cvs commit: src/sys/i386/i386 local_apic.c src/sys/amd64/amd64 local_apic.c

[Maxim Sobolev]

John Baldwin wrote:
> On Tuesday 05 September 2006 16:02, Maxim Sobolev wrote:
>> John Baldwin wrote:
>>>>>  (That is, are there any such places.   If so, you
>>>>> just broke them.)
>>>> No, I believe that I did not, unless you can provide example of the 
>>>> contrary.
>>> linprocfs, but it lies anyway.  I've engaged in hacks like this in 4.x,
>> That's what I mean - I can't imagine how can you get any useful 
>> statistics out of CPU times by combining it with number of processors.
>>
>>> but I think they are just that: hacks.  I think a real fix is to support 
>>> turning off CPUs in the MI code and allow userland to query via a 
> non-hackish 
>>> interface how many CPUs are actually enabled and get appropriate load 
> stats, 
>>> etc. based on that.
>> Yes, that's would be nice. But in the meantime my goal is to resolve 
>> obvious regression we have in the 6.x release in the presence of the HTT 
>> CPU.
> 
> It's not a regression I think as 4.x and 5.x both do the same as before this 
> commit (IIRC), but that's ok.

Yes, this problem was introduced by the fix to the famous "HTT 
vulnerability".

-Maxim