*/
#include <rdma/ib_cm.h>
-#include <rdma/ib_cache.h>
#include <net/dst.h>
#include <net/icmp.h>
#include <linux/icmpv6.h>
#include <linux/delay.h>
+#include <linux/slab.h>
#include <linux/vmalloc.h>
+#include <linux/moduleparam.h>
#include "ipoib.h"
ib_dma_unmap_single(priv->ca, mapping[0], IPOIB_CM_HEAD_SIZE, DMA_FROM_DEVICE);
for (i = 0; i < frags; ++i)
- ib_dma_unmap_single(priv->ca, mapping[i + 1], PAGE_SIZE, DMA_FROM_DEVICE);
+ ib_dma_unmap_page(priv->ca, mapping[i + 1], PAGE_SIZE, DMA_FROM_DEVICE);
}
static int ipoib_cm_post_receive_srq(struct net_device *dev, int id)
goto partial_error;
skb_fill_page_desc(skb, i, page, 0, PAGE_SIZE);
- mapping[i + 1] = ib_dma_map_page(priv->ca, skb_shinfo(skb)->frags[i].page,
+ mapping[i + 1] = ib_dma_map_page(priv->ca, page,
0, PAGE_SIZE, DMA_FROM_DEVICE);
if (unlikely(ib_dma_mapping_error(priv->ca, mapping[i + 1])))
goto partial_error;
ib_dma_unmap_single(priv->ca, mapping[0], IPOIB_CM_HEAD_SIZE, DMA_FROM_DEVICE);
for (; i > 0; --i)
- ib_dma_unmap_single(priv->ca, mapping[i], PAGE_SIZE, DMA_FROM_DEVICE);
+ ib_dma_unmap_page(priv->ca, mapping[i], PAGE_SIZE, DMA_FROM_DEVICE);
dev_kfree_skb_any(skb);
return NULL;
dev_kfree_skb_any(rx_ring[i].skb);
}
- kfree(rx_ring);
+ vfree(rx_ring);
}
static void ipoib_cm_start_rx_drain(struct ipoib_dev_priv *priv)
sge[i].length = PAGE_SIZE;
wr->next = NULL;
- wr->sg_list = priv->cm.rx_sge;
+ wr->sg_list = sge;
wr->num_sge = priv->cm.num_frags;
}
int ret;
int i;
- rx->rx_ring = kcalloc(ipoib_recvq_size, sizeof *rx->rx_ring, GFP_KERNEL);
- if (!rx->rx_ring)
+ rx->rx_ring = vzalloc(ipoib_recvq_size * sizeof *rx->rx_ring);
+ if (!rx->rx_ring) {
+ printk(KERN_WARNING "%s: failed to allocate CM non-SRQ ring (%d entries)\n",
+ priv->ca->name, ipoib_recvq_size);
return -ENOMEM;
+ }
t = kmalloc(sizeof *t, GFP_KERNEL);
if (!t) {
if (length == 0) {
/* don't need this page */
- skb_fill_page_desc(toskb, i, frag->page, 0, PAGE_SIZE);
+ skb_fill_page_desc(toskb, i, skb_frag_page(frag),
+ 0, PAGE_SIZE);
--skb_shinfo(skb)->nr_frags;
} else {
size = min(length, (unsigned) PAGE_SIZE);
- frag->size = size;
+ skb_frag_size_set(frag, size);
skb->data_len += size;
skb->truesize += size;
skb->len += size;
skb_reset_mac_header(skb);
skb_pull(skb, IPOIB_ENCAP_LEN);
- dev->last_rx = jiffies;
++dev->stats.rx_packets;
dev->stats.rx_bytes += skb->len;
void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx)
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
- struct ipoib_tx_buf *tx_req;
+ struct ipoib_cm_tx_buf *tx_req;
u64 addr;
+ int rc;
if (unlikely(skb->len > tx->mtu)) {
ipoib_warn(priv, "packet len %d (> %d) too long to send, dropping\n",
return;
}
- tx_req->mapping[0] = addr;
+ tx_req->mapping = addr;
- if (unlikely(post_send(priv, tx, tx->tx_head & (ipoib_sendq_size - 1),
- addr, skb->len))) {
- ipoib_warn(priv, "post_send failed\n");
+ rc = post_send(priv, tx, tx->tx_head & (ipoib_sendq_size - 1),
+ addr, skb->len);
+ if (unlikely(rc)) {
+ ipoib_warn(priv, "post_send failed, error %d\n", rc);
++dev->stats.tx_errors;
