Version 4.13, February 22, 2017

This version is a relatively minor update mainly consisting of bug fixes
and Microsoft/GNU/Clang compatibility improvements.  It does include
the following features and notable changes:

1. More C++17 features have been implemented:
     - Remove deprecated register keyword (P0001R1)
     - Remove deprecated operator++(bool) (P0002R1)
     - [[fallthrough]], [[nodiscard]], and [[maybe_unused]] (P0188R1,
       P0189R1, P0212R1)
     - Using attribute namespaces without repetition (P0028R4)
     - Standard and nonstandard attributes (P0283R2)
     - Hexadecimal floating-point literals (P0245R1)
     - __has_include (P0061R1)

2. __float80 and __float128 are now supported.

3. Compatibility with Microsoft Visual Studio 15 (--microsoft_version=1910).
   This includes enabling the appropriate language features and also fixes
   for some bugs encountered in handling the Microsoft header files.

4. C++11 constexpr support has been reworked to make use of the interpreter
   that was developed for C++14 constexpr.  This was done to make C++11
   constexpr more robust and to eliminate the need to maintain two
   versions of parts of constexpr support.

As usual, we have fixed bugs, added many minor features, and continued to
improve our various compatibility modes.  See the Changes_4.13 file for
full details (about 3,200 lines).

This release includes new target-dependent variables.  If you are using
run-time target configurations (other than the default ones we provide), you
will need to update those to include new target-specific configuration macros
with appropriate values for each of your target configurations.  The new
target configuration variables are:

  targ_sizeof_float128
  targ_alignof_float128
  targ_sizeof_float80
  targ_alignof_float80
  targ_float128_mant_dig
  targ_float80_mant_dig
  targ_float128_max_expnn
  targ_float80_max_exp
  targ_float128_min_exp
  targ_float80_min_exp
  targ_float128_field_alignment
  targ_float80_field_alignment

Note that if you do not provide the necessary configuration macros, you
will get build errors along the lines of "TARG_SIZEOF_FLOAT128_my_config
undefined".


