committed, thanks.
On Thu, Dec 29, 2022 at 11:34 PM wrote:
> From: Ju-Zhe Zhong
>
> Currently we use pred_mov to to do the codegen for vse intrinsics.
> However, it
> generates inferior codegen when I am testing AVL model of VSETVL PASS
> using vse intrinsics.
>
> Consider this following code:
From: Ju-Zhe Zhong
Currently we use pred_mov to to do the codegen for vse intrinsics. However, it
generates inferior codegen when I am testing AVL model of VSETVL PASS using vse
intrinsics.
Consider this following code:
void f2 (int * restrict in, int * restrict out, void * restrict mask_in, in
