On Fri, 17 Jul 2020, Patrick Palka wrote:

> On Fri, 17 Jul 2020, Patrick Palka wrote:
> 
> > On Wed, 15 Jul 2020, Patrick Palka wrote:
> > 
> > > On Tue, 14 Jul 2020, Patrick Palka wrote:
> > > 
> > > > This implements the floating-point std::to_chars overloads for float,
> > > > double and long double.  We use the Ryu library to compute the shortest
> > > > round-trippable fixed and scientific forms of a number for float, double
> > > > and long double.  We also use Ryu for performing fixed and scientific
> > > > formatting of float and double. For formatting long double with an
> > > > explicit precision argument we use a printf fallback.  Hexadecimal
> > > > formatting for float, double and long double is implemented from
> > > > scratch.
> > > > 
> > > > The supported long double binary formats are float64 (same as double),
> > > > float80 (x86 extended precision), float128 and ibm128.
> > > > 
> > > > Much of the complexity of the implementation is in computing the exact
> > > > output length before handing it off to Ryu (which doesn't do bounds
> > > > checking).  In some cases it's hard to compute the output length before
> > > > the fact, so in these cases we instead compute an upper bound on the
> > > > output length and use a sufficiently-sized intermediate buffer (if the
> > > > output range is smaller than the upper bound).
> > > > 
> > > > Another source of complexity is in the general-with-precision formatting
> > > > mode, where we need to do zero-trimming of the string returned by Ryu, 
> > > > and
> > > > where we also take care to avoid having to format the string a second
> > > > time when the general formatting mode resolves to fixed.
> > > > 
> > > > Tested on x86_64-pc-linux-gnu, aarch64-unknown-linux-gnu,
> > > > s390x-ibm-linux-gnu, and powerpc64-unknown-linux-gnu.
> > > > 
> > > > libstdc++-v3/ChangeLog:
> > > > 
> > > >         * acinclude.m4 (libtool_VERSION): Bump to 6:29:0.
> > > >         * config/abi/pre/gnu.ver: Add new exports.
> > > >         * configure: Regenerate.
> > > >         * include/std/charconv (to_chars): Declare the floating-point
> > > >         overloads for float, double and long double.
> > > >         * src/c++17/Makefile.am (sources): Add floating_to_chars.cc.
> > > >         * src/c++17/Makefile.in: Regenerate.
> > > >         * src/c++17/floating_to_chars.cc: New file.
> > > >         * testsuite/20_util/to_chars/long_double.cc: New test.
> > > >         * testsuite/util/testsuite_abi.cc: Add new symbol version.
> > > 
> > > Here is v2 of this patch, which fixes a build failure on i386 due to
> > > __int128 being unavailable, by refactoring the long double binary format
> > > selection to avoid referring to __int128 when it doesn't exist.  The
> > > patch also makes the hex formatting for 80-bit long double use uint64_t
> > > instead of __int128 since the mantissa has exactly 64 bits in this case.
> > 
> > Here's v3 which just makes some minor stylistic adjustments, and most
> > notably replaces the use of _GLIBCXX_DEBUG with _GLIBCXX_ASSERTIONS
> > since we just want to enable __glibcxx_assert and not all of debug mode.
> 
> Here's v4, which should now correctly support using <charconv> with
> -mlong-double-64 on targets with a large default long double type.
> This is done by defining the long double to_chars overloads as inline
> wrappers around the double overloads within <charconv> whenever
> __DBL_MANT_DIG__ equals __LDBL_MANT_DIG__.

> 
> -- >8 --
> 
> Subject: [PATCH 3/4] libstdc++: Add floating-point std::to_chars
>  implementation
> 
> This implements the floating-point std::to_chars overloads for float,
> double and long double.  We use the Ryu library to compute the shortest
> round-trippable fixed and scientific forms of a number for float, double
> and long double.  We also use Ryu for performing explicit-precision
> fixed and scientific formatting of float and double. For
> explicit-precision formatting of long double we fall back to using
> printf.  Hexadecimal formatting for float, double and long double is
> implemented from scratch.
> 
> The supported long double binary formats are binary64, binary80 (x86
> 80-bit extended precision), binary128 and ibm128.
