NMI on shutdown

[Hiroki Sato]

Sean Bruno <sean_bruno at yahoo.com> wrote
  in <1374942794.1431.37.camel at localhost>:

se> :-)  I just spent the week looking at something that looks like this on
se> my Dell machines.  In my testing the NMI EISA problem seems to be coming
se> from the fact that IPMI pokes at both the ACPI and ISA interfaces to the
se> IPMI controller resulting in an attempt to create /dev/ipmi0
se> and /dev/ipmi1.
se>
se> Somewhere in the recent past (affects 9 as well) the ACPI and ISA IPMI
se> device nodes where children of the same parent and
se> ipmi_isa.c::ipmi_isa_identitfy() would see the ACPI attachment and do
se> nothing.  Now the two interfaces have different parents in the device
se> tree.
se>
se> On bce(4) based systems, even if not using IPMI, this seems to
se> crash/confuse older version of the management firmware and yield results
se> similar to what you see.
se>
se> I've just commited a tested fix from Yahoo on this that Peter and I
se> worked out.  see svn R 253708

 Hmm, in my case the symptom does not depend on ipmi(4) (ipmi.ko was
 not loaded).  I tried the patch in r253708 on stable/9 in both cases
 of loading the ipmi.ko or not loading it, but there was no
 difference.

 Did NMI happen on you box only when ipmi.ko was loaded?

-- Hiroki
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://lists.freebsd.org/pipermail/freebsd-current/attachments/20130729/9b32715e/attachment.sig>