On Fri, Dec 7, 2012 at 12:36 AM, Alan Cox <alan@xxxxxxxxxxxxxxx> wrote: > What about just using 0, 1, 2, 3 to indicate states similar to D0-D3 > which as a base concept everyone is familiar with ? Sorry, I might be something of an oddity, but ACPI is not used on ARM systems as of today, and I was completely confused by these: uart_change_pm(state, 0); uart_change_pm(state, 3); I was just like ... eh ... 0? 3? No ACPI associations existed in my head until I finally found the note in the documentation saying that it corresponds to ACPI states. We are a few kernel developers these days who never ever worked on Intel, sorry for this, but that's what we are like :-) > If we need something more than that, then can we talk about what the > actual extra non 0/3 states needed are and how to get a generic > meaningful definition, or indeed whether a number rather than flag bits > makes sense ? Sure, I think my colleague Rickard will soon post another patch for controlling the state 1,2 bottom-up from drivers. > Start with the driver/platform that causes the concern and then the > rest is probably blindingly obvious. It's the introduction of runtime power management for the ARM amba-pl011.c UART driver that confuse us. It is atleast obvious that none of the 5 ARM platforms using that driver do not use ACPI, and I expect the situation to be the same for most of the ARM drivers. Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
