[12]RFR: 8211034: OnStackReplacePercentage option checking has bugs
Vladimir Kozlov
vladimir.kozlov at oracle.com
Thu Dec 13 01:27:30 UTC 2018
Looks good.
Thanks,
Vladimir
On 12/11/18 4:14 PM, Jamsheed wrote:
> Thank you for the review,Vladimir. please find the updated webrev
>
> http://cr.openjdk.java.net/~jcm/8211034/webrev.05/
>
> Best regards,
>
> Jamsheed
>
> On 29/11/18 11:56 PM, Vladimir Kozlov wrote:
>> I think you need to add downcast (int) to expressions you changed when the destination has int type. You may get C++
>> warning ("loose precision" or something). Please test it.
>>
>> Otherwise looks good.
>>
>> Thanks,
>> Vladimir
>>
>> On 11/28/18 7:01 PM, Jamsheed C M wrote:
>>> Hi Vladimir,
>>> Thank you for the feedback.
>>> Revised webrev: http://cr.openjdk.java.net/~jcm/8211034/webrev.04/
>>> Best regards,
>>> Jamsheed
>>>
>>> ----- Original Message -----
>>> From: vladimir.kozlov at oracle.com
>>> To: jamsheed.c.m at oracle.com, hotspot-compiler-dev at openjdk.java.net
>>> Sent: Thursday, 22 November, 2018 3:23:35 AM GMT +05:30 Chennai, Kolkata, Mumbai, New Delhi
>>> Subject: Re: [12]RFR: 8211034: OnStackReplacePercentage option checking has bugs
>>>
>>> It is not 32-bit specific code.
>>>
>>> http://hg.openjdk.java.net/jdk/jdk/file/74cf02d5f6e2/src/hotspot/share/oops/methodCounters.hpp#l100
>>>
>>> CompileThreshold is used only when TierCompilation is off but it is not 32-bit VM.
>>>
>>> Vladimir
>>>
>>> On 11/21/18 1:04 PM, jamsheed wrote:
>>>> Hi Vladimir,
>>>>
>>>> found some issues with the code.
>>>>
>>>> 1) ProfileInterpreter on, doesn't use count_shift for limit computation.
>>>>
>>>> 2) 32 bit vm, lower compile threshold value will overflow.
>>>>
>>>> fixed both of them,
>>>>
>>>> revised webrev: http://cr.openjdk.java.net/~jcm/8211034/webrev.03/
>>>>
>>>> Best regards,
>>>>
>>>> Jamsheed
>>>>
>>>>
>>>> On Wednesday 21 November 2018 10:20 PM, Vladimir Kozlov wrote:
>>>>> Good.
>>>>>
>>>>> Thanks,
>>>>> Vladimir
>>>>>
>>>>> On 11/21/18 6:59 AM, jamsheed wrote:
>>>>>> Hi Vladimir,
>>>>>>
>>>>>> agree. revised webrev: http://cr.openjdk.java.net/~jcm/8211034/webrev.01/
>>>>>>
>>>>>> Best regards,
>>>>>>
>>>>>> Jamsheed
>>>>>>
>>>>>>
>>>>>> On Wednesday 21 November 2018 01:24 AM, Vladimir Kozlov wrote:
>>>>>>> Hi Jamsheed,
>>>>>>>
>>>>>>> I think messages for over limit cases should be changed to use max_percentage_limit.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Vladimir
>>>>>>>
>>>>>>> On 11/20/18 10:50 AM, jamsheed wrote:
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> Request for review the changes made in OnStackReplacePercentageConstraintFunc for overflow handling.
>>>>>>>>
>>>>>>>> webrev: http://cr.openjdk.java.net/~jcm/8211034/webrev.00/
>>>>>>>>
>>>>>>>> jbs:https://bugs.openjdk.java.net/browse/JDK-8211034
>>>>>>>>
>>>>>>>> Best regards,
>>>>>>>>
>>>>>>>> Jamsheed
>>>>>>>>
>>>>>>
>>>>
More information about the hotspot-compiler-dev
mailing list