Jürgen Keil wrote:
Both ac[] and al[] don't look this interesting to me...
Yep, there are no threshold temperatures defined,
so that tzp->current_level will never change and
always remains at -1 (so the code will never call
tzmon_set_power() to enable / disable any fans).
On Lenovo T61:
ac = [ 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff ]
al = [
{
Length = 0
Pointer = 0
}
Not having temperatures defined in the "ac" array could be ok,
when the BIOS is able to control the fans without help from
the operating system.
By the way, a BIOS that *depends* on the OS to control fans is a
system that *will* eventually overheat (such as just being stopped
in the kernel debugger). So far, I've heard of only one system that
truly depended on the OS to control fans - an obsolete Compaq notebook.
What I would first suspect in the case of overheating is a heatsink
clog (usually dust). I'll use a 'dust-off' spray can periodically to blow
out the dust from the CPU fan/heatsink on my notebooks.
Is your T61 overheating, too?
and the following ACPI errors in the log:
http://defect.opensolaris.org/bz/show_bug.cgi?id=6964
http://bugs.opensolaris.org/view_bug.do?bug_id=6820829
These look bad. Is this from a debug kernel?
Or from a standard release kernel?
I'd like to see the ACPI tables for this machine, to understand why the
subsystem doesn't think the EC handler is registered. (output of iasl -g).
Dana
_______________________________________________
opensolaris-discuss mailing list
[email protected]
