On Wed, 10 Apr 2024 17:36:32 GMT, Alex Menkov <amen...@openjdk.org> wrote:
>> The test starts listening on dynamic port and calls stopListening with
>> incorrect (cArgs1) and correct (cArgs2) argument maps.
>> Incorrect map is created by finding "free" port (`(new
>> ServerSocket(0)).getLocalPort()`)
>> The test fails if the same port is selected later when the listening starts.
>> To avoid this free port should be picked after listening on dynamic port
>> started.
>>
>> Additionally removed unnecessary `exclusiveAccess.dirs=.` test property and
>> incorrect comment in the description.
>> The test uses `SocketListen` connectors, the coment is about
>> `SharedMemoryListen` connector.
>>
>> Testing: run the test on all Oracle-supported platforms 100 times
>
> Alex Menkov has updated the pull request incrementally with one additional
> commit since the last revision:
>
> Update
> test/hotspot/jtreg/vmTestbase/nsk/jdi/ListeningConnector/stopListening/stoplis001.java
>
> Co-authored-by: Chris Plummer <chris.plum...@oracle.com>
You might want to consider adding the following diff. It will help in case this
issue (or similar) reproduces again:
---
a/test/hotspot/jtreg/vmTestbase/nsk/jdi/ListeningConnector/stopListening/stoplis001.java
+++
b/test/hotspot/jtreg/vmTestbase/nsk/jdi/ListeningConnector/stopListening/stoplis001.java
@@ -92,6 +92,8 @@ private int runIt(String argv[], PrintStream out) {
}
else
log.display("TEST: start listening the address " + addr);
+ log.display("cArgs1: " + cArgs1);
+ log.display("cArgs2: " + cArgs2);
/* Check that an Exception is thrown if ListeningConnector.stopListening
has been invoked with argument map different from the one given for
TEST: start listening the address localhost:42473
cArgs1: {timeout=timeout=, port=port=45521, localAddress=localAddress=}
cArgs2: {timeout=timeout=, port=port=42473, localAddress=localAddress=}
-------------
PR Comment: https://git.openjdk.org/jdk/pull/18705#issuecomment-2052232161
