Re: ACPI battery state and resume not working on Inspiron 5150

> From: Joe Marcus Clarke <marcus_at_marcuscom.com>
> Date: Wed, 06 Aug 2003 16:01:37 -0400
> Sender: owner-freebsd-current_at_freebsd.org
> 
> 
> --=-MHp9eSkqmbnyoWl+2a1w
> Content-Type: text/plain
> Content-Transfer-Encoding: quoted-printable
> 
> On Wed, 2003-08-06 at 15:56, Barney Wolff wrote:
> > On Wed, Aug 06, 2003 at 03:31:01PM -0400, Joe Marcus Clarke wrote:
> > >     ACPI-1287: *** Error: Method execution failed
> > > [\\_SB_.PCI0.LPCB.BAT1._STA] (Node 0xc6137640), AE_NOT_EXIST
> >=20
> > I would not expect BAT1 to exist unless you have 2 batteries installed.
> 
> Ah, good point.  However, I don't see any battery sysctls, and I do have
> at least one battery installed.
> 
> > As to resume, on my I5000 it takes almost a minute to come back from
> > S3, but does eventually come back on a -current from 7/30.
> 
> Thanks for the suggestion.  I'll wait a bit longer.

Well, I did some experimenting yesterday with the ACPI code on my IBM
T30 and learned one thing...if you plan on suspending, you need to set
a sleep delay. Before I set the delay I had some nasty problems
because power went away immediately and the disk cache did not have a
chance to flush (ouch!) and left the display where it should not be.

I noticed that Windows XP has a delay of about 5 seconds. I set the
sysctl and tried again and things went MUCH better. The suspend didn't
leave the disk corrupt (whew!) and the display dropped to low
resolution before the graphics was shut down and switched back on
resume! My Radeon M7 even retained sync.

Of course, the USB driver simply does not recover from a suspend on
ACPI and this should be fixed before too long. Also, the backlight
stays on making the suspend NVU (not very useful). But it is a huge
improvement and adding a delay MAY help a lot of other laptop
suspend/resume areas. 

Whether this will help th I5000 problems, I can't say, but it seems
like suspend/resume is the most common show-stopper for ACPI on
laptops, so it's worth a shot. If there is a trend that indicates that
a short delay in suspending fixes a number of problems, the default
delay should probably be modified from 0 to 4 or 5.
-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman_at_es.net			Phone: +1 510 486-8634