^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) .. SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) =====================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) Softnet Driver Issues
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) =====================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) Transmit path guidelines:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) 1) The ndo_start_xmit method must not return NETDEV_TX_BUSY under
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) any normal circumstances. It is considered a hard error unless
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) there is no way your device can tell ahead of time when it's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) transmit function will become busy.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) Instead it must maintain the queue properly. For example,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) for a driver implementing scatter-gather this means::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) static netdev_tx_t drv_hard_start_xmit(struct sk_buff *skb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) struct net_device *dev)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) struct drv *dp = netdev_priv(dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) lock_tx(dp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) ...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) /* This is a hard error log it. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) if (TX_BUFFS_AVAIL(dp) <= (skb_shinfo(skb)->nr_frags + 1)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) netif_stop_queue(dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) unlock_tx(dp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) printk(KERN_ERR PFX "%s: BUG! Tx Ring full when queue awake!\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) dev->name);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) return NETDEV_TX_BUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) ... queue packet to card ...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) ... update tx consumer index ...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) if (TX_BUFFS_AVAIL(dp) <= (MAX_SKB_FRAGS + 1))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) netif_stop_queue(dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) ...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) unlock_tx(dp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) ...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) return NETDEV_TX_OK;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) And then at the end of your TX reclamation event handling::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) if (netif_queue_stopped(dp->dev) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) TX_BUFFS_AVAIL(dp) > (MAX_SKB_FRAGS + 1))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) netif_wake_queue(dp->dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) For a non-scatter-gather supporting card, the three tests simply become::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) /* This is a hard error log it. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) if (TX_BUFFS_AVAIL(dp) <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) and::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) if (TX_BUFFS_AVAIL(dp) == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) and::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) if (netif_queue_stopped(dp->dev) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) TX_BUFFS_AVAIL(dp) > 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) netif_wake_queue(dp->dev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) 2) An ndo_start_xmit method must not modify the shared parts of a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) cloned SKB.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) 3) Do not forget that once you return NETDEV_TX_OK from your
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) ndo_start_xmit method, it is your driver's responsibility to free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) up the SKB and in some finite amount of time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) For example, this means that it is not allowed for your TX
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) mitigation scheme to let TX packets "hang out" in the TX
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) ring unreclaimed forever if no new TX packets are sent.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) This error can deadlock sockets waiting for send buffer room
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) to be freed up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) If you return NETDEV_TX_BUSY from the ndo_start_xmit method, you
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) must not keep any reference to that SKB and you must not attempt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) to free it up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) Probing guidelines:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) 1) Any hardware layer address you obtain for your device should
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) be verified. For example, for ethernet check it with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) linux/etherdevice.h:is_valid_ether_addr()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) Close/stop guidelines:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) 1) After the ndo_stop routine has been called, the hardware must
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) not receive or transmit any data. All in flight packets must
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) be aborted. If necessary, poll or wait for completion of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) any reset commands.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) 2) The ndo_stop routine will be called by unregister_netdevice
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) if device is still UP.
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags