https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122483

--- Comment #36 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Richard Sandiford from comment #20)
> But the point is that a choice is always being made.  The choice is silent
> for normal functions and noisy for streaming and streaming-compatible
> functions.  But there's a choice nonetheless.  There is no concept of an
> "uncertain", "non-commital" or "deferential" type.

How is that different from any other ABI attributes?
We also have normal calling convention, or say regparm(1), or regparm(2), or
regparm(3) and similarly with many other ABI attributes.  No special attribute
for the normal case, and yet we do accept
void f(void) [[that_ABI_attribute]];
void f(void);
as being the same as
void f(void) [[that_ABI_attribute]];
void f(void) [[that_ABI_attribute]];
because there is no risk of it breaking anything, that is a user friendly
choice (and sometimes really needed, when you don't have control over what
attributes some header imposes on standard APIs).  The function already has
some specific calling convention, so there is no reason to harass the user that
they haven't provided it again.

Reply via email to