[PATCH 0/8] power_supply: Add API for safe access of get_property-like function attrs

[PATCH 0/8] power_supply: Add API for safe access of get_property-like function attrs

[Krzysztof Kozlowski]

Hi,


After fixing issue with referencing old power supply after driver
unbind in charger manager [1] I noticed that the race condition in such
case may still exist. It would be harder to trigger but still possible.

The race is between using a reference to power supply (obtained
with power_supply_get_by_name()) and removing the driver.

This patchset aims to fix the race by introducing wrappers for
accessing the power supply function attributes.

Patch 1 introduces new API. Other patches replace direct calls in
drivers.

Rebased on next-20141007.
Tested on Trats2 board (max77693 + charger manager).


Kindly asking for reviewing/testing the drivers, although the changes
are straightforward.


Best regards,
Krzysztof


[1] https://lkml.org/lkml/2014/10/13/272



Krzysztof Kozlowski (8):
  power_supply: Add API for safe access of power supply function attrs
  power_supply: sysfs: Use power_supply_*() API for accessing function
    attrs
  power: 88pm860x_charger: Use power_supply_*() API for accessing
    function attrs
  power: ab8500: Use power_supply_*() API for accessing function attrs
  mfd: ab8500: Use power_supply_*() API for accessing function attrs
  power: apm_power: Use power_supply_*() API for accessing function
    attrs
  power: bq2415x_charger: Use power_supply_*() API for accessing
    function attrs
  power: charger-manager: Use power_supply_*() API for accessing
    function attrs

 drivers/mfd/ab8500-sysctrl.c       |  7 +++--
 drivers/power/88pm860x_charger.c   | 13 +++++----
 drivers/power/ab8500_btemp.c       |  2 +-
 drivers/power/ab8500_charger.c     |  2 +-
 drivers/power/ab8500_fg.c          |  2 +-
 drivers/power/abx500_chargalg.c    |  4 +--
 drivers/power/apm_power.c          |  4 +--
 drivers/power/bq2415x_charger.c    |  3 +-
 drivers/power/charger-manager.c    | 37 +++++++++++++------------
 drivers/power/power_supply_core.c  | 57 ++++++++++++++++++++++++++++++++++++--
 drivers/power/power_supply_sysfs.c |  6 ++--
 include/linux/power_supply.h       | 16 +++++++++++
 12 files changed, 115 insertions(+), 38 deletions(-)

-- 
1.9.1

[PATCH 1/8] power_supply: Add API for safe access of power supply function attrs

[Krzysztof Kozlowski]

Add simple wrappers for accessing power supply's function attributes:
 - get_property -> power_supply_get_property
 - set_property -> power_supply_set_property
 - property_is_writeable -> power_supply_property_is_writeable
 - external_power_changed -> power_supply_external_power_changed
 - set_charged -> power_supply_set_charged

This API adds a safe way of accessing a power supply from another
driver. Particularly it solves invalid memory references (and lockup) in
following race condition scenario:

Thread 1: charger manager
Thread 2: power supply driver, used by charger manager

THREAD 1 (charger manager)         THREAD 2 (power supply driver)
==========================         ==============================
psy = power_supply_get_by_name()
                                   Driver unbind, .remove
                                     power_supply_unregister
                                     Device fully removed
psy->get_property()

The 'get_property' callback is still valid and leads to actual calling of
Thread2->get_property() but now in invalid context because the driver was
unbound.

This could be observed easily with charger manager driver (here compiled
with max17042 fuel gauge):
$ echo "12-0036" > /sys/bus/i2c/drivers/max17042/unbind
$ cat /sys/devices/virtual/power_supply/battery/temp
[  240.505998] INFO: task cat:1394 blocked for more than 120 seconds.
[  240.510762]       Not tainted 3.17.0-next-20141007-00028-ge60b6dd79570-dirty #183
[  240.518208] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  240.526025] cat             D c0469548     0  1394      1 0x00000000
[  240.532384] [<c0469548>] (__schedule) from [<c0469d54>] (schedule_preempt_disabled+0x14/0x20)
[  240.540885] [<c0469d54>] (schedule_preempt_disabled) from [<c046af20>] (mutex_lock_nested+0x1bc/0x458)
[  240.550171] [<c046af20>] (mutex_lock_nested) from [<c0287a98>] (regmap_read+0x30/0x60)
[  240.558079] [<c0287a98>] (regmap_read) from [<c032238c>] (max17042_get_property+0x2e8/0x350)
[  240.566490] [<c032238c>] (max17042_get_property) from [<c0324b60>] (charger_get_property+0x238/0x31c)
[  240.575688] [<c0324b60>] (charger_get_property) from [<c0320764>] (power_supply_show_property+0x48/0x1e0)
[  240.585241] [<c0320764>] (power_supply_show_property) from [<c027308c>] (dev_attr_show+0x1c/0x48)
[  240.594100] [<c027308c>] (dev_attr_show) from [<c0141fb0>] (sysfs_kf_seq_show+0x84/0x104)
[  240.602251] [<c0141fb0>] (sysfs_kf_seq_show) from [<c0140b18>] (kernfs_seq_show+0x24/0x28)
[  240.610497] [<c0140b18>] (kernfs_seq_show) from [<c0104574>] (seq_read+0x1b0/0x484)
[  240.618138] [<c0104574>] (seq_read) from [<c00e1e24>] (vfs_read+0x88/0x144)
[  240.625127] [<c00e1e24>] (vfs_read) from [<c00e1f20>] (SyS_read+0x40/0x8c)
[  240.631941] [<c00e1f20>] (SyS_read) from [<c000e760>] (ret_fast_syscall+0x0/0x48)

