Hi, On 2/6/23 15:36, Jonathan Cameron wrote: > On Mon, 6 Feb 2023 14:36:16 +0800 > Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx> wrote: > >> If a second client that talks to the actual I2C address was created in >> probe(), there should be a corresponding cleanup in remove() to avoid >> leakage. >> >> So if the "client" is not the same one used by I2C core, unregister it >> accordingly. >> >> Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2152281 >> Fixes: c1e62062ff54 ("iio: light: cm32181: Handle CM3218 ACPI devices with 2 I2C resources") >> Signed-off-by: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx> > > Whilst this fixes the leak, it introduces a race between unregistering > some of the hardware and the later removal of userspace interfaces. > This needs to be handled via a devm_add_action_or_reset() > registered handler so it's automatically cleaned up at the right place > in the driver remove flow. Good point, I missed that issue. Regards, Hans >> --- >> drivers/iio/light/cm32181.c | 10 ++++++++++ >> 1 file changed, 10 insertions(+) >> >> diff --git a/drivers/iio/light/cm32181.c b/drivers/iio/light/cm32181.c >> index b1674a5bfa368..37439e103d273 100644 >> --- a/drivers/iio/light/cm32181.c >> +++ b/drivers/iio/light/cm32181.c >> @@ -488,6 +488,15 @@ static int cm32181_probe(struct i2c_client *client) >> return 0; >> } >> >> +static void cm32181_remove(struct i2c_client *client) >> +{ >> + struct cm32181_chip *cm32181 = iio_priv(i2c_get_clientdata(client)); >> + >> + /* Unregister the dummy client */ >> + if (cm32181->client != client) >> + i2c_unregister_device(cm32181->client); >> +} >> + >> static int cm32181_suspend(struct device *dev) >> { >> struct cm32181_chip *cm32181 = iio_priv(dev_get_drvdata(dev)); >> @@ -531,6 +540,7 @@ static struct i2c_driver cm32181_driver = { >> .pm = pm_sleep_ptr(&cm32181_pm_ops), >> }, >> .probe_new = cm32181_probe, >> + .remove = cm32181_remove, >> }; >> >> module_i2c_driver(cm32181_driver); >
