RFR(XS): 8204136: jtreg: Fix failing RTM test RTMSpinLoopCount

[Gustavo Romero]

Hi Goetz,

On 06/03/2018 03:39 AM, Lindenmaier, Goetz wrote:
> Looks good.
> I will sponsor this and others for you.

Thanks for reviewing and sponsoring the changes!

Igor, thanks for reviewing the change!


Best regards,
Gustavo
  
> Best regards,
>    Goetz.
> 
>> -----Original Message-----
>> From: Gustavo Romero <gromero at linux.vnet.ibm.com>
>> Sent: Friday, June 1, 2018 5:22 PM
>> To: Igor Ignatyev <igor.ignatyev at oracle.com>
>> Cc: hotspot-compiler-dev at openjdk.java.net; Doerr, Martin
>> <martin.doerr at sap.com>; Lindenmaier, Goetz
>> <goetz.lindenmaier at sap.com>
>> Subject: Re: RFR(XS): 8204136: jtreg: Fix failing RTM test RTMSpinLoopCount
>>
>> Hi Igor,
>>
>> On 05/31/2018 12:37 AM, Igor Ignatyev wrote:
>>> Hi Gustavo,
>>>
>>> the fix looks reasonable to me.
>>
>> Thanks a lot for reviewing it.
>>
>> I think I still need one additional review for that change?
>>
>>
>> Best regards,
>> Gustavo
>>
>>> Thanks,
>>> -- Igor
>>>
>>>> On May 30, 2018, at 7:05 PM, Gustavo Romero
>> <gromero at linux.vnet.ibm.com> wrote:
>>>>
>>>> Hi,
>>>>
>>>> Could the following change be reviewed please?
>>>>
>>>> webrev: http://cr.openjdk.java.net/~gromero/8204136/v1
>>>> bug   : https://bugs.openjdk.java.net/browse/JDK-8204136
>>>>
>>>> It throttles down RTMSpinLoopCount deltas in TestRTMSpinLoopCount to
>> avoid
>>>> a test timeout. Using current RTMSpinLoopCount values mostly makes
>> the last
>>>> tries fail (RTMSpinLoopCount = 1_000_000 or 10_000_000). The proposed
>>>> increment is to make the next RTMSpinLoopCount value 10x greater than
>> the
>>>> previous one.
>>>>
>>>> Currently RTMSpinLoopCount can fail (error) sometimes due to timeout.
>>>> (Please see:
>> http://cr.openjdk.java.net/~gromero/misc/moog_spincountloop_timeout.lo
>> g)
>>>>
>>>> The JVM behavior seems correct and indeed high RTMSpinLoopCount
>> values
>>>> like 10_000_000 should take a long time to complete (> 2 minutes).
>>>>
>>>> Thank you.
>>>>
>>>>
>>>> Best regards,
>>>> Gustavo
>>>>
>>>
>