On Thursday 05 of December 2013 10:35:20 Doug Anderson wrote: > Tomasz, > > On Thu, Dec 5, 2013 at 10:30 AM, Tomasz Figa <tomasz.figa@xxxxxxxxx> wrote: > >> I'd vote for using "pmu-system-registers". We end up using the > >> "syscon" subsystem but really we're describing pmu registers. > >> > >> I'd even say that you don't need to formally specify the "name" in the > >> bindings (though I'm not up with all the latest device tree > >> requirements). ...still you'd want to use "pmu-system-registers" in > >> the DTS changes. > > > > Well, since the name should specify the class of device, I would say that > > pmu-system-registers is too specific. If we want to change this, I'd say > > we should go with system-controller. > > ...but the "compatible" is "samsung,exynos5250-pmu", "syscon", right? > That means that the class of the device is "exynos5250-pmu", right? Nope. "samsung,exynos5250-pmu" is the specific device (or hardware programming interface) this device is compatible with. With class I mean the generic kind of device, such as system-controller, i2c, pinctrl, display, etc., as specified by sections 2.2.1 and 2.2.2 of ePAPR. Anyway, node names are just a matter of coding style, as they don't have any semantical meaning in most cases (such as this one). > It is also compatible with the generic "syscon" class of devices. It is also compatible with the generic "syscon" programming interface, which represents a set of loosely related registers that control various aspects of other IP blocks. Best regards, Tomasz -- 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