RFR: 8355719: Reduce memory consumption of BigInteger.pow() [v49]

Raffaello Giulietti rgiulietti at openjdk.org
Wed May 7 11:57:20 UTC 2025


On Wed, 30 Apr 2025 15:59:05 GMT, fabioromano1 <duke at openjdk.org> wrote:

>> This PR optimizes `BigInteger.pow(int)` method. The primary enhancement in `pow()` is not concerned most on execution time, but rather in memory optimization, because the PR implementation does the "shift of the exponent" squaring the result rather than the base, so the base is not squared like in the current implementation, and this permits to save about half of the memory.
>
> fabioromano1 has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Simplify long power computing

test/micro/org/openjdk/bench/java/math/BigIntegerPow.java line 72:

> 70: 
> 71:         new Runner(opt).run();
> 72:     }

There's no need for these lines.

test/micro/org/openjdk/bench/java/math/BigIntegerPow.java line 117:

> 115:     @OperationsPerInvocation(TESTSIZE)
> 116:     public void testPowXS(Blackhole bh) {
> 117:         for (BigInteger s : xsArray) {

Suggestion:

        for (BigInteger xs : xsArray) {


and similarly for the other arrays.

test/micro/org/openjdk/bench/java/math/BigIntegerPow.java line 118:

> 116:     public void testPowXS(Blackhole bh) {
> 117:         for (BigInteger s : xsArray) {
> 118:             bh.consume(s.pow(xsExp));

Suggestion:

            bh.consume(s.pow(xsExp & ~0b0111_0101));
            bh.consume(s.pow(xsExp));
            bh.consume(s.pow(xsExp + 1));

or something similar, just to measure other than "all-ones" exponents as well.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/24690#discussion_r2077443258
PR Review Comment: https://git.openjdk.org/jdk/pull/24690#discussion_r2077449324
PR Review Comment: https://git.openjdk.org/jdk/pull/24690#discussion_r2077452722


More information about the core-libs-dev mailing list