Re: [PATCH] x86: mtrr cleanup for converting continuous to discrete layout v5

[ebiederm@xmission.com (Eric W. Biederman)]

"Yinghai Lu" <yhlu.kernel@gmail.com> writes:

> On Tue, Apr 29, 2008 at 1:29 PM, Eric W. Biederman
> <ebiederm@xmission.com> wrote:
>> "Yinghai Lu" <yhlu.kernel@gmail.com> writes:
>>
>>  > (less memory + fast X) or  (more 8M RAM + slow...)
>>
>>  Yes. That is the basic question.  Not all X drivers need it and
>>  potentially the current kernel drm modules can use the
>>  PAT infrastructure that has been merged.
>>
>>  Further a SMM monitor running 100 times or more slower may cause
>>  problems if SMM mode is entered frequently, slowing down the entire
>>  system not just X.
>>
>>  So if you don't have X or you have a crazy SMM monitor this can
>>  be an issue.
>
> agreed. so that feature is compiled in but disable by default.
>
> BTW: is any chance for OS to disable SMI etc? to verify is the
> unstatbility is caused by SMI?

Not in general no.  Frequently you can get at the registers that
will enable/disable an SMI but that is chipset specific.

Think of SMM mode is a lightweight hypervisor that we can't get rid
of, if you want to understand the worst case.

In theory SMM mode is completely unnecessary as soon as we enable
ACPI.  In practice ACPI appears to frequently trap into SMM mode.

Eric