Re: [PATCH 13/14] Blackfin SPI driver: Move cs_chg_udelay to cs_deactive to fix bug when some SPI LCD driver needs delay after cs_deactive

[Bryan Wu <bryan.wu@analog.com>]

On Tue, 2007-10-30 at 13:18 -0700, David Brownell wrote:
> On Tuesday 30 October 2007, Bryan Wu wrote:
> > @@ -211,6 +211,10 @@ static void cs_deactive(struct driver_data *drv_data, struct chip_data *chip)
> >         flag |= (chip->flag << 8);
> >  
> >         write_FLAG(drv_data, flag);
> > +
> > +       /* Move delay here for consistency */
> > +       if (chip->cs_chg_udelay)
> > +               udelay(chip->cs_chg_udelay);
> >  }
> >  
> 
> By the way, if this is something needed very often, that mechanism
> should be moved into the SPI framework.  It wouldn't be polite if
> such LCD drivers could only work on Blackfin boards.  :)

I am not sure whether this LCD needs cs_chg_udelay on other boards.
It is found by Cameron Barfield. Maybe it is some timing requirement for
specific hardware.

-Bryan Wu