mad.c

来自「LINUX 2.6.17.4的源码」· C语言 代码 · 共 2,279 行 · 第 1/5 页

			 sizeof(struct ib_mad_private) -
			 sizeof(struct ib_mad_private_header),
			 DMA_FROM_DEVICE);

	/* Setup MAD receive work completion from "normal" work completion */
	recv->header.wc = *wc;
	recv->header.recv_wc.wc = &recv->header.wc;
	recv->header.recv_wc.mad_len = sizeof(struct ib_mad);
	recv->header.recv_wc.recv_buf.mad = &recv->mad.mad;
	recv->header.recv_wc.recv_buf.grh = &recv->grh;

	if (atomic_read(&qp_info->snoop_count))
		snoop_recv(qp_info, &recv->header.recv_wc, IB_MAD_SNOOP_RECVS);

	/* Validate MAD */
	if (!validate_mad(&recv->mad.mad, qp_info->qp->qp_num))
		goto out;

	if (recv->mad.mad.mad_hdr.mgmt_class ==
	    IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) {
		if (!smi_handle_dr_smp_recv(&recv->mad.smp,
					    port_priv->device->node_type,
					    port_priv->port_num,
					    port_priv->device->phys_port_cnt))
			goto out;
		if (!smi_check_forward_dr_smp(&recv->mad.smp))
			goto local;
		if (!smi_handle_dr_smp_send(&recv->mad.smp,
					    port_priv->device->node_type,
					    port_priv->port_num))
			goto out;
		if (!smi_check_local_smp(&recv->mad.smp, port_priv->device))
			goto out;
	}

local:
	/* Give driver "right of first refusal" on incoming MAD */
	if (port_priv->device->process_mad) {
		int ret;

		if (!response) {
			printk(KERN_ERR PFX "No memory for response MAD\n");
			/*
			 * Is it better to assume that
			 * it wouldn't be processed ?
			 */
			goto out;
		}

		ret = port_priv->device->process_mad(port_priv->device, 0,
						     port_priv->port_num,
						     wc, &recv->grh,
						     &recv->mad.mad,
						     &response->mad.mad);
		if (ret & IB_MAD_RESULT_SUCCESS) {
			if (ret & IB_MAD_RESULT_CONSUMED)
				goto out;
			if (ret & IB_MAD_RESULT_REPLY) {
				agent_send_response(&response->mad.mad,
						    &recv->grh, wc,
						    port_priv->device,
						    port_priv->port_num,
						    qp_info->qp->qp_num);
				goto out;
			}
		}
	}

	mad_agent = find_mad_agent(port_priv, &recv->mad.mad);
	if (mad_agent) {
		ib_mad_complete_recv(mad_agent, &recv->header.recv_wc);
		/*
		 * recv is freed up in error cases in ib_mad_complete_recv
		 * or via recv_handler in ib_mad_complete_recv()
		 */
		recv = NULL;
	}

out:
	/* Post another receive request for this QP */
	if (response) {
		ib_mad_post_receive_mads(qp_info, response);
		if (recv)
			kmem_cache_free(ib_mad_cache, recv);
	} else
		ib_mad_post_receive_mads(qp_info, recv);
}

static void adjust_timeout(struct ib_mad_agent_private *mad_agent_priv)
{
	struct ib_mad_send_wr_private *mad_send_wr;
	unsigned long delay;

	if (list_empty(&mad_agent_priv->wait_list)) {
		cancel_delayed_work(&mad_agent_priv->timed_work);
	} else {
		mad_send_wr = list_entry(mad_agent_priv->wait_list.next,
					 struct ib_mad_send_wr_private,
					 agent_list);

		if (time_after(mad_agent_priv->timeout,
			       mad_send_wr->timeout)) {
			mad_agent_priv->timeout = mad_send_wr->timeout;
			cancel_delayed_work(&mad_agent_priv->timed_work);
			delay = mad_send_wr->timeout - jiffies;
			if ((long)delay <= 0)
				delay = 1;
			queue_delayed_work(mad_agent_priv->qp_info->
					   port_priv->wq,
					   &mad_agent_priv->timed_work, delay);
		}
	}
}

static void wait_for_response(struct ib_mad_send_wr_private *mad_send_wr)
{
	struct ib_mad_agent_private *mad_agent_priv;
	struct ib_mad_send_wr_private *temp_mad_send_wr;
	struct list_head *list_item;
	unsigned long delay;

	mad_agent_priv = mad_send_wr->mad_agent_priv;
	list_del(&mad_send_wr->agent_list);

	delay = mad_send_wr->timeout;
	mad_send_wr->timeout += jiffies;

	if (delay) {
		list_for_each_prev(list_item, &mad_agent_priv->wait_list) {
			temp_mad_send_wr = list_entry(list_item,
						struct ib_mad_send_wr_private,
						agent_list);
			if (time_after(mad_send_wr->timeout,
				       temp_mad_send_wr->timeout))
				break;
		}
	}
	else
		list_item = &mad_agent_priv->wait_list;
	list_add(&mad_send_wr->agent_list, list_item);

