[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Hibernate fails



On 04/07/11 06:37, Carl Fink wrote:
> On Fri, Jul 01, 2011 at 06:48:33PM +1000, Scott Ferguson wrote:
> 
>> If ACPI is broken you'll see messages in dmesg.
> 
> A quick scan shows this:
>    26.446021] [Firmware Bug]: powernow-k8: No compatible ACPI _PSS objects found.
> [   26.446024] [Firmware Bug]: powernow-k8: Try again with latest BIOS.
> 
> There are no updates available for my BIOS.


The Microsoft supports it but Linux doesn't scenario.... might be that
"Linux" isn't a Gold Standard OEM partner :-/

Probably/possibly three paths from here:-

;1. try other hibernation methods
;2. pinch ACPI support from somewhere else
;3. dodge the BIOS and look for Coreboot support

The first two approaches seem like a lot of work for something I never
use (even well supported netbooks shutdown and restart quicker with a
little tuning than a tweaked hibernation scheme) :-(

Approach #3 is always my preferred approach, it's probably the most
difficult too.

Approach #2 might solve some other (possibly) ACPI related issues you
mentioned. Gateway tend to re/use other manufactures hardware, generally
with little modification (they also tend to bastardise DSDT too). If you
can find another netbook that uses the Phoenix BIOS equivalent of
Version Gateway LT31 v1.3307 - then "borrow" the custom acpi module.
If you feel particularly brave - you can try a modified BIOS (eg.from
Kizwan). If you decide you'd like to fix the missing ACPI support a
modified BIOS that supports DSDT might be a useful tool ;-p

Approach #1a is to use APM(d) for (swsusp) hibernation, or:-
#1b TuxOnIce. AFAIK TuxOnIce is not prebuilt for Wheezy, but it (IMHO)
is superior to swsusp, and as it's more similar to Microsoft's hibernate
process - possible it will work on your netbook without tweaking.
> 
>> Thanks for that
>> Do you have a /var/log/wpa_action.log??
>> If so please post
> 
> No such file.

K. I was "hoping" to find the location of wpa control interface there,
but chances are the broken (non-spec) ACPI interface is the problem.
Next time I'll think to ask for any pertinent dmesg errors right at the
start of a thread (mea culpa etc). The LT31xx have a history of
weirdness with hibernation eg. whether the power cable is plugged in or
not causing problems resuming from hibernation.

If I was stuck with your netbook - I'd create a list of core hardware
components, put them into a nice boolean search term and google up a
list of netbooks with matching components. Then shorten that list to
those with a Phoenix BIOS, shorten again to the same BIOS features
(except Nbus functionality - Gateway tend to skimp in those areas).
Finally I'd take that short list as see which ones have an OEM Linux OS
available. Then pull out the ACPI module and any udev and kernel
customisations.
If nothing else, you'll gain an insight into your netbook - Linux, and a
new respect for manufacturers who conform to published specification ;-p

Please check what I've written - and not just the spelling and grammar
:-D  Apologies for any mistakes - pressed for time and jetlagged.

Cheers

-- 
I love talking about the Kennedy assassination.
The reason I do is because I'm fascinated by it.
I'm fascinated that our government could lie to us so blatantly, so
obviously for so long, and we do absolutely nothing about it.
I think that's interesting in what is ostensibly a democracy.
Sarcasm - come on in.
People say "Bill, quit talking about Kennedy man.
It was a long time ago, just let it go, alright?
It's a long time ago, just forget it."
I'm like, alright, then don't bring up Jesus to me.
As long as we're talking shelf life here...
~ Bill Hicks


Reply to: