On Fri, May 29, 2026 at 02:17:00PM +0300, Svyatoslav Ryhel wrote:

пт, 29 трав. 2026 р. о 14:10 Daniel Thompson [off-list ref] пише:

quoted

On Thu, May 28, 2026 at 04:51:22PM +0300, Svyatoslav Ryhel wrote:

quoted

Add support to obtain the initial mapping mode from DT instead of leaving
it unconfigured. Additionally, update the linear sysfs code, which uses a
similar coding pattern.

Words like "additionally" in a patch description can be a sign the patch
should actually be two patches. In this case the patch would be a lot
easier to read if you cleaned up the linear sysfs code (patch N) and then
added the new DT logic (patch N+1).

I looked into this in reverse. My goal was to add DT logic I don't
case how sysfs works. My code matched with what sysfs does I just
included sysfs change as well. I might better drop sysfs changes
entirely since with such pace this patchset will inflate from 6 to 15
and beyond.

Not sure about that. The clean up is good and introducing the new
BIT(2 * n + 1) macro does need to be used pervasively or there's no
point in adding it.


Daniel.