https://bugzilla.kernel.org/show_bug.cgi?id=216230 --- Comment #23 from madcatx@xxxxxxxx --- (In reply to Mario Limonciello (AMD) from comment #22) > Were you still using the initramfs method when 33000 came through? > If so; please switch to built-in. That should allow the driver to load as > early as possible. Just tried that and the number of unserviced interrupts dropped down to less than 8000; that's the lowest value I've seen so far. pinctrl_amd gets loaded about 0.48 secs into boot, it loaded at about 0.67 sec when it was a module. Anyway, it looks like we pinned the problem down pretty well. Thanks! -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