Or:
[   17.277605] Division by zero in kernel.
[   17.280043] CPU: 2 PID: 1384 Comm: cat Not tainted 3.17.0-next-20141007-00028-ge60b6dd79570-dirty #181
[   17.289348] [<c00140f0>] (unwind_backtrace) from [<c0011228>] (show_stack+0x10/0x14)
[   17.297055] [<c0011228>] (show_stack) from [<c04680d0>] (dump_stack+0x70/0xbc)
[   17.304264] [<c04680d0>] (dump_stack) from [<c01f7a28>] (Ldiv0+0x8/0x10)
[   17.310933] [<c01f7a28>] (Ldiv0) from [<c01f79f8>] (__aeabi_uidivmod+0x8/0x18)
[   17.318132] [<c01f79f8>] (__aeabi_uidivmod) from [<c0287a84>] (regmap_read+0x1c/0x60)
[   17.325956] [<c0287a84>] (regmap_read) from [<c0322260>] (max17042_get_property+0x1bc/0x350)
[   17.334361] [<c0322260>] (max17042_get_property) from [<c0324734>] (charger_get_property+0x198/0x328)
[   17.343560] [<c0324734>] (charger_get_property) from [<c0320764>] (power_supply_show_property+0x48/0x1e0)
[   17.353108] [<c0320764>] (power_supply_show_property) from [<c03209d4>] (power_supply_uevent+0x9c/0x1c4)
[   17.362575] [<c03209d4>] (power_supply_uevent) from [<c02745d0>] (dev_uevent+0xb8/0x1c8)
[   17.370639] [<c02745d0>] (dev_uevent) from [<c0272c4c>] (uevent_show+0xa8/0x104)
[   17.378015] [<c0272c4c>] (uevent_show) from [<c027308c>] (dev_attr_show+0x1c/0x48)
[   17.385579] [<c027308c>] (dev_attr_show) from [<c0141fb0>] (sysfs_kf_seq_show+0x84/0x104)
[   17.393731] [<c0141fb0>] (sysfs_kf_seq_show) from [<c0140b18>] (kernfs_seq_show+0x24/0x28)
[   17.401979] [<c0140b18>] (kernfs_seq_show) from [<c0104574>] (seq_read+0x1b0/0x484)
[   17.409619] [<c0104574>] (seq_read) from [<c00e1e24>] (vfs_read+0x88/0x144)
[   17.416557] [<c00e1e24>] (vfs_read) from [<c00e1f20>] (SyS_read+0x40/0x8c)
[   17.423416] [<c00e1f20>] (SyS_read) from [<c000e760>] (ret_fast_syscall+0x0/0x48)
[   17.430880] power_supply battery: driver failed to report `voltage_now' property: -22

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/power_supply_core.c | 57 +++++++++++++++++++++++++++++++++++++--
 include/linux/power_supply.h      | 16 +++++++++++
 2 files changed, 71 insertions(+), 2 deletions(-)

diff --git a/drivers/power/power_supply_core.c b/drivers/power/power_supply_core.c
index 6cb7fe5c022d..8e36986a73e1 100644
--- a/drivers/power/power_supply_core.c
+++ b/drivers/power/power_supply_core.c
@@ -366,6 +366,56 @@ struct power_supply *power_supply_get_by_phandle(struct device_node *np,
 EXPORT_SYMBOL_GPL(power_supply_get_by_phandle);
 #endif /* CONFIG_OF */
 
+int power_supply_get_property(struct power_supply *psy,
+			    enum power_supply_property psp,
+			    union power_supply_propval *val)
+{
+	if (!psy || !psy->dev)
+		return -ENODEV;
+
+	return psy->get_property(psy, psp, val);
+}
+EXPORT_SYMBOL_GPL(power_supply_get_property);
+
+int power_supply_set_property(struct power_supply *psy,
+			    enum power_supply_property psp,
+			    const union power_supply_propval *val)
+{
+	if (!psy || !psy->dev || !psy->set_property)
+		return -ENODEV;
+
+	return psy->set_property(psy, psp, val);
+}
+EXPORT_SYMBOL_GPL(power_supply_set_property);
+
+int power_supply_property_is_writeable(struct power_supply *psy,
+					enum power_supply_property psp)
+{
+	if (!psy || !psy->dev || !psy->property_is_writeable)
+		return -ENODEV;
+
+	return psy->property_is_writeable(psy, psp);
+}
+EXPORT_SYMBOL_GPL(power_supply_property_is_writeable);
+
+void power_supply_external_power_changed(struct power_supply *psy)
+{
+	if (!psy || !psy->dev || !psy->external_power_changed)
+		return;
+
+	psy->external_power_changed(psy);
+}
+EXPORT_SYMBOL_GPL(power_supply_external_power_changed);
+
+void power_supply_set_charged(struct power_supply *psy)
+{
+	if (!psy || !psy->dev || !psy->set_charged)
+		return;
+
+	psy->set_charged(psy);
+}
+EXPORT_SYMBOL_GPL(power_supply_set_charged);
+
 int power_supply_powers(struct power_supply *psy, struct device *dev)
 {
 	return sysfs_create_link(&psy->dev->kobj, &dev->kobj, "powers");
@@ -616,13 +666,16 @@ EXPORT_SYMBOL_GPL(power_supply_register_no_ws);
 
 void power_supply_unregister(struct power_supply *psy)
 {
+	struct device *dev = psy->dev;
+
 	cancel_work_sync(&psy->changed_work);
 	sysfs_remove_link(&psy->dev->kobj, "powers");
+	psy->dev = NULL;
 	power_supply_remove_triggers(psy);
 	psy_unregister_cooler(psy);
 	psy_unregister_thermal(psy);
-	device_init_wakeup(psy->dev, false);
-	device_unregister(psy->dev);
+	device_init_wakeup(dev, false);
+	device_unregister(dev);
 }
 EXPORT_SYMBOL_GPL(power_supply_unregister);
 
diff --git a/include/linux/power_supply.h b/include/linux/power_supply.h
index 3ed049673022..44e749c65505 100644
--- a/include/linux/power_supply.h
+++ b/include/linux/power_supply.h
@@ -189,6 +189,12 @@ struct power_supply {
 	size_t num_supplies;
 	struct device_node *of_node;
 
