On Wed, 15 Jul 2026 03:46:08 GMT, Kuai Wei <[email protected]> wrote:

>> I recently noticed a behavioral discrepancy in 
>> jdk.internal.util.ArraysSupport.vectorizedMismatch between the Java 
>> implementation and the platform intrinsic implementations.
>> 
>> Current behavior
>> 
>> The Java implementation may leave a tail of elements unchecked, returning 
>> the bitwise complement of the number of remaining elements (i.e., 
>> ~remaining).
>> The x86_64 intrinsic, by contrast, compares all elements and simply returns 
>> -1 when no mismatch is found.
>> 
>> Proposed change
>> 
>> This PR refines the Java implementation so that it always compares all 
>> elements and returns -1 when no mismatch is found, matching the x86_64 
>> intrinsic behavior.
>> 
>> A regression test is included at 
>> `test/hotspot/jtreg/compiler/intrinsics/VectorizedMismatchReturnDiffTest.java`
>>  which demonstrates the original behavioral difference.
>> 
>> ## Test
>> - [x] tier1 test suites on linux x86_64
>> - [x] tier1 test suites on linux aarch64
>> 
>> ---------
>> - [x] I confirm that I make this contribution in accordance with the 
>> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai).
>
> Kuai Wei has updated the pull request incrementally with one additional 
> commit since the last revision:
> 
>   Fix copyright header

Ideally we need to ensure intrinsics return no negative values besides -1.

src/java.base/share/classes/jdk/internal/util/ArraysSupport.java line 114:

> 112:      * and {@code length} (exclusive), of the first mismatching pair of 
> elements
> 113:      * in the two arrays.  Otherwise, if a mismatch is not found the 
> bitwise
> 114:      * compliment of the number of remaining pairs of elements to be 
> checked in

I think now we never return a negative value indicating the number of element 
to check; we only ever return -1. We need to reflect that fact in the comments 
here.

src/java.base/share/classes/jdk/internal/util/ArraysSupport.java line 748:

> 746:                 // Fall back to slow mechanism
> 747:                 // ISSUE: Consider looping over vectorizedMismatch 
> adjusting ranges
> 748:                 // However, requires that returned value be relative to 
> input ranges

You updated the float[] code but not the double[] code. Please keep their code 
consistent.

test/hotspot/jtreg/compiler/intrinsics/VectorizedMismatchReturnDiffTest.java 
line 42:

> 40:  *                   
> -XX:CompileCommand=compileonly,compiler.intrinsics.VectorizedMismatchReturnDiffTest::testComp*
> 41:  *                   -Xbatch -XX:-TieredCompilation
> 42:  *                    compiler.intrinsics.VectorizedMismatchReturnDiffTest

Suggestion:

 *                   compiler.intrinsics.VectorizedMismatchReturnDiffTest

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

PR Review: https://git.openjdk.org/jdk/pull/31802#pullrequestreview-4709832637
PR Review Comment: https://git.openjdk.org/jdk/pull/31802#discussion_r3592113148
PR Review Comment: https://git.openjdk.org/jdk/pull/31802#discussion_r3592124706
PR Review Comment: https://git.openjdk.org/jdk/pull/31802#discussion_r3592129797

Reply via email to