From: Pradeep Satyanarayana Date: Fri, 21 Dec 2007 21:08:23 +0000 (-0800) Subject: IPoIB/CM: Enable SRQ support on HCAs that support fewer than 16 SG entries X-Git-Tag: v2.6.25-rc1~1234^2~37 X-Git-Url: http://pilppa.com/gitweb/?a=commitdiff_plain;h=586a693448676de5174e752426ced69ec79ab174;p=linux-2.6-omap-h63xx.git IPoIB/CM: Enable SRQ support on HCAs that support fewer than 16 SG entries Some HCAs (such as ehca2) support SRQ, but only support fewer than 16 SG entries for SRQs. Currently IPoIB/CM implicitly assumes all HCAs will support 16 SG entries for SRQs (to handle a 64K MTU with 4K pages). This patch removes that restriction by limiting the maximum MTU in connected mode to what the maximum number of SRQ SG entries allows. This patch addresses Signed-off-by: Pradeep Satyanarayana Signed-off-by: Roland Dreier --- diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index d35025f0652..fe250c60607 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -238,6 +238,8 @@ struct ipoib_cm_dev_priv { struct ib_sge rx_sge[IPOIB_CM_RX_SG]; struct ib_recv_wr rx_wr; int nonsrq_conn_qp; + int max_cm_mtu; + int num_frags; }; /* @@ -503,6 +505,12 @@ static inline int ipoib_cm_has_srq(struct net_device *dev) return !!priv->cm.srq; } +static inline unsigned int ipoib_cm_max_mtu(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = netdev_priv(dev); + return priv->cm.max_cm_mtu; +} + void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx); int ipoib_cm_dev_open(struct net_device *dev); void ipoib_cm_dev_stop(struct net_device *dev); @@ -552,6 +560,11 @@ static inline int ipoib_cm_has_srq(struct net_device *dev) return 0; } +static inline unsigned int ipoib_cm_max_mtu(struct net_device *dev) +{ + return 0; +} + static inline void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx) { diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index fdf33cecc6d..1818f958c25 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -96,13 +96,13 @@ static int ipoib_cm_post_receive_srq(struct net_device *dev, int id) priv->cm.rx_wr.wr_id = id | IPOIB_OP_CM | IPOIB_OP_RECV; - for (i = 0; i < IPOIB_CM_RX_SG; ++i) + for (i = 0; i < priv->cm.num_frags; ++i) priv->cm.rx_sge[i].addr = priv->cm.srq_ring[id].mapping[i]; ret = ib_post_srq_recv(priv->cm.srq, &priv->cm.rx_wr, &bad_wr); if (unlikely(ret)) { ipoib_warn(priv, "post srq failed for buf %d (%d)\n", id, ret); - ipoib_cm_dma_unmap_rx(priv, IPOIB_CM_RX_SG - 1, + ipoib_cm_dma_unmap_rx(priv, priv->cm.num_frags - 1, priv->cm.srq_ring[id].mapping); dev_kfree_skb_any(priv->cm.srq_ring[id].skb); priv->cm.srq_ring[id].skb = NULL; @@ -1399,13 +1399,13 @@ int ipoib_cm_add_mode_attr(struct net_device *dev) return device_create_file(&dev->dev, &dev_attr_mode); } -static void ipoib_cm_create_srq(struct net_device *dev) +static void ipoib_cm_create_srq(struct net_device *dev, int max_sge) { struct ipoib_dev_priv *priv = netdev_priv(dev); struct ib_srq_init_attr srq_init_attr = { .attr = { .max_wr = ipoib_recvq_size, - .max_sge = IPOIB_CM_RX_SG + .max_sge = max_sge } }; @@ -1431,7 +1431,8 @@ static void ipoib_cm_create_srq(struct net_device *dev) int ipoib_cm_dev_init(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); - int i; + int i, ret; + struct ib_device_attr attr; INIT_LIST_HEAD(&priv->cm.passive_ids); INIT_LIST_HEAD(&priv->cm.reap_list); @@ -1448,22 +1449,40 @@ int ipoib_cm_dev_init(struct net_device *dev) skb_queue_head_init(&priv->cm.skb_queue); - for (i = 0; i < IPOIB_CM_RX_SG; ++i) + ret = ib_query_device(priv->ca, &attr); + if (ret) { + printk(KERN_WARNING "ib_query_device() failed with %d\n", ret); + return ret; + } + + ipoib_dbg(priv, "max_srq_sge=%d\n", attr.max_srq_sge); + + attr.max_srq_sge = min_t(int, IPOIB_CM_RX_SG, attr.max_srq_sge); + ipoib_cm_create_srq(dev, attr.max_srq_sge); + if (ipoib_cm_has_srq(dev)) { + priv->cm.max_cm_mtu = attr.max_srq_sge * PAGE_SIZE - 0x10; + priv->cm.num_frags = attr.max_srq_sge; + ipoib_dbg(priv, "max_cm_mtu = 0x%x, num_frags=%d\n", + priv->cm.max_cm_mtu, priv->cm.num_frags); + } else { + priv->cm.max_cm_mtu = IPOIB_CM_MTU; + priv->cm.num_frags = IPOIB_CM_RX_SG; + } + + for (i = 0; i < priv->cm.num_frags; ++i) priv->cm.rx_sge[i].lkey = priv->mr->lkey; priv->cm.rx_sge[0].length = IPOIB_CM_HEAD_SIZE; - for (i = 1; i < IPOIB_CM_RX_SG; ++i) + for (i = 1; i < priv->cm.num_frags; ++i) priv->cm.rx_sge[i].length = PAGE_SIZE; priv->cm.rx_wr.next = NULL; priv->cm.rx_wr.sg_list = priv->cm.rx_sge; - priv->cm.rx_wr.num_sge = IPOIB_CM_RX_SG; - - ipoib_cm_create_srq(dev); + priv->cm.rx_wr.num_sge = priv->cm.num_frags; if (ipoib_cm_has_srq(dev)) { for (i = 0; i < ipoib_recvq_size; ++i) { if (!ipoib_cm_alloc_rx_skb(dev, priv->cm.srq_ring, i, - IPOIB_CM_RX_SG - 1, + priv->cm.num_frags - 1, priv->cm.srq_ring[i].mapping)) { ipoib_warn(priv, "failed to allocate " "receive buffer %d\n", i); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 3bfc2ef1303..d7330451685 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -182,17 +182,20 @@ static int ipoib_change_mtu(struct net_device *dev, int new_mtu) struct ipoib_dev_priv *priv = netdev_priv(dev); /* dev->mtu > 2K ==> connected mode */ - if (ipoib_cm_admin_enabled(dev) && new_mtu <= IPOIB_CM_MTU) { + if (ipoib_cm_admin_enabled(dev)) { + if (new_mtu > ipoib_cm_max_mtu(dev)) + return -EINVAL; + if (new_mtu > priv->mcast_mtu) ipoib_warn(priv, "mtu > %d will cause multicast packet drops.\n", priv->mcast_mtu); + dev->mtu = new_mtu; return 0; } - if (new_mtu > IPOIB_PACKET_SIZE - IPOIB_ENCAP_LEN) { + if (new_mtu > IPOIB_PACKET_SIZE - IPOIB_ENCAP_LEN) return -EINVAL; - } priv->admin_mtu = new_mtu;