+	/*
+	 * Functions for drivers implementing power supply class.
+	 * These shouldn't be called directly by other drivers for accessing
+	 * this power supply. Instead use power_supply_*() functions (for
+	 * example power_supply_get_property()).
+	 */
 	int (*get_property)(struct power_supply *psy,
 			    enum power_supply_property psp,
 			    union power_supply_propval *val);
@@ -268,6 +274,16 @@ extern int power_supply_is_system_supplied(void);
 static inline int power_supply_is_system_supplied(void) { return -ENOSYS; }
 #endif
 
+extern int power_supply_get_property(struct power_supply *psy,
+			    enum power_supply_property psp,
+			    union power_supply_propval *val);
+extern int power_supply_set_property(struct power_supply *psy,
+			    enum power_supply_property psp,
+			    const union power_supply_propval *val);
+extern int power_supply_property_is_writeable(struct power_supply *psy,
+					enum power_supply_property psp);
+extern void power_supply_external_power_changed(struct power_supply *psy);
+extern void power_supply_set_charged(struct power_supply *psy);
 extern int power_supply_register(struct device *parent,
 				 struct power_supply *psy);
 extern int power_supply_register_no_ws(struct device *parent,
-- 
1.9.1

[PATCH 2/8] power_supply: sysfs: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property
 - set_property -> power_supply_set_property
 - property_is_writeable -> power_supply_property_is_writeable

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/power_supply_sysfs.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/power/power_supply_sysfs.c b/drivers/power/power_supply_sysfs.c
index 62653f50a524..f817aab80813 100644
--- a/drivers/power/power_supply_sysfs.c
+++ b/drivers/power/power_supply_sysfs.c
@@ -76,7 +76,7 @@ static ssize_t power_supply_show_property(struct device *dev,
 	if (off == POWER_SUPPLY_PROP_TYPE) {
 		value.intval = psy->type;
 	} else {
-		ret = psy->get_property(psy, off, &value);
+		ret = power_supply_get_property(psy, off, &value);
 
 		if (ret < 0) {
 			if (ret == -ENODATA)
@@ -125,7 +125,7 @@ static ssize_t power_supply_store_property(struct device *dev,
 
 	value.intval = long_val;
 
-	ret = psy->set_property(psy, off, &value);
+	ret = power_supply_set_property(psy, off, &value);
 	if (ret < 0)
 		return ret;
 
@@ -223,7 +223,7 @@ static umode_t power_supply_attr_is_visible(struct kobject *kobj,
 
 		if (property == attrno) {
 			if (psy->property_is_writeable &&
-			    psy->property_is_writeable(psy, property) > 0)
+			    power_supply_property_is_writeable(psy, property) > 0)
 				mode |= S_IWUSR;
 
 			return mode;
-- 
1.9.1

[PATCH 3/8] power: 88pm860x_charger: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property
 - set_property -> power_supply_set_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/88pm860x_charger.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/drivers/power/88pm860x_charger.c b/drivers/power/88pm860x_charger.c
index de029bbc1cc1..534b3e2e3487 100644
--- a/drivers/power/88pm860x_charger.c
+++ b/drivers/power/88pm860x_charger.c
@@ -296,12 +296,13 @@ static int set_charging_fsm(struct pm860x_charger_info *info)
 	psy = power_supply_get_by_name(pm860x_supplied_to[0]);
 	if (!psy)
 		return -EINVAL;
-	ret = psy->get_property(psy, POWER_SUPPLY_PROP_VOLTAGE_NOW, &data);
+	ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_VOLTAGE_NOW,
+			&data);
 	if (ret)
 		return ret;
 	vbatt = data.intval / 1000;
 
-	ret = psy->get_property(psy, POWER_SUPPLY_PROP_PRESENT, &data);
+	ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_PRESENT, &data);
 	if (ret)
 		return ret;
 
@@ -430,7 +431,7 @@ static irqreturn_t pm860x_temp_handler(int irq, void *data)
 	psy = power_supply_get_by_name(pm860x_supplied_to[0]);
 	if (!psy)
 		goto out;
-	ret = psy->get_property(psy, POWER_SUPPLY_PROP_TEMP, &temp);
+	ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_TEMP, &temp);
 	if (ret)
 		goto out;
 	value = temp.intval / 10;
@@ -485,7 +486,8 @@ static irqreturn_t pm860x_done_handler(int irq, void *data)
 	psy = power_supply_get_by_name(pm860x_supplied_to[0]);
 	if (!psy)
 		goto out;
-	ret = psy->get_property(psy, POWER_SUPPLY_PROP_VOLTAGE_NOW, &val);
+	ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_VOLTAGE_NOW,
+			&val);
 	if (ret)
 		goto out;
 	vbatt = val.intval / 1000;
@@ -500,7 +502,8 @@ static irqreturn_t pm860x_done_handler(int irq, void *data)
 	if (ret < 0)
 		goto out;
 	if (vbatt > CHARGE_THRESHOLD && ret & STATUS2_CHG)
-		psy->set_property(psy, POWER_SUPPLY_PROP_CHARGE_FULL, &val);
+		power_supply_set_property(psy, POWER_SUPPLY_PROP_CHARGE_FULL,
+				&val);
 
 out:
 	mutex_unlock(&info->lock);
-- 
1.9.1

[PATCH 4/8] power: ab8500: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/ab8500_btemp.c    | 2 +-
 drivers/power/ab8500_charger.c  | 2 +-
 drivers/power/ab8500_fg.c       | 2 +-
 drivers/power/abx500_chargalg.c | 4 ++--
 4 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/power/ab8500_btemp.c b/drivers/power/ab8500_btemp.c
index 7f9a4547dccd..8161960ee78c 100644
--- a/drivers/power/ab8500_btemp.c
+++ b/drivers/power/ab8500_btemp.c
@@ -938,7 +938,7 @@ static int ab8500_btemp_get_ext_psy_data(struct device *dev, void *data)
 		enum power_supply_property prop;
 		prop = ext->properties[j];
 
