Currently, if the no_console_suspend isn't set, as soon as system begins the suspend, at core printk level output is suppressed. However, I want to have a simple and clean way to have all console messages in the memory, because on my laptop memory isn't cleared on reboots. My first approach was just to put printk buffer at predefined address. This wasn't favored by kernel developers, and even I myself found this to be not such great solution. Other approach is to create a console driver, however it won't be used by printk ether during suspend and resume. So, maybe we can add a flag that console should always be used, regardless of no_console_suspend? (This is the particular console is safe for that purpose) I can even make the new console pm aware, thus make is possible to avoid collisions between every stage of the hibernation. ( Which happen today because resume kernel overwrites the printk predefined buffer, and that is annoying because my system works just fine now, but I want to keep that feature on) Best regards, Maxim Levitsky _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/linux-pm
