Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 10 May 2010 01:28:44 +0200
From:      "Julian H. Stacey" <jhs@berklix.com>
To:        Nathan BIAGINI <nathan.open@gmail.com>
Cc:        freebsd-mobile@freebsd.org
Subject:   Re: high cpu temp and fan speed problem 
Message-ID:  <201005092329.o49NSiVI045698@fire.js.berklix.net>
In-Reply-To: Your message "Sun, 09 May 2010 14:45:22 %2B0200." <AANLkTimtNdxE9q9Mgc7VhMrrpanFl07Hn0rTs0P2ws5T@mail.gmail.com> 

next in thread | previous in thread | raw e-mail | index | archive | help
Hi Nathan,
I restored cc freebsd-mobile@freebsd.org
& converted top post to bottom post.

Reference:
> From:		Nathan BIAGINI <nathan.open@gmail.com> 
> Date:		Sun, 9 May 2010 14:45:22 +0200 
> To: "Julian H. Stacey" <jhs@berklix.com>	

> 
> 2010/5/9, Julian H. Stacey <jhs@berklix.com>:
> > Nathan BIAGINI wrote:
> >> Hi,
> >>
> >> i bought a hp pavilion dv6-1123ef laptop and i've installed FreeBSD 8.0 on
> >> it. Everything work except one think : i can't do high-cpu task. The
> >> problem
> >> is when i run a high-cpu task (no very high in fact), my cpu temperature
> >> can
> >> increase to 90°C and when it's the case, system go down. Further, i heard
> >> the fans are running. When i work on windows (short time), i never had
> >> this
> >> kind of problem so i think is may be a kernel config problem.
> >>
> >> I precise that when i want to run mbmon to minitor cpu temp and fan speed
> >> (compiled form the ports), it return an unknow error like what no hardware
> >> monitor is found, the cause of my problem?
> >>
> >> Thanks for helping.
> >
> >> PS : sorry if i'm not really clear but english isn't my native language.
> >

> "Julian H. Stacey" <jhs@berklix.com> wrote: 

> > All understandable, except 2nd para. needs:   s/I precise/I presume/	:-)
> >
> > Others know power better than me, but here's a first answer,
> > mouse copied from my laptop, to search for keyword to read manuals for,
> > while you wait for better answers :-)
> >
> > man powerd	# try powerd -v
> > sysctl -a | grep acpi
> > man acpi
> >
> > My /boot/loader.conf	# man loader.conf
> > 	acpi_toshiba_load="YES"
> > 	# apm_load="YES"
> > 	boot_verbose="yes"            # Collect more info to later tune
> > 	debug.bootverbose="1"         # Variable shown by sysctl -a.
> > 	# one of those 2 verbose is old syntax I think
> >
> > My /etc/rc.conf	# See also /etc/defaults/rc.conf
> > 	# apm_enable="NO"       # Set to YES to enable APM BIOS functions (or NO).
> > 	# apmd_enable="NO"      # Run apmd to handle APM event from userland.
> > 	# apmd_flags=""         # Flags to apmd (if enabled).
> > 	
> > 	powerd_enable="YES"      # /etc/defaults/rc.conf = "NO"
> > 	powerd_flags="-a hiadaptive -b adaptive -n adaptive"
> > 	#       -a maximum # too hot underneath, & USB2 cardbus might cook.
> > 	#       -b minimum      # just for max life if editing & not processing
> > 	# man powerd: Default adaptive for battery and hiadaptive for the rest.
> > 	# sysctl dev.cpu.0.freq         # shows 300 to 1900
> > 	case $hostname in       #{
> > 		homehostname)
> > 	   powerd_flags="-a adaptive -b minimum  -n minimum"  # Less Heat
> > 			;
> > 		no_net_hostname)
> > 	   # powerd_flags="-a maximum  -b adaptive -n adaptive" # More Perf
> > 	                  powerd_flags="-a adaptive -b minimum  -n minimum"  # Less
> > Heat
> > 	                # powerd I may be at a lake. No power.
> > 			;
> > 	esac    # }
> >
> > Cheers,
> > Julian

Nathan BIAGINI wrote Sun, 9 May 2010 14:45:22 +0200:
> Thanks for it Julian. Powerd in adaptive mode don't solve anything. I

I suppose not, if you've given it a long compile intensive task.
   ( Though it keeps my laptop cooler to the touch when idling, &
   sysctl dev.cpu.0.freq reports 300 instead of 1900, ie 16%, ( &
   if one assumes CMOS power is a square of frequency, I suppose
   my CPU might be consuming 300^2 / 1900^2 ie 2.5% ))


> add acpi_intel_load="YES" in my loader.conf file and i'm trying to
> understand most of the command specified on acpi manual.
> 
> Other idea, it is maybe a problem with the fans. The fans frequency is
> may be to small, i will try to learn more about that.

A lots of sysctl values are read only, you can not set them.  I
assume one cannot set fan speed by sysctl (though I've seen tower
PCs.  with manual speed control knobs at the back) I assume one
also cannot set some temperatures, except eg power down threshold,

Maybe your fan has a stiff bearing when hot, or fluff in air line.
Try turning off & spinning fan with a small stick (or flexible
plastic refill of a Biro (ball point pen).

When not mobile, (& thus powered, from mains 220V, & may be running
long at full speed) I place my laptop on larger feet, to allow more
air under (especially as cantilevered laptop stand has raised folded
metal sides that block air).

My cardbus also gets very hot, & has crashed my laptop before, I
use an external powered USB hub to avoid my cardbus also having to
power external USB devices (Each USB socket can provide up to 5V x
0.5A = 2.5W).

Some cheap laptops [used to] use CPUs sold for towers, they needed
more air cooling. I guess CPUs still come in ranges of power
consumption that may tempt laptop manufacturers to design for one
CPU, then get desperate on price, & build in cheaper hotter chips.


> The mbmon soft alway return the same error and i don't find anything
> on google about this error (only halt answer).

Mbmon doesnt work on my laptop either (but fan rarely goes on in
idel with powerd)

PS I see quite a lot of "cdev=apm" & "Processing event '!system=ACPI
subsystem=Thermal type=\_TZ_.THRM notify=0x80'" entries with devd
(replacement of USBD on older FreeBSD).  Maybe check if you have
devd running (devd fails if there is a syntax error in config file).

> So now i will read manual on my noisy laptop :-)

I too should again read man acpi_thermal


Cheers,
Julian
-- 
Julian Stacey: BSD Unix Linux C Sys Eng Consultants Munich http://berklix.com
Mail plain text,  Not HTML quoted-printable Base64 http://www.asciiribbon.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201005092329.o49NSiVI045698>