Skip to contentSkip to conversation

i3c: mipi-i3c-hci-pci: Add System Suspend support

PATCH Superseded Adrian Hunter 3 participants 10 messages Jan 16 – Mar 11 I3C

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Adrian Hunter

author
#0 lore raw

cover i3c: mipi-i3c-hci-pci: Add System Suspend support

Hi

Here are 3 small patches to add System Suspend support to mipi-i3c-hci and mipi-i3c-hci-pci.

Adrian Hunter (3):
i3c: master: Add i3c_master_restore_daa() for post-hibernation address recovery
i3c: mipi-i3c-hci: Add optional System Suspend support
i3c: mipi-i3c-hci-pci: Add System Suspend support

 drivers/i3c/master.c                               | 60 ++++++++++++++++------
 drivers/i3c/master/mipi-i3c-hci/core.c             | 53 +++++++++++++++++++
 drivers/i3c/master/mipi-i3c-hci/mipi-i3c-hci-pci.c |  7 +++
 include/linux/i3c/master.h                         |  1 +
 4 files changed, 104 insertions(+), 17 deletions(-)

Regards
Adrian

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
author
#1 lore raw

PATCH 1/3 i3c: master: Add i3c_master_restore_daa() for post-hibernation address recovery

After system hibernation, I3C Dynamic Addresses may be reassigned at boot and no longer match the values recorded before suspend. Introduce i3c_master_restore_daa() to handle this situation.

The restore procedure is straightforward: issue a Reset Dynamic Address Assignment (RSTDAA), then run the standard DAA sequence. The existing DAA logic already supports detecting and updating devices whose dynamic addresses differ from previously known values.

Refactor the DAA implementation by introducing a shared helper used by both i3c_master_do_daa() and the new restore function, and correct the kernel-doc in the process.

Export i3c_master_restore_daa() so that master drivers can invoke it from their PM restore callbacks.

Signed-off-by Adrian Hunter <[email protected]>
---
 drivers/i3c/master.c       | 60 +++++++++++++++++++++++++++-----------
 include/linux/i3c/master.h |  1 +
 2 files changed, 44 insertions(+), 17 deletions(-)
drivers/i3c/master.c +43 -17
index 594d61edcef4..1ba5d08365b4 100644
@@ -1734,23 +1734,9 @@ i3c_master_register_new_i3c_devs(struct i3c_master_controller *master)
 	}
 }
 
-/**
- * i3c_master_do_daa() - do a DAA (Dynamic Address Assignment)
- * @master: master doing the DAA
- *
- * This function is instantiating an I3C device object and adding it to the
- * I3C device list. All device information are automatically retrieved using
- * standard CCC commands.
- *
- * The I3C device object is returned in case the master wants to attach
- * private data to it using i3c_dev_set_master_data().
- *
- * This function must be called with the bus lock held in write mode.
- *
- * Return: a 0 in case of success, an negative error code otherwise.
- */
-int i3c_master_do_daa(struct i3c_master_controller *master)
+static int __i3c_master_do_daa(struct i3c_master_controller *master, bool reset)
 {
+	int rstret;
 	int ret;
 
 	ret = i3c_master_rpm_get(master);

@@ -1758,7 +1744,13 @@ int i3c_master_do_daa(struct i3c_master_controller *master)
 		return ret;
 
 	i3c_bus_maintenance_lock(&master->bus);
+
+	rstret = i3c_master_rstdaa_locked(master, I3C_BROADCAST_ADDR);
+	if (rstret == I3C_ERROR_M2)
+		rstret = 0;
+
 	ret = master->ops->do_daa(master);
+
 	i3c_bus_maintenance_unlock(&master->bus);
 
 	if (ret)

@@ -1769,10 +1761,44 @@ int i3c_master_do_daa(struct i3c_master_controller *master)
 	i3c_bus_normaluse_unlock(&master->bus);
 out:
 	i3c_master_rpm_put(master);
-	return ret;
+
+	return rstret ?: ret;
+}
+
+/**
+ * i3c_master_do_daa() - do a DAA (Dynamic Address Assignment)
+ * @master: master doing the DAA
+ *
+ * This function instantiates I3C device objects and adds them to the
+ * I3C device list. All device information is automatically retrieved using
+ * standard CCC commands.
+ *
+ * Return: a 0 in case of success, an negative error code otherwise.
+ */
+int i3c_master_do_daa(struct i3c_master_controller *master)
+{
+	return __i3c_master_do_daa(master, false);
 }
 EXPORT_SYMBOL_GPL(i3c_master_do_daa);
 
