From acfa9e94e2f06f8911a1d2f257880af4ad945eef Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 2 Jul 2012 08:36:12 +0000 Subject: net: dont use __netdev_alloc_skb for bounce buffer commit a1c7fff7e1 (net: netdev_alloc_skb() use build_skb()) broke b44 on some 64bit machines. It appears b44 and b43 use __netdev_alloc_skb() instead of alloc_skb() for their bounce buffers. There is no need to add an extra NET_SKB_PAD reservation for bounce buffers : - In TX path, NET_SKB_PAD is useless - In RX path in b44, we force a copy of incoming frames if GFP_DMA allocations were needed. Reported-and-bisected-by: Stefan Bader Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/b44.c | 4 ++-- drivers/net/wireless/b43legacy/dma.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/broadcom/b44.c b/drivers/net/ethernet/broadcom/b44.c index 46b8b7d81633..d09c6b583d17 100644 --- a/drivers/net/ethernet/broadcom/b44.c +++ b/drivers/net/ethernet/broadcom/b44.c @@ -656,7 +656,7 @@ static int b44_alloc_rx_skb(struct b44 *bp, int src_idx, u32 dest_idx_unmasked) dma_unmap_single(bp->sdev->dma_dev, mapping, RX_PKT_BUF_SZ, DMA_FROM_DEVICE); dev_kfree_skb_any(skb); - skb = __netdev_alloc_skb(bp->dev, RX_PKT_BUF_SZ, GFP_ATOMIC|GFP_DMA); + skb = alloc_skb(RX_PKT_BUF_SZ, GFP_ATOMIC | GFP_DMA); if (skb == NULL) return -ENOMEM; mapping = dma_map_single(bp->sdev->dma_dev, skb->data, @@ -967,7 +967,7 @@ static netdev_tx_t b44_start_xmit(struct sk_buff *skb, struct net_device *dev) dma_unmap_single(bp->sdev->dma_dev, mapping, len, DMA_TO_DEVICE); - bounce_skb = __netdev_alloc_skb(dev, len, GFP_ATOMIC | GFP_DMA); + bounce_skb = alloc_skb(len, GFP_ATOMIC | GFP_DMA); if (!bounce_skb) goto err_out; diff --git a/drivers/net/wireless/b43legacy/dma.c b/drivers/net/wireless/b43legacy/dma.c index f1f8bd09bd87..c8baf020c20f 100644 --- a/drivers/net/wireless/b43legacy/dma.c +++ b/drivers/net/wireless/b43legacy/dma.c @@ -1072,7 +1072,7 @@ static int dma_tx_fragment(struct b43legacy_dmaring *ring, meta->dmaaddr = map_descbuffer(ring, skb->data, skb->len, 1); /* create a bounce buffer in zone_dma on mapping failure. */ if (b43legacy_dma_mapping_error(ring, meta->dmaaddr, skb->len, 1)) { - bounce_skb = __dev_alloc_skb(skb->len, GFP_ATOMIC | GFP_DMA); + bounce_skb = alloc_skb(skb->len, GFP_ATOMIC | GFP_DMA); if (!bounce_skb) { ring->current_slot = old_top_slot; ring->used_slots = old_used_slots; -- cgit v1.2.3 From b94e52f62683dc0b00c6d1b58b80929a078c0fd5 Mon Sep 17 00:00:00 2001 From: Cloud Ren Date: Tue, 3 Jul 2012 16:51:48 +0000 Subject: atl1c: fix issue of transmit queue 0 timed out some people report atl1c could cause system hang with following kernel trace info: --------------------------------------- WARNING: at.../net/sched/sch_generic.c:258 dev_watchdog+0x1db/0x1d0() ... NETDEV WATCHDOG: eth0 (atl1c): transmit queue 0 timed out ... --------------------------------------- This is caused by netif_stop_queue calling when cable Link is down. So remove netif_stop_queue, because link_watch will take it over. Signed-off-by: xiong Cc: stable Signed-off-by: Cloud Ren Signed-off-by: David S. Miller --- drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c index 9cc15701101b..1f78b63d5efe 100644 --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c @@ -261,7 +261,6 @@ static void atl1c_check_link_status(struct atl1c_adapter *adapter) if ((phy_data & BMSR_LSTATUS) == 0) { /* link down */ netif_carrier_off(netdev); - netif_stop_queue(netdev); hw->hibernate = true; if (atl1c_reset_mac(hw) != 0) if (netif_msg_hw(adapter)) -- cgit v1.2.3 From d4e41649434cd6db2e69783130cba81886dac97f Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Wed, 4 Jul 2012 02:00:25 +0000 Subject: ixgbe: DCB and SR-IOV can not co-exist and will cause hangs DCB and SR-IOV cannot currently be enabled at the same time as the queueing schemes are incompatible. If they are both enabled it will result in Tx hangs since only the first Tx queue will be able to transmit any traffic. This simple fix for this is to block us from enabling TCs in ixgbe_setup_tc if SR-IOV is enabled. This change will be reverted once we can support SR-IOV and DCB coexistence. Signed-off-by: Alexander Duyck Acked-by: John Fastabend Tested-by: Phil Schmitt Tested-by: Ross Brattain Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'drivers') diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 18ca3bcadf0c..e242104ab471 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -6647,6 +6647,11 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc) return -EINVAL; } + if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) { + e_err(drv, "Enable failed, SR-IOV enabled\n"); + return -EINVAL; + } + /* Hardware supports up to 8 traffic classes */ if (tc > adapter->dcb_cfg.num_tcs.pg_tcs || (hw->mac.type == ixgbe_mac_82598EB && -- cgit v1.2.3 From b93984c9afacd4fe32b785d52a93660d91202b10 Mon Sep 17 00:00:00 2001 From: David Daney Date: Wed, 4 Jul 2012 12:06:16 +0000 Subject: netdev/phy: Fixup lockdep warnings in mdio-mux.c With lockdep enabled we get: ============================================= [ INFO: possible recursive locking detected ] 3.4.4-Cavium-Octeon+ #313 Not tainted --------------------------------------------- kworker/u:1/36 is trying to acquire lock: (&bus->mdio_lock){+.