On Sat, Jan 09, 2010 at 09:34:58PM -0500, Alex Goncharov wrote: > It may help one addressing the problem in NetBSD to see the FreeBSD > behavior and an appropriate resolution. > > I just filed a detailed report: > > http://www.freebsd.org/cgi/query-pr.cgi?pr=142561 > > with both the data in unmangled form (the size of it being modest) and > a certain code patch, which may prove helpful to cover this situation > in NetBSD.
Well, with that PR the problem is finally clear. I don't see why the patch has any effect. What is happening is that one of the methods called during the init of the ACPI EC accesses an undefined variable. Due to that the call ultimately fails and the EC is not attached. I don't know why that changes with the patch in FreeBSD land though. Joerg