	/* Reschedule a work item if we have a shorter timeout */
	if (mad_agent_priv->wait_list.next == &mad_send_wr->agent_list) {
		cancel_delayed_work(&mad_agent_priv->timed_work);
		queue_delayed_work(mad_agent_priv->qp_info->port_priv->wq,
				   &mad_agent_priv->timed_work, delay);
	}
}

void ib_reset_mad_timeout(struct ib_mad_send_wr_private *mad_send_wr,
			  int timeout_ms)
{
	mad_send_wr->timeout = msecs_to_jiffies(timeout_ms);
	wait_for_response(mad_send_wr);
}

/*
 * Process a send work completion
 */
void ib_mad_complete_send_wr(struct ib_mad_send_wr_private *mad_send_wr,
			     struct ib_mad_send_wc *mad_send_wc)
{
	struct ib_mad_agent_private	*mad_agent_priv;
	unsigned long			flags;
	int				ret;

	mad_agent_priv = mad_send_wr->mad_agent_priv;
	spin_lock_irqsave(&mad_agent_priv->lock, flags);
	if (mad_agent_priv->agent.rmpp_version) {
		ret = ib_process_rmpp_send_wc(mad_send_wr, mad_send_wc);
		if (ret == IB_RMPP_RESULT_CONSUMED)
			goto done;
	} else
		ret = IB_RMPP_RESULT_UNHANDLED;

	if (mad_send_wc->status != IB_WC_SUCCESS &&
	    mad_send_wr->status == IB_WC_SUCCESS) {
		mad_send_wr->status = mad_send_wc->status;
		mad_send_wr->refcount -= (mad_send_wr->timeout > 0);
	}

	if (--mad_send_wr->refcount > 0) {
		if (mad_send_wr->refcount == 1 && mad_send_wr->timeout &&
		    mad_send_wr->status == IB_WC_SUCCESS) {
			wait_for_response(mad_send_wr);
		}
		goto done;
	}

	/* Remove send from MAD agent and notify client of completion */
	list_del(&mad_send_wr->agent_list);
	adjust_timeout(mad_agent_priv);
	spin_unlock_irqrestore(&mad_agent_priv->lock, flags);

	if (mad_send_wr->status != IB_WC_SUCCESS )
		mad_send_wc->status = mad_send_wr->status;
	if (ret == IB_RMPP_RESULT_INTERNAL)
		ib_rmpp_send_handler(mad_send_wc);
	else
		mad_agent_priv->agent.send_handler(&mad_agent_priv->agent,
						   mad_send_wc);

	/* Release reference on agent taken when sending */
	deref_mad_agent(mad_agent_priv);
	return;
done:
	spin_unlock_irqrestore(&mad_agent_priv->lock, flags);
}

static void ib_mad_send_done_handler(struct ib_mad_port_private *port_priv,
				     struct ib_wc *wc)
{
	struct ib_mad_send_wr_private	*mad_send_wr, *queued_send_wr;
	struct ib_mad_list_head		*mad_list;
	struct ib_mad_qp_info		*qp_info;
	struct ib_mad_queue		*send_queue;
	struct ib_send_wr		*bad_send_wr;
	struct ib_mad_send_wc		mad_send_wc;
	unsigned long flags;
	int ret;

	mad_list = (struct ib_mad_list_head *)(unsigned long)wc->wr_id;
	mad_send_wr = container_of(mad_list, struct ib_mad_send_wr_private,
				   mad_list);
	send_queue = mad_list->mad_queue;
	qp_info = send_queue->qp_info;

retry:
	dma_unmap_single(mad_send_wr->send_buf.mad_agent->device->dma_device,
			 pci_unmap_addr(mad_send_wr, header_mapping),
			 mad_send_wr->sg_list[0].length, DMA_TO_DEVICE);
	dma_unmap_single(mad_send_wr->send_buf.mad_agent->device->dma_device,
			 pci_unmap_addr(mad_send_wr, payload_mapping),
			 mad_send_wr->sg_list[1].length, DMA_TO_DEVICE);
	queued_send_wr = NULL;
	spin_lock_irqsave(&send_queue->lock, flags);
	list_del(&mad_list->list);

