devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
@ 2024-06-11 11:01 Sebastian Ene
  0 siblings, 0 replies; 8+ messages in thread
From: Sebastian Ene @ 2024-06-11 11:01 UTC (permalink / raw)
  To: arnd, gregkh, will, maz, Rob Herring, Dragan Cvetic,
	Guenter Roeck
  Cc: linux-kernel, devicetree, kernel-team, Sebastian Ene

Hello,


This is a small update of the previously introduced vcpu stall detector
which adds an interrupt to the virtual device to notify the guest VM in
case it stalls. This lets the guest VM to handle the reboot and to
panic in case it expires. 

Changelog from v1:

* 1/2 : collected the Ack from Conor Dooley, thank you Conor !

* 2/2 : applied the feedback received from Conor and used
	platform_get_irq_optional. Removed the error messages during
	probe
   

Thanks, 

Sebastian Ene (2):
  dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual
    device
  misc: Register a PPI for the vcpu stall detection virtual device

 .../misc/qemu,vcpu-stall-detector.yaml        |  6 ++++
 drivers/misc/vcpu_stall_detector.c            | 31 +++++++++++++++++--
 2 files changed, 35 insertions(+), 2 deletions(-)

-- 
2.45.2.505.gda0bf45e8d-goog


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
@ 2024-06-13 14:13 Sebastian Ene
  2024-06-13 14:13 ` [PATCH v2 1/2] dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual device Sebastian Ene
                   ` (2 more replies)
  0 siblings, 3 replies; 8+ messages in thread
From: Sebastian Ene @ 2024-06-13 14:13 UTC (permalink / raw)
  To: arnd, gregkh, will, maz, Rob Herring, Dragan Cvetic,
	Guenter Roeck
  Cc: linux-kernel, devicetree, kernel-team, Sebastian Ene

Hello,

I respined the v2 version to address an issue previously found here:
https://lore.kernel.org/all/202406132132.FBKSWFav-lkp@intel.com/
 
This is a small update of the previously introduced vcpu stall detector
which adds an interrupt to the virtual device to notify the guest VM in
case it stalls. This lets the guest VM to handle the reboot and to
panic in case it expires. 

Changelog from v1:

* 1/2 : collected the Ack from Conor Dooley, thank you Conor !

* 2/2 : applied the feedback received from Conor and used
	platform_get_irq_optional. Removed the error messages during
	probe
   

Thanks, 

Sebastian Ene (2):
  dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual
    device
  misc: Register a PPI for the vcpu stall detection virtual device

 .../misc/qemu,vcpu-stall-detector.yaml        |  6 ++++
 drivers/misc/vcpu_stall_detector.c            | 31 +++++++++++++++++--
 2 files changed, 35 insertions(+), 2 deletions(-)

-- 
2.45.2.505.gda0bf45e8d-goog


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH v2 1/2] dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual device
  2024-06-13 14:13 [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Sebastian Ene
@ 2024-06-13 14:13 ` Sebastian Ene
  2024-06-13 14:13 ` [PATCH v2 2/2] misc: Register a PPI for the vcpu stall detection " Sebastian Ene
  2024-06-13 14:52 ` [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Marc Zyngier
  2 siblings, 0 replies; 8+ messages in thread
From: Sebastian Ene @ 2024-06-13 14:13 UTC (permalink / raw)
  To: arnd, gregkh, will, maz, Rob Herring, Dragan Cvetic,
	Guenter Roeck
  Cc: linux-kernel, devicetree, kernel-team, Sebastian Ene,
	Conor Dooley

The vcpu stall detector allows the host to monitor the availability of a
guest VM. Introduce a PPI interrupt which can be injected from the host
into the virtual gic to let the guest reboot itself.

