* [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead dma_request_slave_channel()
@ 2019-12-17 7:51 Peter Ujfalusi
2019-12-23 15:42 ` Jonathan Cameron
0 siblings, 1 reply; 3+ messages in thread
From: Peter Ujfalusi @ 2019-12-17 7:51 UTC (permalink / raw)
To: jic23, mcoquelin.stm32, alexandre.torgue
Cc: linux-iio, vkoul, linux-kernel, linux-arm-kernel, linux-stm32
dma_request_slave_channel() is a wrapper on top of dma_request_chan()
eating up the error code.
By using dma_request_chan() directly the driver can support deferred
probing against DMA.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
drivers/iio/adc/stm32-adc.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/iio/adc/stm32-adc.c b/drivers/iio/adc/stm32-adc.c
index 3b291d72701c..5dab23f1fdee 100644
--- a/drivers/iio/adc/stm32-adc.c
+++ b/drivers/iio/adc/stm32-adc.c
@@ -1746,9 +1746,15 @@ static int stm32_adc_dma_request(struct iio_dev *indio_dev)
struct dma_slave_config config;
int ret;
- adc->dma_chan = dma_request_slave_channel(&indio_dev->dev, "rx");
- if (!adc->dma_chan)
+ adc->dma_chan = dma_request_chan(&indio_dev->dev, "rx");
+ if (IS_ERR(adc->dma_chan)) {
+ if (PTR_ERR(adc->dma_chan) == -EPROBE_DEFER)
+ return -EPROBE_DEFER;
+
+ /* Ignore errors to fall back to IRQ mode */
+ adc->dma_chan = NULL;
return 0;
+ }
adc->rx_buf = dma_alloc_coherent(adc->dma_chan->device->dev,
STM32_DMA_BUFFER_SIZE,
--
Peter
Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki.
Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead dma_request_slave_channel()
2019-12-17 7:51 [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead dma_request_slave_channel() Peter Ujfalusi
@ 2019-12-23 15:42 ` Jonathan Cameron
2020-01-06 10:25 ` Fabrice Gasnier
0 siblings, 1 reply; 3+ messages in thread
From: Jonathan Cameron @ 2019-12-23 15:42 UTC (permalink / raw)
To: Peter Ujfalusi
Cc: alexandre.torgue, linux-iio, linux-kernel, vkoul,
mcoquelin.stm32, Fabrice Gasnier, linux-stm32, linux-arm-kernel
On Tue, 17 Dec 2019 09:51:53 +0200
Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> dma_request_slave_channel() is a wrapper on top of dma_request_chan()
> eating up the error code.
>
> By using dma_request_chan() directly the driver can support deferred
> probing against DMA.
>
> Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Hi Peter,
Change looks sensible to me, but I'd like to leave a little longer
for others to look at this.
Give me a poke if I seem to have lost it by the end of the first
week in Jan.
Thanks,
Jonathan
> ---
> drivers/iio/adc/stm32-adc.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/iio/adc/stm32-adc.c b/drivers/iio/adc/stm32-adc.c
> index 3b291d72701c..5dab23f1fdee 100644
> --- a/drivers/iio/adc/stm32-adc.c
> +++ b/drivers/iio/adc/stm32-adc.c
> @@ -1746,9 +1746,15 @@ static int stm32_adc_dma_request(struct iio_dev *indio_dev)
> struct dma_slave_config config;
> int ret;
>
> - adc->dma_chan = dma_request_slave_channel(&indio_dev->dev, "rx");
> - if (!adc->dma_chan)
> + adc->dma_chan = dma_request_chan(&indio_dev->dev, "rx");
> + if (IS_ERR(adc->dma_chan)) {
> + if (PTR_ERR(adc->dma_chan) == -EPROBE_DEFER)
> + return -EPROBE_DEFER;
> +
> + /* Ignore errors to fall back to IRQ mode */
> + adc->dma_chan = NULL;
> return 0;
> + }
>
> adc->rx_buf = dma_alloc_coherent(adc->dma_chan->device->dev,
> STM32_DMA_BUFFER_SIZE,
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead dma_request_slave_channel()
2019-12-23 15:42 ` Jonathan Cameron
@ 2020-01-06 10:25 ` Fabrice Gasnier
0 siblings, 0 replies; 3+ messages in thread
From: Fabrice Gasnier @ 2020-01-06 10:25 UTC (permalink / raw)
To: Jonathan Cameron, Peter Ujfalusi
Cc: Olivier MOYSAN, alexandre.torgue, linux-iio, linux-kernel, vkoul,
mcoquelin.stm32, linux-stm32, linux-arm-kernel
On 12/23/19 4:42 PM, Jonathan Cameron wrote:
> On Tue, 17 Dec 2019 09:51:53 +0200
> Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
>
>> dma_request_slave_channel() is a wrapper on top of dma_request_chan()
>> eating up the error code.
