Is hw.pci.do_powerstate expected to be casually tuned?

[Hiroharu Tamaru]

I think I spoke too early...

OK. I retried RELENG_6 kernel and learned that

$FreeBSD: src/sys/dev/pci/pci.c,v 1.298 2005/09/21 19:47:00 imp Exp $

and a subsequent MFC

$FreeBSD: src/sys/dev/pci/pci.c,v 1.292.2.3 2005/09/27 05:57:47 imp Exp $

has split hw.pci.do_powerstate=1 into
hw.pci.do_power_resume=1
hw.pci.do_power_nodriver=0

FIVA 206VL needs hw.pci.do_power_resume=0, while do_power_nodriver can
be > 0.

So I'm now seeing that these are in fact expected to be tuned by an
user depending on their hardware. ;-)

Maybe, a better question would then be...
Would it make sense to force do_power_resume = 0 when doing a S4BIOS?
Is the current behavior tested on any other S4BIOS capable machine?

Thanks, always.
-- 
Hiroharu Tamaru