On Tue, Oct 06, 2015 at 01:38:10PM +0200, Martin Sperl wrote: > On 06.10.2015 13:23, Mark Brown wrote: > >On Fri, Sep 11, 2015 at 11:22:04AM +0000, kernel@xxxxxxxxxxxxxxxx wrote: > >>+ bs->irq = irq_of_parse_and_map(pdev->dev.of_node, 0); > >Why are we using this unusual DT specific API here? I'd expect to see > >just use platform_get_irq() here and this seems to be the only DT > >dependency in the driver. > I copied it from spi-bcm2835.c, where it is also used in exactly that > fashion. > You want me to move to your recommended api instead for both drivers? > >I'd expect to request the IRQ after we reset the hardware in order to > >ensure that the hardware is in a known good state > > Same copy/paste in this case - i can change the order for both > drivers if you want. Yes, that'd be better. I'll just apply this now and you can send followup patches fixing these for both drivers.
Attachment:
signature.asc
Description: Digital signature
