Re: [PATCH] Add gnu::diagnose_as attribute

[Jason Merrill <jason@redhat.com>]

On 7/1/21 5:28 AM, Matthias Kretz wrote:
> On Tuesday, 22 June 2021 22:12:42 CEST Jason Merrill wrote:
>> On 6/22/21 4:01 PM, Matthias Kretz wrote:
>>> On Tuesday, 22 June 2021 21:52:16 CEST Jason Merrill wrote:
>>>> For alias templates, you probably want the attribute only on the
>>>> templated class, not on the instantiations.
>>>
>>> Oh good point. My current patch does not allow the attribute on alias
>>> templates. Consider:
>>>
>>> template <class T, class U>
>>>
>>>     struct X {};
>>>
>>> template <class T>
>>>
>>>     using foo [[gnu::diagnose_as]] = X<T, int>;
>>>
>>> I have no idea how this could work. I would have to set the attribute for
>>> an implicit partial specialization (not that I know of the existence of
>>> such a thing)? I.e. X<int, int> would have to be diagnosed as foo<int>,
>>> but X<int, float> would have to be diagnosed as X<int, float>, not foo.
>>>
>>> So if anything it should only support alias templates if they are strictly
>>> "renaming" the type. I.e. their template parameters must match up exactly.
>>> Can I constrain the attribute like this?
>>
>> Yes.  You can check that with get_underlying_template.
>>
>> Or you could support the above by putting the attribute on the
>> instantiation with the TEMPLATE_INFO for foo<A> rather than a simple name.
> 
> Question, given:
> 
>    template <class T> using foo = bar<T>;
> 
> The diagnose_as attribute handler isn't called until e.g. `foo<int>` is
> instantiated.

You probably want to adjust is_late_template_attribute to change that.

> Meaning that even after the declaration of the alias template
> `bar<int>` will not be diagnosed as `foo<int>`, which happens only after the
> first use of `foo<int>`. I find that more confusing than helpful, even if the
> expectation would be that users only use the alias template.
> 
> So do you still expect alias templates to support diagnose_as? And if yes, how
> could I handle the attribute so that the diagnose_as attribute is applied to
> `bar` on declaration of `foo`?
>