All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ian Campbell <Ian.Campbell@citrix.com>
To: Thomas Gleixner <tglx@linutronix.de>, <stable@kernel.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Jeremy Fitzhardinge <Jeremy.Fitzhardinge@citrix.com>,
	xen-devel <xen-devel@lists.xensource.com>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Subject: Re: [Xen-devel] Re: xen: IPI interrupts not resumed early enough on suspend/resume
Date: Mon, 17 Oct 2011 14:55:10 +0100	[thread overview]
Message-ID: <1318859710.16132.15.camel@zakaz.uk.xensource.com> (raw)
In-Reply-To: <alpine.LFD.2.02.1110171049440.3240@ionos>

Hi Greg,

On Mon, 2011-10-17 at 09:51 +0100, Thomas Gleixner wrote:
> On Sat, 15 Oct 2011, Ian Campbell wrote:
> > I noticed that the backport to longterm 2.6.32 won't work since syscore
> > isn't available back then. I think one approach for the backport would
> > be to call irq_pm_syscore_resume() directly from sysdev_resume(), does
> > that sound sensible?
> 
> I think so. [...]

The following is a backport of 9bab0b7fbaceec47d32db51cd9e59c82fb071f5a
(currently in tip.git#irq/core tree and queued for 3.2) to 2.6.32 as
described above.

8<-------------------------------------------------

>From 0e3acef83438e502ca5a1985cc5231c325ca73e2 Mon Sep 17 00:00:00 2001
From: Ian Campbell <ian.campbell@citrix.com>
Date: Mon, 3 Oct 2011 15:37:00 +0100
Subject: [PATCH] genirq: Add IRQF_RESUME_EARLY and resume such IRQs earlier

commit 9bab0b7fbaceec47d32db51cd9e59c82fb071f5a upstream

This adds a mechanism to resume selected IRQs during syscore_resume
instead of dpm_resume_noirq.

Under Xen we need to resume IRQs associated with IPIs early enough
that the resched IPI is unmasked and we can therefore schedule
ourselves out of the stop_machine where the suspend/resume takes
place.

This issue was introduced by 676dc3cf5bc3 "xen: Use IRQF_FORCE_RESUME".

Back ported to 2.6.32 (which lacks syscore support) by calling the relavant
resume function directly from sysdev_resume).

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Jeremy Fitzhardinge <Jeremy.Fitzhardinge@citrix.com>
Cc: xen-devel <xen-devel@lists.xensource.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Link: http://lkml.kernel.org/r/1318713254.11016.52.camel@dagon.hellion.org.uk
Cc: stable@kernel.org (at least to 2.6.32.y)
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 drivers/base/sys.c        |    6 ++++++
 drivers/xen/events.c      |    2 +-
 include/linux/interrupt.h |    5 +++++
 kernel/irq/pm.c           |   35 ++++++++++++++++++++++++++++-------
 4 files changed, 40 insertions(+), 8 deletions(-)

diff --git a/drivers/base/sys.c b/drivers/base/sys.c
index 0d90390..3f202f7 100644
--- a/drivers/base/sys.c
+++ b/drivers/base/sys.c
@@ -471,6 +471,12 @@ int sysdev_resume(void)
 {
 	struct sysdev_class *cls;
 
+	/*
+	 * Called from syscore in mainline but called directly here
+	 * since syscore does not exist in this tree.
+	 */
+	irq_pm_syscore_resume();
+
 	WARN_ONCE(!irqs_disabled(),
 		"Interrupts enabled while resuming system devices\n");
 
diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 009ca4e..15ed43e 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -536,7 +536,7 @@ int bind_ipi_to_irqhandler(enum ipi_vector ipi,
 	if (irq < 0)
 		return irq;
 
