RFR: 7231: JMC createReport for JMC8 Automated Analysis fails to evaluate several rules [v3]

[Marcus Hirt]

On Wed, 29 Sep 2021 17:14:18 GMT, Alex Macdonald <aptmac at openjdk.org> wrote:

>> This PR addresses JMC-7231 [[0]](https://bugs.openjdk.java.net/browse/JMC-7231), in which using `JfrHtmlRulesReport` throws a number of exceptions and fails to evaluate a handful of rules.
>> 
>> I used some of the review comments [[1]](https://github.com/openjdk/jmc/pull/302#pullrequestreview-726780649) that I made in the other open PR to address the same issue.
>> 
>> As far as I can tell, there are two main issues here.
>> 
>> The first is some rules now (after the Rules 2.0 update) have dependencies, these are (4) `GcLockerRule`, `GcStallRule`, `HeapInspectionRule`, and `SystemGcRule` which depend on `GarbageCollectionInfoRule`. The problem here is that if `GarbageCollectionInfoRule` isn't evaluated then the other four rules end up throwing an NPE while trying to evaluate.
>> 
>> The second issue is that as of Rules 2.0, on the jmc.ui Automated Analysis page we use `RulesToolkit.matchesEventAvailabilityMap()` to verify is an event is is available and if the rule should be evaluated. This check wasn't replicated on the `core` implementation, so I've added it in. What's interesting here is that in the test jfr I was using, `GarbageCollectionInfoRule` was filtered out by `matchesEventAvailabilityMap()`, which was causing the above 4 Gc rules to explode.
>> 
>> The solution I've put together uses two lists to track 1). rules that are unavailable due to being filtered out by `matchesEventAvailabilityMap()`, and 2). rules that have dependencies. It later loops through the list of rules with dependencies, and checks to see if they exist in the list with unavailable rules. If a rule is not able to be evaluated by either 1). not being available or 2). by having a dependency that is not available, it returns a completed future `IResult` containing "not available", similar to how on this is handled on the jmc.ui side. Then the rest of the rules report proceeds as usual.
>> 
>> I think the easiest way to verify/try this fix is by creating a run configuration for JfrHtmlRulesReport in Eclipse, and feeding it a jfr file.
>> 
>> [0] https://bugs.openjdk.java.net/browse/JMC-7231
>> [1] https://github.com/openjdk/jmc/pull/302#pullrequestreview-726780649
>
> Alex Macdonald has updated the pull request incrementally with one additional commit since the last revision:
> 
>   run mvn spotless:apply

LGTM, but I'd prefer @Gunde to review this.

It could be worth getting this particular fix into an 8.1.1 release.

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

PR: https://git.openjdk.java.net/jmc/pull/311