[PATCH 1/2 v2] Introduce __builtin_is_virtual_base_of

[Giuseppe D'Angelo]

Hi,

Thank you for the review.

Il 25/09/24 20:13, Jason Merrill ha scritto:
> 
> Supporting the ambiguous case seems pointless to me but, that is what
> the accepted proposal specifies, so indeed that's what we should implement.

It's fundamentally the same with is_base_of, also supporting ambiguous 
and unaccesible bases.

>> +/* Nonzero iff TYPE is derived virtually from PARENT. Ignores accessibility and
>> +   ambiguity issues.  */
>> +#define DERIVED_VIRTUALLY_FROM_P(PARENT, TYPE) \
>> +  (lookup_base ((TYPE), (PARENT), ba_require_virtual, NULL, tf_none) != NULL_TREE)
> I don't think we need a macro for this.

I've added it to match the existing DERIVED_FROM_P macro, but can 
certainly remove it, since it's got only one user. Should I remove it or 
keep it for symmetry?


>> +  ba_require_virtual = 1 << 3 /* Require a virtual base */
> Comment should end with period and two spaces.


>> +	    /* Skip this result if we require virtual inheritance
>> +	       and this is not a virtual base. */
> Likewise.

Will fix.

> 
>> -      data.want_any = access == ba_any;
>> +      data.want_any = (access & ~ba_require_virtual) == ba_any;
>>         data.offset = offset;
>> +      data.require_virtual = (access & ba_require_virtual);
> It seems like you aren't using ba_require_virtual|ba_any, so do you need
> to mess with bitwise &?

base_access_flags is meant to be a bitmask, but for some reason ba_any 
is 0, and I didn't want to change the semantics of that == check.


> 
>> +@defbuiltin{bool __builtin_is_virtual_base_of (@var{base_type}, @var{derived_type})}
>> +If @var{base_type} is a virtual base class of @var{derived_type}
>> +([class.derived]) then the trait is @code{true}, otherwise it is @code{false}.
>> +Top-level cv-qualifications of @var{base_type} and
>> +@var{derived_type} are ignored.
> We might mention here that this ignores access and ambiguity.

Will add. This was adapted from __is_base_of's docs; do you want me to 
add the same remark there?

Thanks,
-- 
Giuseppe D'Angelo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4244 bytes
Desc: Firma crittografica S/MIME
URL: <https://gcc.gnu.org/pipermail/libstdc++/attachments/20240926/c611e915/attachment.p7s>