rx_buff_failed; + hw_csum_rx_error += rx_ring->rx_stats.csum_err; bytes += rx_ring->stats.bytes; packets += rx_ring->stats.packets; } adapter->non_eop_descs = non_eop_descs; adapter->alloc_rx_page_failed = alloc_rx_page_failed; adapter->alloc_rx_buff_failed = alloc_rx_buff_failed; + adapter->hw_csum_rx_error = hw_csum_rx_error; netdev->stats.rx_bytes = bytes; netdev->stats.rx_packets = packets; @@ -5945,7 +5363,7 @@ static void ixgbe_watchdog_link_is_up(struct ixgbe_adapter *adapter) * print link down message * @adapter - pointer to the adapter structure **/ -static void ixgbe_watchdog_link_is_down(struct ixgbe_adapter* adapter) +static void ixgbe_watchdog_link_is_down(struct ixgbe_adapter *adapter) { struct net_device *netdev = adapter->netdev; struct ixgbe_hw *hw = &adapter->hw; @@ -6190,41 +5608,32 @@ static void ixgbe_service_timer(unsigned long data) unsigned long next_event_offset; bool ready = true; -#ifdef CONFIG_PCI_IOV - ready = false; + /* poll faster when waiting for link */ + if (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE) + next_event_offset = HZ / 10; + else + next_event_offset = HZ * 2; +#ifdef CONFIG_PCI_IOV /* * don't bother with SR-IOV VF DMA hang check if there are * no VFs or the link is down */ if (!adapter->num_vfs || - (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE)) { - ready = true; + (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE)) goto normal_timer_service; - } /* If we have VFs allocated then we must check for DMA hangs */ ixgbe_check_for_bad_vf(adapter); next_event_offset = HZ / 50; adapter->timer_event_accumulator++; - if (adapter->timer_event_accumulator >= 100) { - ready = true; + if (adapter->timer_event_accumulator >= 100) adapter->timer_event_accumulator = 0; - } - - goto schedule_event; - -normal_timer_service: -#endif - /* poll faster when waiting for link */ - if (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE) - next_event_offset = HZ / 10; else - next_event_offset = HZ * 2; + ready = false; -#ifdef CONFIG_PCI_IOV -schedule_event: +normal_timer_service: #endif /* Reset the timer */ mod_timer(&adapter->service_timer, next_event_offset + jiffies); @@ -6273,30 +5682,11 @@ static void ixgbe_service_task(struct work_struct *work) ixgbe_service_event_complete(adapter); } -void ixgbe_tx_ctxtdesc(struct ixgbe_ring *tx_ring, u32 vlan_macip_lens, - u32 fcoe_sof_eof, u32 type_tucmd, u32 mss_l4len_idx) -{ - struct ixgbe_adv_tx_context_desc *context_desc; - u16 i = tx_ring->next_to_use; - - context_desc = IXGBE_TX_CTXTDESC_ADV(tx_ring, i); - - i++; - tx_ring->next_to_use = (i < tx_ring->count) ? i : 0; - - /* set bits to identify this as an advanced context descriptor */ - type_tucmd |= IXGBE_TXD_CMD_DEXT | IXGBE_ADVTXD_DTYP_CTXT; - - context_desc->vlan_macip_lens = cpu_to_le32(vlan_macip_lens); - context_desc->seqnum_seed = cpu_to_le32(fcoe_sof_eof); - context_desc->type_tucmd_mlhl = cpu_to_le32(type_tucmd); - context_desc->mss_l4len_idx = cpu_to_le32(mss_l4len_idx); -} - -static int ixgbe_tso(struct ixgbe_ring *tx_ring, struct sk_buff *skb, - u32 tx_flags, __be16 protocol, u8 *hdr_len) +static int ixgbe_tso(struct ixgbe_ring *tx_ring, + struct ixgbe_tx_buffer *first, + u8 *hdr_len) { - int err; + struct sk_buff *skb = first->skb; u32 vlan_macip_lens, type_tucmd; u32 mss_l4len_idx, l4len; @@ -6304,7 +5694,7 @@ static int ixgbe_tso(struct ixgbe_ring *tx_ring, struct sk_buff *skb, return 0; if (skb_header_cloned(skb)) { - err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC); + int err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC); if (err) return err; } @@ -6312,7 +5702,7 @@ static int ixgbe_tso(struct ixgbe_ring *tx_ring, struct sk_buff *skb, /* ADV DTYP TUCMD MKRLOC/ISCSIHEDLEN */ type_tucmd = IXGBE_ADVTXD_TUCMD_L4T_TCP; - if (protocol == __constant_htons(ETH_P_IP)) { + if (first->protocol == __constant_htons(ETH_P_IP)) { struct iphdr *iph = ip_hdr(skb); iph->tot_len = 0; iph->check = 0; @@ -6321,17 +5711,27 @@ static int ixgbe_tso(struct ixgbe_ring *tx_ring, struct sk_buff *skb, IPPROTO_TCP, 0); type_tucmd |= IXGBE_ADVTXD_TUCMD_IPV4; + first->tx_flags |= IXGBE_TX_FLAGS_TSO | + IXGBE_TX_FLAGS_CSUM | + IXGBE_TX_FLAGS_IPV4; } else if (skb_is_gso_v6(skb)) { ipv6_hdr(skb)->payload_len = 0; tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr, 0, IPPROTO_TCP, 0); + first->tx_flags |= IXGBE_TX_FLAGS_TSO | + IXGBE_TX_FLAGS_CSUM; } + /* compute header lengths */ l4len = tcp_hdrlen(skb); *hdr_len = skb_transport_offset(skb) + l4len; + /* update gso size and bytecount with header size */ + first->gso_segs = skb_shinfo(skb)->gso_segs; + first->bytecount += (first->gso_segs - 1) * *hdr_len; + /* mss_l4len_id: use 1 as index for TSO */ mss_l4len_idx = l4len << IXGBE_ADVTXD_L4LEN_SHIFT; mss_l4len_idx |= skb_shinfo(skb)->gso_size << IXGBE_ADVTXD_MSS_SHIFT; @@ -6340,29 +5740,29 @@ static int ixgbe_tso(struct ixgbe_ring *tx_ring, struct sk_buff *skb, /* vlan_macip_lens: HEADLEN, MACLEN, VLAN tag */ vlan_macip_lens = skb_network_header_len(skb); vlan_macip_lens |= skb_network_offset(skb) << IXGBE_ADVTXD_MACLEN_SHIFT; - vlan_macip_lens |= tx_flags & IXGBE_TX_FLAGS_VLAN_MASK; + vlan_macip_lens |= first->tx_flags & IXGBE_TX_FLAGS_VLAN_MASK; ixgbe_tx_ctxtdesc(tx_ring, vlan_macip_lens, 0, type_tucmd, - mss_l4len_idx); + mss_l4len_idx); return 1; } -static bool ixgbe_tx_csum(struct ixgbe_ring *tx_ring, - struct sk_buff *skb, u32 tx_flags, - __be16 protocol) +static void ixgbe_tx_csum(struct ixgbe_ring *tx_ring, + struct ixgbe_tx_buffer *first) { + struct sk_buff *skb = first->skb; u32 vlan_macip_lens = 0; u32 mss_l4len_idx = 0; u32 type_tucmd = 0; if (skb->ip_summed != CHECKSUM_PARTIAL) { - if (!(tx_flags & IXGBE_TX_FLAGS_HW_VLAN) && - !(tx_flags & IXGBE_TX_FLAGS_TXSW)) - return false; + if (!(first->tx_flags & IXGBE_TX_FLAGS_HW_VLAN) && + !(first->tx_flags & IXGBE_TX_FLAGS_TXSW)) + return; } else { u8 l4_hdr = 0; - switch (protocol) { + switch (first->protocol) { case __constant_htons(ETH_P_IP): vlan_macip_lens |= skb_network_header_len(skb); type_tucmd |= IXGBE_ADVTXD_TUCMD_IPV4; @@ -6376,7 +5776,7 @@ static bool ixgbe_tx_csum(struct ixgbe_ring *tx_ring, if (unlikely(net_ratelimit())) { dev_warn(tx_ring->dev, "partial checksum but proto=%x!\n", - skb->protocol); + first->protocol); } break; } @@ -6400,19 +5800,21 @@ static bool ixgbe_tx_csum(struct ixgbe_ring *tx_ring, if (unlikely(net_ratelimit())) { dev_warn(tx_ring->dev, "partial checksum but l4 proto=%x!\n", - skb->protocol); + l4_hdr); } break; } + + /* update TX checksum flag */ + first->tx_flags |= IXGBE_TX_FLAGS_CSUM; } + /* vlan_macip_lens: MACLEN, VLAN tag */ vlan_macip_lens |= skb_network_offset(skb) << IXGBE_ADVTXD_MACLEN_SHIFT; - vlan_macip_lens |= tx_flags & IXGBE_TX_FLAGS_VLAN_MASK; + vlan_macip_lens |= first->tx_flags & IXGBE_TX_FLAGS_VLAN_MASK; ixgbe_tx_ctxtdesc(tx_ring, vlan_macip_lens, 0, type_tucmd, mss_l4len_idx); - - return (skb->ip_summed == CHECKSUM_PARTIAL); } static __le32 ixgbe_tx_cmd_type(u32 tx_flags) @@ -6428,7 +5830,7 @@ static __le32 ixgbe_tx_cmd_type(u32 tx_flags) /* set segmentation enable bits for TSO/FSO */ #ifdef IXGBE_FCOE - if ((tx_flags & IXGBE_TX_FLAGS_TSO) || (tx_flags & IXGBE_TX_FLAGS_FSO)) + if (tx_flags & (IXGBE_TX_FLAGS_TSO | IXGBE_TX_FLAGS_FSO)) #else if (tx_flags & IXGBE_TX_FLAGS_TSO) #endif @@ -6437,200 +5839,192 @@ static __le32 ixgbe_tx_cmd_type(u32 tx_flags) return cmd_type; } -static __le32 ixgbe_tx_olinfo_status(u32 tx_flags, unsigned int paylen) +static void ixgbe_tx_olinfo_status(union ixgbe_adv_tx_desc *tx_desc, + u32 tx_flags, unsigned int paylen) { - __le32 olinfo_status = - cpu_to_le32(paylen << IXGBE_ADVTXD_PAYLEN_SHIFT); - - if (tx_flags & IXGBE_TX_FLAGS_TSO) { - olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_POPTS_TXSM | - (1 << IXGBE_ADVTXD_IDX_SHIFT)); - /* enble IPv4 checksum for TSO */ - if (tx_flags & IXGBE_TX_FLAGS_IPV4) - olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_POPTS_IXSM); - } + __le32 olinfo_status = cpu_to_le32(paylen << IXGBE_ADVTXD_PAYLEN_SHIFT); /* enable L4 checksum for TSO and TX checksum offload */ if (tx_flags & IXGBE_TX_FLAGS_CSUM) olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_POPTS_TXSM); -#ifdef IXGBE_FCOE - /* use index 1 context for FCOE/FSO */ - if (tx_flags & IXGBE_TX_FLAGS_FCOE) - olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_CC | - (1 << IXGBE_ADVTXD_IDX_SHIFT)); + /* enble IPv4 checksum for TSO */ + if (tx_flags & IXGBE_TX_FLAGS_IPV4) + olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_POPTS_IXSM); + /* use index 1 context for TSO/FSO/FCOE */ +#ifdef IXGBE_FCOE + if (tx_flags & (IXGBE_TX_FLAGS_TSO | IXGBE_TX_FLAGS_FCOE)) +#else + if (tx_flags & IXGBE_TX_FLAGS_TSO) #endif + olinfo_status |= cpu_to_le32(1 << IXGBE_ADVTXD_IDX_SHIFT); + /* * Check Context must be set if Tx switch is enabled, which it * always is for case where virtual functions are running */ +#ifdef IXGBE_FCOE + if (tx_flags & (IXGBE_TX_FLAGS_TXSW | IXGBE_TX_FLAGS_FCOE)) +#else if (tx_flags & IXGBE_TX_FLAGS_TXSW) +#endif olinfo_status |= cpu_to_le32(IXGBE_ADVTXD_CC); - return olinfo_status; + tx_desc->read.olinfo_status = olinfo_status; } #define IXGBE_TXD_CMD (IXGBE_TXD_CMD_EOP | \ IXGBE_TXD_CMD_RS) static void ixgbe_tx_map(struct ixgbe_ring *tx_ring, - struct sk_buff *skb, struct ixgbe_tx_buffer *first, - u32 tx_flags, const u8 hdr_len) { - struct device *dev = tx_ring->dev; - struct ixgbe_tx_buffer *tx_buffer_info; - union ixgbe_adv_tx_desc *tx_desc; dma_addr_t dma; - __le32 cmd_type, olinfo_status; - struct skb_frag_struct *frag; - unsigned int f = 0; + struct sk_buff *skb = first->skb; + struct ixgbe_tx_buffer *tx_buffer; + union ixgbe_adv_tx_desc *tx_desc; + struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[0]; unsigned int data_len = skb->data_len; unsigned int size = skb_headlen(skb); - u32 offset = 0; - u32 paylen = skb->len - hdr_len; + unsigned int paylen = skb->len - hdr_len; + u32 tx_flags = first->tx_flags; + __le32 cmd_type; u16 i = tx_ring->next_to_use; - u16 gso_segs; + + tx_desc = IXGBE_TX_DESC(tx_ring, i); + + ixgbe_tx_olinfo_status(tx_desc, tx_flags, paylen); + cmd_type = ixgbe_tx_cmd_type(tx_flags); #ifdef IXGBE_FCOE if (tx_flags & IXGBE_TX_FLAGS_FCOE) { - if (data_len >= sizeof(struct fcoe_crc_eof)) { - data_len -= sizeof(struct fcoe_crc_eof); - } else { + if (data_len < sizeof(struct fcoe_crc_eof)) { size -= sizeof(struct fcoe_crc_eof) - data_len; data_len = 0; + } else { + data_len -= sizeof(struct fcoe_crc_eof); } } #endif - dma = dma_map_single(dev, skb->data, size, DMA_TO_DEVICE); - if (dma_mapping_error(dev, dma)) + dma = dma_map_single(tx_ring->dev, skb->data, size, DMA_TO_DEVICE); + if (dma_mapping_error(tx_ring->dev, dma)) goto dma_error; - cmd_type = ixgbe_tx_cmd_type(tx_flags); - olinfo_status = ixgbe_tx_olinfo_status(tx_flags, paylen); + /* record length, and DMA address */ + dma_unmap_len_set(first, len, size); + dma_unmap_addr_set(first, dma, dma); - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, i); + tx_desc->read.buffer_addr = cpu_to_le64(dma); for (;;) { - while (size > IXGBE_MAX_DATA_PER_TXD) { - tx_desc->read.buffer_addr = cpu_to_le64(dma + offset); + while (unlikely(size > IXGBE_MAX_DATA_PER_TXD)) { tx_desc->read.cmd_type_len = cmd_type | cpu_to_le32(IXGBE_MAX_DATA_PER_TXD); - tx_desc->read.olinfo_status = olinfo_status; - - offset += IXGBE_MAX_DATA_PER_TXD; - size -= IXGBE_MAX_DATA_PER_TXD; - tx_desc++; i++; + tx_desc++; if (i == tx_ring->count) { - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, 0); + tx_desc = IXGBE_TX_DESC(tx_ring, 0); i = 0; } + + dma += IXGBE_MAX_DATA_PER_TXD; + size -= IXGBE_MAX_DATA_PER_TXD; + + tx_desc->read.buffer_addr = cpu_to_le64(dma); + tx_desc->read.olinfo_status = 0; } - tx_buffer_info = &tx_ring->tx_buffer_info[i]; - tx_buffer_info->length = offset + size; - tx_buffer_info->tx_flags = tx_flags; - tx_buffer_info->dma = dma; + if (likely(!data_len)) + break; - tx_desc->read.buffer_addr = cpu_to_le64(dma + offset); + if (unlikely(skb->no_fcs)) + cmd_type &= ~(cpu_to_le32(IXGBE_ADVTXD_DCMD_IFCS)); tx_desc->read.cmd_type_len = cmd_type | cpu_to_le32(size); - tx_desc->read.olinfo_status = olinfo_status; - if (!data_len) - break; + i++; + tx_desc++; + if (i == tx_ring->count) { + tx_desc = IXGBE_TX_DESC(tx_ring, 0); + i = 0; + } - frag = &skb_shinfo(skb)->frags[f]; #ifdef IXGBE_FCOE size = min_t(unsigned int, data_len, skb_frag_size(frag)); #else size = skb_frag_size(frag); #endif data_len -= size; - f++; - - offset = 0; - tx_flags |= IXGBE_TX_FLAGS_MAPPED_AS_PAGE; - dma = skb_frag_dma_map(dev, frag, 0, size, DMA_TO_DEVICE); - if (dma_mapping_error(dev, dma)) + dma = skb_frag_dma_map(tx_ring->dev, frag, 0, size, + DMA_TO_DEVICE); + if (dma_mapping_error(tx_ring->dev, dma)) goto dma_error; - tx_desc++; - i++; - if (i == tx_ring->count) { - tx_desc = IXGBE_TX_DESC_ADV(tx_ring, 0); - i = 0; - } - } + tx_buffer = &tx_ring->tx_buffer_info[i]; + dma_unmap_len_set(tx_buffer, len, size); + dma_unmap_addr_set(tx_buffer, dma, dma); - tx_desc->read.cmd_type_len |= cpu_to_le32(IXGBE_TXD_CMD); + tx_desc->read.buffer_addr = cpu_to_le64(dma); + tx_desc->read.olinfo_status = 0; - i++; - if (i == tx_ring->count) - i = 0; - - tx_ring->next_to_use = i; + frag++; + } - if (tx_flags & IXGBE_TX_FLAGS_TSO) - gso_segs = skb_shinfo(skb)->gso_segs; -#ifdef IXGBE_FCOE - /* adjust for FCoE Sequence Offload */ - else if (tx_flags & IXGBE_TX_FLAGS_FSO) - gso_segs = DIV_ROUND_UP(skb->len - hdr_len, - skb_shinfo(skb)->gso_size); -#endif /* IXGBE_FCOE */ - else - gso_segs = 1; + /* write last descriptor with RS and EOP bits */ + cmd_type |= cpu_to_le32(size) | cpu_to_le32(IXGBE_TXD_CMD); + tx_desc->read.cmd_type_len = cmd_type; - /* multiply data chunks by size of headers */ - tx_buffer_info->bytecount = paylen + (gso_segs * hdr_len); - tx_buffer_info->gso_segs = gso_segs; - tx_buffer_info->skb = skb; + netdev_tx_sent_queue(txring_txq(tx_ring), first->bytecount); /* set the timestamp */ first->time_stamp = jiffies; /* - * Force memory writes to complete before letting h/w - * know there are new descriptors to fetch. (Only - * applicable for weak-ordered memory model archs, - * such as IA-64). + * Force memory writes to complete before letting h/w know there + * are new descriptors to fetch. (Only applicable for weak-ordered + * memory model archs, such as IA-64). + * + * We also need this memory barrier to make certain all of the + * status bits have been updated before next_to_watch is written. */ wmb(); /* set next_to_watch value indicating a packet is present */ first->next_to_watch = tx_desc; + i++; + if (i == tx_ring->count) + i = 0; + + tx_ring->next_to_use = i; + /* notify HW of packet */ writel(i, tx_ring->tail); return; dma_error: - dev_err(dev, "TX DMA map failed\n"); + dev_err(tx_ring->dev, "TX DMA map failed\n"); /* clear dma mappings for failed tx_buffer_info map */ for (;;) { - tx_buffer_info = &tx_ring->tx_buffer_info[i]; - ixgbe_unmap_tx_resource(tx_ring, tx_buffer_info); - if (tx_buffer_info == first) + tx_buffer = &tx_ring->tx_buffer_info[i]; + ixgbe_unmap_and_free_tx_resource(tx_ring, tx_buffer); + if (tx_buffer == first) break; if (i == 0) i = tx_ring->count; i--; } - dev_kfree_skb_any(skb); - tx_ring->next_to_use = i; } -static void ixgbe_atr(struct ixgbe_ring *ring, struct sk_buff *skb, - u32 tx_flags, __be16 protocol) +static void ixgbe_atr(struct ixgbe_ring *ring, + struct ixgbe_tx_buffer *first) { struct ixgbe_q_vector *q_vector = ring->q_vector; union ixgbe_atr_hash_dword input = { .dword = 0 }; @@ -6654,16 +6048,16 @@ static void ixgbe_atr(struct ixgbe_ring *ring, struct sk_buff *skb, ring->atr_count++; /* snag network header to get L4 type and address */ - hdr.network = skb_network_header(skb); + hdr.network = skb_network_header(first->skb); /* Currently only IPv4/IPv6 with TCP is supported */ - if ((protocol != __constant_htons(ETH_P_IPV6) || + if ((first->protocol != __constant_htons(ETH_P_IPV6) || hdr.ipv6->nexthdr != IPPROTO_TCP) && - (protocol != __constant_htons(ETH_P_IP) || + (first->protocol != __constant_htons(ETH_P_IP) || hdr.ipv4->protocol != IPPROTO_TCP)) return; - th = tcp_hdr(skb); + th = tcp_hdr(first->skb); /* skip this packet since it is invalid or the socket is closing */ if (!th || th->fin) @@ -6676,7 +6070,7 @@ static void ixgbe_atr(struct ixgbe_ring *ring, struct sk_buff *skb, /* reset sample count */ ring->atr_count = 0; - vlan_id = htons(tx_flags >> IXGBE_TX_FLAGS_VLAN_SHIFT); + vlan_id = htons(first->tx_flags >> IXGBE_TX_FLAGS_VLAN_SHIFT); /* * src and dst are inverted, think how the receiver sees them @@ -6691,13 +6085,13 @@ static void ixgbe_atr(struct ixgbe_ring *ring, struct sk_buff *skb, * since src port and flex bytes occupy the same word XOR them together * and write the value to source port portion of compressed dword */ - if (tx_flags & (IXGBE_TX_FLAGS_SW_VLAN | IXGBE_TX_FLAGS_HW_VLAN)) + if (first->tx_flags & (IXGBE_TX_FLAGS_SW_VLAN | IXGBE_TX_FLAGS_HW_VLAN)) common.port.src ^= th->dest ^ __constant_htons(ETH_P_8021Q); else - common.port.src ^= th->dest ^ protocol; + common.port.src ^= th->dest ^ first->protocol; common.port.dst ^= th->source; - if (protocol == __constant_htons(ETH_P_IP)) { + if (first->protocol == __constant_htons(ETH_P_IP)) { input.formatted.flow_type = IXGBE_ATR_FLOW_TYPE_TCPV4; common.ip ^= hdr.ipv4->saddr ^ hdr.ipv4->daddr; } else { @@ -6785,7 +6179,7 @@ netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb, /* * need: 1 descriptor per page * PAGE_SIZE/IXGBE_MAX_DATA_PER_TXD, - * + 1 desc for skb_head_len/IXGBE_MAX_DATA_PER_TXD, + * + 1 desc for skb_headlen/IXGBE_MAX_DATA_PER_TXD, * + 2 desc gap to keep tail from touching head, * + 1 desc for context descriptor, * otherwise try next time @@ -6801,11 +6195,12 @@ netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb, return NETDEV_TX_BUSY; } -#ifdef CONFIG_PCI_IOV - if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) - tx_flags |= IXGBE_TX_FLAGS_TXSW; + /* record the location of the first descriptor for this packet */ + first = &tx_ring->tx_buffer_info[tx_ring->next_to_use]; + first->skb = skb; + first->bytecount = skb->len; + first->gso_segs = 1; -#endif /* if we have a HW VLAN tag being added default to the HW one */ if (vlan_tx_tag_present(skb)) { tx_flags |= vlan_tx_tag_get(skb) << IXGBE_TX_FLAGS_VLAN_SHIFT; @@ -6818,10 +6213,20 @@ netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb, goto out_drop; protocol = vhdr->h_vlan_encapsulated_proto; - tx_flags |= ntohs(vhdr->h_vlan_TCI) << IXGBE_TX_FLAGS_VLAN_SHIFT; + tx_flags |= ntohs(vhdr->h_vlan_TCI) << + IXGBE_TX_FLAGS_VLAN_SHIFT; tx_flags |= IXGBE_TX_FLAGS_SW_VLAN; } +#ifdef CONFIG_PCI_IOV + /* + * Use the l2switch_enable flag - would be false if the DMA + * Tx switch had been disabled. + */ + if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) + tx_flags |= IXGBE_TX_FLAGS_TXSW; + +#endif /* DCB maps skb priorities 0-7 onto 3 bit PCP of VLAN tag. */ if ((adapter->flags & IXGBE_FLAG_DCB_ENABLED) && ((tx_flags & (IXGBE_TX_FLAGS_HW_VLAN | IXGBE_TX_FLAGS_SW_VLAN)) || @@ -6842,61 +6247,69 @@ netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb, } } - /* record the location of the first descriptor for this packet */ - first = &tx_ring->tx_buffer_info[tx_ring->next_to_use]; + /* record initial flags and protocol */ + first->tx_flags = tx_flags; + first->protocol = protocol; #ifdef IXGBE_FCOE /* setup tx offload for FCoE */ if ((protocol == __constant_htons(ETH_P_FCOE)) && (adapter->flags & IXGBE_FLAG_FCOE_ENABLED)) { - tso = ixgbe_fso(tx_ring, skb, tx_flags, &hdr_len); + tso = ixgbe_fso(tx_ring, first, &hdr_len); if (tso < 0) goto out_drop; - else if (tso) - tx_flags |= IXGBE_TX_FLAGS_FSO | - IXGBE_TX_FLAGS_FCOE; - else - tx_flags |= IXGBE_TX_FLAGS_FCOE; goto xmit_fcoe; } #endif /* IXGBE_FCOE */ - /* setup IPv4/IPv6 offloads */ - if (protocol == __constant_htons(ETH_P_IP)) - tx_flags |= IXGBE_TX_FLAGS_IPV4; - - tso = ixgbe_tso(tx_ring, skb, tx_flags, protocol, &hdr_len); + tso = ixgbe_tso(tx_ring, first, &hdr_len); if (tso < 0) goto out_drop; - else if (tso) - tx_flags |= IXGBE_TX_FLAGS_TSO; - else if (ixgbe_tx_csum(tx_ring, skb, tx_flags, protocol)) - tx_flags |= IXGBE_TX_FLAGS_CSUM; + else if (!tso) + ixgbe_tx_csum(tx_ring, first); /* add the ATR filter if ATR is on */ if (test_bit(__IXGBE_TX_FDIR_INIT_DONE, &tx_ring->state)) - ixgbe_atr(tx_ring, skb, tx_flags, protocol); + ixgbe_atr(tx_ring, first); #ifdef IXGBE_FCOE xmit_fcoe: #endif /* IXGBE_FCOE */ - ixgbe_tx_map(tx_ring, skb, first, tx_flags, hdr_len); + ixgbe_tx_map(tx_ring, first, hdr_len); ixgbe_maybe_stop_tx(tx_ring, DESC_NEEDED); return NETDEV_TX_OK; out_drop: - dev_kfree_skb_any(skb); + dev_kfree_skb_any(first->skb); + first->skb = NULL; + return NETDEV_TX_OK; } -static netdev_tx_t ixgbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev) +static netdev_tx_t ixgbe_xmit_frame(struct sk_buff *skb, + struct net_device *netdev) { struct ixgbe_adapter *adapter = netdev_priv(netdev); struct ixgbe_ring *tx_ring; + if (skb->len <= 0) { + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + + /* + * The minimum packet size for olinfo paylen is 17 so pad the skb + * in order to meet this minimum size requirement. + */ + if (skb->len < 17) { + if (skb_padto(skb, 17)) + return NETDEV_TX_OK; + skb->len = 17; + } + tx_ring = adapter->tx_ring[skb->queue_mapping]; return ixgbe_xmit_frame_ring(skb, adapter, tx_ring); } @@ -7029,8 +6442,8 @@ static void ixgbe_netpoll(struct net_device *netdev) } adapter->flags &= ~IXGBE_FLAG_IN_NETPOLL; } -#endif +#endif static struct rtnl_link_stats64 *ixgbe_get_stats64(struct net_device *netdev, struct rtnl_link_stats64 *stats) { @@ -7079,6 +6492,7 @@ static struct rtnl_link_stats64 *ixgbe_get_stats64(struct net_device *netdev, return stats; } +#ifdef CONFIG_IXGBE_DCB /* ixgbe_validate_rtr - verify 802.1Qp to Rx packet buffer mapping is valid. * #adapter: pointer to ixgbe_adapter * @tc: number of traffic classes currently enabled @@ -7115,7 +6529,6 @@ static void ixgbe_validate_rtr(struct ixgbe_adapter *adapter, u8 tc) return; } - /* ixgbe_setup_tc - routine to configure net_device for multiple traffic * classes. * @@ -7135,7 +6548,8 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc) /* Hardware supports up to 8 traffic classes */ if (tc > adapter->dcb_cfg.num_tcs.pg_tcs || - (hw->mac.type == ixgbe_mac_82598EB && tc < MAX_TRAFFIC_CLASS)) + (hw->mac.type == ixgbe_mac_82598EB && + tc < MAX_TRAFFIC_CLASS)) return -EINVAL; /* Hardware has to reinitialize queues and interrupts to @@ -7149,7 +6563,6 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc) if (tc) { netdev_set_num_tc(dev, tc); adapter->last_lfc_mode = adapter->hw.fc.current_mode; - adapter->flags |= IXGBE_FLAG_DCB_ENABLED; adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE; @@ -7157,7 +6570,6 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc) adapter->hw.fc.requested_mode = ixgbe_fc_none; } else { netdev_reset_tc(dev); - adapter->hw.fc.requested_mode = adapter->last_lfc_mode; adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED; @@ -7175,6 +6587,7 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc) return 0; } +#endif /* CONFIG_IXGBE_DCB */ void ixgbe_do_reset(struct net_device *netdev) { struct ixgbe_adapter *adapter = netdev_priv(netdev); @@ -7186,59 +6599,52 @@ void ixgbe_do_reset(struct net_device *netdev) } static netdev_features_t ixgbe_fix_features(struct net_device *netdev, - netdev_features_t data) + netdev_features_t features) { struct ixgbe_adapter *adapter = netdev_priv(netdev); #ifdef CONFIG_DCB if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) - data &= ~NETIF_F_HW_VLAN_RX; + features &= ~NETIF_F_HW_VLAN_RX; #endif /* return error if RXHASH is being enabled when RSS is not supported */ if (!