Re: [PATCH] iio: adc: stm32-adc: Add check on overrun interrupt

From: Jonathan Cameron
Date: Sun Dec 01 2019 - 06:31:40 EST

Next message: Jonathan Cameron: "Re: [PATCH v4] dt-bindings: iio: accel: add binding documentation for ADIS16240"
Previous message: syzbot: "Re: BUG: sleeping function called from invalid context in __alloc_pages_nodemask"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Mon, 25 Nov 2019 09:42:01 +0100
Fabrice Gasnier <fabrice.gasnier@xxxxxx> wrote:

> On 11/23/19 12:36 PM, Jonathan Cameron wrote:
> > On Fri, 22 Nov 2019 16:08:14 +0100
> > Fabrice Gasnier <fabrice.gasnier@xxxxxx> wrote:
> >
> >> Enable overrun interrupt on STM32 ADC. In case data register hasn't been
> >> read (by CPU or DMA), overrun condition is detected when there's new
> >> conversion data available. Stop grabbing data and log an error message.
> >> Use a threaded irq to avoid printing the error message from hard irq
> >> context.
> >>
> >> Signed-off-by: Fabrice Gasnier <fabrice.gasnier@xxxxxx>
> >
> > Absolutely makes sense to notify the user this occurred and to stop
> > the current process. It may be useful to give a hint on how
> > to restart capture?
>
> Hi Jonathan,
>
> Restarting the capture can simply be done by disabling/re-enabling, e.g.
> from userland:
> cd /sys/bus/iio/devices/iio:deviceX
> echo 0 > buffer/enable
> echo 1 > buffer/enable
>
> Do you think such a hint should be added somewhere? (in the commit
> message or in a comment...) ?

Perhaps it should put something in the kernel log if
the overflow occurs?

Otherwise comment in code is always better than the commit log as chances
of a random user finding the commit very low indeed ;)