+/**
+ * i3c_master_restore_daa() - Restore Dynamic Addresses
+ * @master: controller
+ *
+ * After System Hibernation, Dynamic Addresses can have been reassigned at boot
+ * time to different values. A simple strategy is followed to handle that.
+ * Perform a Reset of Dynamic Address (RSTDAA) followed by the normal DAA
+ * procedure which has provision for reassigning addresses that differ from the
+ * previously recorded addresses.
+ *
+ * Return: a 0 in case of success, an negative error code otherwise.
+ */
+int i3c_master_restore_daa(struct i3c_master_controller *master)
+{
+	return __i3c_master_do_daa(master, true);
+}
+EXPORT_SYMBOL_GPL(i3c_master_restore_daa);
+
 /**
  * i3c_master_dma_map_single() - Map buffer for single DMA transfer
  * @dev: device object of a device doing DMA
include/linux/i3c/master.h +1 -0
index c1ec597f655c..ea757b92e440 100644
@@ -598,6 +598,7 @@ int i3c_master_get_free_addr(struct i3c_master_controller *master,
 int i3c_master_add_i3c_dev_locked(struct i3c_master_controller *master,
 				  u8 addr);
 int i3c_master_do_daa(struct i3c_master_controller *master);
+int i3c_master_restore_daa(struct i3c_master_controller *master);
 struct i3c_dma *i3c_master_dma_map_single(struct device *dev, void *ptr,
 					  size_t len, bool force_bounce,
 					  enum dma_data_direction dir);
Signature · 7 lines
2.51.0


-- 
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
reviewer
#2 lore raw

Re: [PATCH 1/3] i3c: master: Add i3c_master_restore_daa() for post-hibernation address recovery

Quoted 9 lines · Adrian Hunter

After system hibernation, I3C Dynamic Addresses may be reassigned at boot and no longer match the values recorded before suspend. Introduce i3c_master_restore_daa() to handle this situation.

The restore procedure is straightforward: issue a Reset Dynamic Address Assignment (RSTDAA), then run the standard DAA sequence. The existing DAA logic already supports detecting and updating devices whose dynamic addresses differ from previously known values.

Quoted 6 lines

Refactor the DAA implementation by introducing a shared helper used by both i3c_master_do_daa() and the new restore function, and correct the kernel-doc in the process.

Export i3c_master_restore_daa() so that master drivers can invoke it from their PM restore callbacks.

It is good. I3C Hub also need it.

Quoted 38 lines
Signed-off-by Adrian Hunter <[email protected]>
---
 drivers/i3c/master.c       | 60 +++++++++++++++++++++++++++-----------
 include/linux/i3c/master.h |  1 +
 2 files changed, 44 insertions(+), 17 deletions(-)
drivers/i3c/master.c +2 -16
index 594d61edcef4..1ba5d08365b4 100644
@@ -1734,23 +1734,9 @@ i3c_master_register_new_i3c_devs(struct i3c_master_controller *master)
 	}
 }
 
-/**
- * i3c_master_do_daa() - do a DAA (Dynamic Address Assignment)
- * @master: master doing the DAA
- *
- * This function is instantiating an I3C device object and adding it to the
- * I3C device list. All device information are automatically retrieved using
- * standard CCC commands.
- *
- * The I3C device object is returned in case the master wants to attach
- * private data to it using i3c_dev_set_master_data().
- *
- * This function must be called with the bus lock held in write mode.
- *
- * Return: a 0 in case of success, an negative error code otherwise.
- */
-int i3c_master_do_daa(struct i3c_master_controller *master)
+static int __i3c_master_do_daa(struct i3c_master_controller *master, bool reset)
 {
+	int rstret;
 	int ret;
 
 	ret = i3c_master_rpm_get(master);

drivers/i3c/master.c int i3c_master_do_daa(struct i3c_master_controller *master) · line 1750

Quoted 7 lines
drivers/i3c/master.c +4 -0
@@ -1758,7 +1744,13 @@ int i3c_master_do_daa(struct i3c_master_controller *master)
 		return ret;
 
 	i3c_bus_maintenance_lock(&master->bus);
+
+	rstret = i3c_master_rstdaa_locked(master, I3C_BROADCAST_ADDR);
+	if (rstret == I3C_ERROR_M2)
+		rstret = 0;

looks like you missed check 'reset' parameter.

