Re: e1000_netpoll(): disable_irq() triggers might_sleep() on linux-next
From: Thomas Gleixner <hidden>
Date: 2014-10-29 19:49:31
Also in:
lkml
On Wed, 29 Oct 2014, Peter Zijlstra wrote:
On Wed, Oct 29, 2014 at 07:33:00PM +0100, Thomas Gleixner wrote:quoted
Yuck. No. You are just papering over the problem. What happens if you add 'threadirqs' to the kernel command line? Or if the interrupt line is shared with a real threaded interrupt user? The proper solution is to have a poll_lock for e1000 which serializes the hardware interrupt against netpoll instead of using disable/enable_irq(). In fact that's less expensive than the disable/enable_irq() dance and the chance of contention is pretty low. If done right it will be a NOOP for the CONFIG_NET_POLL_CONTROLLER=n case.OK a little something like so then I suppose.. But I suspect most all the network drivers will need this and maybe more, disable_irq() is a popular little thing and we 'just' changed semantics on them.
We changed that almost 4 years ago :) What we 'just' did was to add a prominent warning into the code.
quoted hunk ↗ jump to hunk
--- drivers/net/ethernet/intel/e1000/e1000.h | 2 ++ drivers/net/ethernet/intel/e1000/e1000_main.c | 22 +++++++++++++++++----- kernel/irq/manage.c | 2 +- 3 files changed, 20 insertions(+), 6 deletions(-)diff --git a/drivers/net/ethernet/intel/e1000/e1000.h b/drivers/net/ethernet/intel/e1000/e1000.h index 69707108d23c..3f48609f2318 100644 --- a/drivers/net/ethernet/intel/e1000/e1000.h +++ b/drivers/net/ethernet/intel/e1000/e1000.h@@ -323,6 +323,8 @@ struct e1000_adapter { struct delayed_work watchdog_task; struct delayed_work fifo_stall_task; struct delayed_work phy_info_task; + + spinlock_t irq_lock; }; enum e1000_state_t {diff --git a/drivers/net/ethernet/intel/e1000/e1000_main.c b/drivers/net/ethernet/intel/e1000/e1000_main.c index 5f6aded512f5..d12cbffe2149 100644 --- a/drivers/net/ethernet/intel/e1000/e1000_main.c +++ b/drivers/net/ethernet/intel/e1000/e1000_main.c@@ -1310,6 +1310,7 @@ static int e1000_sw_init(struct e1000_adapter *adapter) e1000_irq_disable(adapter); spin_lock_init(&adapter->stats_lock); + spin_lock_init(&adapter->irq_lock); set_bit(__E1000_DOWN, &adapter->flags);@@ -3748,10 +3749,8 @@ void e1000_update_stats(struct e1000_adapter *adapter) * @irq: interrupt number * @data: pointer to a network interface device structure **/ -static irqreturn_t e1000_intr(int irq, void *data) +static irqreturn_t __e1000_intr(int irq, struct e1000_adapter *adapter) { - struct net_device *netdev = data; - struct e1000_adapter *adapter = netdev_priv(netdev); struct e1000_hw *hw = &adapter->hw; u32 icr = er32(ICR);@@ -3793,6 +3792,19 @@ static irqreturn_t e1000_intr(int irq, void *data) return IRQ_HANDLED; } +static irqreturn_t e1000_intr(int irq, void *data) +{ + struct net_device *netdev = data; + struct e1000_adapter *adapter = netdev_priv(netdev); + irqreturn_t ret; + + spin_lock(&adapter->irq_lock); + ret = __e1000_intr(irq, adapter); + spin_unlock(&adapter->irq_lock); + + return ret; +} + /** * e1000_clean - NAPI Rx polling callback * @adapter: board private structure@@ -5217,9 +5229,9 @@ static void e1000_netpoll(struct net_device *netdev) { struct e1000_adapter *adapter = netdev_priv(netdev); - disable_irq(adapter->pdev->irq); + spin_lock(&adapter->irq_lock) e1000_intr(adapter->pdev->irq, netdev); - enable_irq(adapter->pdev->irq); + spin_unlock(&adapter->irq_lock) } #endifdiff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 0a9104b4608b..b5a4a06bf2fd 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c@@ -427,7 +427,7 @@ EXPORT_SYMBOL(disable_irq_nosync); * to complete before returning. If you use this function while * holding a resource the IRQ handler may need you will deadlock. * - * This function may be called - with care - from IRQ context. + * This function may _NOT_ be called from IRQ context.
It can only be called from preemptible thread context. Thanks, tglx