ib_dma_unmap_single(priv->ca, addr, skb->len, DMA_TO_DEVICE);
dev_kfree_skb_any(skb);
if (++priv->tx_outstanding == ipoib_sendq_size) {
ipoib_dbg(priv, "TX ring 0x%x full, stopping kernel net queue\n",
tx->qp->qp_num);
+ if (ib_req_notify_cq(priv->send_cq, IB_CQ_NEXT_COMP))
+ ipoib_warn(priv, "request notify on send CQ failed\n");
netif_stop_queue(dev);
}
}
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct ipoib_cm_tx *tx = wc->qp->qp_context;
unsigned int wr_id = wc->wr_id & ~IPOIB_OP_CM;
- struct ipoib_tx_buf *tx_req;
+ struct ipoib_cm_tx_buf *tx_req;
unsigned long flags;
ipoib_dbg_data(priv, "cm send completion: id %d, status: %d\n",
tx_req = &tx->tx_ring[wr_id];
- ib_dma_unmap_single(priv->ca, tx_req->mapping[0], tx_req->skb->len, DMA_TO_DEVICE);
+ ib_dma_unmap_single(priv->ca, tx_req->mapping, tx_req->skb->len, DMA_TO_DEVICE);
/* FIXME: is this right? Shouldn't we only increment on success? */
++dev->stats.tx_packets;
dev_kfree_skb_any(tx_req->skb);
- spin_lock_irqsave(&priv->tx_lock, flags);
+ netif_tx_lock(dev);
+
++tx->tx_tail;
if (unlikely(--priv->tx_outstanding == ipoib_sendq_size >> 1) &&
netif_queue_stopped(dev) &&
"(status=%d, wrid=%d vend_err %x)\n",
wc->status, wr_id, wc->vendor_err);
- spin_lock(&priv->lock);
+ spin_lock_irqsave(&priv->lock, flags);
neigh = tx->neigh;
if (neigh) {
clear_bit(IPOIB_FLAG_OPER_UP, &tx->flags);
- spin_unlock(&priv->lock);
+ spin_unlock_irqrestore(&priv->lock, flags);
}
- spin_unlock_irqrestore(&priv->tx_lock, flags);
+ netif_tx_unlock(dev);
}
int ipoib_cm_dev_open(struct net_device *dev)
struct ipoib_dev_priv *priv = netdev_priv(p->dev);
int ret;
- p->tx_ring = vmalloc(ipoib_sendq_size * sizeof *p->tx_ring);
+ p->tx_ring = vzalloc(ipoib_sendq_size * sizeof *p->tx_ring);
if (!p->tx_ring) {
ipoib_warn(priv, "failed to allocate tx ring\n");
ret = -ENOMEM;
goto err_tx;
}
- memset(p->tx_ring, 0, ipoib_sendq_size * sizeof *p->tx_ring);
p->qp = ipoib_cm_create_tx_qp(p->dev, p);
if (IS_ERR(p->qp)) {
goto err_send_cm;
}
- ipoib_dbg(priv, "Request connection 0x%x for gid " IPOIB_GID_FMT " qpn 0x%x\n",
- p->qp->qp_num, IPOIB_GID_ARG(pathrec->dgid), qpn);
+ ipoib_dbg(priv, "Request connection 0x%x for gid %pI6 qpn 0x%x\n",
+ p->qp->qp_num, pathrec->dgid.raw, qpn);
return 0;
static void ipoib_cm_tx_destroy(struct ipoib_cm_tx *p)
{
struct ipoib_dev_priv *priv = netdev_priv(p->dev);
- struct ipoib_tx_buf *tx_req;
- unsigned long flags;
+ struct ipoib_cm_tx_buf *tx_req;
unsigned long begin;
ipoib_dbg(priv, "Destroy active connection 0x%x head 0x%x tail 0x%x\n",
while ((int) p->tx_tail - (int) p->tx_head < 0) {
tx_req = &p->tx_ring[p->tx_tail & (ipoib_sendq_size - 1)];
- ib_dma_unmap_single(priv->ca, tx_req->mapping[0], tx_req->skb->len,
+ ib_dma_unmap_single(priv->ca, tx_req->mapping, tx_req->skb->len,
DMA_TO_DEVICE);
dev_kfree_skb_any(tx_req->skb);
++p->tx_tail;
- spin_lock_irqsave(&priv->tx_lock, flags);
+ netif_tx_lock_bh(p->dev);
if (unlikely(--priv->tx_outstanding == ipoib_sendq_size >> 1) &&
netif_queue_stopped(p->dev) &&