+...}, at: [] mdio_mux_read+0x38/0xa0 but task is already holding lock: (&bus->mdio_lock){+.+...}, at: [] mdiobus_read+0x44/0x88 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&bus->mdio_lock); lock(&bus->mdio_lock); *** DEADLOCK *** May be due to missing lock nesting notation . . . This is a false positive, since we are indeed using 'nested' locking, we need to use mutex_lock_nested(). Now in theory we can stack multiple MDIO multiplexers, but that would require passing the nesting level (which is difficult to know) to mutex_lock_nested(). Instead we assume the simple case of a single level of nesting. Since these are only warning messages, it isn't so important to solve the general case. Signed-off-by: David Daney Signed-off-by: David S. Miller --- drivers/net/phy/mdio-mux.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/net/phy/mdio-mux.c b/drivers/net/phy/mdio-mux.c index 39ea0674dcde..5c120189ec86 100644 --- a/drivers/net/phy/mdio-mux.c +++ b/drivers/net/phy/mdio-mux.c @@ -46,7 +46,13 @@ static int mdio_mux_read(struct mii_bus *bus, int phy_id, int regnum) struct mdio_mux_parent_bus *pb = cb->parent; int r; - mutex_lock(&pb->mii_bus->mdio_lock); + /* In theory multiple mdio_mux could be stacked, thus creating + * more than a single level of nesting. But in practice, + * SINGLE_DEPTH_NESTING will cover the vast majority of use + * cases. We use it, instead of trying to handle the general + * case. + */ + mutex_lock_nested(&pb->mii_bus->mdio_lock, SINGLE_DEPTH_NESTING); r = pb->switch_fn(pb->current_child, cb->bus_number, pb->switch_data); if (r) goto out; @@ -71,7 +77,7 @@ static int mdio_mux_write(struct mii_bus *bus, int phy_id, int r; - mutex_lock(&pb->mii_bus->mdio_lock); + mutex_lock_nested(&pb->mii_bus->mdio_lock, SINGLE_DEPTH_NESTING); r = pb->switch_fn(pb->current_child, cb->bus_number, pb->switch_data); if (r) goto out; -- cgit v1.2.3 From 6fecd35d4cd79fc75e8290abb86734c18500d2a2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Thu, 5 Jul 2012 01:13:33 +0000 Subject: net: qmi_wwan: add ZTE MF60 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Adding a device with limited QMI support. It does not support normal QMI_WDS commands for connection management. Instead, sending a QMI_CTL SET_INSTANCE_ID command is required to enable the network interface: 01 0f 00 00 00 00 00 00 20 00 04 00 01 01 00 00 A number of QMI_DMS and QMI_NAS commands are also supported for optional device management. Signed-off-by: Bjørn Mork Signed-off-by: David S. Miller --- drivers/net/usb/qmi_wwan.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'drivers') diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index b01960fcfbc9..a051cedd64bd 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -346,6 +346,15 @@ static const struct driver_info qmi_wwan_force_int1 = { .data = BIT(1), /* interface whitelist bitmap */ }; +static const struct driver_info qmi_wwan_force_int2 = { + .description = "Qualcomm WWAN/QMI device", + .flags = FLAG_WWAN, + .bind = qmi_wwan_bind_shared, + .unbind = qmi_wwan_unbind_shared, + .manage_power = qmi_wwan_manage_power, + .data = BIT(2), /* interface whitelist bitmap */ +}; + static const struct driver_info qmi_wwan_force_int3 = { .description = "Qualcomm WWAN/QMI device", .flags = FLAG_WWAN, @@ -498,6 +507,15 @@ static const struct usb_device_id products[] = { .bInterfaceProtocol = 0xff, .driver_info = (unsigned long)&qmi_wwan_force_int4, }, + { /* ZTE MF60 */ + .match_flags = USB_DEVICE_ID_MATCH_DEVICE | USB_DEVICE_ID_MATCH_INT_INFO, + .idVendor = 0x19d2, + .idProduct = 0x1402, + .bInterfaceClass = 0xff, + .bInterfaceSubClass = 0xff, + .bInterfaceProtocol = 0xff, + .driver_info = (unsigned long)&qmi_wwan_force_int2, + }, { /* Sierra Wireless MC77xx in QMI mode */ .match_flags = USB_DEVICE_ID_MATCH_DEVICE | USB_DEVICE_ID_MATCH_INT_INFO, .idVendor = 0x1199, -- cgit v1.2.3 From 054581e6c1eb314d54d4747fba545e9802be29da Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Thu, 5 Jul 2012 14:21:55 +0000 Subject: cnic: Don't use netdev->base_addr commit c0357e975afdbbedab5c662d19bef865f02adc17 bnx2: stop using net_device.