-		if (ext->get_property(ext, prop, &ret))
+		if (power_supply_get_property(ext, prop, &ret))
 			continue;
 
 		switch (prop) {
diff --git a/drivers/power/ab8500_charger.c b/drivers/power/ab8500_charger.c
index 19110aa613a1..38d61523227f 100644
--- a/drivers/power/ab8500_charger.c
+++ b/drivers/power/ab8500_charger.c
@@ -1957,7 +1957,7 @@ static int ab8500_charger_get_ext_psy_data(struct device *dev, void *data)
 		enum power_supply_property prop;
 		prop = ext->properties[j];
 
-		if (ext->get_property(ext, prop, &ret))
+		if (power_supply_get_property(ext, prop, &ret))
 			continue;
 
 		switch (prop) {
diff --git a/drivers/power/ab8500_fg.c b/drivers/power/ab8500_fg.c
index 217da4b2ca86..ed2aca42f968 100644
--- a/drivers/power/ab8500_fg.c
+++ b/drivers/power/ab8500_fg.c
@@ -2199,7 +2199,7 @@ static int ab8500_fg_get_ext_psy_data(struct device *dev, void *data)
 		enum power_supply_property prop;
 		prop = ext->properties[j];
 
-		if (ext->get_property(ext, prop, &ret))
+		if (power_supply_get_property(ext, prop, &ret))
 			continue;
 
 		switch (prop) {
diff --git a/drivers/power/abx500_chargalg.c b/drivers/power/abx500_chargalg.c
index 6d2723664a01..f3d9dbac3568 100644
--- a/drivers/power/abx500_chargalg.c
+++ b/drivers/power/abx500_chargalg.c
@@ -1001,7 +1001,7 @@ static int abx500_chargalg_get_ext_psy_data(struct device *dev, void *data)
 	 * property because of handling that sysfs entry on its own, this is
 	 * the place to get the battery capacity.
 	 */
-	if (!ext->get_property(ext, POWER_SUPPLY_PROP_CAPACITY, &ret)) {
+	if (!power_supply_get_property(ext, POWER_SUPPLY_PROP_CAPACITY, &ret)) {
 		di->batt_data.percent = ret.intval;
 		capacity_updated = true;
 	}
@@ -1019,7 +1019,7 @@ static int abx500_chargalg_get_ext_psy_data(struct device *dev, void *data)
 			ext->type == POWER_SUPPLY_TYPE_USB)
 			di->usb_chg = psy_to_ux500_charger(ext);
 
-		if (ext->get_property(ext, prop, &ret))
+		if (power_supply_get_property(ext, prop, &ret))
 			continue;
 		switch (prop) {
 		case POWER_SUPPLY_PROP_PRESENT:
-- 
1.9.1

[PATCH 5/8] mfd: ab8500: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/mfd/ab8500-sysctrl.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/mfd/ab8500-sysctrl.c b/drivers/mfd/ab8500-sysctrl.c
index 8e0dae59844d..93b2d2c32ca3 100644
--- a/drivers/mfd/ab8500-sysctrl.c
+++ b/drivers/mfd/ab8500-sysctrl.c
@@ -49,7 +49,8 @@ static void ab8500_power_off(void)
 		if (!psy)
 			continue;
 
-		ret = psy->get_property(psy, POWER_SUPPLY_PROP_ONLINE, &val);
+		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_ONLINE,
+				&val);
 
 		if (!ret && val.intval) {
 			charger_present = true;
@@ -63,8 +64,8 @@ static void ab8500_power_off(void)
 	/* Check if battery is known */
 	psy = power_supply_get_by_name("ab8500_btemp");
 	if (psy) {
-		ret = psy->get_property(psy, POWER_SUPPLY_PROP_TECHNOLOGY,
-					&val);
+		ret = power_supply_get_property(psy,
+				POWER_SUPPLY_PROP_TECHNOLOGY, &val);
 		if (!ret && val.intval != POWER_SUPPLY_TECHNOLOGY_UNKNOWN) {
 			printk(KERN_INFO
 			       "Charger \"%s\" is connected with known battery."
-- 
1.9.1

[PATCH 6/8] power: apm_power: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/apm_power.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/power/apm_power.c b/drivers/power/apm_power.c
index 39763015b360..2206f9ff6488 100644
--- a/drivers/power/apm_power.c
+++ b/drivers/power/apm_power.c
@@ -15,10 +15,10 @@
 #include <linux/apm-emulation.h>
 
 
-#define PSY_PROP(psy, prop, val) (psy->get_property(psy, \
+#define PSY_PROP(psy, prop, val) (power_supply_get_property(psy, \
 			 POWER_SUPPLY_PROP_##prop, val))
 
-#define _MPSY_PROP(prop, val) (main_battery->get_property(main_battery, \
+#define _MPSY_PROP(prop, val) (power_supply_get_property(main_battery, \
 							 prop, val))
 
 #define MPSY_PROP(prop, val) _MPSY_PROP(POWER_SUPPLY_PROP_##prop, val)
-- 
1.9.1

[PATCH 7/8] power: bq2415x_charger: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/bq2415x_charger.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/power/bq2415x_charger.c b/drivers/power/bq2415x_charger.c
index d9c457217b6a..81a3a78b92d8 100644
--- a/drivers/power/bq2415x_charger.c
+++ b/drivers/power/bq2415x_charger.c
@@ -816,7 +816,8 @@ static int bq2415x_notifier_call(struct notifier_block *nb,
 
 	dev_dbg(bq->dev, "notifier call was called\n");
 
