Re: [PATCH 1/3] spi: Re-do the returning value of rspi_dma_check_then_transfer

[Cao Minh Hiep <cm-hiep@jinso.co.jp>]

Hi Geert-san,

On 2015年04月28日 18:02, Geert Uytterhoeven wrote:
> Hi Hiep-san,
>
> On Tue, Apr 28, 2015 at 7:55 AM, Cao Minh Hiep <cm-hiep@jinso.co.jp> wrote:
>> To reduce indentation and complex of code, insteeds of returning zero,
>> the function rspi_dma_check_then_transfer should be returned
>> rspi_dma_transfer directly after checking error.
> Thanks for the update!
>
> Unfortunately it doesn't compile...
Sorry to bother you again.
>> --- a/drivers/spi/spi-rspi.c
>> +++ b/drivers/spi/spi-rspi.c
>> @@ -665,15 +665,12 @@ static bool rspi_can_dma(struct spi_master *master, struct spi_device *spi,
>>   static int rspi_dma_check_then_transfer(struct rspi_data *rspi,
>>                                           struct spi_transfer *xfer)
>>   {
>> -       if (rspi->master->can_dma && __rspi_can_dma(rspi, xfer)) {
>> -               /* rx_buf can be NULL on RSPI on SH in TX-only Mode */
>> -               int ret = rspi_dma_transfer(rspi, &xfer->tx_sg,
>> -                                       xfer->rx_buf ? &xfer->rx_sg : NULL);
>> -               if (ret != -EAGAIN)
>> -                       return 0;
>> -       }
>> +       if (!rspi->master->can_dma || !__rspi_can_dma(rspi, xfer)) {
> Superfluous opening curly brace.
> Please compile and run test your patches.
> Thanks!

Thanks, I'll fix it and send you at a moment.

>> +               return -EAGAIN;
>>
>> -       return -EAGAIN;
>> +       /* rx_buf can be NULL on RSPI on SH in TX-only Mode */
>> +       return rspi_dma_transfer(rspi, &xfer->tx_sg,
>> +                                       xfer->rx_buf ? &xfer->rx_sg : NULL);
>>   }
>>
>>   static int rspi_common_transfer(struct rspi_data *rspi,
>
>
>

Hiep.