[PATCH 0/2] acl: avoid vector spills in the RVV search loops

Konstantin Ananyev konstantin.v.ananyev at yandex.ru
Mon Nov 2 09:41:50 CET 2026



01.11.2026 20:57, Stephen Hemminger пишет:
> The compiler is doing what we asked. Why workaround it.

Same thought: why simply not to use -O3 then?

>
> On Sun, Nov 1, 2026, 9:04 AM Sun Yuechi <sunyuechi at iscas.ac.cn> wrote:
>
>     At -O2 the ACL match checks are not inlined into the RVV search
>     loops, so
>     a vector value live across those calls is spilled and the
>     remaining locals
>     are addressed through vlenb.  -O3 inlines them already and is
>     unchanged.
>
>     Add __rte_flatten, and apply it to the two search functions.
>
>     Sun Yuechi (2):
>       eal: add flatten function attribute
>       acl: inline the match checks into the RVV search loops
>
>      lib/acl/acl_run_rvv.h        |  8 +++++---
>      lib/eal/include/rte_common.h | 13 +++++++++++++
>      2 files changed, 18 insertions(+), 3 deletions(-)
>
>     -- 
>     2.56.0
>



More information about the dev mailing list