test_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags))
netif_wake_queue(p->dev);
- spin_unlock_irqrestore(&priv->tx_lock, flags);
+ netif_tx_unlock_bh(p->dev);
}
if (p->qp)
struct ipoib_dev_priv *priv = netdev_priv(tx->dev);
struct net_device *dev = priv->dev;
struct ipoib_neigh *neigh;
+ unsigned long flags;
int ret;
switch (event->event) {
case IB_CM_REJ_RECEIVED:
case IB_CM_TIMEWAIT_EXIT:
ipoib_dbg(priv, "CM error %d.\n", event->event);
- spin_lock_irq(&priv->tx_lock);
- spin_lock(&priv->lock);
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
neigh = tx->neigh;
if (neigh) {
queue_work(ipoib_workqueue, &priv->cm.reap_task);
}
- spin_unlock(&priv->lock);
- spin_unlock_irq(&priv->tx_lock);
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
break;
default:
break;
if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &tx->flags)) {
list_move(&tx->list, &priv->cm.reap_list);
queue_work(ipoib_workqueue, &priv->cm.reap_task);
- ipoib_dbg(priv, "Reap connection for gid " IPOIB_GID_FMT "\n",
- IPOIB_GID_ARG(tx->neigh->dgid));
+ ipoib_dbg(priv, "Reap connection for gid %pI6\n",
+ tx->neigh->dgid.raw);
tx->neigh = NULL;
}
}
struct ib_sa_path_rec pathrec;
u32 qpn;
- spin_lock_irqsave(&priv->tx_lock, flags);
- spin_lock(&priv->lock);
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
+
while (!list_empty(&priv->cm.start_list)) {
p = list_entry(priv->cm.start_list.next, typeof(*p), list);
list_del_init(&p->list);
neigh = p->neigh;
qpn = IPOIB_QPN(neigh->neighbour->ha);
memcpy(&pathrec, &p->path->pathrec, sizeof pathrec);
- spin_unlock(&priv->lock);
- spin_unlock_irqrestore(&priv->tx_lock, flags);
+
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
+
ret = ipoib_cm_tx_init(p, qpn, &pathrec);
- spin_lock_irqsave(&priv->tx_lock, flags);
- spin_lock(&priv->lock);
+
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
+
if (ret) {
neigh = p->neigh;
if (neigh) {
kfree(p);
}
}
- spin_unlock(&priv->lock);
- spin_unlock_irqrestore(&priv->tx_lock, flags);
+
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
}
static void ipoib_cm_tx_reap(struct work_struct *work)
{
struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv,
cm.reap_task);
+ struct net_device *dev = priv->dev;
struct ipoib_cm_tx *p;
+ unsigned long flags;
+
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
- spin_lock_irq(&priv->tx_lock);
- spin_lock(&priv->lock);
while (!list_empty(&priv->cm.reap_list)) {
p = list_entry(priv->cm.reap_list.next, typeof(*p), list);
list_del(&p->list);
- spin_unlock(&priv->lock);
- spin_unlock_irq(&priv->tx_lock);
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
ipoib_cm_tx_destroy(p);
- spin_lock_irq(&priv->tx_lock);
- spin_lock(&priv->lock);
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
}
- spin_unlock(&priv->lock);
- spin_unlock_irq(&priv->tx_lock);
+
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
}
static void ipoib_cm_skb_reap(struct work_struct *work)
{
struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv,
cm.skb_task);
+ struct net_device *dev = priv->dev;
