This is the mail archive of the
gcc-help@gcc.gnu.org
mailing list for the GCC project.
RE: target_clones constexpr
> -----Original Message-----
> From: Matthias Kretz [mailto:m.kretz@gsi.de]
> Sent: Monday, January 15, 2018 7:24 AM
> To: gcc-help@gcc.gnu.org
> Cc: Schulz, Roland <roland.schulz@intel.com>; Victor Rodriguez
> <vm.rod25@gmail.com>; Mason <slash.tmp@free.fr>; O Mahony, Billy
> <billy.o.mahony@intel.com>; Evgeny Stupachenko <evstupac@gmail.com>;
> Jakub Jelinek <jakub@redhat.com>
> Subject: Re: target_clones constexpr
>
> On Donnerstag, 11. Januar 2018 20:37:55 CET Schulz, Roland wrote:
> > I would like to do something like: [...]
>
> +1
>
> > I would like to use width to do manual strip-mining or instantiate
> > SIMD C++ templates. The problem is that this doesn't compile because
> > __builtin_cpu_supports isn't constexpr. For every clone of foo, the
> > compiler knows the SIMD support at compile time. I'm looking for a
> > builtin which lets me query that in a constexpr way.
>
> AFAIK there is nothing (I'd be happy to be proven wrong). I also can't think of
> any hack that would produce the value as a constant expression. (I tried
> `using v64 [[gnu::vector_size(64)]] = float; alignof(v64)`, but its alignment is
> 64, independent of the register size it uses.)
I filed it as feature request: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83875
Roland