Re: [PATCH] sky2: safer transmit ring cleaning
From: Michael Breuer <hidden>
Date: 2010-01-12 18:36:01
On 1/12/2010 11:15 AM, Stephen Hemminger wrote:
quoted hunk ↗ jump to hunk
This code makes transmit path and transmit reset safer by: * adding memory barrier before checking available ring slots * reseting state of tx ring elements after free * seperate cleanup function from ring done function * removing mostly unused tx_next element Signed-off-by: Stephen Hemminger<redacted> --- Please apply this instead of the various bits and pieces flying around labeled as sky2 panic under load--- a/drivers/net/sky2.c 2010-01-11 10:49:50.907113126 -0800 +++ b/drivers/net/sky2.c 2010-01-11 17:36:22.027429875 -0800@@ -1596,6 +1596,9 @@ static inline int tx_inuse(const struct /* Number of list elements available for next tx */ static inline int tx_avail(const struct sky2_port *sky2) { + /* Makes sure update of tx_prod from start_xmit and + tx_cons from tx_done are seen. */ + smp_mb(); return sky2->tx_pending - tx_inuse(sky2); }@@ -1618,8 +1621,7 @@ static unsigned tx_le_req(const struct s return count; } -static void sky2_tx_unmap(struct pci_dev *pdev, - const struct tx_ring_info *re) +static void sky2_tx_unmap(struct pci_dev *pdev, struct tx_ring_info *re) { if (re->flags& TX_MAP_SINGLE) pci_unmap_single(pdev, pci_unmap_addr(re, mapaddr),@@ -1629,6 +1631,7 @@ static void sky2_tx_unmap(struct pci_dev pci_unmap_page(pdev, pci_unmap_addr(re, mapaddr), pci_unmap_len(re, maplen), PCI_DMA_TODEVICE); + re->flags = 0; } /*@@ -1804,7 +1807,8 @@ mapping_error: } /* - * Free ring elements from starting at tx_cons until "done" + * Transmit complete processing + * Free ring elements from starting at tx_cons until done index * * NB: * 1. The hardware will tell us about partial completion of multi-part@@ -1813,9 +1817,9 @@ mapping_error: * looks at the tail of the queue of FIFO (tx_cons), not * the head (tx_prod) */ -static void sky2_tx_complete(struct sky2_port *sky2, u16 done) +static void sky2_tx_done(struct net_device *dev, u16 done) { - struct net_device *dev = sky2->netdev; + struct sky2_port *sky2 = netdev_priv(dev); unsigned idx; BUG_ON(done>= sky2->tx_ring_size);@@ -1828,6 +1832,8 @@ static void sky2_tx_complete(struct sky2 sky2_tx_unmap(sky2->hw->pdev, re); if (skb) { + re->skb = NULL; + if (unlikely(netif_msg_tx_done(sky2))) printk(KERN_DEBUG "%s: tx done %u\n", dev->name, idx);@@ -1836,13 +1842,10 @@ static void sky2_tx_complete(struct sky2 dev->stats.tx_bytes += skb->len; dev_kfree_skb_any(skb); - - sky2->tx_next = RING_NEXT(idx, sky2->tx_ring_size); } } sky2->tx_cons = idx; - smp_mb(); if (tx_avail(sky2)> MAX_SKB_TX_LE + 4) netif_wake_queue(dev);@@ -1870,6 +1873,21 @@ static void sky2_tx_reset(struct sky2_hw sky2_write8(hw, SK_REG(port, TX_GMF_CTRL_T), GMF_RST_SET); } +static void sky2_tx_clean(struct sky2_port *sky2) +{ + u16 idx; + + for (idx = 0; idx< sky2->tx_ring_size; idx++) { + struct tx_ring_info *re = sky2->tx_ring + idx; + + sky2_tx_unmap(sky2->hw->pdev, re); + if (re->skb) { + dev_kfree_skb_any(re->skb); + re->skb = NULL; + } + } +} + /* Network shutdown */ static int sky2_down(struct net_device *dev) {@@ -1933,8 +1951,7 @@ static int sky2_down(struct net_device * sky2_tx_reset(hw, port); /* Free any pending frames stuck in HW queue */ - sky2_tx_complete(sky2, sky2->tx_prod); - + sky2_tx_clean(sky2); sky2_rx_clean(sky2); sky2_free_buffers(sky2);@@ -2411,15 +2428,6 @@ error: goto resubmit; } -/* Transmit complete */ -static inline void sky2_tx_done(struct net_device *dev, u16 last) -{ - struct sky2_port *sky2 = netdev_priv(dev); - - if (netif_running(dev)) - sky2_tx_complete(sky2, last); -} - static inline void sky2_skb_rx(const struct sky2_port *sky2, u32 status, struct sk_buff *skb) {@@ -4201,7 +4209,7 @@ static int sky2_debug_show(struct seq_fi /* Dump contents of tx ring */ sop = 1; - for (idx = sky2->tx_next; idx != sky2->tx_prod&& idx< sky2->tx_ring_size; + for (idx = sky2->tx_cons; idx != sky2->tx_prod&& idx< sky2->tx_ring_size; idx = RING_NEXT(idx, sky2->tx_ring_size)) { const struct sky2_tx_le *le = sky2->tx_le + idx; u32 a = le32_to_cpu(le->addr); --- a/drivers/net/sky2.h 2010-01-11 17:29:22.817088617 -0800 +++ b/drivers/net/sky2.h 2010-01-11 17:29:28.197120484 -0800@@ -2187,7 +2187,6 @@ struct sky2_port { u16 tx_ring_size; u16 tx_cons; /* next le to check */ u16 tx_prod; /* next le to use */ - u16 tx_next; /* debug only */ u16 tx_pending; u16 tx_last_mss;
Test observations: 1. DHCP request/response sequence having some issues... can't confirm that it's a result of this patch, but I don't see this with the prior patch. Prior to this patch, if I connect a new device (Blackberry in this case) I see DHCPDISCOVER;DHCPOFFER;DHCPREQUEST;DHCPACK (just the four messages). With this patch I'm seeing repeated attempts - i.e., DISCOVER/OFFER 6 times before the REQUEST/ACK. This is repeatable and happening whether or not under load. As my original problem started with DHCP packets, this seems interesting. I don't see any errors logged (dmesg, messages, etc.). 2. Probably not related to this patch, but perhaps to the driver - I've finally tracked the source of my dropped RX packets. It's happening when a sending data to a CIFS client on a MacOS system. The RX drop rate seems proportional to the TX rate for SMB to that client - at a tx rate of about 200Kb/s I see about 20 dropped RX packets/sec - at 400 I see about 40. I'm thinking therefore it's ACKs being dropped on RX. Why? no idea (yet). Nothing reported by ethtool or netstat -s remotely correlates to the number of dropped RX packets.