Jonathan

>
> >
> > Is this a condition that can occur under 'normal conditions' such
> > as something else grabbing the CPU for too long?
>
> Indeed, this may happen for instance when DMA isn't used (e.g. no DMA
> channel as been assigned to the ADC, so using CPU/interrupts), and CPU
> is busy for too long... I don't expect this may happen otherwise.
>
> >
> > I'd just like to understand if we are dealing with a 'fault'
> > case of just one where things go wrong under weird conditions.
>
> Of course something weird, wrong, faulty... happening on the DMA side
> for example, may cause overrun. But it's quite unlikely.
>
> >
> > Patch itself looks fine.
>
> Please let me know if I need to improve description.
>
> Thanks for reviewing,
> Fabrice
>
> >
> > Thanks,
> >
> > Jonathan
> >
> >
> >> ---
> >> drivers/iio/adc/stm32-adc-core.c | 14 +++++------
> >> drivers/iio/adc/stm32-adc-core.h | 9 +++++++
> >> drivers/iio/adc/stm32-adc.c | 53 ++++++++++++++++++++++++++++++++++++++--
> >> 3 files changed, 67 insertions(+), 9 deletions(-)
> >>
> >> diff --git a/drivers/iio/adc/stm32-adc-core.c b/drivers/iio/adc/stm32-adc-core.c
> >> index 6537f4f..97655d7 100644
> >> --- a/drivers/iio/adc/stm32-adc-core.c
> >> +++ b/drivers/iio/adc/stm32-adc-core.c
> >> @@ -280,21 +280,21 @@ static int stm32h7_adc_clk_sel(struct platform_device *pdev,
> >> static const struct stm32_adc_common_regs stm32f4_adc_common_regs = {
> >> .csr = STM32F4_ADC_CSR,
> >> .ccr = STM32F4_ADC_CCR,
> >> - .eoc1_msk = STM32F4_EOC1,
> >> - .eoc2_msk = STM32F4_EOC2,
> >> - .eoc3_msk = STM32F4_EOC3,
> >> + .eoc1_msk = STM32F4_EOC1 | STM32F4_OVR1,
> >> + .eoc2_msk = STM32F4_EOC2 | STM32F4_OVR2,
> >> + .eoc3_msk = STM32F4_EOC3 | STM32F4_OVR3,
> >> .ier = STM32F4_ADC_CR1,
> >> - .eocie_msk = STM32F4_EOCIE,
> >> + .eocie_msk = STM32F4_EOCIE | STM32F4_OVRIE,
> >> };
> >>
> >> /* STM32H7 common registers definitions */
> >> static const struct stm32_adc_common_regs stm32h7_adc_common_regs = {
> >> .csr = STM32H7_ADC_CSR,
> >> .ccr = STM32H7_ADC_CCR,
> >> - .eoc1_msk = STM32H7_EOC_MST,
> >> - .eoc2_msk = STM32H7_EOC_SLV,
> >> + .eoc1_msk = STM32H7_EOC_MST | STM32H7_OVR_MST,
> >> + .eoc2_msk = STM32H7_EOC_SLV | STM32H7_OVR_SLV,
> >> .ier = STM32H7_ADC_IER,
> >> - .eocie_msk = STM32H7_EOCIE,
> >> + .eocie_msk = STM32H7_EOCIE | STM32H7_OVRIE,
> >> };
> >>
> >> static const unsigned int stm32_adc_offset[STM32_ADC_MAX_ADCS] = {
> >> diff --git a/drivers/iio/adc/stm32-adc-core.h b/drivers/iio/adc/stm32-adc-core.h
> >> index 2579d51..2322809 100644
> >> --- a/drivers/iio/adc/stm32-adc-core.h
> >> +++ b/drivers/iio/adc/stm32-adc-core.h
> >> @@ -51,10 +51,12 @@
> >> #define STM32F4_ADC_CCR (STM32_ADCX_COMN_OFFSET + 0x04)
> >>
> >> /* STM32F4_ADC_SR - bit fields */
> >> +#define STM32F4_OVR BIT(5)
> >> #define STM32F4_STRT BIT(4)
> >> #define STM32F4_EOC BIT(1)
> >>
> >> /* STM32F4_ADC_CR1 - bit fields */
> >> +#define STM32F4_OVRIE BIT(26)
> >> #define STM32F4_RES_SHIFT 24
> >> #define STM32F4_RES_MASK GENMASK(25, 24)
> >> #define STM32F4_SCAN BIT(8)
> >> @@ -72,8 +74,11 @@
> >> #define STM32F4_ADON BIT(0)
> >>
> >> /* STM32F4_ADC_CSR - bit fields */
> >> +#define STM32F4_OVR3 BIT(21)
> >> #define STM32F4_EOC3 BIT(17)
