| Message ID | 20230130093742.838577-1-marten.lindahl@axis.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | iio: light: vcnl4000: Fix WARN_ON on uninitialized lock | expand |
On Mon, Jan 30, 2023 at 10:37:42AM +0100, Mårten Lindahl wrote: > There are different init functions for the sensors in this driver in > which only one initialize the generic vcnl4000_lock. With commit > e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > the vcnl4040 sensor started to depend on the lock, but it was missed to > initialize it in vcnl4040's init function. This has not been visible > until we run lockdep on it: > DEBUG_LOCKS_WARN_ON(lock->magic != lock) > WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 > ... > Call trace: > __mutex_lock > mutex_lock_nested > vcnl4200_set_power_state > vcnl4200_init > vcnl4000_probe > i2c_device_probe > really_probe > __driver_probe_device > driver_probe_device > __driver_attach > bus_for_each_dev > driver_attach > bus_add_driver > driver_register > i2c_register_driver > vcnl4000_driver_init > do_one_initcall > do_init_module > load_module > __do_sys_finit_module > ... https://www.kernel.org/doc/html/latest/process/submitting-patches.html#backtraces-in-commit-messages Otherwise looks good to me, Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > Fix this by adding mutex_init on the lock in the init function used for > vcnl4040. > > Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> > --- > drivers/iio/light/vcnl4000.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c > index cc1a2062e76d..a8a9fc3b1a02 100644 > --- a/drivers/iio/light/vcnl4000.c > +++ b/drivers/iio/light/vcnl4000.c > @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) > } > mutex_init(&data->vcnl4200_al.lock); > mutex_init(&data->vcnl4200_ps.lock); > + mutex_init(&data->vcnl4000_lock); > > ret = data->chip_spec->set_power_state(data, true); > if (ret < 0) > -- > 2.30.2 >
On Mon, Jan 30, 2023 at 02:39:12PM +0200, Andy Shevchenko wrote: > On Mon, Jan 30, 2023 at 10:37:42AM +0100, Mårten Lindahl wrote: ... > > Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") Actually one more, please provide a proper Fixes tag. https://www.kernel.org/doc/html/latest/process/submitting-patches.html#describe-your-changes It's explained there.
On Mon, 30 Jan 2023 10:37:42 +0100 Mårten Lindahl <marten.lindahl@axis.com> wrote: > There are different init functions for the sensors in this driver in > which only one initialize the generic vcnl4000_lock. With commit > e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > the vcnl4040 sensor started to depend on the lock, but it was missed to > initialize it in vcnl4040's init function. This has not been visible > until we run lockdep on it: > > DEBUG_LOCKS_WARN_ON(lock->magic != lock) > WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 > ... > Call trace: > __mutex_lock > mutex_lock_nested > vcnl4200_set_power_state > vcnl4200_init > vcnl4000_probe > i2c_device_probe > really_probe > __driver_probe_device > driver_probe_device > __driver_attach > bus_for_each_dev > driver_attach > bus_add_driver > driver_register > i2c_register_driver > vcnl4000_driver_init > do_one_initcall > do_init_module > load_module > __do_sys_finit_module > ... > > Fix this by adding mutex_init on the lock in the init function used for > vcnl4040. > > Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> Perhaps better to pull the lock out of the device specific setup? > --- > drivers/iio/light/vcnl4000.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c > index cc1a2062e76d..a8a9fc3b1a02 100644 > --- a/drivers/iio/light/vcnl4000.c > +++ b/drivers/iio/light/vcnl4000.c > @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) > } > mutex_init(&data->vcnl4200_al.lock); > mutex_init(&data->vcnl4200_ps.lock); > + mutex_init(&data->vcnl4000_lock); > > ret = data->chip_spec->set_power_state(data, true); > if (ret < 0)
On 1/30/23 13:40, Andy Shevchenko wrote: > On Mon, Jan 30, 2023 at 02:39:12PM +0200, Andy Shevchenko wrote: >> On Mon, Jan 30, 2023 at 10:37:42AM +0100, Mårten Lindahl wrote: > ... > >>> Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > Actually one more, please provide a proper Fixes tag. > > https://www.kernel.org/doc/html/latest/process/submitting-patches.html#describe-your-changes > > It's explained there. > Hi Andy! Sorry, I'll fix those. Kind regards Mårten