Frank

Quoted 72 lines
drivers/i3c/master.c +37 -1
+
 	ret = master->ops->do_daa(master);
+
 	i3c_bus_maintenance_unlock(&master->bus);
 
 	if (ret)

@@ -1769,10 +1761,44 @@ int i3c_master_do_daa(struct i3c_master_controller *master)
 	i3c_bus_normaluse_unlock(&master->bus);
 out:
 	i3c_master_rpm_put(master);
-	return ret;
+
+	return rstret ?: ret;
+}
+
+/**
+ * i3c_master_do_daa() - do a DAA (Dynamic Address Assignment)
+ * @master: master doing the DAA
+ *
+ * This function instantiates I3C device objects and adds them to the
+ * I3C device list. All device information is automatically retrieved using
+ * standard CCC commands.
+ *
+ * Return: a 0 in case of success, an negative error code otherwise.
+ */
+int i3c_master_do_daa(struct i3c_master_controller *master)
+{
+	return __i3c_master_do_daa(master, false);
 }
 EXPORT_SYMBOL_GPL(i3c_master_do_daa);
 
+/**
+ * i3c_master_restore_daa() - Restore Dynamic Addresses
+ * @master: controller
+ *
+ * After System Hibernation, Dynamic Addresses can have been reassigned at boot
+ * time to different values. A simple strategy is followed to handle that.
+ * Perform a Reset of Dynamic Address (RSTDAA) followed by the normal DAA
+ * procedure which has provision for reassigning addresses that differ from the
+ * previously recorded addresses.
+ *
+ * Return: a 0 in case of success, an negative error code otherwise.
+ */
+int i3c_master_restore_daa(struct i3c_master_controller *master)
+{
+	return __i3c_master_do_daa(master, true);
+}
+EXPORT_SYMBOL_GPL(i3c_master_restore_daa);
+
 /**
  * i3c_master_dma_map_single() - Map buffer for single DMA transfer
  * @dev: device object of a device doing DMA
include/linux/i3c/master.h +1 -1
index c1ec597f655c..ea757b92e440 100644
@@ -598,6 +598,7 @@ int i3c_master_get_free_addr(struct i3c_master_controller *master,
 int i3c_master_add_i3c_dev_locked(struct i3c_master_controller *master,
 				  u8 addr);
 int i3c_master_do_daa(struct i3c_master_controller *master);
+int i3c_master_restore_daa(struct i3c_master_controller *master);
 struct i3c_dma *i3c_master_dma_map_single(struct device *dev, void *ptr,
 					  size_t len, bool force_bounce,
 					  enum dma_data_direction dir);
--

2.51.0

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
author
#3 lore raw

PATCH 2/3 i3c: mipi-i3c-hci: Add optional System Suspend support

In reply to Adrian Hunter

Add system suspend callbacks. Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Signed-off-by Adrian Hunter <[email protected]>
---
 drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++
 1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +53 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
 	return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+	int ret;
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret)
+		return ret;
+
+	if (restore)
+		ret = i3c_master_restore_daa(&hci->master);
+	else
+		ret = i3c_master_do_daa(&hci->master);
+
+	if (ret)
+		dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+	/*
+	 * I3C devices may have retained their dynamic address anyway. Do not
+	 * fail the resume because of DAA error.
+	 */
+	return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+	.suspend  = pm_sleep_ptr(i3c_hci_suspend),
+	.resume   = pm_sleep_ptr(i3c_hci_resume),
+	.freeze   = pm_sleep_ptr(i3c_hci_suspend),
+	.thaw     = pm_sleep_ptr(i3c_hci_resume),
+	.poweroff = pm_sleep_ptr(i3c_hci_suspend),
+	.restore  = pm_sleep_ptr(i3c_hci_restore),
 	RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };
Signature · 7 lines
2.51.0


-- 
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
reviewer
#4 lore raw

Re: [PATCH 2/3] i3c: mipi-i3c-hci: Add optional System Suspend support

Quoted 37 lines · Adrian Hunter

Add system suspend callbacks. Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Signed-off-by Adrian Hunter <[email protected]>
---
 drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++
 1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +14 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
 	return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+	int ret;

drivers/i3c/master/mipi-i3c-hci/core.c static int i3c_hci_runtime_resume(struct device *dev) · line 795

Quoted 10 lines
drivers/i3c/master/mipi-i3c-hci/core.c +10 -0
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret)
+		return ret;
+
+	if (restore)
+		ret = i3c_master_restore_daa(&hci->master);

