Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-22 Thread Joseph Myers
On Wed, 22 Sep 2021, Iain Sandoe wrote: > However, note that the use of __MACH__ to guard the Mach-O code has been > there for a long time (it is present in all open branches). So it's possible > that has > been silently doing the wrong thing for some time, That's "wrong thing" as in having pre

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-22 Thread Iain Sandoe
Hi Joseph, Thomas, > On 21 Sep 2021, at 21:11, Joseph Myers wrote: > > On Fri, 3 Sep 2021, Iain Sandoe wrote: > >> given that: >> >> a) this fixes Darwin x86-64 bootstrap which has been broken for more than 24h >> b) the patch is now Darwin-local. > > Actually, it's not Darwin-local. It uses

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-21 Thread Iain Sandoe
Hello Joseph, > On 21 Sep 2021, at 21:11, Joseph Myers wrote: > > On Fri, 3 Sep 2021, Iain Sandoe wrote: > >> given that: >> >> a) this fixes Darwin x86-64 bootstrap which has been broken for more than 24h >> b) the patch is now Darwin-local. > > Actually, it's not Darwin-local. It uses __MA

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-21 Thread Joseph Myers
On Fri, 3 Sep 2021, Iain Sandoe wrote: > given that: > > a) this fixes Darwin x86-64 bootstrap which has been broken for more than 24h > b) the patch is now Darwin-local. Actually, it's not Darwin-local. It uses __MACH__, which is also defined for Hurd. And because sfp-machine.h gets included

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-03 Thread Iain Sandoe
> On 3 Sep 2021, at 08:51, Iain Sandoe wrote: > > >> On 2 Sep 2021, at 21:03, Joseph Myers wrote: >> >> On Thu, 2 Sep 2021, Iain Sandoe via Gcc-patches wrote: >> >>> diff --git a/libgcc/soft-fp/eqdf2.c b/libgcc/soft-fp/eqdf2.c >>> index 2a44ee377ce..a3bb664f5f1 100644 >>> --- a/libgcc/soft

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-03 Thread Iain Sandoe via Gcc-patches
Hi Joseph, > On 2 Sep 2021, at 21:03, Joseph Myers wrote: > > On Thu, 2 Sep 2021, Iain Sandoe via Gcc-patches wrote: > >> diff --git a/libgcc/soft-fp/eqdf2.c b/libgcc/soft-fp/eqdf2.c >> index 2a44ee377ce..a3bb664f5f1 100644 >> --- a/libgcc/soft-fp/eqdf2.c >> +++ b/libgcc/soft-fp/eqdf2.c >> @@ -

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Joseph Myers
On Thu, 2 Sep 2021, Iain Sandoe via Gcc-patches wrote: > diff --git a/libgcc/soft-fp/eqdf2.c b/libgcc/soft-fp/eqdf2.c > index 2a44ee377ce..a3bb664f5f1 100644 > --- a/libgcc/soft-fp/eqdf2.c > +++ b/libgcc/soft-fp/eqdf2.c > @@ -28,6 +28,7 @@ > License along with the GNU C Library; if not, see >

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Joseph Myers
One of the committed changes breaks the build of libgcc for 32-bit x86 configurations without SSE2 enabled by default: In file included from /scratch/jmyers/glibc-bot/src/gcc/libgcc/soft-fp/extendhfsf2.c:31: /scratch/jmyers/glibc-bot/src/gcc/libgcc/soft-fp/half.h:62:1: error: unable to emulate

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Iain Sandoe via Gcc-patches
Patch below fixes bootstrap, OK if it passes testing on x86_64 darwin/linux? (if !OK .. then suggestions welcome) thanks Iain > On 2 Sep 2021, at 16:18, Hongtao Liu wrote: > > > > On Thursday, September 2, 2021, Iain Sandoe wrote: > Hi Hongtao. > > > On 2 Sep 2021, at 07:06, Hongtao Liu vi

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Hongtao Liu via Gcc-patches
On Thursday, September 2, 2021, H.J. Lu wrote: > On Wed, Sep 1, 2021 at 11:00 PM Hongtao Liu wrote: > > > > I'm going to check in the first 3 patches which are already approved. > > > > Update hf soft-fp from glibc. > > [i386] Enable _Float16 type for TARGET_SSE2 and above. > > [i386] libg

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread H.J. Lu via Gcc-patches
On Wed, Sep 1, 2021 at 11:00 PM Hongtao Liu wrote: > > I'm going to check in the first 3 patches which are already approved. > > Update hf soft-fp from glibc. > [i386] Enable _Float16 type for TARGET_SSE2 and above. > [i386] libgcc: Enable hfmode soft-sf/df/xf/tf extensions and > truncat

[PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Hongtao Liu via Gcc-patches
On Thursday, September 2, 2021, Iain Sandoe wrote: > Hi Hongtao. > > > On 2 Sep 2021, at 07:06, Hongtao Liu via Gcc-patches < > gcc-patches@gcc.gnu.org> wrote: > > > > I'm going to check in the first 3 patches which are already approved. > > > > Update hf soft-fp from glibc. > > [i386] Enable _

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-02 Thread Iain Sandoe via Gcc-patches
Hi Hongtao. > On 2 Sep 2021, at 07:06, Hongtao Liu via Gcc-patches > wrote: > > I'm going to check in the first 3 patches which are already approved. > > Update hf soft-fp from glibc. > [i386] Enable _Float16 type for TARGET_SSE2 and above. > [i386] libgcc: Enable hfmode soft-sf/df/xf/tf ex

Re: [PATCH V3 0/6] Initial support for AVX512FP16

2021-09-01 Thread Hongtao Liu via Gcc-patches
I'm going to check in the first 3 patches which are already approved. Update hf soft-fp from glibc. [i386] Enable _Float16 type for TARGET_SSE2 and above. [i386] libgcc: Enable hfmode soft-sf/df/xf/tf extensions and truncations. On Mon, Aug 2, 2021 at 2:31 PM liuhongt wrote: > > Update

[PATCH V3 0/6] Initial support for AVX512FP16

2021-08-01 Thread liuhongt via Gcc-patches
Update from v2: 1. Support -fexcess-precision=16 which will enable FLT_EVAL_METHOD_PROMOTE_TO_FLOAT16 when backend supports _Float16. 2. Update ix86_get_excess_precision, so -fexcess-precision=standard should not do anything different from -fexcess-precision=fast regarding _Float16. 3. Avoiding