RFR: 8370370: Add still more cases to WorstCaseTests [v2]

[Raffaello Giulietti]

On Mon, 27 Oct 2025 16:30:44 GMT, Joe Darcy <darcy at openjdk.org> wrote:

>>> @jddarcy Can you check the discrepancy in the argument?
>> 
>> This test case is commented out in WorstCaseTests since it is larger than 1 ulp. To compensate, I made sure the hyperbolics had these values added to their StrictMath tests.
>> 
>> (There are utility methods in the testing library than can test for multiple ulp errors, but we'd have to so something like ceil(ulps) plus one or two to make that work.)
>
> I could add a comment in the hyperbolics saying "See StrictMath tests" if you think that would clarify the situation.

Yes, the test case is commented.
However, the discrepancy is really huge (sign, exponent, and some significand hexdigits), so this might be a copy & paste error here. Or is the big error done on purpose?

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

PR Review Comment: https://git.openjdk.org/jdk/pull/27939#discussion_r2466336730