RFR: JDK-8189664: Stop producing ARCHIVE_BUNDLE.zip from test makefiles unless asked to

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

RFR: JDK-8189664: Stop producing ARCHIVE_BUNDLE.zip from test makefiles unless asked to

Erik Joelsson
In test/TestCommon.gmk the jtreg invocation will unconditionally produce
an archive bundle, a zip file consisting of all the output from jtreg.
When testing in JPRT, JPRT sets JPRT_ARCHIVE_BUNDLE to override the name
of this zip. In all other situation, I can't see the point of creating
it. It just adds unnecessary test execution time as well as bloating up
the test results dir when we now save it using different means.

I propose to stop producing any zip file unless either ARCHIVE_BUNDLE or
JPRT_ARCHIVE_BUNDLE is set by the user.

Webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.01/

Bug: https://bugs.openjdk.java.net/browse/JDK-8189664

/Erik

Reply | Threaded
Open this post in threaded view
|

Re: RFR: JDK-8189664: Stop producing ARCHIVE_BUNDLE.zip from test makefiles unless asked to

David Holmes
Looks okay.

Thanks,
David

On 19/10/2017 10:32 PM, Erik Joelsson wrote:

> In test/TestCommon.gmk the jtreg invocation will unconditionally produce
> an archive bundle, a zip file consisting of all the output from jtreg.
> When testing in JPRT, JPRT sets JPRT_ARCHIVE_BUNDLE to override the name
> of this zip. In all other situation, I can't see the point of creating
> it. It just adds unnecessary test execution time as well as bloating up
> the test results dir when we now save it using different means.
>
> I propose to stop producing any zip file unless either ARCHIVE_BUNDLE or
> JPRT_ARCHIVE_BUNDLE is set by the user.
>
> Webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.01/
>
> Bug: https://bugs.openjdk.java.net/browse/JDK-8189664
>
> /Erik
>
Reply | Threaded
Open this post in threaded view
|

Re: RFR: JDK-8189664: Stop producing ARCHIVE_BUNDLE.zip from test makefiles unless asked to

Erik Joelsson
In reply to this post by Erik Joelsson
This didn't quite work out. The clean and prep recipes broke down when
ARCHIVE_BUNDLE was empty.

New webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.02/

/Erik


On 2017-10-19 14:32, Erik Joelsson wrote:

> In test/TestCommon.gmk the jtreg invocation will unconditionally
> produce an archive bundle, a zip file consisting of all the output
> from jtreg. When testing in JPRT, JPRT sets JPRT_ARCHIVE_BUNDLE to
> override the name of this zip. In all other situation, I can't see the
> point of creating it. It just adds unnecessary test execution time as
> well as bloating up the test results dir when we now save it using
> different means.
>
> I propose to stop producing any zip file unless either ARCHIVE_BUNDLE
> or JPRT_ARCHIVE_BUNDLE is set by the user.
>
> Webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.01/
>
> Bug: https://bugs.openjdk.java.net/browse/JDK-8189664
>
> /Erik
>

Reply | Threaded
Open this post in threaded view
|

Re: RFR: JDK-8189664: Stop producing ARCHIVE_BUNDLE.zip from test makefiles unless asked to

Magnus Ihse Bursie
On 2017-10-20 15:30, Erik Joelsson wrote:
> This didn't quite work out. The clean and prep recipes broke down when
> ARCHIVE_BUNDLE was empty.
>
> New webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.02/
Looks good to me.

/Magnus

>
> /Erik
>
>
> On 2017-10-19 14:32, Erik Joelsson wrote:
>> In test/TestCommon.gmk the jtreg invocation will unconditionally
>> produce an archive bundle, a zip file consisting of all the output
>> from jtreg. When testing in JPRT, JPRT sets JPRT_ARCHIVE_BUNDLE to
>> override the name of this zip. In all other situation, I can't see
>> the point of creating it. It just adds unnecessary test execution
>> time as well as bloating up the test results dir when we now save it
>> using different means.
>>
>> I propose to stop producing any zip file unless either ARCHIVE_BUNDLE
>> or JPRT_ARCHIVE_BUNDLE is set by the user.
>>
>> Webrev: http://cr.openjdk.java.net/~erikj/8189664/webrev.01/
>>
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8189664
>>
>> /Erik
>>
>