From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jay Rolette Subject: Couple of PMD questions Date: Tue, 19 Apr 2016 15:16:37 -0500 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 To: DPDK Return-path: Received: from mail-vk0-f42.google.com (mail-vk0-f42.google.com [209.85.213.42]) by dpdk.org (Postfix) with ESMTP id DD3AA2C5A for ; Tue, 19 Apr 2016 22:16:37 +0200 (CEST) Received: by mail-vk0-f42.google.com with SMTP id t129so34055323vkg.2 for ; Tue, 19 Apr 2016 13:16:37 -0700 (PDT) List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" In ixgbe_dev_rx_init(), there is this bit of code: /* * Configure the RX buffer size in the BSIZEPACKET field of * the SRRCTL register of the queue. * The value is in 1 KB resolution. Valid values can be from * 1 KB to 16 KB. */ buf_size = (uint16_t)(rte_pktmbuf_data_room_size(rxq->mb_pool) - RTE_PKTMBUF_HEADROOM); srrctl |= ((buf_size >> IXGBE_SRRCTL_BSIZEPKT_SHIFT) & IXGBE_SRRCTL_BSIZEPKT_MASK); IXGBE_WRITE_REG(hw, IXGBE_SRRCTL(rxq->reg_idx), srrctl); buf_size = (uint16_t) ((srrctl & IXGBE_SRRCTL_BSIZEPKT_MASK) << IXGBE_SRRCTL_BSIZEPKT_SHIFT); /* It adds dual VLAN length for supporting dual VLAN */ if (dev->data->dev_conf.rxmode.max_rx_pkt_len + 2 * IXGBE_VLAN_TAG_SIZE > buf_size) dev->data->scattered_rx = 1; Couple of questions I have about it: 1) If the caller configured the MBUF memory pool data room size to be something other than a multiple of 1K (+ RTE_PKTMBUF_HEADROOM), then the driver ends up silently programming the NIC to use a smaller max RX size than the caller specified. Should the driver error out in that case instead of only "sort of" working? If not, it seems like it should be logging an error message. 2) Second question is about the "/* It adds dual VLAN length for supporting dual VLAN */" bit... As I understand it, dev_conf.rxmode.max_rx_pkt_len is supposed to be set to the max frame size you support (although it says it's only used if jumbo frame is enabled). That same value is generally used when calculating the size that mbuf elements should be created at. The description for the data_room_size parameter of rte_pktmbuf_pool_create(): "Size of data buffer in each mbuf, including RTE_PKTMBUF_HEADROOM." If I support a max frame size of 9216 bytes (exactly a 1K multiple to make the NIC happy), then max_rx_pkt_len is going to be 9216 and data_room_size will be 9216 + RTE_PKTMBUF_HEADROOM. ixgbe_dev_rx_init() will calculate normalize that back to 9216, which will fail the dual VLAN length check. The really nasty part about that is it has a side-effect of enabling scattered RX regardless of the fact that I didn't enable scattered RX in dev_conf.rxmode. The code in the e1000 PMD is similar, so nothing unique to ixgbe. Is that check correct? It seems wrong to be adding space for q-in-q on top of your specified max frame size... Thanks, Jay