RFR: 8252842: Extend jmap to support parallel heap dump [v18]

[Chris Plummer]

On Mon, 22 Mar 2021 12:35:52 GMT, Lin Zang <lzang at openjdk.org> wrote:

>> 8252842: Extend jmap to support parallel heap dump
>
> Lin Zang has updated the pull request incrementally with two additional commits since the last revision:
> 
>  - remove parallel option and dumpheapext command
>  - Revert "hide the dumpheapext error message"
>    
>    This reverts commit 1af0e1e2bfab4f5d079c03ff0cb25067acacdac4.

I've only had a chance to skim over the code, so I just noted a few comment and wording related fixes while doing that. I'll need to set aside some more time for a more in depth review since the changes are pretty extensive.

src/hotspot/share/services/heapDumper.cpp line 556:

> 554:                          (u4) (position() - dump_segment_header_size));
> 555:     } else {
> 556:       // Finished process huge sub record

Is this suppose to be "Finished processing" or maybe "Finish processing".

src/hotspot/share/services/heapDumper.cpp line 557:

> 555:     } else {
> 556:       // Finished process huge sub record
> 557:       // Set _is_huge_sub_record to false so the parallel dump writer could flush data to file.

"can flush"

src/hotspot/share/services/heapDumper.cpp line 1665:

> 1663: // To avoid memory consumption, when dumping large objects such as huge array and
> 1664: // large objects whose size are larger than LARGE_OBJECT_DUMP_THRESHOLD, the scanned
> 1665: // partial object/array data will be send to backend directly instead of caching the

"will be sent to the backend..."

src/hotspot/share/services/heapDumper.cpp line 1666:

> 1664: // large objects whose size are larger than LARGE_OBJECT_DUMP_THRESHOLD, the scanned
> 1665: // partial object/array data will be send to backend directly instead of caching the
> 1666: // whole object/array internal buffer.

Should this be "...in the internal buffer"?

src/hotspot/share/services/heapDumper.cpp line 1668:

> 1666: // whole object/array internal buffer.
> 1667: // The HeapDumpLargeObjectList is used to save the large object when dumper scanning
> 1668: // the heap. The large objects could be added (push) parallelly by multiple dumpers.

Should this be "...when the dumper scans the heap."?

src/hotspot/share/services/heapDumper.cpp line 1884:

> 1882:   // parallel heap dump support
> 1883:   uint                    _num_dumper_threads;
> 1884:   uint                    _num_writter_threads;

Should be "writer", not "writter". This appears in a few fields and comments.

-------------

PR: https://git.openjdk.java.net/jdk/pull/2261