> >> +#define STM32F4_OVR2 BIT(13)
> >> #define STM32F4_EOC2 BIT(9)
> >> +#define STM32F4_OVR1 BIT(5)
> >> #define STM32F4_EOC1 BIT(1)
> >>
> >> /* STM32F4_ADC_CCR - bit fields */
> >> @@ -103,10 +108,12 @@
> >>
> >> /* STM32H7_ADC_ISR - bit fields */
> >> #define STM32MP1_VREGREADY BIT(12)
> >> +#define STM32H7_OVR BIT(4)
> >> #define STM32H7_EOC BIT(2)
> >> #define STM32H7_ADRDY BIT(0)
> >>
> >> /* STM32H7_ADC_IER - bit fields */
> >> +#define STM32H7_OVRIE STM32H7_OVR
> >> #define STM32H7_EOCIE STM32H7_EOC
> >>
> >> /* STM32H7_ADC_CR - bit fields */
> >> @@ -155,7 +162,9 @@ enum stm32h7_adc_dmngt {
> >> #define STM32H7_LINCALFACT_MASK GENMASK(29, 0)
> >>
> >> /* STM32H7_ADC_CSR - bit fields */
> >> +#define STM32H7_OVR_SLV BIT(20)
> >> #define STM32H7_EOC_SLV BIT(18)
> >> +#define STM32H7_OVR_MST BIT(4)
> >> #define STM32H7_EOC_MST BIT(2)
> >>
> >> /* STM32H7_ADC_CCR - bit fields */
> >> diff --git a/drivers/iio/adc/stm32-adc.c b/drivers/iio/adc/stm32-adc.c
> >> index 3b291d7..9361f92 100644
> >> --- a/drivers/iio/adc/stm32-adc.c
> >> +++ b/drivers/iio/adc/stm32-adc.c
> >> @@ -117,7 +117,9 @@ struct stm32_adc_regs {
> >> * struct stm32_adc_regspec - stm32 registers definition
> >> * @dr: data register offset
> >> * @ier_eoc: interrupt enable register & eocie bitfield
> >> + * @ier_ovr: interrupt enable register & overrun bitfield
> >> * @isr_eoc: interrupt status register & eoc bitfield
> >> + * @isr_ovr: interrupt status register & overrun bitfield
> >> * @sqr: reference to sequence registers array
> >> * @exten: trigger control register & bitfield
> >> * @extsel: trigger selection register & bitfield
> >> @@ -128,7 +130,9 @@ struct stm32_adc_regs {
> >> struct stm32_adc_regspec {
> >> const u32 dr;
> >> const struct stm32_adc_regs ier_eoc;
> >> + const struct stm32_adc_regs ier_ovr;
> >> const struct stm32_adc_regs isr_eoc;
> >> + const struct stm32_adc_regs isr_ovr;
> >> const struct stm32_adc_regs *sqr;
> >> const struct stm32_adc_regs exten;
> >> const struct stm32_adc_regs extsel;
> >> @@ -337,7 +341,9 @@ static const unsigned int stm32f4_adc_smp_cycles[STM32_ADC_MAX_SMP + 1] = {
> >> static const struct stm32_adc_regspec stm32f4_adc_regspec = {
> >> .dr = STM32F4_ADC_DR,
> >> .ier_eoc = { STM32F4_ADC_CR1, STM32F4_EOCIE },
> >> + .ier_ovr = { STM32F4_ADC_CR1, STM32F4_OVRIE },
> >> .isr_eoc = { STM32F4_ADC_SR, STM32F4_EOC },
> >> + .isr_ovr = { STM32F4_ADC_SR, STM32F4_OVR },
> >> .sqr = stm32f4_sq,
> >> .exten = { STM32F4_ADC_CR2, STM32F4_EXTEN_MASK, STM32F4_EXTEN_SHIFT },
> >> .extsel = { STM32F4_ADC_CR2, STM32F4_EXTSEL_MASK,
> >> @@ -429,7 +435,9 @@ static const unsigned int stm32h7_adc_smp_cycles[STM32_ADC_MAX_SMP + 1] = {
> >> static const struct stm32_adc_regspec stm32h7_adc_regspec = {
> >> .dr = STM32H7_ADC_DR,
> >> .ier_eoc = { STM32H7_ADC_IER, STM32H7_EOCIE },
> >> + .ier_ovr = { STM32H7_ADC_IER, STM32H7_OVRIE },
> >> .isr_eoc = { STM32H7_ADC_ISR, STM32H7_EOC },
> >> + .isr_ovr = { STM32H7_ADC_ISR, STM32H7_OVR },
> >> .sqr = stm32h7_sq,
> >> .exten = { STM32H7_ADC_CFGR, STM32H7_EXTEN_MASK, STM32H7_EXTEN_SHIFT },
> >> .extsel = { STM32H7_ADC_CFGR, STM32H7_EXTSEL_MASK,
> >> @@ -506,6 +514,18 @@ static void stm32_adc_conv_irq_disable(struct stm32_adc *adc)
