From Randy.Crihfield at Oracle.com Mon Dec 18 16:55:52 2017 From: Randy.Crihfield at Oracle.com (Randy Crihfield) Date: Mon, 18 Dec 2017 11:55:52 -0500 Subject: RFR: 8192837 Need new test for release file info Message-ID: <5A37F318.5020207@Oracle.com> I have created an OpenJDK negative test that confirms the closed source files are not included in the SOURCE. Version of the actual test for review: http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ Any comments/suggestions are welcome, also I will need a sponsor for it at the end? Randy From martinrb at google.com Mon Dec 18 19:40:48 2017 From: martinrb at google.com (Martin Buchholz) Date: Mon, 18 Dec 2017 11:40:48 -0800 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: <5A37F318.5020207@Oracle.com> References: <5A37F318.5020207@Oracle.com> Message-ID: Hi Randy, Your bugid link in the webrev is broken - points to bogus https://bugs.openjdk.java.net/browse/https://bugs.openjdk.java.net/browse/JDK-8192837 Please fix. OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 is not public - please fix. 92 String sJdkPath = System.getProperty("test.jdk"); 93 String sRuntime = System.getProperty("java.runtime.name"); This looks like Hungarian Notation, which is not normal in openjdk code, so s/sJdkPath/jdkPath/ etc. On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield wrote: > I have created an OpenJDK negative test that confirms the closed source > files are not included in the SOURCE. > > Version of the actual test for review: > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ > > Any comments/suggestions are welcome, also I will need a sponsor for it at > the end? > > Randy > > From Randy.Crihfield at Oracle.com Tue Dec 19 00:05:49 2017 From: Randy.Crihfield at Oracle.com (Randy Crihfield) Date: Mon, 18 Dec 2017 19:05:49 -0500 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: References: <5A37F318.5020207@Oracle.com> Message-ID: <5A3857DD.2080402@Oracle.com> On 12/18/17 02:40 PM, Martin Buchholz wrote: Here it is Mr Martin http://cr.openjdk.java.net/~shurailine/8192837/webrev.01/ Fixed the bugID link and changed the variables to what I believe you are asking for. Also made the bug public. Thanks for the help! Randy > Hi Randy, > > Your bugid link in the webrev is broken - points to bogus > https://bugs.openjdk.java.net/browse/https://bugs.openjdk.java.net/browse/JDK-8192837 > > Please fix. > > OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 is not public - > please fix. > > 92 String sJdkPath = System.getProperty("test.jdk"); > 93 String sRuntime = System.getProperty("java.runtime.name "); > This looks like Hungarian Notation, which is not normal in openjdk > code, so s/sJdkPath/jdkPath/ etc. > > > On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield > > wrote: > > I have created an OpenJDK negative test that confirms the closed > source files are not included in the SOURCE. > > Version of the actual test for review: > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ > > > Any comments/suggestions are welcome, also I will need a sponsor > for it at the end? > > Randy > > From martinrb at google.com Tue Dec 19 00:23:09 2017 From: martinrb at google.com (Martin Buchholz) Date: Mon, 18 Dec 2017 16:23:09 -0800 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: <5A3857DD.2080402@Oracle.com> References: <5A37F318.5020207@Oracle.com> <5A3857DD.2080402@Oracle.com> Message-ID: 28 * @summary Test to verify release file not contains close repo info if it's open bundle "closed" would be much clearer than "close" --- 58 readIn.trim(); This is a classic bug that should really be caught by static analysis somewhere (e.g. http://errorprone.info/bugpattern/CheckReturnValue) On Mon, Dec 18, 2017 at 4:05 PM, Randy Crihfield wrote: > On 12/18/17 02:40 PM, Martin Buchholz wrote: > > Here it is Mr Martin > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.01/ > > Fixed the bugID link and changed the variables to what I believe you are > asking for. > Also made the bug public. > Thanks for the help! > > Randy > > > Hi Randy, > > Your bugid link in the webrev is broken - points to bogus > https://bugs.openjdk.java.net/browse/https://bugs.openjdk. > java.net/browse/JDK-8192837 > Please fix. > > OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 