(adapter->flags & IXGBE_FLAG_RSS_ENABLED)) - data &= ~NETIF_F_RXHASH; + features &= ~NETIF_F_RXHASH; /* If Rx checksum is disabled, then RSC/LRO should also be disabled */ - if (!(data & NETIF_F_RXCSUM)) - data &= ~NETIF_F_LRO; + if (!(features & NETIF_F_RXCSUM)) + features &= ~NETIF_F_LRO; - /* Turn off LRO if not RSC capable or invalid ITR settings */ - if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)) { - data &= ~NETIF_F_LRO; - } else if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) && - (adapter->rx_itr_setting != 1 && - adapter->rx_itr_setting > IXGBE_MAX_RSC_INT_RATE)) { - data &= ~NETIF_F_LRO; - e_info(probe, "rx-usecs set too low, not enabling RSC\n"); - } + /* Turn off LRO if not RSC capable */ + if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)) + features &= ~NETIF_F_LRO; + - return data; + return features; } static int ixgbe_set_features(struct net_device *netdev, - netdev_features_t data) + netdev_features_t features) { struct ixgbe_adapter *adapter = netdev_priv(netdev); + netdev_features_t changed = netdev->features ^ features; bool need_reset = false; - /* If Rx checksum is disabled, then RSC/LRO should also be disabled */ - if (!(data & NETIF_F_RXCSUM)) - adapter->flags &= ~IXGBE_FLAG_RX_CSUM_ENABLED; - else - adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED; - /* Make sure RSC matches LRO, reset if change */ - if (!!(data & NETIF_F_LRO) != - !!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)) { - adapter->flags2 ^= IXGBE_FLAG2_RSC_ENABLED; - switch (adapter->hw.mac.type) { - case ixgbe_mac_X540: - case ixgbe_mac_82599EB: + if (!(features & NETIF_F_LRO)) { + if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) need_reset = true; - break; - default: - break; + adapter->flags2 &= ~IXGBE_FLAG2_RSC_ENABLED; + } else if ((adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE) && + !(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)) { + if (adapter->rx_itr_setting == 1 || + adapter->rx_itr_setting > IXGBE_MIN_RSC_ITR) { + adapter->flags2 |= IXGBE_FLAG2_RSC_ENABLED; + need_reset = true; + } else if ((changed ^ features) & NETIF_F_LRO) { + e_info(probe, "rx-usecs set too low, " + "disabling RSC\n"); } } @@ -7246,27 +6652,30 @@ static int ixgbe_set_features(struct net_device *netdev, * Check if Flow Director n-tuple support was enabled or disabled. If * the state changed, we need to reset. */ - if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)) { - /* turn off ATR, enable perfect filters and reset */ - if (data & NETIF_F_NTUPLE) { - adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE; - adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE; + if (!(features & NETIF_F_NTUPLE)) { + if (adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE) { + /* turn off Flow Director, set ATR and reset */ + if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) && + !(adapter->flags & IXGBE_FLAG_DCB_ENABLED)) + adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE; need_reset = true; } - } else if (!(data & NETIF_F_NTUPLE)) { - /* turn off Flow Director, set ATR and reset */ adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE; - if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) && - !