On Sun, 2022-09-11 at 12:26 +0100, Jonathan Cameron wrote: > On Fri, 9 Sep 2022 17:14:11 +0200 > Nuno Sá <nuno.sa@xxxxxxxxxx> wrote: > > > Some of the supported devices have 4 or 2 LSB trailing bits that > > should > > not be taken into account. Hence we need to shift these bits out > > which > > fits perfectly on the scan type shift property. This change fixes > > both > > raw and buffered reads. > > Hi Nuno, Hi Jonathan, > > Seems that all the values of shift are 12 - realbits. > If that's the case, can we reduce the noise this patch creates by > just > updating AD7923_V_CHAN() to set .shift = 12 - (bits) ? > Yes, it should be pretty much the same... As I don't have any strong feelings I can do as you suggest. > I guess that's not as flexible if anyone adds support for a device > with different shifts, but I suspect that may never happen. > Or a device with realbits > 12. But yeah, I'm also fairly positive we won't see that happening... - Nuno Sá
