On Thu, 23 Feb 2023 05:16:52 GMT, Sergey Bylokhov <s...@openjdk.org> wrote:
> Implementation of the "ManagementAgent.status" indirectly depends on the size > of "O_BUFLEN". > > The root cause is usage of > [print_cr()](https://github.com/openjdk/jdk/blob/master/src/hotspot/share/utilities/ostream.cpp#L153) > which may > [truncate](https://github.com/openjdk/jdk/blob/master/src/hotspot/share/utilities/ostream.cpp#L101) > the output. > > The solution is similar to the > [JDK-8263640](https://bugs.openjdk.org/browse/JDK-8263640). see discussion > about that approach https://github.com/openjdk/jdk/pull/4616 Change seems reasonable. Can we get an actual regression test to verify this? test/jdk/sun/management/jmxremote/startstop/JMXStatusTest.java line 36: > 34: /** > 35: * @test > 36: * @bug 8023093 8138748 8142398 8303102 As this test doesn't explicitly test/check the issue in 8303102 I don't think it needs to belisted on the @bug line. (And if it does then the copyright year needs t be updated too). ------------- PR: https://git.openjdk.org/jdk/pull/12724
