Hang on boot with 4.9-STABLE

[Peter Radcliffe]

Doug White <dwhite at gumbysoft.com> probably said:
> > The reason I'm not using ACPI is that is previously wasn't available
> > with 4.x and I have no need for it.
> I believe it uses the improved interrupt and CPU tables, but I might be
> confusing it with -current.

First I've heard of any benifit to using ACPI on a motherboard that
does APM, particularly with how broken a lot of ACPI code (in FreeBSD
and on laptops and other machines) has been and ACPI in 4.x being
labelled experimental. None of it has sounded like something I wanted
to mess with on production machines.

> > I can only test new kernels when I'm around to go power cycle them
> > when they hang (until I can put the remote controlled power strip in
> > that rack to do it remotely), so I can test ACPI but not today.
> That doesn't help much :(

These are production machines, I can't just pull them out of service
whenever I feel like it. None of the other machines I have that are
running 4.9-R or 4.9-STABLE are exibiting the same problem.

> Intel is Intel. ServerWorks is owned by Broadcom.  They are separate
> companies.

Intel sell motherboards they label as ServerWorks;
  http://www.intel.com/design/servers/buildingblocks/srsh4_spsh4/index.htm

I was told that this motherboard was a serverworks motherboard in part
because it had the serial console features I needed. I may have been
misinformed - I don't keep track of most PC things.

This is the motherboard I have;

  http://www.supermicro.com/PRODUCT/MotherBoards/E7501/X5DMS-6GM.htm

> OK, so Intel E7500 chipset.  I've run -stable and -current on that
> in the past without issue, so it could be recent breakage.

I've been running 4.x on several of these motherboards for months with
no problems, until 4.9.

P.

-- 
pir