Date: Thu, 30 Jun 2011 23:59:08 +0300 From: Andriy Gapon <avg@FreeBSD.org> To: Ed VanderPloeg <edv@agile.bc.ca> Cc: freebsd-acpi@FreeBSD.org Subject: Re: Atom N270 - ACPI Error: [RTMP] Namespace lookup failure Message-ID: <4E0CE39C.5050307@FreeBSD.org> In-Reply-To: <4E0CA533.5030104@agile.bc.ca> References: <4E0A50AA.5000003@agile.bc.ca> <4E0AF27B.3030600@FreeBSD.org> <4E0B6873.6010901@agile.bc.ca> <4E0B80BE.6080605@FreeBSD.org> <4E0CA533.5030104@agile.bc.ca>
next in thread | previous in thread | raw e-mail | index | archive | help
on 30/06/2011 19:32 Ed VanderPloeg said the following: > I updated to 8-stable but am still getting ACPI error messages to console every > 10 seconds. Just for my curiosity - has anything changed with respect to est driver attachment? > What happens when sysctl hw.acpi.thermal.polling_rate=0? Does this disable > polling? I noticed that when I set it to zero, the error messages seem to stop, > but then setting it to a non-zero value never brings the messages back again. I've just the code in acpi_thermal driver and it doesn't have any validation for polling_rate and thus no special treatment for the value of zero. So, it passes timeout of zero to msleep() function, which does have a special meaning for zero - it means sleep forever until wakeup event. Essentially when you did sysctl hw.acpi.thermal.polling_rate=0, you made the thermal zone handling thread to sleep "forever". This can be considered a bug in FreeBSD ACPI TZ driver. The only thing that seems to be able to wake up the thread after that is a change of power profile. So switching from AC to batter or vice versa could wake up the thread and make it use a new value of polling_rate. > On 2011-06-29 12:45 PM, Andriy Gapon wrote: >> on 29/06/2011 21:01 Ed VanderPloeg said the following: >>> On 2011-06-29 2:38 AM, Andriy Gapon wrote: >>>> on 29/06/2011 01:07 Ed VanderPloeg said the following: >>>>> I'm using an Aaeon AEC-6831 embedded system based on an Intel Atom N270, which >>>>> uses their GENE-9455 motherboard. After updating the BIOS to enable ACPI, I'm >>>>> now getting the following (verbose) console message during boot and every 10 >>>>> seconds thereafter: >>>>> >>>>> ACPI Error: [RTMP] Namespace lookup failure, AE_NOT_FOUND >>>>> (20101013/psargs-464) >>>>> ACPI Error: Method parse/execution failed [\_TZ_.THRM._TMP] (Node 0xc56b0760), >>>>> AE_NOT_FOUND (20101013/psparse-633) >>>>> acpi_tz0: error fetching current temperature -- AE_NOT_FOUND >>>> >>>> The problem is that RTMP is defined as an external object: >>>> External (RTMP, IntObj) >>>> so it's supposed to come from an additional table, but apparently either no >>>> additional table defines it or a necessary additional table is not loaded. >>>> This could be either a BIOS problem or... something else :) >>> >>> Aaeon tech support has now stated that the errors are from a faulty BIOS, and >>> that AWARD will eventually release an update to fix this. >> >> OK. >> >>>>> The unit seems to run very warm which makes me wonder if this problem is >>>>> preventing lower power states, if such things are related. >>>>> >>>>> I've collected the outputs from a verbose dmesg, from sysctl hw.acpi, and from >>>>> acpidump. They are zipped up over here: >>>>> >>>>> http://www.agilecontrols.com/post/aec6831_acpi.zip >>>> >>>> Try either recent stable/8 or head (aka CURRENT) and see if it helps. They >>>> contain a change that may be a work-around for a BIOS (ACPI tables) like yours. >>> >>> I'll give stable/8 a try. >>> >>> Does this error indicate something potentially harmful, or if it is benign? I >>> can silence the messages easy enough until a BIOS update comes out. >> >> >> Actually I was speaking about potentially making est driver attach on your >> system. I also suspected that RTMP might have gotten loaded from the same >> dynamic table that is related to est attachment issue, but apparently that was >> not going to happen. >> -- Andriy Gapon
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4E0CE39C.5050307>