Signed-off-by: Sebastian Ene <sebastianene@google.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
---
 .../devicetree/bindings/misc/qemu,vcpu-stall-detector.yaml  | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/devicetree/bindings/misc/qemu,vcpu-stall-detector.yaml b/Documentation/devicetree/bindings/misc/qemu,vcpu-stall-detector.yaml
index 1aebeb696ee0..e12d80be00cd 100644
--- a/Documentation/devicetree/bindings/misc/qemu,vcpu-stall-detector.yaml
+++ b/Documentation/devicetree/bindings/misc/qemu,vcpu-stall-detector.yaml
@@ -29,6 +29,9 @@ properties:
       Defaults to 10 if unset.
     default: 10
 
+  interrupts:
+    maxItems: 1
+
   timeout-sec:
     description: |
       The stall detector expiration timeout measured in seconds.
@@ -43,9 +46,12 @@ additionalProperties: false
 
 examples:
   - |
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+
     vmwdt@9030000 {
       compatible = "qemu,vcpu-stall-detector";
       reg = <0x9030000 0x10000>;
       clock-frequency = <10>;
       timeout-sec = <8>;
+      interrupts = <GIC_PPI 15 IRQ_TYPE_EDGE_RISING>;
     };
-- 
2.45.2.505.gda0bf45e8d-goog


^ permalink raw reply related	[flat|nested] 8+ messages in thread

* [PATCH v2 2/2] misc: Register a PPI for the vcpu stall detection virtual device
  2024-06-13 14:13 [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Sebastian Ene
  2024-06-13 14:13 ` [PATCH v2 1/2] dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual device Sebastian Ene
@ 2024-06-13 14:13 ` Sebastian Ene
  2024-06-13 14:52 ` [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Marc Zyngier
  2 siblings, 0 replies; 8+ messages in thread
From: Sebastian Ene @ 2024-06-13 14:13 UTC (permalink / raw)
  To: arnd, gregkh, will, maz, Rob Herring, Dragan Cvetic,
	Guenter Roeck
  Cc: linux-kernel, devicetree, kernel-team, Sebastian Ene

Request a PPI for each vCPU during probe which will be used by the host
to communicate a stall detected event on the vCPU. When the host raises
this interrupt from the virtual machine monitor, the guest is expected to
handle the interrupt and panic.

Signed-off-by: Sebastian Ene <sebastianene@google.com>
---
 drivers/misc/vcpu_stall_detector.c | 31 ++++++++++++++++++++++++++++--
 1 file changed, 29 insertions(+), 2 deletions(-)

diff --git a/drivers/misc/vcpu_stall_detector.c b/drivers/misc/vcpu_stall_detector.c
index e2015c87f03f..41b8c2119e20 100644
--- a/drivers/misc/vcpu_stall_detector.c
+++ b/drivers/misc/vcpu_stall_detector.c
@@ -32,6 +32,7 @@
 struct vcpu_stall_detect_config {
 	u32 clock_freq_hz;
 	u32 stall_timeout_sec;
+	int ppi_irq;
 
 	void __iomem *membase;
 	struct platform_device *dev;
@@ -77,6 +78,12 @@ vcpu_stall_detect_timer_fn(struct hrtimer *hrtimer)
 	return HRTIMER_RESTART;
 }
 
+static irqreturn_t vcpu_stall_detector_irq(int irq, void *dev)
+{
+	panic("vCPU stall detector");
+	return IRQ_HANDLED;
+}
+
 static int start_stall_detector_cpu(unsigned int cpu)
 {
 	u32 ticks, ping_timeout_ms;
@@ -132,7 +139,7 @@ static int stop_stall_detector_cpu(unsigned int cpu)
 
 static int vcpu_stall_detect_probe(struct platform_device *pdev)
 {
-	int ret;
+	int ret, irq;
 	struct resource *r;
 	void __iomem *membase;
 	u32 clock_freq_hz = VCPU_STALL_DEFAULT_CLOCK_HZ;
@@ -169,9 +176,22 @@ static int vcpu_stall_detect_probe(struct platform_device *pdev)
 	vcpu_stall_config = (struct vcpu_stall_detect_config) {
 		.membase		= membase,
 		.clock_freq_hz		= clock_freq_hz,
-		.stall_timeout_sec	= stall_timeout_sec
+		.stall_timeout_sec	= stall_timeout_sec,
+		.ppi_irq		= -1,
 	};
 
+	irq = platform_get_irq_optional(pdev, 0);
+	if (irq > 0) {
+		ret = request_percpu_irq(irq,
+					 vcpu_stall_detector_irq,
+					 "vcpu_stall_detector",
+					 vcpu_stall_detectors);
+		if (ret)
+			goto err;
+
+		vcpu_stall_config.ppi_irq = irq;
+	}
+
 	ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN,
 				"virt/vcpu_stall_detector:online",
 				start_stall_detector_cpu,
@@ -184,6 +204,9 @@ static int vcpu_stall_detect_probe(struct platform_device *pdev)
 	vcpu_stall_config.hp_online = ret;
 	return 0;
 err:
+	if (vcpu_stall_config.ppi_irq > 0)
+		free_percpu_irq(vcpu_stall_config.ppi_irq,
+				vcpu_stall_detectors);
 	return ret;
 }
 
@@ -193,6 +216,10 @@ static void vcpu_stall_detect_remove(struct platform_device *pdev)
 
 	cpuhp_remove_state(vcpu_stall_config.hp_online);
 
+	if (vcpu_stall_config.ppi_irq > 0)
+		free_percpu_irq(vcpu_stall_config.ppi_irq,
+				vcpu_stall_detectors);
+
 	for_each_possible_cpu(cpu)
 		stop_stall_detector_cpu(cpu);
 }
-- 
2.45.2.505.gda0bf45e8d-goog


^ permalink raw reply related	[flat|nested] 8+ messages in thread

* Re: [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
  2024-06-13 14:13 [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Sebastian Ene
  2024-06-13 14:13 ` [PATCH v2 1/2] dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual device Sebastian Ene
  2024-06-13 14:13 ` [PATCH v2 2/2] misc: Register a PPI for the vcpu stall detection " Sebastian Ene
@ 2024-06-13 14:52 ` Marc Zyngier
  2024-06-13 16:44   ` Sebastian Ene
  2 siblings, 1 reply; 8+ messages in thread
From: Marc Zyngier @ 2024-06-13 14:52 UTC (permalink / raw)
  To: Sebastian Ene
  Cc: arnd, gregkh, will, Rob Herring, Dragan Cvetic, Guenter Roeck,
	linux-kernel, devicetree, kernel-team

On Thu, 13 Jun 2024 15:13:33 +0100,
Sebastian Ene <sebastianene@google.com> wrote:
> 
> Hello,
> 
> I respined the v2 version to address an issue previously found here:
> https://lore.kernel.org/all/202406132132.FBKSWFav-lkp@intel.com/

So is this v2 or v3? Having two v2s on the list is... confusing.

	M.

-- 
Without deviation from the norm, progress is not possible.

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
  2024-06-13 14:52 ` [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Marc Zyngier
@ 2024-06-13 16:44   ` Sebastian Ene
  2024-07-03 14:34     ` Greg KH
  0 siblings, 1 reply; 8+ messages in thread
From: Sebastian Ene @ 2024-06-13 16:44 UTC (permalink / raw)
  To: Marc Zyngier
  Cc: arnd, gregkh, will, Rob Herring, Dragan Cvetic, Guenter Roeck,
	linux-kernel, devicetree, kernel-team

On Thu, Jun 13, 2024 at 03:52:08PM +0100, Marc Zyngier wrote:
> On Thu, 13 Jun 2024 15:13:33 +0100,
> Sebastian Ene <sebastianene@google.com> wrote:
> > 
> > Hello,
> > 
> > I respined the v2 version to address an issue previously found here:
> > https://lore.kernel.org/all/202406132132.FBKSWFav-lkp@intel.com/
> 
> So is this v2 or v3? Having two v2s on the list is... confusing.
> 
> 	M.
> 

There is a small change in the patch 2/2 so you are right it should be v3,
sorry for the confusion.

