Message ID | 20200826120042.200364-1-alexandru.ardelean@analog.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | [v2] iio: stm32-dac: Replace indio_dev->mlock with own device lock | expand |
On Wed, Aug 26, 2020 at 3:03 PM Alexandru Ardelean <alexandru.ardelean@analog.com> wrote: > > From: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > > As part of the general cleanup of indio_dev->mlock, this change replaces > it with a local lock. The lock protects against potential races when > reading the CR reg and then updating, so that the state of pm_runtime > is consistent between the two operations. > > Signed-off-by: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> > --- Forgot the changelog here. Apologies. Changelog v1 -> v2: * removed whitespace change for 'common' field * updated comment about the lock usage > drivers/iio/dac/stm32-dac.c | 12 ++++++++---- > 1 file changed, 8 insertions(+), 4 deletions(-) > > diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c > index 092c796fa3d9..7a8aed476850 100644 > --- a/drivers/iio/dac/stm32-dac.c > +++ b/drivers/iio/dac/stm32-dac.c > @@ -26,9 +26,11 @@ > /** > * struct stm32_dac - private data of DAC driver > * @common: reference to DAC common data > + * @lock: lock to protect the data buffer during regmap ops > */ > struct stm32_dac { > struct stm32_dac_common *common; > + struct mutex lock; > }; > > static int stm32_dac_is_enabled(struct iio_dev *indio_dev, int channel) > @@ -58,10 +60,10 @@ static int stm32_dac_set_enable_state(struct iio_dev *indio_dev, int ch, > int ret; > > /* already enabled / disabled ? */ > - mutex_lock(&indio_dev->mlock); > + mutex_lock(&dac->lock); > ret = stm32_dac_is_enabled(indio_dev, ch); > if (ret < 0 || enable == !!ret) { > - mutex_unlock(&indio_dev->mlock); > + mutex_unlock(&dac->lock); > return ret < 0 ? ret : 0; > } > > @@ -69,13 +71,13 @@ static int stm32_dac_set_enable_state(struct iio_dev *indio_dev, int ch, > ret = pm_runtime_get_sync(dev); > if (ret < 0) { > pm_runtime_put_noidle(dev); > - mutex_unlock(&indio_dev->mlock); > + mutex_unlock(&dac->lock); > return ret; > } > } > > ret = regmap_update_bits(dac->common->regmap, STM32_DAC_CR, msk, en); > - mutex_unlock(&indio_dev->mlock); > + mutex_unlock(&dac->lock); > if (ret < 0) { > dev_err(&indio_dev->dev, "%s failed\n", en ? > "Enable" : "Disable"); > @@ -327,6 +329,8 @@ static int stm32_dac_probe(struct platform_device *pdev) > indio_dev->info = &stm32_dac_iio_info; > indio_dev->modes = INDIO_DIRECT_MODE; > > + mutex_init(&dac->lock); > + > ret = stm32_dac_chan_of_init(indio_dev); > if (ret < 0) > return ret; > -- > 2.25.1 >
On 8/27/20 10:55 AM, Alexandru Ardelean wrote: > On Wed, Aug 26, 2020 at 3:03 PM Alexandru Ardelean > <alexandru.ardelean@analog.com> wrote: >> From: Sergiu Cuciurean <sergiu.cuciurean@analog.com> >> >> As part of the general cleanup of indio_dev->mlock, this change replaces >> it with a local lock. The lock protects against potential races when >> reading the CR reg and then updating, so that the state of pm_runtime >> is consistent between the two operations. >> >> Signed-off-by: Sergiu Cuciurean <sergiu.cuciurean@analog.com> >> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> >> --- > Forgot the changelog here. > Apologies. > > Changelog v1 -> v2: > * removed whitespace change for 'common' field > * updated comment about the lock usage Hi Alexandru, Sorry if I missed it... is there an update on the comment :-) ? Best Regards, Fabrice > >> drivers/iio/dac/stm32-dac.c | 12 ++++++++---- >> 1 file changed, 8 insertions(+), 4 deletions(-) >> >> diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c >> index 092c796fa3d9..7a8aed476850 100644 >> --- a/drivers/iio/dac/stm32-dac.c >> +++ b/drivers/iio/dac/stm32-dac.c >> @@ -26,9 +26,11 @@ >> /** >> * struct stm32_dac - private data of DAC driver >> * @common: reference to DAC common data >> + * @lock: lock to protect the data buffer during regmap ops >> */ >> struct stm32_dac { >> struct stm32_dac_common *common; >> + struct mutex lock; >> };
