[11u] RFR(S): 8241234: Unify monitor enter/exit runtime entries.
Lindenmaier, Goetz
goetz.lindenmaier at sap.com
Tue Sep 8 12:01:38 UTC 2020
Hi
> On the wider point:
>
> A while ago I announced that we should not continue to independently
> approve backports that had already been approved by Oracle. My
> reasoning was that they'd done the work of balancing risk and reward,
> and that for us to do it again when the decision would almost
> inevitably be the same was not a good use of our time. After some very
> noisy protests which insisted that this project must continue to
> approve backports independently, I relented. But there is no point
> retaining that right if we're never going to use it!
I would appreciate this a lot, as it would reduce the effort of
downporting considerably. Thanks!
> We should keep a close watch on Oracle backports for a while to try to
> understand what their criteria are, and whether those criteria are a
> good fit for our mission. If we see a significant number of minor
> performance tweaks and "cleanups" we should reconsider our policy.
One possibility to watch what Oracle is doing is that maintainers have
an eye on the according filters of 11.0.x-oracle changes and flag them
jdk11u-fix-no ad-hoc, before someone starts to actually work and
downport the bug.
Others, that also have an opinion on this, could also watch the
filter and then add comments why they recommend not to downport
the change.
Then again, if someone has a concrete interest to downport, besides
the motivation that Oracle downported it, this can be discussed.
In the case of this change, it required a close look at the coding,
which is why I had asked Martin to address it. Martin is an
expert in this area. The objections came just after he had completed
his work. I's really not helpful if a change is turned down just after
all the work is done. This could be avoided by above approach.
Great it is approved.
Best regards,
Goetz.
>
> --
> Andrew Haley (he/him)
> Java Platform Lead Engineer
> Red Hat UK Ltd. <https://www.redhat.com>
> https://keybase.io/andrewhaley
> EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671
More information about the jdk-updates-dev
mailing list