Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 63b66d12 authored Feb 26, 2007 by Ron Mercer Committed by Jeff Garzik Feb 27, 2007

qla3xxx: bugfix tx reset after stress conditions.



To Reproduce the Problem:
To reproduce this panic consistently, we run an intensive network
application like 'netperf' and then switch to a different console.
After waiting for a couple of seconds, you will see a tx reset has occured.

Reason:
We enable interrupts even if we were not running.

Solution:
Now we will enable interrupts only after we are ready to give up the poll
routine.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

parent 0f8ab89e

drivers/net/qla3xxx.c

+41 −28

Original line number	Diff line number	Diff line
		@@ -1921,10 +1921,11 @@ static int ql_tx_rx_clean(struct ql3_adapter *qdev,
		struct net_rsp_iocb *net_rsp;
		struct net_device *ndev = qdev->ndev;
		unsigned long hw_flags;
		int work_done = 0;

		/* While there are entries in the completion queue. */
		while ((cpu_to_le32(*(qdev->prsp_producer_index)) !=
		qdev->rsp_consumer_index) && (*rx_cleaned < work_to_do)) {
		qdev->rsp_consumer_index) && (work_done < work_to_do)) {

		net_rsp = qdev->rsp_current;
		switch (net_rsp->opcode) {
		@@ -1975,8 +1976,11 @@ static int ql_tx_rx_clean(struct ql3_adapter *qdev,
		} else {
		qdev->rsp_current++;
		}

		work_done = tx_cleaned + rx_cleaned;
		}

		if(work_done) {
		spin_lock_irqsave(&qdev->hw_lock, hw_flags);

		ql_update_lrg_bufq_prod_index(qdev);
		@@ -1991,22 +1995,23 @@ static int ql_tx_rx_clean(struct ql3_adapter *qdev,
		qdev->small_buf_release_cnt -= 8;
		}

		wmb();
		ql_write_common_reg(qdev,
		&port_regs->CommonRegs.
		rxSmallQProducerIndex,
		qdev->small_buf_q_producer_index);

		}

		ql_write_common_reg(qdev,
		&port_regs->CommonRegs.rspQConsumerIndex,
		qdev->rsp_consumer_index);
		spin_unlock_irqrestore(&qdev->hw_lock, hw_flags);

		if (unlikely(netif_queue_stopped(qdev->ndev))) {
		if (netif_queue_stopped(qdev->ndev) &&
		(atomic_read(&qdev->tx_count) > (NUM_REQ_Q_ENTRIES / 4)))
		(atomic_read(&qdev->tx_count) >
		(NUM_REQ_Q_ENTRIES / 4)))
		netif_wake_queue(qdev->ndev);
		}
		}

		return tx_cleaned + rx_cleaned;
		}
		@@ -2016,6 +2021,8 @@ static int ql_poll(struct net_device ndev, int budget)
		struct ql3_adapter *qdev = netdev_priv(ndev);
		int work_to_do = min(*budget, ndev->quota);
		int rx_cleaned = 0, tx_cleaned = 0;
		unsigned long hw_flags;
		struct ql3xxx_port_registers __iomem *port_regs = qdev->mem_map_registers;

		if (!netif_carrier_ok(ndev))
		goto quit_polling;
		@@ -2027,6 +2034,13 @@ static int ql_poll(struct net_device ndev, int budget)
		if ((!tx_cleaned && !rx_cleaned) \|\| !netif_running(ndev)) {
		quit_polling:
		netif_rx_complete(ndev);

		spin_lock_irqsave(&qdev->hw_lock, hw_flags);
		ql_write_common_reg(qdev,
		&port_regs->CommonRegs.rspQConsumerIndex,
		qdev->rsp_consumer_index);
		spin_unlock_irqrestore(&qdev->hw_lock, hw_flags);

		ql_enable_interrupts(qdev);
		return 0;
		}
		@@ -2079,11 +2093,10 @@ static irqreturn_t ql3xxx_isr(int irq, void *dev_id)
		queue_delayed_work(qdev->workqueue, &qdev->reset_work, 0);
		spin_unlock(&qdev->adapter_lock);
		} else if (value & ISP_IMR_DISABLE_CMPL_INT) {
		if (likely(netif_rx_schedule_prep(ndev))) {
		ql_disable_interrupts(qdev);
		if (likely(netif_rx_schedule_prep(ndev)))
		__netif_rx_schedule(ndev);
		else
		ql_enable_interrupts(qdev);
		}
		} else {
		return IRQ_NONE;
		}