{base_addr, irq}. removed netdev->base_addr so we need to update cnic to get the MMIO base address from pci_resource_start(). Otherwise, mmap of the uio device will fail. Signed-off-by: Michael Chan Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/cnic.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/broadcom/cnic.c b/drivers/net/ethernet/broadcom/cnic.c index c95e7b5e2b85..3c95065e0def 100644 --- a/drivers/net/ethernet/broadcom/cnic.c +++ b/drivers/net/ethernet/broadcom/cnic.c @@ -1053,12 +1053,13 @@ static int cnic_init_uio(struct cnic_dev *dev) uinfo = &udev->cnic_uinfo; - uinfo->mem[0].addr = dev->netdev->base_addr; + uinfo->mem[0].addr = pci_resource_start(dev->pcidev, 0); uinfo->mem[0].internal_addr = dev->regview; - uinfo->mem[0].size = dev->netdev->mem_end - dev->netdev->mem_start; uinfo->mem[0].memtype = UIO_MEM_PHYS; if (test_bit(CNIC_F_BNX2_CLASS, &dev->flags)) { + uinfo->mem[0].size = MB_GET_CID_ADDR(TX_TSS_CID + + TX_MAX_TSS_RINGS + 1); uinfo->mem[1].addr = (unsigned long) cp->status_blk.gen & PAGE_MASK; if (cp->ethdev->drv_state & CNIC_DRV_STATE_USING_MSIX) @@ -1068,6 +1069,8 @@ static int cnic_init_uio(struct cnic_dev *dev) uinfo->name = "bnx2_cnic"; } else if (test_bit(CNIC_F_BNX2X_CLASS, &dev->flags)) { + uinfo->mem[0].size = pci_resource_len(dev->pcidev, 0); + uinfo->mem[1].addr = (unsigned long) cp->bnx2x_def_status_blk & PAGE_MASK; uinfo->mem[1].size = sizeof(*cp->bnx2x_def_status_blk); -- cgit v1.2.3 From b3190466b0b6d336eddd10319c64a3ce3029b3ff Mon Sep 17 00:00:00 2001 From: Bing Zhao Date: Tue, 3 Jul 2012 15:53:13 -0700 Subject: mwifiex: fix Coverity SCAN CID 709078: Resource leak (RESOURCE_LEAK) > *. CID 709078: Resource leak (RESOURCE_LEAK) > - drivers/net/wireless/mwifiex/cfg80211.c, line: 935 > Assigning: "bss_cfg" = storage returned from "kzalloc(132UL, 208U)" > - but was not free > drivers/net/wireless/mwifiex/cfg80211.c:935 Signed-off-by: Bing Zhao Signed-off-by: John W. Linville --- drivers/net/wireless/mwifiex/cfg80211.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers') diff --git a/drivers/net/wireless/mwifiex/cfg80211.c b/drivers/net/wireless/mwifiex/cfg80211.c index ce61b6fae1c9..5c7fd185373c 100644 --- a/drivers/net/wireless/mwifiex/cfg80211.c +++ b/drivers/net/wireless/mwifiex/cfg80211.c @@ -958,6 +958,7 @@ static int mwifiex_cfg80211_start_ap(struct wiphy *wiphy, case NL80211_HIDDEN_SSID_ZERO_CONTENTS: /* firmware doesn't support this type of hidden SSID */ default: + kfree(bss_cfg); return -EINVAL; } -- cgit v1.2.3 From efd821182cec8c92babef6e00a95066d3252fda4 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 4 Jul 2012 13:10:02 +0200 Subject: rt2x00usb: fix indexes ordering on RX queue kick On rt2x00_dmastart() we increase index specified by Q_INDEX and on rt2x00_dmadone() we increase index specified by Q_INDEX_DONE. So entries between Q_INDEX_DONE and Q_INDEX are those we currently process in the hardware. Entries between Q_INDEX and Q_INDEX_DONE are those we can submit to the hardware. According to that fix rt2x00usb_kick_queue(), as we need to submit RX entries that are not processed by the hardware. It worked before only for empty queue, otherwise was broken. Note that for TX queues indexes ordering are ok. We need to kick entries that have filled skb, but was not submitted to the hardware, i.e. started from Q_INDEX_DONE and have ENTRY_DATA_PENDING bit set. From practical standpoint this fixes RX queue stall, usually reproducible in AP mode, like for example reported here: https://bugzilla.redhat.com/show_bug.cgi?id=828824 Reported-and-tested-by: Franco Miceli Reported-and-tested-by: Tom Horsley Cc: stable@vger.kernel.org Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/rt2x00/rt2x00usb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/net/wireless/rt2x00/rt2x00usb.c b/drivers/net/wireless/rt2x00/rt2x00usb.c index d357d1ed92f6..74ecc33fdd90 100644 --- a/drivers/net/wireless/rt2x00/rt2x00usb.c +++ b/drivers/net/wireless/rt2x00/rt2x00usb.c @@ -436,8 +436,8 @@ void rt2x00usb_kick_queue(struct data_queue *queue) case QID_RX: if (!rt2x00queue_full(queue)) rt2x00queue_for_each_entry(queue, - Q_INDEX_DONE, Q_INDEX, + Q_INDEX_DONE, NULL, rt2x00usb_kick_rx_entry); break; -- cgit v1.2.3 From c2ca7d92ed4bbd779516beb6eb226e19f7f7ab0f Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 4 Jul 2012 13:20:20 +0200 Subject: iwlegacy: always monitor for stuck queues This is iwlegacy version of: