RFR: 8255343: Problemlist java/util/stream/SpliteratorTest.java on 32-bit platforms until JDK-8254162

[Aleksey Shipilev]

On Fri, 23 Oct 2020 14:57:06 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

>> Thing is, I don't think there is a way to problem list the _TestNG_ testcase, is it? There is only a way to problem list the entire jtreg test or its separate `@run` with ID. Which is what the patch does. I would think that `generic-i586` only touches `x86_32`, and thus the test, while disabled there, is still operational on every other platform.
>> 
>> I would love if there is was pointed fix that resolves the actual issue, though.
>
> Can you see if this helps?
> 
> diff --git a/test/jdk/java/util/stream/test/org/openjdk/tests/java/util/stream/SegmentTestDataProvider.java b/test/jdk/java/util/stream/test/org/openjdk/tests/java/util/stream/SegmentTestDataProvider.java
> index 2d2d33d0ab7..c5410865a2b 100644
> --- a/test/jdk/java/util/stream/test/org/openjdk/tests/java/util/stream/SegmentTestDataProvider.java
> +++ b/test/jdk/java/util/stream/test/org/openjdk/tests/java/util/stream/SegmentTestDataProvider.java
> @@ -40,12 +40,12 @@ import org.testng.annotations.DataProvider;
>  public class SegmentTestDataProvider {
>  
>      static VarHandle BYTE_HANDLE = MemoryLayouts.JAVA_BYTE.varHandle(byte.class);
> -    static VarHandle CHAR_HANDLE = MemoryLayouts.JAVA_CHAR.varHandle(char.class);
> -    static VarHandle SHORT_HANDLE = MemoryLayouts.JAVA_SHORT.varHandle(short.class);
> -    static VarHandle INT_HANDLE = MemoryLayouts.JAVA_INT.varHandle(int.class);
> -    static VarHandle LONG_HANDLE = MemoryLayouts.JAVA_LONG.varHandle(long.class);
> -    static VarHandle FLOAT_HANDLE = MemoryLayouts.JAVA_FLOAT.varHandle(float.class);
> -    static VarHandle DOUBLE_HANDLE = MemoryLayouts.JAVA_DOUBLE.varHandle(double.class);
> +    static VarHandle CHAR_HANDLE = MemoryLayouts.JAVA_CHAR.withBitAlignment(8).varHandle(char.class);
> +    static VarHandle SHORT_HANDLE = MemoryLayouts.JAVA_SHORT.withBitAlignment(8).varHandle(short.class);
> +    static VarHandle INT_HANDLE = MemoryLayouts.JAVA_INT.withBitAlignment(8).varHandle(int.class);
> +    static VarHandle LONG_HANDLE = MemoryLayouts.JAVA_LONG.withBitAlignment(8).varHandle(long.class);
> +    static VarHandle FLOAT_HANDLE = MemoryLayouts.JAVA_FLOAT.withBitAlignment(8).varHandle(float.class);
> +    static VarHandle DOUBLE_HANDLE = MemoryLayouts.JAVA_DOUBLE.withBitAlignment(8).varHandle(double.class);
>  
>      static boolean compareSegmentsByte(Collection<MemorySegment> segments1, Collection<MemorySegment> segments2, boolean isOrdered) {
> :
> ass);
> -    static VarHandle INT_HANDLE = MemoryLayouts.JAVA_INT.varHandle(int.class);
> -    static VarHandle LONG_HANDLE = MemoryLayouts.JAVA_LONG.varHandle(long.class);
> -    static VarHandle FLOAT_HANDLE = MemoryLayouts.JAVA_FLOAT.varHandle(float.class);
> -    static VarHandle DOUBLE_HANDLE = MemoryLayouts.JAVA_DOUBLE.varHandle(double.class);
> +    static VarHandle CHAR_HANDLE = MemoryLayouts.JAVA_CHAR.withBitAlignment(8).varHandle(char.class);
> +    static VarHandle SHORT_HANDLE = MemoryLayouts.JAVA_SHORT.withBitAlignment(8).varHandle(short.class);
> +    static VarHandle INT_HANDLE = MemoryLayouts.JAVA_INT.withBitAlignment(8).varHandle(int.class);
> +    static VarHandle LONG_HANDLE = MemoryLayouts.JAVA_LONG.withBitAlignment(8).varHandle(long.class);
> +    static VarHandle FLOAT_HANDLE = MemoryLayouts.JAVA_FLOAT.withBitAlignment(8).varHandle(float.class);
> +    static VarHandle DOUBLE_HANDLE = MemoryLayouts.JAVA_DOUBLE.withBitAlignment(8).varHandle(double.class);
>  
>      static boolean compareSegmentsByte(Collection<MemorySegment> segments1, Collection<MemorySegment> segments2, boolean isOrdered) {
>          Function<MemorySegment, Byte> mapper = segment -> (byte)BYTE_HANDLE.get(segment.baseAddress());

> Can you see if this helps?
>      static boolean compareSegmentsByte(Collection<MemorySegment> segments1, Collection<MemorySegment> segments2, boolean isOrdered) {
> :
> ass);

I can, but the patch is corrupted near this line. Ah, that seems to be the stray copy of the hunk above.

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

PR: https://git.openjdk.java.net/jdk/pull/836