On 1/30/23 14:04, Jonathan Cameron wrote: > On Mon, 30 Jan 2023 10:37:42 +0100 > Mårten Lindahl <marten.lindahl@axis.com> wrote: > >> There are different init functions for the sensors in this driver in >> which only one initialize the generic vcnl4000_lock. With commit >> e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") >> the vcnl4040 sensor started to depend on the lock, but it was missed to >> initialize it in vcnl4040's init function. This has not been visible >> until we run lockdep on it: >> >> DEBUG_LOCKS_WARN_ON(lock->magic != lock) >> WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 >> ... >> Call trace: >> __mutex_lock >> mutex_lock_nested >> vcnl4200_set_power_state >> vcnl4200_init >> vcnl4000_probe >> i2c_device_probe >> really_probe >> __driver_probe_device >> driver_probe_device >> __driver_attach >> bus_for_each_dev >> driver_attach >> bus_add_driver >> driver_register >> i2c_register_driver >> vcnl4000_driver_init >> do_one_initcall >> do_init_module >> load_module >> __do_sys_finit_module >> ... >> >> Fix this by adding mutex_init on the lock in the init function used for >> vcnl4040. >> >> Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") >> Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> > Perhaps better to pull the lock out of the device specific setup? Hi Jonathan! Ok, having a static DEFINE_MUTEX(vcnl4000_lock); will make sure initializing it wont be forgotten for specific setups. I'll test it. Kind regards Mårten > >> --- >> drivers/iio/light/vcnl4000.c | 1 + >> 1 file changed, 1 insertion(+) >> >> diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c >> index cc1a2062e76d..a8a9fc3b1a02 100644 >> --- a/drivers/iio/light/vcnl4000.c >> +++ b/drivers/iio/light/vcnl4000.c >> @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) >> } >> mutex_init(&data->vcnl4200_al.lock); >> mutex_init(&data->vcnl4200_ps.lock); >> + mutex_init(&data->vcnl4000_lock); >> >> ret = data->chip_spec->set_power_state(data, true); >> if (ret < 0)
On Mon, 30 Jan 2023 15:10:40 +0100 Mårten Lindahl <martenli@axis.com> wrote: > On 1/30/23 14:04, Jonathan Cameron wrote: > > On Mon, 30 Jan 2023 10:37:42 +0100 > > Mårten Lindahl <marten.lindahl@axis.com> wrote: > > > >> There are different init functions for the sensors in this driver in > >> which only one initialize the generic vcnl4000_lock. With commit > >> e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > >> the vcnl4040 sensor started to depend on the lock, but it was missed to > >> initialize it in vcnl4040's init function. This has not been visible > >> until we run lockdep on it: > >> > >> DEBUG_LOCKS_WARN_ON(lock->magic != lock) > >> WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 > >> ... > >> Call trace: > >> __mutex_lock > >> mutex_lock_nested > >> vcnl4200_set_power_state > >> vcnl4200_init > >> vcnl4000_probe > >> i2c_device_probe > >> really_probe > >> __driver_probe_device > >> driver_probe_device > >> __driver_attach > >> bus_for_each_dev > >> driver_attach > >> bus_add_driver > >> driver_register > >> i2c_register_driver > >> vcnl4000_driver_init > >> do_one_initcall > >> do_init_module > >> load_module > >> __do_sys_finit_module > >> ... > >> > >> Fix this by adding mutex_init on the lock in the init function used for > >> vcnl4040. > >> > >> Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") > >> Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> > > Perhaps better to pull the lock out of the device specific setup? > > Hi Jonathan! > > Ok, having a > > static DEFINE_MUTEX(vcnl4000_lock); > No. Definitely don't do that. We need one per device instance. I just meant move the mutex_init() into the main probe function rather than the chip specific init that is called from there. > will make sure initializing it wont be forgotten for specific setups. > > I'll test it. > > Kind regards > > Mårten > > > > >> --- > >> drivers/iio/light/vcnl4000.c | 1 + > >> 1 file changed, 1 insertion(+) > >> > >> diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c > >> index cc1a2062e76d..a8a9fc3b1a02 100644 > >> --- a/drivers/iio/light/vcnl4000.c > >> +++ b/drivers/iio/light/vcnl4000.c > >> @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) > >> } > >> mutex_init(&data->vcnl4200_al.lock); > >> mutex_init(&data->vcnl4200_ps.lock); > >> + mutex_init(&data->vcnl4000_lock); > >> > >> ret = data->chip_spec->set_power_state(data, true); > >> if (ret < 0)