> 
> Much of the complexity of the implementation is in computing the exact
> output length before handing it off to Ryu (which doesn't do bounds
> checking).  In some cases it's hard to compute the output length
> beforehand, so in these cases we instead compute an upper bound on the
> output length and use a sufficiently-sized intermediate buffer if
> necessary.
> 
> Another source of complexity is in the general-with-precision formatting
> mode, where we need to do zero-trimming of the string returned by Ryu,
> and where we also take care to avoid having to format the string a
> second time when the general formatting mode resolves to fixed.
> 
> This implementation is non-conforming in a couple of ways:
> 
> 1. For the shortest hexadecimal formatting, we currently follow the
>    Microsoft implementation's approach of being consistent with the
>    output of printf's '%a' specifier at the expense of sometimes not
>    printing the shortest representation.  For example, the shortest hex
>    form of 1.08p+0 is 2.1p-1, but we output the former instead of the
>    latter, as does printf.
> 
> 2. The Ryu routines for doing shortest formatting on types larger than
>    binary64 use the __int128 type, and some targets (e.g. i386) have a
>    large long double type but lack __int128.  For such targets we make
>    the long double to_chars overloads go through the double overloads,
>    which means we lose precision in the output.  (The mantissa of long
>    double is 64 bits on i386, so I think we could potentially fix this
>    by writing a specialized version of the generic Ryu formatting
>    routine which works with uint64_t instead of __int128.)
> 
> 3. The __ibm128 shortest formatting routines don't guarantee
>    round-trippability if the difference between the high- and low-order
>    exponent is too large.  This is because we treat the type as if it
>    has a contiguous 105-bit mantissa by merging the high- and low-order
>    mantissas, so we potentially lose precision from the low-order part.
>    Although this precision-dropping behavior is non-conforming, it seems
>    consistent with how printf formats __ibm128.
> 
> libstdc++-v3/ChangeLog:
> 
>       * acinclude.m4 (libtool_VERSION): Bump to 6:29:0.
>       * config/abi/pre/gnu.ver: Add new exports.
>       * configure: Regenerate.
>       * include/std/charconv (to_chars): Declare the floating-point
>       overloads for float, double and long double.
>       * src/c++17/Makefile.am (sources): Add floating_to_chars.cc.
>       * src/c++17/Makefile.in: Regenerate.
>       * src/c++17/floating_to_chars.cc: New file.
>       * testsuite/20_util/to_chars/long_double.cc: New test.
>       * testsuite/util/testsuite_abi.cc: Add new symbol version.
> ---
>  libstdc++-v3/acinclude.m4                     |    2 +-
>  libstdc++-v3/config/abi/pre/gnu.ver           |   12 +
>  libstdc++-v3/configure                        |    2 +-
>  libstdc++-v3/include/std/charconv             |   43 +
>  libstdc++-v3/src/c++17/Makefile.am            |    1 +
>  libstdc++-v3/src/c++17/Makefile.in            |    5 +-
>  libstdc++-v3/src/c++17/floating_to_chars.cc   | 1514 +++++++++++++++++
>  .../testsuite/20_util/to_chars/long_double.cc |  197 +++
>  libstdc++-v3/testsuite/util/testsuite_abi.cc  |    3 +-
>  9 files changed, 1774 insertions(+), 5 deletions(-)
>  create mode 100644 libstdc++-v3/src/c++17/floating_to_chars.cc
>  create mode 100644 libstdc++-v3/testsuite/20_util/to_chars/long_double.cc
> 
> diff --git a/libstdc++-v3/acinclude.m4 b/libstdc++-v3/acinclude.m4
> index ee5e0336f2c..e3926e1c9c2 100644
> --- a/libstdc++-v3/acinclude.m4
> +++ b/libstdc++-v3/acinclude.m4
> @@ -3846,7 +3846,7 @@ changequote([,])dnl
>  fi
>  
>  # For libtool versioning info, format is CURRENT:REVISION:AGE
> -libtool_VERSION=6:28:0
> +libtool_VERSION=6:29:0
>  
>  # Everything parsed; figure out what files and settings to use.
