RFR: 8223147: JFR Backport (to jdk8u)

[Mario Torre]

On Mon, May 13, 2019 at 7:04 PM Andrey Petushkov <andrey at azul.com> wrote:
>
> Hi Mario,
>
> Apparently the situation is not that bad, just a few things went wrong on our sides.
> My fault I did not switch --enable-jfr to false. Fixed now
> And you seem did not make clean after changing configure options :) With proper clean build the only wrong behavior I get from
> your list is the presence of jfc/template files in the distribution. I've fixed that as well. I don't see any other JFR artefacts in the images
> directory when building with JFR disabled (there are few intermediate files generated but these do not get into final image)
> Your test passes
> The webrevs are updated, please could you check again

Hi Andrey,

Thanks again for the patch. I did perform a bunch of tests and
although there is still work necessary, I think this part of the patch
is good to go, so please do commit it.

As for the next steps, I'm going to see each of the patches that both
Azul and Alibaba submitted for review, I think we should create
subsequent umbrella bugs as we did with this first step so they are
easier to track.

There is a number of fixes that is necessary once all of this has been
merged, for instance when running jtreg tests for the JFR directory I
noticed that many tests (about 30 at this first count) don't seem to
complete properly (one such example TestBiasedLockRevocationEvents).

I didn't yet look at the specific of each test that fails but I think
we should be able to execute every test (or exclude the ones that fail
for obvious reason if necessary).

Also, there's a number of differences between the jdk11 and onward
metadata.xml with what's in this backport patch. Denghui already
mentioned some unsupported events, but also the event definition in
some case is a mismatch. I'm going through each one of those and see
where this is part of the natural evolution of the events and what
should be instead changed to reflect the final definition.

Nevertheless, I think it makes sense to start from here rather than
wait more to have one perfect patch, this will also give us a little
bit more exposure for doing performance test and the likes.

So, bottom line, thanks for your patch, please go ahead and push to
the incubator repository!

Cheers,
Mario

-- 
Mario Torre
Associate Manager, Software Engineering
Red Hat GmbH <https://www.redhat.com>
9704 A60C B4BE A8B8 0F30  9205 5D7E 4952 3F65 7898