-	ret = psy->get_property(psy, POWER_SUPPLY_PROP_CURRENT_MAX, &prop);
+	ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_CURRENT_MAX,
+			&prop);
 	if (ret != 0)
 		return NOTIFY_OK;
 
-- 
1.9.1

[PATCH 8/8] power: charger-manager: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

Replace direct calls to power supply function attributes with wrappers.
Wrappers provide safe access access in case of unregistering the power
supply (.e.g by removing the driver). Replace:
 - get_property -> power_supply_get_property

Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
---
 drivers/power/charger-manager.c | 37 ++++++++++++++++++++-----------------
 1 file changed, 20 insertions(+), 17 deletions(-)

diff --git a/drivers/power/charger-manager.c b/drivers/power/charger-manager.c
index b5d272cb7c4a..9e01074d3169 100644
--- a/drivers/power/charger-manager.c
+++ b/drivers/power/charger-manager.c
@@ -112,8 +112,8 @@ static bool is_batt_present(struct charger_manager *cm)
 		if (!psy)
 			break;
 
-		ret = psy->get_property(psy,
-				POWER_SUPPLY_PROP_PRESENT, &val);
+		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_PRESENT,
+				&val);
 		if (ret == 0 && val.intval)
 			present = true;
 		break;
@@ -127,8 +127,8 @@ static bool is_batt_present(struct charger_manager *cm)
 				continue;
 			}
 
-			ret = psy->get_property(psy, POWER_SUPPLY_PROP_PRESENT,
-					&val);
+			ret = power_supply_get_property(psy,
+				POWER_SUPPLY_PROP_PRESENT, &val);
 			if (ret == 0 && val.intval) {
 				present = true;
 				break;
@@ -164,7 +164,8 @@ static bool is_ext_pwr_online(struct charger_manager *cm)
 			continue;
 		}
 
-		ret = psy->get_property(psy, POWER_SUPPLY_PROP_ONLINE, &val);
+		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_ONLINE,
+				&val);
 		if (ret == 0 && val.intval) {
 			online = true;
 			break;
@@ -192,7 +193,7 @@ static int get_batt_uV(struct charger_manager *cm, int *uV)
 	if (!fuel_gauge)
 		return -ENODEV;
 
-	ret = fuel_gauge->get_property(fuel_gauge,
+	ret = power_supply_get_property(fuel_gauge,
 				POWER_SUPPLY_PROP_VOLTAGE_NOW, &val);
 	if (ret)
 		return ret;
@@ -232,7 +233,8 @@ static bool is_charging(struct charger_manager *cm)
 		}
 
 		/* 2. The charger should be online (ext-power) */
-		ret = psy->get_property(psy, POWER_SUPPLY_PROP_ONLINE, &val);
+		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_ONLINE,
+				&val);
 		if (ret) {
 			dev_warn(cm->dev, "Cannot read ONLINE value from %s\n",
 				 cm->desc->psy_charger_stat[i]);
@@ -245,7 +247,8 @@ static bool is_charging(struct charger_manager *cm)
 		 * 3. The charger should not be FULL, DISCHARGING,
 		 * or NOT_CHARGING.
 		 */
-		ret = psy->get_property(psy, POWER_SUPPLY_PROP_STATUS, &val);
+		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_STATUS,
+				&val);
 		if (ret) {
 			dev_warn(cm->dev, "Cannot read STATUS value from %s\n",
 				 cm->desc->psy_charger_stat[i]);
@@ -288,7 +291,7 @@ static bool is_full_charged(struct charger_manager *cm)
 		val.intval = 0;
 
 		/* Not full if capacity of fuel gauge isn't full */
-		ret = fuel_gauge->get_property(fuel_gauge,
+		ret = power_supply_get_property(fuel_gauge,
 				POWER_SUPPLY_PROP_CHARGE_FULL, &val);
 		if (!ret && val.intval > desc->fullbatt_full_capacity)
 			return true;
@@ -305,7 +308,7 @@ static bool is_full_charged(struct charger_manager *cm)
 	if (desc->fullbatt_soc > 0) {
 		val.intval = 0;
 
-		ret = fuel_gauge->get_property(fuel_gauge,
+		ret = power_supply_get_property(fuel_gauge,
 				POWER_SUPPLY_PROP_CAPACITY, &val);
 		if (!ret && val.intval >= desc->fullbatt_soc)
 			return true;
@@ -589,7 +592,7 @@ static int cm_get_battery_temperature_by_psy(struct charger_manager *cm,
 	if (!fuel_gauge)
 		return -ENODEV;
 
-	return fuel_gauge->get_property(fuel_gauge,
+	return power_supply_get_property(fuel_gauge,
 				POWER_SUPPLY_PROP_TEMP,
 				(union power_supply_propval *)temp);
 }
@@ -909,7 +912,7 @@ static int charger_get_property(struct power_supply *psy,
 			ret = -ENODEV;
 			break;
 		}
-		ret = fuel_gauge->get_property(fuel_gauge,
+		ret = power_supply_get_property(fuel_gauge,
 				POWER_SUPPLY_PROP_CURRENT_NOW, val);
 		break;
 	case POWER_SUPPLY_PROP_TEMP:
@@ -928,7 +931,7 @@ static int charger_get_property(struct power_supply *psy,
 			break;
 		}
 
-		ret = fuel_gauge->get_property(fuel_gauge,
+		ret = power_supply_get_property(fuel_gauge,
 					POWER_SUPPLY_PROP_CAPACITY, val);
 		if (ret)
 			break;
@@ -984,7 +987,7 @@ static int charger_get_property(struct power_supply *psy,
 				break;
 			}
 
-			ret = fuel_gauge->get_property(fuel_gauge,
+			ret = power_supply_get_property(fuel_gauge,
 						POWER_SUPPLY_PROP_CHARGE_NOW,
 						val);
 			if (ret) {
@@ -1553,7 +1556,7 @@ static int cm_init_thermal_data(struct charger_manager *cm,
 	int ret;
 
