Hi Marcus, On Fri, Nov 20, 2015 at 7:45 PM, Marcus Weseloh <mweseloh42@xxxxxxxxx> wrote: > Hi Julian, > > 2015-11-19 23:59 GMT+01:00 Julian Calaby <julian.calaby@xxxxxxxxx>: >> Should you possibly hide the 3 clock periods from the user? >> >> I.e. they set whatever they want for the wdelay, we set it to the >> closest number we can that's greater or equal to what they ask for. > > That's a good idea and much better than having to remember to subtract > 3 cycles from the desired wait time! > > But it would mean that this magic number becomes part of the driver > code. I have found no official documentation that mentions those > additional cycles. While I have checked many different transmission > speeds using both CDR1 and CDR2 divider configurations, there is still > the possibility that the behaviour changes with weird SPI module > configurations... And I've only tested it on A20 hardware. So it would > be great if somebody else with access to A10 hardware and an > oscilloscope could check if we have a consistent 3 cycle overhead. Having magic numbers is kind-of a drivers' job. (and the wdelay should arguably be a core-spi thing, not a sunxi thing, but that's a separate discussion) If it is different for other SoCs, then you might have to move that constant somewhere else and introduce new compatible strings for them. Thanks, -- Julian Calaby Email: julian.calaby@xxxxxxxxx Profile: http://www.google.com/profiles/julian.calaby/ -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html