-	irqflags |= IRQF_NO_SUSPEND | IRQF_FORCE_RESUME;
+	irqflags |= IRQF_NO_SUSPEND | IRQF_FORCE_RESUME | IRQF_EARLY_RESUME;
 	retval = request_irq(irq, handler, irqflags, devname, dev_id);
 	if (retval != 0) {
 		unbind_from_irq(irq);
diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index 4528f29..c7e1aa5 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -54,6 +54,8 @@
  *                irq line disabled until the threaded handler has been run.
  * IRQF_NO_SUSPEND - Do not disable this IRQ during suspend
  * IRQF_FORCE_RESUME - Force enable it on resume even if IRQF_NO_SUSPEND is set
+ * IRQF_EARLY_RESUME - Resume IRQ early during syscore instead of at device
+ *                resume time.
  */
 #define IRQF_DISABLED		0x00000020
 #define IRQF_SAMPLE_RANDOM	0x00000040
@@ -66,6 +68,7 @@
 #define IRQF_ONESHOT		0x00002000
 #define IRQF_NO_SUSPEND		0x00004000
 #define IRQF_FORCE_RESUME	0x00008000
+#define IRQF_EARLY_RESUME	0x00020000
 
 #define IRQF_TIMER		(__IRQF_TIMER | IRQF_NO_SUSPEND)
 
@@ -196,6 +199,7 @@ extern void enable_irq(unsigned int irq);
 #ifdef CONFIG_GENERIC_HARDIRQS
 extern void suspend_device_irqs(void);
 extern void resume_device_irqs(void);
+extern void irq_pm_syscore_resume(void);
 #ifdef CONFIG_PM_SLEEP
 extern int check_wakeup_irqs(void);
 #else
@@ -204,6 +208,7 @@ static inline int check_wakeup_irqs(void) { return 0; }
 #else
 static inline void suspend_device_irqs(void) { };
 static inline void resume_device_irqs(void) { };
+static inline void irq_pm_syscore_resume(void) { };
 static inline int check_wakeup_irqs(void) { return 0; }
 #endif
 
diff --git a/kernel/irq/pm.c b/kernel/irq/pm.c
index 0067abb..b1fc3dd 100644
--- a/kernel/irq/pm.c
+++ b/kernel/irq/pm.c
@@ -39,25 +39,46 @@ void suspend_device_irqs(void)
 }
 EXPORT_SYMBOL_GPL(suspend_device_irqs);
 
-/**
- * resume_device_irqs - enable interrupt lines disabled by suspend_device_irqs()
- *
- * Enable all interrupt lines previously disabled by suspend_device_irqs() that
- * have the IRQ_SUSPENDED flag set.
- */
-void resume_device_irqs(void)
+static void resume_irqs(bool want_early)
 {
 	struct irq_desc *desc;
 	int irq;
 
 	for_each_irq_desc(irq, desc) {
 		unsigned long flags;
+		bool is_early = desc->action &&
+			desc->action->flags & IRQF_EARLY_RESUME;
+
+		if (is_early != want_early)
+			continue;
 
 		spin_lock_irqsave(&desc->lock, flags);
 		__enable_irq(desc, irq, true);
 		spin_unlock_irqrestore(&desc->lock, flags);
 	}
 }
+
+/**
+ * irq_pm_syscore_ops - enable interrupt lines early
+ *
+ * Enable all interrupt lines with %IRQF_EARLY_RESUME set.
+ */
+void irq_pm_syscore_resume(void)
+{
+	resume_irqs(true);
+}
+
+/**
+ * resume_device_irqs - enable interrupt lines disabled by suspend_device_irqs()
+ *
+ * Enable all non-%IRQF_EARLY_RESUME interrupt lines previously
+ * disabled by suspend_device_irqs() that have the IRQS_SUSPENDED flag
+ * set as well as those with %IRQF_FORCE_RESUME.
+ */
+void resume_device_irqs(void)
+{
+	resume_irqs(false);
+}
 EXPORT_SYMBOL_GPL(resume_device_irqs);
 
 /**
-- 
1.7.2.5




WARNING: multiple messages have this Message-ID (diff)
From: Ian Campbell <Ian.Campbell@citrix.com>
To: Thomas Gleixner <tglx@linutronix.de>, stable@kernel.org
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>,
	xen-devel <xen-devel@lists.xensource.com>,
	Jeremy Fitzhardinge <Jeremy.Fitzhardinge@citrix.com>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Subject: Re: Re: xen: IPI interrupts not resumed early enough on suspend/resume
Date: Mon, 17 Oct 2011 14:55:10 +0100	[thread overview]
Message-ID: <1318859710.16132.15.camel@zakaz.uk.xensource.com> (raw)
In-Reply-To: <alpine.LFD.2.02.1110171049440.3240@ionos>

Hi Greg,

On Mon, 2011-10-17 at 09:51 +0100, Thomas Gleixner wrote:
> On Sat, 15 Oct 2011, Ian Campbell wrote:
> > I noticed that the backport to longterm 2.6.32 won't work since syscore
> > isn't available back then. I think one approach for the backport would
> > be to call irq_pm_syscore_resume() directly from sysdev_resume(), does
> > that sound sensible?
> 
> I think so. [...]

The following is a backport of 9bab0b7fbaceec47d32db51cd9e59c82fb071f5a
(currently in tip.git#irq/core tree and queued for 3.2) to 2.6.32 as
described above.

8<-------------------------------------------------

>From 0e3acef83438e502ca5a1985cc5231c325ca73e2 Mon Sep 17 00:00:00 2001
From: Ian Campbell <ian.campbell@citrix.com>
Date: Mon, 3 Oct 2011 15:37:00 +0100
Subject: [PATCH] genirq: Add IRQF_RESUME_EARLY and resume such IRQs earlier

commit 9bab0b7fbaceec47d32db51cd9e59c82fb071f5a upstream

This adds a mechanism to resume selected IRQs during syscore_resume
instead of dpm_resume_noirq.

Under Xen we need to resume IRQs associated with IPIs early enough
that the resched IPI is unmasked and we can therefore schedule
ourselves out of the stop_machine where the suspend/resume takes
place.

This issue was introduced by 676dc3cf5bc3 "xen: Use IRQF_FORCE_RESUME".

Back ported to 2.6.32 (which lacks syscore support) by calling the relavant
resume function directly from sysdev_resume).

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Jeremy Fitzhardinge <Jeremy.Fitzhardinge@citrix.com>
Cc: xen-devel <xen-devel@lists.xensource.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Link: http://lkml.kernel.org/r/1318713254.11016.52.camel@dagon.hellion.org.uk
Cc: stable@kernel.org (at least to 2.6.32.y)
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 drivers/base/sys.c        |    6 ++++++
 drivers/xen/events.c      |    2 +-
 include/linux/interrupt.h |    5 +++++
 kernel/irq/pm.c           |   35 ++++++++++++++++++++++++++++-------
 4 files changed, 40 insertions(+), 8 deletions(-)

diff --git a/drivers/base/sys.c b/drivers/base/sys.c
index 0d90390..3f202f7 100644
--- a/drivers/base/sys.c
+++ b/drivers/base/sys.c
@@ -471,6 +471,12 @@ int sysdev_resume(void)
 {
 	struct sysdev_class *cls;
 
+	/*
+	 * Called from syscore in mainline but called directly here
+	 * since syscore does not exist in this tree.
+	 */
+	irq_pm_syscore_resume();
+
 	WARN_ONCE(!irqs_disabled(),
 		"Interrupts enabled while resuming system devices\n");
 
diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 009ca4e..15ed43e 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -536,7 +536,7 @@ int bind_ipi_to_irqhandler(enum ipi_vector ipi,
 	if (irq < 0)
 		return irq;
 
-	irqflags |= IRQF_NO_SUSPEND | IRQF_FORCE_RESUME;
+	irqflags |= IRQF_NO_SUSPEND | IRQF_FORCE_RESUME | IRQF_EARLY_RESUME;
 	retval = request_irq(irq, handler, irqflags, devname, dev_id);
 	if (retval != 0) {
 		unbind_from_irq(irq);
diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index 4528f29..c7e1aa5 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -54,6 +54,8 @@
  *                irq line disabled until the threaded handler has been run.
  * IRQF_NO_SUSPEND - Do not disable this IRQ during suspend
  * IRQF_FORCE_RESUME - Force enable it on resume even if IRQF_NO_SUSPEND is set
+ * IRQF_EARLY_RESUME - Resume IRQ early during syscore instead of at device
+ *                resume time.
  */
 #define IRQF_DISABLED		0x00000020
 #define IRQF_SAMPLE_RANDOM	0x00000040
@@ -66,6 +68,7 @@
 #define IRQF_ONESHOT		0x00002000
 #define IRQF_NO_SUSPEND		0x00004000
 #define IRQF_FORCE_RESUME	0x00008000
+#define IRQF_EARLY_RESUME	0x00020000
 
 #define IRQF_TIMER		(__IRQF_TIMER | IRQF_NO_SUSPEND)
 
