[csmith-project/csmith] b138a4: Start a new branch for floating point work

[Yang Chen <chenyang@cs.utah.edu>]

  Branch: refs/heads/floating-point
  Home:   https://github.com/csmith-project/csmith
  Commit: b138a40da6c385023567c0f9f61d46d45286f524
      https://github.com/csmith-project/csmith/commit/b138a40da6c385023567c0f9f61d46d45286f524
  Author: Yang Chen <chenyang@cs.utah.edu>
  Date:   2014-09-20 (Sat, 20 Sep 2014)

  Changed paths:
    M src/ArrayVariable.cpp
    M src/CGOptions.cpp
    M src/CGOptions.h
    M src/Constant.cpp
    M src/ExpressionVariable.cpp
    M src/Function.cpp
    M src/FunctionInvocation.cpp
    M src/FunctionInvocation.h
    M src/Lhs.cpp
    M src/Probabilities.cpp
    M src/Probabilities.h
    M src/RandomProgramGenerator.cpp
    M src/StatementArrayOp.cpp
    M src/StatementAssign.cpp
    M src/StatementAssign.h
    M src/Type.cpp
    M src/Type.h
    M src/VariableSelector.cpp

  Log Message:
  -----------
  Start a new branch for floating point work

* It's based on the patch provided by Alastair Donaldson

* Don't generate floating-point operands for bit-operators

* Don't generate floating-point index for arrays

* Don't generate floating-point operands for shifts

Currently csmith is able to generate programs that can be compiled successfully, but full with
UBs caused by convertion from floating-point to integers where the floating-point value is
out of the range of integers.