This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: SSE and SSE2 intrinsics
> > > To get some benefits, you need to eighter use the intrisc and then the
> > > code would not compile w/o those -m options or use -mfp-math=sse
> > > to enable use of SSE instructions for floating point that should
> > > improve performance of FP code but not due to use of paralelization.
> >
> > Indeed; I already have some code which relies on the ICC intrinsics,
> > which I'd rather like to be able to compile with the normal gcc tools.
>
> I would be very interested in the results.
Regrettably, the code turns out to use only SSE2 intrinsics, so even with
the new snapshot it doesn't compile.
It's unhelpful that the result of #include <xmmintrin.h> without an
explicit -msse gives a huge screed of "__builtin_ia32_whatever not declared"
errors ... I was using -msse2 and expecting that, since all chips supporting
SSE2 also support SSE and MMX, -msse2 implied -msse and -mmmx.
Tom