On Thu, Aug 27, 2020 at 12:03 PM Fabrice Gasnier <fabrice.gasnier@st.com> wrote: > > On 8/27/20 10:55 AM, Alexandru Ardelean wrote: > > On Wed, Aug 26, 2020 at 3:03 PM Alexandru Ardelean > > <alexandru.ardelean@analog.com> wrote: > >> From: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > >> > >> As part of the general cleanup of indio_dev->mlock, this change replaces > >> it with a local lock. The lock protects against potential races when > >> reading the CR reg and then updating, so that the state of pm_runtime > >> is consistent between the two operations. > >> > >> Signed-off-by: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > >> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> > >> --- > > Forgot the changelog here. > > Apologies. > > > > Changelog v1 -> v2: > > * removed whitespace change for 'common' field > > * updated comment about the lock usage > > Hi Alexandru, > > Sorry if I missed it... is there an update on the comment :-) ? For a moment there, I thought I didn't. GMail's threading is confusing. ---------------------------------------------------------------------------- As part of the general cleanup of indio_dev->mlock, this change replaces it with a local lock. The lock protects against potential races when reading the CR reg and then updating, so that the state of pm_runtime is consistent between the two operations. ---------------------------------------------------------------------------- > > Best Regards, > Fabrice > > > >> drivers/iio/dac/stm32-dac.c | 12 ++++++++---- > >> 1 file changed, 8 insertions(+), 4 deletions(-) > >> > >> diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c > >> index 092c796fa3d9..7a8aed476850 100644 > >> --- a/drivers/iio/dac/stm32-dac.c > >> +++ b/drivers/iio/dac/stm32-dac.c > >> @@ -26,9 +26,11 @@ > >> /** > >> * struct stm32_dac - private data of DAC driver > >> * @common: reference to DAC common data > >> + * @lock: lock to protect the data buffer during regmap ops > >> */ > >> struct stm32_dac { > >> struct stm32_dac_common *common; > >> + struct mutex lock; > >> };
On Thu, 27 Aug 2020 13:00:36 +0300 Alexandru Ardelean <ardeleanalex@gmail.com> wrote: > On Thu, Aug 27, 2020 at 12:03 PM Fabrice Gasnier <fabrice.gasnier@st.com> wrote: > > > > On 8/27/20 10:55 AM, Alexandru Ardelean wrote: > > > On Wed, Aug 26, 2020 at 3:03 PM Alexandru Ardelean > > > <alexandru.ardelean@analog.com> wrote: > > >> From: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > > >> > > >> As part of the general cleanup of indio_dev->mlock, this change replaces > > >> it with a local lock. The lock protects against potential races when > > >> reading the CR reg and then updating, so that the state of pm_runtime > > >> is consistent between the two operations. > > >> > > >> Signed-off-by: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > > >> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> > > >> --- > > > Forgot the changelog here. > > > Apologies. > > > > > > Changelog v1 -> v2: > > > * removed whitespace change for 'common' field > > > * updated comment about the lock usage > > > > Hi Alexandru, > > > > Sorry if I missed it... is there an update on the comment :-) ? > > For a moment there, I thought I didn't. > GMail's threading is confusing. > > ---------------------------------------------------------------------------- > As part of the general cleanup of indio_dev->mlock, this change replaces > it with a local lock. The lock protects against potential races when > reading the CR reg and then updating, so that the state of pm_runtime > is consistent between the two operations. > ---------------------------------------------------------------------------- I think this got confused... see below. > > > > > Best Regards, > > Fabrice > > > > > >> drivers/iio/dac/stm32-dac.c | 12 ++++++++---- > > >> 1 file changed, 8 insertions(+), 4 deletions(-) > > >> > > >> diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c > > >> index 092c796fa3d9..7a8aed476850 100644 > > >> --- a/drivers/iio/dac/stm32-dac.c > > >> +++ b/drivers/iio/dac/stm32-dac.c > > >> @@ -26,9 +26,11 @@ > > >> /** > > >> * struct stm32_dac - private data of DAC driver > > >> * @common: reference to DAC common data > > >> + * @lock: lock to protect the data buffer during regmap ops The original comment was: In this particular case I'm not sure that's what mlock was being used for. I think it's about avoiding races around checking if powered down and actually doing it. And Fabrice's reply: Hi Sergiu, Indeed, purpose is to protect against a race here when reading CR, and updating it via regmap (this also makes the subsequent pm_runtime calls to be balanced based on this). (Side note: there is no data buffer involved for the DAC.) Could you please update the comment ? Thanks, Fabrice > > >> */ > > >> struct stm32_dac { > > >> struct stm32_dac_common *common; > > >> + struct mutex lock; > > >> };
