RFR: 8305913: com/sun/jdi/JdbLastErrorTest.java failed with "'lastError = 42' missing from stdout/stderr"

[Kevin Walls]

On Fri, 14 Apr 2023 21:20:46 GMT, Chris Plummer <cjplummer at openjdk.org> wrote:

> I guess I don't understand the purpose of having this test if it is just going to work around the underlying issue it is suppose to be testing for. Does it serve a purpose still? Does it still correctly detect the fix that was done for [JDK-8292302](https://bugs.openjdk.org/browse/JDK-8292302)?

Right, and I do wonder about removing it.  It is still verifying that you can try and debug a java app that uses panama, and can set and fetch last error using panama, in a debugged target (after you've called getLastError once).  Now I'm admitting that various other causes make the first attempt commonly fail, and it is testing that things don't get any _worse_ than the first iteration failing, but no it would not detect if we reverted 8292302.  I think it has some value, but in my note above I hinted that if it causes more trouble it should probably go.  If anybody thinks it should go now...

-------------

PR Comment: https://git.openjdk.org/jdk/pull/13481#issuecomment-1509328958