 	/* Verify whether fuel gauge provides battery temperature */
-	ret = fuel_gauge->get_property(fuel_gauge,
+	ret = power_supply_get_property(fuel_gauge,
 					POWER_SUPPLY_PROP_TEMP, &val);
 
 	if (!ret) {
@@ -1845,13 +1848,13 @@ static int charger_manager_probe(struct platform_device *pdev)
 	cm->charger_psy.num_properties = psy_default.num_properties;
 
 	/* Find which optional psy-properties are available */
-	if (!fuel_gauge->get_property(fuel_gauge,
+	if (!power_supply_get_property(fuel_gauge,
 					  POWER_SUPPLY_PROP_CHARGE_NOW, &val)) {
 		cm->charger_psy.properties[cm->charger_psy.num_properties] =
 				POWER_SUPPLY_PROP_CHARGE_NOW;
 		cm->charger_psy.num_properties++;
 	}
-	if (!fuel_gauge->get_property(fuel_gauge,
+	if (!power_supply_get_property(fuel_gauge,
 					  POWER_SUPPLY_PROP_CURRENT_NOW,
 					  &val)) {
 		cm->charger_psy.properties[cm->charger_psy.num_properties] =
-- 
1.9.1

[PATCH 0/8] power_supply: Add API for safe access of get_property-like function attrs

[jonghwa3.lee at samsung.com]

Hi,
On 2014? 10? 14? 21:20, Krzysztof Kozlowski wrote:

> Hi,
> 
> 
> After fixing issue with referencing old power supply after driver
> unbind in charger manager [1] I noticed that the race condition in such
> case may still exist. It would be harder to trigger but still possible.
> 
> The race is between using a reference to power supply (obtained
> with power_supply_get_by_name()) and removing the driver.
> 
> This patchset aims to fix the race by introducing wrappers for
> accessing the power supply function attributes.
> 
> Patch 1 introduces new API. Other patches replace direct calls in
> drivers.
> 
> Rebased on next-20141007.
> Tested on Trats2 board (max77693 + charger manager).
> 
> 
> Kindly asking for reviewing/testing the drivers, although the changes
> are straightforward.
> 


Looks good to me, but need someone to comment on this.

Acked-by : Jonghwa Lee <jonghwa3.lee@samusng.com>

Thanks,
Jonghwa

> 
> Best regards,
> Krzysztof
> 
> 
> [1] https://lkml.org/lkml/2014/10/13/272
> 
> 
> 
> Krzysztof Kozlowski (8):
>   power_supply: Add API for safe access of power supply function attrs
>   power_supply: sysfs: Use power_supply_*() API for accessing function
>     attrs
>   power: 88pm860x_charger: Use power_supply_*() API for accessing
>     function attrs
>   power: ab8500: Use power_supply_*() API for accessing function attrs
>   mfd: ab8500: Use power_supply_*() API for accessing function attrs
>   power: apm_power: Use power_supply_*() API for accessing function
>     attrs
>   power: bq2415x_charger: Use power_supply_*() API for accessing
>     function attrs
>   power: charger-manager: Use power_supply_*() API for accessing
>     function attrs
> 
>  drivers/mfd/ab8500-sysctrl.c       |  7 +++--
>  drivers/power/88pm860x_charger.c   | 13 +++++----
>  drivers/power/ab8500_btemp.c       |  2 +-
>  drivers/power/ab8500_charger.c     |  2 +-
>  drivers/power/ab8500_fg.c          |  2 +-
>  drivers/power/abx500_chargalg.c    |  4 +--
>  drivers/power/apm_power.c          |  4 +--
>  drivers/power/bq2415x_charger.c    |  3 +-
>  drivers/power/charger-manager.c    | 37 +++++++++++++------------
>  drivers/power/power_supply_core.c  | 57 ++++++++++++++++++++++++++++++++++++--
>  drivers/power/power_supply_sysfs.c |  6 ++--
>  include/linux/power_supply.h       | 16 +++++++++++
>  12 files changed, 115 insertions(+), 38 deletions(-)
> 

[PATCH 5/8] mfd: ab8500: Use power_supply_*() API for accessing function attrs

[Lee Jones]

On Tue, 14 Oct 2014, Krzysztof Kozlowski wrote:

> Replace direct calls to power supply function attributes with wrappers.
> Wrappers provide safe access access in case of unregistering the power
> supply (.e.g by removing the driver). Replace:
>  - get_property -> power_supply_get_property
> 
> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
> ---
>  drivers/mfd/ab8500-sysctrl.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)

Acked-by: Lee Jones <lee.jones@linaro.org>

> diff --git a/drivers/mfd/ab8500-sysctrl.c b/drivers/mfd/ab8500-sysctrl.c
> index 8e0dae59844d..93b2d2c32ca3 100644
> --- a/drivers/mfd/ab8500-sysctrl.c
> +++ b/drivers/mfd/ab8500-sysctrl.c
> @@ -49,7 +49,8 @@ static void ab8500_power_off(void)
>  		if (!psy)
>  			continue;
>  
> -		ret = psy->get_property(psy, POWER_SUPPLY_PROP_ONLINE, &val);
> +		ret = power_supply_get_property(psy, POWER_SUPPLY_PROP_ONLINE,
> +				&val);
>  
>  		if (!ret && val.intval) {
>  			charger_present = true;
> @@ -63,8 +64,8 @@ static void ab8500_power_off(void)
>  	/* Check if battery is known */
>  	psy = power_supply_get_by_name("ab8500_btemp");
>  	if (psy) {
> -		ret = psy->get_property(psy, POWER_SUPPLY_PROP_TECHNOLOGY,
> -					&val);
> +		ret = power_supply_get_property(psy,
> +				POWER_SUPPLY_PROP_TECHNOLOGY, &val);
>  		if (!ret && val.intval != POWER_SUPPLY_TECHNOLOGY_UNKNOWN) {
>  			printk(KERN_INFO
>  			       "Charger \"%s\" is connected with known battery."