> -- 
> Without deviation from the norm, progress is not possible.

Thanks,
Seb

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
  2024-06-13 16:44   ` Sebastian Ene
@ 2024-07-03 14:34     ` Greg KH
  2024-07-03 15:00       ` Sebastian Ene
  0 siblings, 1 reply; 8+ messages in thread
From: Greg KH @ 2024-07-03 14:34 UTC (permalink / raw)
  To: Sebastian Ene
  Cc: Marc Zyngier, arnd, will, Rob Herring, Dragan Cvetic,
	Guenter Roeck, linux-kernel, devicetree, kernel-team

On Thu, Jun 13, 2024 at 04:44:23PM +0000, Sebastian Ene wrote:
> On Thu, Jun 13, 2024 at 03:52:08PM +0100, Marc Zyngier wrote:
> > On Thu, 13 Jun 2024 15:13:33 +0100,
> > Sebastian Ene <sebastianene@google.com> wrote:
> > > 
> > > Hello,
> > > 
> > > I respined the v2 version to address an issue previously found here:
> > > https://lore.kernel.org/all/202406132132.FBKSWFav-lkp@intel.com/
> > 
> > So is this v2 or v3? Having two v2s on the list is... confusing.
> > 
> > 	M.
> > 
> 
> There is a small change in the patch 2/2 so you are right it should be v3,
> sorry for the confusion.

Please send a new version, nothing we can do with duplicate series on
the lists...

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt
  2024-07-03 14:34     ` Greg KH
@ 2024-07-03 15:00       ` Sebastian Ene
  0 siblings, 0 replies; 8+ messages in thread
From: Sebastian Ene @ 2024-07-03 15:00 UTC (permalink / raw)
  To: Greg KH
  Cc: Marc Zyngier, arnd, will, Rob Herring, Dragan Cvetic,
	Guenter Roeck, linux-kernel, devicetree, kernel-team

On Wed, Jul 03, 2024 at 04:34:08PM +0200, Greg KH wrote:
> On Thu, Jun 13, 2024 at 04:44:23PM +0000, Sebastian Ene wrote:
> > On Thu, Jun 13, 2024 at 03:52:08PM +0100, Marc Zyngier wrote:
> > > On Thu, 13 Jun 2024 15:13:33 +0100,
> > > Sebastian Ene <sebastianene@google.com> wrote:
> > > > 
> > > > Hello,
> > > > 
> > > > I respined the v2 version to address an issue previously found here:
> > > > https://lore.kernel.org/all/202406132132.FBKSWFav-lkp@intel.com/
> > > 
> > > So is this v2 or v3? Having two v2s on the list is... confusing.
> > > 
> > > 	M.
> > > 
> > 
> > There is a small change in the patch 2/2 so you are right it should be v3,
> > sorry for the confusion.

Hello,

> 
> Please send a new version, nothing we can do with duplicate series on
> the lists...


Thanks for letting me know, I will spin a new version.

Seb

^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2024-07-03 15:00 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-06-13 14:13 [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Sebastian Ene
2024-06-13 14:13 ` [PATCH v2 1/2] dt-bindings: vcpu_stall_detector: Add a PPI interrupt to the virtual device Sebastian Ene
2024-06-13 14:13 ` [PATCH v2 2/2] misc: Register a PPI for the vcpu stall detection " Sebastian Ene
2024-06-13 14:52 ` [PATCH v2 0/2] misc: vcpu_stall_detector: Add a PPI interrupt Marc Zyngier
2024-06-13 16:44   ` Sebastian Ene
2024-07-03 14:34     ` Greg KH
2024-07-03 15:00       ` Sebastian Ene
  -- strict thread matches above, loose matches on Subject: below --
2024-06-11 11:01 Sebastian Ene

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).