Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH] iio: adc: stm32-adc: Use dma_request_chan() instead
 dma_request_slave_channel()
To:     Jonathan Cameron <jic23@kernel.org>,
        Peter Ujfalusi <peter.ujfalusi@ti.com>
CC:     <mcoquelin.stm32@gmail.com>, <alexandre.torgue@st.com>,
        <vkoul@kernel.org>, <linux-iio@vger.kernel.org>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>,
        <linux-stm32@st-md-mailman.stormreply.com>,
        Olivier MOYSAN <olivier.moysan@st.com>
References: <20191217075153.23766-1-peter.ujfalusi@ti.com>
 <20191223154230.7fb2a988@archlinux>
From:   Fabrice Gasnier <fabrice.gasnier@st.com>
Message-ID: <9f190583-824c-01a6-7da3-43065fde49c7@st.com>
Date:   Mon, 6 Jan 2020 11:25:11 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.9.0
MIME-Version: 1.0
In-Reply-To: <20191223154230.7fb2a988@archlinux>
Content-Type: text/plain; charset="utf-8"
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

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,
>