>>
>> By using dma_request_chan() directly the driver can support deferred
>> probing against DMA.
>>
>> Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
> Hi Peter,
>
> Change looks sensible to me, but I'd like to leave a little longer
> for others to look at this.
>
> Give me a poke if I seem to have lost it by the end of the first
> week in Jan.
>
> Thanks,
>
> Jonathan
>
>> ---
>> drivers/iio/adc/stm32-adc.c | 10 ++++++++--
>> 1 file changed, 8 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/iio/adc/stm32-adc.c b/drivers/iio/adc/stm32-adc.c
>> index 3b291d72701c..5dab23f1fdee 100644
>> --- a/drivers/iio/adc/stm32-adc.c
>> +++ b/drivers/iio/adc/stm32-adc.c
>> @@ -1746,9 +1746,15 @@ static int stm32_adc_dma_request(struct iio_dev *indio_dev)
>> struct dma_slave_config config;
>> int ret;
>>
>> - adc->dma_chan = dma_request_slave_channel(&indio_dev->dev, "rx");
>> - if (!adc->dma_chan)
>> + adc->dma_chan = dma_request_chan(&indio_dev->dev, "rx");
>> + if (IS_ERR(adc->dma_chan)) {
>> + if (PTR_ERR(adc->dma_chan) == -EPROBE_DEFER)
>> + return -EPROBE_DEFER;
Hi Peter,
Thanks for the patch and sorry for the late reply.
I'd rather prefer to check explicitly for -ENODEV (as the DMA is
optional) to use the IRQ mode and treat all other codes (including
EPROBE_DEFER) as errors.
Rationale is: This can hide other errors e.g. like all DMA channels are
busy/reserved for other usage. So the user may wrongly think it's
probing the driver, with DMA. This may be a corner case, but still...
DMA channels are assigned via device tree. I'd rather prefer to avoid
depending of runtime (probe ordering) in such a case.
Can you update the patch considering this ?
Please find here an alternate proposal:
adc->dma_chan = dma_request_chan(&indio_dev->dev, "rx");
if (IS_ERR(adc->dma_chan)) {
if (PTR_ERR(adc->dma_chan) != -ENODEV)
return PTR_ERR(adc->dma_chan);
/* DMA is optional: fall back to IRQ mode */
adc->dma_chan = NULL;
return 0;
}
Best regards,
Fabrice
>> +
>> + /* Ignore errors to fall back to IRQ mode */
>> + adc->dma_chan = NULL;
>> return 0;
>> + }
>>
>> adc->rx_buf = dma_alloc_coherent(adc->dma_chan->device->dev,
>> STM32_DMA_BUFFER_SIZE,
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2020-01-06 10:25 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-12-17 7:51 [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead dma_request_slave_channel() Peter Ujfalusi
2019-12-23 15:42 ` Jonathan Cameron
2020-01-06 10:25 ` Fabrice Gasnier
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).