- Enable support for Green Hills Software compiler
- Define weak implementations only once except on Windows

The user can specify UNITY_OUTPUT_CHAR_HEADER_DECLARATION and
UNITY_OUTPUT_FLUSH_HEADER_DECLARATION when he would like to declare
UNITY_OUTPUT_CHAT or UNITY_OUTPUT_FLUSH respectivly

Fixed copy-paste error for:
- UNITY_TEST_ASSERT_EACH_EQUAL_INT64
- UNITY_TEST_ASSERT_EACH_EQUAL_UINT64
- UNITY_TEST_ASSERT_EACH_EQUAL_HEX64

So that the comparison is done on the expected _value_ instead of the _array_.

Turn UNITY_OUTPUT_FLUSH off by default. Added a quick-define for the most common case: UNITY_USE_FLUSH_STDOUT. Clarified documentation. Fixes issue #269

There was a mismatch of definition and declaration of UnityPrintFloat
and this commit aligns both on const UNITY_DOUBLE, but then inside it
actually uses a positive version of the variable only.

With a wider integer width than a pointer, gcc can still warn
when you cast a pointer. Strange but true. To see these warnings
compile with gcc -m32, 32-bit target, and enable UNITY_SUPPORT_64

Using this option changes the control flow of Unity, but is useful on
constrained embedded systems. You can't fully simulate the power of
'longjmp' with just 'return', but Unity still works well, all tests pass.

The setjmp calls are all in TEST_PROTECT() already

Originated from commit c6dc96f3 but are rarely used and not in math.h
Removes float to double promotion warning and simplifies the code
'gcc -std=gnu99 -Wdouble-promotion' spits out these warnings

Remove UNITY_FIXTURES option, which was probably not doing what was
originally indended. It was suppressing the test name even in quiet mode.

The old type was really just used as a cast-to-int to do pointer math
in char size.

Revised internal type naming scheme to better sandbox Unity away from everything else. Sure, short was nice, but not at the expense of naming collisions.

Remove UNITY_FLOAT_VERBOSE entirely, add option UNITY_EXCLUDE_FLOAT_PRINT
Remove some questionable float casts from doubles
Default to Round Ties to Even behavior, add option to Round Ties Away from Zero

 Change the parameter of UnityPrintFloat from float to double (_UD)
 The call to snprintf uses implict default promotion to a double anyway

 Delete old comments from before math.h was the default isinf() method

 The C preprocessor cannot use sizeof in #if etc.

Change-Id: I901bf192d3c06457d67c6d3a64bdb9aea22c53c2

Script verification hasn't been added yet.

add UNITY_OUTPUT_FLUSH to make sure we get the output we need on aborted tests and whatnot.

 This solves the warnings from -Wredundant-decls when overriding the
  OUTPUT function with a function declared in another header.
 It's better this is the non-default option, since using it requires either
  a declaration of the new function OR a preprocessor guard on declaring
  UNITY_OUTPUT_CHAR (as in unity.c here), in every file using the function.
 See Pull Request #185 for more.

 These were aimed at preventing the internals header from declaring
  a prototype for 'int UNITY_OUTPUT_CHAR(int);'. The second part fixed
  an omitted declaration in the tests caused by the first commit.
 Will replace the default behavior & add an option for this in next commit

 Revert the previous commit. Add tests for extended enum cases.
 Fix crash due to accessing 'trait_names' array out of bounds.
 Adding an extra invalid value to the end of an enum causes '-Wswitch' flag
  to warn unless there is a switch default case - also enabled by '-Wall'.

 Add extern UnityFixture in internals header. Add C89 target in Makefile
 clang will show these warnings but can suppress "//" with -Wno-comment.