(adapter->flags & IXGBE_FLAG_DCB_ENABLED)) - adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE; + } else if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)) { + /* turn off ATR, enable perfect filters and reset */ + adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE; + adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE; need_reset = true; } + if (changed & NETIF_F_RXALL) + need_reset = true; + + netdev->features = features; if (need_reset) ixgbe_do_reset(netdev); return 0; - } static const struct net_device_ops ixgbe_netdev_ops = { @@ -7274,7 +6683,7 @@ static const struct net_device_ops ixgbe_netdev_ops = { .ndo_stop = ixgbe_close, .ndo_start_xmit = ixgbe_xmit_frame, .ndo_select_queue = ixgbe_select_queue, - .ndo_set_rx_mode = ixgbe_set_rx_mode, + .ndo_set_rx_mode = ixgbe_set_rx_mode, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = ixgbe_set_mac, .ndo_change_mtu = ixgbe_change_mtu, @@ -7285,10 +6694,12 @@ static const struct net_device_ops ixgbe_netdev_ops = { .ndo_set_vf_mac = ixgbe_ndo_set_vf_mac, .ndo_set_vf_vlan = ixgbe_ndo_set_vf_vlan, .ndo_set_vf_tx_rate = ixgbe_ndo_set_vf_bw, - .ndo_set_vf_spoofchk = ixgbe_ndo_set_vf_spoofchk, + .ndo_set_vf_spoofchk = ixgbe_ndo_set_vf_spoofchk, .ndo_get_vf_config = ixgbe_ndo_get_vf_config, .ndo_get_stats64 = ixgbe_get_stats64, +#ifdef CONFIG_IXGBE_DCB .ndo_setup_tc = ixgbe_setup_tc, +#endif #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = ixgbe_netpoll, #endif @@ -7306,7 +6717,7 @@ static const struct net_device_ops ixgbe_netdev_ops = { }; static void __devinit ixgbe_probe_vf(struct ixgbe_adapter *adapter, - const struct ixgbe_info *ii) + const struct ixgbe_info *ii) { #ifdef CONFIG_PCI_IOV struct ixgbe_hw *hw = &adapter->hw; @@ -7493,6 +6904,9 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, e_crit(probe, "Fan has stopped, replace the adapter\n"); } + if (allow_unsupported_sfp) + hw->allow_unsupported_sfp = allow_unsupported_sfp; + /* reset_hw fills in the perm_addr as well */ hw->phy.reset_if_overtemp = true; err = hw->mac.ops.reset_hw(hw); @@ -7537,6 +6951,8 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, break; } + netdev->hw_features |= NETIF_F_RXALL; + netdev->vlan_features |= NETIF_F_TSO; netdev->vlan_features |= NETIF_F_TSO6; netdev->vlan_features |= NETIF_F_IP_CSUM; @@ -7544,6 +6960,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, netdev->vlan_features |= NETIF_F_SG; netdev->priv_flags |= IFF_UNICAST_FLT; + netdev->priv_flags |= IFF_SUPP_NOFCS; if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) adapter->flags &= ~(IXGBE_FLAG_RSS_ENABLED | @@ -7581,7 +6998,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, if (hw->eeprom.ops.validate_checksum(hw, NULL) < 0) { e_dev_err("The EEPROM Checksum Is Not Valid\n"); err = -EIO; - goto err_eeprom; + goto err_sw_init; } memcpy(netdev->dev_addr, hw->mac.perm_addr, netdev->addr_len); @@ -7590,11 +7007,11 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, if (ixgbe_validate_mac_addr(netdev->perm_addr)) { e_dev_err("invalid MAC address\n"); err = -EIO; - goto err_eeprom; + goto err_sw_init; } setup_timer(&adapter->service_timer, &ixgbe_service_timer, - (unsigned long) adapter); + (unsigned long) adapter); INIT_WORK(&adapter->service_task, ixgbe_service_task); clear_bit(__IXGBE_SERVICE_SCHED, &adapter->state); @@ -7682,7 +7099,6 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev, /* reset the hardware with the new settings */ err = hw->mac.ops.start_hw(hw); - if (err == IXGBE_ERR_EEPROM_VERSION) { /* We are running on a pre-production device, log a warning */ e_dev_warn("This device is a pre-production adapter/LOM. " @@ -7737,7 +7153,6 @@ err_register: ixgbe_release_hw_control(adapter); ixgbe_clear_interrupt_scheme(adapter); err_sw_init: -err_eeprom: if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) ixgbe_disable_sriov(adapter); adapter->flags2 &= ~IXGBE_FLAG2_SEARCH_FOR_SFP; |