[
https://issues.apache.org/jira/browse/ARROW-15739?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yibo Cai updated ARROW-15739:
-----------------------------
Description:
Besides improvements of Neon any/all implementation, we can simplify below code
{code:cpp}
#ifdef ARROW_HAVE_NEON
using simd_batch = xsimd::batch<int8_t, xsimd::neon64>;
#else
using simd_batch = xsimd::batch<int8_t, xsimd::sse4_2>;
#endif
{code}
to
{code:cpp}
using simd_batch = xsimd::make_sized_batch_t<int8_t, 16>;
{code}
was:
Besides improvements of Neon any/all implementation, we can simply below code
{code:cpp}
#ifdef ARROW_HAVE_NEON
using simd_batch = xsimd::batch<int8_t, xsimd::neon64>;
#else
using simd_batch = xsimd::batch<int8_t, xsimd::sse4_2>;
#endif
{code}
to
{code:cpp}
using simd_batch = xsimd::make_sized_batch_t<int8_t, 16>;
{code}
> [C++] Bump xsimd to latest version
> ----------------------------------
>
> Key: ARROW-15739
> URL: https://issues.apache.org/jira/browse/ARROW-15739
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++
> Reporter: Yibo Cai
> Assignee: Yibo Cai
> Priority: Major
>
> Besides improvements of Neon any/all implementation, we can simplify below
> code
> {code:cpp}
> #ifdef ARROW_HAVE_NEON
> using simd_batch = xsimd::batch<int8_t, xsimd::neon64>;
> #else
> using simd_batch = xsimd::batch<int8_t, xsimd::sse4_2>;
> #endif
> {code}
> to
> {code:cpp}
> using simd_batch = xsimd::make_sized_batch_t<int8_t, 16>;
> {code}
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
