| Message ID | 20210324125548.45983-4-aardelean@deviqon.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | platform/x86: toshiba_acpi: move acpi add/remove to device-managed routines | expand |
On Wed, 24 Mar 2021 14:55:41 +0200 Alexandru Ardelean <aardelean@deviqon.com> wrote: > This change moves the registration of the Toshiba ACPI miscdev to be > handled by the devm_add_action_or_reset() hook. This way, the miscdev will > be unregistered when the reference count of the parent device object goes > to zero. > > This also changes the order of cleanup in toshiba_acpi_remove(), where the > miscdev was deregistered first. Now it will be deregistered right before > the toshiba_acpi_dev object is free'd. > > Signed-off-by: Alexandru Ardelean <aardelean@deviqon.com> Reorder looks right to me, but maybe I'm missing something subtle. Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> One unrelated comment inline. Jonathan > --- > drivers/platform/x86/toshiba_acpi.c | 14 ++++++++++++-- > 1 file changed, 12 insertions(+), 2 deletions(-) > > diff --git a/drivers/platform/x86/toshiba_acpi.c b/drivers/platform/x86/toshiba_acpi.c > index c5284601bc2a..53ef565378ef 100644 > --- a/drivers/platform/x86/toshiba_acpi.c > +++ b/drivers/platform/x86/toshiba_acpi.c > @@ -2963,8 +2963,6 @@ static int toshiba_acpi_remove(struct acpi_device *acpi_dev) > { > struct toshiba_acpi_dev *dev = acpi_driver_data(acpi_dev); > > - misc_deregister(&dev->miscdev); > - > remove_toshiba_proc_entries(dev); > > if (dev->accelerometer_supported && dev->indio_dev) { > @@ -3014,6 +3012,13 @@ static void toshiba_acpi_singleton_clear(void *data) > toshiba_acpi = NULL; > } > > +static void toshiba_acpi_misc_deregister(void *data) > +{ > + struct miscdevice *miscdev = data; > + > + misc_deregister(miscdev); > +} > + > static int toshiba_acpi_add(struct acpi_device *acpi_dev) > { > struct device *parent = &acpi_dev->dev; > @@ -3056,6 +3061,11 @@ static int toshiba_acpi_add(struct acpi_device *acpi_dev) > return ret; > } > > + ret = devm_add_action_or_reset(parent, toshiba_acpi_misc_deregister, > + &dev->miscdev); > + if (ret) > + return ret; > + > acpi_dev->driver_data = dev; > dev_set_drvdata(&acpi_dev->dev, dev); Why are we carrying two copies of the same thing? (obviously unrelated to your patch :) >
diff --git a/drivers/platform/x86/toshiba_acpi.c b/drivers/platform/x86/toshiba_acpi.c index c5284601bc2a..53ef565378ef 100644 --- a/drivers/platform/x86/toshiba_acpi.c +++ b/drivers/platform/x86/toshiba_acpi.c @@ -2963,8 +2963,6 @@ static int toshiba_acpi_remove(struct acpi_device *acpi_dev) { struct toshiba_acpi_dev *dev = acpi_driver_data(acpi_dev); - misc_deregister(&dev->miscdev); - remove_toshiba_proc_entries(dev); if (dev->accelerometer_supported && dev->indio_dev) { @@ -3014,6 +3012,13 @@ static void toshiba_acpi_singleton_clear(void *data) toshiba_acpi = NULL; } +static void toshiba_acpi_misc_deregister(void *data) +{ + struct miscdevice *miscdev = data; + + misc_deregister(miscdev); +} + static int toshiba_acpi_add(struct acpi_device *acpi_dev) { struct device *parent = &acpi_dev->dev; @@ -3056,6 +3061,11 @@ static int toshiba_acpi_add(struct acpi_device *acpi_dev) return ret; } + ret = devm_add_action_or_reset(parent, toshiba_acpi_misc_deregister, + &dev->miscdev); + if (ret) + return ret; + acpi_dev->driver_data = dev; dev_set_drvdata(&acpi_dev->dev, dev);
This change moves the registration of the Toshiba ACPI miscdev to be handled by the devm_add_action_or_reset() hook. This way, the miscdev will be unregistered when the reference count of the parent device object goes to zero. This also changes the order of cleanup in toshiba_acpi_remove(), where the miscdev was deregistered first. Now it will be deregistered right before the toshiba_acpi_dev object is free'd. Signed-off-by: Alexandru Ardelean <aardelean@deviqon.com> --- drivers/platform/x86/toshiba_acpi.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-)