Fan resume not working on nx6325 without userland assistance

Fan resume not working on nx6325 without userland assistance

[Rafael J. Wysocki]

Hi,

It turns out that the following script (from openSUSE 10.3):

#############################################################
# triggers the ACPI fan(s) after resume. Since ACPI drivers
# have no suspend support, this is sometimes necessary.
# see http://article.gmane.org/gmane.linux.acpi.devel/16643
kick-fan()
{
        local FAN DUMMY STATE
        for FAN in /proc/acpi/fan/*/state; do
                [ ! -e $FAN ] && continue
                read DUMMY STATE < $FAN
                if [ "$STATE" = "on" ]; then
                        echo "kicking $FAN"
                        echo -n 3 > $FAN
                        echo -n 0 > $FAN
                fi
        done
}

case $1 in
        thaw|resume)
                kick-fan
                ;;
esac

is necessary to make the fan behave appropriately after a resume from RAM
(I haven't checked resume from hibernation, but I guess the same thing happens)
on HP nx6325 with 2.6.24-rc8 (and with previous kernels probably too).

It wasn't needed at one point, so something has regressed.  Sigh.

Greetings,
Rafael

Re: Fan resume not working on nx6325 without userland assistance

[Rafael J. Wysocki]

On Tuesday, 22 of January 2008, Rafael J. Wysocki wrote:
> Hi,
> 
> It turns out that the following script (from openSUSE 10.3):
> 
> #############################################################
> # triggers the ACPI fan(s) after resume. Since ACPI drivers
> # have no suspend support, this is sometimes necessary.
> # see http://article.gmane.org/gmane.linux.acpi.devel/16643
> kick-fan()
> {
>         local FAN DUMMY STATE
>         for FAN in /proc/acpi/fan/*/state; do
>                 [ ! -e $FAN ] && continue
>                 read DUMMY STATE < $FAN
>                 if [ "$STATE" = "on" ]; then
>                         echo "kicking $FAN"
>                         echo -n 3 > $FAN
>                         echo -n 0 > $FAN
>                 fi
>         done
> }
> 
> case $1 in
>         thaw|resume)
>                 kick-fan
>                 ;;
> esac
> 
> is necessary to make the fan behave appropriately after a resume from RAM
> (I haven't checked resume from hibernation, but I guess the same thing happens)
> on HP nx6325 with 2.6.24-rc8 (and with previous kernels probably too).
> 
> It wasn't needed at one point, so something has regressed.  Sigh.

Ah, I didn't say that thermal management is completely busted after a resume
if the fan is not kicked from the user land (using the above script).  Usually
the fan is 100% on (that corresponds to all ACPI "fans" being on), but once it
had gone off and I was unable to turn it on by any means (including rising the
temperature to a dangerous level).

Greetings,
Rafael

Re: Fan resume not working on nx6325 without userland assistance

[Alexey Starikovskiy]

Does revert of 93ad7c07ad487b036add8760dabcc35666a550ef helps?



Rafael J. Wysocki wrote:
> On Tuesday, 22 of January 2008, Rafael J. Wysocki wrote:
>> Hi,
>>
>> It turns out that the following script (from openSUSE 10.3):
>>
>> #############################################################
>> # triggers the ACPI fan(s) after resume. Since ACPI drivers
>> # have no suspend support, this is sometimes necessary.
>> # see http://article.gmane.org/gmane.linux.acpi.devel/16643
>> kick-fan()
>> {
>>         local FAN DUMMY STATE
>>         for FAN in /proc/acpi/fan/*/state; do
>>                 [ ! -e $FAN ] && continue
>>                 read DUMMY STATE < $FAN
>>                 if [ "$STATE" = "on" ]; then
>>                         echo "kicking $FAN"
>>                         echo -n 3 > $FAN
>>                         echo -n 0 > $FAN
>>                 fi
>>         done
>> }
>>
>> case $1 in
>>         thaw|resume)
>>                 kick-fan
>>                 ;;
>> esac
>>
>> is necessary to make the fan behave appropriately after a resume from RAM
>> (I haven't checked resume from hibernation, but I guess the same thing happens)
>> on HP nx6325 with 2.6.24-rc8 (and with previous kernels probably too).
>>
>> It wasn't needed at one point, so something has regressed.  Sigh.
> 
> Ah, I didn't say that thermal management is completely busted after a resume
> if the fan is not kicked from the user land (using the above script).  Usually
> the fan is 100% on (that corresponds to all ACPI "fans" being on), but once it
> had gone off and I was unable to turn it on by any means (including rising the
> temperature to a dangerous level).
> 
> Greetings,
> Rafael

Re: Fan resume not working on nx6325 without userland assistance

[Rafael J. Wysocki]

On Tuesday, 22 of January 2008, Alexey Starikovskiy wrote:
> Does revert of 93ad7c07ad487b036add8760dabcc35666a550ef helps?

Yes, it does, thanks.


> Rafael J. Wysocki wrote:
> > On Tuesday, 22 of January 2008, Rafael J. Wysocki wrote:
> >> Hi,
> >>
> >> It turns out that the following script (from openSUSE 10.3):
> >>
> >> #############################################################
> >> # triggers the ACPI fan(s) after resume. Since ACPI drivers
> >> # have no suspend support, this is sometimes necessary.
> >> # see http://article.gmane.org/gmane.linux.acpi.devel/16643
> >> kick-fan()
> >> {
> >>         local FAN DUMMY STATE
> >>         for FAN in /proc/acpi/fan/*/state; do
> >>                 [ ! -e $FAN ] && continue
> >>                 read DUMMY STATE < $FAN
> >>                 if [ "$STATE" = "on" ]; then
> >>                         echo "kicking $FAN"
> >>                         echo -n 3 > $FAN
> >>                         echo -n 0 > $FAN
> >>                 fi
> >>         done
> >> }
> >>
> >> case $1 in
> >>         thaw|resume)
> >>                 kick-fan
> >>                 ;;
> >> esac
> >>
> >> is necessary to make the fan behave appropriately after a resume from RAM
> >> (I haven't checked resume from hibernation, but I guess the same thing happens)
> >> on HP nx6325 with 2.6.24-rc8 (and with previous kernels probably too).
> >>
> >> It wasn't needed at one point, so something has regressed.  Sigh.
> > 
> > Ah, I didn't say that thermal management is completely busted after a resume
> > if the fan is not kicked from the user land (using the above script).  Usually
> > the fan is 100% on (that corresponds to all ACPI "fans" being on), but once it
> > had gone off and I was unable to turn it on by any means (including rising the
> > temperature to a dangerous level).
> > 
> > Greetings,
> > Rafael
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 



-- 
"Premature optimization is the root of all evil." - Donald Knuth

Re: Fan resume not working on nx6325 without userland assistance

[Rafael J. Wysocki]

On Tuesday, 22 of January 2008, Rafael J. Wysocki wrote:
> On Tuesday, 22 of January 2008, Alexey Starikovskiy wrote:
> > Does revert of 93ad7c07ad487b036add8760dabcc35666a550ef helps?
> 
> Yes, it does, thanks.

Well, reverting commit 93ad7c07ad487b036add8760dabcc35666a550ef is the only
thing that makes it work in all situations, AFAICS.

Len, can you push the revert to Linus for 2.6.24, please?

Thanks,
Rafael