On Tue, Sep 7, 2021 at 3:18 PM Jakub Jelinek <ja...@redhat.com> wrote:
>
> On Tue, Sep 07, 2021 at 09:52:57AM +0800, Hongtao Liu wrote:
> > Adjust the wording for x86 _Float16 type.
> >
> > gcc/ChangeLog:
> >
> > * doc/extend.texi: (@node Floating Types): Adjust the wording.
> > (@node Half-Precision): Ditto.
> >
> > 1 file changed, 15 insertions(+), 13 deletions(-)
> > gcc/doc/extend.texi | 28 +++++++++++++++-------------
> >
> > modified gcc/doc/extend.texi
> > @@ -1076,9 +1076,10 @@ systems where @code{__float128} is supported.
> > The @code{_Float32}
> > type is supported on all systems supporting IEEE binary32; the
> > @code{_Float64} and @code{_Float32x} types are supported on all systems
> > supporting IEEE binary64. The @code{_Float16} type is supported on AArch64
> > -systems by default, and on ARM systems when the IEEE format for 16-bit
> > -floating-point types is selected with @option{-mfp16-format=ieee}.
> > -GCC does not currently support @code{_Float128x} on any systems.
> > +systems by default when the IEEE format for 16-bit floating-point types is
>
> The AArch64 case now has the ARM case restriction and ARM is lost. It
> should be
>
> +systems by default, on ARM systems when the IEEE format for 16-bit
> +floating-point-types is
>
> > +selected with @option{-mfp16-format=ieee} and, for both C and C++, on x86
> > +systems with SSE2 enabled. GCC does not currently support
> > +@code{_Float128x} on any systems.
> >
> > On the i386, x86_64, IA-64, and HP-UX targets, you can declare complex
> > types using the corresponding internal complex type, @code{XCmode} for
> > @@ -1108,6 +1109,12 @@ On ARM and AArch64 targets, GCC supports
> > half-precision (16-bit) floating
> > point via the @code{__fp16} type defined in the ARM C Language Extensions.
> > On ARM systems, you must enable this type explicitly with the
> > @option{-mfp16-format} command-line option in order to use it.
> > +On x86 targets with SSE2 enabled, GCC supports half-precision (16-bit)
> > +floating point via the @code{_Float16} type, there are many ways to enable
> > +SSE2, @option{-msse2, -mavx, -mavx512f, ...} on the command line, or
> > various
> > +target attributes.
>
> The ", there are many ways ... attributes" was just meant as explanation for
> the "with SSE2 enabled" wording, not something that should be literally in
> the documentation. It is documented elsewhere...
>
> > +For C++, x86 provides a builtin type named @code{_Float16} which contains
> > +same data format as C.
> >
> > ARM targets support two incompatible representations for half-precision
> > floating-point values. You must choose one of the representations and
> > @@ -1151,16 +1158,11 @@ calls.
> > It is recommended that portable code use the @code{_Float16} type defined
> > by ISO/IEC TS 18661-3:2015. @xref{Floating Types}.
> >
> > -On x86 targets with @code{target("sse2")} and above, GCC supports
> > half-precision
> > -(16-bit) floating point via the @code{_Float16} type which is defined by
> > -18661-3:2015. For C++, x86 provide a builtin type named @code{_Float16}
> > -which contains same data format as C.
> > -
> > -Without @option{-mavx512fp16}, @code{_Float16} type is storage only, all
> > -operations will be emulated by software emulation and the @code{float}
> > -instructions. The default behavior for @code{FLT_EVAL_METHOD} is to keep
> > -the intermediate result of the operation as 32-bit precision. This may lead
> > -to inconsistent behavior between software emulation and AVX512-FP16
> > +On x86 targets, without @option{-mavx512fp16}, @code{_Float16} type is
> > +storage only, all operations will be emulated by software emulation and the
> > +@code{float} instructions. The default behavior for @code{FLT_EVAL_METHOD}
> > is
> > +to keep the intermediate result of the operation as 32-bit precision. This
> > may
> > +lead to inconsistent behavior between software emulation and AVX512-FP16
> > instructions.
> >
> > @node Decimal Float
>
> Jakub
>
Like this?
1 file changed, 12 insertions(+), 13 deletions(-)
gcc/doc/extend.texi | 25 ++++++++++++-------------
modified gcc/doc/extend.texi
@@ -1076,9 +1076,10 @@ systems where @code{__float128} is supported.
The @code{_Float32}
type is supported on all systems supporting IEEE binary32; the
@code{_Float64} and @code{_Float32x} types are supported on all systems
supporting IEEE binary64. The @code{_Float16} type is supported on AArch64
-systems by default, and on ARM systems when the IEEE format for 16-bit
-floating-point types is selected with @option{-mfp16-format=ieee}.
-GCC does not currently support @code{_Float128x} on any systems.
+systems by default, on ARM systems when the IEEE format for 16-bit
+floating-point types is selected with @option{-mfp16-format=ieee} and,
+for both C and C++, on x86 systems with SSE2 enabled. GCC does not currently
+support @code{_Float128x} on any systems.
On the i386, x86_64, IA-64, and HP-UX targets, you can declare complex
types using the corresponding internal complex type, @code{XCmode} for
@@ -1108,6 +1109,9 @@ On ARM and AArch64 targets, GCC supports
half-precision (16-bit) floating
point via the @code{__fp16} type defined in the ARM C Language Extensions.
On ARM systems, you must enable this type explicitly with the
@option{-mfp16-format} command-line option in order to use it.
+On x86 targets with SSE2 enabled, GCC supports half-precision (16-bit)
+floating point via the @code{_Float16} type. For C++, x86 provides a builtin
+type named @code{_Float16} which contains the same data format as C.
ARM targets support two incompatible representations for half-precision
floating-point values. You must choose one of the representations and
@@ -1151,16 +1155,11 @@ calls.
It is recommended that portable code use the @code{_Float16} type defined
by ISO/IEC TS 18661-3:2015. @xref{Floating Types}.
-On x86 targets with @code{target("sse2")} and above, GCC supports
half-precision
-(16-bit) floating point via the @code{_Float16} type which is defined by
-18661-3:2015. For C++, x86 provide a builtin type named @code{_Float16}
-which contains same data format as C.
-
-Without @option{-mavx512fp16}, @code{_Float16} type is storage only, all
-operations will be emulated by software emulation and the @code{float}
-instructions. The default behavior for @code{FLT_EVAL_METHOD} is to keep
-the intermediate result of the operation as 32-bit precision. This may lead
-to inconsistent behavior between software emulation and AVX512-FP16
+On x86 targets, without @option{-mavx512fp16}, @code{_Float16} type is
+storage only, all operations will be emulated by software emulation and the
+@code{float} instructions. The default behavior for @code{FLT_EVAL_METHOD} is
+to keep the intermediate result of the operation as 32-bit precision. This may
+lead to inconsistent behavior between software emulation and AVX512-FP16
instructions.
--
BR,
Hongtao
