On Wed, 16 Apr 2025 19:29:07 GMT, Vladimir Ivanov <vliva...@openjdk.org> wrote:
>> Migrate Vector API math library (SVML and SLEEF) linkage from native code >> (in JVM) to Java FFM API. >> >> Since FFM API doesn't support vector calling conventions yet, migration >> affects only symbol lookup for now. But it still enables significant >> simplifications on JVM side. >> >> The patch consists of the following parts: >> * on-demand symbol lookup in Java code replaces eager lookup from native >> code during JVM startup; >> * 2 new VM intrinsics for vector calls (support unary and binary shapes) >> (code separated from unary/binary vector operations); >> * new internal interface to query supported CPU ISA extensions >> (`jdk.incubator.vector.CPUFeatures`) used for CPU dispatching. >> >> `java.lang.foreign` API is used to perform symbol lookup in vector math >> library, then the address is cached and fed into corresponding JVM >> intrinsic, so C2 can turn it into a direct vector call in generated code. >> >> Once `java.lang.foreign` supports vectors & vector calling conventions, VM >> intrinsics can go away. >> >> Performance is on par with original implementation (tested with >> microbenchmarks on linux-x64 and macosx-aarch64). >> >> Testing: hs-tier1 - hs-tier6, microbenchmarks (on linux-x64 and >> macosx-aarch64) >> >> Thanks! > > Vladimir Ivanov has updated the pull request with a new target base due to a > merge or a rebase. The pull request now contains 21 commits: > > - Merge branch 'master' into vector.math.01.java > - Fix debugName handling > - Merge branch 'master' into vector.math.01.java > - RVV and SVE adjustments > - Merge branch 'master' into vector.math.01.java > - Fix windows-aarch64 build failure > - features_string -> cpu_info_string > - Reviews and Float64Vector-related fix > - Misc fixes and cleanups > - CPU features support > - ... and 11 more: https://git.openjdk.org/jdk/compare/98dac46a...a288cbbf Both `CPUFeatures.validateFeatures` and the checks it performs are assertions, so it has to be explicitly enabled by `-esa`. The assert validates some assumptions the code has about the format of features string VM produces. And feature names are controlled by JVM (JVM assigns names itself). If there's a mismatch, something is really broken. Why do you think it makes sense to be defensive here and try to recover if the checks fail? ------------- PR Comment: https://git.openjdk.org/jdk/pull/24462#issuecomment-2810842069
