RFR(xs): 8149184: os::is_server_class_machine() could return incorrect result if a host's cpu have a few logical cores

[David Holmes]

On 9/02/2016 9:40 AM, Kirill Zhaldybin wrote:
> David,
>
> On 02/09/2016 01:41 AM, David Holmes wrote:
>> Hi Kirill,
>>
>> On 9/02/2016 3:09 AM, Kirill Zhaldybin wrote:
>>> Dear all,
>>>
>>> Could you please review a small fix for JDK-8149184?
>>>
>>> It fixes server class machine discovery in case if a host has 2 physical
>>> cpu with 2 threads for core each.
>>>
>>> CR: https://bugs.openjdk.java.net/browse/JDK-8149184
>>>
>>> WebRev:
>>> http://cr.openjdk.java.net/~kzhaldyb/webrevs/JDK-8149184/webrev.00/
>>
>> Fix seems logically correct, but as I noted in the bug report this
>> seems to be dead code at the moment.
> I am looking into JDK-8149185 now. If I find a way to fix it
> os::is_server_class_machine will work as expected.

Yes - though it is really a minor issue and I suspect difficult to solve 
because of the use of the stub generator.

> Anyway since patch for JDK-8149184 is pretty simple I think it is worth
> to fix it even it is not used right now.

Sure - it is good experience. :) You will need a sponsor for this change 
- I can do that for you.

Thanks,
David


> Thank you.
> Regards, Kirill
>>
>>
>> Thanks,
>> David
>>
>>> Thank you.
>>>
>>> Regards, Kirill
>