This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH], PR 71806, Fix -mfloat128/-mfloat128-hardware defaults on power9
- From: Michael Meissner <meissner at linux dot vnet dot ibm dot com>
- To: Segher Boessenkool <segher at kernel dot crashing dot org>
- Cc: Michael Meissner <meissner at linux dot vnet dot ibm dot com>, gcc-patches at gcc dot gnu dot org, David Edelsohn <dje dot gcc at gmail dot com>, Bill Schmidt <wschmidt at linux dot vnet dot ibm dot com>
- Date: Fri, 8 Jul 2016 10:17:14 -0400
- Subject: Re: [PATCH], PR 71806, Fix -mfloat128/-mfloat128-hardware defaults on power9
- Authentication-results: sourceware.org; auth=none
- References: <20160708133133.GA31893@ibm-tiger.the-meissners.org> <20160708141350.GE12770@gate.crashing.org>
On Fri, Jul 08, 2016 at 09:13:50AM -0500, Segher Boessenkool wrote:
> On Fri, Jul 08, 2016 at 09:31:33AM -0400, Michael Meissner wrote:
> > * gcc.target/powerpc/p9-lxvx-stxvx-3.c: Add -mfloat128 option.
>
> Is that the only testcase that needs updating?
>
> > --- gcc/config/rs6000/rs6000-cpus.def (revision 238127)
> > +++ gcc/config/rs6000/rs6000-cpus.def (working copy)
> > @@ -63,7 +63,6 @@
> > /* Add ISEL back into ISA 3.0, since it is supposed to be a win. Do not add
> > P9_MINMAX until the hardware that supports it is available. */
> > #define ISA_3_0_MASKS_SERVER (ISA_2_7_MASKS_SERVER \
> > - | OPTION_MASK_FLOAT128_HW \
>
> Please add a comment for this as well?
Ok.
> > /* IEEE 128-bit floating point hardware instructions imply enabling
> > __float128. */
> > if (TARGET_FLOAT128_HW
> > - && (rs6000_isa_flags & (OPTION_MASK_P9_VECTOR
> > - | OPTION_MASK_DIRECT_MOVE
> > - | OPTION_MASK_UPPER_REGS_DI
> > - | OPTION_MASK_UPPER_REGS_DF
> > - | OPTION_MASK_UPPER_REGS_SF)) == 0)
> > + && (rs6000_isa_flags & ISA_3_0_MASKS_IEEE) != ISA_3_0_MASKS_IEEE)
> > {
> > if ((rs6000_isa_flags_explicit & OPTION_MASK_FLOAT128_HW) != 0)
> > error ("-mfloat128-hardware requires full ISA 3.0 support");
>
> That is not the same thing... New one looks better, is this a bugfix?
> The changelog doesn't say.
I just moved the OPTIONS_MASKS_* used here to a common macro that is checked
earlier to enable hardware support if -mfloat128.
> Okay for trunk and 6 with those nits fixed. Thanks,
Thanks.
--
Michael Meissner, IBM
IBM, M/S 2506R, 550 King Street, Littleton, MA 01460-6245, USA
email: meissner@linux.vnet.ibm.com, phone: +1 (978) 899-4797