Dear Russell On Thu, Apr 3, 2014 at 11:27 PM, Russell King - ARM Linux <linux@xxxxxxxxxxxxxxxx> wrote: > On Wed, Apr 02, 2014 at 11:21:45AM +0200, Arnd Bergmann wrote: >> - As David Laight pointed out earlier, you must also ensure that >> you don't have too much /data/ pending in the descriptor ring >> when you stop the queue. For a 10mbit connection, you have already >> tested (as we discussed on IRC) that 64 descriptors with 1500 byte >> frames gives you a 68ms round-trip ping time, which is too much. >> Conversely, on 1gbit, having only 64 descriptors actually seems >> a little low, and you may be able to get better throughput if >> you extend the ring to e.g. 512 descriptors. > > You don't manage that by stopping the queue - there's separate interfaces > where you report how many bytes you've queued (netdev_sent_queue()) and > how many bytes/packets you've sent (netdev_tx_completed_queue()). This > allows the netdev schedulers to limit how much data is held in the queue, > preserving interactivity while allowing the advantages of larger rings. My god, it's awesome. The latency can be solved via adding netdev_sent_queue in xmit, and netdev_completed_queue in reclaim. In the experiment, iperf -P 3 could get 930M, and ping could get response within 0.4 ms in the meantime. Is that mean the timer -> reclaim should be removed at all? The background are. 1. No xmit_complete interrupt. 2. Only xmit call reclaim used buffer can achieve best throughput. Adding timer in case no next xmit for reclaim. > >> > + phys = dma_map_single(&ndev->dev, skb->data, skb->len, DMA_TO_DEVICE); >> > + if (dma_mapping_error(&ndev->dev, phys)) { >> > + dev_kfree_skb(skb); >> > + return NETDEV_TX_OK; >> > + } >> > + >> > + priv->tx_skb[tx_head] = skb; >> > + priv->tx_phys[tx_head] = phys; >> > + desc->send_addr = cpu_to_be32(phys); >> > + desc->send_size = cpu_to_be16(skb->len); >> > + desc->cfg = cpu_to_be32(DESC_DEF_CFG); >> > + phys = priv->tx_desc_dma + tx_head * sizeof(struct tx_desc); >> > + desc->wb_addr = cpu_to_be32(phys); >> >> One detail: since you don't have cache-coherent DMA, "desc" will >> reside in uncached memory, so you try to minimize the number of accesses. >> It's probably faster if you build the descriptor on the stack and >> then atomically copy it over, rather than assigning each member at >> a time. > > DMA coherent memory is write combining, so multiple writes will be > coalesced. This also means that barriers may be required to ensure the > descriptors are pushed out in a timely manner if something like writel() > is not used in the transmit-triggering path. > Currently writel is used in xmit. And regmap_write -> writel is used in poll. Thanks -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html