@@ -196,6 +199,7 @@ extern void enable_irq(unsigned int irq);
 #ifdef CONFIG_GENERIC_HARDIRQS
 extern void suspend_device_irqs(void);
 extern void resume_device_irqs(void);
+extern void irq_pm_syscore_resume(void);
 #ifdef CONFIG_PM_SLEEP
 extern int check_wakeup_irqs(void);
 #else
@@ -204,6 +208,7 @@ static inline int check_wakeup_irqs(void) { return 0; }
 #else
 static inline void suspend_device_irqs(void) { };
 static inline void resume_device_irqs(void) { };
+static inline void irq_pm_syscore_resume(void) { };
 static inline int check_wakeup_irqs(void) { return 0; }
 #endif
 
diff --git a/kernel/irq/pm.c b/kernel/irq/pm.c
index 0067abb..b1fc3dd 100644
--- a/kernel/irq/pm.c
+++ b/kernel/irq/pm.c
@@ -39,25 +39,46 @@ void suspend_device_irqs(void)
 }
 EXPORT_SYMBOL_GPL(suspend_device_irqs);
 
-/**
- * resume_device_irqs - enable interrupt lines disabled by suspend_device_irqs()
- *
- * Enable all interrupt lines previously disabled by suspend_device_irqs() that
- * have the IRQ_SUSPENDED flag set.
- */
-void resume_device_irqs(void)
+static void resume_irqs(bool want_early)
 {
 	struct irq_desc *desc;
 	int irq;
 
 	for_each_irq_desc(irq, desc) {
 		unsigned long flags;
+		bool is_early = desc->action &&
+			desc->action->flags & IRQF_EARLY_RESUME;
+
+		if (is_early != want_early)
+			continue;
 
 		spin_lock_irqsave(&desc->lock, flags);
 		__enable_irq(desc, irq, true);
 		spin_unlock_irqrestore(&desc->lock, flags);
 	}
 }
+
+/**
+ * irq_pm_syscore_ops - enable interrupt lines early
+ *
+ * Enable all interrupt lines with %IRQF_EARLY_RESUME set.
+ */
+void irq_pm_syscore_resume(void)
+{
+	resume_irqs(true);
+}
+
+/**
+ * resume_device_irqs - enable interrupt lines disabled by suspend_device_irqs()
+ *
+ * Enable all non-%IRQF_EARLY_RESUME interrupt lines previously
+ * disabled by suspend_device_irqs() that have the IRQS_SUSPENDED flag
+ * set as well as those with %IRQF_FORCE_RESUME.
+ */
+void resume_device_irqs(void)
+{
+	resume_irqs(false);
+}
 EXPORT_SYMBOL_GPL(resume_device_irqs);
 
 /**
-- 
1.7.2.5

  parent reply	other threads:[~2011-10-17 13:55 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-10-03 15:10 xen: IPI interrupts not resumed early enough on suspend/resume Ian Campbell
2011-10-03 18:42 ` Thomas Gleixner
2011-10-03 18:42   ` Thomas Gleixner
2011-10-03 19:08   ` Ian Campbell
2011-10-03 20:35     ` Thomas Gleixner
2011-10-03 20:35       ` Thomas Gleixner
2011-10-07 15:32       ` Ian Campbell
2011-10-07 15:32         ` Ian Campbell
2011-10-07 16:29         ` Thomas Gleixner
2011-10-10 13:06           ` Ian Campbell
2011-10-14 13:23             ` Ian Campbell
2011-10-15 21:14               ` [Xen-devel] " Ian Campbell
2011-10-15 21:14                 ` Ian Campbell
2011-10-17  8:51                 ` Thomas Gleixner
2011-10-17  9:25                   ` Ian Campbell
2011-10-17  9:39                     ` Thomas Gleixner
2011-10-17 13:55                   ` Ian Campbell [this message]
2011-10-17 13:55                     ` Ian Campbell
2011-10-03 20:02   ` Rafael J. Wysocki
2011-10-03 20:28     ` Thomas Gleixner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1318859710.16132.15.camel@zakaz.uk.xensource.com \
    --to=ian.campbell@citrix.com \
    --cc=Jeremy.Fitzhardinge@citrix.com \
    --cc=konrad.wilk@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rjw@sisk.pl \
    --cc=stable@kernel.org \
    --cc=tglx@linutronix.de \
    --cc=xen-devel@lists.xensource.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.