commit 342bbf3fee2fa9a18147e74b2e3c4229a4564912 Author: Johannes Berg Date: Sun Mar 4 08:50:46 2012 -0800 iwlwifi: always monitor for stuck queues If we only monitor while associated, the following can happen: - we're associated, and the queue stuck check runs, setting the queue "touch" time to X - we disassociate, stopping the monitoring, which leaves the time set to X - almost 2s later, we associate, and enqueue a frame - before the frame is transmitted, we monitor for stuck queues, and find the time set to X, although it is now later than X + 2000ms, so we decide that the queue is stuck and erroneously restart the device Cc: stable@vger.kernel.org Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/iwlegacy/common.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) (limited to 'drivers') diff --git a/drivers/net/wireless/iwlegacy/common.c b/drivers/net/wireless/iwlegacy/common.c index cbf2dc18341f..5d4807c2b56d 100644 --- a/drivers/net/wireless/iwlegacy/common.c +++ b/drivers/net/wireless/iwlegacy/common.c @@ -4767,14 +4767,12 @@ il_bg_watchdog(unsigned long data) return; /* monitor and check for other stuck queues */ - if (il_is_any_associated(il)) { - for (cnt = 0; cnt < il->hw_params.max_txq_num; cnt++) { - /* skip as we already checked the command queue */ - if (cnt == il->cmd_queue) - continue; - if (il_check_stuck_queue(il, cnt)) - return; - } + for (cnt = 0; cnt < il->hw_params.max_txq_num; cnt++) { + /* skip as we already checked the command queue */ + if (cnt == il->cmd_queue) + continue; + if (il_check_stuck_queue(il, cnt)) + return; } mod_timer(&il->watchdog, -- cgit v1.2.3 From b48d96652626b315229b1b82c6270eead6a77a6d Mon Sep 17 00:00:00 2001 From: Emmanuel Grumbach Date: Wed, 4 Jul 2012 13:59:08 +0200 Subject: iwlegacy: don't mess up the SCD when removing a key When we remove a key, we put a key index which was supposed to tell the fw that we are actually removing the key. But instead the fw took that index as a valid index and messed up the SRAM of the device. This memory corruption on the device mangled the data of the SCD. The impact on the user is that SCD queue 2 got stuck after having removed keys. Reported-by: Paul Bolle Cc: stable@vger.kernel.org Signed-off-by: Emmanuel Grumbach Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/iwlegacy/4965-mac.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/net/wireless/iwlegacy/4965-mac.c b/drivers/net/wireless/iwlegacy/4965-mac.c index 509301a5e7e2..ff5d689e13f3 100644 --- a/drivers/net/wireless/iwlegacy/4965-mac.c +++ b/drivers/net/wireless/iwlegacy/4965-mac.c @@ -3405,7 +3405,7 @@ il4965_remove_dynamic_key(struct il_priv *il, return 0; } - if (il->stations[sta_id].sta.key.key_offset == WEP_INVALID_OFFSET) { + if (il->stations[sta_id].sta.key.key_flags & STA_KEY_FLG_INVALID) { IL_WARN("Removing wrong key %d 0x%x\n", keyconf->keyidx, key_flags); spin_unlock_irqrestore(&il->sta_lock, flags); @@ -3420,7 +3420,7 @@ il4965_remove_dynamic_key(struct il_priv *il, memset(&il->stations[sta_id].sta.key, 0, sizeof(struct il4965_keyinfo)); il->stations[sta_id].sta.key.key_flags = STA_KEY_FLG_NO_ENC | STA_KEY_FLG_INVALID; - il->stations[sta_id].sta.key.key_offset = WEP_INVALID_OFFSET; + il->stations[sta_id].sta.key.key_offset = keyconf->hw_key_idx; il->stations[sta_id].sta.sta.modify_mask = STA_MODIFY_KEY_MASK; il->stations[sta_id].sta.mode = STA_CONTROL_MODIFY_MSK; -- cgit v1.2.3 From 8e83989106562326bfd6aaf92174fe138efd026b Mon Sep 17 00:00:00 2001 From: Deepak Sikri Date: Sun, 8 Jul 2012 21:14:45 +0000 Subject: stmmac: Fix for nfs hang on multiple reboot It was observed that during multiple reboots nfs hangs. The status of receive descriptors shows that all the descriptors were in control of CPU, and none were assigned to DMA. Also the DMA status register confirmed that the Rx buffer is unavailable. This patch adds the fix for the same by adding the memory barriers to ascertain that the all instructions before enabling the Rx or Tx DMA are completed which involves the proper setting of the ownership bit in DMA descriptors. Signed-off-by: Deepak Sikri Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers') diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 51b3b68528ee..ea3003edde18 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -1212,6 +1212,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) priv->hw->desc->prepare_tx_desc(desc, 0, len, csum_insertion); wmb(); priv->hw->desc->set_tx_owner(desc); + wmb(); } /* Interrupt on completition only for the latest segment */ @@ -1227,6 +1228,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) /* To