svn commit: r259641 - in head/sys/amd64: amd64 include vmm/intel

[Benjamin Kaduk]

On Fri, 20 Dec 2013, Neel Natu wrote:

> Hi Ben,
>
> On Fri, Dec 20, 2013 at 10:27 AM, Benjamin Kaduk <bjkfbsd at gmail.com> wrote:
>> On Thu, Dec 19, 2013 at 9:50 PM, Neel Natu <neel at freebsd.org> wrote:
>>>
>>> Author: neel
>>> Date: Fri Dec 20 05:50:22 2013
>>> New Revision: 259641
>>> URL: http://svnweb.freebsd.org/changeset/base/259641
>>>
>>> Log:
>>>   Re-arrange bits in the amd64/pmap 'pm_flags' field.
>>>
>>>   The least significant 8 bits of 'pm_flags' are now used for the IPI
>>> vector
>>>   to use for nested page table TLB shootdown.
>>>
>>>   Previously we used IPI_AST to interrupt the host cpu which is
>>> functionally
>>>   correct but could lead to misleading interrupt counts for AST handler.
>>> The
>>>   AST handler was also doing a lot more than what is required for the
>>> nested
>>>   page table TLB shootdown (EOI and IRET).
>>
>>
>> This represents a KBI change, does it not?  Should __FreeBSD_version be
>> bumped?
>>
>
> I was under the impression that a bump would be required only if this
> change is MFCed to a stable branch. I could be wrong about this and if
> so would be happy to rectify it.

I guess KBI is probably okay, that's true.  I have had a lot of headaches 
with KPI changes on -current and third-party modules, so I am a bit 
sensitive.  I'm also recovering from a laptop theft while on vacation, and 
thus a bit discombobulated.

Thanks,

Ben