On Sat, Aug 29, 2020 at 6:46 PM Jonathan Cameron <jic23@kernel.org> wrote: > > On Thu, 27 Aug 2020 13:00:36 +0300 > Alexandru Ardelean <ardeleanalex@gmail.com> wrote: > > > On Thu, Aug 27, 2020 at 12:03 PM Fabrice Gasnier <fabrice.gasnier@st.com> wrote: > > > > > > On 8/27/20 10:55 AM, Alexandru Ardelean wrote: > > > > On Wed, Aug 26, 2020 at 3:03 PM Alexandru Ardelean > > > > <alexandru.ardelean@analog.com> wrote: > > > >> From: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > > > >> > > > >> As part of the general cleanup of indio_dev->mlock, this change replaces > > > >> it with a local lock. The lock protects against potential races when > > > >> reading the CR reg and then updating, so that the state of pm_runtime > > > >> is consistent between the two operations. > > > >> > > > >> Signed-off-by: Sergiu Cuciurean <sergiu.cuciurean@analog.com> > > > >> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> > > > >> --- > > > > Forgot the changelog here. > > > > Apologies. > > > > > > > > Changelog v1 -> v2: > > > > * removed whitespace change for 'common' field > > > > * updated comment about the lock usage > > > > > > Hi Alexandru, > > > > > > Sorry if I missed it... is there an update on the comment :-) ? > > > > For a moment there, I thought I didn't. > > GMail's threading is confusing. > > > > ---------------------------------------------------------------------------- > > As part of the general cleanup of indio_dev->mlock, this change replaces > > it with a local lock. The lock protects against potential races when > > reading the CR reg and then updating, so that the state of pm_runtime > > is consistent between the two operations. > > ---------------------------------------------------------------------------- > I think this got confused... > > see below. > > > > > > > > > > Best Regards, > > > Fabrice > > > > > > > >> drivers/iio/dac/stm32-dac.c | 12 ++++++++---- > > > >> 1 file changed, 8 insertions(+), 4 deletions(-) > > > >> > > > >> diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c > > > >> index 092c796fa3d9..7a8aed476850 100644 > > > >> --- a/drivers/iio/dac/stm32-dac.c > > > >> +++ b/drivers/iio/dac/stm32-dac.c > > > >> @@ -26,9 +26,11 @@ > > > >> /** > > > >> * struct stm32_dac - private data of DAC driver > > > >> * @common: reference to DAC common data > > > >> + * @lock: lock to protect the data buffer during regmap ops oh, silly me; it's about this comment; will re-spin > > The original comment was: > > > In this particular case I'm not sure that's what mlock was being used for. > I think it's about avoiding races around checking if powered down and > actually doing it. > > And Fabrice's reply: > > Hi Sergiu, > > Indeed, purpose is to protect against a race here when reading CR, and > updating it via regmap (this also makes the subsequent pm_runtime calls > to be balanced based on this). > (Side note: there is no data buffer involved for the DAC.) > Could you please update the comment ? > > Thanks, > Fabrice > > > > >> */ > > > >> struct stm32_dac { > > > >> struct stm32_dac_common *common; > > > >> + struct mutex lock; > > > >> }; >
diff --git a/drivers/iio/dac/stm32-dac.c b/drivers/iio/dac/stm32-dac.c index 092c796fa3d9..7a8aed476850 100644 --- a/drivers/iio/dac/stm32-dac.c +++ b/drivers/iio/dac/stm32-dac.c @@ -26,9 +26,11 @@ /** * struct stm32_dac - private data of DAC driver * @common: reference to DAC common data + * @lock: lock to protect the data buffer during regmap ops */ struct stm32_dac { struct stm32_dac_common *common; + struct mutex lock; }; static int stm32_dac_is_enabled(struct iio_dev *indio_dev, int channel) @@ -58,10 +60,10 @@ static int stm32_dac_set_enable_state(struct iio_dev *indio_dev, int ch, int ret; /* already enabled / disabled ? */ - mutex_lock(&indio_dev->mlock); + mutex_lock(&dac->lock); ret = stm32_dac_is_enabled(indio_dev, ch); if (ret < 0 || enable == !!ret) { - mutex_unlock(&indio_dev->mlock); + mutex_unlock(&dac->lock); return ret < 0 ? ret : 0; } @@ -69,13 +71,13 @@ static int stm32_dac_set_enable_state(struct iio_dev *indio_dev, int ch, ret = pm_runtime_get_sync(dev); if (ret < 0) { pm_runtime_put_noidle(dev); - mutex_unlock(&indio_dev->mlock); + mutex_unlock(&dac->lock); return ret; } } ret = regmap_update_bits(dac->common->regmap, STM32_DAC_CR, msk, en); - mutex_unlock(&indio_dev->mlock); + mutex_unlock(&dac->lock); if (ret < 0) { dev_err(&indio_dev->dev, "%s failed\n", en ? "Enable" : "Disable"); @@ -327,6 +329,8 @@ static int stm32_dac_probe(struct platform_device *pdev) indio_dev->info = &stm32_dac_iio_info; indio_dev->modes = INDIO_DIRECT_MODE; + mutex_init(&dac->lock); + ret = stm32_dac_chan_of_init(indio_dev); if (ret < 0) return ret;