On 1/30/23 18:30, Jonathan Cameron wrote: > On Mon, 30 Jan 2023 15:10:40 +0100 > Mårten Lindahl <martenli@axis.com> wrote: > >> On 1/30/23 14:04, Jonathan Cameron wrote: >>> On Mon, 30 Jan 2023 10:37:42 +0100 >>> Mårten Lindahl <marten.lindahl@axis.com> wrote: >>> >>>> There are different init functions for the sensors in this driver in >>>> which only one initialize the generic vcnl4000_lock. With commit >>>> e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") >>>> the vcnl4040 sensor started to depend on the lock, but it was missed to >>>> initialize it in vcnl4040's init function. This has not been visible >>>> until we run lockdep on it: >>>> >>>> DEBUG_LOCKS_WARN_ON(lock->magic != lock) >>>> WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 >>>> ... >>>> Call trace: >>>> __mutex_lock >>>> mutex_lock_nested >>>> vcnl4200_set_power_state >>>> vcnl4200_init >>>> vcnl4000_probe >>>> i2c_device_probe >>>> really_probe >>>> __driver_probe_device >>>> driver_probe_device >>>> __driver_attach >>>> bus_for_each_dev >>>> driver_attach >>>> bus_add_driver >>>> driver_register >>>> i2c_register_driver >>>> vcnl4000_driver_init >>>> do_one_initcall >>>> do_init_module >>>> load_module >>>> __do_sys_finit_module >>>> ... >>>> >>>> Fix this by adding mutex_init on the lock in the init function used for >>>> vcnl4040. >>>> >>>> Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") >>>> Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> >>> Perhaps better to pull the lock out of the device specific setup? >> Hi Jonathan! >> >> Ok, having a >> >> static DEFINE_MUTEX(vcnl4000_lock); >> > No. Definitely don't do that. We need one per device instance. > > I just meant move the mutex_init() into the main probe function rather > than the chip specific init that is called from there. Ok, sorry for the missunderstanding. I''ll fix it. Kind regards Mårten > >> will make sure initializing it wont be forgotten for specific setups. >> >> I'll test it. >> >> Kind regards >> >> Mårten >> >>> >>>> --- >>>> drivers/iio/light/vcnl4000.c | 1 + >>>> 1 file changed, 1 insertion(+) >>>> >>>> diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c >>>> index cc1a2062e76d..a8a9fc3b1a02 100644 >>>> --- a/drivers/iio/light/vcnl4000.c >>>> +++ b/drivers/iio/light/vcnl4000.c >>>> @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) >>>> } >>>> mutex_init(&data->vcnl4200_al.lock); >>>> mutex_init(&data->vcnl4200_ps.lock); >>>> + mutex_init(&data->vcnl4000_lock); >>>> >>>> ret = data->chip_spec->set_power_state(data, true); >>>> if (ret < 0)
diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c index cc1a2062e76d..a8a9fc3b1a02 100644 --- a/drivers/iio/light/vcnl4000.c +++ b/drivers/iio/light/vcnl4000.c @@ -316,6 +316,7 @@ static int vcnl4200_init(struct vcnl4000_data *data) } mutex_init(&data->vcnl4200_al.lock); mutex_init(&data->vcnl4200_ps.lock); + mutex_init(&data->vcnl4000_lock); ret = data->chip_spec->set_power_state(data, true); if (ret < 0)
There are different init functions for the sensors in this driver in which only one initialize the generic vcnl4000_lock. With commit e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") the vcnl4040 sensor started to depend on the lock, but it was missed to initialize it in vcnl4040's init function. This has not been visible until we run lockdep on it: DEBUG_LOCKS_WARN_ON(lock->magic != lock) WARNING: CPU: 1 PID: 8800 at kernel/locking/mutex.c:575 __mutex_lock+0x4f8/0x890 ... Call trace: __mutex_lock mutex_lock_nested vcnl4200_set_power_state vcnl4200_init vcnl4000_probe i2c_device_probe really_probe __driver_probe_device driver_probe_device __driver_attach bus_for_each_dev driver_attach bus_add_driver driver_register i2c_register_driver vcnl4000_driver_init do_one_initcall do_init_module load_module __do_sys_finit_module ... Fix this by adding mutex_init on the lock in the init function used for vcnl4040. Fixes: e21b5b1f2 ("iio: light: vcnl4000: Preserve conf bits when toggle power") Signed-off-by: Mårten Lindahl <marten.lindahl@axis.com> --- drivers/iio/light/vcnl4000.c | 1 + 1 file changed, 1 insertion(+)