	/* Move queued send to the send queue */
	if (send_queue->count-- > send_queue->max_active) {
		mad_list = container_of(qp_info->overflow_list.next,
					struct ib_mad_list_head, list);
		queued_send_wr = container_of(mad_list,
					struct ib_mad_send_wr_private,
					mad_list);
		list_del(&mad_list->list);
		list_add_tail(&mad_list->list, &send_queue->list);
	}
	spin_unlock_irqrestore(&send_queue->lock, flags);

	mad_send_wc.send_buf = &mad_send_wr->send_buf;
	mad_send_wc.status = wc->status;
	mad_send_wc.vendor_err = wc->vendor_err;
	if (atomic_read(&qp_info->snoop_count))
		snoop_send(qp_info, &mad_send_wr->send_buf, &mad_send_wc,
			   IB_MAD_SNOOP_SEND_COMPLETIONS);
	ib_mad_complete_send_wr(mad_send_wr, &mad_send_wc);

	if (queued_send_wr) {
		ret = ib_post_send(qp_info->qp, &queued_send_wr->send_wr,
				   &bad_send_wr);
		if (ret) {
			printk(KERN_ERR PFX "ib_post_send failed: %d\n", ret);
			mad_send_wr = queued_send_wr;
			wc->status = IB_WC_LOC_QP_OP_ERR;
			goto retry;
		}
	}
}

static void mark_sends_for_retry(struct ib_mad_qp_info *qp_info)
{
	struct ib_mad_send_wr_private *mad_send_wr;
	struct ib_mad_list_head *mad_list;
	unsigned long flags;

	spin_lock_irqsave(&qp_info->send_queue.lock, flags);
	list_for_each_entry(mad_list, &qp_info->send_queue.list, list) {
		mad_send_wr = container_of(mad_list,
					   struct ib_mad_send_wr_private,
					   mad_list);
		mad_send_wr->retry = 1;
	}
	spin_unlock_irqrestore(&qp_info->send_queue.lock, flags);
}

static void mad_error_handler(struct ib_mad_port_private *port_priv,
			      struct ib_wc *wc)
{
	struct ib_mad_list_head *mad_list;
	struct ib_mad_qp_info *qp_info;
	struct ib_mad_send_wr_private *mad_send_wr;
	int ret;

	/* Determine if failure was a send or receive */
	mad_list = (struct ib_mad_list_head *)(unsigned long)wc->wr_id;
	qp_info = mad_list->mad_queue->qp_info;
	if (mad_list->mad_queue == &qp_info->recv_queue)
		/*
		 * Receive errors indicate that the QP has entered the error
		 * state - error handling/shutdown code will cleanup
		 */
		return;

	/*
	 * Send errors will transition the QP to SQE - move
	 * QP to RTS and repost flushed work requests
	 */
	mad_send_wr = container_of(mad_list, struct ib_mad_send_wr_private,
				   mad_list);
	if (wc->status == IB_WC_WR_FLUSH_ERR) {
		if (mad_send_wr->retry) {
			/* Repost send */
			struct ib_send_wr *bad_send_wr;

			mad_send_wr->retry = 0;
			ret = ib_post_send(qp_info->qp, &mad_send_wr->send_wr,
					&bad_send_wr);
			if (ret)
				ib_mad_send_done_handler(port_priv, wc);
		} else
			ib_mad_send_done_handler(port_priv, wc);
	} else {
		struct ib_qp_attr *attr;

		/* Transition QP to RTS and fail offending send */
		attr = kmalloc(sizeof *attr, GFP_KERNEL);
		if (attr) {
			attr->qp_state = IB_QPS_RTS;
			attr->cur_qp_state = IB_QPS_SQE;
			ret = ib_modify_qp(qp_info->qp, attr,
					   IB_QP_STATE | IB_QP_CUR_STATE);
			kfree(attr);
			if (ret)
				printk(KERN_ERR PFX "mad_error_handler - "
				       "ib_modify_qp to RTS : %d\n", ret);
			else
				mark_sends_for_retry(qp_info);
		}
		ib_mad_send_done_handler(port_priv, wc);
	}
}