Is it simpler if design API like

	i3c_master_do_daa_ext(&hci->master, restore);

Frank

Quoted 47 lines
drivers/i3c/master/mipi-i3c-hci/core.c +29 -1
+	else
+		ret = i3c_master_do_daa(&hci->master);
+
+	if (ret)
+		dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+	/*
+	 * I3C devices may have retained their dynamic address anyway. Do not
+	 * fail the resume because of DAA error.
+	 */
+	return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+	.suspend  = pm_sleep_ptr(i3c_hci_suspend),
+	.resume   = pm_sleep_ptr(i3c_hci_resume),
+	.freeze   = pm_sleep_ptr(i3c_hci_suspend),
+	.thaw     = pm_sleep_ptr(i3c_hci_resume),
+	.poweroff = pm_sleep_ptr(i3c_hci_suspend),
+	.restore  = pm_sleep_ptr(i3c_hci_restore),
 	RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };
 
--

2.51.0

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
#5 lore raw

Re: [PATCH 2/3] i3c: mipi-i3c-hci: Add optional System Suspend support

In reply to Adrian Hunter
Quoted 9 lines · Adrian Hunter

Add system suspend callbacks. Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Why do you make this distinction ?
RSTDAA+DAA during boot is motivated by that uboot might have assigned addresses. Why would we need to behave differently in suspend and suspend-to-disk? Its fair to assume that I3C devices might have been powered off in suspend too, right ?

Quoted 85 lines
Signed-off-by Adrian Hunter <[email protected]>

---
drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++

1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +53 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
 	return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+	struct i3c_hci *hci = dev_get_drvdata(dev);
+	int ret;
+
+	if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+		return 0;
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret)
+		return ret;
+
+	if (restore)
+		ret = i3c_master_restore_daa(&hci->master);
+	else
+		ret = i3c_master_do_daa(&hci->master);
+
+	if (ret)
+		dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+	/*
+	 * I3C devices may have retained their dynamic address anyway. Do not
+	 * fail the resume because of DAA error.
+	 */
+	return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+	return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+	.suspend  = pm_sleep_ptr(i3c_hci_suspend),
+	.resume   = pm_sleep_ptr(i3c_hci_resume),
+	.freeze   = pm_sleep_ptr(i3c_hci_suspend),
+	.thaw     = pm_sleep_ptr(i3c_hci_resume),
+	.poweroff = pm_sleep_ptr(i3c_hci_suspend),
+	.restore  = pm_sleep_ptr(i3c_hci_restore),
 	RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };
Signature · 7 lines
2.51.0


-- 
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
author
#6 lore raw

Re: [PATCH 2/3] i3c: mipi-i3c-hci: Add optional System Suspend support

Quoted 10 lines · David Nyström
Quoted 6 lines · Adrian Hunter

Add system suspend callbacks.  Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

Quoted 7 lines
Quoted 3 lines

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Why do you make this distinction ?
RSTDAA+DAA during boot is motivated by that uboot might have assigned addresses. Why would we need to behave differently in suspend and suspend-to-disk?

Because, in the case of suspend-to-disk, there is a complete boot process in between.

Quoted 1 line

Its fair to assume that I3C devices might have been powered off in suspend too, right ?

Hibernation is quite different. Briefly:

Entry

	ops->freeze
	create image
	ops->thaw
	write image
	ops->poweroff

Exit

	boot			<- address assignment not guaranteed to be the same as previous boot
	ops->freeze
	restore image
	ops->restore		<- deal with I3C devices that may have different addresses
Quoted 86 lines
Quoted 86 lines
Signed-off-by Adrian Hunter <[email protected]>

---
drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++

1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +53 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
     return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+    int ret;
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    ret = pm_runtime_force_resume(dev);
+    if (ret)
+        return ret;
+
+    if (restore)
+        ret = i3c_master_restore_daa(&hci->master);
+    else
+        ret = i3c_master_do_daa(&hci->master);
+
+    if (ret)
+        dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+    /*
+     * I3C devices may have retained their dynamic address anyway. Do not
+     * fail the resume because of DAA error.
+     */
+    return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+    .suspend  = pm_sleep_ptr(i3c_hci_suspend),
+    .resume   = pm_sleep_ptr(i3c_hci_resume),
+    .freeze   = pm_sleep_ptr(i3c_hci_suspend),
+    .thaw     = pm_sleep_ptr(i3c_hci_resume),
+    .poweroff = pm_sleep_ptr(i3c_hci_suspend),
+    .restore  = pm_sleep_ptr(i3c_hci_restore),
     RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };

