RFR(XXS) 8249624: update appcds/sharedStrings/IncompatibleOptions test in view of 8081416 closed as WNF

[Yumin Qi]

Thanks David!

Yumin

On 7/23/20 10:13 PM, David Holmes wrote:
> Looks good!
>
> Thanks,
> David
>
> On 24/07/2020 3:01 pm, Yumin Qi wrote:
>> Hi, David
>>
>>    Sorry I misunderstood the sentence as "without xxxx and run 
>> without (xxxx)".
>>
>>    Updated at new link: 
>> http://cr.openjdk.java.net/~minqi/2020/8249624/webrev-02/
>>
>>    Also re-group similar option tests. Passed local jtreg.
>>
>> Thanks
>>
>> Yumin
>>
>>
>> On 7/23/20 4:31 PM, David Holmes wrote:
>>> Hi Yumin,
>>>
>>> On 24/07/2020 3:10 am, Yumin Qi wrote:
>>>> HI, David
>>>>
>>>>    Thanks for the review. Updated on new link with your suggestion:
>>>>
>>>> http://cr.openjdk.java.net/~minqi/2020/8249624/webrev-01/
>>>
>>>  142         // Explicitly archive without compressed oops and run 
>>> with.
>>>  143         testDump(13, "-XX:+UseG1GC", "-XX:-UseCompressedOops", 
>>> null, false);
>>>  144         testExec(13, "-XX:+UseG1GC", "-XX:-UseCompressedOops", 
>>> null, false);
>>>
>>> That should be using +UseCompressedOops for testExec and expect 
>>> failure.
>>>
>>>  145         // Implicitly archive without compressed oops and run 
>>> with.
>>>  146         testDump(14, "-XX:+UseG1GC", "-Xmx32G", null, false);
>>>  147         testExec(14, "-XX:+UseG1GC", "-Xmx32G", null, false);
>>>
>>> That should be using -Xmx1G for testExec and expect failure.
>>>
>>> David
>>> -----
>>>
>>>>
>>>> Thanks
>>>>
>>>> Yumin
>>>>
>>>>
>>>> On 7/22/20 9:30 PM, David Holmes wrote:
>>>>> Hi Yumin,
>>>>>
>>>>> Given we have the earlier test:
>>>>>
>>>>>  112         // ======= archive with compressed oops, run w/o
>>>>>
>>>>> it would seem better if we had:
>>>>>
>>>>>  112         // Explicitly archive with compressed oops, run without.
>>>>>
>>>>> and:
>>>>>
>>>>>  127         // Implicitly archive with compressed oops, run without.
>>>>>  128         // Max heap size for compressed oops is around 31G.
>>>>>  129         // UseCompressedOops is turned on by default when the 
>>>>> heap
>>>>>  130         // size is under 31G, but will be turned off when the 
>>>>> heap
>>>>>  131         // size is greater than that.
>>>>>
>>>>> And should we also have the opposite test:
>>>>>
>>>>> // Explicitly archive without compressed oops and run with.
>>>>> // Implicitly archive without compressed oops and run with.
>>>>>
>>>>> Thanks,
>>>>> David
>>>>> -----
>>>>>
>>>>> On 23/07/2020 8:55 am, Yumin Qi wrote:
>>>>>> Hi Ioi,
>>>>>>
>>>>>>    I have updated the words as your suggestion, also more 
>>>>>> precisely for the max heap size for compressed oop is around 31G, 
>>>>>> which is calculated by max_heap_for_compressed_oops().
>>>>>>
>>>>>>    updated on same webrev.
>>>>>>
>>>>>> $J6/bin/java -Xshare:on 
>>>>>> -XX:SharedArchiveFile=$J6/lib/server/classes.jsa -Xmx31G -version
>>>>>> java version "16-internal" 2021-03-16
>>>>>> Java(TM) SE Runtime Environment (slowdebug build 
>>>>>> 16-internal+0-adhoc.minqi.open)
>>>>>> Java HotSpot(TM) 64-Bit Server VM (slowdebug build 
>>>>>> 16-internal+0-adhoc.minqi.open, mixed mode, sharing)
>>>>>>
>>>>>> $J6/bin/java -Xshare:on 
>>>>>> -XX:SharedArchiveFile=$J6/lib/server/classes.jsa -Xmx32G -version
>>>>>> An error has occurred while processing the shared archive file.
>>>>>> Unable to use shared archive.
>>>>>> The saved state of UseCompressedOops and 
>>>>>> UseCompressedClassPointers is different from runtime, CDS will be 
>>>>>> disabled.
>>>>>> Error occurred during initialization of VM
>>>>>> Unable to use shared archive.
>>>>>>
>>>>>>
>>>>>> Thanks
>>>>>>
>>>>>> Yumin
>>>>>>
>>>>>> On 7/22/20 2:06 PM, Ioi Lam wrote:
>>>>>>> Hi Yumin,
>>>>>>>
>>>>>>> Just small nits on the comments:
>>>>>>>
>>>>>>> // UseCompressedOops default is turned on when heap is under 32G 
>>>>>>> but will be
>>>>>>>
>>>>>>> -> UseCompressedOops is turned on by default ....
>>>>>>>
>>>>>>> // turned off when heap is greater than 32G. This leads 
>>>>>>> inconsistency
>>>>>>>
>>>>>>> -> This leads to inconsistency ...
>>>>>>>
>>>>>>> // of UseCompressedOops at dump time and runtime.
>>>>>>>
>>>>>>>
>>>>>>> Thanks
>>>>>>> - Ioi
>>>>>>>
>>>>>>> On 7/22/20 1:47 PM, Yumin Qi wrote:
>>>>>>>> Hi, Please review this tiny change on comment:
>>>>>>>>
>>>>>>>> bug: https://bugs.openjdk.java.net/browse/JDK-8249624
>>>>>>>>
>>>>>>>> webrev: http://cr.openjdk.java.net/~minqi/2020/8249624/webrev/
>>>>>>>>
>>>>>>>>
>>>>>>>> Note 8081416 already marked as fixed (thanks Ioi), please read 
>>>>>>>> the comment on https://bugs.openjdk.java.net/browse/JDK-8081416
>>>>>>>>
>>>>>>>> With CDS can be done with UseCompressedOops disabled, the test 
>>>>>>>> already has correct result.
>>>>>>>>
>>>>>>>>
>>>>>>>> Thanks
>>>>>>>>
>>>>>>>> Yumin
>>>>>>>>
>>>>>>>>
>>>>>>>