avoid raise condition */ priv->hw->desc->set_tx_owner(first); + wmb(); priv->cur_tx++; @@ -1290,6 +1292,7 @@ static inline void stmmac_rx_refill(struct stmmac_priv *priv) } wmb(); priv->hw->desc->set_rx_owner(p + entry); + wmb(); } } -- cgit v1.2.3 From 684901a6df1fb91fc9a2bdb89ffbebb241428d78 Mon Sep 17 00:00:00 2001 From: Deepak Sikri Date: Sun, 8 Jul 2012 21:14:46 +0000 Subject: stmmac: Fix for higher mtu size handling For the higher mtu sizes requiring the buffer size greater than 8192, the buffers are sent or received using multiple dma descriptors/ same descriptor with option of multi buffer handling. It was observed during tests that the driver was missing on data packets during the normal ping operations if the data buffers being used catered to jumbo frame handling. The memory barrriers are added in between preparation of dma descriptors in the jumbo frame handling path to ensure all instructions before enabling the dma are complete. Signed-off-by: Deepak Sikri Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/ring_mode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c index fb8377da1687..4b785e10f2ed 100644 --- a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c +++ b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c @@ -51,7 +51,7 @@ static unsigned int stmmac_jumbo_frm(void *p, struct sk_buff *skb, int csum) desc->des3 = desc->des2 + BUF_SIZE_4KiB; priv->hw->desc->prepare_tx_desc(desc, 1, bmax, csum); - + wmb(); entry = (++priv->cur_tx) % txsize; desc = priv->dma_tx + entry; @@ -59,6 +59,7 @@ static unsigned int stmmac_jumbo_frm(void *p, struct sk_buff *skb, int csum) len, DMA_TO_DEVICE); desc->des3 = desc->des2 + BUF_SIZE_4KiB; priv->hw->desc->prepare_tx_desc(desc, 0, len, csum); + wmb(); priv->hw->desc->set_tx_owner(desc); priv->tx_skbuff[entry] = NULL; } else { -- cgit v1.2.3 From a64d49c3dd504b685f9742a2f3dcb11fb8e4345f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 9 Jul 2012 10:51:45 +0000 Subject: bonding: Manage /proc/net/bonding/ entries from the netdev events It was recently reported that moving a bonding device between network namespaces causes warnings from /proc. It turns out after the move we were trying to add and to remove the /proc/net/bonding entries from the wrong network namespace. Move the bonding /proc registration code into the NETDEV_REGISTER and NETDEV_UNREGISTER events where the proc registration and unregistration will always happen at the right time. Signed-off-by: "Eric W. Biederman" Signed-off-by: David S. Miller --- drivers/net/bonding/bond_main.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'drivers') diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index b9c2ae62166d..2ee76993f052 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -3227,6 +3227,12 @@ static int bond_master_netdev_event(unsigned long event, switch (event) { case NETDEV_CHANGENAME: return bond_event_changename(event_bond); + case NETDEV_UNREGISTER: + bond_remove_proc_entry(event_bond); + break; + case NETDEV_REGISTER: + bond_create_proc_entry(event_bond); + break; default: break; } @@ -4411,8 +4417,6 @@ static void bond_uninit(struct net_device *bond_dev) bond_work_cancel_all(bond); - bond_remove_proc_entry(bond); - bond_debug_unregister(bond); __hw_addr_flush(&bond->mc_list); @@ -4814,7 +4818,6 @@ static int bond_init(struct net_device *bond_dev) bond_set_lockdep_class(bond_dev); - bond_create_proc_entry(bond); list_add_tail(&bond->bond_list, &bn->dev_list); bond_prepare_sysfs_group(bond); -- cgit v1.2.3 From 96ca7ffe748bf91f851e6aa4479aa11c8b1122ba Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 9 Jul 2012 10:52:43 +0000 Subject: bonding: debugfs and network namespaces are incompatible The bonding debugfs support has been broken in the presence of network namespaces since it has been added. The debugfs support does not handle multiple bonding devices with the same name in different network namespaces. I haven't had any bug reports, and I'm not interested in getting any. Disable the debugfs support when network namespaces are enabled. Signed-off-by: "Eric W. Biederman" Signed-off-by: David S. Miller --- drivers/net/bonding/bond_debugfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/net/bonding/bond_debugfs.c b/drivers/net/bonding/bond_debugfs.c index 3680aa251dea..2cf084eb9d52 100644 --- a/drivers/net/bonding/bond_debugfs.c +++ b/drivers/net/bonding/bond_debugfs.c @@ -6,7 +6,7 @@ #include "bonding.h" #include "bond_alb.h" -#ifdef CONFIG_DEBUG_FS +#if defined(CONFIG_DEBUG_FS) && !defined(CONFIG_NET_NS) #include #include -- cgit v1.2.3 From 1b9faf5e66bae8428a4ccdc1447d5399d1014581 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sun, 8 Jul 2012 01:37:38 +0000 Subject: drivers/isdn/mISDN/stack.c: remove invalid reference to list iterator variable If list_for_each_entry, etc complete a traversal of the list, the iterator variable ends up pointing to an address at an offset from the list head, and not a meaningful structure. Thus this value should not be used after the end of the iterator. The dereferences are just deleted from the debugging statement. This problem was found using Coccinelle (http://coccinelle.lip6.fr/). Signed-off-by: Julia Lawall Signed-off-by: David S. Miller --- drivers/isdn/mISDN/stack.