--
2.51.0

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
#7 lore raw

Re: [PATCH 2/3] i3c: mipi-i3c-hci: Add optional System Suspend support

Quoted 33 lines · Adrian Hunter
Quoted 10 lines · David Nyström
Quoted 6 lines · Adrian Hunter

Add system suspend callbacks.  Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

Quoted 7 lines
Quoted 3 lines

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Why do you make this distinction ?
RSTDAA+DAA during boot is motivated by that uboot might have assigned addresses. Why would we need to behave differently in suspend and suspend-to-disk?

Because, in the case of suspend-to-disk, there is a complete boot process in between.

Quoted 1 line

Its fair to assume that I3C devices might have been powered off in suspend too, right ?

Hibernation is quite different. Briefly:

Entry

	ops->freeze
	create image
	ops->thaw
	write image
	ops->poweroff
Quoted 5 lines

Exit

	boot			<- address assignment not guaranteed to be the same as previous boot
	ops->freeze
	restore image
	ops->restore		<- deal with I3C devices that may have different addresses

OK, so bootloader might have messed things up for us again. Thanks for elaborating.

Quoted 93 lines
Quoted 86 lines
Quoted 86 lines
Signed-off-by Adrian Hunter <[email protected]>

---
drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++

1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +53 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
     return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+    int ret;
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    ret = pm_runtime_force_resume(dev);
+    if (ret)
+        return ret;
+
+    if (restore)
+        ret = i3c_master_restore_daa(&hci->master);
+    else
+        ret = i3c_master_do_daa(&hci->master);
+
+    if (ret)
+        dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+    /*
+     * I3C devices may have retained their dynamic address anyway. Do not
+     * fail the resume because of DAA error.
+     */
+    return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+    .suspend  = pm_sleep_ptr(i3c_hci_suspend),
+    .resume   = pm_sleep_ptr(i3c_hci_resume),
+    .freeze   = pm_sleep_ptr(i3c_hci_suspend),
+    .thaw     = pm_sleep_ptr(i3c_hci_resume),
+    .poweroff = pm_sleep_ptr(i3c_hci_suspend),
+    .restore  = pm_sleep_ptr(i3c_hci_restore),
     RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };

--
2.51.0

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
author
#8 lore raw

Re: [PATCH 2/3] i3c: mipi-i3c-hci: Add optional System Suspend support

Quoted 37 lines · David Nyström
Quoted 33 lines · Adrian Hunter
Quoted 10 lines · David Nyström
Quoted 6 lines · Adrian Hunter

Add system suspend callbacks.  Implement them by forcing runtime PM. Consequently bail out if Runtime PM is not allowed.

On resume from System Suspend (suspend to RAM), rerun Dynamic Address Assignment to restore addresses for devices that may have lost power.

Quoted 7 lines
Quoted 3 lines

On resume from System Hibernation (suspend to disk), use the new i3c_master_restore_daa() helper which additionally handles the case where devices are assigned different dynamic addresses after a hibernation boot.

Why do you make this distinction ?
RSTDAA+DAA during boot is motivated by that uboot might have assigned addresses. Why would we need to behave differently in suspend and suspend-to-disk?

Because, in the case of suspend-to-disk, there is a complete boot process in between.

Quoted 1 line

Its fair to assume that I3C devices might have been powered off in suspend too, right ?

Hibernation is quite different.  Briefly:

Entry
    ops->freeze
    create image
    ops->thaw
    write image
    ops->poweroff

Quoted 7 lines
Quoted 5 lines

Exit
    boot            <- address assignment not guaranteed to be the same as previous boot ops->freeze
    restore image
    ops->restore        <- deal with I3C devices that may have different addresses

OK, so bootloader might have messed things up for us again.

No, it is the kernel that cannot guarantee that an I3C device will have the same dynamic address from one boot to the next. Generally, the same address will be assigned. However the case I have experienced is the first I3C device does not come back after hibernation, and all the subsequent devices are given different addresses, and then none of them work.

Quoted 95 lines

Thanks for elaborating.

