Implementing wbN suffix for literals (#119058)

[Alejandro Colomar]

Hi Jakub,

On Sun, Jun 29, 2025 at 08:22:01PM +0200, Jakub Jelinek wrote:
> On Sun, Jun 29, 2025 at 06:28:11AM +0200, Alejandro Colomar via Gcc wrote:
> > --- i/libcpp/include/cpplib.h
> > +++ w/libcpp/include/cpplib.h
> > @@ -1363,6 +1363,7 @@ struct cpp_num
> >  #define CPP_N_SIZE_T   0x2000000 /* C++23 size_t literal.  */
> >  #define CPP_N_BFLOAT16 0x4000000 /* std::bfloat16_t type.  */
> >  #define CPP_N_BITINT   0x8000000 /* C23 _BitInt literal.  */
> > +#define CPP_N_BITINT_N 0x10000000 /* C2y _BitInt(N) literal.  */
> 
> This is not correct.
> That overlaps with the lowest bit in CPP_N_WIDTH_FLOATN_NX.
> So, one would need to be extra careful not to treat say f16 suffix
> which is CPP_N_FLOATN | 0x10000000 as CPP_N_BITINT_N or
> wbN or uwbN suffix as the N in fN or fNx.

Oops, sorry.  Should I move FLOATN higher to keep these contiguous?
Thanks!

Have a lovely day!
Alex

> >  
> >  #define CPP_N_WIDTH_FLOATN_NX  0xF0000000 /* _FloatN / _FloatNx value
> >                                               of N, divided by 16.  */
> 
> 	Jakub
> 

-- 
<https://www.alejandro-colomar.es/>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://gcc.gnu.org/pipermail/gcc/attachments/20250629/48c55fd7/attachment.sig>