diff options
Diffstat (limited to 'Documentation/networking/driver.rst')
| -rw-r--r-- | Documentation/networking/driver.rst | 156 |
1 files changed, 93 insertions, 63 deletions
diff --git a/Documentation/networking/driver.rst b/Documentation/networking/driver.rst index 64f7236ff10b..4f5dfa9c022e 100644 --- a/Documentation/networking/driver.rst +++ b/Documentation/networking/driver.rst @@ -4,94 +4,124 @@ Softnet Driver Issues ===================== -Transmit path guidelines: +Probing guidelines +================== -1) The ndo_start_xmit method must not return NETDEV_TX_BUSY under - any normal circumstances. It is considered a hard error unless - there is no way your device can tell ahead of time when its - transmit function will become busy. +Address validation +------------------ - Instead it must maintain the queue properly. For example, - for a driver implementing scatter-gather this means:: +Any hardware layer address you obtain for your device should +be verified. For example, for ethernet check it with +linux/etherdevice.h:is_valid_ether_addr() + +Close/stop guidelines +===================== + +Quiescence +---------- + +After the ndo_stop routine has been called, the hardware must +not receive or transmit any data. All in flight packets must +be aborted. If necessary, poll or wait for completion of +any reset commands. + +Auto-close +---------- + +The ndo_stop routine will be called by unregister_netdevice +if device is still UP. + +Transmit path guidelines +======================== + +Stop queues in advance +---------------------- + +The ndo_start_xmit method must not return NETDEV_TX_BUSY under +any normal circumstances. It is considered a hard error unless +there is no way your device can tell ahead of time when its +transmit function will become busy. + +Instead it must maintain the queue properly. For example, +for a driver implementing scatter-gather this means: + +.. code-block:: c + + static u32 drv_tx_avail(struct drv_ring *dr) + { + u32 used = READ_ONCE(dr->prod) - READ_ONCE(dr->cons); + + return dr->tx_ring_size - (used & bp->tx_ring_mask); + } static netdev_tx_t drv_hard_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct drv *dp = netdev_priv(dev); + struct netdev_queue *txq; + struct drv_ring *dr; + int idx; - lock_tx(dp); - ... - /* This is a hard error log it. */ - if (TX_BUFFS_AVAIL(dp) <= (skb_shinfo(skb)->nr_frags + 1)) { + idx = skb_get_queue_mapping(skb); + dr = dp->tx_rings[idx]; + txq = netdev_get_tx_queue(dev, idx); + + //... + /* This should be a very rare race - log it. */ + if (drv_tx_avail(dr) <= skb_shinfo(skb)->nr_frags + 1) { netif_stop_queue(dev); - unlock_tx(dp); - printk(KERN_ERR PFX "%s: BUG! Tx Ring full when queue awake!\n", - dev->name); + netdev_warn(dev, "Tx Ring full when queue awake!\n"); return NETDEV_TX_BUSY; } - ... queue packet to card ... - ... update tx consumer index ... - - if (TX_BUFFS_AVAIL(dp) <= (MAX_SKB_FRAGS + 1)) - netif_stop_queue(dev); - - ... - unlock_tx(dp); - ... - return NETDEV_TX_OK; - } - - And then at the end of your TX reclamation event handling:: + //... queue packet to card ... - if (netif_queue_stopped(dp->dev) && - TX_BUFFS_AVAIL(dp) > (MAX_SKB_FRAGS + 1)) - netif_wake_queue(dp->dev); + netdev_tx_sent_queue(txq, skb->len); - For a non-scatter-gather supporting card, the three tests simply become:: + //... update tx producer index using WRITE_ONCE() ... - /* This is a hard error log it. */ - if (TX_BUFFS_AVAIL(dp) <= 0) + if (!netif_txq_maybe_stop(txq, drv_tx_avail(dr), + MAX_SKB_FRAGS + 1, 2 * MAX_SKB_FRAGS)) + dr->stats.stopped++; - and:: + //... + return NETDEV_TX_OK; + } - if (TX_BUFFS_AVAIL(dp) == 0) +And then at the end of your TX reclamation event handling: - and:: +.. code-block:: c - if (netif_queue_stopped(dp->dev) && - TX_BUFFS_AVAIL(dp) > 0) - netif_wake_queue(dp->dev); + //... update tx consumer index using WRITE_ONCE() ... -2) An ndo_start_xmit method must not modify the shared parts of a - cloned SKB. + netif_txq_completed_wake(txq, cmpl_pkts, cmpl_bytes, + drv_tx_avail(dr), 2 * MAX_SKB_FRAGS); -3) Do not forget that once you return NETDEV_TX_OK from your - ndo_start_xmit method, it is your driver's responsibility to free - up the SKB and in some finite amount of time. +Lockless queue stop / wake helper macros +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - For example, this means that it is not allowed for your TX - mitigation scheme to let TX packets "hang out" in the TX - ring unreclaimed forever if no new TX packets are sent. - This error can deadlock sockets waiting for send buffer room - to be freed up. +.. kernel-doc:: include/net/netdev_queues.h + :doc: Lockless queue stopping / waking helpers. - If you return NETDEV_TX_BUSY from the ndo_start_xmit method, you - must not keep any reference to that SKB and you must not attempt - to free it up. +No exclusive ownership +---------------------- -Probing guidelines: +An ndo_start_xmit method must not modify the shared parts of a +cloned SKB. -1) Any hardware layer address you obtain for your device should - be verified. For example, for ethernet check it with - linux/etherdevice.h:is_valid_ether_addr() +Timely completions +------------------ -Close/stop guidelines: +Do not forget that once you return NETDEV_TX_OK from your +ndo_start_xmit method, it is your driver's responsibility to free +up the SKB and in some finite amount of time. -1) After the ndo_stop routine has been called, the hardware must - not receive or transmit any data. All in flight packets must - be aborted. If necessary, poll or wait for completion of - any reset commands. +For example, this means that it is not allowed for your TX +mitigation scheme to let TX packets "hang out" in the TX +ring unreclaimed forever if no new TX packets are sent. +This error can deadlock sockets waiting for send buffer room +to be freed up. -2) The ndo_stop routine will be called by unregister_netdevice - if device is still UP. +If you return NETDEV_TX_BUSY from the ndo_start_xmit method, you +must not keep any reference to that SKB and you must not attempt +to free it up. |