Re: [PATCH] x86, apic: Do not increment disabled_cpus from generic_processor_info.

[Cyrill Gorcunov <gorcunov@gmail.com>]

On 01/10/2011 07:06 AM, Rakib Mullick wrote:
> On Mon, Jan 10, 2011 at 12:38 AM, Cyrill Gorcunov <gorcunov@gmail.com> wrote:
>> On 01/09/2011 07:57 PM, Rakib Mullick wrote:
>>> On Sun, Jan 9, 2011 at 3:39 PM, Ingo Molnar <mingo@elte.hu> wrote:
>>
>>  When nr_cpus=n passed from command line and there is N > n physical cpu
>> present we *still* have to increment disabled_cpus in generic_processor_info
>> because:
>>
>> 1) We're priting out the number of cpu which is disabled
>> 2) total_cpus become inconsistent
>>
>> and while (1) is not that important, total_cpus _is_ important (it
>> is used to print out offlined cpus).
>>
> When we use nr_cpus=n, it works as an upper limit. If there are any
> other CPUs beyond that limit those are not counted and we couldn't put
> them back on work. So, when we couldn't use hotpluging feature to back
> them into work, should we care about them?
> 

  Yes we should, the cpus which are present but offlined (due to maxcpus
or whatever reason) are listed in sysfs so i fear such a change would
break compatibility with userspace as well.

  Rakib, don't get me wrong, I don't like to complain but the side effect
of the patch might be pretty inconvenient.

> 
>>  So I still fail to see why we need to drop the former increment in
>> first place.

-- 
    Cyrill