JDK submit repo

jesper.wilhelmsson at oracle.com jesper.wilhelmsson at oracle.com
Wed Jan 24 11:11:21 UTC 2018

> On 24 Jan 2018, at 11:25, David Holmes <david.holmes at oracle.com> wrote:
> On 24/01/2018 8:04 PM, dalibor topic wrote:
>> On 23.01.2018 21:33, mark.reinhold at oracle.com wrote:
>>> If you push a branch whose name starts with "JDK-" into this repo
>>> then it will automatically be built and tested on Oracle's internal
>>> build system, and a summary of the results will be e-mailed back to
>>> you.  This typically takes a couple of hours.  If one or more tests
>>> fail then you can contact an Oracle engineer to get more information
>>> about the failures.
>> Since the wiki
>> * doesn't say how to contact that engineer
>> * doesn't say how to work with them
>> that might be a hurdle too steep in some cases.
>> As the only contact given on the wiki is ops at openjdk, I assume it will get more than its fair share of test failures to forward through email chains in search of that engineer. That would be unfortunate.
>> Instead, I would suggest explicitly advising Comitters with test failures to forward them to the jdk-dev mailing list in search of an Oracle engineer.
> I would not advise that as not all engineers hang out on jdk-dev. Rather they should forward to xxx-dev where xxx relates to the component on which they were working: hotspot-dev, hotspot-runtime-dev, core-libs-dev, net-dev etc
>> Or, ideally, the system could simply send those e-mails announcing failures to the Committer with a CC: to jdk-dev, taking out the manual part of Oracle engineer search and forwarding e-mails around out of the process.
> Please no! :)


I don't think we should assume that every single test failure will need Oracle assistance. It's quite likely that developers will send their code through the system more than once before deciding to go ask for help.

> The way I see this working is as follows:
> - contributor works on a fix and gets it working locally on whatever systems they have available to them
> - contributor then wants to test further and so submits to submit repo
> - contributor gets response email:
> a) all good: contributor sends RFR to xxx-dev
> b) not all good: contributor sends prelim-RFR to xxx-dev and includes failure info, and waits for help on the failures
> Cheers,
> David
> ------
>> cheers,
>> dalibor topic

More information about the jdk-dev mailing list