> >> adc->cfg->regs->ier_eoc.mask);
> >> }
> >>
> >> +static void stm32_adc_ovr_irq_enable(struct stm32_adc *adc)
> >> +{
> >> + stm32_adc_set_bits(adc, adc->cfg->regs->ier_ovr.reg,
> >> + adc->cfg->regs->ier_ovr.mask);
> >> +}
> >> +
> >> +static void stm32_adc_ovr_irq_disable(struct stm32_adc *adc)
> >> +{
> >> + stm32_adc_clr_bits(adc, adc->cfg->regs->ier_ovr.reg,
> >> + adc->cfg->regs->ier_ovr.mask);
> >> +}
> >> +
> >> static void stm32_adc_set_res(struct stm32_adc *adc)
> >> {
> >> const struct stm32_adc_regs *res = &adc->cfg->regs->res;
> >> @@ -1205,6 +1225,19 @@ static int stm32_adc_read_raw(struct iio_dev *indio_dev,
> >> }
> >> }
> >>
> >> +static irqreturn_t stm32_adc_threaded_isr(int irq, void *data)
> >> +{
> >> + struct stm32_adc *adc = data;
> >> + struct iio_dev *indio_dev = iio_priv_to_dev(adc);
> >> + const struct stm32_adc_regspec *regs = adc->cfg->regs;
> >> + u32 status = stm32_adc_readl(adc, regs->isr_eoc.reg);
> >> +
> >> + if (status & regs->isr_ovr.mask)
> >> + dev_err(&indio_dev->dev, "Overrun interrupt, stopping.\n");
> >> +
> >> + return IRQ_HANDLED;
> >> +}
> >> +
> >> static irqreturn_t stm32_adc_isr(int irq, void *data)
> >> {
> >> struct stm32_adc *adc = data;
> >> @@ -1212,6 +1245,17 @@ static irqreturn_t stm32_adc_isr(int irq, void *data)
> >> const struct stm32_adc_regspec *regs = adc->cfg->regs;
> >> u32 status = stm32_adc_readl(adc, regs->isr_eoc.reg);
> >>
> >> + if (status & regs->isr_ovr.mask) {
> >> + /*
> >> + * Overrun occurred on regular conversions: data for wrong
> >> + * channel may be read. Unconditionally disable interrupts
> >> + * to stop processing data and print error message.
> >> + */
> >> + stm32_adc_ovr_irq_disable(adc);
> >> + stm32_adc_conv_irq_disable(adc);
> >> + return IRQ_WAKE_THREAD;
> >> + }
> >> +
> >> if (status & regs->isr_eoc.mask) {
> >> /* Reading DR also clears EOC status flag */
> >> adc->buffer[adc->bufi] = stm32_adc_readw(adc, regs->dr);
> >> @@ -1441,6 +1485,8 @@ static int __stm32_adc_buffer_postenable(struct iio_dev *indio_dev)
> >> /* Reset adc buffer index */
> >> adc->bufi = 0;
> >>
> >> + stm32_adc_ovr_irq_enable(adc);
> >> +
> >> if (!adc->dma_chan)
> >> stm32_adc_conv_irq_enable(adc);
> >>
> >> @@ -1481,6 +1527,8 @@ static void __stm32_adc_buffer_predisable(struct iio_dev *indio_dev)
> >> if (!adc->dma_chan)
> >> stm32_adc_conv_irq_disable(adc);
> >>
> >> + stm32_adc_ovr_irq_disable(adc);
> >> +
> >> if (adc->dma_chan)
> >> dmaengine_terminate_sync(adc->dma_chan);
> >>
> >> @@ -1818,8 +1866,9 @@ static int stm32_adc_probe(struct platform_device *pdev)
> >> if (adc->irq < 0)
> >> return adc->irq;
> >>
> >> - ret = devm_request_irq(&pdev->dev, adc->irq, stm32_adc_isr,
> >> - 0, pdev->name, adc);
> >> + ret = devm_request_threaded_irq(&pdev->dev, adc->irq, stm32_adc_isr,
> >> + stm32_adc_threaded_isr,
> >> + 0, pdev->name, adc);
> >> if (ret) {
> >> dev_err(&pdev->dev, "failed to request IRQ\n");
> >> return ret;
> >

Next message: Jonathan Cameron: "Re: [PATCH v4] dt-bindings: iio: accel: add binding documentation for ADIS16240"
Previous message: syzbot: "Re: BUG: sleeping function called from invalid context in __alloc_pages_nodemask"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]