ColinLeeo commented on issue #1073:
URL: https://github.com/apache/tsfile/issues/1073#issuecomment-6905148122
Resolved by #1078 (commit 18483414a). The RLE bit-unpack kernels now
dispatch to native AVX2 intrinsics at runtime: `cpp/src/common/simd_dispatch.h`
exposes `common::cpu_has_avx2()` (`__builtin_cpu_supports("avx2")`) and
`int32_packer.h` / `int64_packer.h` pick the native kernel per packer instance
instead of relying on the translation unit being compiled with `-mavx2`. The
widths that can cross a ninth byte (59/61/62/63) keep the 128-bit buffered
scalar fallback, and the simde path remains for CPUs without AVX2.
Closing as completed in develop.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]