/*
 * IB MAD completion callback
 */
static void ib_mad_completion_handler(void *data)
{
	struct ib_mad_port_private *port_priv;
	struct ib_wc wc;

	port_priv = (struct ib_mad_port_private *)data;
	ib_req_notify_cq(port_priv->cq, IB_CQ_NEXT_COMP);

	while (ib_poll_cq(port_priv->cq, 1, &wc) == 1) {
		if (wc.status == IB_WC_SUCCESS) {
			switch (wc.opcode) {
			case IB_WC_SEND:
				ib_mad_send_done_handler(port_priv, &wc);
				break;
			case IB_WC_RECV:
				ib_mad_recv_done_handler(port_priv, &wc);
				break;
			default:
				BUG_ON(1);
				break;
			}
		} else
			mad_error_handler(port_priv, &wc);
	}
}

static void cancel_mads(struct ib_mad_agent_private *mad_agent_priv)
{
	unsigned long flags;
	struct ib_mad_send_wr_private *mad_send_wr, *temp_mad_send_wr;
	struct ib_mad_send_wc mad_send_wc;
	struct list_head cancel_list;

	INIT_LIST_HEAD(&cancel_list);

	spin_lock_irqsave(&mad_agent_priv->lock, flags);
	list_for_each_entry_safe(mad_send_wr, temp_mad_send_wr,
				 &mad_agent_priv->send_list, agent_list) {
		if (mad_send_wr->status == IB_WC_SUCCESS) {
 			mad_send_wr->status = IB_WC_WR_FLUSH_ERR;
			mad_send_wr->refcount -= (mad_send_wr->timeout > 0);
		}
	}

	/* Empty wait list to prevent receives from finding a request */
	list_splice_init(&mad_agent_priv->wait_list, &cancel_list);
	/* Empty local completion list as well */
	list_splice_init(&mad_agent_priv->local_list, &cancel_list);
	spin_unlock_irqrestore(&mad_agent_priv->lock, flags);

	/* Report all cancelled requests */
	mad_send_wc.status = IB_WC_WR_FLUSH_ERR;
	mad_send_wc.vendor_err = 0;

	list_for_each_entry_safe(mad_send_wr, temp_mad_send_wr,
				 &cancel_list, agent_list) {
		mad_send_wc.send_buf = &mad_send_wr->send_buf;
		list_del(&mad_send_wr->agent_list);
		mad_agent_priv->agent.send_handler(&mad_agent_priv->agent,
						   &mad_send_wc);
		atomic_dec(&mad_agent_priv->refcount);
	}
}

static struct ib_mad_send_wr_private*
find_send_wr(struct ib_mad_agent_private *mad_agent_priv,
	     struct ib_mad_send_buf *send_buf)
{
	struct ib_mad_send_wr_private *mad_send_wr;

	list_for_each_entry(mad_send_wr, &mad_agent_priv->wait_list,
			    agent_list) {
		if (&mad_send_wr->send_buf == send_buf)
			return mad_send_wr;
	}

	list_for_each_entry(mad_send_wr, &mad_agent_priv->send_list,
			    agent_list) {
		if (is_data_mad(mad_agent_priv, mad_send_wr->send_buf.mad) &&
		    &mad_send_wr->send_buf == send_buf)
			return mad_send_wr;
	}
	return NULL;
}

int ib_modify_mad(struct ib_mad_agent *mad_agent,
		  struct ib_mad_send_buf *send_buf, u32 timeout_ms)
{
	struct ib_mad_agent_private *mad_agent_priv;
	struct ib_mad_send_wr_private *mad_send_wr;
	unsigned long flags;
	int active;

	mad_agent_priv = container_of(mad_agent, struct ib_mad_agent_private,
				      agent);
	spin_lock_irqsave(&mad_agent_priv->lock, flags);
	mad_send_wr = find_send_wr(mad_agent_priv, send_buf);
	if (!mad_send_wr || mad_send_wr->status != IB_WC_SUCCESS) {
		spin_unlock_irqrestore(&mad_agent_priv->lock, flags);
		return -EINVAL;
	}

	active = (!mad_send_wr->timeout || mad_send_wr->refcount > 1);
	if (!timeout_ms) {
		mad_send_wr->status = IB_WC_WR_FLUSH_ERR;
		mad_send_wr->refcount -= (mad_send_wr->timeout > 0);
	}

	mad_s