on 05/11/2012 16:52 Tom Lislegaard said the following: > > >> -----Original Message----- >> From: Andriy Gapon [mailto:a...@freebsd.org] >> Sent: 5. november 2012 15:21 >> To: Tom Lislegaard >> Cc: freebsd-stable@FreeBSD.org; freebsd-a...@freebsd.org >> Subject: Re: 9-Stable panic: resource_list_unreserve: can't find resource >> >> on 05/11/2012 15:54 Tom Lislegaard said the following: >>> Here's the distribution from running devd over 40 minutes >>> >>> 589 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU0 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU1 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU2 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU3 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU4 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU5 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU6 >>> notify=0x81' >>> 590 Processing event '!system=ACPI subsystem=PROCESSOR type=\\_PR_.CPU7 >>> notify=0x81' >>> 1 Processing event '!system=DEVFS subsystem=CDEV type=CREATE >>> cdev=dsp4.1' >>> 1 Processing event '!system=DEVFS subsystem=CDEV type=CREATE cdev=pts/2' >>> 1 Processing event '!system=DEVFS subsystem=CDEV type=CREATE >>> cdev=vboxdrv0' >>> 1 Processing event '!system=DEVFS subsystem=CDEV type=DESTROY >>> cdev=dsp4.1' >>> 1 Processing event '!system=DEVFS subsystem=CDEV type=DESTROY >>> cdev=vboxdrv0' >>> >>> Any use in running it over a longer period of time? >> >> Very interesting. >> So the processors get _CST change notifications rather frequently, but there >> is no obvious >> source/cause for them... >> >> Could you please send to me acpidump -dt output or upload it somewhere and >> post a link? > > Here you go > > https://dl.dropbox.com/u/13263820/acpidump.txt
So, ACPI platform on your machine sends 0x81 notification for processors objects each time "_PSR" method of AC Adapter / Power Source device is queried. There could be a number of reason to invoke the method - either AC line status queries from userland (some battery / line monitoring program/applet) or internal ACPI notifications. Are you willing to go as far as recompiling your kernel with 'options ACPI_DEBUG' to get to the bottom of this issue? If yes, then please do that and also add the following lines to loader.conf: debug.acpi.layer="ACPI_EVENTS ACPI_AC_ADAPTER" debug.acpi.level="ACPI_LV_INFO" I would be interested in all periodically occurring ACPI debug messages (after boot is finished). I suspect that the ACPI platform and/or embedded controller send too many notifications when they are not strictly necessary. Maybe there is a BIOS update for your machine? In any case, I am starting to work on a patch that should fix this problem without resorting to any special configuration. -- Andriy Gapon _______________________________________________ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
