[aarch64-port-dev ] [RFR] 8u152 Upstream Sync

[Andrew Hughes]

On 25 January 2018 at 16:25, Tiago Daitx <tiago.daitx at canonical.com> wrote:
> The tag aarch64-jdk8u152-b16 seems to be missing, although the merge
> is in the tip. The newest tag at
> http://hg.openjdk.java.net/aarch64-port/jdk8u/tags is jdk8u152-b15.
>
> Just a heads up in case this isn't intentional.
>
> Thanks!
>

It's a side-effect of the way Mercurial handles tags. Because tag
creation creates
another changeset, the tag changeset is not part of the tag :/

 $ hg log -R jdk -k jdk8u152-b16
changeset:   12894:df979e82b65c
tag:         aarch64-jdk8u152-b17
user:        andrew
date:        Fri Dec 01 15:58:47 2017 +0000
summary:     Added tag aarch64-jdk8u152-b16 for changeset b1c489be77e2

changeset:   12893:b1c489be77e2
tag:         aarch64-jdk8u152-b16
parent:      12622:fe6a6fe4f4a0
parent:      12892:1cb70967c4d7
user:        andrew
date:        Fri Dec 01 03:30:58 2017 +0000
summary:     Merge jdk8u152-b16

$ hg log -R jdk -r aarch64-jdk8u152-b16
changeset:   12893:b1c489be77e2
tag:         aarch64-jdk8u152-b16
parent:      12622:fe6a6fe4f4a0
parent:      12892:1cb70967c4d7
user:        andrew
date:        Fri Dec 01 03:30:58 2017 +0000
summary:     Merge jdk8u152-b16

When we request to push 'aarch64-jdk8u152-b16', as in this case, what is pushed
is b1c489be77e2 and its predecessors, but not the later changeset df979e82b65c
with the tag addition.

This will be rectified when the next batch [0] of changes is pushed.
Usually I'd push
the tip, which would include the tag, but I can't do this in cases
where there are further
changes on top that we don't want to push yet (u161).

The reason HotSpot did get the tag is because upstream changes occurred in that
repository after the pull I began the u152 merge with. Thus, these had
to be merged
before the changes could be pushed, so the tag changeset ended up being merged
with the changeset from upstream in a further merge changeset, and I pushed the
merge changeset. Mercurial doesn't allow the unmerged aarch64-jdk8u152-b16
head to be pushed in such a situation.

$ hg log -R hotspot -r aarch64-jdk8u152-b16
changeset:   9612:860742065a84
tag:         aarch64-jdk8u152-b16
parent:      9472:c51c309e3c81
parent:      9611:98b4b0661837
user:        andrew
date:        Fri Dec 01 03:30:51 2017 +0000
summary:     Merge jdk8u152-b16

$ hg log -R hotspot -r 9613
changeset:   9613:698f744bb411
user:        andrew
date:        Fri Dec 01 15:58:46 2017 +0000
summary:     Added tag aarch64-jdk8u152-b16 for changeset 860742065a84

$ hg log -R hotspot -r 9614
changeset:   9614:97df73117bbe
parent:      9472:c51c309e3c81
user:        rkennke
date:        Thu Dec 07 17:23:19 2017 +0100
summary:     8193133: Assertion failure because 0xDEADDEAD can be in-heap

$ hg log -R hotspot -r 9615
changeset:   9615:d0281b4e8f2c
tag:         aarch64-jdk8u152-b17
parent:      9613:698f744bb411
parent:      9614:97df73117bbe
user:        andrew
date:        Thu Jan 04 20:19:27 2018 +0000
summary:     Merge

So, although I mention it in [0], the changes for aarch64-jdk8u152-b17
are already
in the HotSpot repository, but not yet tagged as such.

[0] http://mail.openjdk.java.net/pipermail/aarch64-port-dev/2018-January/005304.html

Hope that helps,
-- 
Andrew :)

Senior Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

Web Site: http://fuseyism.com
Twitter: https://twitter.com/gnu_andrew_java
PGP Key: ed25519/0xCFDA0F9B35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222