Quoted 93 lines
Quoted 81 lines
Quoted 81 lines
Signed-off-by Adrian Hunter <[email protected]>

---
drivers/i3c/master/mipi-i3c-hci/core.c | 53 ++++++++++++++++++++++++++

1 file changed, 53 insertions(+)
drivers/i3c/master/mipi-i3c-hci/core.c +53 -0
index 02c5a133e329..b972d8bc81e6 100644
@@ -769,6 +769,53 @@ static int i3c_hci_runtime_resume(struct device *dev)
     return 0;
 }
 
+static int i3c_hci_suspend(struct device *dev)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    return pm_runtime_force_suspend(dev);
+}
+
+static int i3c_hci_resume_common(struct device *dev, bool restore)
+{
+    struct i3c_hci *hci = dev_get_drvdata(dev);
+    int ret;
+
+    if (!(hci->quirks & HCI_QUIRK_RPM_ALLOWED))
+        return 0;
+
+    ret = pm_runtime_force_resume(dev);
+    if (ret)
+        return ret;
+
+    if (restore)
+        ret = i3c_master_restore_daa(&hci->master);
+    else
+        ret = i3c_master_do_daa(&hci->master);
+
+    if (ret)
+        dev_err(dev, "Dynamic Address Assignment failed on resume, error %d\n", ret);
+
+    /*
+     * I3C devices may have retained their dynamic address anyway. Do not
+     * fail the resume because of DAA error.
+     */
+    return 0;
+}
+
+static int i3c_hci_resume(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, false);
+}
+
+static int i3c_hci_restore(struct device *dev)
+{
+    return i3c_hci_resume_common(dev, true);
+}
+
 #define DEFAULT_AUTOSUSPEND_DELAY_MS 1000
 
 static void i3c_hci_rpm_enable(struct device *dev)

@@ -945,6 +992,12 @@ static const struct platform_device_id i3c_hci_driver_ids[] = {
 MODULE_DEVICE_TABLE(platform, i3c_hci_driver_ids);
 
 static const struct dev_pm_ops i3c_hci_pm_ops = {
+    .suspend  = pm_sleep_ptr(i3c_hci_suspend),
+    .resume   = pm_sleep_ptr(i3c_hci_resume),
+    .freeze   = pm_sleep_ptr(i3c_hci_suspend),
+    .thaw     = pm_sleep_ptr(i3c_hci_resume),
+    .poweroff = pm_sleep_ptr(i3c_hci_suspend),
+    .restore  = pm_sleep_ptr(i3c_hci_restore),
     RUNTIME_PM_OPS(i3c_hci_runtime_suspend, i3c_hci_runtime_resume, NULL)
 };

--
2.51.0

Quoted 5 lines
Quoted 5 lines

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

--
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c

Signature · 3 lines
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c
author
#9 lore raw

PATCH 3/3 i3c: mipi-i3c-hci-pci: Add System Suspend support

In reply to Adrian Hunter

Assign the driver PM operations pointer, which is necessary for the PCI subsystem to put the device into a low power state. Refer to pci_pm_suspend_noirq() which bails out if the pointer is NULL, before it has the opportunity to call pci_prepare_to_sleep().

No other actions are necessary as the mipi-i3c-hci driver takes care of controller state.

Signed-off-by Adrian Hunter <[email protected]>
---
 drivers/i3c/master/mipi-i3c-hci/mipi-i3c-hci-pci.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/i3c/master/mipi-i3c-hci/mipi-i3c-hci-pci.c +7 -0
index 1b38771667e5..0f05a15c14c7 100644
@@ -320,6 +320,10 @@ static void mipi_i3c_hci_pci_remove(struct pci_dev *pci)
 	mfd_remove_devices(&pci->dev);
 }
 
+/* PM ops must exist for PCI to put a device to a low power state */
+static const struct dev_pm_ops mipi_i3c_hci_pci_pm_ops = {
+};
+
 static const struct pci_device_id mipi_i3c_hci_pci_devices[] = {
 	/* Wildcat Lake-U */
 	{ PCI_VDEVICE(INTEL, 0x4d7c), (kernel_ulong_t)&intel_mi_1_info},

@@ -342,6 +346,9 @@ static struct pci_driver mipi_i3c_hci_pci_driver = {
 	.id_table = mipi_i3c_hci_pci_devices,
 	.probe = mipi_i3c_hci_pci_probe,
 	.remove = mipi_i3c_hci_pci_remove,
+	.driver = {
+		.pm = pm_ptr(&mipi_i3c_hci_pci_pm_ops)
+	},
 };
 
 module_pci_driver(mipi_i3c_hci_pci_driver);
Signature · 7 lines
2.51.0


-- 
linux-i3c mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-i3c