RFR: 8336274: MutableBigInteger.leftShift(int) optimization [v5]

Raffaello Giulietti rgiulietti at openjdk.org
Tue Sep 3 16:07:23 UTC 2024


On Tue, 3 Sep 2024 15:49:01 GMT, fabioromano1 <duke at openjdk.org> wrote:

>> It would be nice to see some benchmarks where it gives performance benefits.
>
>> It would be nice to see some benchmarks where it gives performance benefits.
> 
> @kuksenko Try to run the benchmark of the `BigInteger`'s square root, maybe there the benefits are more visibles...

@fabioromano1 The `MutableBigIntegersLeftShift` benchmark class in this PR does not work as written.
While the schema with `MutableBigIntegerBox` to get access to package-private methods/fields in package-private types works for unit tests, it does not for benchmarks.

I suggest you show evidence of better performance by restructuring the benchmarks to target public API points that indirectly make use of `MutableBigInteger::leftShift`, for example `BigInteger` square root, as you point out above.

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

PR Comment: https://git.openjdk.org/jdk/pull/20008#issuecomment-2326897259


More information about the core-libs-dev mailing list