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org ? Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog

[PATCH 1/8] power_supply: Add API for safe access of power supply function attrs

[Pavel Machek]

On Tue 2014-10-14 14:20:39, Krzysztof Kozlowski wrote:
> Add simple wrappers for accessing power supply's function attributes:
>  - get_property -> power_supply_get_property
>  - set_property -> power_supply_set_property
>  - property_is_writeable -> power_supply_property_is_writeable
>  - external_power_changed -> power_supply_external_power_changed
>  - set_charged -> power_supply_set_charged
> 
> This API adds a safe way of accessing a power supply from another
> driver. Particularly it solves invalid memory references (and lockup) in
> following race condition scenario:
> 
> Thread 1: charger manager
> Thread 2: power supply driver, used by charger manager
> 
> THREAD 1 (charger manager)         THREAD 2 (power supply driver)
> ==========================         ==============================
> psy = power_supply_get_by_name()
>                                    Driver unbind, .remove
>                                      power_supply_unregister
>                                      Device fully removed
> psy->get_property()
> 
> The 'get_property' callback is still valid and leads to actual calling of
> Thread2->get_property() but now in invalid context because the driver was
> unbound.
> 
> This could be observed easily with charger manager driver (here compiled
> with max17042 fuel gauge):
> $ echo "12-0036" > /sys/bus/i2c/drivers/max17042/unbind
> $ cat /sys/devices/virtual/power_supply/battery/temp
> [  240.505998] INFO: task cat:1394 blocked for more than 120 seconds.
> [  240.510762]       Not tainted 3.17.0-next-20141007-00028-ge60b6dd79570-dirty #183
> [  240.518208] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [  240.526025] cat             D c0469548     0  1394      1 0x00000000
> [  240.532384] [<c0469548>] (__schedule) from [<c0469d54>] (schedule_preempt_disabled+0x14/0x20)
> [  240.540885] [<c0469d54>] (schedule_preempt_disabled) from [<c046af20>] (mutex_lock_nested+0x1bc/0x458)
> [  240.550171] [<c046af20>] (mutex_lock_nested) from [<c0287a98>] (regmap_read+0x30/0x60)
> [  240.558079] [<c0287a98>] (regmap_read) from [<c032238c>] (max17042_get_property+0x2e8/0x350)
> [  240.566490] [<c032238c>] (max17042_get_property) from [<c0324b60>] (charger_get_property+0x238/0x31c)
> [  240.575688] [<c0324b60>] (charger_get_property) from [<c0320764>] (power_supply_show_property+0x48/0x1e0)
> [  240.585241] [<c0320764>] (power_supply_show_property) from [<c027308c>] (dev_attr_show+0x1c/0x48)
> [  240.594100] [<c027308c>] (dev_attr_show) from [<c0141fb0>] (sysfs_kf_seq_show+0x84/0x104)
> [  240.602251] [<c0141fb0>] (sysfs_kf_seq_show) from [<c0140b18>] (kernfs_seq_show+0x24/0x28)
> [  240.610497] [<c0140b18>] (kernfs_seq_show) from [<c0104574>] (seq_read+0x1b0/0x484)
> [  240.618138] [<c0104574>] (seq_read) from [<c00e1e24>] (vfs_read+0x88/0x144)
> [  240.625127] [<c00e1e24>] (vfs_read) from [<c00e1f20>] (SyS_read+0x40/0x8c)
> [  240.631941] [<c00e1f20>] (SyS_read) from [<c000e760>] (ret_fast_syscall+0x0/0x48)
> 
> Or:
> [   17.277605] Division by zero in kernel.
> [   17.280043] CPU: 2 PID: 1384 Comm: cat Not tainted 3.17.0-next-20141007-00028-ge60b6dd79570-dirty #181
> [   17.289348] [<c00140f0>] (unwind_backtrace) from [<c0011228>] (show_stack+0x10/0x14)
> [   17.297055] [<c0011228>] (show_stack) from [<c04680d0>] (dump_stack+0x70/0xbc)
> [   17.304264] [<c04680d0>] (dump_stack) from [<c01f7a28>] (Ldiv0+0x8/0x10)
> [   17.310933] [<c01f7a28>] (Ldiv0) from [<c01f79f8>] (__aeabi_uidivmod+0x8/0x18)
> [   17.318132] [<c01f79f8>] (__aeabi_uidivmod) from [<c0287a84>] (regmap_read+0x1c/0x60)
> [   17.325956] [<c0287a84>] (regmap_read) from [<c0322260>] (max17042_get_property+0x1bc/0x350)
> [   17.334361] [<c0322260>] (max17042_get_property) from [<c0324734>] (charger_get_property+0x198/0x328)
> [   17.343560] [<c0324734>] (charger_get_property) from [<c0320764>] (power_supply_show_property+0x48/0x1e0)
> [   17.353108] [<c0320764>] (power_supply_show_property) from [<c03209d4>] (power_supply_uevent+0x9c/0x1c4)
> [   17.362575] [<c03209d4>] (power_supply_uevent) from [<c02745d0>] (dev_uevent+0xb8/0x1c8)
> [   17.370639] [<c02745d0>] (dev_uevent) from [<c0272c4c>] (uevent_show+0xa8/0x104)
> [   17.378015] [<c0272c4c>] (uevent_show) from [<c027308c>] (dev_attr_show+0x1c/0x48)
> [   17.385579] [<c027308c>] (dev_attr_show) from [<c0141fb0>] (sysfs_kf_seq_show+0x84/0x104)
> [   17.393731] [<c0141fb0>] (sysfs_kf_seq_show) from [<c0140b18>] (kernfs_seq_show+0x24/0x28)
> [   17.401979] [<c0140b18>] (kernfs_seq_show) from [<c0104574>] (seq_read+0x1b0/0x484)
> [   17.409619] [<c0104574>] (seq_read) from [<c00e1e24>] (vfs_read+0x88/0x144)
> [   17.416557] [<c00e1e24>] (vfs_read) from [<c00e1f20>] (SyS_read+0x40/0x8c)
> [   17.423416] [<c00e1f20>] (SyS_read) from [<c000e760>] (ret_fast_syscall+0x0/0x48)
> [   17.430880] power_supply battery: driver failed to report `voltage_now' property: -22
> 
> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>

