Re: RFR JDK-8151913: Fix module dependencies in java/net tests
Hi, Just restarted this job. On 2016/4/29 15:34, Alan Bateman wrote: On 28/04/2016 05:50, John Jiang wrote: Hi, Please review another webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.02 The java.httpclient module declaration is removed from all of java/net/httpclient tests, even though some ones have to declare other modules. If a test has the @modules tag then then the "modules" key in the test suite configuration (TEST.properties) won't be used. That is, @modules overrides rather than augments. I just re-read the tag spec [1] and I think have this right. In that case, the java.httpclient tests that have @modules jdk.httpserver will need to list java.httpclient too. Please review the updated webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.03 Additionally, test/java/net/CookieHandler/CookieManagerTest.java and test/java/net/HttpURLConnection/UnmodifiableMaps.java have defined @modules, so this new patch excludes them. Best regards, John Jiang
Re: (9) RFR: 8157783: Fix module dependencies for /javax/* and /jdk/* tests
Hi Chris, Thanks for your comment! Please review the updated webrev: http://cr.openjdk.java.net/~jjiang/8157783/webrev.01 It excludes the /javax/* tests. Best regards, John Jiang On 2016/5/25 16:02, Chris Hegarty wrote: On 25 May 2016, at 07:44, John Jiang wrote: Hi, Please review this patch on fixing module dependencies for /javax/* and /jdk/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157783 Webrev: http://cr.openjdk.java.net/~jjiang/8157783/webrev.00/ The changes to the non-javax tests look fine. Just a question on scripts, GetCallerClassTest.sh in this case, but the question is more general. You have added '@modules java.base/jdk.internal.reflect’, but what does this actually do? I can see the value of explicitly declaring that the script requires a particular set of modules, which could be used for test selection depending on the runtime. But the qualified export appears meaningless. Also, since the module is java.base, the dependency is implicit and not required. -Chris.
Re: (9) RFR: 8157783: Fix module dependencies for /javax/* and /jdk/* tests
OK, I'll update issue JDK-8157783 and its webrev to focus on /jdk/* only. Thanks! John Jiang On 2016/5/25 15:07, Felix Yang wrote: Hi John, I think changes for javax/* tests are not needed. Please have a look at the bug below. We are waiting for the next jtreg promotion to push the change, otherwise those tests will be skipped silently. https://bugs.openjdk.java.net/browse/JDK-8157135 -Felix On 2016/5/25 14:44, John Jiang wrote: Hi, Please review this patch on fixing module dependencies for /javax/* and /jdk/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157783 Webrev: http://cr.openjdk.java.net/~jjiang/8157783/webrev.00/ Best regards, John Jiang
(9) RFR: 8157783: Fix module dependencies for /javax/* and /jdk/* tests
Hi, Please review this patch on fixing module dependencies for /javax/* and /jdk/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157783 Webrev: http://cr.openjdk.java.net/~jjiang/8157783/webrev.00/ Best regards, John Jiang
Re: (9) RFR: 8157635: Fix module dependencies for /sun/* tests
Hi Alan, On 2016/5/24 14:59, Alan Bateman wrote: On 24/05/2016 07:35, John Jiang wrote: Hi, Please review this patch on fixing module dependencies for /sun/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157635 Webrev: http://cr.openjdk.java.net/~jjiang/8157635/webrev.00/ Is dns/cname.sh really using module java.naming now? Just asking because I thought the name service provider mechanism has been removed so InetAddress can no longer be configured to use the JNDI DNS provider. CanonicalName.java, which is a part of the test, is using java.naming APIs. In sun/security/provider/PolicyFile/Modules.java then it would be good to sort the module names as it will be others hard to maintain this list. Please review the updated webrev [1]. [1] http://cr.openjdk.java.net/~jjiang/8157635/webrev.01 Best regards, John Jiang
Re: (9) RFR: 8157633: Fix module dependencies for /com/* tests
Hi Alan, Thanks for your review. And I have to update the copyright notice year [1]. [1] http://cr.openjdk.java.net/~jjiang/8157633/webrev.01 Best regards, John Jiang On 2016/5/24 14:49, Alan Bateman wrote: On 24/05/2016 04:34, John Jiang wrote: Hi, Please review this small patch on fixing module dependencies for /com/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157633 Webrev: http://cr.openjdk.java.net/~jjiang/8157633/webrev.00/ Looks okay to me. -Alan
(9) RFR: 8157635: Fix module dependencies for /sun/* tests
Hi, Please review this patch on fixing module dependencies for /sun/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157635 Webrev: http://cr.openjdk.java.net/~jjiang/8157635/webrev.00/ Best regards, John Jiang
(9) RFR: 8157633: Fix module dependencies for /com/* tests
Hi, Please review this small patch on fixing module dependencies for /com/* tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8157633 Webrev: http://cr.openjdk.java.net/~jjiang/8157633/webrev.00/ Best regards, John Jiang
Re: RFR JDK-8151913: Fix module dependencies in java/net tests
Hi Amy, That's case to case. If a test is using java.logging APIs directly, I declared the module for the test. Otherwise, I didn't. I think that may be more clear. Although a test is using jdk.httpserver, that doesn't mean it also dependents on java.logging. Best regards, John Jiang On 2016/4/28 13:09, Amy Lu wrote: On 4/28/16 12:50 PM, John Jiang wrote: Hi, Please review another webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.02 The java.httpclient module declaration is removed from all of java/net/httpclient tests, even though some ones have to declare other modules. + * @modules jdk.httpserver + * java.logging Not necessary to declare java.logging as it’s a dependency of jdk.httpserver (it does not hurt, though) Please wait for reviewer's feedback... Thanks, Amy Best regards, John Jiang On 2016/4/27 23:07, John Jiang wrote: Hi Alan, Felix, Thanks for your comments. Please review the updated webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.01/ Best regards, John Jiang On 2016/4/27 15:08, John Jiang wrote: Hi, Please review the fix for explicitly declaring module dependencies for java net tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8151913 Webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.00 Best regards, John Jiang
Re: RFR JDK-8151913: Fix module dependencies in java/net tests
Hi, Please review another webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.02 The java.httpclient module declaration is removed from all of java/net/httpclient tests, even though some ones have to declare other modules. Best regards, John Jiang On 2016/4/27 23:07, John Jiang wrote: Hi Alan, Felix, Thanks for your comments. Please review the updated webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.01/ Best regards, John Jiang On 2016/4/27 15:08, John Jiang wrote: Hi, Please review the fix for explicitly declaring module dependencies for java net tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8151913 Webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.00 Best regards, John Jiang
Re: RFR JDK-8151913: Fix module dependencies in java/net tests
Hi Alan, Felix, Thanks for your comments. Please review the updated webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.01/ Best regards, John Jiang On 2016/4/27 15:08, John Jiang wrote: Hi, Please review the fix for explicitly declaring module dependencies for java net tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8151913 Webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.00 Best regards, John Jiang
RFR JDK-8151913: Fix module dependencies in java/net tests
Hi, Please review the fix for explicitly declaring module dependencies for java net tests. Issue: https://bugs.openjdk.java.net/browse/JDK-8151913 Webrev: http://cr.openjdk.java.net/~jjiang/8151913/webrev.00 Best regards, John Jiang
RFR: 8130054: javax/naming/module/basic.sh needs to be headless
Hi, Please review the fix for JDK-8130054. The test is using java.awt.event.ActionEvent instead of java.awt.Button. Issue: https://bugs.openjdk.java.net/browse/JDK-8130054 Webrev: http://cr.openjdk.java.net/~jjiang/8130054/webrev.00/ Best regards, John Jiang