On Wed, 3 Jun 2026 02:19:30 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 setting of CopyAVX3Threshold > > Signed-off-by: Ashutosh Mehra <[email protected]> Thank you, @ashu-mehra, for doing this refactoring. I have few comments. src/hotspot/cpu/x86/vm_version_x86.cpp line 1105: > 1103: // For AMD Processors use XMM/YMM MOVDQU instructions > 1104: // for Object Initialization as default > 1105: if (is_amd()) { Do we need this? We are in `amd_config()`. src/hotspot/cpu/x86/vm_version_x86.cpp line 1726: > 1724: > 1725: if (is_intel()) { // Intel cpus specific settings > 1726: if (is_knights_family()) { I think we can combine checks into one condition here. There is separate issue not related to these changes which I realized by looking on this code. `-XX:+UseKNLSetting` testing flag checks are guarded by `is_intel()`. Our testing infra check `avx512` to run tests with this flag. Which worked before AMD implemented AVX512. Now we run tests on AMD with avx512 but CPU features are not disabled :( ------------- PR Review: https://git.openjdk.org/jdk/pull/31301#pullrequestreview-4469706204 PR Review Comment: https://git.openjdk.org/jdk/pull/31301#discussion_r3389938744 PR Review Comment: https://git.openjdk.org/jdk/pull/31301#discussion_r3390046206
