**** BEGIN LOGGING AT Mon Jul 04 02:59:58 2011 Jul 04 16:51:52 challenge to the kernel people here... Jul 04 16:52:10 hot to make the pcf50633 driver talk less to the chip? Jul 04 16:52:15 how to Jul 04 16:52:24 without recompiling kernel. Jul 04 16:52:36 is there a sysfs or parameter I can use? Jul 04 16:57:54 mokomoko: but why? Jul 04 16:58:15 mokomoko: it talks to the chip to control charging etc. Jul 04 16:59:57 mokomoko: btw, the driver talks to the chip only when it's asked explicitely by the userspace or when the chip signals an IRQ, so i guess you can't make it talk less to it. Jul 04 17:00:18 mokomoko: but can you please share what do you have i mind asking that? Jul 04 17:00:27 hhmmm.. ok. good. then who in appspace is talking to it? Jul 04 17:01:11 why? because I want the bus mainly for myself. :) real-time robotics without onboard neo battery. Jul 04 17:02:28 there's too mcuh traffic on the bus now, it screws up my system's determinism. Jul 04 17:02:30 mokomoko: some apps that might be accessing /sys/class/power_supply/* or reading/writing to pcf50633 RTC. Jul 04 17:02:44 mokomoko: other than that, it's pcf50633 generating irqs to be handled... Jul 04 17:02:49 AFAICT by a cursory look. Jul 04 17:03:36 well, I've uninstalled pretty much everything, the only app still running is deviced from SHR I think. Jul 04 17:04:36 or some other daemon, I can't check right now, but I remember it was related to device management, not phone. Jul 04 17:05:34 ok, will check exactly which daemon it is and kill it. but then if there is still talk on the bus, it should be IRQ... Jul 04 17:06:24 would I be able to change that without changing kernel code? Jul 04 17:27:51 mokomoko: in fact yes, you can use "pmu" utility to write directly to the chip, that way you can disable irqs :) Jul 05 01:14:57 Okay, I'm here. (courtesy of docscrutinizer) **** ENDING LOGGING AT Tue Jul 05 02:59:57 2011