Worked in RELENG_5, fails in RELENG_6

[David Kelly]

On Tue, Nov 29, 2005 at 02:52:34PM -0500, John Baldwin wrote:
> On Monday 28 November 2005 10:46 pm, David Kelly wrote:
> >
> > Downloaded and FLASHed BIOS from A06 to Dell's latest, A09 this
> > evening. No improvement. Also built and installed a new kernel from
> > cvsup this afternoon. No improvement.
> 
> But 5.4 works fine with ACPI enabled?  This is the machine that even
> acpidump chokes on, yes?  Does acpidump work ok on 5.4?

"acpidump -t -d" stops after 31 lines and says, "acpidump: FACS is
corrupt" on stderr. See 
http://home.hiwaay.net/~dkelly/opus-acpidump-d-t.txt

"acpidump -d" runs without error and emits 2754 lines. See
http://home.hiwaay.net/~dkelly/opus-acpidump-d.txt

ACPI worked with 5.4 so I never had a reason to acpidump then. Could I
boot a 5.4 CDROM to test ACPI? Without much effort have put my hands on
a "5.3-RELEASE Live Filesystem" CDROM that should make a good test.

There was an issue briefly with ACPI in 5.0 or 5.1, very likely in a
prerelease.

Absolultely certain ACPI worked prior to 6.0 because among other things
ACPI installed a power button handler which performed a clean shutdown
(same as ctl-alt-del) rather than brute force yank-the-plug crash.

> > Oh, and apparently the dmesg buffer isn't big enough to hold
> > everything "boot -v" wants to say. Surely there is a way to make it
> > bigger?
> 
> Well, using a serial console is the best method, but there is also a
> kernel option for increasing the message buffer size documented in
> sys/conf/NOTES.

I can work the problem both ways tonight when I can put my hands on the
physical machine.

-- 
David Kelly N4HHE, dkelly at HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.