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/isdn/mISDN/stack.c b/drivers/isdn/mISDN/stack.c index 1a0ae4445ff2..5f21f629b7ae 100644 --- a/drivers/isdn/mISDN/stack.c +++ b/drivers/isdn/mISDN/stack.c @@ -135,8 +135,8 @@ send_layer2(struct mISDNstack *st, struct sk_buff *skb) skb = NULL; else if (*debug & DEBUG_SEND_ERR) printk(KERN_DEBUG - "%s ch%d mgr prim(%x) addr(%x) err %d\n", - __func__, ch->nr, hh->prim, ch->addr, ret); + "%s mgr prim(%x) err %d\n", + __func__, hh->prim, ret); } out: mutex_unlock(&st->lmutex); -- cgit v1.2.3 From 022f09784b85396b4ceba954ce28e50de4882281 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sun, 8 Jul 2012 01:37:43 +0000 Subject: drivers/net/ethernet/broadcom/cnic.c: remove invalid reference to list iterator variable If list_for_each_entry, etc complete a traversal of the list, the iterator variable ends up pointing to an address at an offset from the list head, and not a meaningful structure. Thus this value should not be used after the end of the iterator. There does not seem to be a meaningful value to provide to netdev_warn. Replace with pr_warn, since pr_err is used elsewhere. This problem was found using Coccinelle (http://coccinelle.lip6.fr/). Signed-off-by: Julia Lawall Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/cnic.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/broadcom/cnic.c b/drivers/net/ethernet/broadcom/cnic.c index 3c95065e0def..2c89d17cbb29 100644 --- a/drivers/net/ethernet/broadcom/cnic.c +++ b/drivers/net/ethernet/broadcom/cnic.c @@ -534,7 +534,8 @@ int cnic_unregister_driver(int ulp_type) } if (atomic_read(&ulp_ops->ref_count) != 0) - netdev_warn(dev->netdev, "Failed waiting for ref count to go to zero\n"); + pr_warn("%s: Failed waiting for ref count to go to zero\n", + __func__); return 0; out_unlock: -- cgit v1.2.3 From 313b037cf054ec908de92fb4c085403ffd7420d4 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 5 Jul 2012 11:45:13 +0000 Subject: gianfar: fix potential sk_wmem_alloc imbalance commit db83d136d7f753 (gianfar: Fix missing sock reference when processing TX time stamps) added a potential sk_wmem_alloc imbalance If the new skb has a different truesize than old one, we can get a negative sk_wmem_alloc once new skb is orphaned at TX completion. Now we no longer early orphan skbs in dev_hard_start_xmit(), this probably can lead to fatal bugs. Signed-off-by: Eric Dumazet Tested-by: Paul Gortmaker Cc: Manfred Rudigier Cc: Claudiu Manoil Cc: Jiajun Wu Cc: Andy Fleming Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/gianfar.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/freescale/gianfar.c b/drivers/net/ethernet/freescale/gianfar.c index f2db8fca46a1..ab1d80ff0791 100644 --- a/drivers/net/ethernet/freescale/gianfar.c +++ b/drivers/net/ethernet/freescale/gianfar.c @@ -2063,10 +2063,9 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev) return NETDEV_TX_OK; } - /* Steal sock reference for processing TX time stamps */ - swap(skb_new->sk, skb->sk); - swap(skb_new->destructor, skb->destructor); - kfree_skb(skb); + if (skb->sk) + skb_set_owner_w(skb_new, skb->sk); + consume_skb(skb); skb = skb_new; } -- cgit v1.2.3 From b28ba72665356438e3a6e3be365c3c3071496840 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 10 Jul 2012 10:03:41 +0000 Subject: IPoIB: fix skb truesize underestimatiom Or Gerlitz reported triggering of WARN_ON_ONCE(delta < len); in skb_try_coalesce() This warning tracks drivers that incorrectly set skb->truesize IPoIB indeed allocates a full page to store a fragment, but only accounts in skb->truesize the used part of the page (frame length) This patch fixes skb truesize underestimation, and also fixes a performance issue, because RX skbs have not enough tailroom to allow IP and TCP stacks to pull their header in skb linear part without an expensive call to pskb_expand_head() Signed-off-by: Eric Dumazet Reported-by: Or Gerlitz Cc: Erez Shitrit Cc: Shlomo Pongartz Cc: Roland Dreier Signed-off-by: David S. Miller --- drivers/infiniband/ulp/ipoib/ipoib_ib.