is not public - > please fix. > > 92 String sJdkPath = System.getProperty("test.jdk"); > 93 String sRuntime = System.getProperty("java.runtime.name"); > > This looks like Hungarian Notation, which is not normal in openjdk code, > so s/sJdkPath/jdkPath/ etc. > > > On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield < > Randy.Crihfield at oracle.com> wrote: > >> I have created an OpenJDK negative test that confirms the closed source >> files are not included in the SOURCE. >> >> Version of the actual test for review: >> >> http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ >> >> Any comments/suggestions are welcome, also I will need a sponsor for it >> at the end? >> >> Randy >> >> > > From Randy.Crihfield at Oracle.com Tue Dec 19 00:49:48 2017 From: Randy.Crihfield at Oracle.com (Randy Crihfield) Date: Mon, 18 Dec 2017 19:49:48 -0500 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: References: <5A37F318.5020207@Oracle.com> <5A3857DD.2080402@Oracle.com> Message-ID: <5A38622C.4060708@Oracle.com> If I understand this correctly, this fixes the bug and gives more clarity. http://cr.openjdk.java.net/~shurailine/8192837/webrev.02/ Randy On 12/18/17 07:23 PM, Martin Buchholz wrote: > 28 * @summary Test to verify release file not contains close repo info if it's open bundle > "closed" would be much clearer than "close" > > --- > > 58 readIn.trim(); > This is a classic bug that should really be caught by static analysis > somewhere (e.g. http://errorprone.info/bugpattern/CheckReturnValue) > > > On Mon, Dec 18, 2017 at 4:05 PM, Randy Crihfield > > wrote: > > On 12/18/17 02:40 PM, Martin Buchholz wrote: > > Here it is Mr Martin > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.01/ > > > Fixed the bugID link and changed the variables to what I believe > you are asking for. > Also made the bug public. > Thanks for the help! > > Randy > > >> Hi Randy, >> >> Your bugid link in the webrev is broken - points to bogus >> https://bugs.openjdk.java.net/browse/https://bugs.openjdk.java.net/browse/JDK-8192837 >> >> >> Please fix. >> >> OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 >> is not public >> - please fix. >> >> 92 String sJdkPath = System.getProperty("test.jdk"); >> 93 String sRuntime = System.getProperty("java.runtime.name "); >> This looks like Hungarian Notation, which is not normal in >> openjdk code, so s/sJdkPath/jdkPath/ etc. >> >> >> On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield >> > >> wrote: >> >> I have created an OpenJDK negative test that confirms the >> closed source files are not included in the SOURCE. >> >> Version of the actual test for review: >> >> http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ >> >> >> Any comments/suggestions are welcome, also I will need a >> sponsor for it at the end? >> >> Randy >> >> > > From martinrb at google.com Tue Dec 19 01:10:00 2017 From: martinrb at google.com (Martin Buchholz) Date: Mon, 18 Dec 2017 17:10:00 -0800 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: <5A38622C.4060708@Oracle.com> References: <5A37F318.5020207@Oracle.com> <5A3857DD.2080402@Oracle.com> <5A38622C.4060708@Oracle.com> Message-ID: Thanks for implementing my test suggestions, Randy. Looks good. It looks like all builds should have a "release" file, and all "release" files should have a SOURCE= line, so I might make those checks unconditonally. If there are other sanity checks for the "release" file, then I might try to coalesce the tests so there is less duplication. On Mon, Dec 18, 2017 at 4:49 PM, Randy Crihfield wrote: > > If I understand this correctly, this fixes the bug and gives more clarity. > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.02/ > > Randy > > > On 12/18/17 07:23 PM, Martin Buchholz wrote: > > 28 * @summary Test to verify release file not contains close repo info if it's open bundle > > "closed" would be much clearer than "close" > > --- > > 58 readIn.trim(); > > This is a classic bug that should really be caught by static analysis > somewhere (e.g. http://errorprone.info/bugpattern/CheckReturnValue) > > > On Mon, Dec 18, 2017 at 4:05 PM, Randy Crihfield < > Randy.Crihfield at oracle.com> wrote: > >> On 12/18/17 02:40 PM, Martin Buchholz wrote: >> >> Here it is Mr Martin >> >> http://cr.openjdk.java.net/~shurailine/8192837/webrev.01/ >> >> Fixed the bugID link and changed the variables to what I believe you are >> asking for. >> Also made the bug public. >> Thanks for the help! >> >> Randy >> >> >> Hi Randy, >> >> Your bugid link in the webrev is broken - points to bogus >> https://bugs.openjdk.java.net/browse/https://bugs.openjdk.ja >> va.net/browse/JDK-8192837 >> Please fix. >> >> OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 is not public - >> please fix. >> >> 92 String sJdkPath = System.getProperty("test.jdk"); >> 93 String sRuntime = System.getProperty("java.runtime.name"); >> >> This looks like Hungarian Notation, which is not normal in openjdk code, >> so s/sJdkPath/jdkPath/ etc. >> >> >> On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield < >> Randy.Crihfield at oracle.com> wrote: >> >>> I have created an OpenJDK negative test that confirms the closed source >>> files are not included in the SOURCE. >>> >>> Version of the actual test for review: >>> >>> http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ >>> >>> Any comments/suggestions are welcome, also I will need a sponsor for it >>> at the end? >>> >>> Randy >>> >>> >> >> > > From Randy.Crihfield at Oracle.com Tue Dec 19 13:11:43 2017 From: Randy.Crihfield at Oracle.com (Randy Crihfield) Date: Tue, 19 Dec 2017 08:11:43 -0500 Subject: RFR: 8192837 Need new test for release file info In-Reply-To: References: <5A37F318.5020207@Oracle.com> <5A3857DD.2080402@Oracle.com> <5A38622C.4060708@Oracle.com> Message-ID: <5A39100F.9020503@Oracle.com> Understood, thanks. Because of a JPRT bug I'm reluctant to insist the "line" part is always there, I was told it could cause false failures in many cases. There actually is a positive case in the works but the SQE in me likes to keep the cases separate so one might know what actually failed without deep diving every time, makes triage faster. Thanks for your detailed help! Randy On 12/18/17 08:10 PM, Martin Buchholz wrote: > Thanks for implementing my test suggestions, Randy. Looks good. > > It looks like all builds should have a "release" file, and all > "release" files should have a SOURCE= line, so I might make those > checks unconditonally. If there are other sanity checks for the > "release" file, then I might try to coalesce the tests so there is > less duplication. > > On Mon, Dec 18, 2017 at 4:49 PM, Randy Crihfield > > wrote: > > > If I understand this correctly, this fixes the bug and gives more > clarity. > > http://cr.openjdk.java.net/~shurailine/8192837/webrev.02/ > > > Randy > > > On 12/18/17 07:23 PM, Martin Buchholz wrote: >> 28 * @summary Test to verify release file not contains close repo info if it's open bundle >> "closed" would be much clearer than "close" >> >> --- >> >> 58 readIn.trim(); >> This is a classic bug that should really be caught by static >> analysis somewhere (e.g. >> http://errorprone.info/bugpattern/CheckReturnValue >> ) >> >> >> On Mon, Dec 18, 2017 at 4:05 PM, Randy Crihfield >> > >> wrote: >> >> On 12/18/17 02:40 PM, Martin Buchholz wrote: >> >> Here it is Mr Martin >> >> http://cr.openjdk.java.net/~shurailine/8192837/webrev.01/ >> >> >> Fixed the bugID link and changed the variables to what I >> believe you are asking for. >> Also made the bug public. >> Thanks for the help! >> >> Randy >> >> >>> Hi Randy, >>> >>> Your bugid link in the webrev is broken - points to bogus >>> https://bugs.openjdk.java.net/browse/https://bugs.openjdk.java.net/browse/JDK-8192837 >>> >>> >>> Please fix. >>> >>> OTOH, https://bugs.openjdk.java.net/browse/JDK-8192837 >>> is not >>> public - please fix. >>> >>> 92 String sJdkPath = System.getProperty("test.jdk"); >>> 93 String sRuntime = System.getProperty("java.runtime.name "); >>> This looks like Hungarian Notation, which is not normal in >>> openjdk code, so s/sJdkPath/jdkPath/ etc. >>> >>> >>> On Mon, Dec 18, 2017 at 8:55 AM, Randy Crihfield >>> >> > wrote: >>> >>> I have created an OpenJDK negative test that confirms >>> the closed source files are not included in the SOURCE. >>> >>> Version of the actual test for review: >>> >>> http://cr.openjdk.java.net/~shurailine/8192837/webrev.00/ >>> >>> Any comments/suggestions are welcome, also I will need a >>> sponsor for it at the end? >>> >>> Randy >>> >>> >> >> > >