On Sat, 30 May 2026 19:21:10 GMT, Ashutosh Mehra <[email protected]> wrote:

>> This patch is reorganizing the code to split `get_processor_features()` into 
>> multiple helper functions. Details are present in the issue 
>> https://bugs.openjdk.org/browse/JDK-8380448.
>> It should not change the behavior of any VM flag, except 
>> `UseVectorizedHashCodeIntrinsic`. The earlier code always set 
>> `UseVectorizedHashCodeIntrinsic` to `true` if UseAVX >= 2. That means 
>> specifying `-XX:-UseVectorizedHashCodeIntrinsic` on command line wouldn't 
>> disable use of this intrinsic. This looks like a bug and has been fixed in 
>> this patch.
>> 
>> ---------
>> - [x] I confirm that I make this contribution in accordance with the 
>> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai).
>
> Ashutosh Mehra has updated the pull request incrementally with one additional 
> commit since the last revision:
> 
>   Fix another mistake
>   
>   Signed-off-by: Ashutosh Mehra <[email protected]>

src/hotspot/cpu/x86/globals_x86.hpp line 182:

> 180:   /* condition. When this value is set to zero clear operations that */  
>    \
> 181:   /* work on memory blocks can also use AVX512 intrinsics. */            
>    \
> 182:   product(int, CopyAVX3Threshold, AVX3Threshold, DIAGNOSTIC,             
>    \

I don't think this will work as expected/intended. IIUC this will initialize 
`CopyAVX3Threshold` to the default build-time value of `AVX3Threshold`, not any 
runtime value that might be set on the command-line.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/31301#discussion_r3339085148

Reply via email to