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'drivers') diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 5c1bc995e560..f10221f40803 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -123,7 +123,7 @@ static void ipoib_ud_skb_put_frags(struct ipoib_dev_priv *priv, skb_frag_size_set(frag, size); skb->data_len += size; - skb->truesize += size; + skb->truesize += PAGE_SIZE; } else skb_put(skb, length); @@ -156,14 +156,18 @@ static struct sk_buff *ipoib_alloc_rx_skb(struct net_device *dev, int id) struct ipoib_dev_priv *priv = netdev_priv(dev); struct sk_buff *skb; int buf_size; + int tailroom; u64 *mapping; - if (ipoib_ud_need_sg(priv->max_ib_mtu)) + if (ipoib_ud_need_sg(priv->max_ib_mtu)) { buf_size = IPOIB_UD_HEAD_SIZE; - else + tailroom = 128; /* reserve some tailroom for IP/TCP headers */ + } else { buf_size = IPOIB_UD_BUF_SIZE(priv->max_ib_mtu); + tailroom = 0; + } - skb = dev_alloc_skb(buf_size + 4); + skb = dev_alloc_skb(buf_size + tailroom + 4); if (unlikely(!skb)) return NULL; -- cgit v1.2.3 From c1f5163de417dab01fa9daaf09a74bbb19303f3c Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Tue, 10 Jul 2012 10:04:40 +0000 Subject: bnx2: Fix bug in bnx2_free_tx_skbs(). In rare cases, bnx2x_free_tx_skbs() can unmap the wrong DMA address when it gets to the last entry of the tx ring. We were not using the proper macro to skip the last entry when advancing the tx index. Reported-by: Zongyun Lai Reviewed-by: Jeffrey Huang Signed-off-by: Michael Chan Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/broadcom/bnx2.c b/drivers/net/ethernet/broadcom/bnx2.c index ac7b74488531..1fa4927a45b1 100644 --- a/drivers/net/ethernet/broadcom/bnx2.c +++ b/drivers/net/ethernet/broadcom/bnx2.c @@ -5372,7 +5372,7 @@ bnx2_free_tx_skbs(struct bnx2 *bp) int k, last; if (skb == NULL) { - j++; + j = NEXT_TX_BD(j); continue; } @@ -5384,8 +5384,8 @@ bnx2_free_tx_skbs(struct bnx2 *bp) tx_buf->skb = NULL; last = tx_buf->nr_frags; - j++; - for (k = 0; k < last; k++, j++) { + j = NEXT_TX_BD(j); + for (k = 0; k < last; k++, j = NEXT_TX_BD(j)) { tx_buf = &txr->tx_buf_ring[TX_RING_IDX(j)]; dma_unmap_page(&bp->pdev->dev, dma_unmap_addr(tx_buf, mapping), -- cgit v1.2.3 From d0efa8f23a644f7cb7d1f8e78dd9a223efa412a3 Mon Sep 17 00:00:00 2001 From: Tushar Dave Date: Thu, 12 Jul 2012 08:56:56 +0000 Subject: e1000e: Correct link check logic for 82571 serdes SYNCH bit and IV bit of RXCW register are sticky. Before examining these bits, RXCW should be read twice to filter out one-time false events and have correct values for these bits. Incorrect values of these bits in link check logic can cause weird link stability issues if auto-negotiation fails. CC: stable [2.6.38+] Reported-by: Dean Nelson Signed-off-by: Tushar Dave Reviewed-by: Bruce Allan Tested-by: Jeff Pieper Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/e1000e/82571.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers') diff --git a/drivers/net/ethernet/intel/e1000e/82571.c b/drivers/net/ethernet/intel/e1000e/82571.c index 36db4df09aed..1f063dcd8f85 100644 --- a/drivers/net/ethernet/intel/e1000e/82571.c +++ b/drivers/net/ethernet/intel/e1000e/82571.c @@ -1572,6 +1572,9 @@ static s32 e1000_check_for_serdes_link_82571(struct e1000_hw *hw) ctrl = er32(CTRL); status = er32(STATUS); rxcw = er32(RXCW); + /* SYNCH bit and IV bit are sticky */ + udelay(10); + rxcw = er32(RXCW); if ((rxcw & E1000_RXCW_SYNCH) && !(rxcw & E1000_RXCW_IV)) { -- cgit v1.2.3 From a52359b56c29f55aaadf1dab80a0e1043b982676 Mon Sep 17 00:00:00 2001 From: Bruce Allan Date: Sat, 14 Jul 2012 04:23:58 +0000 Subject: e1000e: fix test for PHY being accessible on 82577/8/9 and I217 Occasionally, the PHY can be initially inaccessible when the first read of a PHY register, e.g. PHY_ID1, happens (signified by the returned value 0xFFFF) but subsequent accesses of the PHY work as expected. Add a retry counter similar to how it is done in the generic e1000_get_phy_id(). Also, when the PHY is completely inaccessible (i.e. when subsequent reads of the PHY_IDx registers returns all F's) and the MDIO access mode must be set to slow before attempting to read the PHY ID again, the functions that do these latter two actions expect the SW/FW/HW semaphore is not already set so the semaphore must be released before and re-acquired after calling them otherwise there is an unnecessarily inordinate amount of delay during device initialization. Reported-by: Heikki Krogerus Signed-off-by: Bruce Allan Tested-by: Heikki Krogerus Tested-by: Aaron Brown Signed-off-by: Jeff Kirsher --- drivers/net/ethernet/intel/e1000e/ich8lan.c | 42 ++++++++++++++++++++++------- 1 file changed, 32 insertions(+), 10 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/intel/e1000e/ich8lan.c b/drivers/net/ethernet/intel/e1000e/ich8lan.c index 238ab2f8a5e7..e3a7b07df629 100644 --- a/drivers/net/ethernet/intel/e1000e/ich8lan.c +++ b/drivers/net/ethernet/intel/e1000e/ich8lan.c @@ -325,24 +325,46 @@ static inline void __ew32flash(struct e1000_hw *hw, unsigned long reg, u32 val) **/ static bool e1000_phy_is_accessible_pchlan(struct e1000_hw *hw) { - u16 phy_reg; - u32 phy_id; + u16 phy_reg = 0; + u32 phy_id = 0; + s32 ret_val; + u16 retry_count; + + for (retry_count = 0; retry_count < 2; retry_count++) { + ret_val = e1e_rphy_locked(hw, PHY_ID1, &phy_reg); + if (ret_val || (phy_reg == 0xFFFF)) + continue; + phy_id = (u32)(phy_reg << 16); - e1e_rphy_locked(hw, PHY_ID1, &phy_reg); - phy_id = (u32)(phy_reg << 16); - e1e_rphy_locked(hw, PHY_ID2, &phy_reg); - phy_id |= (u32)(phy_reg & PHY_REVISION_MASK); + ret_val = e1e_rphy_locked(hw, PHY_ID2, &phy_reg); + if (ret_val || (phy_reg == 0xFFFF)) { + phy_id = 0; + continue; + } + phy_id |= (u32)(phy_reg & PHY_REVISION_MASK); + break; + } if (hw->phy.id) { if (hw->phy.id == phy_id) return true; - } else { - if ((phy_id != 0) && (phy_id != PHY_REVISION_MASK)) - hw->phy.id = phy_id; + } else if (phy_id) { + hw->phy.id = phy_id; + hw->phy.revision = (u32)(phy_reg & ~PHY_REVISION_MASK); return true; } - return false; + /* + * In case the PHY needs to be in mdio slow mode, + * set slow mode and try to get the PHY id again. + */ + hw->phy.ops.release(hw); + ret_val = e1000_set_mdio_slow_mode_hv(hw); + if (!ret_val) + ret_val = e1000e_get_phy_id(hw); + hw->phy.ops.acquire(hw); + + return !ret_val; } /** -- cgit v1.2.3 From 936597631dd310e220544dc5c6075d924efd39b2 Mon Sep 17 00:00:00 2001 From: Narendra K Date: Mon, 16 Jul 2012 15:24:41 +0000 Subject: ixgbevf: Prevent RX/TX statistics getting reset to zero The commit 4197aa7bb81877ebb06e4f2cc1b5fea2da23a7bd implements 64 bit per ring statistics. But the driver resets the 'total_bytes' and 'total_packets' from RX and TX rings in the RX and TX interrupt handlers to zero. This results in statistics being lost and user space reporting RX and TX statistics as zero. This patch addresses the issue by preventing the resetting of RX and TX ring statistics to zero. Signed-off-by: Narendra K Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 12 ------------ 1 file changed, 12 deletions(-) (limited to 'drivers') diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index f69ec4288b10..8b304a43a22d 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -969,8 +969,6 @@ static irqreturn_t ixgbevf_msix_clean_tx(int irq, void *data) r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues); for (i = 0; i < q_vector->txr_count; i++) { tx_ring = &(adapter->tx_ring[r_idx]); - tx_ring->total_bytes = 0; - tx_ring->total_packets = 0; ixgbevf_clean_tx_irq(adapter, tx_ring); r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues, r_idx + 1); @@ -994,16 +992,6 @@ static irqreturn_t ixgbevf_msix_clean_rx(int irq, void *data) struct ixgbe_hw *hw = &adapter->hw; struct ixgbevf_ring *rx_ring; int r_idx; - int i; - - r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues); - for (i = 0; i < q_vector->rxr_count; i++) { - rx_ring = &(adapter->rx_ring[r_idx]); - rx_ring->total_bytes = 0; - rx_ring->total_packets = 0; - r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues, - r_idx + 1); - } if (!q_vector->rxr_count) return IRQ_HANDLED; -- cgit v1.2.3 From 10cc1bdd5ef65f60f570b594c4c066d763c128fb Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 16 Jul 2012 23:44:48 +0000 Subject: ixgbevf: Fix panic when loading driver This patch addresses a kernel panic seen when setting up the interface. Specifically we see a NULL pointer dereference on the Tx descriptor cleanup path when enabling interrupts. This change corrects that so it cannot occur. Signed-off-by: Alexander Duyck Acked-by: Greg Rose Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers') diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index 8b304a43a22d..41e32257a4e8 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -201,6 +201,9 @@ static bool ixgbevf_clean_tx_irq(struct ixgbevf_adapter *adapter, unsigned int i, eop, count = 0; unsigned int total_bytes = 0, total_packets = 0; + if (test_bit(__IXGBEVF_DOWN, &adapter->state)) + return true; + i = tx_ring->next_to_clean; eop = tx_ring->tx_buffer_info[i].next_to_watch; eop_desc = IXGBE_TX_DESC_ADV(*tx_ring, eop); -- cgit v1.2.3