Hello Eli!
I'm referring to the thread below, where you explained the rational of
calling skb_orphan() after post_send() in UD path:
[ewg] Re: [PATCH] call skb_orphan() after sending an SKB
http://lists.openfabrics.org/pipermail/ewg/2008-February/005739.html
Looking at cm_send path I don't see that call. Is there any reason not to
do that here?
void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct
ipoib_cm_tx *tx)
{
...
if (unlikely(post_send(priv, tx, tx->tx_head & (ipoib_sendq_size - 1),
addr, skb->len))) {
ipoib_warn(priv, "post_send failed\n");
++priv->stats.tx_errors;
ib_dma_unmap_single(priv->ca, addr, skb->len, DMA_TO_DEVICE);
dev_kfree_skb_any(skb);
} else {
dev->trans_start = jiffies;
++tx->tx_head;
if (++priv->tx_outstanding == ipoib_sendq_size - 1) {
ipoib_dbg(priv, "%s: TX ring 0x%x full,"
"stopping kernel net queue\n", __func__,
tx->qp->qp_num);
netif_stop_queue(dev);
}
}
}
Thanks and regards
Nam
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
