RFR: 8333813: Seviceability tests fail due to stderr containing Temporarily processing option UseNotificationThread

Jaikiran Pai jpai at openjdk.org
Sat Jun 8 07:38:15 UTC 2024


On Sat, 8 Jun 2024 07:17:55 GMT, Chris Plummer <cjplummer at openjdk.org> wrote:

>> Hello Chris, given these similary named methods in this class, perhaps we should instead just have one single `stderrShouldBeEmptyIgnoring(...)` method which takes the messages that should be ignored. Something like:
>> 
>> 
>> public static final String VM_WARNING_MSG = ".* VM warning:.*";
>> 
>> public static final String VM_WARNING_DEPRECATED_MSG = ".* VM warning:.* deprecated.*";
>> ...
>> 
>> public OutputAnalyzer stderrShouldBeEmptyIgnoring(String ignoreMsg, String... additionalIgnoreMsgs) {
>>     String stdErrContent = getStderr().replaceAll(ignoreMsg + "\\R", "");
>>     if (additionalIgnoreMsgs != null) {
>>     	for (String additionalIgnore : additionalIgnoreMsgs) {
>>     		stdErrContent = stdErrContent.replaceAll(additionalIgnore + "\\R", "");
>>     	}
>>     }
>>     if (!stdErrContent.isEmpty()) {
>>     	reportDiagnosticSummary();
>>     	throw new RuntimeException("stderr was not empty");
>>     }
>>     return this;
>> }
>> 
>> 
>> We make those private fields in OutputAnalyzer public and have the caller pass them:
>> 
>> 
>> oa.stderrShouldBeEmptyIgnoring(OutputAnalyzer.VM_WARNING_DEPRECATED_MSG)
>
> That seems like a lot of busy work and over abstraction for trying to solve a rather trivial issue that already has a very simple solution, but maybe could use a better API name.

In that case, would `stderrShouldBeEmptyIgnoreVMOptionDeprecations` be OK?

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

PR Review Comment: https://git.openjdk.org/jdk/pull/19606#discussion_r1631938967


More information about the serviceability-dev mailing list