Acked-by: Pavel Machek <pavel@ucw.cz>

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[PATCH 2/8] power_supply: sysfs: Use power_supply_*() API for accessing function attrs

[Pavel Machek]

On Tue 2014-10-14 14:20:40, Krzysztof Kozlowski wrote:
> Replace direct calls to power supply function attributes with wrappers.
> Wrappers provide safe access access in case of unregistering the power
> supply (.e.g by removing the driver). Replace:
>  - get_property -> power_supply_get_property
>  - set_property -> power_supply_set_property
>  - property_is_writeable -> power_supply_property_is_writeable
> 
> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>

Acked-by: Pavel Machek <pavel@ucw.cz>

> @@ -76,7 +76,7 @@ static ssize_t power_supply_show_property(struct device *dev,
>  	if (off == POWER_SUPPLY_PROP_TYPE) {
>  		value.intval = psy->type;
>  	} else {
> -		ret = psy->get_property(psy, off, &value);
> +		ret = power_supply_get_property(psy, off, &value);
>  

One thing.. Your power_supply_get_property (and friends) check for psy
== NULL. Is it neccessary / good idea? As far as I can tell, it should
not really be NULL...
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[PATCH 3/8] power: 88pm860x_charger: Use power_supply_*() API for accessing function attrs

[Pavel Machek]

On Tue 2014-10-14 14:20:41, Krzysztof Kozlowski wrote:
> Replace direct calls to power supply function attributes with wrappers.
> Wrappers provide safe access access in case of unregistering the power
> supply (.e.g by removing the driver). Replace:
>  - get_property -> power_supply_get_property
>  - set_property -> power_supply_set_property
> 
> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>

Acked-by: Pavel Machek <pavel@ucw.cz>

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[PATCH 0/8] power_supply: Add API for safe access of get_property-like function attrs

[Pavel Machek]

Hi!

> Kindly asking for reviewing/testing the drivers, although the changes
> are straightforward.

You can add 

Acked-by: Pavel Machek <pavel@ucw.cz>

to whole series.
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[PATCH 2/8] power_supply: sysfs: Use power_supply_*() API for accessing function attrs

[Krzysztof Kozlowski]

On ?ro, 2014-10-15 at 12:32 +0200, Pavel Machek wrote:
> On Tue 2014-10-14 14:20:40, Krzysztof Kozlowski wrote:
> > Replace direct calls to power supply function attributes with wrappers.
> > Wrappers provide safe access access in case of unregistering the power
> > supply (.e.g by removing the driver). Replace:
> >  - get_property -> power_supply_get_property
> >  - set_property -> power_supply_set_property
> >  - property_is_writeable -> power_supply_property_is_writeable
> > 
> > Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
> 
> Acked-by: Pavel Machek <pavel@ucw.cz>

Thanks for looking at patches.


> > @@ -76,7 +76,7 @@ static ssize_t power_supply_show_property(struct device *dev,
> >  	if (off == POWER_SUPPLY_PROP_TYPE) {
> >  		value.intval = psy->type;
> >  	} else {
> > -		ret = psy->get_property(psy, off, &value);
> > +		ret = power_supply_get_property(psy, off, &value);
> >  
> 
> One thing.. Your power_supply_get_property (and friends) check for psy
> == NULL. Is it neccessary / good idea? As far as I can tell, it should
> not really be NULL...

It is not necessary. I thought it would be a good behavior of such
exported function. You're right that this shouldn't be NULL especially
because previously it was dereferenced.

Other existing power supply exported functions don't check this so maybe
I shouldn't introduce inconsistency.

I'll remove the check and re-spin. I'll found ugly typos in commit
message anyway.

Best regards,
Krzysztof

[PATCH 2/8] power_supply: sysfs: Use power_supply_*() API for accessing function attrs

[Pavel Machek]

> > > @@ -76,7 +76,7 @@ static ssize_t power_supply_show_property(struct device *dev,
> > >  	if (off == POWER_SUPPLY_PROP_TYPE) {
> > >  		value.intval = psy->type;
> > >  	} else {
> > > -		ret = psy->get_property(psy, off, &value);
> > > +		ret = power_supply_get_property(psy, off, &value);
> > >  
> > 
> > One thing.. Your power_supply_get_property (and friends) check for psy
> > == NULL. Is it neccessary / good idea? As far as I can tell, it should
> > not really be NULL...
> 
> It is not necessary. I thought it would be a good behavior of such
> exported function. You're right that this shouldn't be NULL especially
> because previously it was dereferenced.
> 
> Other existing power supply exported functions don't check this so maybe
> I shouldn't introduce inconsistency.
> 
> I'll remove the check and re-spin. I'll found ugly typos in commit
> message anyway.

Thanks!
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[PATCH 4/8] power: ab8500: Use power_supply_*() API for accessing function attrs

[Linus Walleij]

On Tue, Oct 14, 2014 at 2:20 PM, Krzysztof Kozlowski
<k.kozlowski@samsung.com> wrote:

> Replace direct calls to power supply function attributes with wrappers.
> Wrappers provide safe access access in case of unregistering the power
> supply (.e.g by removing the driver). Replace:
>  - get_property -> power_supply_get_property
>
> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>

Acked-by: Linus Walleij <linus.walleij@linaro.org>

Yours,
Linus Walleij