On Sun, 29 Aug 2010, Rafael J. Wysocki wrote: > > > It seems to me that the standard for /sys is to use miliseconds. > > > > "45 minutes" expressed in milliseconds isn't very easy to comprehend; > > it comes out as "2700000". Compare this to "45:00.000". I'm not sure > > that's the best way to go. > > Well, that depends on whether or not you intend to make them human-readable. > While the latter is definitely better for humans, the former may be more > convenient for software processing. It seems likely that the most common usage pattern will be via a GUI control, so the GUI can do the formatting and conversion. In that case I think the attribute should be named "suspend_delay_ms". Putting the units into the name like this seems to be a common pattern. Should we continue the convention (used in the USB subsystem) that negative values for suspend_delay will disable autosuspend? I'm not sure how useful that is, given that setting power/control to "on" has the same effect. Oliver, you originally argued strongly in favor of having both mechanisms. Do you still think they are both needed? One last thing... I'm definitely planning to make the suspend_delay value affect suspends initiated by pm_schedule_suspend. (The routine will always use the maximum of the specified delay and the value calculated from last_busy plus suspend_delay. Also, pm_suspend_timer_fn would check to see if last_busy has changed and restart the timer if needed.) I'm not sure whether suspends initiated from pm_runtime_suspend or pm_request_suspend should behave the same way. What do you think? Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
