On Mon, Oct 10, 2022 at 08:29:05PM +0200, Rafael J. Wysocki wrote: > > It failed 3/10 times. > > This is still not acceptable. > Agreed. > > That's less than the previous 5/10 failures but I > > cannot be certain it helped without running a lot more boot tests. The > > failure happens in the same function as before. > > I've overlooked the fact that acpi_install_fixed_event_handler() > enables the event on success, so it is a bug to call it when the > handler is not ready. > > It should help to only enable the event after running cmos_do_probe() > where the driver data pointer is set, so please try the attached > patch. Looks good and it booted 10 times successfully. -- Mel Gorman SUSE Labs
