[RFC,v6,1/3] genirq: introduce CONFIG_GENERIC_IRQ_MIGRATION and kernel/irq/cpuhotplug.c

Commit Message

Yang Yingliang Sept. 24, 2015, 9:32 a.m. UTC

Add migrating interrupts code to a new file kernel/irq/cpuhotplug.c and
make it depends on CONFIG_GENERIC_IRQ_MIGRATION. So we can use it to migrate
interrupts, before cpu is offline.

Cc: Jiang Liu <[e-mail]>
Cc: Thomas Gleixner <[e-mail]>
Cc: Marc Zyngier <[e-mail]>
Cc: Mark Rutland <[e-mail]>
Cc: Will Deacon <[e-mail]>
Cc: Russell King - ARM Linux <[e-mail]>
Cc: Hanjun Guo <[e-mail]>
Signed-off-by: Yang Yingliang <[e-mail]>
---
 include/linux/irq.h     |  2 ++
 kernel/irq/Kconfig      |  4 +++
 kernel/irq/Makefile     |  1 +
 kernel/irq/cpuhotplug.c | 68 +++++++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 75 insertions(+)
 create mode 100644 kernel/irq/cpuhotplug.c

Comments

Thomas Gleixner Oct. 1, 2015, 2:24 p.m. UTC | #1

On Thu, 1 Oct 2015, tip-bot for Yang Yingliang wrote:

> Commit-ID:  f1e0bb0ad473a32d1b7e6d285ae9f7e47710bb5e
> Gitweb:     http://git.kernel.org/tip/f1e0bb0ad473a32d1b7e6d285ae9f7e47710bb5e
> Author:     Yang Yingliang <[e-mail]>
> AuthorDate: Thu, 24 Sep 2015 17:32:13 +0800
> Committer:  Thomas Gleixner <[e-mail]>
> CommitDate: Thu, 1 Oct 2015 14:51:15 +0200
> 
> genirq: Introduce generic irq migration for cpu hotunplug
> 
> ARM and ARM64 have almost identical code for migrating interrupts on
> cpu hotunplug. Provide a generic version which can be used by both.
> 
> The new code addresses a shortcoming in the ARM[64] variants which
> fails to update the affinity change in some cases. The solution for
> this is to use the core function irq_do_set_affinity() instead of open
> coding it.
> 
> [ tglx: Added copyright notice and license boilerplate. Rewrote
>   	subject and changelog. ]
> 
> Signed-off-by: Yang Yingliang <[e-mail]>
> Acked-by: Russell King - ARM Linux <[e-mail]>
> Cc: Jiang Liu <[e-mail]>
> Cc: Marc Zyngier <[e-mail]>
> Cc: Mark Rutland <[e-mail]>
> Cc: Will Deacon <[e-mail]>
> Cc: Hanjun Guo <[e-mail]>
> Cc: <[e-mail]>
> Link: http://lkml.kernel.org/r/[e-mail]
> Signed-off-by: Thomas Gleixner <[e-mail]>

Russell, Will,

I applied that to a seperate branch

  git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git irq/for-arm

You can pull that into arm[64] so you can apply the architecture
specific part.

Thanks,

	tglx

Catalin Marinas Oct. 1, 2015, 2:47 p.m. UTC | #2

On Thu, Oct 01, 2015 at 04:24:24PM +0200, Thomas Gleixner wrote:
> On Thu, 1 Oct 2015, tip-bot for Yang Yingliang wrote:
> 
> > Commit-ID:  f1e0bb0ad473a32d1b7e6d285ae9f7e47710bb5e
> > Gitweb:     http://git.kernel.org/tip/f1e0bb0ad473a32d1b7e6d285ae9f7e47710bb5e
> > Author:     Yang Yingliang <[e-mail]>
> > AuthorDate: Thu, 24 Sep 2015 17:32:13 +0800
> > Committer:  Thomas Gleixner <[e-mail]>
> > CommitDate: Thu, 1 Oct 2015 14:51:15 +0200
> > 
> > genirq: Introduce generic irq migration for cpu hotunplug
> > 
> > ARM and ARM64 have almost identical code for migrating interrupts on
> > cpu hotunplug. Provide a generic version which can be used by both.
> > 
> > The new code addresses a shortcoming in the ARM[64] variants which
> > fails to update the affinity change in some cases. The solution for
> > this is to use the core function irq_do_set_affinity() instead of open
> > coding it.
> > 
> > [ tglx: Added copyright notice and license boilerplate. Rewrote
> >   	subject and changelog. ]
> > 
> > Signed-off-by: Yang Yingliang <[e-mail]>
> > Acked-by: Russell King - ARM Linux <[e-mail]>
> > Cc: Jiang Liu <[e-mail]>
> > Cc: Marc Zyngier <[e-mail]>
> > Cc: Mark Rutland <[e-mail]>
> > Cc: Will Deacon <[e-mail]>
> > Cc: Hanjun Guo <[e-mail]>
> > Cc: <[e-mail]>
> > Link: http://lkml.kernel.org/r/[e-mail]
> > Signed-off-by: Thomas Gleixner <[e-mail]>
> 
> Russell, Will,
> 
> I applied that to a seperate branch
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git irq/for-arm
> 
> You can pull that into arm[64] so you can apply the architecture
> specific part.

Thanks. I'll queue the corresponding arm64 patch for 4.4.

Patch

diff --git a/include/linux/irq.h b/include/linux/irq.h
index 11bf092..45cc729 100644
--- a/include/linux/irq.h
+++ b/include/linux/irq.h
@@ -452,6 +452,8 @@  extern int irq_set_affinity_locked(struct irq_data *data,
 				   const struct cpumask *cpumask, bool force);
 extern int irq_set_vcpu_affinity(unsigned int irq, void *vcpu_info);
 
+extern void irq_migrate_all_off_this_cpu(void);
+
 #if defined(CONFIG_SMP) && defined(CONFIG_GENERIC_PENDING_IRQ)
 void irq_move_irq(struct irq_data *data);
 void irq_move_masked_irq(struct irq_data *data);
diff --git a/kernel/irq/Kconfig b/kernel/irq/Kconfig
index 9a76e3b..3b48dab 100644
--- a/kernel/irq/Kconfig
+++ b/kernel/irq/Kconfig
@@ -30,6 +30,10 @@  config GENERIC_IRQ_LEGACY_ALLOC_HWIRQ
 config GENERIC_PENDING_IRQ
 	bool
 
+# Support for generic irq migrating off cpu before the cpu is offline.
+config GENERIC_IRQ_MIGRATION
+	bool
+
 # Alpha specific irq affinity mechanism
 config AUTO_IRQ_AFFINITY
        bool
diff --git a/kernel/irq/Makefile b/kernel/irq/Makefile
index d121235..2fc9cbd 100644
--- a/kernel/irq/Makefile
+++ b/kernel/irq/Makefile
@@ -5,5 +5,6 @@  obj-$(CONFIG_GENERIC_IRQ_PROBE) += autoprobe.o
 obj-$(CONFIG_IRQ_DOMAIN) += irqdomain.o
 obj-$(CONFIG_PROC_FS) += proc.o
 obj-$(CONFIG_GENERIC_PENDING_IRQ) += migration.o
+obj-$(CONFIG_GENERIC_IRQ_MIGRATION) += cpuhotplug.o
 obj-$(CONFIG_PM_SLEEP) += pm.o
 obj-$(CONFIG_GENERIC_MSI_IRQ) += msi.o
diff --git a/kernel/irq/cpuhotplug.c b/kernel/irq/cpuhotplug.c
new file mode 100644
index 0000000..8f2e9c0
--- /dev/null
+++ b/kernel/irq/cpuhotplug.c
@@ -0,0 +1,68 @@ 
+#include <linux/irq.h>
+#include <linux/interrupt.h>
+#include <linux/ratelimit.h>
+
+#include "internals.h"
+
+static bool migrate_one_irq(struct irq_desc *desc)
+{
+	struct irq_data *d = irq_desc_get_irq_data(desc);
+	const struct cpumask *affinity = d->common->affinity;
+	struct irq_chip *c;
+	bool ret = false;
+
+	/*
+	 * If this is a per-CPU interrupt, or the affinity does not
+	 * include this CPU, then we have nothing to do.
+	 */
+	if (irqd_is_per_cpu(d) || !cpumask_test_cpu(smp_processor_id(), affinity))
+		return false;
+
+	if (cpumask_any_and(affinity, cpu_online_mask) >= nr_cpu_ids) {
+		affinity = cpu_online_mask;
+		ret = true;
+	}
+
+	c = irq_data_get_irq_chip(d);
+	if (!c->irq_set_affinity) {
+		pr_warn_ratelimited("IRQ%u: unable to set affinity\n", d->irq);
+	} else {
+		int r = irq_do_set_affinity(d, affinity, false);
+		if (r)
+			pr_warn_ratelimited("IRQ%u: set affinity failed(%d).\n", d->irq, r);
+	}
+
+	return ret;
+}
+
+/*
+ * The current CPU has been marked offline.  Migrate IRQs off this CPU.
+ * If the affinity settings do not allow other CPUs, force them onto any
+ * available CPU.
+ *
+ * Note: we must iterate over all IRQs, whether they have an attached
+ * action structure or not, as we need to get chained interrupts too.
+ */
+void irq_migrate_all_off_this_cpu(void)
+{
+	unsigned int irq;
+	struct irq_desc *desc;
+	unsigned long flags;
+
+	local_irq_save(flags);
+
+	for_each_active_irq(irq) {
+		bool affinity_broken;
+
+		desc = irq_to_desc(irq);
+		raw_spin_lock(&desc->lock);
+		affinity_broken = migrate_one_irq(desc);
+		raw_spin_unlock(&desc->lock);
+
+		if (affinity_broken)
+			pr_warn_ratelimited("IRQ%u no longer affine to CPU%u\n",
+					    irq, smp_processor_id());
+	}
+
+	local_irq_restore(flags);
+}