Ping.

On Sat, 2024-06-15 at 21:47 +0800, Xi Ruoyao wrote:
> The first form has a lower latency (due to the special handling of
> "move" in LA464 and LA664) despite it's longer.
> 
> gcc/ChangeLog:
> 
>       * config/loongarch/loongarch.md (define_peephole2): Require
>       optimize_insn_for_size_p () for move/move/bstrins =>
>       srai/bstrins transform.
> ---
> 
> Bootstrapped and regtested on loongarch64-linux-gnu.  Ok for trunk?
> 
>  gcc/config/loongarch/loongarch.md | 9 ++++++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
> 
> diff --git a/gcc/config/loongarch/loongarch.md
> b/gcc/config/loongarch/loongarch.md
> index 25c1d323ba0..e4434c3bd4e 100644
> --- a/gcc/config/loongarch/loongarch.md
> +++ b/gcc/config/loongarch/loongarch.md
> @@ -1617,20 +1617,23 @@ (define_insn_and_split
> "*bstrins_<mode>_for_ior_mask"
>    })
>  
>  ;; We always avoid the shift operation in bstrins_<mode>_for_ior_mask
> -;; if possible, but the result may be sub-optimal when one of the
> masks
> +;; if possible, but the result may be larger when one of the masks
>  ;; is (1 << N) - 1 and one of the src register is the dest register.
>  ;; For example:
>  ;;     move          t0, a0
>  ;;     move          a0, a1
>  ;;     bstrins.d     a0, t0, 42, 0
>  ;;     ret
> -;; using a shift operation would be better:
> +;; using a shift operation would be smaller:
>  ;;     srai.d                t0, a1, 43
>  ;;     bstrins.d     a0, t0, 63, 43
>  ;;     ret
>  ;; unfortunately we cannot figure it out in split1: before reload we
> cannot
>  ;; know if the dest register is one of the src register.  Fix it up
> in
>  ;; peephole2.
> +;;
> +;; Note that the first form has a lower latency so this should only
> be
> +;; done when optimizing for size.
>  (define_peephole2
>    [(set (match_operand:GPR 0 "register_operand")
>       (match_operand:GPR 1 "register_operand"))
> @@ -1639,7 +1642,7 @@ (define_peephole2
>                         (match_operand:SI 3 "const_int_operand")
>                         (const_int 0))
>       (match_dup 0))]
> -  "peep2_reg_dead_p (3, operands[0])"
> +  "peep2_reg_dead_p (3, operands[0]) && optimize_insn_for_size_p ()"
>    [(const_int 0)]
>    {
>      int len = GET_MODE_BITSIZE (<MODE>mode) - INTVAL (operands[3]);

-- 
Xi Ruoyao <xry...@xry111.site>
School of Aerospace Science and Technology, Xidian University

Reply via email to