Skip to contentSkip to conversation

soc: qcom: smp2p: Add wakeup capability to SMP2P IRQ

PATCH Superseded Deepak Kumar Singh 2 participants 3 messages Sep 18, 2021 – Sep 21, 2021

Subsystems (2)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH v4 1/1 soc: qcom: smp2p: Add wakeup capability to SMP2P IRQ

Remote susbsystems notify fatal crash throught smp2p interrupt. When remoteproc crashes it can cause soc to come out of low power state and may not allow again to enter in low power state until crash is handled.

Mark smp2p interrupt wakeup capable so that interrupt handler is executed and remoteproc crash can be handled in system resume path. This patch marks interrupt wakeup capable but keeps wakeup disabled by default. User space can enable it based on its requirement for wakeup from suspend.

---
 drivers/soc/qcom/smp2p.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/soc/qcom/smp2p.c +21 -0
index 2df4883..60ad632 100644
@@ -14,6 +14,7 @@
 #include <linux/mfd/syscon.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/pm_wakeirq.h>
 #include <linux/regmap.h>
 #include <linux/soc/qcom/smem.h>
 #include <linux/soc/qcom/smem_state.h>
@@ -538,9 +539,26 @@ static int qcom_smp2p_probe(struct platform_device *pdev)
 		goto unwind_interfaces;
 	}
 
+	/*
+	 * Treat smp2p interrupt as wakeup source, but keep it disabled
+	 * by default. User space can decide enabling it depending on its
+	 * use cases. For example if remoteproc crashes and device wants
+	 * to handle it immediatedly (e.g. to not miss phone calls) it can
+	 * enable wakeup source from user space, while other devices which
+	 * do not have proper autosleep feature may want to handle it with
+	 * other wakeup events (e.g. Power button) instead waking up immediately.
+	 */
+	device_set_wakeup_capable(&pdev->dev, true);
+
+	ret = dev_pm_set_wake_irq(&pdev->dev, irq);
+	if (ret)
+		goto set_wake_irq_fail;
 
 	return 0;
 
+set_wake_irq_fail:
+	dev_pm_clear_wake_irq(&pdev->dev);
+
 unwind_interfaces:
 	list_for_each_entry(entry, &smp2p->inbound, node)
 		irq_domain_remove(entry->domain);
@@ -565,6 +583,9 @@ static int qcom_smp2p_remove(struct platform_device *pdev)
 	struct qcom_smp2p *smp2p = platform_get_drvdata(pdev);
 	struct smp2p_entry *entry;
 
+	dev_pm_clear_wake_irq(&pdev->dev);
+	device_init_wakeup(&pdev->dev, false);
+
 	list_for_each_entry(entry, &smp2p->inbound, node)
 		irq_domain_remove(entry->domain);
 
Signature · 2 lines
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project
#1 lore raw

Quoting Deepak Kumar Singh (2021-09-18 12:02:15)

Quoted 40 lines
drivers/soc/qcom/smp2p.c +18 -0
index 2df4883..60ad632 100644
@@ -14,6 +14,7 @@
 #include <linux/mfd/syscon.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/pm_wakeirq.h>
 #include <linux/regmap.h>
 #include <linux/soc/qcom/smem.h>
 #include <linux/soc/qcom/smem_state.h>
@@ -538,9 +539,26 @@ static int qcom_smp2p_probe(struct platform_device *pdev)
                goto unwind_interfaces;
        }
 
+       /*
+        * Treat smp2p interrupt as wakeup source, but keep it disabled
+        * by default. User space can decide enabling it depending on its
+        * use cases. For example if remoteproc crashes and device wants
+        * to handle it immediatedly (e.g. to not miss phone calls) it can
+        * enable wakeup source from user space, while other devices which
+        * do not have proper autosleep feature may want to handle it with
+        * other wakeup events (e.g. Power button) instead waking up immediately.
+        */
+       device_set_wakeup_capable(&pdev->dev, true);
+
+       ret = dev_pm_set_wake_irq(&pdev->dev, irq);
+       if (ret)
+               goto set_wake_irq_fail;
 
        return 0;
 
+set_wake_irq_fail:
+       dev_pm_clear_wake_irq(&pdev->dev);
+
 unwind_interfaces:
        list_for_each_entry(entry, &smp2p->inbound, node)
                irq_domain_remove(entry->domain);

drivers/soc/qcom/smp2p.c static int qcom_smp2p_remove(struct platform_device *pdev) · line 587

Quoted 5 lines
drivers/soc/qcom/smp2p.c +2 -0
@@ -565,6 +583,9 @@ static int qcom_smp2p_remove(struct platform_device *pdev)
        struct qcom_smp2p *smp2p = platform_get_drvdata(pdev);
        struct smp2p_entry *entry;
 
+       dev_pm_clear_wake_irq(&pdev->dev);
+       device_init_wakeup(&pdev->dev, false);

Is this device_init_wakeup() call necessary? It looks like we can get away without it and then once this driver probes the device will have the wakeup capability set on it. Future binding/unbinding of the driver will keep working.

#2 lore raw
Quoted 44 lines · Stephen Boyd

Quoting Deepak Kumar Singh (2021-09-18 12:02:15)

Quoted 43 lines
drivers/soc/qcom/smp2p.c +18 -0
index 2df4883..60ad632 100644
@@ -14,6 +14,7 @@
  #include <linux/mfd/syscon.h>
  #include <linux/module.h>
  #include <linux/platform_device.h>
+#include <linux/pm_wakeirq.h>
  #include <linux/regmap.h>
  #include <linux/soc/qcom/smem.h>
  #include <linux/soc/qcom/smem_state.h>
@@ -538,9 +539,26 @@ static int qcom_smp2p_probe(struct platform_device *pdev)
                 goto unwind_interfaces;
         }
 
+       /*
+        * Treat smp2p interrupt as wakeup source, but keep it disabled
+        * by default. User space can decide enabling it depending on its
+        * use cases. For example if remoteproc crashes and device wants
+        * to handle it immediatedly (e.g. to not miss phone calls) it can
+        * enable wakeup source from user space, while other devices which
+        * do not have proper autosleep feature may want to handle it with
+        * other wakeup events (e.g. Power button) instead waking up immediately.
+        */
+       device_set_wakeup_capable(&pdev->dev, true);
+
+       ret = dev_pm_set_wake_irq(&pdev->dev, irq);
+       if (ret)
+               goto set_wake_irq_fail;
 
         return 0;
 
+set_wake_irq_fail:
+       dev_pm_clear_wake_irq(&pdev->dev);
+
  unwind_interfaces:
         list_for_each_entry(entry, &smp2p->inbound, node)
                 irq_domain_remove(entry->domain);
@@ -565,6 +583,9 @@ static int qcom_smp2p_remove(struct platform_device *pdev)
         struct qcom_smp2p *smp2p = platform_get_drvdata(pdev);
         struct smp2p_entry *entry;
 
Quoted 6 lines

drivers/soc/qcom/smp2p.c static int qcom_smp2p_remove(struct platform_device *pdev) · line 587

Quoted 2 lines
drivers/soc/qcom/smp2p.c +2 -0
+       dev_pm_clear_wake_irq(&pdev->dev);
+       device_init_wakeup(&pdev->dev, false);

Is this device_init_wakeup() call necessary? It looks like we can get away without it and then once this driver probes the device will have the wakeup capability set on it. Future binding/unbinding of the driver will keep working.

ok, cleanup is handled in device_unregister() path. So it is redundant here.

I will remove this in next patch.