>  case $enable_symvers in
> diff --git a/libstdc++-v3/config/abi/pre/gnu.ver 
> b/libstdc++-v3/config/abi/pre/gnu.ver
> index edf4485e607..9a1bcfd25d1 100644
> --- a/libstdc++-v3/config/abi/pre/gnu.ver
> +++ b/libstdc++-v3/config/abi/pre/gnu.ver
> @@ -2299,6 +2299,18 @@ GLIBCXX_3.4.28 {
>  
>  } GLIBCXX_3.4.27;
>  
> +GLIBCXX_3.4.29 {
> +    # to_chars(char*, char*, [float|double|long double])
> +    _ZSt8to_charsPcS_[fdeg];
> +
> +    # to_chars(char*, char*, [float|double|long double], chars_format)
> +    _ZSt8to_charsPcS_[fdeg]St12chars_format;
> +
> +    # to_chars(char*, char*, [float|double|long double], chars_format, int)
> +    _ZSt8to_charsPcS_[fdeg]St12chars_formati;
> +
> +} GLIBCXX_3.4.28;
> +
>  # Symbols in the support library (libsupc++) have their own tag.
>  CXXABI_1.3 {
>  
> diff --git a/libstdc++-v3/configure b/libstdc++-v3/configure
> index dd54bd406a9..73f771e7335 100755
> --- a/libstdc++-v3/configure
> +++ b/libstdc++-v3/configure
> @@ -75231,7 +75231,7 @@ $as_echo "$as_me: WARNING: === Symbol versioning will 
> be disabled." >&2;}
>  fi
>  
>  # For libtool versioning info, format is CURRENT:REVISION:AGE
> -libtool_VERSION=6:28:0
> +libtool_VERSION=6:29:0
>  
>  # Everything parsed; figure out what files and settings to use.
>  case $enable_symvers in
> diff --git a/libstdc++-v3/include/std/charconv 
> b/libstdc++-v3/include/std/charconv
> index cc7dd0e3758..bd59924f7e7 100644
> --- a/libstdc++-v3/include/std/charconv
> +++ b/libstdc++-v3/include/std/charconv
> @@ -688,6 +688,49 @@ namespace __detail
>    operator^=(chars_format& __lhs, chars_format __rhs) noexcept
>    { return __lhs = __lhs ^ __rhs; }
>  
> +  // Floating-point std::to_chars
> +
> +  // Overloads for float.
> +  to_chars_result to_chars(char* __first, char* __last, float __value) 
> noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, float __value,
> +                        chars_format __fmt) noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, float __value,
> +                        chars_format __fmt, int __precision) noexcept;
> +
> +  // Overloads for double.
> +  to_chars_result to_chars(char* __first, char* __last, double __value) 
> noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, double __value,
> +                        chars_format __fmt) noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, double __value,
> +                        chars_format __fmt, int __precision) noexcept;
> +
> +  // Overloads for long double.
> +  to_chars_result to_chars(char* __first, char* __last, long double __value)
> +    noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, long double __value,
> +                        chars_format __fmt) noexcept;
> +  to_chars_result to_chars(char* __first, char* __last, long double __value,
> +                        chars_format __fmt, int __precision) noexcept;
> +
> +  // If long double has the same binary format as double, then we just define
> +  // the long double overloads as wrappers around the corresponding double
> +  // overloads.
> +#if __LDBL_MANT_DIG__ == __DBL_MANT_DIG__
> +  inline to_chars_result
> +  to_chars(char* __first, char* __last, long double __value) noexcept
> +  { return to_chars(__first, __last, double(__value)); }
> +
> +  inline to_chars_result
> +  to_chars(char* __first, char* __last, long double __value,
> +        chars_format __fmt) noexcept
> +  { return to_chars(__first, __last, double(__value), __fmt); }
> +
> +  inline to_chars_result
> +  to_chars(char* __first, char* __last, long double __value,
> +        chars_format __fmt, int __precision) noexcept
> +  { return to_chars(__first, __last, double(__value), __fmt, __precision); }
> +#endif

Hmm, I think this approach for supporting -mlong-double-64 might
introduce an ODR violation because each long double to_chars overload
could potentially have two different definitions available in a program,
one out-of-line in floating_to_chars.cc (compiled without
-mlong-double-64) and another inline in <charconv> (compiled with
-mlong-double-64)..

Reply via email to