Re: [PATCH] spi: core: use gpio_request_one for configured gpio cs

[Chris Packham <Chris.Packham@xxxxxxxxxxxxxxxxxxx>]

On 25/07/17 00:42, Mark Brown wrote:
> On Mon, Jul 24, 2017 at 01:49:14PM +0300, Andy Shevchenko wrote:
>> On Mon, Jul 24, 2017 at 12:53 PM, Martin Hundebøll <mnhu@xxxxxxxxx> wrote:
>>> Some gpio controllers requires lines to be configured as outputs to make
>>> subsequent calls to `gpio_set_value()` have any effect.
> 
>>> Add a call to `gpio_request_one()` for configured gpio chip-selects to
>>> make it work on (at least) mvebu gpio controllers.
> 
>> More and more hacks around plain GPIO integers... :-(
> 
>> Perhaps we need to discuss how to convert SPI core to use GPIO
>> descriptors everywhere.
> 
> The best way would be to test the work that Chris Packham submitted in
> May to do this, it'll need a refresh now and it ought to get some
> testing on ep93xx but sadly nobody working on ep93xx has responded - I
> think at this point I'd probably just apply and let people test -next.
> 

My patchset kind of stalled. I can do a refresh if there's interest. But 
I really do need some help from people with access to hardware (I can 
cover spi-orion and spi-fsl-espi but that's about it). After the initial 
RFC series I posted in May I found that iMX platforms use the integer 
cs_gpio field for both native chipselects and gpio based ones (values < 
0 indicate native CS) so again I'd be coding blind trying to disentangle 
that.
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html