On 28/05/2013, at 1:15 AM, Arnd Bergmann <a...@arndb.de> wrote: > On Monday 27 May 2013, Daniel Tang wrote: >> Before any peripheral is accessed. I.e. before the clocksource and irqchip >> drivers. > > The irqchip comes first, and by that time, you can actually call > of_iomap(). > >> The write to the port is supposed to ensure all mmio peripherals can be >> accessed. Without it, access to certain peripherals will result in >> undefined reads or ignored writes. >> >> On second thoughts, would this actually be the job of the boot loader? > > Doing it in the boot loader would certainly simplify things. I wonder > about the dynamic aspects of power management though: It might be > better to expose the individual bits of this register through a proper > driver. The boot loader can start out enabling everything, but then > you turn off everything that is not needed when that driver gets > loaded.
That's the idea for the long term. For now though, I'll probably just let the bootloader enable everything and work on a proper driver for power management later. > > I'm still not sure what the register actually does: Does it > control reset lines, clock signals, voltage regulators or something > else? These things all have their own subsystems, and then there > is also the power domain framework. To be perfectly honest, I'm not too sure. The documentation for the TI-Nspire is all gathered from reverse engineering and all it says is that register "disables bus access to peripherals". > > Arnd Cheers, Daniel Tang-- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