struct sk_buff *skb;
-
+ unsigned long flags;
unsigned mtu = priv->mcast_mtu;
- spin_lock_irq(&priv->tx_lock);
- spin_lock(&priv->lock);
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
+
while ((skb = skb_dequeue(&priv->cm.skb_queue))) {
- spin_unlock(&priv->lock);
- spin_unlock_irq(&priv->tx_lock);
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
+
if (skb->protocol == htons(ETH_P_IP))
icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
else if (skb->protocol == htons(ETH_P_IPV6))
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, priv->dev);
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
#endif
dev_kfree_skb_any(skb);
- spin_lock_irq(&priv->tx_lock);
- spin_lock(&priv->lock);
+
+ netif_tx_lock_bh(dev);
+ spin_lock_irqsave(&priv->lock, flags);
}
- spin_unlock(&priv->lock);
- spin_unlock_irq(&priv->tx_lock);
+
+ spin_unlock_irqrestore(&priv->lock, flags);
+ netif_tx_unlock_bh(dev);
}
void ipoib_cm_skb_too_long(struct net_device *dev, struct sk_buff *skb,
struct ipoib_dev_priv *priv = netdev_priv(dev);
int e = skb_queue_empty(&priv->cm.skb_queue);
- if (skb->dst)
- skb->dst->ops->update_pmtu(skb->dst, mtu);
+ if (skb_dst(skb))
+ skb_dst(skb)->ops->update_pmtu(skb_dst(skb), mtu);
skb_queue_tail(&priv->cm.skb_queue, skb);
if (e)
struct net_device *dev = to_net_dev(d);
struct ipoib_dev_priv *priv = netdev_priv(dev);
+ if (!rtnl_trylock())
+ return restart_syscall();
+
/* flush paths if we switch modes so that connections are restarted */
if (IPOIB_CM_SUPPORTED(dev->dev_addr) && !strcmp(buf, "connected\n")) {
set_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags);
ipoib_warn(priv, "enabling connected mode "
"will cause multicast packet drops\n");
-
- rtnl_lock();
- dev->features &= ~(NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_TSO);
+ netdev_update_features(dev);
rtnl_unlock();
priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM;
if (!strcmp(buf, "datagram\n")) {
clear_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags);
-
- rtnl_lock();
- if (test_bit(IPOIB_FLAG_CSUM, &priv->flags)) {
- dev->features |= NETIF_F_IP_CSUM | NETIF_F_SG;
- if (priv->hca_caps & IB_DEVICE_UD_TSO)
- dev->features |= NETIF_F_TSO;
- }
+ netdev_update_features(dev);
dev_set_mtu(dev, min(priv->mcast_mtu, dev->mtu));
rtnl_unlock();
ipoib_flush_paths(dev);
return count;
}
+ rtnl_unlock();
return -EINVAL;
}
{
struct ipoib_dev_priv *priv = netdev_priv(dev);
struct ib_srq_init_attr srq_init_attr = {
+ .srq_type = IB_SRQT_BASIC,
.attr = {
.max_wr = ipoib_recvq_size,
.max_sge = max_sge
return;
}
- priv->cm.srq_ring = kzalloc(ipoib_recvq_size * sizeof *priv->cm.srq_ring,
- GFP_KERNEL);
+ priv->cm.srq_ring = vzalloc(ipoib_recvq_size * sizeof *priv->cm.srq_ring);
if (!priv->cm.srq_ring) {
printk(KERN_WARNING "%s: failed to allocate CM SRQ ring (%d entries)\n",
priv->ca->name, ipoib_recvq_size);
ib_destroy_srq(priv->cm.srq);
priv->cm.srq = NULL;
+ return;
}
+
}
int ipoib_cm_dev_init(struct net_device *dev)
projects / linux-flexiantxendom0-3.2.10.git / blobdiff