On Sat, Dec 17, 2011 at 7:54 AM, Joshua C. <joshuacov(a)googlemail.com> wrote:
I saw in the change log for the kernel-3.1.5-7 that the Intel IOMMU
has been disabled by default. Can someone explain this decision or
point me to the right discussion on the internet?
Dave or Matthew can answer more fully than I can, but there is an
on-going patch series to reenable Sandybridge GPU power savings and
DMAR conflicts with it some how. I don't recall the details, but you
can find them in the discussions between Matthew and Keith Packard on
LKML.
Defaulting this to on has caused a number of odd issues in the past,
and it is mostly used by people using VT-d to assign physical
resources directly to virtual guests. Those doing that can still boot
with intel_iommu=on.
josh