[ibm-acpi-devel] [ltp] /proc/acpi/ibm/bay gone with ThinkPad ACPI Extras v0.19-20080107/2.6.24

[Henrique de Moraes Holschuh]

On Thu, 31 Jan 2008, Henrique de Moraes Holschuh wrote:
> > I had a similar problem with 2.6.23 and thinkpad_acpi before I loaded 
> > thinkpad_acpi before most other modules (including libata). Is it possible 
> > that libata, which now has been extended to make use of ACPI, causes this?
> 
> Sounds very strange, it shouldn't.

I stand corrected.  I guess that's it.  We may have finally hit a case where
the linux ACPI model of only one "notify handler"/driver per node fails, and
fails so bad it is not funny.

But better to be sure about it first.  Here's how to test it for people with
the new libata ACPI support:

Please boot the kernel with "libata.noacpi=1".  If either bay or
thinkpad-acpi bay handling starts working *for disks insterted in the
ultrabay, NOT for batteries*, then we have pinpointed the problem.

For batteries, ACPI generic bay should be working already, since they use a
different node.  Thinkpad-acpi won't, because it supports the batteries just
by accident(!).  Indeed there IS a reason why I deprecated that code, and
would have to rewrite it from scratch if it were to survive :-)

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh