Re: [PATCH v7] c++: Move consteval folding to cp_fold_r

[Jason Merrill <jason@redhat.com>]

On 9/19/23 09:01, Marek Polacek wrote:
> On Mon, Sep 18, 2023 at 09:36:31PM -0400, Jason Merrill wrote:
>> On 9/18/23 17:42, Marek Polacek wrote:
>>> +  /* The purpose of this is not to emit errors for mce_unknown.  */
>>> +  const tsubst_flags_t complain = (data->flags == ff_fold_immediate
>>> +				   ? tf_none : tf_error);
>>
>> Maybe check flags & ff_mce_false, instead?  OK with that change.
> 
> Thanks!  And what do you think about:
> 
> --- a/gcc/cp/cp-gimplify.cc
> +++ b/gcc/cp/cp-gimplify.cc
> @@ -1162,7 +1162,8 @@ cp_fold_r (tree *stmt_p, int *walk_subtrees, void *data_)
>     tree stmt = *stmt_p;
>     enum tree_code code = TREE_CODE (stmt);
> 
> -  cp_fold_immediate_r (stmt_p, walk_subtrees, data);
> +  if (cxx_dialect > cxx17)
> +    cp_fold_immediate_r (stmt_p, walk_subtrees, data);
> 
>     *stmt_p = stmt = cp_fold (*stmt_p, data->flags);
> 
> since we can recurse on ?:, this could save some time.  It's sad that
> it checks cxx_dialect in every invocation of cp_fold_r but still, it
> should help.

Sure.

Jason