Hi Paolo, > On 5/4/25 4:55 PM, Lukasz Majewski wrote: > > + /* This does 16 byte alignment, exactly what we > > need. > > + * The packet length includes FCS, but we don't > > want to > > + * include that when passing upstream as it messes > > up > > + * bridging applications. > > + */ > > + skb = netdev_alloc_skb(pndev, pkt_len + > > NET_IP_ALIGN); > > + if (unlikely(!skb)) { > > + dev_dbg(&fep->pdev->dev, > > + "%s: Memory squeeze, dropping > > packet.\n", > > + pndev->name); > > + pndev->stats.rx_dropped++; > > + goto err_mem; > > + } else { > > + skb_reserve(skb, NET_IP_ALIGN); > > + skb_put(skb, pkt_len); /* Make room */ > > + skb_copy_to_linear_data(skb, data, > > pkt_len); > > + skb->protocol = eth_type_trans(skb, pndev); > > + napi_gro_receive(&fep->napi, skb); > > + } > > + > > + bdp->cbd_bufaddr = dma_map_single(&fep->pdev->dev, > > data, > > + bdp->cbd_datlen, > > + DMA_FROM_DEVICE); > > + if (unlikely(dma_mapping_error(&fep->pdev->dev, > > + bdp->cbd_bufaddr))) > > { > > + dev_err(&fep->pdev->dev, > > + "Failed to map descriptor rx > > buffer\n"); > > + pndev->stats.rx_errors++; > > + pndev->stats.rx_dropped++; > > + dev_kfree_skb_any(skb); > > + goto err_mem; > > + } > > This is doing the mapping and ev. dropping the skb _after_ pushing the > skb up the stack, you must attempt the mapping first. > > > +static void mtip_free_buffers(struct net_device *dev) > > +{ > > + struct mtip_ndev_priv *priv = netdev_priv(dev); > > + struct switch_enet_private *fep = priv->fep; > > + struct sk_buff *skb; > > + struct cbd_t *bdp; > > + int i; > > + > > + bdp = fep->rx_bd_base; > > + for (i = 0; i < RX_RING_SIZE; i++) { > > + skb = fep->rx_skbuff[i]; > > + > > + if (bdp->cbd_bufaddr) > > + dma_unmap_single(&fep->pdev->dev, > > bdp->cbd_bufaddr, > > + MTIP_SWITCH_RX_FRSIZE, > > + DMA_FROM_DEVICE); > > + if (skb) > > + dev_kfree_skb(skb); > > I suspect that on error paths mtip_free_buffers() can be invoked > multiple consecutive times with any successful allocation in between: > skb will be freed twice. Likely you need to clear fep->rx_skbuff[i] > here. I don't know what I shall say now.... really... > > Side note: this patch is way too big for a proper review: you need to > break it in multiple smaller ones, introducing the basic features > separately. > This code is a basic version of the driver as discussed on March with the community (Andrew). It provides the basic functionality - like separate ports support and then, if required, configures the IP block to perform L2 switching in HW. > Cheers, > > Paolo > Best regards, Lukasz Majewski -- DENX Software Engineering GmbH, Managing Director: Erika Unter HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lukma@denx.de