[PATCH v2 10/51] tcg/optimize: Use fold_masks_zs in fold_count_zeros

Richard Henderson posted 51 patches 3 days, 19 hours ago
There is a newer version of this series
[PATCH v2 10/51] tcg/optimize: Use fold_masks_zs in fold_count_zeros
Posted by Richard Henderson 3 days, 19 hours ago
Avoid the use of the OptContext slots. Find TempOptInfo once.
Compute s_mask from the union of the maximum count and the
op2 fallback for op1 being zero.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 tcg/optimize.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/tcg/optimize.c b/tcg/optimize.c
index b8e9ded821..270051acd1 100644
--- a/tcg/optimize.c
+++ b/tcg/optimize.c
@@ -1583,10 +1583,12 @@ static bool fold_call(OptContext *ctx, TCGOp *op)
 
 static bool fold_count_zeros(OptContext *ctx, TCGOp *op)
 {
-    uint64_t z_mask;
+    uint64_t z_mask, s_mask;
+    TempOptInfo *t1 = arg_info(op->args[1]);
+    TempOptInfo *t2 = arg_info(op->args[2]);
 
-    if (arg_is_const(op->args[1])) {
-        uint64_t t = arg_info(op->args[1])->val;
+    if (t1->is_const) {
+        uint64_t t = t1->val;
 
         if (t != 0) {
             t = do_constant_folding(op->opc, ctx->type, t, 0);
@@ -1605,8 +1607,11 @@ static bool fold_count_zeros(OptContext *ctx, TCGOp *op)
     default:
         g_assert_not_reached();
     }
-    ctx->z_mask = arg_info(op->args[2])->z_mask | z_mask;
-    return false;
+    s_mask = smask_from_zmask(z_mask);
+    z_mask |= t2->z_mask;
+    s_mask &= t2->s_mask;
+
+    return fold_masks_zs(ctx, op, z_mask, s_mask);
 }
 
 static bool fold_ctpop(OptContext *ctx, TCGOp *op)
-- 
2.43.0
Re: [PATCH v2 10/51] tcg/optimize: Use fold_masks_zs in fold_count_zeros
Posted by Pierrick Bouvier 3 days, 3 hours ago
On 12/19/24 20:10, Richard Henderson wrote:
> Avoid the use of the OptContext slots. Find TempOptInfo once.
> Compute s_mask from the union of the maximum count and the
> op2 fallback for op1 being zero.
> 
> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
> ---
>   tcg/optimize.c | 15 ++++++++++-----
>   1 file changed, 10 insertions(+), 5 deletions(-)
> 
> diff --git a/tcg/optimize.c b/tcg/optimize.c
> index b8e9ded821..270051acd1 100644
> --- a/tcg/optimize.c
> +++ b/tcg/optimize.c
> @@ -1583,10 +1583,12 @@ static bool fold_call(OptContext *ctx, TCGOp *op)
>   
>   static bool fold_count_zeros(OptContext *ctx, TCGOp *op)
>   {
> -    uint64_t z_mask;
> +    uint64_t z_mask, s_mask;
> +    TempOptInfo *t1 = arg_info(op->args[1]);
> +    TempOptInfo *t2 = arg_info(op->args[2]);
>   
> -    if (arg_is_const(op->args[1])) {
> -        uint64_t t = arg_info(op->args[1])->val;
> +    if (t1->is_const) {
> +        uint64_t t = t1->val;
>   
>           if (t != 0) {
>               t = do_constant_folding(op->opc, ctx->type, t, 0);
> @@ -1605,8 +1607,11 @@ static bool fold_count_zeros(OptContext *ctx, TCGOp *op)
>       default:
>           g_assert_not_reached();
>       }
> -    ctx->z_mask = arg_info(op->args[2])->z_mask | z_mask;
> -    return false;
> +    s_mask = smask_from_zmask(z_mask);
> +    z_mask |= t2->z_mask;
> +    s_mask &= t2->s_mask;
> +
> +    return fold_masks_zs(ctx, op, z_mask, s_mask);
>   }
>   
>   static bool fold_ctpop(OptContext *ctx, TCGOp *op)

Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>