ACPI: battery: Check for error code from devm_mutex_init() call
PATCH 1/1 ACPI: battery: Check for error code from devm_mutex_init() call
Even if it's not critical, the avoidance of checking the error code from devm_mutex_init() call today diminishes the point of using devm variant of it. Tomorrow it may even leak something. Add the missed check.
--- drivers/acpi/battery.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/acpi/battery.c +6 -2
@@ -1226,8 +1226,12 @@ static int acpi_battery_add(struct acpi_device *device)strscpy (acpi_device_name (device),ACPI_BATTERY_DEVICE_NAME );strscpy (acpi_device_class (device),ACPI_BATTERY_CLASS ); device->driver_data = battery; -devm_mutex_init (&device->dev, &battery->lock); -devm_mutex_init (&device->dev, &battery->sysfs_lock); + result =devm_mutex_init (&device->dev, &battery->lock); +if (result) +return result; + result =devm_mutex_init (&device->dev, &battery->sysfs_lock); +if (result) +return result;if (acpi_has_method (battery->device->handle,"_BIX" ))set_bit (ACPI_BATTERY_XINFO_PRESENT , &battery->flags);
Signature · 1 line
2.43.0.rc1.1336.g36b5255a03ac
Hi Andy,
Oct 30, 2024 10:28:03 Andy Shevchenko <[email protected]>:
Quoted 4 lines
Even if it's not critical, the avoidance of checking the error code from
devm_mutex_init()call today diminishes the point of using devm variant of it. Tomorrow it may even leak something. Add the missed check.
Thanks!
Assuming you found this via some sort of tool and you already fixed up all the other places in the tree missing these checks, wouldn't it make sense to mark devm_mutex_init() as __must_check?
Quoted 2 lines
Fixes 0710c1ce5045 ("ACPI: battery: initialize mutexes through devm_ APIs")Signed-off-by Andy Shevchenko <[email protected]>
Quoted 25 lines
---
drivers/acpi/battery.c| 8 ++++++--1 file changed, 6 insertions(+), 2 deletions(-)drivers/acpi/battery.c +6
-3@@ -1226,8 +1226,12 @@ static int acpi_battery_add(struct acpi_device *device) strscpy(acpi_device_name(device), ACPI_BATTERY_DEVICE_NAME); strscpy(acpi_device_class(device), ACPI_BATTERY_CLASS); device->driver_data = battery; - devm_mutex_init(&device->dev, &battery->lock); - devm_mutex_init(&device->dev, &battery->sysfs_lock); + result = devm_mutex_init(&device->dev, &battery->lock); + if (result) + return result; + result = devm_mutex_init(&device->dev, &battery->sysfs_lock); + if (result) + return result; if (acpi_has_method(battery->device->handle, "_BIX")) set_bit(ACPI_BATTERY_XINFO_PRESENT, &battery->flags); --2.43.0.rc1.1336.g36b5255a03ac
Quoted 1 line · Thomas Weißschuh
Oct 30, 2024 10:28:03 Andy Shevchenko <[email protected]>:
...
Quoted 2 lines
Assuming you found this via some sort of tool and you already fixed up all the other places in the tree missing these checks,
The tool is called git grep.
Quoted 1 line
wouldn't it make sense to mark
devm_mutex_init()as__must_check?
It's macro, any idea how to do that for the macros?
...
Quoted 1 line
Reviewed-by Thomas Weißschuh <[email protected]>
Thanks!
Signature · 2 lines
With Best Regards, Andy Shevchenko
Oct 30, 2024 11:31:21 Andy Shevchenko <[email protected]>:
Quoted 8 lines · Thomas Weißschuh
Quoted 1 line
Oct 30, 2024 10:28:03 Andy Shevchenko <[email protected]>:
...
Quoted 1 line
wouldn't it make sense to mark
devm_mutex_init()as__must_check?It's macro, any idea how to do that for the macros?
It should work on __devm_mutex_init().
I don't think the expression macro in between should interfere. Unfortunately I can't test it myself right now.
Quoted 3 lines · Thomas Weißschuh
Oct 30, 2024 11:31:21 Andy Shevchenko <[email protected]>:
Quoted 2 lines · Thomas Weißschuh
Quoted 1 line
Oct 30, 2024 10:28:03 Andy Shevchenko <[email protected]>:
...
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
wouldn't it make sense to mark
devm_mutex_init()as__must_check?It's macro, any idea how to do that for the macros?
It should work on
__devm_mutex_init().
I don't think the expression macro in between should interfere. Unfortunately I can't test it myself right now.
Okay, when you have a patch, feel free to Cc me for review.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 12 lines · Thomas Weißschuh
Hi Andy,
Oct 30, 2024 10:28:03 Andy Shevchenko <[email protected]>:
Quoted 4 lines
Even if it's not critical, the avoidance of checking the error code from
devm_mutex_init()call today diminishes the point of using devm variant of it. Tomorrow it may even leak something. Add the missed check.Thanks!
Quoted 7 lines
Assuming you found this via some sort of tool and you already fixed up all the other places in the tree missing these checks, wouldn't it make sense to mark
devm_mutex_init()as__must_check?Quoted 2 lines
Fixes 0710c1ce5045 ("ACPI: battery: initialize mutexes through devm_ APIs")Signed-off-by Andy Shevchenko <[email protected]>Reviewed-by Thomas Weißschuh <[email protected]>
Applied, thanks!
Quoted 26 lines
Quoted 25 lines
---
drivers/acpi/battery.c| 8 ++++++--1 file changed, 6 insertions(+), 2 deletions(-)drivers/acpi/battery.c +6
-3@@ -1226,8 +1226,12 @@ static int acpi_battery_add(struct acpi_device *device)strscpy (acpi_device_name (device),ACPI_BATTERY_DEVICE_NAME );strscpy (acpi_device_class (device),ACPI_BATTERY_CLASS ); device->driver_data = battery; -devm_mutex_init (&device->dev, &battery->lock); -devm_mutex_init (&device->dev, &battery->sysfs_lock); + result =devm_mutex_init (&device->dev, &battery->lock); +if (result) +return result; + result =devm_mutex_init (&device->dev, &battery->sysfs_lock); +if (result) +return result;if (acpi_has_method (battery->device->handle,"_BIX" ))set_bit (ACPI_BATTERY_XINFO_PRESENT , &battery->flags); --2.43.0.rc1.1336.g36b5255a03ac