RFR: 8222001: JFR event for heap dumps written

[Erik Gahlin]

Thanks for the reviews.

I will keep events.size(), but remove the Files.delete before pushing.

Erik

On 2020-01-21 22:45, mikhailo.seledtsov at oracle.com wrote:
>
> Overall looks good; several of comments though:
>
>   - TestHeapDump.java: lines 62-64:
>
>    62             if (events.size() != 1) {
>    63                 throw new Exception("Expected one event, got " + events.size());
>    64             }
>
>    You could use jdk.test.lib.Asserts.assertNotEquals(events.size(), 1, message)
>    A matter of style, up to you
>
>   - TestHeapDump.java:72
>     Files.delete(path);
>     As far as I understand the path is inside a "scratch" test directory. If this is correct, then there is no need to remove the files explicitly; the JTreg framework will take care of them.
>
>   - TestHeapDump.java: new file, Copyright year should be 2020
>
>   - also, please update copyright years for heapDumper.cpp and EventNames.java
>
>
> Thank you,
> Misha
> On 1/16/20 12:49 PM, Erik Gahlin wrote:
>> |Hi,
>>
>> Could I have a review of a change that writes an event when a heap 
>> dump has been written.|
>>
>> |The tests checks that it works by invoking the dump over JMX, but I 
>> have also verified manually that it works on command line 
>> (-XX:+HeapDumpOnOutOfMemoryError) and using jcmd (GC.heap_dump).
>>
>> Bug:
>> https://bugs.openjdk.java.net/browse/JDK-8222001
>>
>> Webrev:
>> http://cr.openjdk.java.net/~egahlin/8222001/
>>
>> Testing:
>> tier1 + tier2, jdk/jdk/jfr
>>
>> Thanks
>> Erik
>> |
>>